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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
c1ff2e24-c307-40d3-9fae-68339268180f | ia-gm-a-deep-bidirectional-learning-method | null | null | https://ojs.aaai.org/index.php/AAAI/article/view/16461 | https://ojs.aaai.org/index.php/AAAI/article/view/16461/16268 | IA-GM: A Deep Bidirectional Learning Method for Graph Matching | Existing deep learning methods for graph matching(GM) problems usually considered affinity learningto assist combinatorial optimization in a feedforward pipeline, and parameter learning is executed by back-propagating the gradients of the matching loss. Such a pipeline pays little attention to the possible complementary benefit from the optimization layer to the learning component. In this paper, we overcome the above limitation under a deep bidirectional learning framework.Our method circulates the output of the GM optimization layer to fuse with the input for affinity learning. Such direct feedback enhances the input by a feature enrichment and fusion technique, which exploits andintegrates the global matching patterns from the deviation of the similarity permuted by the current matching estimate. As a result, the circulation enables the learning component to benefit from the optimization process, taking advantage of both global feature and the embedding result which is calculated by local propagationthrough node-neighbors. Moreover, circulation consistency induces an unsupervised loss that can be implemented individually or jointly to regularize the supervised loss. Experiments on challenging datasets demonstrate the effectiveness of our methods for both supervised learning and unsupervised learning. | ['Lei Xu', 'Shikui Tu', 'Kaixuan Zhao'] | 2021-05-18 | null | null | null | aaai-2021-5 | ['graph-matching'] | ['graphs'] | [ 1.75092034e-02 3.71657908e-01 -4.40431803e-01 -5.36870480e-01
-6.48044109e-01 -3.17206502e-01 6.21979535e-01 6.33999825e-01
-5.62752843e-01 4.81310219e-01 1.77595362e-01 1.04937829e-01
-2.62057394e-01 -1.02028453e+00 -9.09861684e-01 -8.28302085e-01
5.19347563e-02 3.60229343e-01 4.04609442e-02 -6.15139157e-02
7.93211237e-02 3.64298493e-01 -1.30725920e+00 2.98240960e-01
1.11068511e+00 1.31601191e+00 1.01102889e-02 2.72583842e-01
-4.12529379e-01 7.55393326e-01 -9.44103897e-02 -5.70957303e-01
4.07603204e-01 -2.68425167e-01 -5.25816321e-01 -1.05131185e-02
4.08685237e-01 -2.58910835e-01 -1.47322118e-01 1.02210307e+00
4.66495484e-01 -7.25066150e-03 2.22699970e-01 -1.25578833e+00
-2.98760772e-01 8.45376909e-01 -3.33395958e-01 -1.90555423e-01
1.81112230e-01 2.08369762e-01 1.52606177e+00 -9.89120007e-01
5.36984444e-01 8.83284450e-01 7.67440021e-01 1.54208258e-01
-1.36896443e+00 -5.72559178e-01 4.94851947e-01 1.43016219e-01
-1.21942782e+00 -3.76493067e-01 1.24185669e+00 -2.33268231e-01
7.40577638e-01 4.00659256e-02 8.89556766e-01 4.79010224e-01
-3.94495316e-02 9.87431645e-01 5.66552579e-01 -2.61574745e-01
1.94055364e-01 3.78586888e-01 1.55363634e-01 9.54132080e-01
-5.83484620e-02 3.48964572e-01 -6.34315193e-01 -6.13147905e-03
4.10999328e-01 6.25365376e-02 -1.80967227e-01 -7.10971892e-01
-6.45802617e-01 7.42749691e-01 1.18877399e+00 2.16439635e-01
-4.30950671e-01 3.16636354e-01 1.15780279e-01 5.71897686e-01
3.82967353e-01 5.24973392e-01 -3.20847273e-01 3.33058953e-01
-1.02536213e+00 -7.22465143e-02 6.35787129e-01 5.32167554e-01
1.49430263e+00 -3.11100930e-01 -2.70228595e-01 7.19009340e-01
5.87494671e-01 2.23051369e-01 1.95370525e-01 -6.36446893e-01
4.26321268e-01 1.06922758e+00 -2.30657011e-01 -1.17852437e+00
-4.75233674e-01 -8.41240764e-01 -6.83607638e-01 3.43897134e-01
4.13515687e-01 9.33551490e-02 -5.78095615e-01 1.89093232e+00
6.75183654e-01 3.33870351e-02 -3.49364161e-01 1.08783388e+00
6.97656929e-01 2.78163135e-01 -4.32770513e-02 1.43843874e-01
7.80245781e-01 -1.23143888e+00 -4.14229482e-01 -8.33530724e-02
9.65311110e-01 -5.22123456e-01 1.00898361e+00 -7.19553754e-02
-1.18641829e+00 -4.20806527e-01 -1.19292533e+00 -1.53860494e-01
-2.60125965e-01 -7.89771322e-03 6.58045113e-01 1.54109061e-01
-1.11844301e+00 1.00432718e+00 -7.43290305e-01 1.63151068e-03
5.13787806e-01 6.60767436e-01 -1.89338923e-01 8.34926516e-02
-1.25752366e+00 6.02858484e-01 2.69338250e-01 3.01009059e-01
-4.64386404e-01 -1.01130784e+00 -7.53594220e-01 3.22740138e-01
2.04108238e-01 -8.07812929e-01 5.47861934e-01 -1.44193697e+00
-1.53166533e+00 6.51597559e-01 1.00372143e-01 -4.30704325e-01
6.74940348e-01 -1.70493007e-01 1.68385103e-01 5.58647923e-02
-1.62296951e-01 7.18874931e-01 8.88708353e-01 -8.39338422e-01
-3.95026207e-01 -4.94123757e-01 2.25705758e-01 1.98134571e-01
-5.83161891e-01 -5.03531575e-01 -5.93258560e-01 -6.78346872e-01
1.69362247e-01 -7.43539453e-01 -2.61527687e-01 5.07476151e-01
-4.21729386e-01 -7.33805299e-02 4.67970520e-01 -5.04608214e-01
1.23347437e+00 -2.26083469e+00 5.12748837e-01 7.86504328e-01
3.93870682e-01 -2.60800421e-02 -3.45173895e-01 4.47365284e-01
-5.95919834e-03 -7.77589828e-02 -2.43728250e-01 -5.53201437e-01
7.29805753e-02 -9.31599140e-02 -6.20541275e-02 6.21466219e-01
5.77301383e-01 1.13501573e+00 -1.03454936e+00 -4.74014580e-01
1.45274654e-01 4.50285584e-01 -8.49937975e-01 4.19617653e-01
-3.44047964e-01 4.45467442e-01 -2.99841285e-01 3.66476566e-01
6.45365417e-01 -4.26453590e-01 2.94254184e-01 -4.62087929e-01
4.49440554e-02 4.62322325e-01 -1.08266354e+00 1.96872079e+00
-5.49920261e-01 2.22354650e-01 3.94505203e-01 -1.12474680e+00
8.85480881e-01 -1.13062702e-01 5.03163755e-01 -6.41587019e-01
1.33207902e-01 3.25190723e-01 -1.21797018e-01 -5.30715473e-02
1.16670065e-01 2.61569858e-01 3.20311129e-01 6.43791795e-01
1.15482636e-01 2.56680310e-01 3.69796390e-03 3.32319349e-01
8.63422811e-01 1.89395636e-01 -1.87133938e-01 -2.23962307e-01
7.72818208e-01 -2.57151335e-01 3.83211315e-01 5.43066025e-01
2.02216774e-01 3.22370440e-01 4.58011955e-01 -4.17279661e-01
-7.73362398e-01 -9.54130650e-01 6.04914911e-02 1.18885839e+00
3.05301160e-01 -3.92011762e-01 -5.07903099e-01 -1.07626283e+00
2.26551220e-01 2.16857269e-01 -6.49483621e-01 -6.30255044e-01
-6.03569508e-01 -5.54322004e-01 1.43428952e-01 5.72846651e-01
5.23726404e-01 -8.46380174e-01 -6.46820143e-02 1.87224373e-01
1.68337956e-01 -5.79662502e-01 -7.50295520e-01 2.42094874e-01
-8.63484561e-01 -8.58926713e-01 -3.90281647e-01 -8.21575344e-01
8.94843221e-01 -8.90232176e-02 8.31232369e-01 5.95761418e-01
-1.09139919e-01 9.28285271e-02 2.82464009e-02 1.86449036e-01
-2.48388797e-01 3.46200794e-01 -3.87099445e-01 3.71090740e-01
3.76693346e-02 -8.97494793e-01 -6.67384267e-01 1.35803565e-01
-5.76451838e-01 1.94472909e-01 6.74367070e-01 1.11404502e+00
5.56069314e-01 -3.47647578e-01 3.42079550e-01 -7.33872414e-01
3.52952987e-01 -3.74141663e-01 -9.74083066e-01 3.29855382e-01
-9.62498724e-01 7.22218990e-01 5.99173665e-01 -2.55265355e-01
-7.36122429e-01 2.76348919e-01 -1.38856336e-01 -5.04944801e-01
4.00550306e-01 6.72432542e-01 -5.42480111e-01 -2.93359041e-01
3.05196077e-01 6.63324967e-02 1.13818586e-01 -4.68782187e-01
7.18646705e-01 1.88235462e-01 2.96160102e-01 -4.94807720e-01
9.89989042e-01 5.03480375e-01 5.05952649e-02 -2.03549460e-01
-5.84039390e-01 -3.31607878e-01 -5.43500721e-01 -3.55866194e-01
6.06053591e-01 -6.05600417e-01 -6.54228687e-01 5.35799146e-01
-1.00563467e+00 -2.39452735e-01 -5.41533411e-01 4.62265611e-01
-1.00513905e-01 1.38667628e-01 -5.44572115e-01 -5.13966739e-01
-5.11348426e-01 -1.07092440e+00 9.53621149e-01 1.41810253e-01
5.98238073e-02 -1.33978724e+00 6.63090423e-02 4.34111021e-02
4.56361324e-01 -7.95375835e-03 1.01053119e+00 -6.74527287e-01
-9.02040958e-01 -4.25662547e-01 -3.53774160e-01 2.06149966e-01
1.81461032e-02 -1.00830756e-01 -1.06061518e+00 -2.94543028e-01
-3.16324949e-01 -4.39417571e-01 1.21589696e+00 1.71425924e-01
1.13290739e+00 -3.01823646e-01 -2.35669568e-01 9.98903930e-01
1.35956538e+00 -5.55788517e-01 3.49096715e-01 2.98109710e-01
9.01429713e-01 7.09693849e-01 2.75432318e-01 2.81835794e-01
4.89987582e-01 7.46695936e-01 5.98026276e-01 -4.74872649e-01
-2.34240577e-01 -5.53952277e-01 3.25969279e-01 8.33202362e-01
3.45856637e-01 1.90399468e-01 -6.57717407e-01 2.01205075e-01
-2.02997112e+00 -7.66032517e-01 2.19902813e-01 2.42177367e+00
9.69394326e-01 2.35133678e-01 -2.34275516e-02 7.13116899e-02
6.37269795e-01 3.64558131e-01 -6.97665811e-01 -1.80745617e-01
-4.64173453e-03 5.54664582e-02 3.40091288e-01 8.81568730e-01
-1.00281084e+00 8.03344846e-01 4.89320040e+00 6.39576614e-01
-1.43185377e+00 -1.27746552e-01 5.49826622e-01 2.81270929e-02
-8.97547901e-01 1.61206007e-01 -4.54429239e-01 3.37576002e-01
4.32737947e-01 1.08396128e-01 6.53927863e-01 7.28456318e-01
-1.40551925e-01 2.22974002e-01 -1.31245267e+00 5.99810779e-01
-3.13144296e-01 -1.39328444e+00 2.71236710e-02 -6.07324168e-02
5.25112331e-01 4.55495976e-02 1.23332262e-01 1.24853596e-01
2.14459494e-01 -7.17982411e-01 6.75351560e-01 4.16504562e-01
2.86800236e-01 -7.68736780e-01 6.94832444e-01 2.12953597e-01
-1.35835171e+00 -1.34904414e-01 -1.69889972e-01 1.22403428e-01
5.84038571e-02 9.30191696e-01 -9.31766152e-01 6.12999201e-01
3.97859246e-01 9.53620613e-01 -5.05396426e-01 9.71852899e-01
-4.81041700e-01 2.51963884e-01 -4.18967873e-01 1.77638186e-03
3.00470412e-01 -4.91280556e-01 5.63250244e-01 1.08147490e+00
-2.30422869e-01 -5.39897621e-01 3.65874320e-01 1.15127122e+00
-3.84327352e-01 3.14549118e-01 -1.90864414e-01 -2.38067031e-01
4.03804451e-01 1.38887036e+00 -4.45147336e-01 -1.13604784e-01
-2.54050702e-01 8.93252969e-01 9.02577460e-01 2.78859228e-01
-7.93304324e-01 -6.11828744e-01 5.16040087e-01 1.50665000e-01
3.67441505e-01 1.24955155e-01 -5.75205266e-01 -1.12956214e+00
2.60323614e-01 -4.22032177e-01 4.71005708e-01 -1.06238335e-01
-1.21802306e+00 3.73768002e-01 -4.13980186e-01 -9.41025853e-01
-3.15920077e-02 -2.53818184e-01 -7.89677024e-01 9.33731556e-01
-1.93100357e+00 -1.25387669e+00 -3.66542339e-01 3.60078633e-01
-9.72191710e-03 5.40445745e-02 4.42008585e-01 6.16073191e-01
-6.73228502e-01 1.11707747e+00 -2.46329799e-01 2.75880918e-02
6.55645967e-01 -1.29675555e+00 1.32359639e-01 6.88848794e-01
8.83880258e-02 5.79923272e-01 2.37279966e-01 -4.94192481e-01
-1.34077728e+00 -1.11633289e+00 1.09214866e+00 3.25588440e-03
8.32212865e-01 -4.08232629e-01 -1.02744269e+00 9.23761651e-02
-3.88891138e-02 2.20728695e-01 5.61843991e-01 2.03842193e-01
-7.16077864e-01 -4.97851670e-01 -9.74113941e-01 2.83740222e-01
1.09571421e+00 -7.06025362e-01 -1.17424630e-01 1.67800397e-01
7.13811040e-01 -4.11590450e-02 -8.87155473e-01 4.54698294e-01
7.46350646e-01 -8.41837704e-01 8.12627971e-01 -7.01585174e-01
4.05885190e-01 -2.94008315e-01 8.08355957e-02 -1.10559249e+00
-2.55115241e-01 -5.14087141e-01 -1.97271839e-01 1.21138680e+00
8.05742562e-01 -6.85347736e-01 8.85506690e-01 6.87533081e-01
-1.77326277e-01 -1.16120744e+00 -6.81926608e-01 -4.09874290e-01
-1.99287347e-02 -1.92817435e-01 8.00377369e-01 9.83466804e-01
5.82897477e-02 1.92169696e-01 -2.32409537e-01 1.68804541e-01
7.18299270e-01 3.28219444e-01 7.05433249e-01 -1.11359060e+00
-6.94583356e-01 -7.04356611e-01 -4.28136408e-01 -1.25876498e+00
2.71519035e-01 -1.53206182e+00 -8.74447636e-03 -1.35148931e+00
1.46056697e-01 -7.03810334e-01 -5.87609291e-01 5.28309941e-01
-5.11799753e-01 -8.86997730e-02 1.51142910e-01 2.65365243e-01
-5.08876026e-01 8.89700592e-01 1.13937163e+00 -3.82440209e-01
-3.55032772e-01 -1.60612121e-01 -4.48902696e-01 4.64424521e-01
6.98114812e-01 -5.58485508e-01 -3.34476441e-01 -6.64532661e-01
5.03301442e-01 -3.34785104e-01 4.07774299e-01 -6.05282664e-01
5.82534194e-01 1.57508165e-01 2.92823583e-01 -2.29587749e-01
1.40656143e-01 -1.02739000e+00 -4.59425300e-02 5.92697144e-01
-6.29306674e-01 -1.16353631e-01 -1.71954334e-01 6.42441571e-01
-1.28995135e-01 3.07624210e-02 7.86093771e-01 1.97879896e-01
-2.99030393e-01 4.52910066e-01 1.93678215e-01 2.71721743e-02
6.28333747e-01 -1.20350726e-01 -1.94351017e-01 -3.12748462e-01
-6.79144681e-01 6.90599024e-01 6.49221659e-01 1.01152956e-01
5.11534631e-01 -1.44892824e+00 -5.83082318e-01 5.53969979e-01
7.53961876e-02 -7.52716511e-02 -1.47765860e-01 1.41389215e+00
-1.41441077e-01 -3.31173390e-02 5.46853356e-02 -6.63900316e-01
-8.07652414e-01 5.17662823e-01 5.61987758e-01 -5.16694427e-01
-4.04420704e-01 1.03055155e+00 4.96530123e-02 -7.56632328e-01
6.03971362e-01 -1.78752512e-01 -3.14556174e-02 1.71962827e-01
3.06163669e-01 3.00521970e-01 2.20829919e-01 -2.21872434e-01
-4.88650769e-01 4.54938740e-01 -1.82699576e-01 6.23203106e-02
1.36209369e+00 2.22522821e-02 -2.72360891e-01 1.50241762e-01
1.65003765e+00 1.07601218e-01 -1.26371646e+00 -4.42295283e-01
1.49607554e-01 -3.75798851e-01 1.86389297e-01 -5.50855279e-01
-1.55062735e+00 7.37241268e-01 4.98149604e-01 -2.00407341e-01
1.16484964e+00 4.26830165e-02 8.02315235e-01 4.03872281e-01
1.03306301e-01 -1.04390407e+00 2.16971844e-01 9.06865597e-02
7.09569097e-01 -1.30864394e+00 -4.38498333e-02 -4.03496206e-01
-2.63139904e-01 1.12955379e+00 5.05810499e-01 -1.74055248e-01
9.09130692e-01 2.21971288e-01 -4.20570411e-02 -1.61704674e-01
-7.95706928e-01 -1.44664958e-01 4.85834062e-01 1.11250952e-01
3.55296463e-01 -2.81657755e-01 -4.12962317e-01 5.58519125e-01
7.55016133e-02 -3.79759312e-01 -3.09945077e-01 8.52659643e-01
-4.13377643e-01 -1.35690260e+00 2.39842772e-01 2.96321899e-01
1.50317431e-03 -1.70919657e-01 -6.93457365e-01 4.58357900e-01
1.85327977e-02 5.90641022e-01 7.83523247e-02 -5.75654566e-01
1.19701110e-01 -1.66083425e-01 2.99076349e-01 -3.06132853e-01
-9.73427355e-01 1.57938935e-02 1.88162308e-02 -9.36764002e-01
-2.61577100e-01 -3.09044063e-01 -1.43168187e+00 -1.04425997e-01
-7.27870286e-01 3.88334811e-01 6.17682219e-01 8.35761726e-01
4.62150633e-01 3.98966283e-01 1.00042903e+00 -7.20783234e-01
-7.09100425e-01 -4.34082448e-01 -1.02038331e-01 4.59284931e-01
4.62998450e-01 -4.19234872e-01 -5.77053666e-01 -4.34267223e-01] | [7.19616174697876, 6.256736755371094] |
8425dd6f-d6d2-4893-b88d-8688dbae27f3 | end-to-end-detection-and-re-identification | 1804.00376 | null | http://arxiv.org/abs/1804.00376v1 | http://arxiv.org/pdf/1804.00376v1.pdf | End-to-End Detection and Re-identification Integrated Net for Person Search | This paper proposes a pedestrian detection and re-identification (re-id)
integration net (I-Net) in an end-to-end learning framework. The I-Net is used
in real-world video surveillance scenarios, where the target person needs to be
searched in the whole scene videos, while the annotations of pedestrian
bounding boxes are unavailable. By comparing to the OIM which is a work for
joint detection and re-id, we have three distinct contributions. First, we
introduce a Siamese architecture of I-Net instead of 1 stream, such that a
verification task can be implemented. Second, we propose a novel on-line
pairing loss (OLP) and hard example priority softmax loss (HEP), such that only
the hard negatives are posed much attention in loss computation. Third, an
on-line dictionary for negative samples storage is designed in I-Net without
recording the positive samples. We show our result on person search datasets,
the gap between detection and re-identification is narrowed. The superior
performance can be achieved. | ['Wei Jia', 'Zhenwei He', 'Lei Zhang'] | 2018-04-02 | null | null | null | null | ['person-search'] | ['computer-vision'] | [-1.84651300e-01 -3.71785849e-01 -8.68262649e-02 -4.09521252e-01
-5.82356155e-01 -2.95784861e-01 4.57686245e-01 -1.12788351e-02
-1.12303829e+00 4.72589999e-01 -1.37984842e-01 1.15043230e-01
3.53685260e-01 -6.19468451e-01 -6.71854496e-01 -5.33017039e-01
-1.38987884e-01 2.61609942e-01 6.29622042e-01 1.01644255e-01
-2.54228443e-01 3.94930780e-01 -1.48019123e+00 2.64826059e-01
5.75514913e-01 1.41804063e+00 4.78714705e-02 6.92257524e-01
4.93176520e-01 7.51505077e-01 -6.62871480e-01 -1.09846497e+00
7.60092020e-01 -8.33130926e-02 -3.25629801e-01 -1.04538456e-01
9.33964252e-01 -9.48060870e-01 -9.21473265e-01 1.13847983e+00
1.02100325e+00 2.19454587e-01 3.85559738e-01 -1.64768898e+00
-1.63577512e-01 2.02360973e-01 -5.75483859e-01 1.64265633e-01
2.66762435e-01 4.48450148e-01 7.89970934e-01 -1.12865126e+00
2.93589324e-01 1.51636958e+00 9.48873878e-01 6.93995595e-01
-8.81513834e-01 -9.19801354e-01 2.77278751e-01 6.62206829e-01
-1.68781698e+00 -4.28534538e-01 3.96627069e-01 -4.48518455e-01
8.05293381e-01 3.76960546e-01 7.80868471e-01 1.28003824e+00
-2.61726350e-01 1.17846584e+00 4.26387817e-01 -2.63787746e-01
5.51731661e-02 3.15222293e-01 2.75645196e-01 6.78628147e-01
3.71754855e-01 4.09125447e-01 -6.91218317e-01 -9.94852632e-02
4.39381987e-01 3.39036644e-01 -1.88728228e-01 -3.48154426e-01
-1.05665433e+00 4.57877040e-01 4.49071258e-01 -3.44141871e-01
-7.59593993e-02 1.12822361e-01 9.32128787e-01 1.32580832e-01
1.55493289e-01 -3.82000208e-01 -1.02379560e-01 -3.76573466e-02
-1.04140055e+00 4.61247504e-01 6.33439124e-01 1.26535594e+00
4.30042237e-01 -2.60508806e-01 -7.24078953e-01 9.33431447e-01
1.99519843e-01 8.22800815e-01 3.18813115e-01 -6.77327752e-01
7.03579783e-01 2.66174525e-01 3.47032875e-01 -9.87340569e-01
-1.60606414e-01 -4.59115177e-01 -1.13811052e+00 1.19312070e-01
4.42577749e-01 -2.26296097e-01 -6.01036012e-01 1.80708551e+00
2.63839304e-01 3.91793251e-01 -1.34633541e-01 1.27691841e+00
7.77949512e-01 3.21334958e-01 1.43427178e-01 1.68245006e-02
1.48763561e+00 -1.31987154e+00 -4.99803603e-01 -1.43133402e-01
3.26719761e-01 -4.58249629e-01 6.78705335e-01 4.65656757e-01
-1.06986749e+00 -8.50211799e-01 -9.62491035e-01 -1.39795497e-01
-3.40873480e-01 5.65145791e-01 2.15750426e-01 8.09796035e-01
-9.79520679e-01 1.81679606e-01 -7.39345372e-01 -4.54031467e-01
4.85883981e-01 4.66888428e-01 -4.45950717e-01 1.39235770e-02
-1.06605554e+00 7.07927585e-01 4.18553531e-01 4.16659325e-01
-9.98550534e-01 -6.01238728e-01 -9.22424197e-01 1.61165267e-01
6.28997147e-01 -7.05460370e-01 1.02894032e+00 -7.83804238e-01
-1.16617250e+00 8.83364558e-01 -3.63011748e-01 -7.19272971e-01
1.08159840e+00 -4.61938262e-01 -3.20802003e-01 7.13647082e-02
2.31416479e-01 7.24044025e-01 6.89656675e-01 -1.01875567e+00
-1.00627553e+00 -3.89406383e-01 -2.18691621e-02 4.42127697e-02
-5.33942878e-01 2.66895056e-01 -1.13897443e+00 -8.67179811e-01
-4.67647940e-01 -8.43319774e-01 -7.70843327e-02 5.57071388e-01
-5.48567533e-01 -2.94555426e-01 8.73268902e-01 -8.13030183e-01
1.44978714e+00 -2.29067302e+00 -1.95572764e-01 3.07954013e-01
2.81780571e-01 7.63177931e-01 -1.38986886e-01 9.77295190e-02
3.27474363e-02 -2.20878705e-01 2.76787244e-02 -8.64241779e-01
2.87111044e-01 -1.29066303e-01 -1.85036331e-01 5.55953622e-01
1.05598390e-01 7.91249752e-01 -7.08394527e-01 -6.51048064e-01
2.62577564e-01 2.67926276e-01 -6.50554955e-01 3.76928478e-01
2.70854801e-01 -1.22021399e-01 -3.60221788e-02 7.59996891e-01
1.09014881e+00 -2.92204246e-02 -1.45328537e-01 -5.69048762e-01
-2.15848505e-01 -1.93379372e-01 -1.51375139e+00 1.33207250e+00
-8.15060064e-02 3.48500133e-01 1.66418672e-01 -8.22181284e-01
5.42692006e-01 3.81383412e-02 2.57130772e-01 -8.27547729e-01
-4.16790321e-02 2.27115918e-02 -3.54005218e-01 -3.20727348e-01
4.43546206e-01 4.79389757e-01 1.50228469e-02 2.67696716e-02
1.28099337e-01 9.67080116e-01 4.29597169e-01 2.44979933e-01
7.80400097e-01 -1.33987725e-01 6.81748986e-02 9.99177527e-03
9.71913815e-01 -2.95820326e-01 9.22402978e-01 1.08846498e+00
-5.84504128e-01 4.83898550e-01 2.18294814e-01 -5.21454990e-01
-1.06613135e+00 -1.07337332e+00 -9.55522507e-02 1.14160526e+00
5.44171929e-01 -4.65228796e-01 -7.96760023e-01 -7.41287172e-01
1.71099365e-01 -8.70479085e-03 -3.05109173e-01 -6.85949996e-02
-7.99245298e-01 -8.14957678e-01 8.71477783e-01 6.59853816e-01
1.01653409e+00 -4.83019859e-01 -5.81410885e-01 -5.91607392e-02
-2.73041815e-01 -1.26970601e+00 -1.08623183e+00 -1.46553904e-01
-2.95392901e-01 -1.16022730e+00 -1.06696689e+00 -8.76902699e-01
5.84093451e-01 4.21601593e-01 7.88369179e-01 -1.08247876e-01
-4.21085656e-01 4.68758821e-01 -1.83142975e-01 -2.50388861e-01
2.95254320e-01 -2.22902253e-01 4.21665609e-01 4.53884095e-01
6.69363678e-01 -5.85298501e-02 -8.54102492e-01 6.00016117e-01
-3.56039405e-01 -3.00506949e-01 2.79094726e-01 1.14216959e+00
6.21853828e-01 -1.23684198e-01 3.08340073e-01 -3.01944137e-01
3.36794317e-01 -1.73778474e-01 -8.56951237e-01 5.25403440e-01
-4.00067031e-01 -3.60742718e-01 5.30845702e-01 -4.83645111e-01
-8.90388668e-01 1.33830413e-01 -3.37942451e-01 -5.89964390e-01
4.69208658e-02 -2.61895865e-01 -5.51882267e-01 -2.63221323e-01
2.38609493e-01 2.70743668e-01 -3.22942249e-02 -5.42093992e-01
4.19669077e-02 7.70417929e-01 8.20097089e-01 -3.39757413e-01
6.29036188e-01 5.00182033e-01 -1.58345833e-01 -8.56171727e-01
-6.46350265e-01 -8.10468078e-01 -4.72269773e-01 -2.23595664e-01
9.19622123e-01 -1.28883958e+00 -1.29014277e+00 8.31141233e-01
-1.29149818e+00 3.82228929e-04 -2.44652584e-01 6.22670650e-01
-1.61852866e-01 8.64015102e-01 -6.72613919e-01 -1.09385979e+00
-5.39257348e-01 -1.19010520e+00 1.13188732e+00 2.54101187e-01
2.02920452e-01 -5.09801209e-01 -2.82776535e-01 1.18754596e-01
1.45786896e-01 -7.22043067e-02 1.38659880e-01 -8.04754794e-01
-6.90192342e-01 -4.19386029e-01 -6.60629034e-01 5.77003360e-01
-3.05478096e-01 -4.52225953e-01 -1.05283141e+00 -8.13826442e-01
-2.35231176e-01 -2.78189301e-01 1.07624304e+00 2.77216583e-01
1.30723906e+00 -5.05404770e-01 -5.40465295e-01 9.73775566e-01
1.21098280e+00 1.88001171e-01 4.83765185e-01 3.73832464e-01
6.73330367e-01 3.75362337e-01 6.80337012e-01 8.00546587e-01
7.48899519e-01 1.08555806e+00 3.22160393e-01 -3.08723629e-01
-1.95804968e-01 -3.39522928e-01 6.09069407e-01 2.40330577e-01
8.17907695e-03 -3.62476259e-01 -5.12763739e-01 4.71668512e-01
-2.20272326e+00 -1.35136926e+00 3.59034725e-02 2.56743002e+00
4.77050900e-01 3.83145921e-02 8.00739050e-01 2.70982534e-02
9.50068533e-01 -6.47803396e-02 -7.00271726e-01 4.16050613e-01
-2.59228915e-01 -1.84980690e-01 8.11319172e-01 5.00398874e-01
-1.49947596e+00 6.77876770e-01 5.68968105e+00 1.03545618e+00
-9.80447948e-01 2.93354958e-01 5.97044945e-01 -3.59158039e-01
5.21604419e-01 -3.45311940e-01 -1.38092101e+00 8.57452452e-01
5.30809343e-01 8.00470486e-02 8.34363326e-02 9.76343572e-01
2.18094379e-01 -6.78340495e-02 -1.28671706e+00 1.78714228e+00
3.59229386e-01 -9.79782343e-01 2.12022793e-02 -2.53391504e-01
3.97872776e-02 -1.86490238e-01 7.71202296e-02 3.51384133e-01
-4.99261951e-04 -5.50414801e-01 1.10576487e+00 5.96208334e-01
9.09713387e-01 -8.30216527e-01 9.38430309e-01 3.66304755e-01
-1.44306505e+00 -3.36578012e-01 -5.20176947e-01 3.16860795e-01
4.02050167e-01 4.58304405e-01 -4.34580386e-01 5.39964914e-01
1.12222505e+00 7.12843478e-01 -6.86615288e-01 1.62806582e+00
1.19199388e-01 3.92981529e-01 -5.01756668e-01 6.47293627e-02
-4.47452292e-02 -2.02471148e-02 5.87817311e-01 1.86459017e+00
1.01338893e-01 -1.39505535e-01 5.77482343e-01 7.00354218e-01
-1.09551987e-02 -1.73104256e-01 -1.66699260e-01 6.58435225e-01
3.26345384e-01 9.78419781e-01 -1.24164850e-01 -5.60842216e-01
-4.57516134e-01 1.42416763e+00 9.62885544e-02 3.55947614e-01
-1.10404980e+00 -6.16527438e-01 6.84232771e-01 1.10797659e-01
4.04990822e-01 4.28859890e-02 1.33043081e-01 -1.44753206e+00
4.42274690e-01 -6.99752152e-01 6.00209892e-01 -2.02857539e-01
-1.63256693e+00 4.20211136e-01 1.13515584e-02 -1.33101273e+00
8.23627561e-02 -6.51387334e-01 -3.11221570e-01 9.87537086e-01
-1.59486842e+00 -1.22403181e+00 -5.64406395e-01 8.74817908e-01
4.55074161e-01 -4.14174587e-01 3.24197292e-01 1.14655113e+00
-9.16505754e-01 1.53990149e+00 -1.51110470e-01 7.41330028e-01
9.50011492e-01 -9.32112575e-01 4.09461737e-01 1.25370967e+00
-2.73874730e-01 5.42894542e-01 2.02293932e-01 -5.97871363e-01
-1.24865234e+00 -1.36365545e+00 9.02306557e-01 -1.92553028e-01
2.70537794e-01 -7.94839621e-01 -4.83113170e-01 4.58614469e-01
-1.06547341e-01 4.01389152e-01 5.75170934e-01 -2.28071019e-01
-4.67295974e-01 -5.56620002e-01 -1.32427502e+00 4.48145092e-01
1.29080927e+00 -4.81600016e-01 -1.40424132e-01 4.11038816e-01
6.22581005e-01 -3.47314149e-01 -5.03781021e-01 2.61037827e-01
6.60816491e-01 -1.05809176e+00 1.42277050e+00 -4.85011786e-01
-1.88230097e-01 -5.25540590e-01 -1.08576261e-01 -5.35093546e-01
-3.43736500e-01 -5.84399223e-01 -2.96389878e-01 1.29714310e+00
1.01831317e-01 -6.29666030e-01 7.66503870e-01 7.70547926e-01
1.09875850e-01 -5.33110499e-01 -1.19683373e+00 -1.07297492e+00
-5.16218901e-01 -3.45832705e-01 4.64102447e-01 4.27151352e-01
-3.33469808e-01 -1.34700898e-03 -9.31257784e-01 3.58561546e-01
1.15467381e+00 -4.42370713e-01 9.36707675e-01 -9.51025009e-01
-5.04304409e-01 -2.15035483e-01 -6.77621186e-01 -1.75566995e+00
-9.11315680e-02 -7.52458692e-01 -1.06419124e-01 -9.79670584e-01
5.26656508e-01 -5.00289202e-01 -3.44359696e-01 3.66928458e-01
-1.66755572e-01 2.78167546e-01 6.80734634e-01 3.39278281e-01
-1.07331383e+00 6.78691208e-01 6.08813643e-01 -4.31012660e-01
7.00610727e-02 2.51351953e-01 -2.96593487e-01 7.22344637e-01
2.65990824e-01 -2.70885646e-01 -9.88757163e-02 -4.49973732e-01
-2.92149514e-01 -1.27525389e-01 1.02198851e+00 -1.25605941e+00
8.13719988e-01 2.48765245e-01 5.94178557e-01 -9.45864081e-01
5.29496372e-01 -8.72195184e-01 -2.31518865e-01 5.90731621e-01
-2.98952609e-01 2.22972140e-01 6.33253381e-02 6.95946038e-01
-3.28438282e-01 -2.07076639e-01 9.08239365e-01 -9.04829055e-03
-8.35161030e-01 6.62219524e-01 1.35359243e-01 5.79600446e-02
1.16261804e+00 -4.53352153e-01 -2.21932754e-01 -9.83687416e-02
-5.47705472e-01 6.02216125e-01 2.61743456e-01 2.35277638e-01
8.16678047e-01 -1.55664432e+00 -7.42796540e-01 2.44232342e-01
7.52523914e-02 -1.40650839e-01 4.03202832e-01 7.62168467e-01
-3.75468403e-01 4.20393258e-01 1.41030671e-02 -5.80770195e-01
-1.56234837e+00 7.29557574e-01 4.34695691e-01 -2.13530630e-01
-6.62169397e-01 1.06390107e+00 3.05329293e-01 -2.01394528e-01
1.05830073e+00 1.47433933e-02 4.09758501e-02 4.32307720e-02
1.05178642e+00 7.52384603e-01 -1.61047816e-01 -6.81564569e-01
-5.23264408e-01 3.18026125e-01 -2.10175946e-01 -1.52516570e-02
9.94941652e-01 -1.77714586e-01 -5.89636574e-03 -1.18465863e-01
1.12557042e+00 -1.69185966e-01 -1.35443389e+00 -3.49503845e-01
-2.48887137e-01 -7.72822559e-01 -3.00541252e-01 -6.34565771e-01
-1.04590714e+00 8.12142670e-01 1.23125982e+00 -3.18397224e-01
1.15498435e+00 -3.66068870e-01 1.23411596e+00 5.32270432e-01
3.25518757e-01 -1.28610396e+00 -9.75293166e-04 4.26474929e-01
6.69842064e-01 -1.45094573e+00 -2.27292582e-01 -2.72655606e-01
-6.20171845e-01 8.19492757e-01 8.96295309e-01 3.51458713e-02
6.08489931e-01 2.62268662e-01 -4.52608109e-01 3.07415247e-01
-3.63684863e-01 -2.99074709e-01 1.08244397e-01 8.91558051e-01
-1.77127212e-01 -1.95304453e-01 -1.38324574e-01 9.12361622e-01
9.28571448e-02 -2.35531740e-02 1.01560138e-01 5.80606282e-01
-2.55648971e-01 -1.03452682e+00 -4.74366069e-01 3.13412398e-01
-3.33602846e-01 -1.75437406e-01 -1.12355798e-01 4.30000424e-01
5.71400821e-01 8.00303221e-01 1.42602012e-01 -5.12805045e-01
6.76774561e-01 -2.46851996e-01 1.74216449e-01 -9.79256481e-02
-6.76345170e-01 -1.47235751e-01 -6.11336669e-03 -7.69032657e-01
-2.11886629e-01 -4.38033044e-01 -7.08812714e-01 -4.22824621e-01
-3.46830457e-01 -6.24829950e-03 1.84902638e-01 4.37165588e-01
4.61845964e-01 1.28956437e-01 4.40235436e-01 -8.46071422e-01
-8.77683282e-01 -6.11914337e-01 -3.68815035e-01 4.10278171e-01
6.29130840e-01 -6.80055261e-01 -2.23164439e-01 1.42406717e-01] | [14.792764663696289, 0.8504324555397034] |
e552c6f6-76f6-4031-a2f4-05fddfa9880b | cross-referencing-self-training-network-for | 2105.13392 | null | https://arxiv.org/abs/2105.13392v1 | https://arxiv.org/pdf/2105.13392v1.pdf | Cross-Referencing Self-Training Network for Sound Event Detection in Audio Mixtures | Sound event detection is an important facet of audio tagging that aims to identify sounds of interest and define both the sound category and time boundaries for each sound event in a continuous recording. With advances in deep neural networks, there has been tremendous improvement in the performance of sound event detection systems, although at the expense of costly data collection and labeling efforts. In fact, current state-of-the-art methods employ supervised training methods that leverage large amounts of data samples and corresponding labels in order to facilitate identification of sound category and time stamps of events. As an alternative, the current study proposes a semi-supervised method for generating pseudo-labels from unsupervised data using a student-teacher scheme that balances self-training and cross-training. Additionally, this paper explores post-processing which extracts sound intervals from network prediction, for further improvement in sound event detection performance. The proposed approach is evaluated on sound event detection task for the DCASE2020 challenge. The results of these methods on both "validation" and "public evaluation" sets of DESED database show significant improvement compared to the state-of-the art systems in semi-supervised learning. | ['Mounya Elhilali', 'David K. Han', 'Sangwook Park'] | 2021-05-27 | null | null | null | null | ['audio-tagging'] | ['audio'] | [ 3.58679801e-01 -1.61901470e-02 2.37473205e-01 -5.44986010e-01
-1.11811578e+00 -5.35746872e-01 2.95777291e-01 6.18832588e-01
-3.86550874e-01 4.94461656e-01 2.92909145e-01 -7.31154680e-02
-8.38870481e-02 -6.82749987e-01 -3.83622408e-01 -3.44103068e-01
-3.37145060e-01 4.24767584e-02 5.36084235e-01 1.99087694e-01
-5.28521910e-02 2.17300832e-01 -1.91114843e+00 5.54554045e-01
2.52028197e-01 1.31067371e+00 -6.44362494e-02 8.53631735e-01
-5.87747581e-02 8.91056538e-01 -8.96549642e-01 9.53896567e-02
-4.63813394e-02 -4.68955517e-01 -7.74651170e-01 -2.76657373e-01
2.92198718e-01 2.71612369e-02 -2.62206614e-01 8.03235531e-01
8.73279870e-01 3.04252446e-01 4.27225530e-01 -1.23229229e+00
5.95108271e-02 1.08880115e+00 -1.99276134e-02 5.41908026e-01
3.16326112e-01 -3.83749843e-01 1.08395553e+00 -6.33208156e-01
1.45257086e-01 6.14151776e-01 1.03866374e+00 2.38688692e-01
-8.84544015e-01 -9.64092553e-01 -2.41523907e-01 3.01489204e-01
-1.39561081e+00 -5.40429294e-01 1.00020289e+00 -6.32807612e-01
1.18307495e+00 2.52803117e-01 5.46069920e-01 8.37492168e-01
-1.43604517e-01 4.69139099e-01 8.15601528e-01 -8.20276260e-01
4.98919934e-01 -1.85142718e-02 2.07156733e-01 4.60710198e-01
-4.74473596e-01 3.42621803e-01 -9.67326641e-01 -2.83484370e-01
3.38460177e-01 -8.89819935e-02 1.74773093e-02 3.94209236e-01
-9.56439614e-01 5.47731578e-01 -1.51465787e-02 3.23992908e-01
-4.90771919e-01 -5.50300926e-02 9.74976242e-01 9.82859805e-02
8.32421720e-01 2.01063707e-01 -5.64774990e-01 -4.78808105e-01
-1.55833268e+00 1.50374144e-01 9.32515025e-01 6.52496457e-01
4.31694090e-01 3.46038818e-01 -3.82199705e-01 1.02101636e+00
2.19628930e-01 3.46681550e-02 5.48515201e-01 -6.56474650e-01
3.94671410e-01 4.68006223e-01 -7.99485594e-02 -7.74194658e-01
-6.66066051e-01 -7.22765446e-01 -4.97584909e-01 -1.14550488e-02
2.18983129e-01 -3.80408913e-01 -7.95400381e-01 1.71686947e+00
3.20601732e-01 7.79922724e-01 -1.49482936e-01 5.56156814e-01
7.93558955e-01 7.54147232e-01 2.64891297e-01 -2.23812819e-01
1.34982264e+00 -7.33767569e-01 -7.07346916e-01 1.15350842e-01
2.10775033e-01 -7.21525371e-01 8.87687802e-01 6.00985706e-01
-7.29280293e-01 -8.38450968e-01 -9.81004596e-01 4.50941592e-01
-4.80818123e-01 4.49312150e-01 4.67880368e-01 7.80719280e-01
-4.85566020e-01 5.19003689e-01 -9.53340769e-01 -2.52942413e-01
4.47834343e-01 2.46517211e-01 6.10093735e-02 6.52739406e-01
-1.44407320e+00 2.25208551e-01 5.84436297e-01 -8.46400857e-02
-1.22303772e+00 -1.09993601e+00 -5.49718022e-01 2.67617553e-01
5.45674786e-02 3.83101851e-02 1.72834671e+00 -7.70547748e-01
-1.36787891e+00 7.77261496e-01 1.49167418e-01 -1.11781645e+00
1.99917108e-01 -4.50517505e-01 -1.11309314e+00 3.31032455e-01
5.75860962e-02 4.43058878e-01 8.94528747e-01 -6.04547799e-01
-1.34951174e+00 1.11650668e-01 -2.37586185e-01 -2.04433113e-01
-4.93505806e-01 4.59564090e-01 -9.58572477e-02 -8.77181947e-01
-1.26793697e-01 -7.39136398e-01 5.63787222e-02 -4.37496006e-01
-5.00891089e-01 -5.49272954e-01 8.71621728e-01 -6.60019040e-01
1.73416793e+00 -2.34626722e+00 -7.24695385e-01 2.87166592e-02
-5.89357577e-02 2.55087882e-01 3.34987164e-01 5.25626838e-01
-3.67451638e-01 -3.28700423e-01 -7.22429007e-02 -4.62971032e-01
1.61722198e-01 -5.11617288e-02 -8.77222598e-01 2.12153926e-01
1.64532438e-01 1.17094688e-01 -9.50743377e-01 -6.03572547e-01
1.91052958e-01 4.72552985e-01 -4.54298794e-01 4.80587512e-01
-2.24269018e-01 4.06178117e-01 -1.54492212e-02 3.82362753e-01
9.59221870e-02 1.71056762e-01 -1.32831722e-01 -4.21967268e-01
-3.98846030e-01 8.29165399e-01 -1.35796916e+00 1.67860639e+00
-4.86408472e-01 7.39361525e-01 -4.45226848e-01 -1.09608853e+00
8.50415111e-01 1.05997264e+00 8.19854498e-01 -3.10294241e-01
9.23482049e-03 1.17995001e-01 -1.61493093e-01 -5.62948704e-01
4.26556855e-01 -1.07333034e-01 -2.40898699e-01 5.89771032e-01
3.96590710e-01 1.64921299e-01 3.85883540e-01 -1.19510002e-01
1.26203430e+00 2.48609781e-01 1.79644078e-01 7.15415478e-02
4.22220916e-01 -1.43105596e-01 6.80249035e-01 6.97484374e-01
-2.00820088e-01 6.84662700e-01 1.55133173e-01 -3.52109373e-01
-6.85477614e-01 -1.10747480e+00 -2.55336255e-01 1.48673820e+00
-5.17088771e-01 -5.64077139e-01 -9.30964351e-01 -6.23476863e-01
-2.68611550e-01 8.19848478e-01 -4.71107990e-01 -1.38108313e-01
-4.84191358e-01 -6.81579173e-01 1.37306225e+00 6.46765292e-01
3.32226604e-01 -1.47414494e+00 -9.60643351e-01 7.03037024e-01
-1.62688851e-01 -1.15308809e+00 -3.45272839e-01 7.07575262e-01
-5.30768394e-01 -8.62457275e-01 -2.28696585e-01 -8.89628351e-01
2.42030978e-01 -2.74179935e-01 1.20730543e+00 -4.77435410e-01
-4.66778576e-01 3.07088584e-01 -7.07596600e-01 -9.58635628e-01
-5.23557723e-01 2.36726910e-01 1.79182306e-01 1.70714334e-01
5.04697263e-01 -9.12661493e-01 -5.99730015e-01 2.18032971e-01
-8.16879213e-01 -2.80354023e-01 1.38724804e-01 4.11169589e-01
7.55358160e-01 5.06514907e-01 1.22130263e+00 -6.99965894e-01
5.32304049e-01 -5.40939867e-01 -4.73880351e-01 -1.04933865e-02
-5.44057190e-01 -3.43430430e-01 8.08530152e-01 -5.42373836e-01
-9.92246926e-01 2.78577775e-01 -4.09649462e-01 -1.66253835e-01
-5.52437723e-01 3.68658841e-01 1.55966952e-01 5.27245104e-01
7.26220191e-01 1.28483295e-01 -6.08574092e-01 -8.75732899e-01
7.57058337e-02 1.14351118e+00 9.00652409e-01 -4.38604414e-01
4.02995735e-01 3.63097429e-01 -3.00004095e-01 -7.19611287e-01
-1.33291304e+00 -7.31686652e-01 -5.42009592e-01 -3.23674619e-01
8.21269810e-01 -9.27922964e-01 -3.56515706e-01 4.62286323e-01
-9.84707117e-01 7.90680479e-03 -6.78912461e-01 6.37202144e-01
-3.90508384e-01 -5.77963032e-02 -4.12414640e-01 -1.01808667e+00
-6.36829734e-01 -4.90119338e-01 9.20754731e-01 2.25999281e-01
-4.86227155e-01 -7.27158666e-01 4.67552423e-01 -9.68204625e-03
3.37804258e-01 1.95714533e-01 5.75913250e-01 -1.19487453e+00
6.68122023e-02 -5.17833054e-01 1.46121398e-01 6.31117702e-01
3.06539506e-01 -7.42521212e-02 -1.49728608e+00 4.98691052e-02
-9.33956802e-02 -2.09099740e-01 6.18028462e-01 3.13019961e-01
1.15037310e+00 -2.24151418e-01 -1.20344110e-01 2.91043103e-01
1.05570638e+00 5.47233820e-01 2.20793247e-01 8.32654238e-02
3.70628238e-01 3.80809188e-01 6.38459980e-01 8.44097495e-01
2.86242634e-01 5.42844236e-01 1.69047907e-01 -2.48383004e-02
-3.84483010e-01 -5.89251041e-01 2.34793544e-01 9.33593750e-01
2.51161695e-01 -1.08900204e-01 -8.97788465e-01 9.06754494e-01
-1.61773491e+00 -1.20249856e+00 3.96516398e-02 2.15992546e+00
1.21022308e+00 3.77945632e-01 4.41644728e-01 1.04796076e+00
7.97687590e-01 -9.96681955e-03 -2.22256243e-01 -1.64369434e-01
2.15745091e-01 6.75503910e-01 1.16740368e-01 -5.88893518e-02
-1.60055935e+00 6.60535574e-01 6.32953310e+00 8.82050633e-01
-1.35343075e+00 4.28883225e-01 2.69270331e-01 -1.39159665e-01
4.91544306e-01 -1.38070777e-01 -9.11130548e-01 5.98585486e-01
1.61215925e+00 1.56712577e-01 1.64144337e-01 6.75824106e-01
5.94936490e-01 1.93155408e-02 -1.27482021e+00 9.84848857e-01
-5.14981486e-02 -1.29426718e+00 -3.88532281e-01 -4.84489769e-01
5.68474412e-01 1.30736873e-01 -2.02455178e-01 3.97367984e-01
-4.02959287e-02 -5.02021074e-01 1.18370092e+00 4.47290033e-01
7.25068331e-01 -9.31693316e-01 4.88306671e-01 1.48722917e-01
-1.48966885e+00 -1.56177789e-01 6.33591637e-02 -2.17220172e-01
3.77113044e-01 9.41397309e-01 -1.15169597e+00 4.26761806e-01
1.01195014e+00 4.84418035e-01 -4.87781227e-01 1.31777859e+00
-1.40524358e-01 1.70023966e+00 -5.39623380e-01 3.05276901e-01
-7.88913891e-02 3.47598076e-01 4.98224169e-01 1.67151320e+00
3.56708348e-01 -2.47245118e-01 2.55204171e-01 6.29611075e-01
-6.64615631e-03 1.33104175e-01 -1.60163447e-01 -1.68051004e-01
7.57662594e-01 1.19782782e+00 -8.56542706e-01 -3.57281953e-01
-2.82693416e-01 3.56812179e-01 2.46843463e-03 5.04147746e-02
-1.04956865e+00 -9.44876015e-01 2.80593753e-01 1.00420363e-01
4.40424114e-01 -4.38088039e-03 -8.68075490e-02 -5.62679172e-01
-2.58184105e-01 -6.76368117e-01 6.20122731e-01 -4.55035776e-01
-1.13694763e+00 7.97185361e-01 3.84659879e-03 -1.66741037e+00
-5.48081458e-01 -6.58964962e-02 -7.97561169e-01 5.35397410e-01
-1.17860687e+00 -1.00276470e+00 -7.48124421e-02 4.26522225e-01
5.83944440e-01 -3.89061511e-01 1.01036739e+00 9.08964097e-01
-5.08800030e-01 6.46706760e-01 -1.15341522e-01 4.49808806e-01
8.01008403e-01 -1.21005583e+00 2.94118166e-01 9.02529001e-01
8.51708770e-01 1.61027387e-01 6.75334811e-01 -4.40532416e-01
-7.84793913e-01 -1.49734545e+00 1.03148246e+00 -2.18435198e-01
7.65537500e-01 -3.51655722e-01 -6.52316988e-01 3.96400839e-01
5.50593808e-02 6.56801909e-02 1.06761038e+00 1.29995018e-01
-2.89632678e-01 -5.26698947e-01 -9.09935713e-01 -2.08311919e-02
5.97335815e-01 -9.79731441e-01 -7.69648969e-01 3.03305894e-01
7.05650568e-01 -2.13147268e-01 -9.23285723e-01 3.07397425e-01
5.88461637e-01 -7.84042418e-01 8.05482924e-01 -4.81756240e-01
1.59018680e-01 -5.12951672e-01 -2.22680867e-01 -1.04293108e+00
1.48866847e-01 -8.84584606e-01 -3.20142984e-01 1.91617405e+00
4.45968449e-01 -8.70039389e-02 6.38641417e-01 4.25997414e-02
-3.53120476e-01 -3.82032186e-01 -1.05944312e+00 -7.21828818e-01
-5.31588137e-01 -1.15653300e+00 5.99458098e-01 7.09920883e-01
3.92281450e-02 3.41907322e-01 -2.73579568e-01 4.52739000e-01
5.10597289e-01 1.64314173e-02 3.93533468e-01 -1.35121953e+00
-4.02288556e-01 -8.31991285e-02 -3.90397102e-01 -4.99304235e-01
6.21976182e-02 -8.95275772e-01 4.14742857e-01 -1.09145808e+00
-2.42466971e-01 -4.12631899e-01 -9.28768396e-01 7.27662444e-01
1.03211336e-01 5.27434647e-01 -1.72165245e-01 -3.60637195e-02
-7.48250306e-01 1.57158822e-01 1.87079847e-01 1.09176390e-01
-5.05537093e-01 5.08111179e-01 -2.48866156e-01 8.78849447e-01
8.64841044e-01 -9.39972997e-01 -5.05171776e-01 -1.34569421e-01
9.16438773e-02 9.94121879e-02 4.23663527e-01 -1.65038478e+00
3.53013009e-01 2.87761569e-01 4.65955250e-02 -9.81949687e-01
1.34117022e-01 -7.90589392e-01 1.78172246e-01 1.47581547e-01
-7.29449213e-01 -1.55372500e-01 2.62015313e-01 6.34671032e-01
-5.37573040e-01 -2.83903301e-01 7.14930356e-01 2.38795564e-01
-4.59189266e-01 4.46497947e-02 -5.22074938e-01 2.78948009e-01
6.53257310e-01 9.83146206e-02 1.16376750e-01 -1.95493549e-01
-9.57031429e-01 -3.30002904e-01 -5.16862631e-01 4.83220696e-01
2.88331151e-01 -1.29085958e+00 -6.06626093e-01 1.93319306e-01
5.71450777e-02 -1.09321095e-01 1.81570262e-01 6.72241569e-01
-3.01219702e-01 3.19968581e-01 3.66888195e-02 -5.68139434e-01
-1.20895493e+00 1.59752533e-01 1.85192540e-01 -2.96830148e-01
-6.35413289e-01 9.09210443e-01 -2.65556693e-01 -1.57505274e-01
8.59992802e-01 -6.29730105e-01 -3.83658886e-01 3.38180900e-01
6.53625250e-01 5.93248904e-01 4.40189451e-01 -4.25010085e-01
-3.70912641e-01 9.84368995e-02 1.03032604e-01 -3.10170084e-01
1.39890671e+00 7.03258663e-02 1.36902839e-01 7.72308171e-01
1.01834810e+00 1.78854773e-03 -1.10436141e+00 -2.92038888e-01
3.05622071e-01 1.03079379e-01 3.20740223e-01 -1.02293074e+00
-1.02544069e+00 1.03477478e+00 1.03896117e+00 5.55451095e-01
1.40993524e+00 -1.57472968e-01 9.78991389e-01 1.84066206e-01
1.38961464e-01 -1.25504899e+00 -2.06063107e-01 4.61277187e-01
4.75243777e-01 -8.51312816e-01 -3.13545316e-01 -4.70421799e-02
-2.91336447e-01 9.62834656e-01 3.22212756e-01 -1.67020679e-01
1.12949812e+00 6.00263536e-01 1.29472688e-01 -1.44094881e-02
-6.94664657e-01 -1.65832937e-01 4.58883941e-01 4.60318595e-01
6.41502023e-01 4.18789797e-02 -1.36500418e-01 1.18423998e+00
-4.49649185e-01 2.02660233e-01 1.38345286e-02 1.04796624e+00
-4.36353266e-01 -1.03864694e+00 -3.25608879e-01 5.18795609e-01
-9.77927446e-01 -2.10037276e-01 -6.89458996e-02 2.38065451e-01
5.59754491e-01 1.29564345e+00 2.48033002e-01 -4.42254543e-01
4.94737983e-01 4.97057408e-01 1.46436781e-01 -8.26778352e-01
-1.08812213e+00 2.67546773e-01 2.60219425e-01 -1.14228100e-01
-5.76572478e-01 -6.96624815e-01 -1.58315444e+00 6.78089201e-01
-3.26854795e-01 5.54363430e-01 6.90885365e-01 9.40830946e-01
4.03292567e-01 1.05486715e+00 7.55014420e-01 -7.34784067e-01
-5.07042825e-01 -1.19379675e+00 -6.15320623e-01 3.72012317e-01
3.11981410e-01 -4.19183224e-01 -3.34879726e-01 7.98464358e-01] | [15.219409942626953, 5.16130256652832] |
a467b341-9442-4395-a9de-bb3d22d925ee | tridonet-a-triple-domain-model-driven-network | 2211.07190 | null | https://arxiv.org/abs/2211.07190v1 | https://arxiv.org/pdf/2211.07190v1.pdf | TriDoNet: A Triple Domain Model-driven Network for CT Metal Artifact Reduction | Recent deep learning-based methods have achieved promising performance for computed tomography metal artifact reduction (CTMAR). However, most of them suffer from two limitations: (i) the domain knowledge is not fully embedded into the network training; (ii) metal artifacts lack effective representation models. The aforementioned limitations leave room for further performance improvement. Against these issues, we propose a novel triple domain model-driven CTMAR network, termed as TriDoNet, whose network training exploits triple domain knowledge, i.e., the knowledge of the sinogram, CT image, and metal artifact domains. Specifically, to explore the non-local repetitive streaking patterns of metal artifacts, we encode them as an explicit tight frame sparse representation model with adaptive thresholds. Furthermore, we design a contrastive regularization (CR) built upon contrastive learning to exploit clean CT images and metal-affected images as positive and negative samples, respectively. Experimental results show that our TriDoNet can generate superior artifact-reduced CT images. | ['Yanwei Qin', 'Qiusheng Lian', 'Shaolei Zhang', 'Ke Jiang', 'Baoshun Shi'] | 2022-11-14 | null | null | null | null | ['metal-artifact-reduction'] | ['medical'] | [ 3.69504213e-01 -2.04978332e-01 -5.68182170e-02 -2.25867584e-01
-9.90033984e-01 1.35023475e-01 2.18613461e-01 -3.24453920e-01
-1.28210217e-01 7.33974457e-01 5.53377509e-01 -4.04996499e-02
-2.86025405e-01 -6.83461666e-01 -6.62327886e-01 -8.40652943e-01
9.38251913e-02 9.80634764e-02 4.23683286e-01 5.84539063e-02
4.55950275e-02 4.61054713e-01 -8.22098017e-01 4.39232528e-01
1.13615549e+00 1.04409504e+00 3.05390567e-01 3.61455902e-02
1.36667892e-01 1.34846950e+00 -4.10919219e-01 1.76674187e-01
2.66017616e-01 -7.31478989e-01 -7.16166019e-01 2.28576124e-01
5.36004901e-02 -6.03453279e-01 -7.69436955e-01 8.67467344e-01
7.28210688e-01 -9.04458051e-04 5.94462693e-01 -7.05007076e-01
-5.86519659e-01 5.52392662e-01 -9.04446185e-01 4.47599798e-01
4.24470790e-02 3.34430307e-01 2.92374492e-01 -1.09052682e+00
6.22994304e-01 7.20399618e-01 7.67505944e-01 4.82847750e-01
-1.02449572e+00 -7.88465679e-01 -2.58389749e-02 2.81480491e-01
-1.24742556e+00 -1.52660936e-01 1.22319555e+00 -4.64354366e-01
5.46048641e-01 2.07389578e-01 7.01659560e-01 1.05001819e+00
3.04148555e-01 8.01143050e-01 1.24054492e+00 -3.65492344e-01
2.43432701e-01 -1.32652059e-01 -6.62974492e-02 7.91607022e-01
2.61586905e-01 3.25475216e-01 -4.00874317e-01 -2.96379894e-01
1.39687788e+00 1.58690244e-01 -7.51517057e-01 -4.65138435e-01
-9.76286471e-01 6.83145106e-01 5.74207962e-01 5.44857323e-01
-5.98795354e-01 1.31223142e-01 5.39839447e-01 -1.32165149e-01
4.33672100e-01 2.73219317e-01 -1.11442834e-01 2.38658920e-01
-9.07948077e-01 -1.21485591e-01 1.75271735e-01 7.53057837e-01
6.24756873e-01 5.50390720e-01 -2.26483554e-01 9.82977509e-01
2.90617466e-01 3.60610932e-01 9.74901319e-01 -6.06002986e-01
3.85511726e-01 5.57633579e-01 -1.23678058e-01 -8.89599979e-01
-4.64671880e-01 -5.97330928e-01 -1.11540592e+00 1.73931688e-01
1.29822046e-01 -1.80655256e-01 -1.26402652e+00 1.42044687e+00
2.71610320e-01 7.17004716e-01 -2.61552960e-01 1.42312491e+00
9.83841300e-01 3.30969691e-01 8.84073526e-02 -4.33367729e-01
1.13832080e+00 -9.08651948e-01 -9.49843347e-01 -2.53610730e-01
6.27334058e-01 -6.90577269e-01 1.12877512e+00 4.54672962e-01
-1.33503175e+00 -2.66810775e-01 -1.31077886e+00 1.55459959e-02
3.47588718e-01 1.14714436e-01 6.95317686e-01 4.67672467e-01
-6.02946579e-01 6.16365135e-01 -1.27827990e+00 3.16346943e-01
7.01588929e-01 2.76180357e-01 5.02062775e-02 -3.94816905e-01
-1.14993119e+00 7.62713552e-01 -1.35204300e-01 1.63847446e-01
-1.12046957e+00 -9.34932709e-01 -8.96980405e-01 -2.02272058e-01
4.46957916e-01 -7.08882034e-01 1.14400101e+00 -9.30744529e-01
-1.45366359e+00 5.57995915e-01 1.85204279e-02 -3.19953799e-01
6.54581845e-01 -1.42729491e-01 -5.64653993e-01 4.78919744e-01
9.47091430e-02 -3.76025401e-02 9.74259555e-01 -1.38588214e+00
-7.96623006e-02 -2.92072117e-01 -3.89101565e-01 3.48489955e-02
-3.39272797e-01 5.02860127e-03 -4.06829447e-01 -1.24290538e+00
5.11150479e-01 -8.01424682e-01 -5.59742868e-01 3.30731213e-01
-3.05251360e-01 2.13419884e-01 6.49539173e-01 -6.74821556e-01
1.36229539e+00 -2.12845922e+00 -1.46907851e-01 3.59272450e-01
4.87097561e-01 2.89084435e-01 -4.29701665e-03 -1.17553040e-01
-4.12645251e-01 -1.03917897e-01 -5.62050641e-01 7.06579164e-02
-5.17545581e-01 2.21494511e-01 -8.99466425e-02 6.73664927e-01
1.66997999e-01 7.85896063e-01 -9.87026691e-01 -7.00721622e-01
4.68529612e-01 4.57786858e-01 -5.79167008e-01 1.62428781e-01
8.59395936e-02 7.77396560e-01 -9.84311819e-01 3.75957191e-01
1.10656333e+00 -6.48626745e-01 1.98025718e-01 -5.91060698e-01
-1.39627457e-01 3.46443057e-01 -9.13304269e-01 1.94993055e+00
-5.34153640e-01 2.30580851e-01 -1.86192598e-02 -8.39691341e-01
6.15945935e-01 5.35034359e-01 1.06193447e+00 -9.97455716e-01
1.30506039e-01 4.82087970e-01 -2.70115603e-02 -7.73004353e-01
-4.53003589e-03 -5.20349860e-01 5.09917557e-01 4.48008299e-01
-2.35675648e-01 -1.64655387e-01 -3.50733966e-01 2.12393701e-02
1.30009949e+00 -5.23951985e-02 -3.75877731e-02 -2.61069030e-01
4.27606702e-01 -3.92423570e-03 8.13202977e-01 6.42773688e-01
-3.34655344e-01 1.16275704e+00 9.46410075e-02 -6.19957209e-01
-8.91098499e-01 -1.10529637e+00 -3.74464363e-01 3.85726184e-01
3.57766062e-01 -1.16619833e-01 -4.90133524e-01 -6.78445935e-01
-4.79160607e-01 2.53933460e-01 -6.02578759e-01 -4.17023540e-01
-1.18466735e+00 -1.10404718e+00 2.71645635e-01 8.78525794e-01
6.37857437e-01 -9.27302599e-01 -5.99846363e-01 3.11867982e-01
-4.64566320e-01 -1.05803132e+00 -5.46563625e-01 1.82315990e-01
-1.22102928e+00 -1.06653214e+00 -1.05936170e+00 -8.35193455e-01
7.33516455e-01 4.14707094e-01 1.07881844e+00 2.52543628e-01
-5.11251032e-01 2.22484753e-01 -3.54935557e-01 -6.19171672e-02
-2.09949747e-01 -5.65693557e-01 -2.44419664e-01 -6.52168095e-02
-7.20310882e-02 -7.45076478e-01 -1.07621026e+00 3.25809240e-01
-1.02426112e+00 2.45426714e-01 8.01522613e-01 1.08583724e+00
8.54252160e-01 1.01570800e-01 4.49674606e-01 -9.94514108e-01
4.14672196e-01 -4.05349344e-01 -1.52508944e-01 1.62801176e-01
-5.17442644e-01 -3.63589860e-02 5.79682052e-01 -4.11463380e-01
-1.38758647e+00 7.96402618e-03 -1.34589061e-01 -7.27183938e-01
1.54417053e-01 5.70883691e-01 1.28520042e-01 -3.88109893e-01
8.95110309e-01 5.06937265e-01 -1.79405019e-01 -3.59880060e-01
-2.15948135e-01 2.70867586e-01 4.10272896e-01 -6.59970105e-01
6.69416606e-01 8.28513086e-01 1.04981493e-02 -6.42763197e-01
-7.92217970e-01 -3.42520177e-01 -2.49279484e-01 -8.97312090e-02
7.91298628e-01 -9.61558163e-01 -1.66572645e-01 5.56805372e-01
-1.03435576e+00 -3.03884029e-01 -3.27420145e-01 7.54879236e-01
-5.33396006e-01 8.49664927e-01 -9.25770044e-01 -4.67517197e-01
-5.38116932e-01 -1.32880521e+00 7.72933781e-01 -1.37061812e-02
4.57592495e-02 -7.58394182e-01 -4.74008545e-02 2.80725151e-01
5.06960154e-01 4.47661787e-01 1.09572148e+00 -1.48676500e-01
-7.54530013e-01 4.58111465e-02 -4.56934035e-01 4.05176878e-01
4.09730375e-01 -5.24304152e-01 -7.30476618e-01 -1.92212164e-01
7.12147593e-01 -3.03756565e-01 6.61077857e-01 7.50138819e-01
1.74835944e+00 -7.57269412e-02 -2.69677579e-01 8.29714477e-01
1.34253991e+00 2.64011681e-01 8.79314065e-01 1.95829093e-01
7.58540809e-01 3.75700295e-02 3.66849989e-01 7.75542617e-01
8.65830667e-03 5.83162725e-01 3.99239928e-01 -5.43627501e-01
-5.91460347e-01 -6.97026923e-02 -2.58704990e-01 9.30206895e-01
-2.90532023e-01 1.02223717e-01 -9.00604188e-01 5.63959599e-01
-1.66576993e+00 -4.86057788e-01 -3.16893399e-01 1.97124982e+00
7.29970455e-01 -1.41569867e-03 -3.10196787e-01 8.50202963e-02
5.95554471e-01 1.38594687e-01 -7.55222142e-01 4.03288662e-01
-2.85219811e-02 5.66636384e-01 3.60763848e-01 -1.13798436e-02
-8.53667676e-01 4.49783921e-01 6.19970322e+00 9.47896183e-01
-1.36648893e+00 4.56095487e-01 5.46553314e-01 3.49514559e-02
-5.28920889e-01 -2.57839173e-01 2.70995311e-02 6.77940845e-01
4.00056481e-01 6.46664016e-03 2.17329875e-01 6.01781130e-01
5.41955948e-01 -3.02507840e-02 -7.02373922e-01 1.13075173e+00
-2.85521075e-02 -1.39809883e+00 1.76569447e-02 -8.99293721e-02
7.55921543e-01 4.64791898e-03 8.45474303e-02 1.24303691e-01
1.76822469e-01 -8.42995107e-01 5.08109748e-01 3.63710254e-01
8.60653222e-01 -5.07434130e-01 5.73923528e-01 -5.75381145e-03
-9.37728584e-01 1.47734642e-01 -2.68481165e-01 4.63631898e-01
2.25409567e-01 8.65416169e-01 -5.60317338e-01 8.69430304e-01
6.37820899e-01 8.67009103e-01 -2.47775152e-01 1.17523193e+00
-2.10587665e-01 7.84066558e-01 -1.90631524e-01 5.89457750e-01
1.31321684e-01 -9.50439349e-02 4.35478866e-01 9.86823678e-01
2.89889127e-01 4.90454882e-01 1.89769581e-01 1.04671311e+00
5.54702783e-05 -1.55300125e-01 -2.17907295e-01 4.10998553e-01
2.70728558e-01 9.40023541e-01 -5.89325905e-01 -3.16369534e-01
-6.87253714e-01 8.60720634e-01 1.63836461e-02 4.94172126e-01
-9.44424152e-01 4.88143004e-02 2.32573405e-01 4.61888701e-01
5.75005226e-02 -1.17673516e-01 -7.48845935e-01 -1.37558091e+00
1.06914632e-01 -8.13360929e-01 3.99450660e-01 -8.00000548e-01
-1.43669319e+00 6.90509379e-01 -2.00580806e-01 -1.69092250e+00
1.40730426e-01 -3.49395037e-01 -7.34956443e-01 6.31631553e-01
-1.73843718e+00 -1.12956119e+00 -6.14009857e-01 9.01403785e-01
4.46182787e-01 1.29802927e-01 4.66740608e-01 7.90114880e-01
-4.67737824e-01 5.30923665e-01 -5.09584136e-02 3.37750077e-01
5.81357718e-01 -8.73472989e-01 -1.81018993e-01 7.55026579e-01
-3.98757488e-01 4.53877509e-01 3.77871841e-01 -7.53925085e-01
-1.36928785e+00 -1.17676020e+00 1.25879124e-01 -4.57962751e-02
5.75978518e-01 7.25347549e-02 -1.12020695e+00 5.79815447e-01
-4.39827479e-02 5.92194438e-01 5.67488968e-01 -2.79115349e-01
-9.66672972e-02 1.83109418e-02 -1.23085618e+00 4.87084687e-01
9.68924880e-01 -3.47265810e-01 -4.90647525e-01 4.83752966e-01
5.60204208e-01 -7.85906136e-01 -9.08056736e-01 7.62855887e-01
2.24184126e-01 -9.67239857e-01 1.04541945e+00 -2.34324396e-01
8.43040228e-01 -1.79298133e-01 5.54881208e-02 -1.13749158e+00
-4.93147731e-01 -2.87606001e-01 -1.79561824e-01 6.73738062e-01
4.88045299e-03 -4.67450351e-01 1.01528943e+00 6.62198186e-01
-8.28186333e-01 -8.88145924e-01 -1.13618004e+00 -7.00459480e-01
2.10961491e-01 -4.36690897e-01 5.11334121e-01 1.23332071e+00
-1.94007799e-01 -8.17895383e-02 -5.10035336e-01 2.75500119e-01
5.68941891e-01 1.31119519e-01 1.51526257e-01 -7.53521204e-01
-3.94194424e-01 -1.91329688e-01 -8.86745527e-02 -9.85911965e-01
-2.40795687e-01 -7.45237112e-01 1.42224608e-02 -1.40440059e+00
4.29230869e-01 -8.01015079e-01 -6.03068173e-01 4.41029400e-01
-2.09897906e-01 2.34815642e-01 -2.02589050e-01 3.78397882e-01
-3.45484048e-01 8.29660952e-01 1.96783578e+00 -1.42724067e-01
-2.04697382e-02 -2.04063669e-01 -5.20754695e-01 9.40485239e-01
4.80493486e-01 -5.30303299e-01 -3.86000723e-01 -7.35457480e-01
-4.31183875e-02 4.95317131e-01 5.41097283e-01 -1.26730776e+00
9.48174894e-02 -2.95598567e-01 5.85047483e-01 -5.10558844e-01
2.60860533e-01 -9.17499125e-01 7.77593851e-02 6.13479078e-01
-8.65744427e-02 -3.81643265e-01 1.79234505e-01 5.43832719e-01
-3.48036319e-01 -1.63833052e-01 1.13943946e+00 -4.74130392e-01
-4.14208442e-01 5.10564148e-01 -4.37850863e-01 9.67812538e-02
6.00136220e-01 -2.56048709e-01 -9.24906954e-02 -1.08991876e-01
-6.21511459e-01 3.11585255e-02 2.43325248e-01 6.91774487e-02
9.66463327e-01 -1.43259156e+00 -6.22172177e-01 3.08118075e-01
1.22849457e-02 2.97120392e-01 8.52843344e-01 1.10346627e+00
-5.22536337e-01 1.17937252e-01 -6.10013083e-02 -6.89385116e-01
-7.02558219e-01 5.39592862e-01 5.52067161e-01 -5.67165852e-01
-1.10062397e+00 6.80761993e-01 6.48380339e-01 -1.34188160e-01
-5.25656156e-02 -4.30949003e-01 2.01942027e-01 -6.41880929e-01
4.40403879e-01 2.19881088e-01 3.99339467e-01 -2.17371494e-01
-4.02895957e-01 6.47733748e-01 -3.33452672e-01 2.66467899e-01
1.45251584e+00 1.04368934e-02 7.09020793e-02 -1.95657372e-01
1.02276659e+00 -1.25713512e-01 -1.30987215e+00 -4.73591208e-01
-8.86991993e-02 -7.00301945e-01 3.51156056e-01 -9.36404407e-01
-1.51275730e+00 7.38311410e-01 9.08437848e-01 -4.71525043e-01
1.41777897e+00 -2.19045997e-01 1.22119570e+00 -1.98222011e-01
4.12717760e-01 -9.29581583e-01 5.54069161e-01 2.14649066e-01
9.21752036e-01 -1.25117004e+00 3.60904098e-01 -7.48952210e-01
-5.63809574e-01 9.23251331e-01 7.57527530e-01 -3.83751869e-01
6.78001165e-01 3.90009552e-01 8.17869976e-02 -4.77413684e-01
-3.69681507e-01 1.40844434e-01 1.94605201e-01 6.19376481e-01
5.15701115e-01 -8.52579325e-02 -4.84193325e-01 7.93017805e-01
3.50008458e-01 4.49669540e-01 5.39746702e-01 1.11393309e+00
-2.54542798e-01 -7.73463011e-01 -4.34750348e-01 5.64405739e-01
-5.30884445e-01 -1.77013397e-01 2.48739608e-02 6.05138540e-01
1.15421429e-01 7.44626164e-01 -3.64473701e-01 -1.85378805e-01
4.19172853e-01 -4.84251559e-01 6.39212072e-01 -6.47341013e-01
-5.52130044e-01 3.86457324e-01 -2.48971850e-01 -5.54671288e-01
-3.33155930e-01 -2.97462642e-01 -1.43008685e+00 4.64373529e-02
-4.32331324e-01 4.89850668e-03 3.73175472e-01 9.62099731e-01
2.03732550e-01 9.57049906e-01 7.56627619e-01 -7.23521054e-01
-4.85704750e-01 -8.75212967e-01 -6.90783501e-01 5.90212166e-01
4.17328566e-01 -7.57119119e-01 -2.43299320e-01 -6.83424622e-02] | [13.526820182800293, -2.5396037101745605] |
f7122521-3bd4-4144-960c-ce52869bb4d7 | improving-accuracy-and-speeding-up-document | 2006.09141 | null | https://arxiv.org/abs/2006.09141v1 | https://arxiv.org/pdf/2006.09141v1.pdf | Improving accuracy and speeding up Document Image Classification through parallel systems | This paper presents a study showing the benefits of the EfficientNet models compared with heavier Convolutional Neural Networks (CNNs) in the Document Classification task, essential problem in the digitalization process of institutions. We show in the RVL-CDIP dataset that we can improve previous results with a much lighter model and present its transfer learning capabilities on a smaller in-domain dataset such as Tobacco3482. Moreover, we present an ensemble pipeline which is able to boost solely image input by combining image model predictions with the ones generated by BERT model on extracted text by OCR. We also show that the batch size can be effectively increased without hindering its accuracy so that the training process can be sped up by parallelizing throughout multiple GPUs, decreasing the computational time needed. Lastly, we expose the training performance differences between PyTorch and Tensorflow Deep Learning frameworks. | ['Mateo Valero', 'Jordi Cortada', 'Daniel Garrido', 'Juan Luis Dominguez', 'Jordi Torres', 'Javier Ferrando', 'Raul Garcia', 'David Garcia'] | 2020-06-16 | null | null | null | null | ['document-image-classification'] | ['computer-vision'] | [ 4.40823659e-02 6.68678358e-02 9.57142711e-02 -2.57326663e-01
-9.51948240e-02 -6.60823107e-01 8.15847456e-01 4.81320322e-02
-7.42229760e-01 4.43926305e-01 4.17916551e-02 -7.31756628e-01
-1.24192499e-01 -9.09910858e-01 -7.39709496e-01 -3.58043343e-01
2.54722506e-01 4.23815310e-01 -1.11944780e-01 -1.31748840e-01
4.07258272e-01 8.14827740e-01 -1.30183589e+00 6.91924989e-01
3.70033711e-01 9.60580647e-01 1.23346858e-01 1.20160496e+00
-2.75798321e-01 1.00747979e+00 -6.80351436e-01 -8.29679191e-01
4.63985145e-01 2.33526126e-01 -9.73526955e-01 -3.97969224e-02
8.78686845e-01 -6.44492507e-01 -4.00315106e-01 8.18270206e-01
4.96642351e-01 -1.08489417e-01 4.46315736e-01 -8.84206772e-01
-6.16380811e-01 7.55454600e-01 -1.62445441e-01 1.79322258e-01
-2.33906895e-01 1.86494499e-01 9.29438770e-01 -8.35960329e-01
8.73186648e-01 1.18762815e+00 8.12323570e-01 4.06026691e-01
-1.05547011e+00 -7.55529642e-01 -3.10961045e-02 2.72197068e-01
-8.77758324e-01 -1.74374238e-01 3.76329720e-01 -4.76203382e-01
1.20445359e+00 2.77445257e-01 5.12699842e-01 1.25875139e+00
-3.11144777e-02 8.19341004e-01 8.18004370e-01 -4.64380980e-01
-5.49818240e-02 -4.28076275e-02 4.57743675e-01 8.74784291e-01
3.70714486e-01 -2.36176327e-01 -1.74661949e-01 2.38424152e-01
8.03137064e-01 1.24955691e-01 -1.92010514e-02 2.27055341e-01
-1.13192248e+00 6.82660043e-01 5.72394609e-01 8.02919388e-01
-2.39776522e-01 4.86349523e-01 6.62315130e-01 2.25413799e-01
5.89871705e-01 6.21925414e-01 -5.62118173e-01 -2.37918019e-01
-1.43470752e+00 2.89150447e-01 9.98467624e-01 9.06860709e-01
7.41735637e-01 2.11028233e-02 -2.33068109e-01 5.17427742e-01
-3.76861989e-02 2.50306487e-01 4.49744165e-01 -9.64462459e-01
5.53052902e-01 8.18899035e-01 -2.31456190e-01 -8.19829464e-01
-4.56808865e-01 -7.23477662e-01 -9.79273677e-01 3.34915698e-01
7.41772771e-01 -2.69648045e-01 -1.09853971e+00 8.73115182e-01
-2.92665184e-01 1.36444673e-01 -1.81507349e-01 5.33821166e-01
8.50920081e-01 5.35338819e-01 1.67748198e-01 6.57451570e-01
1.36152375e+00 -1.45866907e+00 -5.97647369e-01 -3.56674157e-02
1.17809606e+00 -7.38948584e-01 8.46202254e-01 6.59985185e-01
-9.78344619e-01 -7.09048450e-01 -1.15997529e+00 -5.52925289e-01
-8.52668166e-01 5.31044841e-01 9.46422696e-01 9.32218730e-01
-1.29938626e+00 1.19033206e+00 -8.12113702e-01 -4.56991702e-01
7.91177332e-01 5.05311131e-01 -4.96584058e-01 2.93222740e-02
-7.08398283e-01 7.15351403e-01 6.15854323e-01 1.56147396e-02
-7.54613221e-01 -9.61957216e-01 -3.12213689e-01 6.42744660e-01
-1.37567818e-01 -4.80531186e-01 1.29386699e+00 -1.17684424e+00
-1.50241566e+00 9.38583016e-01 1.60249427e-01 -7.91766465e-01
9.38262939e-01 -2.00048909e-01 -1.39133140e-01 1.59786105e-01
-5.23064375e-01 8.28790843e-01 6.22648060e-01 -8.07098925e-01
-4.83877987e-01 -2.68436551e-01 1.83197990e-01 -2.37652749e-01
-9.49198842e-01 5.37924431e-02 -6.12707734e-01 -3.92783344e-01
-5.58700502e-01 -1.02679050e+00 -2.47744337e-01 5.98150305e-03
-4.74277198e-01 -9.55255106e-02 9.20520008e-01 -1.00881732e+00
8.58504891e-01 -2.08284044e+00 -8.59726071e-02 2.65233874e-01
2.64935225e-01 6.64662957e-01 -2.80037701e-01 2.21767932e-01
-2.78875172e-01 5.57993114e-01 1.13394253e-01 -7.07490563e-01
-3.03868540e-02 1.49446771e-01 -1.92838296e-01 2.70763606e-01
1.91355675e-01 1.07434893e+00 -4.45701212e-01 -3.41083556e-01
2.09045455e-01 5.69577157e-01 -5.92510939e-01 -2.00108215e-02
-1.64339408e-01 1.01785243e-01 -1.69992268e-01 4.89557028e-01
7.91075706e-01 -4.77716327e-01 1.68436319e-01 -2.12189760e-02
-2.07053572e-01 1.82628632e-03 -7.91039109e-01 1.89111745e+00
-6.21414363e-01 1.03537786e+00 1.44398585e-01 -1.08760774e+00
9.18133080e-01 2.34695867e-01 2.02199668e-01 -5.82971871e-01
3.28584313e-01 1.37501687e-01 4.15715203e-02 -2.87158310e-01
7.36851394e-01 4.05411541e-01 4.19209182e-01 4.74065632e-01
4.24365222e-01 3.34303901e-02 4.75610942e-01 6.68942109e-02
1.07403314e+00 2.05169946e-01 -3.33846122e-01 -3.62020075e-01
5.85471988e-01 1.34697720e-01 -1.63728938e-01 9.51068461e-01
1.04228050e-01 6.00725293e-01 7.47517407e-01 -8.03230882e-01
-1.59731877e+00 -3.69405508e-01 -1.53359383e-01 1.25574517e+00
-6.75587416e-01 -4.55315024e-01 -1.07224238e+00 -9.07209754e-01
-1.04275733e-01 5.77819049e-01 -7.62375116e-01 3.14192981e-01
-6.76781058e-01 -9.30154204e-01 9.45069909e-01 6.84370041e-01
7.47453332e-01 -9.44478869e-01 -2.84581721e-01 4.08129357e-02
3.28515321e-01 -1.30927145e+00 2.72064418e-01 3.36829394e-01
-1.06001067e+00 -8.90655160e-01 -7.54952967e-01 -6.84859872e-01
5.81727862e-01 -8.20564106e-02 1.03596115e+00 5.59644341e-01
-4.44001198e-01 1.20448537e-01 -2.32255876e-01 -6.28034949e-01
-5.05767286e-01 6.74874902e-01 -5.51591992e-01 -2.45078191e-01
3.72222990e-01 -3.13236058e-01 -6.42803431e-01 -3.70274693e-01
-1.08324182e+00 3.91323119e-01 3.73157978e-01 6.92135334e-01
-9.18937102e-02 -9.14126188e-02 1.60594583e-02 -1.15700483e+00
4.91724640e-01 -2.75040537e-01 -6.82364821e-01 7.97801539e-02
-7.71872163e-01 8.29696730e-02 7.61688590e-01 -1.92460209e-01
-1.03135669e+00 -2.02650996e-03 -2.56546140e-01 -3.21778387e-01
-2.41342351e-01 2.56654263e-01 3.95032763e-01 -1.09653503e-01
5.92214406e-01 1.52845532e-02 3.06230225e-03 -8.02541077e-01
4.18376446e-01 7.01168299e-01 4.57910120e-01 -5.08616567e-01
6.05076730e-01 6.23041451e-01 1.94085985e-01 -6.23306334e-01
-6.34979784e-01 -3.50770533e-01 -7.93981016e-01 -1.27457961e-01
9.31719959e-01 -8.94495666e-01 -9.95854676e-01 5.49511790e-01
-1.44571698e+00 -4.52188432e-01 -9.14225727e-02 3.25373590e-01
-2.22490162e-01 3.10962409e-01 -1.11630511e+00 -4.38453406e-01
-6.73672140e-01 -1.11773336e+00 1.02366376e+00 2.62122482e-01
5.99848218e-02 -1.23616815e+00 -2.63680488e-01 4.09069449e-01
8.20412576e-01 1.00981500e-02 1.00458813e+00 -1.14548337e+00
-5.07905245e-01 -5.34396112e-01 -6.87502027e-01 7.53729463e-01
-4.43891555e-01 4.60451066e-01 -1.37184918e+00 -3.22487831e-01
-4.06100661e-01 -8.98156464e-02 1.18348432e+00 1.57702059e-01
1.79188395e+00 -1.10133149e-01 -1.77032560e-01 7.91714728e-01
1.50521708e+00 -2.11247191e-01 8.65665376e-01 8.22348416e-01
8.62315238e-01 4.84022349e-01 -1.01967767e-01 3.40637654e-01
-1.60681471e-01 2.67234981e-01 6.57997072e-01 -5.82571387e-01
-3.90295148e-01 3.71568510e-03 1.12519562e-01 5.80912292e-01
-5.38346410e-01 -4.29820538e-01 -1.11527228e+00 3.98124784e-01
-1.79364169e+00 -7.68964827e-01 -6.21422529e-01 1.59843969e+00
4.19682473e-01 1.23815887e-01 -1.38734594e-01 1.55965820e-01
6.70400321e-01 -6.84185401e-02 -1.08862743e-01 -7.56132960e-01
-1.19101241e-01 7.16529071e-01 8.38925302e-01 3.01154912e-01
-1.13913345e+00 9.61035132e-01 6.75333595e+00 7.85889030e-01
-1.31484783e+00 1.01057731e-01 9.94286716e-01 -2.13421673e-01
2.04554021e-01 -1.00595362e-01 -9.67428446e-01 3.18987668e-01
1.47728610e+00 2.30663955e-01 3.14383745e-01 1.10609055e+00
2.62082797e-02 2.82305837e-01 -1.16609251e+00 6.63816094e-01
-1.08345106e-01 -1.87074733e+00 9.92920101e-02 2.95364827e-01
6.94105804e-01 4.10677671e-01 4.91670258e-02 4.24362838e-01
4.51470733e-01 -1.17212677e+00 5.08210063e-01 4.74029541e-01
5.91761291e-01 -6.87374949e-01 1.01380742e+00 1.10203899e-01
-4.92051661e-01 -2.74536252e-01 -5.73132575e-01 -1.76603824e-01
-5.44194400e-01 6.98794544e-01 -1.27264941e+00 6.56738698e-01
8.20844352e-01 7.96067953e-01 -1.15420282e+00 9.26078022e-01
-1.02822952e-01 7.33582020e-01 -1.21694282e-01 -1.84858009e-01
5.17800152e-01 -2.05753937e-01 1.52800083e-01 1.90972304e+00
5.76426923e-01 -7.10694551e-01 -3.76495093e-01 9.52493012e-01
-5.82365751e-01 2.42683932e-01 -3.82923245e-01 -1.24554358e-01
-1.20474555e-01 1.84497058e+00 -7.41170764e-01 -7.14361310e-01
-3.65402460e-01 9.53418076e-01 3.09128016e-01 2.11537614e-01
-6.67441010e-01 -5.05920112e-01 3.86402428e-01 1.42550766e-01
5.33255458e-01 -1.91508308e-01 -5.71136355e-01 -1.16284883e+00
-3.22988361e-01 -6.41210735e-01 5.18533625e-02 -9.06390846e-01
-9.49682832e-01 5.05955040e-01 -5.57519913e-01 -7.49251604e-01
-2.10983735e-02 -1.49922061e+00 -5.37882328e-01 8.82435441e-01
-1.73930442e+00 -1.34095585e+00 -4.93694603e-01 4.45115715e-01
3.75506163e-01 -1.64027348e-01 9.27984416e-01 5.04692852e-01
-8.09345722e-01 4.57752943e-01 4.38179672e-01 6.14428580e-01
6.81653440e-01 -1.44179893e+00 5.82928956e-01 6.45935714e-01
-5.62791340e-02 4.92378622e-01 3.57428133e-01 -1.65799499e-01
-1.06444836e+00 -1.19089150e+00 9.71874893e-01 -2.48720527e-01
8.76924098e-01 -5.86829364e-01 -7.57592916e-01 8.69033635e-01
5.94470739e-01 8.78299400e-02 5.23782074e-01 1.92240790e-01
-5.76544523e-01 9.50611383e-02 -1.00894666e+00 5.53666890e-01
7.74792731e-01 -4.31679964e-01 -7.76521787e-02 5.93314409e-01
5.50733566e-01 -3.04155558e-01 -1.08309782e+00 3.14039215e-02
5.91633379e-01 -1.01641071e+00 7.71213055e-01 -9.23709750e-01
1.06756771e+00 2.29319170e-01 7.91551247e-02 -9.26053643e-01
-3.98494840e-01 -3.54509473e-01 -4.61045280e-02 1.23335433e+00
4.34428781e-01 -3.12473744e-01 9.99518812e-01 6.23922586e-01
-4.66700904e-02 -3.20511580e-01 -7.29381144e-01 -6.64061606e-01
4.31441605e-01 -5.67365289e-01 6.98161244e-01 8.74073446e-01
-3.00658643e-01 1.21570408e-01 -9.81890559e-02 -2.65408128e-01
2.06668869e-01 -1.85198978e-01 9.22156334e-01 -1.50769353e+00
-2.37402722e-01 -6.30745351e-01 -3.95107269e-01 -5.99350393e-01
2.41236746e-01 -1.21003771e+00 -5.28515279e-01 -1.29401422e+00
3.19280207e-01 -1.68324709e-01 -3.75959992e-01 7.91534424e-01
1.26090765e-01 4.27818567e-01 5.83332419e-01 2.00092837e-01
-3.08940589e-01 -1.60422668e-01 1.34029627e+00 -2.05846846e-01
3.68821383e-01 -4.06992376e-01 -5.12860000e-01 7.35711277e-01
6.88269258e-01 -3.48600000e-01 2.30145723e-01 -6.53294563e-01
4.71044868e-01 -3.59371811e-01 5.54930985e-01 -1.21981764e+00
1.23957120e-01 5.52775741e-01 9.50045347e-01 -3.66595924e-01
4.56763841e-02 -8.91001642e-01 -1.92484707e-01 7.71748662e-01
-5.29308319e-01 2.37907600e-02 5.24943411e-01 1.94730118e-01
9.01379436e-02 -6.85188115e-01 5.98896265e-01 -1.96224689e-01
-3.19694489e-01 1.49499208e-01 -4.14936036e-01 -7.22613513e-01
6.25371993e-01 4.71392013e-02 -7.09523082e-01 -8.17839056e-02
-8.15690637e-01 -1.63418218e-01 3.92895013e-01 4.01661098e-01
-9.22700837e-02 -9.86018658e-01 -8.37825119e-01 6.63391501e-02
-2.97773868e-01 -1.32274359e-01 1.72719374e-01 7.32408047e-01
-1.19769466e+00 7.82429397e-01 -4.14035827e-01 -5.15699804e-01
-1.33383811e+00 4.80219275e-01 4.56143886e-01 -9.73102450e-01
-5.16550839e-01 7.15368986e-01 -1.92838296e-01 -5.04743576e-01
2.32386664e-01 -4.73971456e-01 -3.50887030e-01 -7.24101514e-02
5.90794027e-01 6.01180673e-01 7.28053629e-01 1.00153564e-02
5.48107661e-02 1.36101440e-01 -4.84630108e-01 1.69727013e-01
1.86310482e+00 3.69723648e-01 -3.74091238e-01 -8.45752880e-02
1.36140895e+00 -1.59130231e-01 -1.21143997e+00 4.31010053e-02
1.50013700e-01 -3.30843963e-02 4.72953349e-01 -9.10341084e-01
-1.19662189e+00 1.17729425e+00 8.21030557e-01 3.16291511e-01
1.03888726e+00 -5.24717391e-01 5.79546273e-01 7.07493305e-01
-3.09097081e-01 -1.25042331e+00 1.27676986e-02 7.41148174e-01
5.84238887e-01 -1.03212464e+00 1.45125657e-01 -2.30724335e-01
-1.54958308e-01 1.85355496e+00 4.78402764e-01 -4.25664276e-01
4.97754335e-01 4.84333664e-01 -2.87161201e-01 -2.31310129e-01
-6.77631617e-01 1.01015093e-02 7.76397586e-02 2.64776558e-01
7.96318293e-01 -9.40762460e-02 -1.52363524e-01 4.13829595e-01
-4.17477846e-01 2.73820907e-01 7.29419291e-01 7.47878850e-01
-8.70430171e-02 -1.25128722e+00 -3.36299270e-01 4.72096592e-01
-1.04720926e+00 -4.86810625e-01 -4.67216134e-01 8.80548239e-01
1.13002844e-01 5.35246789e-01 4.54535246e-01 -2.24471062e-01
1.40342578e-01 3.23586226e-01 3.79676998e-01 -4.45994139e-01
-1.24626803e+00 -2.27437586e-01 1.72167242e-01 -4.50276822e-01
-2.61851549e-01 -2.91142553e-01 -9.52119410e-01 -7.09953249e-01
-1.88923538e-01 -2.80421317e-01 1.34319699e+00 9.01330888e-01
5.60693860e-01 8.37526798e-01 2.05006406e-01 -1.07594883e+00
-2.98861831e-01 -1.29928446e+00 -3.98632914e-01 3.61819297e-01
6.35475069e-02 9.96360779e-02 -2.70941496e-01 3.42604578e-01] | [11.446207046508789, 2.6088814735412598] |
9d2f7922-020f-497f-9e8d-8100de4f20d1 | cascaded-continuous-regression-for-real-time | 1608.01137 | null | http://arxiv.org/abs/1608.01137v2 | http://arxiv.org/pdf/1608.01137v2.pdf | Cascaded Continuous Regression for Real-time Incremental Face Tracking | This paper introduces a novel real-time algorithm for facial landmark
tracking. Compared to detection, tracking has both additional challenges and
opportunities. Arguably the most important aspect in this domain is updating a
tracker's models as tracking progresses, also known as incremental (face)
tracking. While this should result in more accurate localisation, how to do
this online and in real time without causing a tracker to drift is still an
important open research question. We address this question in the cascaded
regression framework, the state-of-the-art approach for facial landmark
localisation. Because incremental learning for cascaded regression is costly,
we propose a much more efficient yet equally accurate alternative using
continuous regression. More specifically, we first propose cascaded continuous
regression (CCR) and show its accuracy is equivalent to the Supervised Descent
Method. We then derive the incremental learning updates for CCR (iCCR) and show
that it is an order of magnitude faster than standard incremental learning for
cascaded regression, bringing the time required for the update from seconds
down to a fraction of a second, thus enabling real-time tracking. Finally, we
evaluate iCCR and show the importance of incremental learning in achieving
state-of-the-art performance. Code for our iCCR is available from
http://www.cs.nott.ac.uk/~psxes1 | ['Enrique Sánchez-Lozano', 'Michel Valstar', 'Georgios Tzimiropoulos', 'Brais Martinez'] | 2016-08-03 | null | null | null | null | ['landmark-tracking'] | ['computer-vision'] | [-1.11451685e-01 -2.19015270e-01 -2.54819721e-01 -1.16878174e-01
-1.01355696e+00 -5.43466270e-01 5.76422572e-01 -6.55219033e-02
-4.98323232e-01 6.35318756e-01 -1.76576227e-01 -2.95205057e-01
-1.00698890e-02 -1.98876321e-01 -8.04193258e-01 -6.16782963e-01
-3.29640031e-01 2.86176980e-01 3.28640968e-01 -7.89146870e-02
-2.41804999e-04 9.43014145e-01 -1.42142296e+00 -3.19103599e-01
2.43508115e-01 8.78855824e-01 -1.96807697e-01 1.14188480e+00
1.34364143e-01 5.01666367e-01 -5.67000687e-01 -5.07943511e-01
4.55096990e-01 -3.71288568e-01 -6.17970228e-01 -1.89049507e-03
9.56342816e-01 -2.66568780e-01 -6.35681301e-02 8.53114247e-01
8.32552671e-01 4.72373553e-02 2.86383778e-01 -1.45764375e+00
-6.89126477e-02 6.18619919e-02 -8.98687840e-01 2.29852572e-01
5.03843009e-01 -3.19861799e-01 6.21286571e-01 -9.09082949e-01
6.04357362e-01 1.16211069e+00 1.19113946e+00 8.21769595e-01
-1.15231395e+00 -9.51119542e-01 4.03028756e-01 -1.48863867e-01
-1.49689043e+00 -9.11217928e-01 6.14170074e-01 -2.95858651e-01
3.37256610e-01 2.59328783e-01 6.64142966e-01 7.25946963e-01
-6.44993037e-02 6.73333108e-01 1.06355453e+00 -6.10290647e-01
-9.16883070e-03 1.20636806e-01 -3.62557203e-01 9.42940831e-01
1.76652700e-01 4.09088016e-01 -5.75932980e-01 -2.06164792e-01
1.03986537e+00 -7.65740275e-02 -5.07402830e-02 -4.05832916e-01
-8.36129308e-01 6.20225251e-01 3.54451150e-01 2.03926787e-01
-1.20511204e-01 6.66930079e-01 3.28219891e-01 4.74630296e-01
7.48140752e-01 -1.20042320e-02 -5.42464674e-01 -2.95396596e-01
-1.35443342e+00 1.23506308e-01 7.59678483e-01 8.12002838e-01
5.44469714e-01 7.21823126e-02 2.21225142e-01 5.69695354e-01
4.48070675e-01 4.67081517e-01 1.15314685e-01 -9.73135710e-01
-2.48138495e-02 1.66648418e-01 3.20395470e-01 -8.81668329e-01
-5.98420441e-01 -6.44863665e-01 -5.63038588e-01 6.48651421e-01
8.23758721e-01 -4.92748976e-01 -5.37569702e-01 1.73662174e+00
8.68143857e-01 8.29376698e-01 -3.81892443e-01 5.40637195e-01
5.18176615e-01 3.46504956e-01 -1.46920547e-01 -4.85720783e-01
1.00735199e+00 -9.13808882e-01 -5.40850461e-01 4.93779872e-03
6.26364350e-01 -9.43419933e-01 2.04085991e-01 2.69240707e-01
-9.49427605e-01 -5.35621941e-01 -7.43024707e-01 1.70511216e-01
-2.47654781e-01 2.21931592e-01 7.05665112e-01 8.38701725e-01
-1.52741933e+00 6.05047762e-01 -1.04288304e+00 -7.04400539e-01
4.47504431e-01 8.36511254e-01 -4.75528359e-01 2.73857087e-01
-8.18316400e-01 9.53556538e-01 -2.58017212e-01 2.60048538e-01
-5.46842575e-01 -8.94457459e-01 -6.88392341e-01 -5.09610116e-01
3.68213505e-01 -3.62136036e-01 1.63947189e+00 -9.04314697e-01
-1.67167020e+00 7.82798231e-01 -7.25975752e-01 -5.81537545e-01
9.36605155e-01 -4.29747343e-01 -2.44116098e-01 -6.65046796e-02
-6.78960755e-02 6.82360053e-01 1.18019414e+00 -9.90058482e-01
-7.20254183e-01 -2.72273839e-01 -4.11365274e-03 8.66290480e-02
-2.00656354e-01 2.82314301e-01 -6.55218184e-01 -4.07402992e-01
-1.30208448e-01 -1.12323976e+00 -2.60571241e-01 6.80129528e-01
2.32767686e-01 -3.08370590e-01 9.49483454e-01 -7.26791382e-01
1.13246322e+00 -2.01423645e+00 -2.16545582e-01 -4.02758718e-02
2.95643002e-01 3.05800259e-01 1.95633825e-02 2.85991907e-01
-6.96676895e-02 -2.88646519e-01 9.25529599e-02 -6.86063290e-01
-2.45753556e-01 -2.15657741e-01 -1.58971399e-01 1.02779698e+00
4.85989638e-02 9.69409168e-01 -1.05979276e+00 -4.66110110e-01
2.24162355e-01 8.79175901e-01 -2.08540007e-01 -1.83810055e-01
9.71320719e-02 7.83528745e-01 1.20946218e-03 8.76144648e-01
6.57861412e-01 4.41964390e-03 -7.78101161e-02 -1.19899912e-02
-4.03949827e-01 -2.01087162e-01 -1.37302136e+00 1.38407063e+00
-5.40999532e-01 9.71174777e-01 2.45376810e-01 -8.66929770e-01
9.18912947e-01 4.19477463e-01 7.86333919e-01 -6.06382012e-01
-9.80674773e-02 2.10601211e-01 -3.87178183e-01 1.22066595e-01
4.65614349e-01 -1.28947452e-01 1.46608457e-01 2.81434089e-01
-1.95487626e-02 2.95685440e-01 -3.56738381e-02 9.22443420e-02
9.10232306e-01 6.52391613e-01 4.46267188e-01 2.71325614e-02
4.83409107e-01 -1.48589298e-01 4.71239269e-01 6.70208395e-01
-4.47466522e-01 3.68773878e-01 3.10112804e-01 -3.82014215e-01
-6.65775418e-01 -6.00487947e-01 -1.60047352e-01 1.22733653e+00
-1.62727777e-02 -4.43171531e-01 -5.65384567e-01 -6.95864499e-01
3.22625846e-01 -9.90946591e-03 -7.58833826e-01 1.63865909e-01
-7.12660551e-01 -5.90911508e-01 5.57461202e-01 6.51360095e-01
2.62216955e-01 -5.54759264e-01 -5.28029084e-01 2.56396621e-01
3.96925479e-01 -9.45585907e-01 -6.10934079e-01 5.91497757e-02
-1.05623019e+00 -1.13205588e+00 -1.01684499e+00 -5.59203506e-01
9.11448181e-01 3.42386723e-01 9.08402145e-01 3.10033411e-01
-3.54205817e-01 7.95838654e-01 -1.34304166e-01 -4.60182220e-01
-2.19820157e-01 1.05927363e-02 3.48871768e-01 3.97041962e-02
5.52418455e-02 -2.02288479e-01 -5.48398614e-01 3.27222526e-01
-2.28451401e-01 -4.45938289e-01 5.54820299e-01 7.15863109e-01
5.69853961e-01 -1.75971761e-01 4.62038606e-01 -1.01453912e+00
3.53815049e-01 -2.17979804e-01 -1.20965970e+00 5.67945428e-02
-6.08651519e-01 -1.77891150e-01 3.75458539e-01 -6.10675514e-01
-7.04637647e-01 6.22681916e-01 -3.83006155e-01 -5.59953868e-01
1.68535277e-01 -9.29516333e-04 3.90921086e-01 -9.76248384e-01
5.63514888e-01 8.13650340e-02 2.10407004e-01 -4.28703249e-01
5.37644923e-01 2.97619462e-01 4.75338846e-01 -3.57285321e-01
1.18841827e+00 8.35295796e-01 2.46153086e-01 -1.02305138e+00
-7.50552118e-01 -7.56781995e-01 -1.06635427e+00 -5.38949490e-01
3.35171163e-01 -9.84954476e-01 -8.88723075e-01 3.41151923e-01
-1.00224340e+00 -5.33592045e-01 -2.68534303e-01 2.24125549e-01
-4.00942415e-01 2.68102288e-01 -2.97094882e-01 -1.13443124e+00
-2.18736514e-01 -6.84394121e-01 1.30601192e+00 3.58313888e-01
-2.38929376e-01 -1.30620587e+00 3.53090256e-01 -2.39148006e-01
4.44603235e-01 5.41455984e-01 -1.81235746e-01 -2.68723905e-01
-2.81367093e-01 -5.88773787e-01 -8.64360780e-02 2.48258132e-02
1.63208380e-01 1.96272001e-01 -9.81093109e-01 -8.42365503e-01
-2.02823475e-01 1.04883566e-01 6.86681449e-01 4.85948354e-01
6.41981125e-01 -2.01028273e-01 -6.97988689e-01 6.38056338e-01
1.42495286e+00 -3.17982733e-02 3.16258460e-01 3.37262452e-01
5.98330319e-01 2.86221147e-01 8.69718552e-01 1.27283737e-01
2.89598584e-01 9.68175590e-01 1.62762046e-01 -3.52154106e-01
-4.19648558e-01 -3.07675511e-01 5.94334483e-01 5.51592767e-01
-3.16628337e-01 5.11255085e-01 -7.51723051e-01 4.51929659e-01
-1.85183263e+00 -1.04723275e+00 -1.27161920e-01 2.56897044e+00
7.09406972e-01 4.00712788e-02 6.30998194e-01 1.34954050e-01
7.50590146e-01 -3.79418992e-02 -4.64492232e-01 -2.64772028e-01
3.49311113e-01 3.19976717e-01 7.39830852e-01 8.48945200e-01
-1.25620711e+00 1.14627147e+00 6.65454817e+00 5.53716838e-01
-1.40236008e+00 4.22654629e-01 3.74856979e-01 -2.13028446e-01
5.27898252e-01 7.33239353e-02 -1.46171105e+00 4.48827237e-01
1.13782251e+00 -6.78768056e-03 2.43028745e-01 9.85110164e-01
1.06760606e-01 -2.30792940e-01 -8.94057274e-01 1.15425456e+00
4.41814810e-02 -1.12315011e+00 -5.84867239e-01 2.13707507e-01
7.78242886e-01 -2.43902416e-03 -3.12135909e-02 3.81452650e-01
2.36459017e-01 -8.15425158e-01 5.50136209e-01 4.79590535e-01
9.88656044e-01 -7.02697158e-01 5.49650311e-01 2.55009621e-01
-1.76948380e+00 9.59877968e-02 -4.43724453e-01 3.08865439e-02
6.98036030e-02 5.97682476e-01 -8.64100277e-01 3.27712595e-01
5.23275673e-01 7.98987985e-01 -8.06058764e-01 1.51772773e+00
-1.89990595e-01 4.43420112e-01 -5.75669348e-01 -1.14482939e-01
2.76035257e-02 1.32815078e-01 4.42775518e-01 1.39375329e+00
2.28434667e-01 -2.27042422e-01 6.83865994e-02 1.89195108e-02
-1.21671252e-01 -8.84714648e-02 -6.50846601e-01 3.29257131e-01
5.77712536e-01 1.46485662e+00 -7.89293766e-01 -8.79163519e-02
-4.88566488e-01 1.05817425e+00 3.98814112e-01 1.55664876e-01
-8.64597023e-01 -2.25289077e-01 5.25411904e-01 2.16411471e-01
5.14433503e-01 -4.13937628e-01 1.48045436e-01 -7.25677490e-01
3.53019498e-03 -5.63381791e-01 3.24380606e-01 -2.45301992e-01
-8.93633485e-01 4.55185175e-01 -1.24146722e-01 -1.29136646e+00
-3.28547776e-01 -5.42307258e-01 -5.05325556e-01 6.03051484e-01
-1.66850603e+00 -1.23045635e+00 -3.86808634e-01 5.77940643e-01
4.95781779e-01 2.37356573e-01 6.94953382e-01 4.51183915e-01
-4.32023048e-01 9.25462365e-01 8.48243535e-02 1.81299850e-01
9.89301443e-01 -1.34305763e+00 6.33584142e-01 7.95997679e-01
4.77623433e-01 5.88232577e-01 6.35727167e-01 -5.73911726e-01
-1.55841136e+00 -1.18054509e+00 7.18646407e-01 -7.86260068e-01
7.15787530e-01 -5.26258826e-01 -5.71251512e-01 8.70305896e-01
-1.95349231e-01 5.65607071e-01 4.22076553e-01 2.59022355e-01
-1.49789631e-01 -3.90685022e-01 -1.02821672e+00 5.01153350e-01
1.13367152e+00 -4.47905868e-01 -3.91629413e-02 4.53787118e-01
3.37505817e-01 -8.13634932e-01 -6.81269884e-01 2.14547336e-01
9.18944001e-01 -7.40370393e-01 9.39025402e-01 -1.68867916e-01
-5.75994372e-01 -4.85444814e-01 2.07915112e-01 -1.03764522e+00
-1.87679514e-01 -1.17859530e+00 -6.18895888e-01 1.20601559e+00
3.24071348e-01 -8.22462499e-01 1.01194596e+00 2.09978774e-01
5.02107739e-01 -1.07364321e+00 -1.08979130e+00 -9.89420474e-01
-7.06735924e-02 -4.96559978e-01 1.76692337e-01 8.12841296e-01
-3.76612842e-01 1.34390399e-01 -4.80130970e-01 2.81266451e-01
8.10100436e-01 -1.96238130e-01 1.09473538e+00 -1.34558868e+00
-1.35946348e-02 -2.61824757e-01 -6.40787423e-01 -1.26077676e+00
7.60831386e-02 -5.82897544e-01 -1.02869935e-01 -1.36653376e+00
-2.92146862e-01 -6.67948067e-01 -1.00575395e-01 4.60596621e-01
-1.85171945e-03 6.89084172e-01 1.34257421e-01 2.62834430e-01
-7.63789475e-01 3.82594243e-02 8.50447595e-01 2.32990667e-01
-2.68688679e-01 5.06099105e-01 -6.24077976e-01 7.18516946e-01
7.69122481e-01 -5.87482452e-01 3.28025892e-02 -2.20421091e-01
1.38704002e-01 -8.76194704e-03 4.20523703e-01 -1.03571892e+00
7.97455251e-01 3.66611481e-01 7.50549793e-01 -6.07868671e-01
4.60347027e-01 -7.71623671e-01 4.02864404e-02 6.16672039e-01
6.95093349e-02 2.60572404e-01 5.64744830e-01 5.73717952e-01
7.74936900e-02 -9.56281424e-02 8.78428578e-01 -1.32144839e-01
-5.94756305e-01 4.84042317e-01 -4.56509739e-02 -2.12280110e-01
1.18129086e+00 -4.82940733e-01 1.72005966e-01 -5.64422905e-01
-7.18272209e-01 1.91382706e-01 4.19998020e-01 2.73037314e-01
4.05134976e-01 -1.32821155e+00 -7.41077840e-01 -2.87561845e-02
-3.93631041e-01 -2.25847334e-01 -2.70057619e-01 1.16467762e+00
-4.27305847e-01 3.24532688e-01 2.79527724e-01 -8.28803301e-01
-1.74981451e+00 1.82927936e-01 4.84453678e-01 -1.50324792e-01
-5.59255958e-01 1.12381840e+00 -3.43423009e-01 -3.94069970e-01
5.16230583e-01 1.74171075e-01 -7.09605590e-02 2.18605995e-01
7.40573347e-01 5.07003486e-01 1.18988492e-01 -6.88403428e-01
-6.25358880e-01 1.12558901e+00 -1.80896595e-01 -1.79890528e-01
1.20136976e+00 -3.12236906e-03 -9.87329111e-02 3.63225758e-01
1.13260972e+00 4.26676065e-01 -1.68874419e+00 -1.45410836e-01
2.28074476e-01 -4.85146403e-01 2.54170448e-02 -5.55132926e-01
-1.03437769e+00 6.33333325e-01 9.52204406e-01 1.28962714e-02
9.14992392e-01 -5.92330024e-02 4.41878617e-01 1.91867128e-01
5.54590821e-01 -8.65481377e-01 -5.98089844e-02 2.57863581e-01
7.92608678e-01 -1.32889152e+00 5.14545143e-01 -3.28414977e-01
-2.54849911e-01 1.20011663e+00 4.30331975e-01 -2.70917535e-01
7.15078831e-01 4.76335526e-01 2.10794017e-01 -7.19093978e-02
-5.81745088e-01 -3.04269582e-01 2.99526870e-01 5.89872539e-01
5.66496193e-01 -3.02858055e-01 2.35742815e-02 -2.31964111e-01
-1.36067197e-01 5.15234917e-02 1.24278657e-01 1.13935089e+00
-2.58699268e-01 -1.14973557e+00 -5.94465733e-01 1.83349743e-01
-4.98472214e-01 2.37962931e-01 -4.26173478e-01 1.03900218e+00
-5.91152422e-02 7.38261998e-01 -9.63632390e-02 -1.05853200e-01
2.77489454e-01 1.10506164e-02 8.92029047e-01 -4.47115183e-01
-4.88628745e-01 1.51380941e-01 -1.34893015e-01 -6.68522239e-01
-5.61519265e-01 -9.50757265e-01 -9.42410707e-01 -3.46058041e-01
-8.00913155e-01 2.04309717e-01 1.02387452e+00 7.56577909e-01
2.82221943e-01 3.67832541e-01 8.18545878e-01 -1.12271261e+00
-3.67851436e-01 -7.46440411e-01 -1.30585104e-01 -2.72996306e-01
7.51348615e-01 -8.06218147e-01 -4.33655500e-01 -4.83440123e-02] | [13.417744636535645, 0.25206828117370605] |
8a9cf26a-c30a-40a5-88fa-6a6a4639ec46 | opinion-based-relational-pivoting-for-cross-1 | null | null | https://aclanthology.org/2022.wassa-1.11 | https://aclanthology.org/2022.wassa-1.11.pdf | Opinion-based Relational Pivoting for Cross-domain Aspect Term Extraction | Domain adaptation methods often exploit domain-transferable input features, a.k.a. pivots. The task of Aspect and Opinion Term Extraction presents a special challenge for domain transfer: while opinion terms largely transfer across domains, aspects change drastically from one domain to another (e.g. from restaurants to laptops). In this paper, we investigate and establish empirically a prior conjecture, which suggests that the linguistic relations connecting opinion terms to their aspects transfer well across domains and therefore can be leveraged for cross-domain aspect term extraction. We present several analyses supporting this conjecture, via experiments with four linguistic dependency formalisms to represent relation patterns. Subsequently, we present an aspect term extraction method that drives models to consider opinion–aspect relations via explicit multitask objectives. This method provides significant performance gains, even on top of a prior state-of-the-art linguistically-informed model, which are shown in analysis to stem from the relational pivoting signal. | ['Ido Dagan', 'Moshe Wasserblat', 'Vasudev Lal', 'Daniel Korat', 'Oren Pereg', 'Ayal Klein'] | null | null | null | null | wassa-acl-2022-5 | ['term-extraction'] | ['natural-language-processing'] | [ 2.64352739e-01 3.06569457e-01 -5.98660111e-01 -6.71736836e-01
-1.10537195e+00 -1.03968179e+00 1.12108982e+00 4.88159657e-01
-3.57780874e-01 9.32079315e-01 4.56147015e-01 -4.50826973e-01
-2.90909231e-01 -6.82173133e-01 -6.98325217e-01 -3.88323724e-01
-8.45941529e-02 8.36763322e-01 5.22542857e-02 -7.02280104e-01
1.76430479e-01 1.29012197e-01 -1.12221444e+00 6.26252353e-01
7.57787883e-01 8.54922712e-01 -3.16995800e-01 2.23731101e-01
-5.70114732e-01 5.86663008e-01 -7.46261895e-01 -6.51790619e-01
4.39528450e-02 -2.72653848e-01 -1.11903715e+00 4.97274399e-02
2.32085869e-01 2.06523523e-01 3.26043904e-01 7.68736064e-01
3.11491191e-01 2.45170832e-01 1.20699787e+00 -1.22722626e+00
-1.14268672e+00 8.08306336e-01 -4.79128152e-01 2.09948733e-01
4.50017482e-01 -2.14539245e-01 1.54460084e+00 -8.70391309e-01
9.43967402e-01 1.26723170e+00 5.97484648e-01 3.18858773e-01
-1.56520152e+00 -4.83327299e-01 7.47057915e-01 8.04788247e-02
-1.01893449e+00 -4.57528204e-01 7.81181514e-01 -6.43534482e-01
1.40623605e+00 -1.44329935e-01 4.38201040e-01 1.28426969e+00
1.63856044e-01 8.79966140e-01 1.49610400e+00 -5.19609392e-01
1.92904636e-01 7.00340569e-01 3.85010839e-01 2.98424363e-01
4.84476119e-01 2.32846513e-02 -1.03328872e+00 -9.27390009e-02
2.32063547e-01 -6.61692977e-01 -4.22098786e-02 -5.36921263e-01
-1.09187114e+00 8.53953183e-01 2.21387953e-01 5.10048807e-01
-2.03284323e-01 -3.82272780e-01 4.62745786e-01 6.91382408e-01
9.41161871e-01 6.42871082e-01 -1.27726638e+00 6.90767989e-02
-6.53525710e-01 4.12054688e-01 1.18927121e+00 1.20885003e+00
1.09092224e+00 -3.25246662e-01 -8.86641443e-02 9.18838263e-01
3.20117325e-01 6.07666552e-01 2.85147458e-01 -6.02957964e-01
6.71357274e-01 5.84914923e-01 6.12675324e-02 -6.87011480e-01
-3.63422453e-01 -2.41023570e-01 -2.30684072e-01 -4.52933572e-02
3.82627696e-01 -3.62257957e-01 -7.84449637e-01 1.83929193e+00
3.81227702e-01 -4.39030439e-01 1.68553293e-01 6.20707393e-01
7.68476605e-01 2.38073394e-01 5.24314880e-01 7.17476802e-03
2.03283072e+00 -7.02127576e-01 -6.52418971e-01 -7.35253036e-01
5.87727904e-01 -7.21215546e-01 1.07934320e+00 2.20156953e-01
-8.23470235e-01 -4.45726514e-01 -8.45809579e-01 -2.65612304e-01
-8.45814049e-01 -1.56612411e-01 9.13765371e-01 5.01465797e-01
-9.05976653e-01 3.33447576e-01 -6.48215950e-01 -6.06930673e-01
2.51275361e-01 2.61185646e-01 -5.31151712e-01 5.53649329e-02
-1.62956989e+00 1.20078731e+00 3.87989014e-01 -4.49927330e-01
-3.28612536e-01 -8.53507400e-01 -1.09761417e+00 -1.67295560e-01
3.22495490e-01 -1.10832632e+00 1.40979779e+00 -1.28187072e+00
-1.41744018e+00 1.30424166e+00 -5.66243351e-01 -2.78675795e-01
-9.91812572e-02 -4.73957658e-01 -6.31085753e-01 -4.86842133e-02
4.99218464e-01 5.58231115e-01 8.88267219e-01 -1.36045253e+00
-9.13472533e-01 -4.13945615e-01 4.90409821e-01 2.99622774e-01
-4.05427039e-01 3.23061645e-01 -1.23687603e-01 -7.34895349e-01
-3.02934468e-01 -9.20966446e-01 4.65189405e-02 -2.09942192e-01
-2.42288575e-01 -4.81186420e-01 4.09898132e-01 -4.02434349e-01
1.01411355e+00 -1.77071476e+00 3.18811476e-01 1.69265181e-01
-5.01030087e-02 6.24297857e-02 -1.15191706e-01 4.40951556e-01
-2.78544039e-01 1.91831231e-01 -4.11525011e-01 -3.79647642e-01
3.04613113e-01 2.61972517e-01 -3.93509537e-01 1.49716094e-01
8.25505435e-01 9.68276143e-01 -7.14660704e-01 -3.90886575e-01
-2.15942696e-01 3.96822393e-01 -6.12876534e-01 5.13094366e-02
-3.73715460e-01 3.77670974e-01 -6.09475076e-01 5.91861010e-01
5.83783329e-01 -3.22135240e-01 1.96839303e-01 -2.83409238e-01
-6.55569881e-02 1.00617814e+00 -6.47556603e-01 1.94021881e+00
-8.38051736e-01 6.10217273e-01 1.05394684e-01 -1.21710539e+00
9.79040921e-01 2.94891268e-01 3.57967138e-01 -5.98906398e-01
6.36214996e-03 1.86096534e-01 -1.17750369e-01 -3.86095852e-01
6.40545785e-01 -6.88890398e-01 -4.70799118e-01 5.78696012e-01
5.14083385e-01 -4.32136059e-01 3.69613886e-01 1.48944616e-01
7.59186149e-01 4.57218200e-01 6.62253141e-01 -6.59995615e-01
6.80683911e-01 4.22766119e-01 3.93337190e-01 3.52503657e-01
-1.49666160e-01 2.46391743e-01 6.97402716e-01 -2.44463652e-01
-6.63862050e-01 -1.10267889e+00 -3.39635253e-01 1.64644778e+00
-3.95201743e-02 -3.81434411e-01 -3.72146130e-01 -1.00401866e+00
2.76451707e-01 9.86396432e-01 -7.09491014e-01 -1.25240371e-01
-5.23771703e-01 -5.91958523e-01 2.35992685e-01 5.11905670e-01
1.68409050e-01 -9.85460460e-01 -1.13747127e-01 2.33552530e-01
-2.24856421e-01 -1.36540568e+00 -1.59480706e-01 5.45395494e-01
-6.49234712e-01 -7.53226638e-01 -6.58558249e-01 -7.91955948e-01
2.09904626e-01 -3.06799375e-02 1.89987695e+00 -6.13243163e-01
3.05502594e-01 5.36929667e-01 -5.41276038e-01 -7.08973527e-01
-2.99246460e-01 6.39306843e-01 -1.95862502e-02 -3.36946756e-01
1.04419172e+00 -7.78887808e-01 -4.39818680e-01 2.56176293e-01
-7.76450694e-01 -4.04338628e-01 6.91001177e-01 5.57742357e-01
4.69837785e-01 -3.53934050e-01 6.80577874e-01 -1.41130006e+00
8.73428822e-01 -8.35871041e-01 -4.09354538e-01 2.32039973e-01
-6.18142486e-01 3.07716131e-01 4.18161929e-01 -2.87676811e-01
-1.40132904e+00 -2.90014267e-01 5.99811785e-02 3.30082297e-01
-4.29197758e-01 7.73227990e-01 -2.50310034e-01 3.00023228e-01
8.64487708e-01 -9.32485834e-02 -3.31888467e-01 -3.55823725e-01
8.39202523e-01 6.41800940e-01 1.31389961e-01 -1.07562840e+00
8.95562351e-01 6.31869912e-01 -2.26275340e-01 -8.72745037e-01
-1.38153219e+00 -5.91071308e-01 -7.97013938e-01 3.15801889e-01
7.85708070e-01 -1.13164353e+00 -8.02719519e-02 8.90177637e-02
-1.26047754e+00 -1.70136511e-01 -4.56249297e-01 2.64042139e-01
-4.80820864e-01 -1.76781081e-02 -4.56986606e-01 -4.84428912e-01
-1.42909303e-01 -6.39646828e-01 1.39601350e+00 2.37263404e-02
-7.34287322e-01 -1.56690598e+00 3.46876413e-01 3.69641066e-01
3.91888767e-01 1.11313000e-01 1.15472269e+00 -1.01656425e+00
-3.03385764e-01 3.00743073e-01 -2.60686964e-01 3.12196314e-01
5.44606805e-01 -3.17280740e-01 -1.08742619e+00 -4.61701639e-02
1.93555020e-02 -2.83964276e-01 8.50098610e-01 2.84997165e-01
2.41828993e-01 -1.82468697e-01 -4.94839430e-01 3.63962948e-01
1.13681686e+00 -8.43411908e-02 1.24392591e-01 6.15272403e-01
5.51523030e-01 9.83903885e-01 7.15514362e-01 8.93897042e-02
8.53463292e-01 5.81969380e-01 -3.87966305e-01 8.71098787e-02
-2.34733582e-01 -1.51778683e-01 5.83719373e-01 8.79613280e-01
1.25298128e-01 -1.53648302e-01 -1.06487584e+00 1.01132786e+00
-1.50025773e+00 -4.69384968e-01 4.61858660e-02 1.77136755e+00
1.26014590e+00 3.33475947e-01 1.15668580e-01 -3.01635474e-01
3.72742116e-01 3.21897328e-01 -5.11158943e-01 -5.51151693e-01
-2.98442751e-01 7.04658806e-01 1.99714661e-01 5.87011218e-01
-1.21061313e+00 1.25052905e+00 6.30366135e+00 5.83926976e-01
-9.21568513e-01 1.72959030e-01 2.59616703e-01 2.27547273e-01
-7.43461430e-01 1.73241392e-01 -9.65082407e-01 -2.86389422e-02
1.05451286e+00 -4.12657291e-01 1.37875989e-01 8.65203917e-01
-1.57933950e-01 3.87341492e-02 -1.54370117e+00 4.01041090e-01
6.78305179e-02 -8.23406100e-01 2.35895917e-01 -4.39907275e-02
9.96348917e-01 1.20215267e-01 1.11718580e-01 5.37478745e-01
6.06089771e-01 -8.25323880e-01 5.39442122e-01 1.94492742e-01
7.52088428e-01 -7.33704090e-01 5.43496370e-01 -3.76557149e-02
-1.26256239e+00 2.70668894e-01 -1.61366507e-01 -7.43565187e-02
4.12053913e-01 8.24401557e-01 -7.06811488e-01 7.65826225e-01
5.99807203e-01 8.90223682e-01 -4.54729229e-01 2.39268348e-01
-7.71154165e-01 6.15184724e-01 -2.01710001e-01 1.29644185e-01
1.53713673e-01 -1.52562425e-01 7.35725701e-01 1.48124516e+00
1.13109373e-01 -1.22708924e-01 3.90448533e-02 8.45258474e-01
7.35494308e-03 2.38415822e-01 -9.64774251e-01 -1.17595181e-01
2.88622946e-01 1.12594342e+00 -4.85278219e-01 -4.40611005e-01
-7.69274533e-01 9.38409507e-01 4.77344751e-01 6.08316362e-01
-4.60336119e-01 -4.71210390e-01 1.16666591e+00 5.08158468e-02
6.49783313e-01 -2.47130647e-01 -4.51234162e-01 -1.33006608e+00
9.99448523e-02 -9.94224131e-01 3.89459819e-01 -4.31325793e-01
-1.80627501e+00 6.84872687e-01 2.77497739e-01 -1.07326198e+00
-6.00608468e-01 -9.03688252e-01 -2.61042804e-01 1.06916809e+00
-2.06348372e+00 -1.39825344e+00 1.81665912e-01 6.84813619e-01
4.87710714e-01 -1.81485176e-01 9.76748228e-01 1.14830948e-01
-4.70220111e-03 5.55244744e-01 -2.81052291e-01 6.22931235e-02
1.14260769e+00 -1.51181972e+00 8.02701712e-01 5.16555488e-01
3.98323864e-01 1.11958086e+00 7.60438800e-01 -4.96270478e-01
-9.82007623e-01 -9.52592075e-01 1.43171775e+00 -1.05789435e+00
1.09957302e+00 -4.08566207e-01 -9.56227362e-01 8.82225692e-01
6.34244263e-01 -1.92046523e-01 1.13300216e+00 9.60489631e-01
-9.03056860e-01 -6.39475212e-02 -9.21583474e-01 5.05757749e-01
1.10746908e+00 -9.68661547e-01 -1.30977142e+00 2.26508573e-01
8.53801131e-01 -1.36335529e-02 -9.45530772e-01 3.05983543e-01
4.78240609e-01 -9.05392051e-01 9.03367043e-01 -1.07414842e+00
5.52546859e-01 -3.08208883e-01 -1.50403902e-01 -1.79160130e+00
-2.39737570e-01 -5.70419967e-01 -1.88853621e-01 1.65303159e+00
9.98737156e-01 -8.37820232e-01 4.05118942e-01 6.66702271e-01
1.74929619e-01 -2.65621632e-01 -8.76513422e-01 -8.46978903e-01
6.97284222e-01 -4.96498078e-01 5.37340939e-01 1.14899647e+00
2.71426797e-01 1.18337846e+00 2.57626683e-01 6.08570129e-02
3.85756642e-01 3.55713367e-01 4.43527251e-01 -1.25545621e+00
-4.82366055e-01 -4.11845088e-01 -2.46453479e-01 -1.16336882e+00
6.68145418e-01 -1.07879257e+00 -4.51424681e-02 -1.53009343e+00
-7.80420378e-02 -4.30991471e-01 -4.91678953e-01 3.68702680e-01
-4.04516727e-01 -9.71561149e-02 -3.70332636e-02 -1.18664443e-01
-5.85021794e-01 5.47086000e-01 1.03294325e+00 -2.90684283e-01
-1.89420909e-01 1.14311770e-01 -1.30927038e+00 8.45992684e-01
8.02620828e-01 -5.84750831e-01 -5.54609120e-01 -6.62171423e-01
4.75629389e-01 -3.61404508e-01 -1.42688870e-01 -4.98123407e-01
-1.55929187e-02 -1.66614726e-02 1.51873278e-02 -1.16704196e-01
3.39544386e-01 -8.70271385e-01 -4.44078952e-01 -1.09510139e-01
-4.27715659e-01 -3.15170512e-02 4.71616864e-01 5.83118975e-01
-4.60988194e-01 9.37687829e-02 4.62449044e-01 -7.39916861e-02
-6.36253178e-01 -2.50462200e-02 -4.38592255e-01 7.56421566e-01
6.12433493e-01 1.70376986e-01 -2.85952449e-01 -3.40115547e-01
-7.90759504e-01 8.84593725e-02 2.20144376e-01 7.27155745e-01
2.17458028e-02 -1.32793450e+00 -8.46328497e-01 4.19464633e-02
5.64600587e-01 -3.02735679e-02 -8.23508650e-02 6.08685434e-01
2.41888717e-01 6.29381835e-01 -1.45205120e-02 -4.73597169e-01
-1.02263999e+00 5.33903658e-01 8.51378497e-03 -6.31021678e-01
-1.22972265e-01 9.40708399e-01 4.68735993e-01 -7.27604210e-01
-3.07173550e-01 -6.23883009e-01 -2.37194881e-01 5.89316428e-01
3.52292985e-01 -1.10919237e-01 2.39739999e-01 -6.40977859e-01
-6.54996634e-01 8.41759443e-01 -2.90599465e-01 -4.49488670e-01
1.35893691e+00 -2.67452776e-01 -2.40346417e-01 6.25213802e-01
1.15580618e+00 1.35419235e-01 -9.08274174e-01 -6.83950365e-01
5.18106163e-01 -9.73120779e-02 -3.35769683e-01 -1.04372609e+00
-6.61127090e-01 7.18030751e-01 5.88605134e-03 3.21229160e-01
9.68453288e-01 4.64016765e-01 5.33053637e-01 5.21418095e-01
3.47729921e-01 -1.14148247e+00 -2.94919312e-01 9.30794477e-01
8.62665892e-01 -1.36701429e+00 1.58469424e-01 -4.50705200e-01
-1.00305319e+00 7.51329124e-01 3.69879246e-01 -1.67499483e-01
9.66801822e-01 3.17743421e-01 2.88728237e-01 -3.07437062e-01
-9.55894530e-01 -6.11019909e-01 5.36398113e-01 1.07619679e+00
8.27423275e-01 1.04648225e-01 -2.95955062e-01 8.20806682e-01
-4.03868377e-01 -1.24221548e-01 -9.56702754e-02 9.02361274e-01
-2.57637054e-01 -1.56728077e+00 -1.71955675e-01 1.83064535e-01
-5.63127995e-01 -4.62992251e-01 -8.01834047e-01 9.17980671e-01
8.84675458e-02 1.03661764e+00 -1.44967079e-01 2.58117467e-02
6.81149065e-01 4.14768040e-01 3.48245442e-01 -9.20000494e-01
-7.14680910e-01 -1.72924042e-01 6.39132977e-01 -3.50684375e-01
-8.85174155e-01 -7.78496742e-01 -8.81492019e-01 -3.04473061e-02
-1.05712719e-01 3.27102572e-01 5.47603190e-01 1.18944895e+00
5.23715854e-01 5.78897953e-01 1.31810546e-01 -6.49461091e-01
-2.76739210e-01 -1.06330550e+00 -4.96054411e-01 5.34605145e-01
4.17703182e-01 -7.39628196e-01 -3.34657937e-01 3.14862162e-01] | [11.334986686706543, 6.872641086578369] |
cd5c82a1-5b94-4070-92b0-bf2a8cf39915 | adversarial-normalization-i-can-visualize | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Choi_Adversarial_Normalization_I_Can_Visualize_Everything_ICE_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Choi_Adversarial_Normalization_I_Can_Visualize_Everything_ICE_CVPR_2023_paper.pdf | Adversarial Normalization: I Can Visualize Everything (ICE) | Vision transformers use [CLS] tokens to predict image classes. Their explainability visualization has been studied using relevant information from [CLS] tokens or focusing on attention scores during self-attention. Such visualization, however, is challenging because of the dependence of the structure of a vision transformer on skip connections and attention operators, the instability of non-linearities in the learning process, and the limited reflection of self-attention scores on relevance. We argue that the output vectors for each input patch token in a vision transformer retain the image information of each patch location, which can facilitate the prediction of an image class. In this paper, we propose ICE (Adversarial Normalization: I Can visualize Everything), a novel method that enables a model to directly predict a class for each patch in an image; thus, advancing the effective visualization of the explainability of a vision transformer. Our method distinguishes background from foreground regions by predicting background classes for patches that do not determine image classes. We used the DeiT-S model, the most representative model employed in studies, on the explainability visualization of vision transformers. On the ImageNet-Segmentation dataset, ICE outperformed all explainability visualization methods for four cases depending on the model size. We also conducted quantitative and qualitative analyses on the tasks of weakly-supervised object localization and unsupervised object discovery. On the CUB-200-2011 and PASCALVOC07/12 datasets, ICE achieved comparable performance to the state-of-the-art methods. We incorporated ICE into the encoder of DeiT-S and improved efficiency by 44.01% on the ImageNet dataset over that achieved by the original DeiT-S model. We showed performance on the accuracy and efficiency comparable to EViT, the state-of-the-art pruning model, demonstrating the effectiveness of ICE. The code is available at https://github.com/Hanyang-HCC-Lab/ICE. | ['Kyungsik Han', 'Seungwan Jin', 'Hoyoung Choi'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['object-discovery', 'object-localization', 'weakly-supervised-object-localization'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 2.53474951e-01 4.95967776e-01 8.54956731e-02 -2.09906742e-01
-2.17055753e-01 -5.40181935e-01 6.36520803e-01 2.51804888e-02
-1.45190760e-01 3.58803511e-01 2.10083071e-02 -5.60696006e-01
-1.08299136e-01 -5.48827827e-01 -1.04538023e+00 -6.72521770e-01
6.70998469e-02 4.52642202e-01 3.61319929e-01 9.94222164e-02
2.40978166e-01 5.10650277e-01 -1.97706699e+00 7.94300556e-01
9.11464036e-01 1.10961640e+00 5.30590534e-01 5.54061055e-01
-2.07469299e-01 9.79414523e-01 -7.41658986e-01 -4.03760821e-01
1.45812929e-01 -4.10273165e-01 -6.55175686e-01 -9.56673324e-02
1.12751663e+00 1.65264308e-03 -1.76682815e-01 1.09795558e+00
1.58432618e-01 -2.51994580e-01 7.85928905e-01 -1.58976364e+00
-9.79970634e-01 7.34535456e-01 -6.01914704e-01 3.58692646e-01
-1.11420015e-02 2.98313290e-01 1.05311286e+00 -1.07180929e+00
7.21170723e-01 1.17341161e+00 6.55656397e-01 5.55751979e-01
-1.50275993e+00 -7.46375561e-01 4.09016281e-01 5.39183319e-01
-1.01706946e+00 4.07178774e-02 6.92299306e-01 -6.42970979e-01
1.02884817e+00 7.63562858e-01 1.10246480e+00 7.60189056e-01
4.87421341e-02 8.54494512e-01 1.18692005e+00 -2.75784016e-01
1.20509446e-01 4.11350697e-01 2.83606648e-01 1.04092073e+00
1.86558247e-01 2.37407252e-01 -7.05345154e-01 2.94058591e-01
6.14871800e-01 2.14298368e-01 -3.62460136e-01 -4.67472881e-01
-1.07349122e+00 6.75624907e-01 1.04839242e+00 1.73852161e-01
-2.71122217e-01 3.58060211e-01 1.25694841e-01 2.57124100e-02
6.25468671e-01 6.29469037e-01 -3.92946005e-01 1.00306451e-01
-1.03146017e+00 1.01138629e-01 6.01086199e-01 7.32801616e-01
6.89337432e-01 1.26392677e-01 -6.21984899e-01 4.27205473e-01
2.12838557e-02 4.72508401e-01 3.00018907e-01 -8.53658736e-01
3.63667011e-01 8.66954625e-01 -2.95142382e-01 -9.00604010e-01
-3.75993043e-01 -8.17331731e-01 -9.09507334e-01 9.05775726e-01
5.32834530e-01 2.16115892e-01 -1.36010587e+00 1.65212286e+00
6.87577799e-02 1.37268379e-01 -2.96839382e-02 1.01867104e+00
1.03425050e+00 5.90750992e-01 1.51643321e-01 2.34434649e-01
1.42779028e+00 -1.11545861e+00 -5.62963068e-01 -3.19691926e-01
2.60089606e-01 -5.27720034e-01 1.53127301e+00 3.78661871e-01
-1.08180416e+00 -7.45215893e-01 -1.05213404e+00 -1.28627911e-01
-7.03896344e-01 3.58093381e-01 6.08156860e-01 3.35803270e-01
-1.13610065e+00 6.61703646e-01 -7.76742280e-01 -1.48295894e-01
1.06325352e+00 3.48310679e-01 -2.08329514e-01 1.77678585e-01
-7.41782486e-01 1.00712645e+00 3.68906148e-02 -2.24733967e-02
-1.11192214e+00 -1.21359360e+00 -6.00856841e-01 6.41768754e-01
1.16094917e-01 -7.40627289e-01 8.24365795e-01 -1.16981089e+00
-7.72510707e-01 9.58155930e-01 -1.14677183e-01 -8.88844788e-01
6.58162117e-01 -3.24161023e-01 1.12516128e-01 4.76549305e-02
5.58412373e-02 1.15015376e+00 9.46435452e-01 -1.35856402e+00
-6.25145674e-01 -3.85417581e-01 1.71341494e-01 6.83846921e-02
-1.45705774e-01 -4.15830940e-01 -4.56508279e-01 -5.55795729e-01
-3.48971821e-02 -8.25887203e-01 -6.74909502e-02 4.05615658e-01
-7.73746192e-01 -2.11442068e-01 1.18887174e+00 -6.22772336e-01
8.23887587e-01 -2.29646373e+00 1.05463527e-01 -8.95495433e-03
6.35032117e-01 4.44411635e-01 -5.98000772e-02 -1.65760115e-01
-4.55038071e-01 4.19750571e-01 -3.78216386e-01 -5.06873190e-01
-6.67761564e-02 1.92294553e-01 -5.42925298e-01 3.53458434e-01
4.88713235e-01 9.19415653e-01 -6.48326397e-01 -4.43769604e-01
5.80272138e-01 8.01378429e-01 -5.43039858e-01 1.12047501e-01
-3.00845265e-01 2.30222419e-01 1.53806172e-02 5.51892519e-01
6.47459865e-01 -4.14688021e-01 -1.78700969e-01 -5.31625986e-01
-2.17053488e-01 1.85642600e-01 -7.06129074e-01 1.31871915e+00
-4.02224690e-01 1.29997444e+00 -3.24219942e-01 -7.86176860e-01
7.23153412e-01 -4.81559373e-02 8.62784907e-02 -9.05358970e-01
-2.05935284e-01 -9.48487148e-02 1.22481368e-01 -4.35217947e-01
2.79517442e-01 4.11416143e-01 5.02498388e-01 2.36940935e-01
-3.36079788e-03 -9.17124227e-02 3.13793808e-01 4.30521935e-01
9.11756694e-01 1.88179135e-01 -1.27700463e-01 -4.51773167e-01
2.03016445e-01 9.35043767e-02 2.09040150e-01 9.46692169e-01
4.17721272e-02 8.69363546e-01 8.69299591e-01 -6.32770121e-01
-9.60346401e-01 -1.11883497e+00 -1.87601492e-01 1.09771347e+00
3.00665408e-01 -4.41638380e-01 -9.31729376e-01 -9.24445808e-01
1.42746851e-01 1.11136734e+00 -1.19433653e+00 -1.32857129e-01
-3.91271502e-01 -4.87007111e-01 4.12382603e-01 5.03073275e-01
4.81714994e-01 -1.37724459e+00 -1.13532925e+00 -2.74190485e-01
-5.72894216e-02 -7.49001980e-01 -2.34939277e-01 6.03298724e-01
-7.78600991e-01 -1.37797439e+00 -4.82583374e-01 -5.68647027e-01
9.89094079e-01 4.07336503e-02 1.34735489e+00 2.59804636e-01
-7.22430885e-01 2.52621412e-01 1.10323906e-01 -6.19567096e-01
-1.93049356e-01 3.34255993e-02 -4.48788077e-01 -2.42615953e-01
1.37476936e-01 -4.65371013e-01 -8.03159058e-01 2.91469455e-01
-7.35422850e-01 6.83479011e-01 6.10382259e-01 8.19977939e-01
8.67216766e-01 -2.80720323e-01 7.16053173e-02 -1.07742131e+00
3.84442478e-01 -1.83900043e-01 -6.05052412e-01 3.50059181e-01
-7.85212517e-01 3.44166011e-01 4.51575518e-01 -5.26537597e-01
-8.95819366e-01 1.79265201e-01 2.69210070e-01 -7.35761464e-01
-1.44645432e-02 3.49959284e-02 -1.34988621e-01 -7.81636834e-02
6.62589192e-01 2.32345000e-01 -9.89025552e-03 -3.78277928e-01
6.85542345e-01 6.44597486e-02 7.06898153e-01 -1.76112846e-01
5.94871581e-01 4.57382172e-01 4.56302501e-02 -5.32398880e-01
-9.00922060e-01 1.44529613e-02 -4.46887970e-01 -3.25641423e-01
9.65780258e-01 -5.65463781e-01 -7.35006988e-01 -2.22109202e-02
-1.33272827e+00 -5.17014742e-01 -7.38068104e-01 -1.97721049e-02
-4.20860857e-01 -1.07837506e-01 -2.80901283e-01 -6.42617047e-01
-3.54405850e-01 -1.35278606e+00 8.79636049e-01 3.24487180e-01
-1.99725628e-01 -7.83913136e-01 -4.42615896e-01 1.51146427e-01
5.39159119e-01 2.98014641e-01 1.12572157e+00 -5.59659004e-01
-8.49605620e-01 2.68966287e-01 -6.55086994e-01 2.31158420e-01
-1.39941558e-01 6.36162683e-02 -1.31269252e+00 1.10744901e-01
-1.63046539e-01 -1.30962357e-01 1.44620740e+00 4.97536302e-01
1.78662336e+00 -5.93246758e-01 -4.55147356e-01 7.67906606e-01
1.25119507e+00 -4.61405180e-02 6.10752761e-01 4.39568043e-01
9.03081357e-01 6.53915465e-01 4.42369938e-01 1.02611937e-01
5.91757856e-02 5.71706951e-01 1.18174684e+00 -8.14101994e-01
-6.82952702e-01 -2.41973758e-01 2.38324013e-02 -1.03563808e-01
-1.59505978e-01 -2.01553270e-01 -8.25471163e-01 5.50713897e-01
-1.87952995e+00 -8.35050046e-01 -2.96398789e-01 2.07631516e+00
6.10235453e-01 1.57974198e-01 -1.78376675e-01 1.15393974e-01
5.61808884e-01 3.39249447e-02 -8.16167355e-01 -3.94484043e-01
-2.70468146e-01 7.00289607e-02 4.49111849e-01 5.14334738e-01
-1.03653753e+00 7.80157089e-01 5.50691986e+00 6.42611444e-01
-1.16725874e+00 1.49553657e-01 1.10886014e+00 -3.81459385e-01
-3.30018789e-01 -1.33407250e-01 -6.18151724e-01 4.52422768e-01
4.29662943e-01 1.71221420e-01 3.80398601e-01 8.94268036e-01
1.73266174e-03 -7.61247054e-02 -1.33212519e+00 9.74723995e-01
1.74179733e-01 -1.60583913e+00 2.27266669e-01 2.37623956e-02
6.03696167e-01 -8.43424816e-03 6.52776897e-01 1.21578395e-01
4.47024070e-02 -1.25054455e+00 1.09571266e+00 5.62477827e-01
9.04447734e-01 -5.69340408e-01 4.55837458e-01 7.31967464e-02
-9.14437473e-01 -2.74802685e-01 -4.56061840e-01 1.44267425e-01
-1.66364804e-01 5.18457770e-01 -9.53600585e-01 4.08133632e-03
1.06668437e+00 8.43422830e-01 -1.13248742e+00 9.80877161e-01
-5.03221393e-01 7.96477020e-01 -2.19091699e-01 9.99204889e-02
3.04605633e-01 5.79699874e-02 5.72901726e-01 1.27660239e+00
1.55908242e-01 -2.13132083e-01 -2.73804098e-01 1.54477310e+00
-3.56024466e-02 -2.23999083e-01 -5.25098920e-01 2.75396615e-01
1.23959601e-01 1.30212677e+00 -8.86588871e-01 -5.19294798e-01
-1.41663611e-01 8.73473704e-01 5.21030724e-01 4.86804992e-01
-1.09665978e+00 -2.27055088e-01 5.70467472e-01 4.57794070e-01
6.90643311e-01 2.71975487e-01 -7.76880324e-01 -5.98015487e-01
2.64992770e-02 -8.15620422e-01 3.11261684e-01 -1.20632446e+00
-8.56320679e-01 9.31219101e-01 -1.04289412e-01 -1.06477904e+00
9.68989916e-04 -7.57561386e-01 -6.87550068e-01 7.51248181e-01
-1.28578174e+00 -1.29425251e+00 -6.44868553e-01 4.02162701e-01
5.61871767e-01 -9.27093998e-02 6.23451471e-01 -1.71717275e-02
-3.99570614e-01 4.58193451e-01 -2.44305357e-01 -1.94746301e-01
3.44356865e-01 -1.58998716e+00 4.09464359e-01 7.81987011e-01
7.06431389e-01 4.54367727e-01 8.95839930e-01 -5.22070825e-01
-8.96258533e-01 -1.04384410e+00 6.44782484e-01 -5.26202977e-01
4.76601034e-01 -5.78484297e-01 -1.07357752e+00 5.53991139e-01
5.12986064e-01 7.96145722e-02 2.82824337e-01 1.00720316e-01
-5.05366027e-01 -2.23262593e-01 -1.04484451e+00 5.98216295e-01
1.16652262e+00 -3.44021916e-01 -4.78169829e-01 2.53545254e-01
8.43525946e-01 -3.98389101e-01 -3.30729336e-01 3.27850342e-01
6.10231340e-01 -1.39472330e+00 1.15176427e+00 -5.08953094e-01
7.29447782e-01 -3.32080275e-01 9.58983302e-02 -1.22095740e+00
-3.27365637e-01 -4.83983122e-02 -2.12310851e-01 1.20915627e+00
7.65522480e-01 -5.38785100e-01 7.27482319e-01 6.41601086e-01
-1.57465994e-01 -1.03216529e+00 -8.41214716e-01 -2.54972219e-01
-2.48249337e-01 -2.71374881e-01 5.56981802e-01 5.54030359e-01
-4.08444673e-01 2.11181834e-01 -1.19978786e-01 1.40741989e-01
5.69119453e-01 2.47191966e-01 5.66666067e-01 -1.21691453e+00
-2.10000396e-01 -7.28650868e-01 -4.78155822e-01 -7.30118096e-01
5.43110333e-02 -9.58457470e-01 -1.29805461e-01 -1.79115140e+00
3.94492865e-01 -2.93276161e-01 -3.86766434e-01 9.42395508e-01
-2.56654471e-01 4.33955193e-01 6.44112349e-01 3.31320971e-01
-5.72144210e-01 3.46809775e-01 1.26093996e+00 -6.30430043e-01
-9.47610736e-02 -9.21198502e-02 -7.33406126e-01 6.10162735e-01
8.24218273e-01 -5.89303017e-01 -5.21815598e-01 -5.01428306e-01
8.75747427e-02 -3.58883858e-01 1.01385963e+00 -1.06493711e+00
2.40482226e-01 3.38258594e-01 7.91951239e-01 -8.13130319e-01
3.34152251e-01 -9.23187435e-01 1.64929017e-01 7.81785488e-01
-6.13112986e-01 1.30086020e-01 6.54981494e-01 4.64288563e-01
-8.22445005e-02 5.12463041e-02 5.56682587e-01 -1.13385720e-02
-6.14279270e-01 1.38783529e-01 -1.43204585e-01 -2.38451213e-02
7.28221416e-01 -2.74122089e-01 -7.29474545e-01 -1.96474150e-01
-7.79222965e-01 8.59827176e-02 4.12742227e-01 3.92790884e-01
7.28873312e-01 -9.86354828e-01 -5.10193646e-01 4.18651819e-01
-5.72877228e-02 4.69578914e-02 2.39843056e-01 7.49771416e-01
-4.00561899e-01 2.23491788e-01 -4.41539556e-01 -1.06722355e+00
-1.63623869e+00 5.80402195e-01 5.82553625e-01 -2.19056129e-01
-8.53097618e-01 8.65838706e-01 9.70972836e-01 -6.20157504e-03
4.80932713e-01 -5.98732531e-01 -3.21996659e-01 -1.70438409e-01
4.82040435e-01 2.71552294e-01 -1.15882397e-01 -3.21359366e-01
-3.67064446e-01 5.10923028e-01 -2.88993935e-03 4.82473755e-03
1.32673395e+00 1.07316352e-01 -5.08952402e-02 3.96549195e-01
8.35392416e-01 -2.60415107e-01 -1.61053371e+00 1.56103611e-01
-2.64014989e-01 -3.88179779e-01 1.81886926e-03 -1.31345809e+00
-1.43923283e+00 1.22296655e+00 1.04482257e+00 3.35128278e-01
1.12270248e+00 2.82493114e-01 -3.93086411e-02 2.57880781e-02
-2.01283500e-01 -4.92351145e-01 8.27023908e-02 1.43461093e-01
1.27910709e+00 -1.20268583e+00 -8.08142796e-02 -4.69061613e-01
-7.66775310e-01 9.45267200e-01 9.09869969e-01 1.59066971e-02
4.95744079e-01 3.48736286e-01 8.20747316e-02 -5.57155013e-01
-7.49860227e-01 -1.65796831e-01 6.60357773e-01 7.38647461e-01
3.35134119e-01 1.09301627e-01 2.08562136e-01 5.52825868e-01
-4.32331085e-01 -3.40257734e-01 2.08726242e-01 3.49208504e-01
-1.91940233e-01 -4.70381975e-01 -2.00398877e-01 5.37910283e-01
-3.56382340e-01 -4.73295838e-01 -6.36548400e-01 9.71369803e-01
4.58088249e-01 6.07253850e-01 3.84537667e-01 -3.21474940e-01
2.62851089e-01 -1.08904220e-01 3.90621901e-01 -3.25845093e-01
-8.27459514e-01 -2.64270633e-01 -1.11538075e-01 -7.98964679e-01
-2.31161639e-01 -4.76350397e-01 -1.24718559e+00 -1.30715603e-02
-1.84751645e-01 -5.30679375e-02 8.40617001e-01 5.72049916e-01
5.11377096e-01 1.00964415e+00 1.10282451e-01 -8.38698328e-01
-1.02217086e-01 -8.60661209e-01 -3.22484761e-01 5.80549479e-01
4.48233843e-01 -7.06898212e-01 -5.24637997e-01 2.36337349e-01] | [9.916299819946289, 1.8866589069366455] |
1e56a6e6-459e-4836-a734-a31a04897505 | an-empirical-study-on-end-to-end-singing | 2108.03008 | null | https://arxiv.org/abs/2108.03008v1 | https://arxiv.org/pdf/2108.03008v1.pdf | An Empirical Study on End-to-End Singing Voice Synthesis with Encoder-Decoder Architectures | With the rapid development of neural network architectures and speech processing models, singing voice synthesis with neural networks is becoming the cutting-edge technique of digital music production. In this work, in order to explore how to improve the quality and efficiency of singing voice synthesis, in this work, we use encoder-decoder neural models and a number of vocoders to achieve singing voice synthesis. We conduct experiments to demonstrate that the models can be trained using voice data with pitch information, lyrics and beat information, and the trained models can produce smooth, clear and natural singing voice that is close to real human voice. As the models work in the end-to-end manner, they allow users who are not domain experts to directly produce singing voice by arranging pitches, lyrics and beats. | ['Cheng-Hao Cai', 'Jing Sun', 'Yanyan Xu', 'Xudong Liu', 'Yuxing Lu', 'Dengfeng Ke'] | 2021-08-06 | null | null | null | null | ['singing-voice-synthesis'] | ['speech'] | [-1.71045721e-01 -1.39406309e-01 -1.62046760e-01 1.50260292e-02
-3.01347554e-01 -4.92724806e-01 -1.29252234e-02 -6.65901601e-01
2.57849216e-01 3.32019150e-01 5.34751892e-01 -1.39745891e-01
2.70992756e-01 -5.07461071e-01 -3.57017130e-01 -2.98355401e-01
8.61124992e-02 -3.25696506e-02 -1.00279510e-01 -4.54217821e-01
-1.30738065e-01 4.16912377e-01 -1.78747416e+00 4.54677433e-01
5.74395776e-01 7.97665954e-01 3.65546674e-01 1.29391038e+00
-1.23630479e-01 8.60055625e-01 -8.46794188e-01 -1.74153708e-02
4.22593594e-01 -9.62077022e-01 -3.16599190e-01 -9.04539973e-02
3.11343998e-01 -5.05701840e-01 -5.44393182e-01 8.29128563e-01
1.01056433e+00 2.30988264e-01 3.94805759e-01 -7.53825784e-01
-7.93770730e-01 8.72803092e-01 9.55789462e-02 -1.83778018e-01
3.12866300e-01 4.73932803e-01 1.23375034e+00 -9.44784999e-01
2.60090083e-01 1.19118333e+00 6.70128942e-01 1.02532387e+00
-8.81106734e-01 -8.51403415e-01 -5.29867053e-01 1.07365854e-01
-1.19410121e+00 -8.68741333e-01 9.93816674e-01 -1.38638124e-01
9.46624935e-01 5.37781954e-01 8.76057267e-01 6.40959978e-01
-5.45542054e-02 8.99542987e-01 5.54768980e-01 -8.01419079e-01
-1.02337971e-01 -1.28411800e-01 -5.22038758e-01 3.65488201e-01
-5.03181458e-01 5.29831409e-01 -9.56024289e-01 2.20051095e-01
1.10851455e+00 -4.49327409e-01 -4.80819613e-01 3.37101460e-01
-1.22848451e+00 5.55054903e-01 3.66252571e-01 4.12841856e-01
-5.61150730e-01 2.86723822e-01 3.38763416e-01 3.27715844e-01
1.97639619e-03 9.96575475e-01 -2.53816724e-01 -5.02188325e-01
-1.36810446e+00 4.65422750e-01 8.84855628e-01 7.79412866e-01
-2.29758039e-01 1.02483082e+00 -3.81386220e-01 1.38637638e+00
2.13538885e-01 5.73366821e-01 7.14003980e-01 -1.22405696e+00
1.72049522e-01 -1.07508879e-02 -2.85456106e-02 -6.86689138e-01
-1.96536347e-01 -4.51068580e-01 -6.86058700e-01 2.66436040e-01
1.96436733e-01 -4.10554081e-01 -7.79560089e-01 1.72158384e+00
-2.76846308e-02 1.63106889e-01 8.35224241e-02 1.08852029e+00
1.06569958e+00 1.22002029e+00 -4.69442666e-01 -4.98311490e-01
1.10447550e+00 -1.44646251e+00 -1.51578760e+00 5.27917221e-02
-4.56147715e-02 -1.12404394e+00 1.46413052e+00 6.66586220e-01
-1.61329901e+00 -1.11711180e+00 -1.07430601e+00 -2.30375469e-01
2.80824900e-01 3.36593866e-01 2.92677402e-01 4.70822841e-01
-9.40332234e-01 9.97701406e-01 -4.57801878e-01 2.10380882e-01
-3.00012305e-02 2.16790289e-01 9.13898721e-02 6.43207133e-01
-1.23310471e+00 5.10551751e-01 3.88453811e-01 1.25614911e-01
-9.36868310e-01 -7.03064919e-01 -5.12926817e-01 2.57685065e-01
8.75576511e-02 -5.70630670e-01 2.05072284e+00 -8.05165589e-01
-2.28572297e+00 1.73806757e-01 -9.32641402e-02 -2.53818661e-01
-1.00776091e-01 -3.16912562e-01 -8.41069818e-01 2.14439817e-02
-3.72135073e-01 6.16338313e-01 1.02615631e+00 -1.02123177e+00
-6.20602071e-01 1.79189444e-01 -3.93386751e-01 3.15090477e-01
-5.07454455e-01 1.80019841e-01 -1.39994472e-01 -9.66057420e-01
1.33082047e-01 -7.79553592e-01 8.67142156e-02 -2.23922983e-01
-4.86508369e-01 -7.28562623e-02 8.49639356e-01 -9.26352739e-01
1.71392262e+00 -2.23360586e+00 2.16635957e-01 -2.23548636e-01
-1.51198909e-01 7.09564865e-01 -1.61899909e-01 4.95736182e-01
8.38516839e-03 -8.99900272e-02 -1.16429597e-01 -2.36744970e-01
1.10724106e-01 1.86645865e-01 -5.75190246e-01 -2.30047002e-01
6.72353432e-02 8.43779027e-01 -8.14770520e-01 -2.56378412e-01
1.85620457e-01 4.67705369e-01 -8.28538120e-01 8.99777293e-01
-3.61822873e-01 4.90636647e-01 1.53125793e-01 5.88113785e-01
4.40153480e-02 3.28858495e-01 7.56529570e-02 -1.12961322e-01
-2.14892298e-01 8.54832292e-01 -1.21704161e+00 1.52065849e+00
-6.85010612e-01 8.13914657e-01 4.39548761e-01 -2.35870689e-01
1.30547631e+00 9.69705582e-01 -4.88711707e-03 -6.61051273e-01
5.16160429e-02 5.39139152e-01 5.39741576e-01 -7.89549649e-01
6.07720017e-01 -7.09894896e-01 2.98417836e-01 3.81721824e-01
1.56074449e-01 -9.12400484e-01 4.00303788e-02 -6.17131650e-01
5.99151015e-01 -1.45263299e-02 1.36454359e-01 2.38388553e-01
3.22186619e-01 -2.15866044e-01 4.99271244e-01 2.21664593e-01
2.03776687e-01 8.80422354e-01 1.22817598e-01 -1.21551141e-01
-1.33143532e+00 -1.07177234e+00 9.51675773e-02 1.14035785e+00
-5.11419058e-01 -5.31991839e-01 -8.72149289e-01 1.93708017e-01
-3.00923258e-01 9.02821124e-01 1.95836782e-01 -1.43809214e-01
-9.83795345e-01 7.67139047e-02 9.61791992e-01 5.80539107e-01
1.94260776e-01 -1.86165917e+00 -2.46674284e-01 6.54546380e-01
-7.28205740e-02 -6.18432224e-01 -1.10130668e+00 2.02154770e-01
-8.04307222e-01 -5.33847034e-01 -1.04303455e+00 -1.10069907e+00
-1.22675113e-01 4.05846909e-02 8.65913630e-01 5.99333569e-02
-1.19480528e-01 -2.91694403e-01 -3.24676722e-01 -6.26670539e-01
-8.08294892e-01 -1.51017075e-02 5.36832869e-01 -2.35428140e-01
-7.90816993e-02 -8.67976546e-01 -3.99945587e-01 1.31604150e-01
-9.43738520e-01 3.08808297e-01 3.73592824e-01 8.11386287e-01
5.43894649e-01 2.04566777e-01 9.87351298e-01 -2.92212188e-01
1.35959637e+00 6.07094504e-02 -4.00703490e-01 -1.56951606e-01
-4.56043392e-01 -2.12315738e-01 1.30434752e+00 -7.08082378e-01
-7.46731579e-01 -8.51336867e-03 -7.33122826e-01 -8.21046948e-01
-7.05088675e-02 5.84781528e-01 -1.66137874e-01 3.69765669e-01
8.23361456e-01 3.73969465e-01 1.73128381e-01 -8.00195873e-01
6.13419116e-01 1.37935591e+00 9.76200640e-01 -1.62627786e-01
7.62138903e-01 -4.03759778e-01 -3.27258319e-01 -1.13933611e+00
-4.92497861e-01 -1.79456189e-01 -3.76114041e-01 -3.01862776e-01
5.78470647e-01 -6.54336870e-01 -8.00721705e-01 4.56017613e-01
-1.23205006e+00 -4.53496754e-01 -5.93058586e-01 7.60707676e-01
-8.23157370e-01 1.72622964e-01 -1.08667219e+00 -1.08082843e+00
-6.55896664e-01 -1.06164074e+00 5.21459937e-01 4.74130899e-01
-5.76560080e-01 -4.48192537e-01 1.49155572e-01 2.12538972e-01
6.96427107e-01 -2.42556512e-01 9.01838362e-01 -6.52952641e-02
-2.74882466e-01 -1.06358685e-01 3.86792898e-01 8.70083511e-01
4.14933681e-01 1.54369131e-01 -1.07114458e+00 2.98563503e-02
7.54319876e-02 -2.71283090e-01 3.98131132e-01 5.55099487e-01
1.12646544e+00 -6.63538456e-01 5.80388546e-01 6.44416690e-01
5.92473328e-01 7.12948978e-01 6.82616234e-01 -3.54283839e-01
6.77029908e-01 6.16974115e-01 3.20326179e-01 2.82826602e-01
-5.44523634e-02 7.48690546e-01 8.75522494e-02 -1.12626113e-01
-8.65268052e-01 -8.64379704e-01 5.00179946e-01 1.76860309e+00
-2.00825915e-01 -6.22807406e-02 -4.18409914e-01 6.00811601e-01
-1.42862248e+00 -1.20550740e+00 -8.64220560e-02 2.16010523e+00
1.27310848e+00 -7.01673552e-02 4.41528052e-01 5.11479914e-01
7.15778530e-01 3.94943506e-01 -5.02645016e-01 -9.94597256e-01
2.03964725e-01 6.73377931e-01 -4.21418808e-02 4.47645575e-01
-5.40894687e-01 1.26182711e+00 7.69962406e+00 9.32188392e-01
-1.50993586e+00 -3.45701098e-01 7.30519241e-04 -4.63205338e-01
-3.00396860e-01 -3.41427624e-01 -5.17808497e-01 2.49329165e-01
1.16539729e+00 -7.90827051e-02 1.33762240e+00 7.77212977e-01
7.54389703e-01 7.28791475e-01 -1.08442187e+00 1.05609429e+00
-1.32113948e-01 -1.47600913e+00 3.01236883e-02 -3.60443890e-01
7.31904387e-01 -4.58723307e-01 1.10195041e-01 4.07098442e-01
-1.95978850e-01 -1.53002453e+00 1.00358057e+00 4.02508825e-01
1.11256886e+00 -7.48819172e-01 2.70448148e-01 5.16801119e-01
-1.19779003e+00 -8.41351375e-02 -4.01640013e-02 -3.96600336e-01
5.41144788e-01 1.08455800e-01 -1.18834805e+00 6.91422075e-02
3.61406177e-01 4.27820653e-01 2.12511465e-01 1.15365672e+00
-5.70162952e-01 1.19939458e+00 -8.41052458e-02 -3.70003432e-01
-4.86443415e-02 8.37770775e-02 5.84387779e-01 1.10399926e+00
5.20860314e-01 2.26468444e-01 -1.58938840e-01 1.12065303e+00
-2.38869518e-01 2.89978683e-01 -4.56125259e-01 -8.15156341e-01
4.82106477e-01 1.03162265e+00 -3.09798145e-03 -9.56813842e-02
1.77774392e-03 6.95431232e-01 -2.81280339e-01 3.73773038e-01
-4.97542083e-01 -7.16512978e-01 6.25495851e-01 2.43641749e-01
2.25316003e-01 -5.21241963e-01 -4.46133792e-01 -5.55227816e-01
-5.74086756e-02 -1.34890938e+00 -4.12051529e-01 -1.13560939e+00
-1.02991390e+00 7.64709115e-01 -5.96368670e-01 -1.32381058e+00
-7.37086356e-01 -5.81153154e-01 -1.06825840e+00 1.04788435e+00
-1.15773642e+00 -7.28893161e-01 2.02033117e-01 2.18824521e-01
9.44815040e-01 -4.80529547e-01 9.71407831e-01 3.69364709e-01
-1.82370797e-01 4.90542412e-01 -1.44213051e-01 1.60370991e-01
5.31184256e-01 -1.12553966e+00 4.64118212e-01 5.12964725e-01
6.40574813e-01 4.88658935e-01 9.00277257e-01 -2.52603233e-01
-1.22207940e+00 -6.52291059e-01 1.03120446e+00 2.04258293e-01
3.51612449e-01 -3.03453654e-01 -9.39801931e-01 5.06445207e-02
4.04703677e-01 -3.69388908e-01 8.26665342e-01 7.19410018e-04
5.62220588e-02 -1.97060928e-01 -6.69837236e-01 8.64301264e-01
8.10913801e-01 -7.72778988e-01 -9.06192660e-01 5.77390455e-02
1.00412416e+00 -5.83941400e-01 -6.99103236e-01 3.55065674e-01
7.35513926e-01 -8.77022862e-01 7.53476739e-01 -6.06881380e-01
7.51799107e-01 -5.39614797e-01 -5.49569093e-02 -1.62775660e+00
-2.75960803e-01 -1.13563931e+00 -2.78759956e-01 1.08455074e+00
6.48765683e-01 9.71430913e-02 4.50244278e-01 3.85433026e-02
-5.91598213e-01 -5.43378532e-01 -6.23830140e-01 -7.74803579e-01
2.98831705e-02 -5.12495577e-01 8.58416915e-01 6.07801378e-01
2.18779922e-01 6.04981422e-01 -9.94969964e-01 -9.44586173e-02
-4.80536893e-02 1.82933688e-01 8.87567520e-01 -8.89060080e-01
-7.78294444e-01 -5.66856563e-01 3.06030035e-01 -1.34301448e+00
-1.08169638e-01 -8.14845979e-01 4.50190187e-01 -1.52086997e+00
-4.53475267e-01 -4.35100384e-02 -1.57282248e-01 4.89541113e-01
3.75993513e-02 1.12526633e-01 6.59821272e-01 6.62928522e-02
1.06678665e-01 8.22786272e-01 1.91945648e+00 -9.36674885e-03
-7.20991492e-01 3.97425979e-01 -4.15299207e-01 7.76030183e-01
8.94916773e-01 -2.81639367e-01 -2.95252472e-01 -2.13664159e-01
-1.50500625e-01 7.59078085e-01 3.57409418e-02 -1.18906116e+00
3.36103082e-01 -9.69416723e-02 1.59158796e-01 -7.00904727e-01
8.29740405e-01 -4.70145762e-01 1.30681351e-01 4.54715699e-01
-6.97260022e-01 -1.12213738e-01 3.11660111e-01 -1.64122935e-02
-5.69615841e-01 -4.27553266e-01 6.99515998e-01 -5.87923229e-02
-2.63693273e-01 6.07940555e-02 -4.04094994e-01 -2.49020323e-01
3.71340841e-01 -5.67211844e-02 1.66106030e-01 -9.09240544e-01
-7.43513584e-01 -1.81914836e-01 -7.98877105e-02 6.90090716e-01
8.01817119e-01 -1.68769789e+00 -7.41219103e-01 4.15319771e-01
-3.08356494e-01 -1.46583870e-01 1.96425185e-01 2.02423394e-01
-6.95495725e-01 5.10128200e-01 -2.36814275e-01 -1.71102837e-01
-1.33885133e+00 4.47726011e-01 4.47337240e-01 2.90865541e-01
-5.96407235e-01 7.14744687e-01 -1.99584037e-01 -6.50445938e-01
4.89007145e-01 -3.15218389e-01 -2.25071713e-01 -4.04971778e-01
6.86706662e-01 2.41783381e-01 -2.75543958e-01 -4.39993709e-01
2.32403442e-01 3.74314189e-01 4.45528030e-01 -3.87154728e-01
1.08933425e+00 2.19752759e-01 9.21917856e-02 7.06762910e-01
8.82405400e-01 5.41602969e-01 -1.13429773e+00 1.41454339e-01
-2.02354014e-01 -4.17591631e-01 3.01513165e-01 -8.73930931e-01
-9.89236176e-01 1.10713756e+00 2.75910079e-01 3.81189972e-01
1.15211844e+00 -2.80971050e-01 1.45999527e+00 2.84547806e-01
-2.50517298e-02 -1.38246477e+00 1.94696099e-01 6.62109911e-01
1.23102450e+00 -6.99858129e-01 -5.77142060e-01 -1.68564752e-01
-7.97793090e-01 1.36628950e+00 4.22179818e-01 -2.10528374e-01
4.69150066e-01 4.03919935e-01 4.47440445e-01 2.35707596e-01
-7.88920760e-01 -1.90336108e-01 5.26935995e-01 4.50311661e-01
9.22727644e-01 3.32599163e-01 -4.22671318e-01 7.78193831e-01
-1.09636521e+00 3.40535849e-01 3.05533946e-01 1.10209063e-01
-8.30225825e-01 -1.35859370e+00 -5.86771667e-01 2.36864418e-01
-6.67843640e-01 -4.02417183e-01 -4.96886104e-01 2.72767276e-01
2.16999263e-01 1.16855407e+00 -5.72881773e-02 -7.80397892e-01
5.96751094e-01 2.08755910e-01 3.15970033e-01 -7.23348737e-01
-8.30100715e-01 5.64422905e-01 2.31385082e-01 -4.53425050e-02
2.03018486e-01 -8.16120133e-02 -1.56438470e+00 -2.08509982e-01
-3.67475510e-01 2.39550531e-01 7.65519321e-01 5.09244800e-01
1.87281340e-01 1.09340823e+00 1.03292906e+00 -8.71121943e-01
-7.85533369e-01 -1.25807440e+00 -1.02346575e+00 1.17246144e-01
4.32673186e-01 -4.94923443e-02 -2.53603280e-01 1.38037130e-01] | [15.53236198425293, 6.182980537414551] |
7eae277a-df5a-4028-8127-0722c297428a | unsupervised-multilingual-sentence-embeddings-1 | 2105.10419 | null | https://arxiv.org/abs/2105.10419v1 | https://arxiv.org/pdf/2105.10419v1.pdf | Unsupervised Multilingual Sentence Embeddings for Parallel Corpus Mining | Existing models of multilingual sentence embeddings require large parallel data resources which are not available for low-resource languages. We propose a novel unsupervised method to derive multilingual sentence embeddings relying only on monolingual data. We first produce a synthetic parallel corpus using unsupervised machine translation, and use it to fine-tune a pretrained cross-lingual masked language model (XLM) to derive the multilingual sentence representations. The quality of the representations is evaluated on two parallel corpus mining tasks with improvements of up to 22 F1 points over vanilla XLM. In addition, we observe that a single synthetic bilingual corpus is able to improve results for other language pairs. | ['Ondřej Bojar', 'Eneko Agirre', 'Gorka Labaka', 'Mikel Artetxe', 'Ivana Kvapilikova'] | 2021-05-21 | unsupervised-multilingual-sentence-embeddings | https://aclanthology.org/2020.acl-srw.34 | https://aclanthology.org/2020.acl-srw.34.pdf | acl-2020-6 | ['unsupervised-machine-translation', 'parallel-corpus-mining'] | ['natural-language-processing', 'natural-language-processing'] | [-1.73467025e-01 -4.57308814e-02 -3.79743695e-01 -4.35711205e-01
-1.35426795e+00 -7.39110112e-01 8.54482353e-01 4.79178876e-01
-9.53307092e-01 9.59406137e-01 4.93610352e-01 -5.83973944e-01
5.24079382e-01 -5.33625543e-01 -8.49290252e-01 -1.48315653e-01
6.57992810e-02 6.26653492e-01 -2.76837349e-01 -5.91694653e-01
-8.73827413e-02 5.44588119e-02 -8.54759872e-01 3.14424962e-01
1.23626029e+00 6.79497793e-02 4.43988413e-01 5.53048551e-01
-3.46657366e-01 1.91718608e-01 -4.78792995e-01 -7.08151340e-01
4.68758732e-01 -4.10264969e-01 -7.65791595e-01 -1.68874100e-01
4.09183174e-01 5.47985621e-02 -1.98966831e-01 1.13499081e+00
4.18783665e-01 -2.09282383e-01 5.45547843e-01 -6.47624552e-01
-9.92429793e-01 1.11212623e+00 -4.12373662e-01 4.78715122e-01
3.39651048e-01 1.59836132e-02 1.21928215e+00 -1.30962241e+00
9.32520270e-01 1.15469182e+00 3.95278335e-01 2.07207784e-01
-1.56418312e+00 -4.71393645e-01 3.75516377e-02 2.58781910e-02
-1.43476677e+00 -5.15114486e-01 6.37592733e-01 -2.61879534e-01
1.57336247e+00 -8.93848464e-02 3.15782875e-01 1.29783678e+00
4.47241127e-01 6.42947018e-01 1.32732606e+00 -9.23001885e-01
-3.29141647e-01 5.46185613e-01 1.68948278e-01 5.87345541e-01
2.25848317e-01 1.32667050e-02 -4.46134180e-01 -5.42025268e-02
3.75241786e-01 -4.52243239e-01 5.74868731e-02 -2.95650139e-02
-1.59203160e+00 1.07210481e+00 1.98228851e-01 7.84319401e-01
-2.37480387e-01 -4.06923294e-01 6.64615691e-01 8.81351292e-01
7.84833550e-01 7.58227527e-01 -6.81588531e-01 4.63798176e-03
-7.00780749e-01 -2.02288121e-01 5.88909626e-01 9.83785987e-01
1.02973127e+00 3.13250870e-02 9.76821408e-02 1.21722770e+00
-1.01396315e-01 8.56152892e-01 8.35868061e-01 -1.15267448e-01
9.95618343e-01 4.57732290e-01 -2.74745554e-01 -6.69813871e-01
-2.01092526e-01 -2.01263398e-01 -5.15724719e-01 -4.76189107e-01
1.24570489e-01 -2.08585307e-01 -3.14750373e-01 1.61376512e+00
7.31291175e-02 -3.57684255e-01 6.72863066e-01 5.62421203e-01
5.09786665e-01 7.54265070e-01 -7.64469132e-02 -1.23041026e-01
1.26766050e+00 -1.25504959e+00 -5.51358044e-01 -3.39972019e-01
1.08392429e+00 -1.17127490e+00 1.41366208e+00 -2.99493875e-02
-9.71224904e-01 -7.78352141e-01 -1.25753736e+00 -3.73162180e-01
-5.26888728e-01 3.47492427e-01 5.38635373e-01 5.66674292e-01
-9.23968613e-01 3.61881346e-01 -6.89467967e-01 -4.47064906e-01
-4.30983379e-02 3.08233380e-01 -8.60624373e-01 -3.07033330e-01
-1.41751277e+00 1.31715262e+00 4.51385200e-01 -9.53478590e-02
-6.15426779e-01 -6.34881556e-01 -1.29028726e+00 -3.78907293e-01
-2.38964781e-01 -4.17197645e-01 8.35298836e-01 -9.77877259e-01
-1.38395405e+00 1.04874527e+00 -2.39817455e-01 -4.92499560e-01
1.64061040e-01 -1.97748140e-01 -7.69446373e-01 -4.91027199e-02
4.77116048e-01 7.57418036e-01 3.71800750e-01 -8.48088086e-01
-4.78133231e-01 -1.02682829e-01 -5.10659181e-02 2.11648285e-01
-7.60919809e-01 3.67353380e-01 -2.25583553e-01 -7.38235354e-01
-3.26746404e-01 -8.08532059e-01 -4.99112129e-01 -7.54975319e-01
-3.18004817e-01 -1.16048671e-01 1.72236472e-01 -9.48986351e-01
9.76478994e-01 -1.83264935e+00 2.14840606e-01 -2.30963126e-01
-3.35578680e-01 1.51035115e-01 -8.01424265e-01 6.72391832e-01
-1.41542926e-01 1.37543097e-01 -2.30857357e-01 -6.79206491e-01
-8.78796205e-02 4.62710798e-01 -2.76600957e-01 3.90989751e-01
6.76822305e-01 9.79093373e-01 -1.00721133e+00 -5.32793880e-01
9.97349024e-02 4.39093411e-01 -4.90592718e-01 1.23696692e-01
-6.83655438e-04 5.23263216e-01 1.28317580e-01 4.52421784e-01
5.90091228e-01 2.85369247e-01 7.77769268e-01 4.07519676e-02
-2.52920330e-01 9.21393275e-01 -5.63286245e-01 2.15614462e+00
-1.15918422e+00 6.00618362e-01 -5.39677024e-01 -7.20444977e-01
1.17496490e+00 3.14998627e-01 7.78864771e-02 -8.82220030e-01
5.92731312e-02 6.43415034e-01 1.42134324e-01 -4.17356163e-01
8.50047529e-01 -3.39906186e-01 -6.17290378e-01 6.81150377e-01
5.20813227e-01 4.99851592e-02 5.10005951e-01 5.73628359e-02
7.36125290e-01 -1.10221747e-02 3.21725190e-01 -7.03761518e-01
6.59672678e-01 1.66861892e-01 4.81390595e-01 1.50373876e-01
2.02317268e-01 4.22695428e-01 2.48651296e-01 -5.33883929e-01
-1.42020392e+00 -1.26425564e+00 -3.09072495e-01 1.10375285e+00
-2.82559901e-01 -4.65205312e-01 -4.90927905e-01 -1.12339377e+00
-1.12759173e-01 8.06602538e-01 -3.74065995e-01 1.36276931e-01
-8.36830437e-01 -1.03581977e+00 6.67969882e-01 4.63957697e-01
-1.74246192e-01 -9.57022548e-01 2.65753776e-01 3.21178824e-01
-4.02191013e-01 -1.41402864e+00 -6.88996255e-01 2.49824062e-01
-7.86669612e-01 -8.64842236e-01 -3.31415623e-01 -1.24435949e+00
8.13650608e-01 -4.57970463e-02 1.39530480e+00 -4.23232287e-01
-1.71223166e-03 -2.81254292e-01 -3.31405252e-01 -1.17593318e-01
-9.11476195e-01 5.31299233e-01 6.04484320e-01 -2.21747145e-01
1.01017988e+00 -5.49714625e-01 8.64708871e-02 -2.34714985e-01
-7.96199381e-01 -1.91416606e-01 7.09073842e-01 1.05189157e+00
6.69605315e-01 -4.59454954e-01 7.87912905e-01 -9.08801854e-01
8.73410165e-01 -4.92598057e-01 -5.07936239e-01 2.79150993e-01
-5.79458952e-01 4.53412861e-01 1.01333022e+00 -6.24388695e-01
-7.10154772e-01 -7.61704445e-02 -2.34283715e-01 -9.98797566e-02
-1.03251614e-01 5.35787642e-01 -8.08052942e-02 3.19031388e-01
6.86095595e-01 2.50338495e-01 -2.12929204e-01 -6.85475647e-01
6.85178041e-01 9.25338149e-01 3.34895194e-01 -6.74685597e-01
8.05972695e-01 -7.31371623e-03 -5.41089952e-01 -8.93343270e-01
-7.02784359e-01 -2.72346795e-01 -1.25968599e+00 4.48738158e-01
8.75391662e-01 -1.26211333e+00 1.23544969e-01 -1.92337662e-01
-1.37103117e+00 -1.17580451e-01 -2.07412153e-01 9.20409381e-01
-1.56878129e-01 3.77249181e-01 -8.57460916e-01 -2.26890549e-01
-4.65233982e-01 -1.33346629e+00 9.07156229e-01 -4.08524185e-01
-3.71725082e-01 -1.43157554e+00 6.67082965e-01 3.27517211e-01
2.82720983e-01 -1.31327331e-01 1.11515522e+00 -8.64067912e-01
-1.04782932e-01 -1.06699750e-01 2.27605924e-02 6.54238284e-01
3.38424385e-01 -2.51014739e-01 -7.69399822e-01 -5.84414184e-01
-9.85831469e-02 -4.45626557e-01 6.84466958e-01 -2.32805654e-01
2.90587515e-01 -2.48690724e-01 -5.41818738e-02 3.84956241e-01
1.50816286e+00 -2.80830562e-01 1.92518979e-01 3.69508862e-01
7.40170598e-01 5.65206647e-01 4.10921931e-01 -1.39795393e-01
7.19471514e-01 5.82447946e-01 -3.60385358e-01 -1.47897214e-01
-6.72148541e-02 -5.56830645e-01 8.85113001e-01 2.22711182e+00
3.68076473e-01 3.71829957e-01 -9.09240365e-01 1.11496937e+00
-1.36557543e+00 -3.97223771e-01 -1.43340945e-01 2.09554911e+00
1.25811839e+00 7.76579678e-02 -9.29409638e-02 -1.48817629e-01
3.76315415e-01 1.07501097e-01 4.73342612e-02 -8.99704576e-01
-4.81985807e-01 6.78772092e-01 5.73318064e-01 8.72072697e-01
-1.04470158e+00 1.49534643e+00 6.69427919e+00 5.10627806e-01
-1.18852425e+00 6.62502527e-01 2.06771940e-01 -7.61103164e-03
-7.80519545e-01 1.80719376e-01 -9.42223430e-01 2.45576322e-01
1.40599990e+00 -3.10191810e-01 3.41329753e-01 5.84644914e-01
2.51413025e-02 2.62575984e-01 -1.26981664e+00 7.16101527e-01
3.34963053e-01 -1.16683018e+00 2.73326784e-01 -7.19337612e-02
1.08020258e+00 5.34161091e-01 -1.57078654e-01 5.89965105e-01
5.38922548e-01 -9.78164077e-01 3.45685810e-01 1.21677227e-01
1.05178952e+00 -1.04979205e+00 8.22908223e-01 4.00072992e-01
-1.09455335e+00 3.55061352e-01 -7.29903579e-01 -1.58378437e-01
2.43952125e-01 4.94637907e-01 -9.64711010e-01 8.95961344e-01
2.57937551e-01 7.20268726e-01 -7.95886755e-01 2.55989432e-01
-4.97397184e-01 4.48037148e-01 -1.60734072e-01 3.83185083e-03
3.22962493e-01 -3.42425764e-01 4.44306225e-01 1.54899526e+00
1.79041848e-01 -8.58756661e-01 3.83556813e-01 5.91055751e-01
-3.31004709e-01 8.33077431e-01 -1.01667368e+00 -4.01474923e-01
3.51548970e-01 1.23675716e+00 -1.94392443e-01 -3.32915455e-01
-7.35650420e-01 1.22449028e+00 8.21589649e-01 1.60880268e-01
-4.65538710e-01 -3.50850463e-01 7.89796650e-01 -1.94444194e-01
1.16941079e-01 -6.60444558e-01 -1.43980995e-01 -1.72140896e+00
8.21421891e-02 -1.15155482e+00 8.75888020e-02 -1.51550069e-01
-1.65494728e+00 1.20907366e+00 -3.01730961e-01 -1.17186856e+00
-6.26550913e-01 -5.74750245e-01 -3.18874270e-01 1.37387824e+00
-1.71628940e+00 -1.37888837e+00 6.39846802e-01 3.40158969e-01
7.36484230e-01 -6.22333527e-01 1.40200067e+00 6.11843467e-01
-5.29857993e-01 1.07398534e+00 2.61174619e-01 4.12587106e-01
9.47854340e-01 -1.21479893e+00 9.21795487e-01 1.04137492e+00
7.85830379e-01 8.70189905e-01 3.97658139e-01 -4.82943594e-01
-1.44863844e+00 -1.25690711e+00 1.84069061e+00 -6.26597345e-01
1.05913806e+00 -8.04093599e-01 -7.53544271e-01 8.70114625e-01
7.44586587e-01 -1.29196256e-01 1.10305142e+00 5.47861755e-01
-6.90713823e-01 -1.43873654e-02 -8.80463302e-01 6.74754560e-01
5.97707629e-01 -9.91188705e-01 -9.01759565e-01 5.10550797e-01
9.98932600e-01 2.37749014e-02 -1.19592190e+00 2.58031756e-01
2.56914079e-01 -3.75600100e-01 7.15295553e-01 -9.23617721e-01
5.69947183e-01 -1.59190983e-01 -5.31030536e-01 -1.69511020e+00
3.77114601e-02 -2.84421861e-01 4.58083242e-01 1.24968743e+00
1.03052819e+00 -7.07354903e-01 2.32758820e-01 -3.23467731e-01
-2.57973094e-02 -6.06100142e-01 -8.88601601e-01 -1.09736681e+00
7.52661049e-01 -4.41888630e-01 5.59495449e-01 1.33571124e+00
3.73569876e-01 9.68140364e-01 -2.40157887e-01 1.83979958e-01
4.91392434e-01 2.51070023e-01 7.33332038e-01 -7.95770049e-01
-4.03954059e-01 -1.84389651e-01 -2.55572379e-01 -9.40008759e-01
6.64061844e-01 -1.76445055e+00 -9.12847966e-02 -1.26338148e+00
2.51535416e-01 -4.23543721e-01 -4.12215054e-01 1.73814699e-01
-2.93875962e-01 4.18307036e-01 4.34498824e-02 3.92710157e-02
-3.00153762e-01 6.83063388e-01 8.81067693e-01 -2.88575262e-01
-6.16083704e-02 -6.27423644e-01 -5.80411851e-01 4.28047419e-01
9.18985903e-01 -7.27093339e-01 -1.27497569e-01 -1.05729914e+00
1.45228570e-02 -1.31799817e-01 -3.06059539e-01 -7.06957817e-01
-2.44157836e-01 1.01827674e-01 2.01541677e-01 -3.01928431e-01
1.36208460e-01 -4.06561911e-01 -3.54511231e-01 3.13478410e-01
-3.28088641e-01 7.13575304e-01 3.25275987e-01 8.53495523e-02
-5.60075581e-01 -3.07253718e-01 4.70533401e-01 -3.72340195e-02
-3.35030645e-01 5.75425662e-02 -3.35649431e-01 1.15912281e-01
6.89290762e-01 5.43358028e-01 -1.64562032e-01 2.31791615e-01
-3.84910822e-01 6.02190755e-02 6.05374932e-01 9.71942902e-01
2.80327082e-01 -1.73749053e+00 -1.08879220e+00 6.03624940e-01
3.75307709e-01 -6.20416641e-01 -1.80189833e-01 7.32399821e-01
-5.12233138e-01 6.52385950e-01 -3.05572152e-01 -4.64287579e-01
-9.85547543e-01 7.42258430e-01 -1.39165707e-02 -5.61467469e-01
-4.13410723e-01 4.95267123e-01 -2.61677861e-01 -1.26635671e+00
-3.29820603e-01 -3.50795597e-01 -5.86755835e-02 -5.15952222e-02
4.16573584e-01 -1.46730870e-01 2.36955270e-01 -1.04550719e+00
-3.91956985e-01 4.25609738e-01 -2.71818757e-01 -4.24295455e-01
1.39676452e+00 -1.10576831e-01 -3.70304197e-01 7.78696299e-01
1.60215056e+00 5.88934541e-01 -4.34337229e-01 -4.12595212e-01
3.37772161e-01 -3.18831384e-01 -3.65872473e-01 -3.16603631e-01
-6.21428311e-01 9.32647109e-01 2.08678886e-01 -3.02794367e-01
7.22375035e-01 -1.24304131e-01 1.04995024e+00 5.26229560e-01
5.00738382e-01 -1.04946065e+00 -2.39294082e-01 9.07308459e-01
7.27377772e-01 -1.49150944e+00 -2.91517794e-01 -1.03061751e-01
-6.57967925e-01 1.10060167e+00 2.94520408e-01 -3.10098678e-01
6.22276425e-01 5.00465572e-01 5.52949011e-01 1.51562303e-01
-8.13892305e-01 -1.00991346e-01 2.86783844e-01 3.70879441e-01
8.70247483e-01 3.69775295e-01 -7.79677451e-01 5.35865843e-01
-5.63763618e-01 -4.75914836e-01 4.27956223e-01 5.21140635e-01
-2.60281432e-02 -1.94133782e+00 -8.05876181e-02 1.70464665e-01
-5.50896466e-01 -6.71760976e-01 -2.36634374e-01 7.12422431e-01
1.48373365e-01 8.94422054e-01 1.36999384e-01 -5.59307933e-01
2.49021322e-01 2.52775729e-01 6.01292312e-01 -8.95538688e-01
-7.15062559e-01 -7.00982437e-02 3.39409560e-01 -1.83991656e-01
-2.75264204e-01 -6.75321817e-01 -8.91269922e-01 -9.95322615e-02
-8.66705477e-02 2.93839157e-01 7.57117033e-01 9.19882119e-01
2.09909484e-01 2.92419493e-01 8.09739292e-01 -6.24958813e-01
-3.79033655e-01 -1.34836364e+00 -3.26950476e-02 3.26991558e-01
-3.58428173e-02 -1.02868102e-01 -1.42543450e-01 -1.79462824e-02] | [11.141098022460938, 10.04642391204834] |
c275a4df-a0d5-4530-ab03-6e104ce9eb63 | facescape-3d-facial-dataset-and-benchmark-for | 2111.01082 | null | https://arxiv.org/abs/2111.01082v1 | https://arxiv.org/pdf/2111.01082v1.pdf | FaceScape: 3D Facial Dataset and Benchmark for Single-View 3D Face Reconstruction | In this paper, we present a large-scale detailed 3D face dataset, FaceScape, and the corresponding benchmark to evaluate single-view facial 3D reconstruction. By training on FaceScape data, a novel algorithm is proposed to predict elaborate riggable 3D face models from a single image input. FaceScape dataset provides 18,760 textured 3D faces, captured from 938 subjects and each with 20 specific expressions. The 3D models contain the pore-level facial geometry that is also processed to be topologically uniformed. These fine 3D facial models can be represented as a 3D morphable model for rough shapes and displacement maps for detailed geometry. Taking advantage of the large-scale and high-accuracy dataset, a novel algorithm is further proposed to learn the expression-specific dynamic details using a deep neural network. The learned relationship serves as the foundation of our 3D face prediction system from a single image input. Different than the previous methods, our predicted 3D models are riggable with highly detailed geometry under different expressions. We also use FaceScape data to generate the in-the-wild and in-the-lab benchmark to evaluate recent methods of single-view face reconstruction. The accuracy is reported and analyzed on the dimensions of camera pose and focal length, which provides a faithful and comprehensive evaluation and reveals new challenges. The unprecedented dataset, benchmark, and code have been released to the public for research purpose. | ['Xun Cao', 'Ruigang Yang', 'Qiu Shen', 'Mingkai Huang', 'Yanru Wang', 'Yidi Zhang', 'Longwei Guo', 'Haotian Yang', 'Hao Zhu'] | 2021-11-01 | null | null | null | null | ['3d-face-reconstruction', 'face-reconstruction'] | ['computer-vision', 'computer-vision'] | [-2.38357782e-01 1.89551592e-01 1.82505608e-01 -9.01062608e-01
-4.55651760e-01 -3.53723466e-01 3.89730126e-01 -8.39950442e-01
3.85241956e-01 2.67138839e-01 5.70964031e-02 3.97049308e-01
2.13605210e-01 -6.42617643e-01 -8.70062053e-01 -4.66932029e-01
-8.10296908e-02 7.98341274e-01 -2.65928835e-01 -4.05294299e-01
-1.94255300e-02 1.15492129e+00 -1.84836245e+00 2.47863173e-01
5.43026403e-02 1.39973652e+00 -3.49008203e-01 3.39540958e-01
1.27324313e-01 -2.27678604e-02 -7.45897964e-02 -7.17014730e-01
6.50058448e-01 -1.26701575e-02 -4.94955003e-01 3.92083257e-01
1.01968956e+00 -6.02826476e-01 -1.65438294e-01 6.97339833e-01
7.74612069e-01 -2.52203763e-01 5.67428172e-01 -1.14151418e+00
-5.13803303e-01 -5.09157360e-01 -7.66239226e-01 -3.46776545e-01
6.88384771e-01 3.04388348e-02 4.00510579e-01 -1.42518878e+00
1.04888034e+00 1.74927461e+00 8.19410443e-01 1.12511599e+00
-1.10591877e+00 -8.71975064e-01 -2.98908586e-03 -3.03615987e-01
-1.44940209e+00 -9.41407502e-01 1.09763670e+00 -5.97137511e-01
6.70591533e-01 1.52233720e-01 9.56385612e-01 1.26963294e+00
3.40658456e-01 2.41867423e-01 1.36352861e+00 -7.05056936e-02
4.79205810e-02 -5.97598553e-02 -4.92393076e-01 1.21902049e+00
-1.08267538e-01 3.82614851e-01 -6.07929111e-01 -9.18286666e-02
1.07871830e+00 8.07941034e-02 -1.18546277e-01 -3.77436727e-01
-6.44403398e-01 4.34540212e-01 2.03403488e-01 -2.67396003e-01
-1.06598347e-01 -7.84537718e-02 1.86875090e-01 1.88555539e-01
9.83691871e-01 -1.51191965e-01 -6.52721286e-01 -6.56412020e-02
-7.37943470e-01 5.31570971e-01 6.22243702e-01 1.06024110e+00
1.08889806e+00 1.34483650e-01 2.23332405e-01 8.00642729e-01
4.72069621e-01 8.53574932e-01 3.61742973e-02 -1.32290840e+00
8.48703682e-02 8.66608500e-01 -1.55770525e-01 -1.31233001e+00
-5.23473203e-01 7.18321279e-02 -9.56605554e-01 4.42262173e-01
1.02440767e-01 1.08331241e-01 -8.69105279e-01 1.61800408e+00
8.24244618e-01 -1.84567682e-02 -4.00048912e-01 1.02697945e+00
1.15955794e+00 4.70224738e-01 -4.74644214e-01 -1.73623413e-01
1.24246383e+00 -5.28416276e-01 -5.26559770e-01 9.17477757e-02
2.24796712e-01 -6.01401150e-01 9.55977976e-01 3.85562003e-01
-1.22799683e+00 -5.00380218e-01 -5.52958429e-01 -1.50485367e-01
-2.05293730e-01 2.00720355e-02 5.22901773e-01 6.94125116e-01
-1.28687942e+00 5.64024568e-01 -7.20556438e-01 -3.79087299e-01
1.00526774e+00 4.23056215e-01 -1.06073058e+00 -7.89253600e-03
-6.43910348e-01 7.17526913e-01 -3.86640400e-01 2.77178526e-01
-9.23395336e-01 -1.00226021e+00 -9.20403063e-01 -3.43541056e-01
-9.41371918e-02 -7.14182079e-01 9.06350493e-01 -9.06527340e-01
-1.81378877e+00 1.61268532e+00 -2.68044889e-01 4.23567921e-01
6.03038907e-01 1.35144532e-01 -2.94727892e-01 3.30589145e-01
-1.31882101e-01 8.92705739e-01 1.04352176e+00 -1.47552633e+00
2.58529603e-01 -1.04006124e+00 -2.59629279e-01 -1.43405155e-03
-1.98362708e-01 1.58932507e-01 -6.76154137e-01 -4.80239302e-01
1.22230656e-01 -9.53632653e-01 1.01213589e-01 7.13718593e-01
-2.40432382e-01 1.10489599e-01 1.03753841e+00 -7.31397450e-01
5.32881260e-01 -2.01419902e+00 2.29785636e-01 1.17387645e-01
3.81229967e-01 6.24114685e-02 -2.09769711e-01 5.30287484e-03
-3.27015489e-01 1.69673964e-01 -8.72068107e-03 -7.99833119e-01
-1.09962180e-01 -2.30553802e-02 -8.22379366e-02 6.25447989e-01
3.75541002e-01 1.03436542e+00 -4.21332657e-01 -5.65106392e-01
2.08680108e-01 8.43937039e-01 -9.10124481e-01 3.83678138e-01
-6.34683520e-02 7.99886942e-01 -5.29121399e-01 1.14272344e+00
1.30674839e+00 4.03125286e-02 -1.01804011e-01 -4.01197970e-01
1.73724100e-01 -4.85101044e-01 -7.38400102e-01 1.77796113e+00
-4.50044930e-01 3.23757976e-01 4.05948639e-01 -6.90159559e-01
1.39979160e+00 2.66184956e-01 7.14010954e-01 -7.44826555e-01
2.44381726e-01 1.82986081e-01 -6.23063982e-01 -5.80302775e-01
3.12638544e-02 -5.59275508e-01 3.17536220e-02 3.04001451e-01
1.61720529e-01 -6.20757699e-01 -5.20473480e-01 -1.50690123e-01
4.86394018e-01 4.28052038e-01 -1.62414670e-01 -2.80398309e-01
3.59153032e-01 -4.79317009e-01 5.16125143e-01 -2.65678406e-01
-1.30068779e-01 9.66852784e-01 5.33715069e-01 -1.09675479e+00
-1.16330922e+00 -1.04921436e+00 -3.74966443e-01 5.96454620e-01
-1.30524039e-01 -5.12108922e-01 -8.91953945e-01 -5.60063541e-01
1.85232878e-01 -1.62089944e-01 -1.08320725e+00 -4.48764451e-02
-5.48267961e-01 -4.86695260e-01 3.85789931e-01 2.79399425e-01
5.26244164e-01 -8.45756471e-01 -2.57500976e-01 -3.35784137e-01
1.33795381e-01 -1.25573778e+00 -5.45529127e-01 -5.20916402e-01
-8.33797932e-01 -1.14235568e+00 -5.75120986e-01 -6.98324859e-01
9.87343848e-01 -4.90845777e-02 1.11723912e+00 1.58147737e-01
-4.71155673e-01 5.30587316e-01 -6.01582006e-02 -3.27828586e-01
-2.63680995e-01 -5.27989686e-01 5.68687439e-01 3.16414237e-01
1.36794252e-02 -9.74625945e-01 -6.53019547e-01 6.82684720e-01
-5.26501715e-01 2.79767781e-01 2.29176342e-01 7.29411364e-01
1.04157794e+00 -2.61025578e-01 3.25399309e-01 -5.86882770e-01
-3.60560603e-02 -2.41908506e-01 -6.06111288e-01 4.12066504e-02
-3.16570193e-01 -3.12539279e-01 5.85195363e-01 -2.26418644e-01
-1.12867904e+00 2.51497597e-01 -4.01537985e-01 -9.63672936e-01
-2.64291108e-01 -1.65352255e-01 -6.13639832e-01 -5.24702132e-01
4.99185205e-01 1.29999891e-01 5.51022112e-01 -6.41340077e-01
1.14143781e-01 5.13429284e-01 2.20280185e-01 -8.21490467e-01
8.69858205e-01 8.80277693e-01 4.04726207e-01 -8.16905022e-01
-7.43931055e-01 2.77118653e-01 -1.04566753e+00 -6.06631219e-01
5.87183475e-01 -1.00149667e+00 -1.03281236e+00 8.45629215e-01
-1.17280066e+00 -4.19124693e-01 1.58710741e-02 -7.19494969e-02
-8.66257668e-01 6.29834458e-02 -4.50505137e-01 -6.64024472e-01
-5.56205869e-01 -1.17757809e+00 1.79729605e+00 1.17931291e-01
4.85671591e-03 -8.01008821e-01 5.39702289e-02 6.51281834e-01
2.66496986e-01 8.56969297e-01 5.83825707e-01 1.29228339e-01
-4.51437652e-01 -1.57758966e-01 -3.31049338e-02 3.05830836e-01
2.13255778e-01 4.50172782e-01 -1.15119672e+00 -2.48151988e-01
-6.23677671e-02 -6.77773535e-01 3.07272434e-01 2.22944573e-01
1.51233482e+00 -3.88029099e-01 -1.74593359e-01 1.27647054e+00
1.11519086e+00 -2.11445913e-01 5.96033573e-01 -3.00040692e-01
8.31288934e-01 9.03854668e-01 3.60727042e-01 6.45352006e-01
4.60405409e-01 1.01371109e+00 5.51874697e-01 -5.62239848e-02
-1.84410378e-01 -5.00506580e-01 3.44264060e-01 7.39397585e-01
-5.50562263e-01 4.17960912e-01 -7.45238185e-01 1.92212418e-03
-1.14576209e+00 -9.84054565e-01 2.32847810e-01 1.84062648e+00
6.94940209e-01 -2.71498948e-01 1.72089025e-01 -2.79885322e-01
4.85735804e-01 4.39485423e-02 -6.91824734e-01 -4.38152164e-01
-1.61035597e-01 4.13204968e-01 -2.61484981e-01 4.53063339e-01
-6.53860688e-01 1.06744576e+00 6.43503571e+00 7.34604597e-01
-1.52265537e+00 -4.46712933e-02 1.09588122e+00 -3.62239778e-01
-3.42794925e-01 -3.46448809e-01 -9.38230753e-01 3.63432586e-01
7.86223471e-01 8.46713334e-02 4.07409161e-01 9.34561133e-01
3.52190912e-01 1.78002581e-01 -1.05966663e+00 1.44210839e+00
2.78418839e-01 -1.58573902e+00 4.00863022e-01 3.57927322e-01
7.90324450e-01 -2.90176690e-01 2.39939719e-01 -2.37279963e-02
-3.03472757e-01 -1.43358016e+00 8.36260021e-01 9.10985351e-01
1.67681634e+00 -7.01264739e-01 2.07333148e-01 1.28804177e-01
-1.13867092e+00 2.55034715e-01 -3.45818907e-01 2.41263695e-02
9.60921496e-02 3.74430090e-01 -5.48231781e-01 4.82331872e-01
1.09776759e+00 9.96872962e-01 -4.82210219e-01 2.90890396e-01
3.03884506e-01 1.72899574e-01 -2.49456361e-01 3.67798924e-01
-4.27389652e-01 -2.39553407e-01 3.74048978e-01 8.59557807e-01
5.03736675e-01 2.89042771e-01 -1.63192600e-01 9.49748278e-01
-3.19597155e-01 1.30493656e-01 -8.13906491e-01 1.14978664e-01
1.78322822e-01 1.59423792e+00 -4.16634738e-01 8.51220861e-02
-2.39323646e-01 8.12700629e-01 4.60869133e-01 3.30639444e-02
-6.35777116e-01 1.89238727e-01 1.08018184e+00 7.00011790e-01
1.73264459e-01 -2.29503870e-01 -7.61633879e-03 -1.22234166e+00
2.03619957e-01 -6.30178154e-01 -1.48003787e-01 -1.08146930e+00
-1.33450139e+00 8.38370979e-01 1.13798402e-01 -1.14636099e+00
-6.02628291e-02 -9.90697801e-01 -5.14420271e-01 6.91622674e-01
-1.46633697e+00 -1.48964059e+00 -6.87451541e-01 8.77169609e-01
2.46947318e-01 -4.20507282e-01 1.06914747e+00 1.92389160e-01
-6.91599965e-01 7.68657684e-01 -4.09537017e-01 9.97053832e-02
7.10968912e-01 -6.53750062e-01 4.09788102e-01 7.20382258e-02
-1.55423731e-01 4.97885585e-01 2.93583602e-01 -4.80963945e-01
-1.97208571e+00 -1.32696056e+00 5.87591827e-01 -9.05329704e-01
1.64068118e-01 -8.25495481e-01 -7.18358517e-01 7.35965908e-01
-4.41333354e-01 7.90678501e-01 6.20280564e-01 -2.81109035e-01
-5.69351196e-01 -5.32432854e-01 -1.59813535e+00 3.78370196e-01
1.59584260e+00 -4.83160496e-01 -2.44930372e-01 3.01941425e-01
4.26408112e-01 -8.78028274e-01 -1.19733870e+00 4.24694985e-01
1.04012990e+00 -1.31525648e+00 8.30113649e-01 -5.75473726e-01
6.22438014e-01 1.26311639e-02 -2.97914445e-01 -1.21903956e+00
1.21956901e-03 -9.32214499e-01 -1.37132332e-01 9.75009739e-01
-8.43217522e-02 -4.48743522e-01 1.06877637e+00 5.63339829e-01
-1.49158269e-01 -1.30771756e+00 -1.15498066e+00 -6.20163441e-01
2.00426906e-01 -2.72361934e-01 1.10214984e+00 7.98866689e-01
-3.80945563e-01 -6.26840070e-02 -4.03953284e-01 -1.18819974e-01
7.20026493e-01 3.48450810e-01 1.10660815e+00 -1.45932734e+00
2.79866159e-01 -1.08537465e-01 -7.38586903e-01 -9.08582211e-01
7.14843452e-01 -9.98312235e-01 -4.51506793e-01 -7.20190227e-01
2.84394979e-01 -4.49989825e-01 5.74360192e-01 5.53133070e-01
4.95882392e-01 7.10710943e-01 -1.10528380e-01 -2.83991713e-02
-1.96613982e-01 9.87786889e-01 1.73378158e+00 5.53942360e-02
8.68747830e-02 -1.70494080e-01 -6.61320329e-01 8.02273214e-01
4.43098694e-01 -1.09138992e-02 -1.68548942e-01 -3.94557476e-01
1.88256770e-01 2.40354687e-01 6.03798032e-01 -5.53912580e-01
-2.65038401e-01 -2.85262495e-01 8.28092456e-01 -4.83906835e-01
9.58442688e-01 -8.32897305e-01 5.56988776e-01 3.65433618e-02
1.89549208e-01 -1.18050408e-02 3.76182884e-01 2.94826984e-01
5.81216179e-02 6.74876988e-01 1.11110294e+00 -1.73309460e-01
-3.70229453e-01 1.24354839e+00 3.78379166e-01 6.01346381e-02
1.09280300e+00 -6.16713524e-01 4.71959151e-02 -3.49307925e-01
-8.52401853e-01 -1.64819151e-01 1.02196729e+00 4.63647187e-01
1.04767895e+00 -1.71319127e+00 -8.40419769e-01 8.99898648e-01
2.13342011e-01 2.47604415e-01 6.45564079e-01 4.81944621e-01
-6.57181323e-01 1.24752998e-01 -8.06898177e-01 -7.23428488e-01
-1.26829183e+00 2.07632780e-01 8.14500213e-01 3.84513646e-01
-6.35083377e-01 7.69865274e-01 2.30417103e-01 -9.15374875e-01
-2.23677039e-01 -1.45008825e-02 5.01261018e-02 -1.54746652e-01
6.46484971e-01 3.56124267e-02 1.93971366e-01 -1.33179784e+00
-4.63852972e-01 1.39880931e+00 3.05669695e-01 2.33939216e-01
1.83214176e+00 -6.02178462e-02 -3.44251275e-01 1.23626322e-01
1.56575716e+00 1.02785267e-01 -1.71542418e+00 1.02117978e-01
-7.82787442e-01 -7.59770274e-01 -2.20432594e-01 -4.48418319e-01
-1.73935986e+00 1.05323052e+00 5.17100692e-01 -5.90844452e-01
1.13292897e+00 1.84109837e-01 6.64028883e-01 5.55953346e-02
6.96137309e-01 -7.19255626e-01 3.49446505e-01 3.98925841e-01
1.46932256e+00 -1.26022398e+00 3.35088782e-02 -6.57099009e-01
-3.41878444e-01 1.37240911e+00 9.41526353e-01 -8.18334520e-02
1.10867727e+00 2.99476147e-01 -7.51522332e-02 -6.07840300e-01
-7.53585696e-01 6.27318323e-01 2.72099048e-01 8.68069530e-01
1.90294221e-01 -1.92321315e-02 4.71847117e-01 7.49391139e-01
-6.10282481e-01 4.77652550e-02 2.80867279e-01 4.09205914e-01
7.26752430e-02 -8.61253321e-01 -3.79902959e-01 2.50349969e-01
-1.71261430e-01 4.86743718e-01 -4.53994095e-01 7.32340634e-01
1.80800438e-01 4.72040266e-01 2.45545000e-01 -6.01529002e-01
4.86191303e-01 4.41258680e-03 9.65857327e-01 -4.78341073e-01
-1.33471027e-01 -2.96284080e-01 -1.74881592e-01 -1.07336259e+00
-2.96529621e-01 -5.20274341e-01 -9.43052888e-01 -7.54079401e-01
2.62365818e-01 -3.52219999e-01 5.92444658e-01 5.80812633e-01
8.47327352e-01 -2.59009957e-01 1.07771182e+00 -1.59806323e+00
-1.39534280e-01 -6.42924666e-01 -7.65822113e-01 6.11843467e-01
3.58386576e-01 -1.10666656e+00 -3.97931159e-01 1.06638432e-01] | [13.110299110412598, -0.014931575395166874] |
519afb1b-8ff4-445d-a5a1-86a12673d5cf | deepsolo-let-transformer-decoder-with | 2211.10772 | null | https://arxiv.org/abs/2211.10772v4 | https://arxiv.org/pdf/2211.10772v4.pdf | DeepSolo: Let Transformer Decoder with Explicit Points Solo for Text Spotting | End-to-end text spotting aims to integrate scene text detection and recognition into a unified framework. Dealing with the relationship between the two sub-tasks plays a pivotal role in designing effective spotters. Although Transformer-based methods eliminate the heuristic post-processing, they still suffer from the synergy issue between the sub-tasks and low training efficiency. In this paper, we present DeepSolo, a simple DETR-like baseline that lets a single Decoder with Explicit Points Solo for text detection and recognition simultaneously. Technically, for each text instance, we represent the character sequence as ordered points and model them with learnable explicit point queries. After passing a single decoder, the point queries have encoded requisite text semantics and locations, thus can be further decoded to the center line, boundary, script, and confidence of text via very simple prediction heads in parallel. Besides, we also introduce a text-matching criterion to deliver more accurate supervisory signals, thus enabling more efficient training. Quantitative experiments on public benchmarks demonstrate that DeepSolo outperforms previous state-of-the-art methods and achieves better training efficiency. In addition, DeepSolo is also compatible with line annotations, which require much less annotation cost than polygons. The code is available at https://github.com/ViTAE-Transformer/DeepSolo. | ['DaCheng Tao', 'Bo Du', 'Tongliang Liu', 'Juhua Liu', 'Shanshan Zhao', 'Jing Zhang', 'Maoyuan Ye'] | 2022-11-19 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Ye_DeepSolo_Let_Transformer_Decoder_With_Explicit_Points_Solo_for_Text_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Ye_DeepSolo_Let_Transformer_Decoder_With_Explicit_Points_Solo_for_Text_CVPR_2023_paper.pdf | cvpr-2023-1 | ['text-spotting', 'scene-text-detection'] | ['computer-vision', 'computer-vision'] | [ 3.53775561e-01 -1.74774572e-01 -2.22487330e-01 -3.37741882e-01
-1.14910376e+00 -5.79289973e-01 4.37185079e-01 1.41663253e-01
-2.94875801e-01 3.04903209e-01 8.11822305e-04 -2.94114053e-01
3.51358503e-01 -6.65596962e-01 -7.88282096e-01 -6.53465271e-01
7.00229764e-01 7.84022689e-01 5.86481988e-01 -2.35419013e-02
6.52313769e-01 1.51758909e-01 -1.13937497e+00 6.10860109e-01
1.01764452e+00 1.07955790e+00 6.61658645e-01 6.09120250e-01
-3.90566945e-01 5.50953686e-01 -6.58481717e-01 -5.39077699e-01
6.82358071e-02 -3.03508461e-01 -4.75271970e-01 3.45015675e-01
6.15810633e-01 -5.07547259e-01 -4.17869151e-01 9.70991194e-01
7.87167370e-01 -9.61697996e-02 5.73923469e-01 -8.52996349e-01
-3.51918399e-01 7.33973384e-01 -7.78907478e-01 -2.55208135e-01
3.32999676e-01 2.20024318e-01 1.35657597e+00 -1.35132492e+00
3.69167745e-01 9.31777239e-01 8.47745955e-01 3.34905773e-01
-1.02117383e+00 -3.55298549e-01 1.49814263e-01 2.55331278e-01
-1.37072659e+00 -6.78607225e-01 5.17750740e-01 -3.72106642e-01
8.66226554e-01 4.84932423e-01 4.16421056e-01 1.06971586e+00
-1.50451079e-01 1.39781690e+00 2.97903061e-01 -4.54509735e-01
9.26050693e-02 6.06601909e-02 1.94989622e-03 8.90189528e-01
-3.01071182e-02 -5.16045332e-01 -8.61532092e-01 2.59121954e-01
6.07612789e-01 5.87880164e-02 -3.81940573e-01 -1.27279952e-01
-1.40699768e+00 5.30059934e-01 2.42466882e-01 1.76258579e-01
-2.84681041e-02 1.34921491e-01 5.85052133e-01 -6.27764314e-02
2.96250492e-01 7.57575035e-02 -3.56693387e-01 -1.84012204e-01
-1.27286482e+00 -6.86585531e-02 5.29908895e-01 1.22665322e+00
7.36920774e-01 -1.51196882e-01 -5.94503045e-01 1.15742230e+00
4.13058400e-01 5.80633819e-01 3.98822069e-01 -2.64394850e-01
1.20305824e+00 7.51329958e-01 -9.74949300e-02 -8.20607245e-01
-1.97024330e-01 -5.19278526e-01 -7.80034244e-01 -1.91177964e-01
3.66001427e-01 2.35734712e-02 -9.14751768e-01 9.63249624e-01
2.41008461e-01 2.12607235e-01 -3.80858421e-01 9.96697128e-01
6.32102549e-01 9.39601541e-01 -4.22448695e-01 2.88874090e-01
1.55758333e+00 -1.37106597e+00 -6.02855861e-01 -7.11263537e-01
8.65304410e-01 -1.06207776e+00 1.29250991e+00 5.27874172e-01
-8.67700994e-01 -3.42784315e-01 -1.11229002e+00 -5.10881901e-01
-1.53006062e-01 8.69103253e-01 2.10250467e-01 3.11092138e-01
-8.58585715e-01 4.18187708e-01 -8.08713675e-01 -2.60274082e-01
4.66091692e-01 2.21158534e-01 4.30996083e-02 9.67653021e-02
-8.78056169e-01 6.02057755e-01 3.42695594e-01 3.20935696e-01
-6.46161973e-01 -4.85032201e-01 -8.39508057e-01 9.15983990e-02
6.44063830e-01 -4.94066626e-01 1.39519227e+00 -6.27701342e-01
-1.55523276e+00 8.03737760e-01 -4.92349148e-01 -2.75757670e-01
7.95343518e-01 -3.80890757e-01 2.41998509e-02 2.17392355e-01
4.26512778e-01 6.35765016e-01 9.58461881e-01 -1.03711951e+00
-9.56676543e-01 -2.36224577e-01 -4.28740561e-01 4.63498890e-01
-4.42198396e-01 -7.85084665e-02 -1.36987102e+00 -8.59104693e-01
3.08301032e-01 -6.06501162e-01 1.43284157e-01 4.09346431e-01
-9.92866993e-01 -3.31977755e-01 9.30980921e-01 -7.52209425e-01
1.41381323e+00 -2.30035830e+00 2.00487033e-01 -1.85472723e-02
2.24355236e-01 1.38554782e-01 9.38366652e-02 6.03061140e-01
2.67360687e-01 4.19798717e-02 -2.57111162e-01 -9.17870879e-01
2.52876997e-01 -7.05062672e-02 -5.71284235e-01 5.44098914e-01
1.26552895e-01 8.96206856e-01 -6.02775335e-01 -7.87768662e-01
5.06357312e-01 2.09927663e-01 -4.14440900e-01 2.37871762e-02
-5.62892795e-01 6.87659532e-02 -5.85224628e-01 8.19583714e-01
5.02547264e-01 -4.10384119e-01 1.02801003e-01 -1.51478469e-01
-2.49404281e-01 4.92183208e-01 -1.15884078e+00 1.75933051e+00
-1.78964600e-01 1.10946155e+00 -7.13806227e-02 -9.94417429e-01
9.56436992e-01 1.06822670e-01 2.15997249e-01 -7.95660317e-01
1.61234275e-01 2.80829459e-01 -5.75342178e-01 -4.86758232e-01
7.86803246e-01 3.95015448e-01 -4.00836170e-02 3.08288574e-01
-1.78270549e-01 -1.19147427e-01 1.38051659e-01 2.11949393e-01
9.12570775e-01 4.05221075e-01 9.14857984e-02 1.20461918e-01
4.42861915e-01 5.69116995e-02 5.01739383e-01 7.16079295e-01
8.33137333e-02 1.02344906e+00 6.53874040e-01 -1.56548500e-01
-1.20894504e+00 -6.84350252e-01 -1.87297642e-01 1.27444673e+00
4.37515765e-01 -6.59595251e-01 -8.43480170e-01 -5.94398618e-01
-6.18472770e-02 7.25336850e-01 -5.29425979e-01 2.84542412e-01
-6.97381735e-01 -4.84356433e-01 8.52466524e-01 5.91942906e-01
5.28941154e-01 -7.31497347e-01 -4.26766306e-01 1.39724523e-01
-5.04201651e-01 -1.19115233e+00 -9.92557108e-01 3.94866467e-01
-7.32444525e-01 -7.76323676e-01 -9.19515371e-01 -1.00830007e+00
7.09963202e-01 4.24834549e-01 7.96112239e-01 3.43337119e-01
-2.39745781e-01 -1.39510259e-01 -5.53356409e-01 -1.61309466e-01
-7.12712407e-02 3.04354459e-01 -5.72543263e-01 7.94887245e-02
1.92554891e-01 1.84273683e-02 -6.58455074e-01 6.24963462e-01
-7.60545552e-01 6.44397974e-01 5.16691566e-01 9.47799742e-01
7.53820240e-01 -2.42362559e-01 9.85427666e-03 -6.66972578e-01
2.16376901e-01 -1.08339615e-01 -6.89244032e-01 4.36907500e-01
-3.18239003e-01 -4.88317348e-02 7.26670146e-01 -9.10889879e-02
-8.24170768e-01 3.33796144e-01 -1.88953713e-01 -3.14130038e-01
-3.55351344e-02 3.83836776e-01 -3.21877867e-01 2.93794185e-01
5.69798172e-01 7.68944800e-01 -3.53087485e-01 -5.66871881e-01
1.12212278e-01 1.00426328e+00 4.87295836e-01 -5.56827426e-01
7.81208694e-01 4.24322754e-01 -4.87125814e-01 -9.54283178e-01
-9.05607581e-01 -6.39885724e-01 -6.94062531e-01 -1.47726685e-01
6.74597263e-01 -8.89117360e-01 -6.10885024e-01 7.91335046e-01
-1.33685482e+00 -7.38215148e-01 3.14707085e-02 7.07282545e-03
-5.67201197e-01 7.56371915e-01 -6.54114068e-01 -5.58724225e-01
-5.13303101e-01 -1.25458908e+00 1.76932478e+00 1.06494479e-01
5.72872460e-02 -7.62253761e-01 -4.02835131e-01 5.62920868e-01
-5.99089749e-02 -3.76525611e-01 5.62416077e-01 -5.85923195e-01
-8.32389176e-01 -3.09705555e-01 -4.53082770e-01 -4.29011695e-02
-1.95051029e-01 1.49823129e-01 -8.83888066e-01 -1.34091362e-01
-3.57636452e-01 -1.87305346e-01 9.87049639e-01 1.14089645e-01
1.35078561e+00 -1.58388525e-01 -4.99080300e-01 9.18137252e-01
1.37481058e+00 7.36311451e-02 7.62619555e-01 6.12403333e-01
8.64036202e-01 3.93488079e-01 8.54399860e-01 7.54400611e-01
4.98078167e-01 9.72192049e-01 4.48235154e-01 -1.55682236e-01
-1.88587964e-01 -4.43015248e-01 4.80344206e-01 7.56214976e-01
4.85364169e-01 -7.98661232e-01 -1.04063690e+00 3.17232072e-01
-2.15310955e+00 -8.29140782e-01 -5.74393511e-01 2.04666090e+00
8.58225703e-01 1.64381176e-01 -1.38568255e-04 3.13059360e-01
9.43999231e-01 2.11188406e-01 -5.66398323e-01 1.29185244e-01
-2.59318441e-01 -1.84686899e-01 6.20704114e-01 5.17158866e-01
-1.15313864e+00 1.38520980e+00 5.37794495e+00 1.28102231e+00
-1.20108140e+00 -7.44093731e-02 5.92124939e-01 -5.63583747e-02
-1.49592042e-01 -3.06201242e-02 -1.17260349e+00 6.22242630e-01
3.15331578e-01 1.63983136e-01 3.19942057e-01 6.85424268e-01
3.61838013e-01 -1.66667700e-01 -1.14564002e+00 1.22017050e+00
1.63676172e-01 -1.44873786e+00 2.71783601e-02 -1.07506856e-01
4.23009336e-01 1.35393843e-01 1.40523920e-02 4.12527192e-03
-2.52344102e-01 -7.82480061e-01 1.32190120e+00 3.29397082e-01
1.05589712e+00 -2.62798756e-01 4.70771432e-01 4.70704645e-01
-1.33815670e+00 5.89481965e-02 -4.12254930e-01 3.18023145e-01
1.81769133e-01 6.95052385e-01 -1.01190591e+00 5.30168176e-01
4.84270602e-01 1.02314854e+00 -5.95257759e-01 1.18592656e+00
-4.94161695e-01 6.82707250e-01 -3.68129492e-01 -2.96642721e-01
3.53192925e-01 -1.73916087e-01 4.17665720e-01 1.51619661e+00
4.30147052e-01 -2.55200893e-01 1.92945451e-01 7.81657875e-01
-4.53007892e-02 2.50701398e-01 -1.92655578e-01 2.33669411e-02
5.84681749e-01 1.09423566e+00 -1.04136586e+00 -5.56857288e-01
-4.31052685e-01 1.44900191e+00 2.35912219e-01 2.19657823e-01
-1.06625915e+00 -6.66193247e-01 2.01469675e-01 2.46016942e-02
5.89017510e-01 -1.68578714e-01 -8.32557380e-01 -1.45192122e+00
2.80269921e-01 -9.69041109e-01 1.27596453e-01 -9.12220836e-01
-8.89773369e-01 3.39902222e-01 -5.92300177e-01 -1.38610196e+00
1.46958828e-01 -7.43979216e-01 -6.15485847e-01 8.25602412e-01
-1.33544695e+00 -1.16483486e+00 -4.12919492e-01 5.51126182e-01
1.22326887e+00 3.62885967e-02 4.03606236e-01 4.61597979e-01
-1.06137431e+00 8.07695568e-01 2.89000958e-01 5.22850692e-01
9.65459704e-01 -1.33746314e+00 7.91701496e-01 9.94848967e-01
1.73864201e-01 2.86686301e-01 4.99239773e-01 -7.49588728e-01
-1.51716709e+00 -1.01544285e+00 8.14675093e-01 -4.20559198e-01
6.26035810e-01 -6.93227172e-01 -8.65339637e-01 5.78786016e-01
7.77108297e-02 -3.07824045e-01 3.22028369e-01 -3.95564735e-02
-1.57918304e-01 -5.09797484e-02 -5.64547479e-01 8.05434942e-01
9.73794281e-01 -5.42175651e-01 -4.25654352e-01 5.29519200e-01
5.52342653e-01 -7.40433633e-01 -3.25703919e-01 -8.41870606e-02
3.45314682e-01 -1.02731562e+00 6.04540169e-01 1.43755704e-01
5.61344504e-01 -5.02855062e-01 -9.25604776e-02 -8.25374007e-01
-1.06084451e-01 -6.56984806e-01 -7.28017688e-02 1.25602245e+00
5.96045554e-01 -4.35812920e-01 1.00728750e+00 1.22117043e-01
-5.53072214e-01 -8.58163178e-01 -8.04999053e-01 -3.58428270e-01
-3.12285036e-01 -6.40994191e-01 4.75492924e-01 7.45259166e-01
4.05386686e-02 5.27974188e-01 -6.20821834e-01 1.80362999e-01
4.55906749e-01 1.74074322e-01 8.74498725e-01 -9.76325333e-01
-3.17029774e-01 -7.56929219e-01 -1.74611509e-01 -1.96191406e+00
-2.43512705e-01 -8.98191869e-01 5.02312899e-01 -1.76005924e+00
1.01440176e-01 -4.32353348e-01 3.40708047e-01 7.94018388e-01
-3.11401755e-01 1.73570886e-01 2.43407637e-01 3.75463367e-01
-8.87043774e-01 6.65592909e-01 1.13051856e+00 -2.47539878e-01
-7.75898173e-02 -5.80189712e-02 -5.30049384e-01 5.97222805e-01
7.84063041e-01 -3.47212791e-01 -1.32261604e-01 -1.03413057e+00
3.11918408e-01 1.27043575e-01 3.06499153e-01 -9.13663507e-01
7.24126518e-01 6.39804155e-02 3.72650892e-01 -1.08027935e+00
3.04680645e-01 -6.76114798e-01 -3.24683756e-01 2.35367715e-01
-3.37090760e-01 -2.19323769e-01 4.60715294e-02 3.88204813e-01
-7.03839958e-02 -6.11650109e-01 6.03615820e-01 2.52858162e-01
-6.96515203e-01 3.48255515e-01 -2.92675257e-01 -1.16922669e-01
8.84074211e-01 -5.50226390e-01 -4.16109532e-01 -1.50514081e-01
-2.85807252e-01 4.62487519e-01 6.17880940e-01 2.93199509e-01
6.87516332e-01 -8.47433746e-01 -6.59159422e-01 1.20526545e-01
1.30773142e-01 4.14272398e-01 1.10339008e-01 9.45152283e-01
-8.85485351e-01 5.40986240e-01 3.50793093e-01 -9.50460851e-01
-1.12339520e+00 2.56806463e-01 3.82214636e-01 -1.58228919e-01
-9.66659963e-01 8.02125931e-01 2.61211842e-01 -3.76258254e-01
6.54164135e-01 -2.48946130e-01 2.08108172e-01 1.46571428e-01
5.26497483e-01 2.77624935e-01 1.72530293e-01 -3.95779550e-01
-2.93651581e-01 7.96734810e-01 -1.43799767e-01 -1.40135989e-01
1.11052179e+00 -2.46234000e-01 -8.47413088e-04 2.87176639e-01
8.39903235e-01 1.76596075e-01 -1.42417502e+00 -3.11653107e-01
1.74593687e-01 -4.40149903e-01 7.38486694e-03 -8.00296545e-01
-8.98538291e-01 1.14496195e+00 1.53692260e-01 1.53447697e-02
1.00172174e+00 -1.85698643e-01 9.18807447e-01 5.08419275e-01
1.17989354e-01 -1.19507515e+00 2.33415306e-01 7.21772492e-01
7.56057799e-01 -1.13785708e+00 -4.75633889e-02 -6.29058719e-01
-6.90186739e-01 1.35525680e+00 6.87698126e-01 2.53501356e-01
2.79767290e-02 5.00735939e-01 -2.02119304e-03 8.63436311e-02
-7.80439377e-01 -1.76831454e-01 2.69480050e-01 2.72159189e-01
4.70638514e-01 -6.75757751e-02 5.69143742e-02 3.81121933e-01
-6.59446046e-02 -1.75908923e-01 2.63236731e-01 6.81619048e-01
-7.42474973e-01 -1.07967103e+00 -5.34033358e-01 5.20974219e-01
-3.11988711e-01 -4.51163769e-01 -4.27208871e-01 4.52251524e-01
-1.23496927e-01 7.12237358e-01 1.56498849e-02 -4.47977334e-01
3.15840364e-01 -1.68208275e-02 6.65571913e-02 -7.42887020e-01
-5.12511194e-01 6.22836769e-01 -1.10646372e-03 -3.81813973e-01
2.95674562e-01 -6.06195092e-01 -1.39339793e+00 -4.02340055e-01
-6.48268223e-01 -5.85297607e-02 6.16565168e-01 8.63478243e-01
4.52636331e-01 5.32560706e-01 6.10475004e-01 -9.38044429e-01
-6.04006052e-01 -9.56025660e-01 -2.70911396e-01 -4.38654833e-02
2.22712338e-01 -1.78585559e-01 -1.38435230e-01 2.20771655e-01] | [11.966286659240723, 2.2420661449432373] |
782bd794-5f81-4e15-a71d-179c46d63590 | disentangling-semantics-in-language-throughs | 2012.13031 | null | https://arxiv.org/abs/2012.13031v2 | https://arxiv.org/pdf/2012.13031v2.pdf | Disentangling semantics in language through VAEs and a certain architectural choice | We present an unsupervised method to obtain disentangled representations of sentences that single out semantic content. Using modified Transformers as building blocks, we train a Variational Autoencoder to translate the sentence to a fixed number of hierarchically structured latent variables. We study the influence of each latent variable in generation on the dependency structure of sentences, and on the predicate structure it yields when passed through an Open Information Extraction model. Our model could separate verbs, subjects, direct objects, and prepositional objects into latent variables we identified. We show that varying the corresponding latent variables results in varying these elements in sentences, and that swapping them between couples of sentences leads to the expected partial semantic swap. | ['Djamé Seddah', 'Joseph Le Roux', 'Ghazi Felhi'] | 2020-12-24 | null | null | null | null | ['open-information-extraction'] | ['natural-language-processing'] | [ 8.57524425e-02 8.24447930e-01 -2.73457140e-01 -4.01217252e-01
-5.71120441e-01 -9.83555913e-01 8.32135618e-01 -3.78369391e-02
-1.47128180e-01 8.81771863e-01 9.67797101e-01 -2.03859672e-01
1.87746003e-01 -1.14206684e+00 -7.33391941e-01 -6.76359773e-01
1.45988435e-01 7.37094522e-01 -1.00459829e-01 -1.64518312e-01
-3.38732630e-01 1.76897764e-01 -1.11770391e+00 4.88916934e-01
5.11931598e-01 1.39523193e-01 1.67770803e-01 5.90138257e-01
-1.36405900e-01 5.90888500e-01 -7.52281725e-01 -8.14091563e-01
1.64983481e-01 -6.34366274e-01 -9.85328317e-01 8.14064443e-02
-1.95954684e-02 -2.90694565e-01 -3.96748990e-01 8.44933510e-01
-4.48339209e-02 -2.25717798e-02 1.05222511e+00 -1.21428907e+00
-1.01666903e+00 1.40393996e+00 -1.33143410e-01 1.72607694e-02
1.95164949e-01 -2.26417676e-01 1.73988056e+00 -6.96386814e-01
1.04534924e+00 1.42479932e+00 4.14744377e-01 7.06903398e-01
-1.94417763e+00 -1.78552821e-01 -4.70217019e-02 -2.04938337e-01
-9.14025068e-01 -6.75621212e-01 1.02449322e+00 -5.92250645e-01
1.07517254e+00 1.36595637e-01 6.64188564e-01 1.50661004e+00
3.50451052e-01 5.90983212e-01 8.52030516e-01 -5.53610623e-01
1.69881701e-01 4.28280711e-01 5.39852083e-01 6.61222041e-01
4.44883913e-01 5.54519445e-02 -6.57550931e-01 -2.45699763e-01
4.85134631e-01 -3.93713832e-01 -2.54712045e-01 -3.84077162e-01
-1.04562676e+00 1.28613877e+00 1.95927143e-01 4.21682507e-01
-3.29719335e-01 4.12851833e-02 1.38543040e-01 2.80764908e-01
4.80334967e-01 6.91454768e-01 -8.20336521e-01 3.15117210e-01
-5.80805063e-01 2.17826620e-01 1.06673086e+00 7.86329925e-01
8.03110778e-01 -1.22692347e-01 -2.71058947e-01 5.47285557e-01
3.71028930e-01 1.64421752e-01 3.92243207e-01 -1.04463530e+00
4.95546430e-01 4.25596178e-01 6.31644763e-03 -8.00819755e-01
-1.76382646e-01 -1.24502704e-01 -4.21957165e-01 -2.72089332e-01
1.62884012e-01 -5.01332819e-01 -7.75385618e-01 2.30769229e+00
-1.46917269e-01 -4.63261157e-01 5.21005630e-01 5.44250488e-01
8.74308228e-01 7.84167290e-01 2.13338882e-01 -5.04354179e-01
1.40679836e+00 -5.85633039e-01 -8.38519573e-01 -2.64243275e-01
4.33153778e-01 -3.56944531e-01 6.71886921e-01 -2.34666348e-01
-1.47735572e+00 -3.37062776e-01 -1.01659024e+00 -4.93882835e-01
-2.20541105e-01 -1.68524906e-01 7.04449296e-01 4.04637694e-01
-1.23036075e+00 8.10974479e-01 -1.01159871e+00 1.12082206e-01
1.48419648e-01 3.45699489e-01 -4.92181212e-01 6.13711119e-01
-1.49764574e+00 1.04509854e+00 4.86163706e-01 -1.49535507e-01
-9.26779807e-01 -4.40814197e-01 -1.31008863e+00 4.31418806e-01
1.45388274e-02 -1.19011760e+00 1.15601468e+00 -8.81075799e-01
-1.33012795e+00 8.38231564e-01 -6.37273908e-01 -4.75404501e-01
9.16998647e-03 7.19248727e-02 -5.19703515e-02 7.92980343e-02
3.24140131e-01 8.25640142e-01 9.92678881e-01 -1.37923026e+00
-6.71678707e-02 -5.10972202e-01 3.01241994e-01 2.19149321e-01
-1.42840594e-01 7.23129809e-02 6.16704337e-02 -5.35852015e-01
3.15476537e-01 -8.43789756e-01 -2.33243573e-02 -5.54516792e-01
-7.53298640e-01 -3.92668009e-01 4.13848966e-01 -8.50411475e-01
7.16197729e-01 -2.07028103e+00 1.01645267e+00 -1.70405179e-01
5.78684628e-01 -5.96860051e-01 -1.94339212e-02 5.00287294e-01
-1.99040294e-01 4.82221514e-01 -2.46021405e-01 -6.56493187e-01
4.10356551e-01 5.88959277e-01 -5.62910259e-01 2.19698906e-01
4.31298912e-01 1.18141937e+00 -6.20818496e-01 -4.61254239e-01
-1.31779015e-01 5.17423213e-01 -9.25467074e-01 2.66500324e-01
-5.57098806e-01 2.58706063e-01 -3.55581194e-01 9.51485857e-02
2.77803779e-01 -2.75992215e-01 5.99331498e-01 -1.61325008e-01
9.21829715e-02 8.11260223e-01 -5.68325758e-01 1.39586425e+00
-3.80084693e-01 8.57660294e-01 5.53075708e-02 -8.58835399e-01
6.45851135e-01 6.57168567e-01 7.21545145e-02 -8.78949091e-02
2.01716736e-01 -3.15192640e-01 1.10597625e-01 -4.93832916e-01
6.30822837e-01 -7.14853108e-01 -5.10573447e-01 6.13707244e-01
4.82967645e-01 -1.35466143e-01 2.67038107e-01 6.15784228e-01
8.71209741e-01 6.81061000e-02 1.15091197e-01 -2.42896616e-01
4.98200282e-02 -2.35163540e-01 6.46475971e-01 5.71470380e-01
1.30573869e-01 3.67077470e-01 1.27100492e+00 -2.04785600e-01
-1.29490018e+00 -1.67315567e+00 -2.55544722e-01 8.99792671e-01
-1.06984206e-01 -5.72927058e-01 -6.88680589e-01 -5.15699625e-01
-1.62369996e-01 1.36350751e+00 -7.78775454e-01 -2.81426847e-01
-5.14265776e-01 -7.77294636e-01 3.96655530e-01 5.71813405e-01
9.73354466e-03 -1.21945763e+00 -3.64837378e-01 -4.26548868e-02
-4.71487075e-01 -1.02281427e+00 -1.73064414e-02 5.90046644e-01
-7.99562097e-01 -7.34219968e-01 -1.42964944e-01 -8.56273293e-01
7.01333225e-01 -4.35219258e-01 1.36528397e+00 -4.14994419e-01
2.73738146e-01 -2.82435142e-03 -5.57351112e-02 -1.18447497e-01
-8.66472304e-01 1.51490584e-01 -1.05215169e-01 -1.53135002e-01
2.94883609e-01 -6.63236082e-01 1.48840211e-02 -3.51131469e-01
-8.77910674e-01 2.75621831e-01 1.83914483e-01 9.09408867e-01
1.50316268e-01 -7.82467499e-02 3.17935720e-02 -1.03382719e+00
9.32553172e-01 -5.03845096e-01 -4.13758576e-01 3.11513156e-01
-2.48754174e-01 8.33771944e-01 3.64428669e-01 -1.24790661e-01
-1.32254648e+00 -3.43999714e-02 8.15423653e-02 -4.30116765e-02
-1.62609771e-01 6.11163557e-01 -4.26859260e-01 8.88946533e-01
5.61999559e-01 1.24734148e-01 -8.71334523e-02 -3.01908433e-01
7.59469628e-01 3.91641855e-01 2.09854797e-01 -6.16040349e-01
7.73385286e-01 3.93968284e-01 -2.01894477e-01 -6.45088434e-01
-8.19662869e-01 3.74682844e-01 -9.51302767e-01 5.08564770e-01
1.50786471e+00 -9.10588145e-01 -4.16234493e-01 -3.73614505e-02
-1.74521101e+00 -7.39148185e-02 -6.27422512e-01 3.08369964e-01
-5.42162120e-01 8.52729306e-02 -8.34917963e-01 -4.17148560e-01
-3.68992575e-02 -1.19207251e+00 1.11624706e+00 -1.24639079e-01
-7.60937393e-01 -1.25833309e+00 2.86652833e-01 3.50504994e-01
-1.47062808e-01 -1.23377316e-01 1.44493973e+00 -7.77046084e-01
-7.17411518e-01 5.86694218e-02 9.80743393e-02 2.82536089e-01
1.86838120e-01 5.56234689e-03 -7.35504389e-01 1.57502461e-02
3.37275028e-01 -2.11722121e-01 1.03729951e+00 4.16352332e-01
4.16681588e-01 -6.99292183e-01 -4.20115262e-01 4.92688954e-01
1.19022894e+00 -1.38242915e-01 4.43550736e-01 -1.46281913e-01
5.83197534e-01 7.08852887e-01 -4.87526000e-01 -1.54945916e-02
3.69566262e-01 4.34061527e-01 -5.17014563e-02 3.05615395e-01
1.54672235e-01 -3.80204260e-01 5.71351647e-01 9.68022346e-01
1.23160318e-01 -3.88711184e-01 -6.18630767e-01 6.32358789e-01
-1.45848238e+00 -1.21237493e+00 2.26699337e-02 1.60275340e+00
1.13237631e+00 1.95596889e-01 -2.79078543e-01 -1.19008772e-01
5.94562650e-01 4.02865171e-01 1.11317113e-02 -5.33281744e-01
-1.66468605e-01 5.48417211e-01 1.32117689e-01 1.09437203e+00
-8.13130438e-01 1.05880737e+00 7.49943399e+00 5.95391728e-02
-5.61185062e-01 2.12176874e-01 1.84720606e-01 -3.67870629e-02
-1.26091516e+00 4.04890001e-01 -7.57143140e-01 2.77237713e-01
9.73910809e-01 -3.70127589e-01 4.01923388e-01 5.04776895e-01
1.34103954e-01 2.16312870e-01 -1.56870270e+00 1.58183500e-01
1.35744587e-01 -1.38316143e+00 4.50993568e-01 8.67427662e-02
6.74571633e-01 -3.07413071e-01 -5.43517508e-02 7.27566853e-02
7.46995330e-01 -9.06189144e-01 7.96278536e-01 3.53790492e-01
7.27016628e-02 -4.04845238e-01 4.50053900e-01 3.79967034e-01
-5.90618014e-01 3.84555943e-02 -1.12012766e-01 -4.60987985e-01
3.19313556e-01 2.28841305e-01 -8.01457405e-01 2.09224150e-01
1.03194863e-01 3.88466716e-01 -2.93426335e-01 -5.69493435e-02
-9.79785621e-01 6.48181558e-01 6.57970086e-02 -1.24384731e-01
-2.22808823e-01 -3.74881238e-01 8.17346215e-01 7.03351736e-01
-1.51895031e-01 2.94646263e-01 -2.48855531e-01 1.47433436e+00
-2.12168902e-01 -2.94220626e-01 -6.82108641e-01 -2.88696080e-01
3.18869174e-01 9.17359889e-01 -7.33937681e-01 -5.84236920e-01
-3.49013776e-01 1.21213603e+00 4.70481426e-01 5.87566435e-01
-7.03595340e-01 -7.62659311e-02 7.46811152e-01 -1.50394037e-01
5.38916111e-01 -2.85828918e-01 -4.29506898e-01 -1.70318246e+00
8.49247053e-02 -4.58445132e-01 8.63942280e-02 -9.70799088e-01
-1.23107409e+00 6.54898405e-01 3.27412367e-01 -3.13288867e-01
-6.67016268e-01 -3.71911645e-01 -5.65050304e-01 1.12128568e+00
-6.63787246e-01 -9.69786048e-01 3.91710669e-01 2.94377387e-01
6.12152994e-01 -2.51882114e-02 1.28901100e+00 -4.40081596e-01
-6.81135595e-01 2.81013399e-01 -8.26652646e-02 4.71055746e-01
5.84655860e-03 -1.33146298e+00 5.43312192e-01 9.03768778e-01
7.32408345e-01 1.16734219e+00 1.09922838e+00 -6.84693635e-01
-9.30599570e-01 -6.63457870e-01 1.55462110e+00 -9.24486935e-01
7.20050752e-01 -8.29711556e-01 -5.82111657e-01 1.39514184e+00
7.17639148e-01 -4.93146271e-01 8.30452979e-01 4.31060433e-01
-4.73489195e-01 3.75899613e-01 -7.58104026e-01 7.63356090e-01
7.80682445e-01 -9.94708538e-01 -1.54521704e+00 2.46545762e-01
1.48144507e+00 -2.46251181e-01 -5.36319673e-01 1.50353312e-01
4.17749494e-01 -8.31090927e-01 7.95987070e-01 -1.26993597e+00
1.08310270e+00 1.67249963e-01 -2.40537390e-01 -1.55612624e+00
-5.37547469e-01 -3.26590002e-01 -1.59368113e-01 1.00277793e+00
9.18578804e-01 -4.38029349e-01 9.09427702e-01 1.01413143e+00
2.72121630e-03 -3.84228975e-01 -8.06384981e-01 -2.23848522e-01
3.46580327e-01 -1.22738175e-01 4.05772150e-01 8.86460245e-01
2.07473144e-01 1.22706163e+00 -1.09212063e-01 2.22205669e-01
5.27005613e-01 4.88108218e-01 2.45739013e-01 -1.20974195e+00
-7.78392971e-01 -9.47150514e-02 -2.27825522e-01 -8.52708161e-01
6.94049656e-01 -1.15924037e+00 -2.47875109e-01 -1.70602190e+00
4.70186979e-01 9.18897837e-02 3.37836370e-02 5.18656611e-01
-7.81414583e-02 -1.42526180e-01 2.10214064e-01 7.71680623e-02
-9.12672654e-02 7.54674137e-01 8.99186909e-01 -3.19752067e-01
-2.94965237e-01 -7.79545531e-02 -1.03632963e+00 7.90921509e-01
7.41855919e-01 -7.35496163e-01 -4.37110722e-01 -6.58108234e-01
4.93506819e-01 5.12527108e-01 4.49838459e-01 -7.03120679e-02
-1.58773437e-01 2.38581914e-02 4.94346291e-01 -3.13059032e-01
5.16438961e-01 -5.41064501e-01 2.03502566e-01 2.45919019e-01
-8.65711868e-01 -4.72400412e-02 -8.47006291e-02 3.54749501e-01
-2.01528430e-01 -4.41717952e-01 3.18638414e-01 -3.73841435e-01
-1.14292115e-01 -2.17117280e-01 -5.91749310e-01 3.56733799e-02
7.75031686e-01 3.44411209e-02 -4.70308922e-02 -2.52971292e-01
-1.42489123e+00 -1.28362760e-01 3.90154988e-01 4.68411058e-01
3.92264456e-01 -1.19822896e+00 -6.71077967e-01 3.56024802e-01
-4.57146674e-01 -1.54613748e-01 -1.23471402e-01 2.51434267e-01
-8.48060176e-02 4.46464151e-01 -2.12391108e-01 -1.72122657e-01
-1.16478717e+00 5.87766886e-01 2.46298194e-01 -2.46795356e-01
-2.40772024e-01 8.84533644e-01 4.00936335e-01 -4.22368705e-01
-1.67532340e-01 -2.70208657e-01 -1.50297508e-01 3.75005037e-01
-2.69829959e-01 -6.05750419e-02 -3.67167652e-01 -7.43597746e-01
-1.21919096e-01 1.97980329e-02 -6.72770217e-02 -6.68341279e-01
1.48838651e+00 -8.72101709e-02 -5.97773612e-01 7.23901629e-01
1.41664410e+00 3.96603793e-01 -8.97713840e-01 9.78969187e-02
-2.42150769e-01 4.24769670e-02 -1.06955431e-01 -3.12518030e-01
-6.96651816e-01 6.50387764e-01 -2.96166539e-01 5.89050472e-01
5.70039392e-01 7.75733829e-01 5.00135779e-01 1.82704672e-01
4.38662283e-02 -5.81061184e-01 -1.91831142e-02 4.65022922e-01
8.11429322e-01 -8.22318554e-01 -3.22388075e-02 -5.19417763e-01
-5.78478158e-01 8.65082145e-01 3.39229256e-01 -1.59692466e-01
5.92107177e-01 4.02602464e-01 -2.18124881e-01 -4.55085397e-01
-1.05585623e+00 1.92809254e-02 1.29963607e-01 3.53539169e-01
5.14315367e-01 4.71592069e-01 -4.30762112e-01 9.31677759e-01
-8.03442776e-01 -5.73969543e-01 7.64193356e-01 5.08811593e-01
-1.36180207e-01 -1.25811458e+00 -6.89963326e-02 3.91613126e-01
-4.82835531e-01 -5.43582082e-01 -7.29841530e-01 6.65075302e-01
2.43755117e-01 7.09481955e-01 4.91427422e-01 -9.57375318e-02
-7.43214134e-03 4.87664431e-01 7.13374257e-01 -9.37921703e-01
-1.95895314e-01 -9.97546241e-02 2.84832507e-01 -1.87125608e-01
-3.03167999e-01 -8.47180128e-01 -1.24979615e+00 6.53706044e-02
-4.94769663e-01 5.77909112e-01 4.78284359e-01 1.18344378e+00
2.06162006e-01 5.08133292e-01 4.44039285e-01 -5.32915771e-01
-6.75133944e-01 -9.14985180e-01 -3.05223793e-01 5.10564864e-01
5.33955336e-01 -3.99593771e-01 -5.32695770e-01 4.16223377e-01] | [10.855985641479492, 8.956674575805664] |
23c6fd15-a8ce-4d4b-9b3a-7eb71a289849 | restricted-boltzmann-machines-for-galaxy | 1911.06259 | null | https://arxiv.org/abs/1911.06259v2 | https://arxiv.org/pdf/1911.06259v2.pdf | Restricted Boltzmann Machines for galaxy morphology classification with a quantum annealer | We present the application of Restricted Boltzmann Machines (RBMs) to the task of astronomical image classification using a quantum annealer built by D-Wave Systems. Morphological analysis of galaxies provides critical information for studying their formation and evolution across cosmic time scales. We compress galaxy images using principal component analysis to fit a representation on the quantum hardware. Then, we train RBMs with discriminative and generative algorithms, including contrastive divergence and hybrid generative-discriminative approaches, to classify different galaxy morphologies. The methods we compare include Quantum Annealing (QA), Markov Chain Monte Carlo (MCMC) Gibbs Sampling, and Simulated Annealing (SA) as well as machine learning algorithms like gradient boosted decision trees. We find that RBMs implemented on D-Wave hardware perform well, and that they show some classification performance advantages on small datasets, but they don't offer a broadly strategic advantage for this task. During this exploration, we analyzed the steps required for Boltzmann sampling with the D-Wave 2000Q, including a study of temperature estimation, and examined the impact of qubit noise by comparing and contrasting the original D-Wave 2000Q to the lower-noise version recently made available. While these analyses ultimately had minimal impact on the performance of the RBMs, we include them for reference. | ['João Caldeira', 'Steven H. Adachi', 'Brian Nord', 'Joshua Job', 'Gabriel N. Perdue'] | 2019-11-14 | null | null | null | null | ['morphology-classification'] | ['computer-vision'] | [ 1.99524343e-01 -4.04423982e-01 3.23706180e-01 -2.49426544e-01
-8.25352788e-01 -7.56173909e-01 1.04847312e+00 -1.62030742e-01
-8.36553216e-01 4.70124006e-01 -9.89050344e-02 -6.91345215e-01
-1.32288292e-01 -1.00961900e+00 -3.87359798e-01 -1.21126962e+00
-4.92875576e-02 1.02947640e+00 2.98507631e-01 -1.93071499e-01
8.25072348e-01 7.56300509e-01 -1.30409241e+00 4.56441008e-02
5.58269083e-01 7.96397984e-01 1.44619077e-01 1.16020048e+00
3.10742915e-01 6.84440315e-01 -4.33465689e-01 -3.81294340e-01
2.33207628e-01 -6.30973518e-01 -9.13223922e-01 -3.42133641e-01
1.66086137e-01 -1.53311267e-01 -6.90211177e-01 1.09447753e+00
6.03086114e-01 4.10598725e-01 1.08034086e+00 -6.30182922e-01
-3.05376351e-01 1.78235695e-01 3.99202704e-02 5.27766824e-01
-1.00721151e-01 5.74799359e-01 7.93196559e-01 -4.50887024e-01
5.63248098e-01 1.02675521e+00 5.50070584e-01 5.79729259e-01
-1.53656197e+00 -5.03986537e-01 -1.08905554e+00 3.28233480e-01
-1.42794478e+00 -6.15571380e-01 3.69565189e-01 -5.24521530e-01
1.57644737e+00 4.17217314e-01 1.03788328e+00 7.82705843e-01
7.21642792e-01 1.07834570e-01 1.70784438e+00 -5.57447493e-01
8.27647626e-01 -7.76695386e-02 1.27905667e-01 9.38543200e-01
1.67037547e-01 4.47754711e-01 -3.82020652e-01 -5.30526340e-01
6.80557430e-01 -5.64047873e-01 1.99411660e-01 -1.01682790e-01
-1.00634229e+00 9.25619185e-01 3.12015682e-01 2.64338136e-01
7.15198228e-04 5.43463826e-01 5.16012013e-02 2.55262256e-02
2.40073875e-01 9.12558973e-01 1.30458757e-01 -5.33648849e-01
-1.10220885e+00 2.63132006e-01 9.33548212e-01 6.30086720e-01
9.81687546e-01 -3.42892222e-02 1.73352405e-01 4.20418978e-01
2.74364948e-01 7.31663644e-01 5.91789424e-01 -1.08089328e+00
-1.73676297e-01 -1.25676349e-01 -8.18997994e-02 -6.42158031e-01
-3.70603323e-01 -2.78280854e-01 -8.82100523e-01 1.82290852e-01
2.69030899e-01 2.48029605e-01 -1.01182175e+00 1.21192622e+00
2.18179226e-02 1.15447246e-01 6.00035023e-03 5.96250355e-01
5.93828738e-01 8.49656641e-01 -2.22621515e-01 6.72547892e-02
1.23947358e+00 -5.37288427e-01 -1.82987172e-02 -2.94473648e-01
8.20738673e-01 -5.96614480e-01 6.38288200e-01 5.99515080e-01
-8.87087464e-01 -3.60052913e-01 -1.30980134e+00 -3.63889243e-03
-2.50056267e-01 -4.25990582e-01 9.31021631e-01 1.42445230e+00
-1.14051533e+00 1.06402898e+00 -1.34324205e+00 -2.60561407e-01
2.48320147e-01 4.29508150e-01 -1.17720641e-01 1.22416265e-01
-9.16031241e-01 1.03934956e+00 2.71912336e-01 -1.83973446e-01
-1.13299525e+00 2.48467326e-01 -4.03423667e-01 -1.72210410e-01
-2.55772829e-01 -8.34921777e-01 1.26827705e+00 -2.68083632e-01
-1.90614295e+00 1.10678852e+00 -7.48100430e-02 -6.00873351e-01
2.06269920e-02 5.29898107e-01 -1.36199221e-01 1.02800436e-01
-4.69485939e-01 4.58782047e-01 6.57786727e-01 -6.20910406e-01
-1.72645539e-01 -4.58952874e-01 -2.33909547e-01 1.00812679e-02
1.63407117e-01 -5.40440381e-02 -2.85225004e-01 3.65672074e-02
3.34175944e-01 -1.46692884e+00 -3.34388345e-01 -9.32642400e-01
-1.63631707e-01 -1.27396107e-01 6.00816607e-02 -5.92896760e-01
8.55815947e-01 -1.93198502e+00 3.20585221e-01 3.66577864e-01
2.25011297e-02 -7.63020515e-02 2.37257466e-01 3.83731604e-01
3.78079593e-01 5.30885644e-02 -3.74352217e-01 -3.21526915e-01
1.47947997e-01 3.86778682e-01 -1.81008682e-01 9.38454688e-01
-2.16532335e-01 9.09246624e-01 -6.67986214e-01 -3.39841783e-01
3.12667549e-01 4.37568575e-02 -7.85462379e-01 -2.37144269e-02
-3.79618667e-02 5.08709252e-01 -1.60568535e-01 4.45862025e-01
4.92037386e-01 -2.41737187e-01 1.09311536e-01 -9.87637267e-02
-1.90297022e-01 6.65977061e-01 -7.66454875e-01 1.72516561e+00
-3.80848795e-01 8.02638412e-01 -7.70112872e-02 -8.69294822e-01
8.83554041e-01 -2.12657511e-01 -1.44360498e-01 -8.95796895e-01
1.77942783e-01 4.93460327e-01 3.68940234e-01 -2.74817765e-01
8.80729020e-01 -5.82650781e-01 -1.52372658e-01 5.94078481e-01
3.71466309e-01 -1.10760999e+00 5.14460802e-02 1.91952422e-01
1.26477575e+00 -1.10972114e-02 1.56425133e-01 -4.52721268e-01
-7.26467445e-02 2.14190051e-01 8.17529857e-02 1.21273756e+00
-1.46724448e-01 7.32778370e-01 2.40723178e-01 -2.59255558e-01
-1.48316109e+00 -1.02247393e+00 -4.58283067e-01 8.45174670e-01
1.84722498e-01 -5.63397408e-01 -7.51702309e-01 -8.22879970e-02
-2.86870152e-01 9.79046583e-01 -3.91460776e-01 -4.56253022e-01
-3.81381810e-01 -1.73132384e+00 8.04239690e-01 -1.08709611e-01
3.92336756e-01 -8.10974479e-01 -8.96711707e-01 5.16814366e-02
2.81914979e-01 -6.06263399e-01 2.21660748e-01 6.72175527e-01
-9.97109830e-01 -6.43462241e-01 -2.58725762e-01 -4.53805506e-01
2.63809174e-01 -1.49368972e-01 1.18698657e+00 -2.95993894e-01
-5.89798748e-01 3.84257317e-01 -2.74846286e-01 -4.40290347e-02
-9.24636662e-01 1.92809939e-01 2.54107356e-01 -5.48280001e-01
3.15140903e-01 -5.92433274e-01 -7.86933303e-01 1.35716215e-01
-7.46755779e-01 2.06385180e-01 6.69186354e-01 9.00326788e-01
3.89782429e-01 1.31567910e-01 -3.88480157e-01 -5.67146599e-01
3.53260428e-01 -2.00589851e-01 -8.03048730e-01 -5.73455617e-02
-7.81884611e-01 4.66553032e-01 3.94865453e-01 2.31083743e-02
-8.01595509e-01 -2.62842566e-01 -6.07168257e-01 1.39826268e-01
-6.08061366e-02 2.25236848e-01 2.82947987e-01 -5.88146925e-01
9.21126246e-01 6.30393565e-01 -1.47631735e-01 -3.92007291e-01
3.22883755e-01 9.38748837e-01 4.97496992e-01 -6.45837545e-01
4.59628284e-01 5.72768867e-01 4.08148617e-01 -1.11022472e+00
-2.55570352e-01 -2.02414721e-01 -3.93370152e-01 -5.14298938e-02
1.00912011e+00 -4.61593270e-01 -6.70808315e-01 6.67817235e-01
-6.63332283e-01 -4.27128017e-01 -2.74069458e-01 7.40658045e-01
-7.85961151e-01 5.39168417e-01 -9.83931601e-01 -8.58253002e-01
-2.44606182e-01 -1.41485929e+00 8.55612278e-01 2.58428305e-01
-7.70991528e-03 -7.28892028e-01 6.42335832e-01 4.42983955e-01
5.18985748e-01 -2.75562376e-01 1.16882539e+00 -5.75131178e-01
-6.69460893e-01 -2.19020143e-01 1.83519065e-01 1.73763409e-01
-3.77311230e-01 5.31787090e-02 -1.03077626e+00 -7.13402212e-01
2.88025379e-01 -5.09905279e-01 1.17121315e+00 3.04755688e-01
9.80222344e-01 8.82245153e-02 -1.96087942e-01 8.72331977e-01
1.41371298e+00 3.30843210e-01 1.02823627e+00 4.90998864e-01
3.29964906e-01 3.16410512e-01 3.05677904e-03 2.38202646e-01
1.14261828e-01 5.56475997e-01 3.93254906e-01 4.30292428e-01
1.15426838e-01 1.46143869e-01 3.86970550e-01 1.34529090e+00
-3.14526320e-01 -1.02082841e-01 -1.20815837e+00 7.09427446e-02
-1.27835190e+00 -1.02269697e+00 -7.67810568e-02 2.44615602e+00
6.17944241e-01 3.39557052e-01 -7.88640156e-02 -2.10810140e-01
3.73720050e-01 5.23640886e-02 -3.82695585e-01 -6.52514815e-01
-1.71569899e-01 8.65755439e-01 8.55810761e-01 3.76557708e-01
-9.22702551e-01 7.01695085e-01 7.28058624e+00 1.44114506e+00
-1.06705117e+00 3.30602258e-01 8.72846842e-01 -1.22049667e-01
-2.71608651e-01 4.85706806e-01 -6.20427728e-01 5.17005444e-01
1.57585800e+00 2.58667856e-01 9.20730412e-01 5.55866063e-01
-3.47449332e-01 -6.49309814e-01 -1.05335391e+00 1.20870078e+00
-1.74946442e-01 -1.53967094e+00 -1.76478818e-01 4.47886258e-01
7.09285498e-01 6.72907114e-01 1.20449997e-01 3.86267722e-01
4.67805982e-01 -1.22805393e+00 7.47270167e-01 5.44593692e-01
8.23807180e-01 -7.60315776e-01 7.00011611e-01 4.19218898e-01
-5.53273499e-01 2.43348062e-01 -8.16009641e-01 -2.14614272e-01
-1.61343530e-01 3.73183906e-01 -8.84414256e-01 2.77519643e-01
5.38282037e-01 4.44938131e-02 -8.74755085e-01 8.41398358e-01
1.36251450e-01 1.02677965e+00 -5.96776128e-01 -6.53007209e-01
3.91118079e-01 -8.10418367e-01 5.28648555e-01 1.28464746e+00
5.65129995e-01 2.85647977e-02 -3.63557696e-01 8.99905205e-01
2.68255621e-01 -2.80326515e-01 -3.88420403e-01 -3.00853431e-01
3.51436883e-01 1.28216374e+00 -1.26283360e+00 -5.12127876e-01
-7.15046003e-02 1.02128625e+00 1.82726070e-01 -1.19442716e-01
-5.67319155e-01 -4.14873809e-01 1.81894541e-01 -7.60446712e-02
2.26624519e-01 -6.13108516e-01 -1.92966864e-01 -1.15338111e+00
-8.04056764e-01 -7.07931817e-01 8.01293924e-02 -9.31690395e-01
-1.04864609e+00 4.80969012e-01 -4.13124599e-02 -6.83883250e-01
-4.66198504e-01 -8.40393305e-01 -3.98797274e-01 8.72998059e-01
-6.91611826e-01 -5.66814244e-01 2.71542072e-02 7.77634531e-02
-1.17528895e-02 -2.38767624e-01 9.99092281e-01 -1.72167227e-01
-4.42550182e-01 1.33910701e-01 1.08456147e+00 -1.15458302e-01
4.09156919e-01 -1.48109925e+00 8.38116229e-01 7.40728736e-01
6.22668684e-01 5.98989010e-01 1.04329777e+00 -5.41619599e-01
-1.85188663e+00 -6.55626774e-01 1.55417204e-01 -5.38678467e-01
4.69994634e-01 -5.76574385e-01 -5.69431663e-01 3.94445807e-01
2.26455748e-01 -2.43537426e-01 6.88068151e-01 -1.73529144e-02
-1.05195157e-01 1.86645374e-01 -1.04860127e+00 4.29608107e-01
7.57621288e-01 -1.15487087e+00 -6.51959896e-01 3.07525277e-01
-7.93903600e-03 -1.58853665e-01 -6.94399416e-01 1.49092162e-02
7.00859308e-01 -1.35618913e+00 8.61558378e-01 -3.54539275e-01
2.47628957e-01 -1.73772052e-01 -3.69240135e-01 -1.32338023e+00
-7.04419494e-01 -6.46279752e-01 2.12815821e-01 6.94955766e-01
1.94224402e-01 -7.22365379e-01 1.08683252e+00 4.12434369e-01
-1.43825099e-01 -4.47072983e-01 -1.46246278e+00 -7.48659134e-01
5.16462803e-01 -4.74517137e-01 2.54630774e-01 6.44668818e-01
-1.52554080e-01 2.94911414e-01 -9.42122415e-02 -1.17057331e-01
7.36798406e-01 3.38899434e-01 6.16572142e-01 -7.45297194e-01
-9.41186488e-01 -5.47806084e-01 -7.13378549e-01 -1.04367197e+00
-2.18667626e-01 -1.34729683e+00 2.06880912e-01 -1.03244054e+00
6.54622138e-01 -5.36186457e-01 -9.60338041e-02 -2.43991300e-01
1.81371614e-01 5.81051826e-01 -1.39392093e-01 5.35183847e-01
-6.18487179e-01 4.24984753e-01 6.57884181e-01 -2.00656831e-01
2.31102020e-01 -2.03427583e-01 -2.31693797e-02 4.71343696e-01
6.69668972e-01 -7.36878872e-01 2.22952619e-01 -1.57690972e-01
6.33077085e-01 1.83359105e-02 4.92142916e-01 -1.42146432e+00
3.31045151e-01 1.08642504e-01 4.17730629e-01 -4.07576948e-01
5.89072108e-01 6.23104535e-02 5.42103708e-01 9.32825446e-01
-4.80300002e-02 -2.29734946e-02 -5.02274595e-02 4.34049159e-01
1.80733576e-01 -8.83882999e-01 1.01464283e+00 -5.81455112e-01
-5.46767235e-01 -1.21032586e-03 -1.12446785e+00 -1.45862788e-01
5.64206302e-01 -1.82759196e-01 -3.03729206e-01 -2.63549596e-01
-4.94179994e-01 -4.35597748e-01 1.04099786e+00 -4.06026870e-01
2.79252589e-01 -9.64969039e-01 -5.24278939e-01 3.31219912e-01
-7.15143681e-02 -3.13547999e-01 2.07469553e-01 7.35711157e-01
-1.11606002e+00 5.40675282e-01 -3.41129601e-01 -7.46673703e-01
-1.05274022e+00 3.67646515e-01 5.53205311e-01 -2.80328363e-01
-1.93770692e-01 1.17328238e+00 -1.31061062e-01 -5.04036546e-01
-4.66179937e-01 -2.90127397e-01 2.06590176e-01 -1.75332323e-01
9.33493823e-02 5.49782574e-01 3.81855816e-01 -5.77195525e-01
-3.14855307e-01 3.89956117e-01 2.93628186e-01 -6.09788358e-01
1.17197835e+00 4.43395264e-02 -4.97712344e-01 6.50365591e-01
1.09544683e+00 9.14827548e-03 -8.29511702e-01 1.14388831e-01
-1.64491355e-01 -2.68595785e-01 3.26670796e-01 -3.54175180e-01
-3.38718742e-01 1.14602518e+00 8.72280061e-01 5.53921700e-01
7.42499590e-01 1.45999178e-01 5.29790044e-01 7.47066379e-01
7.77057588e-01 -1.02947319e+00 -2.28047132e-01 5.76840699e-01
1.26014262e-01 -8.04956794e-01 3.61364305e-01 2.68733948e-01
-3.58550064e-02 1.27408075e+00 -1.31060435e-02 -2.86160767e-01
3.66215676e-01 1.49066418e-01 -6.24032497e-01 -4.08856362e-01
-6.87305510e-01 -4.20756713e-02 6.08397499e-02 2.07889900e-01
6.57403246e-02 3.09285283e-01 7.25713298e-02 7.87815899e-02
-6.39998376e-01 -3.60748351e-01 6.66107237e-01 1.06408978e+00
-5.63110113e-01 -9.73798096e-01 -5.43766439e-01 7.69257903e-01
-1.92170545e-01 -3.35693657e-01 -2.53314495e-01 3.56267601e-01
7.16174841e-02 6.08634412e-01 3.65327895e-01 -7.52268255e-01
-3.73419106e-01 2.52134711e-01 1.28834438e+00 -4.56188023e-01
-4.58531886e-01 -9.80984941e-02 2.41525292e-01 -2.57858008e-01
-2.54172593e-01 -8.56547356e-01 -7.88111389e-01 -7.12028503e-01
-6.49928272e-01 4.22402889e-01 1.04154956e+00 8.88789654e-01
7.02358112e-02 1.85528621e-01 4.39812422e-01 -1.27719104e+00
-8.45912218e-01 -9.82643604e-01 -9.28772926e-01 1.05263650e-01
1.70160364e-02 -2.08315969e-01 -5.89296937e-01 -3.43420506e-01] | [5.577632427215576, 4.8739495277404785] |
4e03f5a2-13af-4523-a9f5-7bd05f3d75c9 | toward-macro-insights-for-suicide-prevention | null | null | https://aclanthology.org/W14-3213 | https://aclanthology.org/W14-3213.pdf | Toward Macro-Insights for Suicide Prevention: Analyzing Fine-Grained Distress at Scale | null | ['Tong Liu', 'Megan Lytle', 'Vincent Silenzio', 'Christopher Homan', 'Ravdeep Johar', 'Cecilia Ovesdotter Alm'] | 2014-06-01 | null | null | null | ws-2014-6 | ['lexical-analysis'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.260907173156738, 3.8529069423675537] |
1af8339c-c731-49d5-94d8-12bbb8147ecd | challenges-and-frontiers-in-abusive-content | null | null | https://aclanthology.org/W19-3509 | https://aclanthology.org/W19-3509.pdf | Challenges and frontiers in abusive content detection | Online abusive content detection is an inherently difficult task. It has received considerable attention from academia, particularly within the computational linguistics community, and performance appears to have improved as the field has matured. However, considerable challenges and unaddressed frontiers remain, spanning technical, social and ethical dimensions. These issues constrain the performance, efficiency and generalizability of abusive content detection systems. In this article we delineate and clarify the main challenges and frontiers in the field, critically evaluate their implications and discuss potential solutions. We also highlight ways in which social scientific insights can advance research. We discuss the lack of support given to researchers working with abusive content and provide guidelines for ethical research. | ['Helen Margetts', 'Scott Hale', 'Rebekah Tromble', 'Dong Nguyen', 'Alex Harris', 'Bertie Vidgen'] | 2019-08-01 | null | null | null | ws-2019-8 | ['abuse-detection'] | ['natural-language-processing'] | [ 6.80707172e-02 1.63923576e-01 -4.15723920e-01 -3.18023562e-01
-4.65915293e-01 -7.46197164e-01 1.57280043e-01 7.28467941e-01
-7.27235973e-01 6.10978365e-01 3.91450167e-01 -3.55833024e-01
-8.32378119e-03 -2.08819270e-01 -7.54944980e-02 -1.66202322e-01
2.34466955e-01 -1.41310781e-01 6.59236237e-02 -2.73022145e-01
7.84149647e-01 4.24990475e-01 -1.33083916e+00 6.01340048e-02
1.06412709e+00 3.46011877e-01 -1.28069609e-01 2.02764675e-01
-4.39620435e-01 9.76732612e-01 -1.03285015e+00 -1.00935793e+00
-2.51200318e-01 -4.80427355e-01 -9.02041912e-01 1.10924326e-01
4.77942079e-01 -7.15579510e-01 -6.32585883e-02 1.28157151e+00
3.82024437e-01 6.20077774e-02 4.25128758e-01 -1.20719814e+00
-1.06609440e+00 5.55046260e-01 -6.46721780e-01 9.10698175e-01
5.57960153e-01 -3.63259651e-02 9.13952887e-01 -6.44804001e-01
9.01258945e-01 1.18809223e+00 4.97935683e-01 4.88847286e-01
-8.59115362e-01 -8.68056178e-01 4.60912198e-01 3.65883678e-01
-1.12774003e+00 -5.25508046e-01 5.96127808e-01 -7.27244616e-01
8.45384955e-01 6.22092560e-02 8.83131981e-01 1.44896412e+00
-1.16220519e-01 6.30387068e-01 9.78629410e-01 -3.64854038e-01
1.43825253e-02 5.66013277e-01 5.53766966e-01 3.98163527e-01
9.31480527e-01 -6.25743091e-01 -5.87181091e-01 -5.47061443e-01
2.71653116e-01 -2.65213281e-01 -2.98860390e-02 2.60347933e-01
-3.76052737e-01 1.29631400e+00 1.07283248e-02 9.10823107e-01
-1.84394643e-01 -2.28710979e-01 9.15559888e-01 9.77234319e-02
1.05532765e+00 4.60736036e-01 2.76032895e-01 -8.02705348e-01
-8.48909974e-01 3.38655591e-01 9.26768780e-01 5.11537910e-01
1.44124866e-01 1.09689429e-01 1.40473589e-01 1.26516771e+00
4.29221809e-01 3.23110610e-01 2.44009271e-01 -9.42522168e-01
4.11279291e-01 4.81601983e-01 9.91274118e-02 -1.72346830e+00
-3.19812655e-01 -2.70022184e-01 4.27508354e-02 -1.10259481e-01
4.86871660e-01 -9.00090560e-02 -1.58701986e-01 1.41459942e+00
3.58540177e-01 -2.51462162e-01 -6.60414398e-01 9.89059210e-01
6.98386133e-01 3.47884566e-01 6.88095331e-01 -6.07682407e-01
1.33241177e+00 -3.35783005e-01 -1.29300845e+00 -6.73082829e-01
7.39771605e-01 -1.08610582e+00 1.07119143e+00 3.09837729e-01
-1.41465628e+00 3.02863330e-01 -1.01758313e+00 -4.63934034e-01
-3.28742802e-01 -4.98335600e-01 4.14714485e-01 1.31114376e+00
-6.90121412e-01 4.51091588e-01 -7.53775477e-01 -8.51454794e-01
6.70657694e-01 -1.79016456e-01 -6.42124638e-02 8.86523649e-02
-1.19592309e+00 1.20175612e+00 1.30495233e-02 5.80859743e-02
-3.09616208e-01 -5.69099903e-01 -7.13522971e-01 -3.60326856e-01
4.55840200e-01 -1.03661269e-01 1.31798613e+00 -8.67114484e-01
-1.20630705e+00 1.37342525e+00 -1.46252409e-01 -2.13333130e-01
4.61214930e-01 -3.24389458e-01 -6.12761021e-01 4.81505781e-01
5.60873151e-01 1.92279853e-02 4.39315349e-01 -9.87512708e-01
-6.18190765e-01 -4.81329739e-01 -1.41628906e-02 3.63513827e-04
-9.51305866e-01 1.14793265e+00 2.69843042e-01 -6.01208270e-01
8.96812081e-02 -8.26532841e-01 1.44369856e-01 1.12288438e-01
-8.73709694e-02 -3.80209386e-01 8.22027981e-01 -8.12507868e-01
1.65002573e+00 -2.32208490e+00 -3.75487953e-01 -5.32115325e-02
5.86408615e-01 3.28612328e-01 5.43631732e-01 7.59805202e-01
2.44730353e-01 7.68734157e-01 9.46212560e-03 -2.99093306e-01
-7.31204525e-02 1.14625104e-01 -2.11650759e-01 1.04535317e+00
3.04582287e-02 6.43384457e-01 -1.20507538e+00 -6.61062002e-01
-8.46393108e-02 3.40728432e-01 -3.41362506e-01 4.82272431e-02
2.10285082e-01 -5.29072098e-02 -5.22685528e-01 7.24817932e-01
7.07752168e-01 -1.76774547e-01 2.10987225e-01 5.07925093e-01
-5.85941076e-01 7.41123319e-01 -7.21146524e-01 9.37564671e-01
-5.65986596e-02 1.14617467e+00 7.24857569e-01 -6.81349277e-01
8.46319437e-01 5.16319796e-02 3.03699434e-01 -3.85477185e-01
3.42641205e-01 3.78992826e-01 3.16663623e-01 -9.44380701e-01
7.58279085e-01 -5.06094337e-01 1.59301981e-01 9.33051050e-01
-3.71306360e-01 2.83276755e-02 -1.93428751e-02 5.40279269e-01
8.65419149e-01 -1.64038852e-01 4.78021622e-01 -2.02645391e-01
2.46547252e-01 4.95704748e-02 4.14766788e-01 6.02211535e-01
-1.09417260e+00 3.76626663e-03 9.41482544e-01 7.33918026e-02
-9.08813477e-01 -4.99078631e-01 -4.16029930e-01 1.36963046e+00
7.36979097e-02 -5.61996937e-01 -8.73682261e-01 -6.22145772e-01
4.45225611e-02 9.37977612e-01 -6.46960437e-01 -1.09579608e-01
-5.78354299e-01 -7.37313867e-01 6.96187913e-01 1.20854378e-02
2.27008507e-01 -1.01220858e+00 -7.16089368e-01 2.15046003e-01
-4.30688471e-01 -1.45188701e+00 -1.31674167e-02 -4.89065230e-01
-7.62195349e-01 -1.04150689e+00 -2.55385637e-01 -4.32613671e-01
2.98910946e-01 9.42467213e-01 7.33146846e-01 4.29919183e-01
-3.11638504e-01 5.07799804e-01 -6.65523589e-01 -7.14976609e-01
-6.75272584e-01 -8.42109993e-02 -5.55161647e-02 -2.30900928e-01
1.04955721e+00 -3.60292822e-01 -3.92538637e-01 1.42319322e-01
-8.52906346e-01 -7.41096556e-01 4.37283516e-02 2.89103836e-01
-1.15022309e-01 -3.50557953e-01 9.29275155e-01 -1.11708045e+00
1.14976704e+00 -1.13865757e+00 1.14223368e-01 -3.19951564e-01
-8.11967850e-01 -1.05019951e+00 6.28610775e-02 -2.78004646e-01
-1.10754526e+00 -9.41481769e-01 -2.37082452e-01 8.24088156e-02
-2.96668679e-01 4.72337842e-01 2.17934012e-01 1.71530731e-02
1.15681076e+00 -5.38437665e-01 3.41363966e-01 -3.88309151e-01
-1.40891209e-01 1.03422976e+00 -1.52732149e-01 -3.38634819e-01
5.27967334e-01 6.36738479e-01 -5.16635358e-01 -1.31634414e+00
-1.11035311e+00 -5.72741091e-01 -3.05214316e-01 -6.14525318e-01
6.89604282e-01 -6.87845170e-01 -6.74169779e-01 2.35547185e-01
-1.08193886e+00 -8.11162665e-02 1.10636614e-01 2.55748868e-01
2.48253532e-03 8.69722247e-01 -9.71919358e-01 -1.17515063e+00
-3.51896524e-01 -7.03614354e-01 4.85081166e-01 1.61079451e-01
-9.67464685e-01 -1.09662569e+00 -2.16029868e-01 9.24317479e-01
3.51874501e-01 3.46112758e-01 5.72373569e-01 -6.25694633e-01
5.98024614e-02 -3.96982521e-01 -1.60828501e-01 1.99663669e-01
4.38552871e-02 5.08640170e-01 -9.01532650e-01 -4.04870212e-02
3.52818251e-01 -5.28817832e-01 2.95617968e-01 1.36055157e-01
6.48187518e-01 -5.22198737e-01 -3.05229127e-01 -1.85667038e-01
1.03237712e+00 5.87033927e-02 2.85677999e-01 7.12938607e-01
4.04074162e-01 1.23733866e+00 8.21700931e-01 5.95776677e-01
3.84601772e-01 2.94645697e-01 3.63998562e-01 5.78820884e-01
1.54040962e-01 -2.96821028e-01 4.68519121e-01 6.24172330e-01
2.56894410e-01 -1.57967135e-01 -9.48132217e-01 6.35737360e-01
-1.58583272e+00 -1.36438572e+00 -7.02849209e-01 1.70548499e+00
6.51272416e-01 3.62232149e-01 7.06391811e-01 3.41093451e-01
9.78338599e-01 3.12689692e-01 -2.84911603e-01 -1.01730311e+00
-1.92969739e-01 -3.28951269e-01 -6.81888089e-02 6.71004117e-01
-5.81827164e-01 1.14603698e+00 7.16528511e+00 5.08110642e-01
-9.71615374e-01 4.71816629e-01 4.47284400e-01 -3.11957717e-01
-3.42211008e-01 -1.44480169e-01 -7.65244305e-01 4.85231161e-01
9.26544070e-01 -5.81424057e-01 1.49619326e-01 1.01612544e+00
6.37458384e-01 -4.07502979e-01 -4.16489482e-01 4.98458862e-01
4.71056163e-01 -8.77795517e-01 -6.95519090e-01 3.27639878e-01
2.98234612e-01 9.38446522e-02 1.71282381e-01 2.61710100e-02
-2.70282533e-02 -7.05823004e-01 7.50961900e-01 -1.64949253e-01
4.11773711e-01 -3.98331463e-01 5.83525181e-01 1.99423000e-01
-1.77792668e-01 -1.43573880e-01 -4.91588831e-01 -6.79989636e-01
3.30164462e-01 6.21338010e-01 -6.77191257e-01 -2.21168846e-01
1.04684687e+00 7.47752488e-01 -5.61998844e-01 8.38589191e-01
-4.95182991e-01 1.11276567e+00 -1.39948443e-01 -5.61530232e-01
2.82507837e-01 -3.46871436e-01 6.52741849e-01 1.48584390e+00
-1.92353293e-01 3.08188707e-01 -2.66961783e-01 7.96410084e-01
1.26694083e-01 5.66285431e-01 -7.21865356e-01 -6.49749517e-01
8.44415903e-01 1.46126604e+00 -1.05808592e+00 -6.47856519e-02
-6.04074955e-01 5.40763855e-01 8.33099782e-01 1.44269884e-01
-6.62474871e-01 -1.21000826e-01 8.73231232e-01 4.83851820e-01
-2.75528580e-01 -2.58251637e-01 -7.37480640e-01 -8.97689819e-01
3.34637053e-02 -1.02803957e+00 6.86060846e-01 -3.26789081e-01
-1.43233109e+00 2.06291914e-01 1.94502115e-01 -7.59332180e-01
-1.34478822e-01 -4.33617264e-01 -2.82353669e-01 5.42867184e-01
-9.99671459e-01 -7.48559833e-01 -2.60153532e-01 -1.80745751e-01
2.71161467e-01 3.34792316e-01 5.50267696e-01 1.97508201e-01
-9.53479528e-01 6.28673077e-01 -4.42880429e-02 -7.60884285e-02
8.54503870e-01 -6.07510090e-01 -1.37069389e-01 7.03019559e-01
-3.03632915e-01 9.48103964e-01 1.13610184e+00 -8.57029974e-01
-9.29962814e-01 -5.11095822e-01 1.17995393e+00 -4.47220355e-01
1.41815901e+00 -4.66014981e-01 -1.20175707e+00 8.93791080e-01
3.68672311e-01 -6.31718159e-01 1.55679214e+00 3.94983083e-01
-6.32488728e-01 4.06559259e-01 -1.43436861e+00 8.77509713e-01
1.17923307e+00 -5.40296733e-01 -5.88545084e-01 6.74652159e-01
2.90803581e-01 -1.90495208e-01 -8.13456118e-01 -4.06529605e-01
5.10440052e-01 -1.11597347e+00 4.26518470e-01 -6.26855373e-01
5.97109854e-01 4.13929015e-01 3.09535146e-01 -7.49705791e-01
-5.02757132e-01 -6.66109443e-01 8.89791697e-02 1.42717707e+00
9.80461091e-02 -8.39113414e-01 7.74776936e-01 1.20577788e+00
-1.61905065e-01 -5.50128281e-01 -8.56305957e-01 -4.87302452e-01
6.21095896e-01 -6.96898162e-01 -8.02758802e-03 1.48879445e+00
9.31545794e-01 3.75216872e-01 -1.90669283e-01 -2.12792307e-01
3.76685262e-01 -2.21322522e-01 5.75111151e-01 -1.09857464e+00
3.54203850e-01 -8.25179935e-01 -3.88767838e-01 -4.84840572e-01
3.64639312e-01 -6.01981878e-01 -5.80149174e-01 -1.48727429e+00
2.99225211e-01 5.59177957e-02 3.06793362e-01 3.10716450e-01
-4.26630408e-01 5.11882305e-01 2.84058541e-01 2.14010373e-01
-3.92188340e-01 1.48512155e-01 9.39664185e-01 3.87601584e-01
-2.44569138e-01 -4.99496222e-01 -1.56538165e+00 7.15664625e-01
1.17207050e+00 -5.71331680e-01 -2.08745822e-01 -2.23660216e-01
6.59877241e-01 -4.92648751e-01 8.54604915e-02 -3.96189451e-01
-2.55649188e-03 -5.80030918e-01 6.01472557e-02 -8.01057667e-02
2.12531701e-01 -5.54069281e-01 -3.37222815e-01 2.73165882e-01
-2.97615618e-01 8.87080804e-02 3.49577457e-01 3.23450536e-01
2.90124733e-02 -9.77836847e-01 9.10831571e-01 7.47175142e-02
-1.29850075e-01 -2.21294478e-01 -9.50639844e-01 2.88788408e-01
1.27553856e+00 -3.93571079e-01 -6.90087914e-01 -7.79731750e-01
-5.96976578e-01 1.01031817e-01 5.22622705e-01 4.57397938e-01
4.26358700e-01 -7.87760854e-01 -7.26483047e-01 -4.88627195e-01
1.83605358e-01 -7.22728431e-01 4.91952330e-01 9.98692036e-01
-5.81436276e-01 1.99312419e-01 -7.48599395e-02 7.93951377e-03
-1.22917986e+00 5.96402109e-01 9.17705819e-02 5.35296261e-01
-6.31382763e-01 5.76247334e-01 -3.21983516e-01 3.10314178e-01
1.02636479e-01 5.41599154e-01 -6.78979337e-01 6.01727784e-01
8.09513748e-01 1.02772546e+00 -2.74682730e-01 -1.17872894e+00
-3.28367949e-01 -9.71435830e-02 -5.97559094e-01 -2.50036716e-01
1.16243851e+00 -5.90713620e-01 -4.81733590e-01 6.99560285e-01
1.15137160e+00 3.33620846e-01 -4.08273131e-01 1.79021209e-02
2.89365619e-01 -9.37611997e-01 -6.68942137e-03 -5.39201200e-01
-5.65406024e-01 7.51825333e-01 -1.75757304e-01 9.47073102e-01
5.79870403e-01 2.05817476e-01 6.57419741e-01 -2.28605449e-01
7.17283562e-02 -1.51819479e+00 3.16826284e-01 3.30981642e-01
1.13327861e+00 -1.06679237e+00 3.57343435e-01 -1.03659129e+00
-5.92012763e-01 1.02579200e+00 7.92026222e-01 -7.82970265e-02
8.51458371e-01 3.27171795e-02 3.33274156e-01 -4.77706283e-01
-3.51961702e-01 -6.34695888e-02 -3.08173418e-01 6.00591302e-01
1.06759012e+00 -2.55584270e-01 -1.53752422e+00 4.54823941e-01
-4.15898502e-01 -2.51385450e-01 1.02203250e+00 1.01205087e+00
-6.32704496e-01 -1.06612086e+00 -5.25152087e-01 5.75195074e-01
-1.19974995e+00 1.59373716e-01 -1.05400133e+00 1.02165604e+00
-1.91993818e-01 1.49678397e+00 -1.21928915e-01 -1.44072786e-01
2.28598729e-01 1.06652744e-01 1.09290808e-01 -9.41102564e-01
-8.19172561e-01 2.16645032e-01 6.68003857e-01 -2.57721126e-01
-2.74290204e-01 -1.26462710e+00 -9.18539941e-01 -8.21565747e-01
-4.83821154e-01 2.24289536e-01 7.99450159e-01 9.50085223e-01
5.34382761e-01 -8.46823081e-02 2.95622587e-01 -1.91426188e-01
-6.36239052e-01 -1.16218638e+00 -6.02488637e-01 2.26336822e-01
1.31267473e-01 -6.04978681e-01 -7.25781143e-01 -2.96705306e-01] | [8.638519287109375, 10.37199592590332] |
cca6d083-46e7-4e40-ba60-4a5362c01b0c | nodule2vec-a-3d-deep-learning-system-for | 2007.07081 | null | https://arxiv.org/abs/2007.07081v1 | https://arxiv.org/pdf/2007.07081v1.pdf | Nodule2vec: a 3D Deep Learning System for Pulmonary Nodule Retrieval Using Semantic Representation | Content-based retrieval supports a radiologist decision making process by presenting the doctor the most similar cases from the database containing both historical diagnosis and further disease development history. We present a deep learning system that transforms a 3D image of a pulmonary nodule from a CT scan into a low-dimensional embedding vector. We demonstrate that such a vector representation preserves semantic information about the nodule and offers a viable approach for content-based image retrieval (CBIR). We discuss the theoretical limitations of the available datasets and overcome them by applying transfer learning of the state-of-the-art lung nodule detection model. We evaluate the system using the LIDC-IDRI dataset of thoracic CT scans. We devise a similarity score and show that it can be utilized to measure similarity 1) between annotations of the same nodule by different radiologists and 2) between the query nodule and the top four CBIR results. A comparison between doctors and algorithm scores suggests that the benefit provided by the system to the radiologist end-user is comparable to obtaining a second radiologist's opinion. | ['Ilia Kravets', 'Tal Heletz', 'Hayit Greenspan'] | 2020-07-11 | null | null | null | null | ['content-based-image-retrieval', 'lung-nodule-detection'] | ['computer-vision', 'medical'] | [-2.18689777e-02 3.33716005e-01 -2.10917443e-01 -3.28181326e-01
-1.50232553e+00 -4.84379083e-01 4.41871464e-01 5.30331492e-01
-6.37716293e-01 1.07651643e-01 4.14066732e-01 -2.92976439e-01
-5.76085448e-01 -6.04973435e-01 -3.38109910e-01 -7.93983161e-01
8.59511569e-02 1.11832583e+00 3.70361745e-01 1.83886290e-01
-8.25713873e-02 7.49382138e-01 -1.18091512e+00 6.10786676e-01
1.73972901e-02 1.19588768e+00 4.05401260e-01 8.29679966e-01
4.37506586e-02 1.12810171e+00 -5.65308809e-01 -2.41513506e-01
3.31908613e-01 -5.48884809e-01 -1.25137377e+00 1.92014873e-01
5.45158207e-01 -5.49382329e-01 -8.82769346e-01 8.79968166e-01
7.64174700e-01 -3.63288373e-01 1.11089909e+00 -7.62945950e-01
-8.18283081e-01 2.95319200e-01 -3.00616413e-01 3.99886847e-01
1.08840957e-01 -2.48098388e-01 1.07377493e+00 -8.53008330e-01
1.15815926e+00 7.52768934e-01 4.82991457e-01 4.32552040e-01
-7.28270471e-01 2.61152852e-02 -5.44458985e-01 4.29286987e-01
-1.37300122e+00 1.98823154e-01 3.95042628e-01 -5.51387012e-01
5.10156929e-01 4.86717045e-01 7.82737255e-01 7.11882412e-01
5.77262402e-01 7.62246788e-01 6.36037648e-01 -3.66468787e-01
1.68831781e-01 3.57122749e-01 -8.58730376e-02 1.10685205e+00
3.30545992e-01 -9.84352231e-02 -1.33054093e-01 -5.40282488e-01
6.43124759e-01 4.17566329e-01 -2.44723678e-01 -7.05832183e-01
-1.42474926e+00 9.29008543e-01 1.07818806e+00 6.26845062e-01
-5.28288722e-01 2.20039770e-01 5.64789832e-01 4.17522579e-01
2.20051631e-01 6.19530678e-01 -1.25508429e-02 5.80793858e-01
-1.06040967e+00 9.42173973e-03 9.59034503e-01 5.51835597e-01
1.55199707e-01 -5.66702485e-01 -5.76967776e-01 6.11455619e-01
2.37446412e-01 4.95526075e-01 1.06175220e+00 -8.26012254e-01
-4.52144146e-01 4.90090609e-01 -6.56837374e-02 -7.41966844e-01
-2.99977213e-01 -5.17777622e-01 -6.08730197e-01 1.77434891e-01
-1.34215858e-02 3.94437164e-01 -9.83576059e-01 1.05870926e+00
1.38833076e-01 -3.63281071e-01 2.58090086e-02 1.21066284e+00
1.41191721e+00 1.85097530e-01 -1.25254914e-01 -3.40767615e-02
1.74686122e+00 -1.14694905e+00 -5.53433478e-01 2.63307154e-01
7.62327790e-01 -8.21136534e-01 6.37241781e-01 3.97974364e-02
-1.11922097e+00 -4.20527011e-01 -9.01395023e-01 -1.15408093e-01
-1.93247244e-01 5.93576312e-01 1.24294363e-01 2.66658098e-01
-1.49945486e+00 3.89153391e-01 -1.05098915e+00 -6.93097353e-01
2.69516110e-01 3.05727571e-01 -5.59097588e-01 -2.73403347e-01
-8.40593457e-01 1.08734977e+00 2.94009149e-01 -4.84832764e-01
-1.17014742e+00 -7.99368918e-01 -5.32774568e-01 2.44403854e-01
4.69228566e-01 -9.61740792e-01 1.70700991e+00 -5.24950027e-01
-8.84603620e-01 1.51344466e+00 3.69138867e-02 -4.68437612e-01
7.88208842e-01 1.62708059e-01 -2.48258680e-01 7.68656015e-01
2.69982487e-01 7.52938330e-01 8.83299947e-01 -9.74956155e-01
-6.07518077e-01 -2.19964236e-01 -2.17889667e-01 3.14731121e-01
-2.85833865e-01 -2.68395841e-01 -7.44565785e-01 -3.87386620e-01
2.82974511e-01 -1.21329737e+00 -3.76830310e-01 8.81423414e-01
-2.81218171e-01 -2.58959025e-01 8.69769514e-01 -5.60164511e-01
8.04010391e-01 -2.08833957e+00 6.27469644e-02 1.90125585e-01
6.70270324e-01 1.70423165e-01 1.35827977e-02 3.58107388e-01
-1.90797523e-01 1.55362144e-01 7.97245875e-02 6.72409683e-02
-2.70614445e-01 2.30316341e-01 2.34216824e-01 6.25544786e-01
9.95155200e-02 1.17709124e+00 -8.24166059e-01 -9.10325348e-01
2.82477587e-01 5.14069080e-01 -3.17084819e-01 4.07552540e-01
3.07311028e-01 1.66427612e-01 -5.40470302e-01 5.50123751e-01
2.30883420e-01 -1.02127862e+00 2.44280770e-01 -5.00963092e-01
3.98105800e-01 9.17108282e-02 -5.02274871e-01 1.72230697e+00
-3.58914763e-01 8.69108856e-01 -1.11111484e-01 -5.94732881e-01
4.61268067e-01 7.76463389e-01 8.89417827e-01 -6.49503708e-01
8.18592459e-02 2.14192808e-01 1.56606644e-01 -5.95225096e-01
1.43804863e-01 2.78057717e-02 1.15517884e-01 9.60724652e-01
1.40850455e-01 -3.02725255e-01 -9.74076763e-02 3.94632339e-01
1.54647970e+00 -6.94822133e-01 8.38711023e-01 -3.00912499e-01
5.53190887e-01 3.00418764e-01 -1.74186304e-01 1.13622952e+00
-4.62968260e-01 8.30832779e-01 2.78825879e-01 -7.22803891e-01
-1.25443590e+00 -1.31632864e+00 -3.97565335e-01 8.73941958e-01
-2.12006737e-02 -6.24631811e-03 -4.56734300e-01 -1.01654422e+00
1.81672797e-01 1.64777875e-01 -1.12151337e+00 -2.65912414e-01
-3.41694325e-01 -2.47896791e-01 2.25376040e-01 5.03460228e-01
5.43965772e-02 -1.02672338e+00 -1.05074131e+00 -5.14418781e-02
-2.01640576e-01 -7.69979715e-01 -4.68115240e-01 2.45296597e-01
-8.14593315e-01 -1.19620895e+00 -1.26381040e+00 -1.14269578e+00
7.84986794e-01 5.52365541e-01 1.37437057e+00 3.64206612e-01
-1.00077069e+00 9.52112734e-01 -3.99319142e-01 -2.84846246e-01
-1.01015699e+00 7.53190517e-02 -2.30843559e-01 -6.05784178e-01
2.80378848e-01 3.08274150e-01 -1.05500698e+00 2.86543638e-01
-1.21193802e+00 -2.04969317e-01 1.04630005e+00 9.11185265e-01
9.32225823e-01 -8.88749063e-02 5.08005507e-02 -9.44183230e-01
4.64817196e-01 -4.41083491e-01 -2.45108157e-01 4.84125048e-01
-7.72246778e-01 2.63868213e-01 -4.21846099e-03 -2.31752276e-01
-7.13897049e-01 2.02276513e-01 8.42537880e-02 -7.12765098e-01
-6.54476583e-02 5.95211327e-01 7.56394684e-01 -1.95995882e-01
9.30266023e-01 2.50237107e-01 1.80967256e-01 -2.48233899e-01
2.94282734e-01 7.41881907e-01 6.23474538e-01 1.03311181e-01
5.96305728e-01 7.00693011e-01 -2.56539658e-02 -5.62187314e-01
-1.06931067e+00 -1.18380368e+00 -6.87567234e-01 -1.48738638e-01
9.08078969e-01 -8.49280953e-01 -5.05526304e-01 -4.58167017e-01
-9.20341134e-01 3.26393157e-01 -7.03360558e-01 5.88537812e-01
-6.43365264e-01 1.80265397e-01 -8.17696571e-01 -1.42632499e-01
-6.67873323e-01 -1.28704298e+00 1.40136147e+00 -1.19511671e-01
-2.04213887e-01 -8.20835173e-01 3.31556618e-01 1.77533194e-01
4.29859251e-01 -3.42420265e-02 1.18930900e+00 -1.09759724e+00
-4.50532317e-01 -6.82519138e-01 -4.57642794e-01 5.68035096e-02
3.95068377e-01 -1.99591637e-01 -9.35417414e-01 -5.76242030e-01
2.50633359e-01 -2.92325169e-01 9.39789712e-01 6.49431765e-01
1.30478954e+00 -1.43120363e-02 -6.39348209e-01 2.66195625e-01
1.58075011e+00 -6.96346397e-03 2.38275588e-01 3.42028290e-01
4.74019974e-01 3.57445627e-01 3.61275047e-01 1.62295893e-01
-1.42370433e-01 4.01396275e-01 6.02022171e-01 -4.65653479e-01
-4.82118666e-01 4.44678105e-02 -4.22987401e-01 8.44359815e-01
2.56346643e-01 -3.91249657e-01 -1.13603675e+00 7.35714734e-01
-1.56887996e+00 -5.46256125e-01 2.74774402e-01 1.92053020e+00
5.86706042e-01 -2.26070896e-01 -2.55785406e-01 -1.49075314e-01
4.81132686e-01 6.80873841e-02 -6.01028085e-01 -1.10727467e-01
3.49459648e-01 1.54588446e-01 5.34133673e-01 1.38185874e-01
-1.26943648e+00 3.28630716e-01 7.16638994e+00 6.85000420e-01
-1.24904525e+00 3.53488892e-01 5.88513911e-01 -1.93696901e-01
1.11232679e-02 -7.32701361e-01 -7.16414005e-02 -1.68361038e-01
8.34968090e-01 -5.53280532e-01 -3.38196725e-01 1.17611146e+00
-2.10383117e-01 7.06060305e-02 -1.77471054e+00 1.12707531e+00
3.97737980e-01 -1.63375044e+00 3.66516322e-01 2.45122463e-01
7.20511675e-01 4.31538582e-01 5.08899689e-01 7.52452612e-02
5.55298142e-02 -1.03032243e+00 1.68934748e-01 4.99074131e-01
9.65636373e-01 -3.40137005e-01 1.14640832e+00 -3.14700529e-02
-8.74148011e-01 1.65030062e-01 -4.59889591e-01 9.33825672e-01
-3.99590760e-01 1.73276842e-01 -1.78153551e+00 6.24631345e-01
7.86443353e-01 5.19430339e-01 -8.30852866e-01 1.28963733e+00
1.37425154e-01 3.87100369e-01 1.44471750e-01 1.31813273e-01
5.90938210e-01 5.13031065e-01 4.92126882e-01 1.24637699e+00
5.05579770e-01 -2.40150355e-02 1.03274383e-01 6.25224173e-01
-3.62567812e-01 3.68958831e-01 -7.93149650e-01 -1.18797548e-01
9.81305093e-02 1.44187856e+00 -8.36878479e-01 -6.38997316e-01
-3.19050312e-01 1.23561215e+00 9.34053436e-02 -1.82654560e-02
-4.23364311e-01 -5.13458019e-03 2.24651113e-01 1.39731556e-01
5.16201317e-01 4.70588297e-01 2.55533993e-01 -7.05672801e-01
-1.27415761e-01 -7.98147857e-01 7.16747284e-01 -9.94278669e-01
-1.44429314e+00 1.00131488e+00 -2.49845654e-01 -1.64146721e+00
-6.16993070e-01 -8.82592380e-01 -2.84991384e-01 5.97768307e-01
-1.42823923e+00 -8.93813193e-01 -4.86340433e-01 4.38533694e-01
5.09015679e-01 -4.04670060e-01 1.15963531e+00 1.01610683e-01
2.97696710e-01 5.24200141e-01 4.23792899e-01 3.88503373e-01
1.08771801e+00 -1.45601499e+00 -1.65924773e-01 9.20965448e-02
2.98704565e-01 1.40392497e-01 4.33038831e-01 -2.45643228e-01
-1.14267302e+00 -1.19878280e+00 7.46294379e-01 -7.21410036e-01
5.04325747e-01 3.53190511e-01 -9.64355230e-01 5.35736084e-01
2.44787186e-01 4.01308626e-01 9.96909022e-01 -4.62812066e-01
-3.32513392e-01 1.43687844e-01 -1.31556726e+00 4.21619952e-01
3.10669601e-01 -8.32360446e-01 -7.71074712e-01 8.59498084e-01
7.19348729e-01 -4.38644826e-01 -1.16713417e+00 2.72707015e-01
6.34514928e-01 -6.97821736e-01 1.16116440e+00 -7.58789957e-01
7.08654046e-01 1.25687063e-01 -1.59095898e-01 -1.06112897e+00
-7.01614439e-01 3.00916940e-01 2.96827435e-01 -9.21377689e-02
3.87531132e-01 -7.26420581e-02 8.53875101e-01 4.87916648e-01
2.64501050e-02 -7.12559581e-01 -8.21159542e-01 -5.07376730e-01
2.10609928e-01 4.49218042e-02 5.50097711e-02 8.13067675e-01
-3.86523068e-01 -2.78367639e-01 1.93305582e-01 2.00091243e-01
4.63066429e-01 3.10150474e-01 1.87559739e-01 -1.38154399e+00
-3.76385301e-01 -4.80065614e-01 -7.66438901e-01 -4.51557994e-01
-2.34804243e-01 -1.40577495e+00 5.06862924e-02 -1.86757147e+00
9.53765988e-01 4.77553494e-02 -8.89259875e-01 2.99167484e-01
1.48122236e-01 4.78989631e-01 6.67513534e-02 7.81033278e-01
-8.58551323e-01 3.37748304e-02 1.48091531e+00 -5.10917068e-01
2.54107922e-01 -1.41004831e-01 -5.23377717e-01 6.47411942e-01
2.83452272e-01 -1.04114032e+00 -2.59076178e-01 -3.79328012e-01
-3.45787667e-02 4.19625133e-01 4.45502013e-01 -9.77734208e-01
4.52998191e-01 4.61204082e-01 6.62213027e-01 -8.21686924e-01
2.79457152e-01 -1.02160311e+00 -2.18093079e-02 1.13030469e+00
-8.04635346e-01 -1.02202050e-01 5.19498847e-02 8.99870157e-01
-5.24553299e-01 -4.07837629e-01 7.00220346e-01 -6.08791530e-01
-4.68136758e-01 5.12853205e-01 -6.61438286e-01 -2.55604059e-01
9.13891912e-01 1.48640573e-01 -1.80363670e-01 -4.86040145e-01
-1.12895334e+00 -1.25624344e-01 3.33097607e-01 8.72075111e-02
6.94321871e-01 -1.39159250e+00 -1.02657902e+00 -1.36704013e-01
6.23759866e-01 -2.96969265e-01 2.60584783e-02 7.84482896e-01
-9.59708035e-01 8.00408304e-01 -1.41298324e-01 -9.94356215e-01
-1.59083211e+00 6.17191017e-01 5.66267431e-01 -7.09635437e-01
-7.88146615e-01 8.01067173e-01 3.63918990e-01 -3.92138302e-01
2.16722086e-01 -3.49617451e-01 -2.61238571e-02 -8.04742277e-02
5.83852708e-01 -3.91566120e-02 5.83311796e-01 -3.84035021e-01
-4.82375860e-01 3.86691183e-01 -7.35902548e-01 -1.50629416e-01
1.01485646e+00 1.41263187e-01 -5.31040914e-02 3.12594086e-01
1.58070946e+00 -4.36642408e-01 -3.72791946e-01 -8.26109767e-01
-7.88293108e-02 -2.90065259e-01 5.13627708e-01 -8.61118138e-01
-1.08405113e+00 7.57307768e-01 1.34518456e+00 2.47804627e-01
9.06510949e-01 6.72401190e-01 5.25074601e-01 1.18729389e+00
-5.00369743e-02 -5.93562365e-01 4.97136474e-01 4.32578400e-02
9.64925885e-01 -1.77975392e+00 2.12356925e-01 -4.02595811e-02
-8.05210114e-01 1.19756758e+00 7.59424567e-02 -2.80493319e-01
9.20032322e-01 -2.41035558e-02 6.22051537e-01 -6.98282897e-01
-8.36621404e-01 -1.48113355e-01 8.07089925e-01 3.31640512e-01
6.57808244e-01 2.54668742e-01 -1.44959539e-01 5.54017574e-02
2.32437775e-02 -5.09465970e-02 4.25716907e-01 1.08905244e+00
-5.01172423e-01 -7.95434177e-01 -1.98816225e-01 8.16404939e-01
-6.93092287e-01 7.71027952e-02 -6.69320822e-01 9.61483419e-01
-4.30204928e-01 4.59416121e-01 1.21665604e-01 -5.50759137e-02
1.58766553e-01 3.49939317e-02 3.15438747e-01 -9.55690205e-01
-5.52714169e-01 3.40137392e-01 -3.82249922e-01 -6.32649660e-01
-3.16105366e-01 -3.56789798e-01 -1.07587874e+00 2.30142653e-01
-2.03288481e-01 3.13060999e-01 7.82957733e-01 5.30547857e-01
2.05578461e-01 8.42169642e-01 5.66012979e-01 -5.70691168e-01
-6.90104842e-01 -6.08391941e-01 -6.35672688e-01 5.67192912e-01
5.30746579e-01 -3.82102370e-01 -2.50214249e-01 9.04802233e-02] | [14.705159187316895, -1.7364832162857056] |
78a141a3-dc79-4f1b-8352-62b1825f2bfd | short-text-topic-modeling-application-to | 2203.11152 | null | https://arxiv.org/abs/2203.11152v1 | https://arxiv.org/pdf/2203.11152v1.pdf | Short Text Topic Modeling: Application to tweets about Bitcoin | Understanding the semantic of a collection of texts is a challenging task. Topic models are probabilistic models that aims at extracting "topics" from a corpus of documents. This task is particularly difficult when the corpus is composed of short texts, such as posts on social networks. Following several previous research papers, we explore in this paper a set of collected tweets about bitcoin. In this work, we train three topic models and evaluate their output with several scores. We also propose a concrete application of the extracted topics. | ['Hugo Schnoering'] | 2022-03-17 | null | null | null | null | ['topic-models'] | ['natural-language-processing'] | [-7.22127706e-02 3.60332459e-01 -4.94856536e-01 -4.88131046e-01
-6.49837315e-01 -5.47459006e-01 1.32302129e+00 6.05392992e-01
-3.12410116e-01 6.74329877e-01 6.44180000e-01 -1.53829068e-01
1.61175311e-01 -9.64631915e-01 -4.78307933e-01 -2.37685338e-01
-1.31970063e-01 7.43379474e-01 4.89800781e-01 -2.20665902e-01
7.85280228e-01 -2.68097371e-01 -1.14711511e+00 3.82302582e-01
5.69155991e-01 9.11335647e-01 2.48654380e-01 3.65045130e-01
-7.24969983e-01 7.03782380e-01 -7.92005956e-01 -5.88325322e-01
-1.58117622e-01 -2.37269312e-01 -1.10577023e+00 4.07591797e-02
-2.41846800e-01 5.76172955e-02 4.43510264e-02 1.22145581e+00
-6.89281970e-02 1.48405805e-01 7.08393633e-01 -1.46634388e+00
2.27986708e-01 1.33534193e+00 -3.17030758e-01 2.45054886e-01
2.54258752e-01 -4.24784362e-01 1.26658762e+00 -8.46203566e-01
8.02423120e-01 1.17390025e+00 2.17137679e-01 1.85997076e-02
-9.53200638e-01 -6.37728870e-01 2.22970366e-01 1.80781245e-01
-1.19525468e+00 1.30563438e-01 1.34128642e+00 -7.05964923e-01
8.32595408e-01 2.39439420e-02 6.14875376e-01 1.49869430e+00
2.00071439e-01 7.02905536e-01 1.23879623e+00 -2.05229089e-01
3.40619355e-01 4.21548307e-01 7.79486060e-01 -1.85624614e-01
2.62352020e-01 -5.93263566e-01 -7.96069562e-01 -6.73331261e-01
3.04959845e-02 1.78055596e-02 -5.81699573e-02 -6.90432042e-02
-1.26957989e+00 1.40870428e+00 1.04744278e-01 5.15418172e-01
-7.06233025e-01 2.65979767e-02 4.35066819e-01 9.63812321e-02
1.09672666e+00 4.50396597e-01 -3.97090197e-01 -4.21543002e-01
-1.36382735e+00 5.33774793e-01 1.61848998e+00 9.21857178e-01
8.03544521e-01 -7.23595798e-01 9.51175988e-02 6.92383051e-01
6.06160939e-01 2.10444078e-01 5.62327981e-01 -4.05824006e-01
7.27767348e-01 5.84206760e-01 2.76047468e-01 -1.21071577e+00
-3.14634889e-01 -3.32215056e-02 -2.10782245e-01 -5.80683827e-01
2.58351564e-01 -6.61673069e-01 -5.21708190e-01 1.16235363e+00
4.94886428e-01 2.08733082e-01 5.19483984e-02 7.81097829e-01
8.52318048e-01 1.13969076e+00 6.40233994e-01 -1.66050777e-01
1.56152666e+00 -8.42163205e-01 -8.84573877e-01 -2.24323496e-01
3.07927541e-02 -1.06155372e+00 5.98654866e-01 5.25274873e-01
-5.47057092e-01 1.03882588e-01 -6.73141479e-01 -3.25353853e-02
-8.07247400e-01 -2.93705314e-01 7.34754682e-01 6.23696864e-01
-4.77500618e-01 4.89540219e-01 -7.20738709e-01 -7.03882039e-01
2.93817997e-01 4.13377061e-02 1.82438567e-01 3.38322848e-01
-1.44343543e+00 5.55948853e-01 8.26541364e-01 -4.16462690e-01
-8.32812309e-01 -4.04751837e-01 -5.85539937e-01 2.49583870e-01
6.75098538e-01 -2.23646864e-01 1.53421307e+00 -8.38572741e-01
-1.55068791e+00 6.48163140e-01 -3.15430641e-01 -7.29525566e-01
2.56691784e-01 -4.24745649e-01 -2.66082466e-01 2.77767211e-01
8.00748840e-02 5.39600849e-01 9.18030322e-01 -1.07605231e+00
-9.39319074e-01 -3.02152008e-01 2.60159135e-01 -1.81335993e-02
-5.43859899e-01 7.74306059e-01 -1.45340174e-01 -7.68782258e-01
-1.06375910e-01 -8.21858287e-01 -2.53611624e-01 -8.95819366e-01
-6.25870883e-01 -9.16428328e-01 1.28536999e+00 -5.97508550e-01
1.06196892e+00 -1.77453959e+00 -1.63543314e-01 3.06086600e-01
3.21347147e-01 -3.07620525e-01 4.70454901e-01 8.22308898e-01
3.39955300e-01 5.78909993e-01 2.26116441e-02 -9.76049155e-02
2.25821748e-01 -4.46421169e-02 -6.63809478e-01 1.56761348e-01
3.42031126e-04 4.05924112e-01 -9.10089731e-01 -7.87832379e-01
-1.69060811e-01 5.31124949e-01 -3.05385798e-01 2.62397259e-01
-8.97305787e-01 1.60528094e-01 -9.25742209e-01 3.10337812e-01
2.92770386e-01 -3.00533414e-01 3.97600859e-01 1.44217283e-01
-5.91877326e-02 8.42902005e-01 -7.51645267e-01 1.53708053e+00
-2.03024343e-01 8.87173712e-01 -1.72387645e-01 -1.00173485e+00
9.62913156e-01 7.18346536e-01 7.29425848e-01 1.53008550e-02
5.27790666e-01 -5.29939495e-02 -2.56256461e-01 -5.43132722e-01
9.55000937e-01 -4.07822907e-01 -4.15221274e-01 1.06520700e+00
1.08019911e-01 -2.68717319e-01 4.39276546e-01 2.86813319e-01
5.57456911e-01 -2.77541250e-01 3.74345452e-01 -5.21957338e-01
1.83831587e-01 3.55676591e-01 3.22060555e-01 5.22282362e-01
-1.70293093e-01 5.26921451e-01 9.36397195e-01 -2.56550342e-01
-1.01651168e+00 -6.51002884e-01 -1.79131880e-01 1.12359095e+00
1.42267391e-01 -9.59603786e-01 -7.44135976e-01 -1.15153146e+00
-1.63653806e-01 8.23224545e-01 -4.88570660e-01 4.51920897e-01
-3.84829104e-01 -8.25207412e-01 -1.19991690e-01 1.04905121e-01
3.46764863e-01 -9.77815211e-01 -7.01387286e-01 4.09591466e-01
-8.43248963e-01 -1.23932540e+00 -1.63590118e-01 -2.09262267e-01
-6.94520652e-01 -1.20924401e+00 -6.86761379e-01 -6.62262678e-01
1.87142611e-01 2.73871690e-01 1.25906360e+00 -3.29933643e-01
3.17305773e-01 1.76904917e-01 -7.46513009e-01 -1.05452955e+00
-2.66845345e-01 5.49850464e-01 -3.20595533e-01 1.96122617e-01
8.69592786e-01 -5.50590158e-01 -6.18565321e-01 4.89909127e-02
-1.00068247e+00 -1.29720971e-01 8.43534842e-02 4.75583285e-01
6.99382834e-03 3.07669193e-01 4.12911564e-01 -1.39381754e+00
1.08964241e+00 -1.37174571e+00 -4.18436408e-01 -1.53411165e-01
-5.12771249e-01 -1.98542356e-01 1.11781456e-01 -4.13478971e-01
-1.03619087e+00 -3.86227071e-01 6.85988665e-02 1.23530425e-01
-3.61955404e-01 9.58799541e-01 1.77914336e-01 6.35878026e-01
5.11556447e-01 -1.82415113e-01 -3.89735699e-01 -6.43212795e-01
2.82041162e-01 1.13892555e+00 -2.18044460e-01 -6.40648127e-01
4.52066749e-01 4.56741631e-01 -4.69450980e-01 -8.71246338e-01
-9.70951259e-01 -1.02699292e+00 -2.62221187e-01 -2.23681554e-01
5.93078315e-01 -6.81206107e-01 -4.56860363e-01 6.75575137e-02
-1.33291984e+00 -1.80841058e-01 -5.01158945e-02 7.75775075e-01
-4.15188700e-01 1.95336163e-01 -3.69096488e-01 -6.94984972e-01
-4.48043376e-01 -8.51321340e-01 8.67249846e-01 1.67436212e-01
-6.60292149e-01 -1.25437737e+00 3.58777374e-01 2.10426033e-01
5.67458868e-01 7.26247653e-02 5.61211944e-01 -1.60236013e+00
-1.74132556e-01 -1.70418113e-01 -2.44243711e-01 -4.20399085e-02
-2.77159233e-02 -8.18700343e-02 -9.42261338e-01 3.52713987e-02
3.08064461e-01 -2.35611811e-01 7.80039668e-01 2.19000190e-01
8.88195455e-01 -7.09430575e-01 -7.01799273e-01 -2.81171322e-01
1.25674462e+00 1.81929946e-01 2.93238521e-01 6.58143103e-01
1.28846243e-01 1.05912340e+00 6.70211732e-01 7.29539156e-01
7.09522903e-01 4.65202659e-01 3.30579877e-01 3.47045600e-01
4.10791725e-01 -2.85381615e-01 1.23705797e-01 6.69939637e-01
1.23670541e-01 -3.44591051e-01 -1.20737445e+00 9.54165995e-01
-1.78217387e+00 -9.63134766e-01 -1.30223453e-01 1.56688178e+00
9.14940774e-01 1.55500770e-01 4.46978599e-01 2.93385889e-02
9.21131372e-01 6.03005707e-01 9.25223827e-02 -3.01328450e-01
3.48905772e-01 1.59534752e-01 9.70226526e-02 3.48020613e-01
-1.39475667e+00 9.20986354e-01 6.52645493e+00 5.06944716e-01
-1.20627189e+00 4.86026049e-01 6.37556851e-01 7.36860037e-02
-3.57861578e-01 5.36229014e-01 -8.28156233e-01 8.50087464e-01
1.24279702e+00 -5.25776088e-01 -2.14354724e-01 1.10165787e+00
1.82810038e-01 -2.47673675e-01 -8.96212459e-01 5.47768533e-01
2.70922512e-01 -1.02033877e+00 -9.41417217e-02 1.58828452e-01
9.17189121e-01 3.56908739e-01 -1.71859741e-01 3.37783098e-01
4.93822038e-01 -8.39320779e-01 6.01326644e-01 3.25070769e-01
-6.74543306e-02 -4.86725003e-01 9.56411242e-01 4.68450159e-01
-5.09665430e-01 1.70027792e-01 -1.63148791e-01 -4.41696960e-03
4.36320394e-01 1.10028577e+00 -1.18952799e+00 3.95308435e-01
5.78203857e-01 6.46078587e-01 -1.82120055e-01 1.31163192e+00
-4.73457277e-01 1.33797050e+00 -6.59867883e-01 -5.36346734e-01
4.75590974e-01 6.21925760e-03 7.10095286e-01 1.64249992e+00
3.40113521e-01 9.87800583e-02 2.90580124e-01 8.97206306e-01
-3.59908313e-01 5.72029531e-01 -5.47069013e-01 -1.93576515e-01
2.79553890e-01 1.49036336e+00 -1.05915713e+00 -7.08706021e-01
-3.38568568e-01 2.97148973e-01 1.22372009e-01 3.83566529e-01
-5.80982268e-01 -5.17941177e-01 3.12096298e-01 3.10469061e-01
3.01293582e-01 -2.23812550e-01 -1.59301281e-01 -1.35129023e+00
-1.47083059e-01 -5.41550875e-01 4.15570170e-01 -4.30937946e-01
-1.25205922e+00 5.43639421e-01 4.31276947e-01 -8.51777613e-01
-3.94519866e-01 -1.91789702e-01 -1.07361710e+00 5.08462667e-01
-1.68224907e+00 -9.81480002e-01 -7.92506263e-02 3.42355907e-01
9.07660306e-01 -4.04940099e-02 3.93449992e-01 -8.02588984e-02
-1.62846774e-01 -3.31604004e-01 -1.26704425e-01 2.06592888e-01
6.85303748e-01 -1.31048906e+00 3.34501982e-01 5.83713174e-01
3.73356313e-01 6.18780434e-01 1.14720702e+00 -7.90596068e-01
-8.48541558e-01 -8.86199236e-01 1.64713812e+00 -3.46012294e-01
1.08172619e+00 -3.19025874e-01 -7.23129809e-01 9.20994580e-01
7.21548557e-01 -4.97197628e-01 1.03712714e+00 6.00224674e-01
-3.41896117e-01 2.80518800e-01 -1.01144898e+00 1.49260536e-01
2.20899895e-01 -3.30041856e-01 -1.12928820e+00 7.09345102e-01
5.51923335e-01 2.93778093e-03 -9.56987262e-01 -2.74588335e-02
3.65875870e-01 -5.47234774e-01 4.33790028e-01 -7.35907495e-01
5.60194194e-01 8.04615095e-02 -6.24739863e-02 -1.52053714e+00
3.48842144e-01 -8.70618582e-01 -1.33201316e-01 1.49535632e+00
5.20289421e-01 -5.16032755e-01 9.12577510e-01 5.88503301e-01
4.39026743e-01 -1.24960706e-01 -7.29470432e-01 -2.97207296e-01
2.21055314e-01 -6.70915008e-01 6.45469069e-01 1.21369112e+00
8.62470150e-01 5.63434541e-01 -7.51403198e-02 -1.38621002e-01
5.52828312e-01 3.11073571e-01 7.71030188e-01 -1.63525140e+00
1.54532120e-01 -5.19246459e-01 -2.61598408e-01 -7.44509816e-01
1.95668414e-01 -6.63090229e-01 1.33136183e-01 -1.48640811e+00
5.58993995e-01 -3.45328182e-01 2.19519641e-02 1.28993154e-01
-8.86358395e-02 -8.57810453e-02 2.06295162e-01 2.69091755e-01
-9.26220179e-01 5.32694876e-01 5.17767549e-01 -2.05132082e-01
-6.25501052e-02 2.24180043e-01 -6.92393780e-01 9.42203581e-01
1.40846264e+00 -9.63403225e-01 -2.51686662e-01 9.83762667e-02
5.06996572e-01 -9.00985207e-03 1.67133454e-02 -4.70127970e-01
2.53807455e-01 -3.80853444e-01 -3.30431610e-01 -8.84599268e-01
2.98986703e-01 -8.61932456e-01 -1.17805287e-01 8.11460540e-02
-4.74302500e-01 -2.64152050e-01 -2.77307093e-01 5.74445724e-01
-6.02640271e-01 -4.24670905e-01 2.04253301e-01 -2.08410248e-01
-3.86151254e-01 1.73095033e-01 -6.43098533e-01 1.87245086e-01
1.14496589e+00 5.16231477e-01 -3.45915794e-01 -6.05996251e-01
-7.89277434e-01 2.34370634e-01 -8.20666254e-02 5.54212034e-01
4.36406225e-01 -1.08486784e+00 -8.07645261e-01 -3.75677079e-01
1.50162488e-01 -6.03430122e-02 -1.80950016e-01 6.97359920e-01
-2.46765569e-01 9.53827500e-01 1.14175633e-01 -3.41240972e-01
-1.10240579e+00 5.70078552e-01 -3.36154521e-01 -4.81385022e-01
-6.52270138e-01 3.51008892e-01 -7.97804096e-04 -6.48212358e-02
1.12106100e-01 -3.35402340e-01 -7.59514689e-01 4.06531483e-01
5.33115923e-01 2.20373824e-01 -3.86462092e-01 -7.08275080e-01
8.52678195e-02 6.55683279e-02 -2.35236049e-01 -4.28605318e-01
1.47295225e+00 -1.90573752e-01 -2.99793839e-01 6.86296463e-01
1.18587208e+00 -5.13287336e-02 -7.84541130e-01 -5.04185379e-01
6.94452465e-01 -3.72047246e-01 7.97404051e-02 -4.11459953e-01
-7.48019338e-01 7.83052742e-01 -5.58604188e-02 1.02034724e+00
4.71179783e-01 3.75222027e-01 9.69031274e-01 1.86995789e-01
1.62028149e-01 -1.33836019e+00 -5.95133267e-02 6.32837415e-01
7.33351469e-01 -1.50244761e+00 -5.07228784e-02 -6.35313332e-01
-6.72094107e-01 1.20018327e+00 1.25659809e-01 -1.76811591e-01
1.37274849e+00 -1.26587972e-01 4.92404588e-02 -6.54019654e-01
-9.07606304e-01 -1.99157625e-01 3.02908987e-01 3.09007853e-01
6.54394448e-01 1.83497831e-01 -1.07336342e+00 1.04011559e+00
-7.35316098e-01 -1.43209979e-01 6.67360127e-01 8.69182646e-01
-7.01547563e-01 -1.11659181e+00 -3.63704860e-01 4.09360945e-01
-1.22773945e+00 -2.86043826e-02 -5.80844820e-01 6.12790108e-01
-3.68538111e-01 1.38468790e+00 1.14035182e-01 -2.49532267e-01
-3.00578713e-01 4.38449681e-01 -5.76105535e-01 -8.87459040e-01
-7.59729624e-01 2.91910261e-01 4.39355820e-01 -1.58049136e-01
-8.15272331e-01 -9.20043409e-01 -9.43930149e-01 -1.34749666e-01
-4.62467074e-01 8.88687253e-01 1.43382120e+00 1.06043375e+00
1.69834509e-01 2.07913555e-02 8.20154369e-01 -5.25621653e-01
-2.38410249e-01 -1.50553489e+00 -5.26858151e-01 5.32857239e-01
-6.20524660e-02 -5.27211547e-01 -1.97614297e-01 2.49813691e-01] | [10.372836112976074, 7.186705589294434] |
f2fc2ce7-c581-48e3-b64f-5d5f8557cfef | phatyp-predicting-the-lifestyle-for | 2206.09693 | null | https://arxiv.org/abs/2206.09693v1 | https://arxiv.org/pdf/2206.09693v1.pdf | PhaTYP: Predicting the lifestyle for bacteriophages using BERT | Bacteriophages (or phages), which infect bacteria, have two distinct lifestyles: virulent and temperate. Predicting the lifestyle of phages helps decipher their interactions with their bacterial hosts, aiding phages' applications in fields such as phage therapy. Because experimental methods for annotating the lifestyle of phages cannot keep pace with the fast accumulation of sequenced phages, computational method for predicting phages' lifestyles has become an attractive alternative. Despite some promising results, computational lifestyle prediction remains difficult because of the limited known annotations and the sheer amount of sequenced phage contigs assembled from metagenomic data. In particular, most of the existing tools cannot precisely predict phages' lifestyles for short contigs. In this work, we develop PhaTYP (Phage TYPe prediction tool) to improve the accuracy of lifestyle prediction on short contigs. We design two different training tasks, self-supervised and fine-tuning tasks, to overcome lifestyle prediction difficulties. We rigorously tested and compared PhaTYP with four state-of-the-art methods: DeePhage, PHACTS, PhagePred, and BACPHLIP. The experimental results show that PhaTYP outperforms all these methods and achieves more stable performance on short contigs. In addition, we demonstrated the utility of PhaTYP for analyzing the phage lifestyle on human neonates' gut data. This application shows that PhaTYP is a useful means for studying phages in metagenomic data and helps extend our understanding of microbial communities. | ['Yanni Sun', 'Xubo Tang', 'Jiayu Shang'] | 2022-06-20 | null | null | null | null | ['type-prediction'] | ['computer-code'] | [ 2.08469525e-01 -4.04168993e-01 1.32125691e-01 3.32623683e-02
8.15983787e-02 -8.25060844e-01 2.10052237e-01 4.17545080e-01
-4.21408355e-01 1.05101168e+00 8.37828144e-02 -5.60847342e-01
-3.38257700e-02 -7.20557451e-01 -7.43135273e-01 -1.15746891e+00
-1.87618241e-01 7.10334837e-01 5.80915868e-01 -3.49004753e-02
2.94209093e-01 1.45068109e-01 -1.64967537e+00 7.19637513e-01
9.57027555e-01 3.62974524e-01 1.14003825e+00 9.10027266e-01
-5.51550508e-01 4.58996482e-02 -2.26095036e-01 -1.27818793e-01
-1.22240074e-01 -2.33734876e-01 -5.93068480e-01 -7.29381859e-01
-6.18252337e-01 2.12134853e-01 3.41061711e-01 7.74487674e-01
5.76977313e-01 -5.98153353e-01 7.85436213e-01 -7.45762348e-01
-2.01639637e-01 3.17109860e-02 -2.31368527e-01 -2.65770823e-01
5.00753939e-01 5.72842807e-02 4.88708586e-01 -5.80887377e-01
1.07009327e+00 1.08008659e+00 1.22043264e+00 4.87310410e-01
-1.33109951e+00 -7.24002719e-02 -4.49905038e-01 3.32980335e-01
-1.14838648e+00 3.13780680e-02 -2.10621837e-03 -1.25655687e+00
9.88206923e-01 6.60796821e-01 7.91087151e-01 1.16358876e+00
4.60783064e-01 4.34564292e-01 1.49272156e+00 -1.01861551e-01
2.08876744e-01 3.49626571e-01 2.12892532e-01 6.47072494e-01
4.54803973e-01 1.80847347e-01 -3.78143698e-01 -6.80719495e-01
2.81210691e-01 -7.09322020e-02 -3.47758830e-01 -4.09305066e-01
-1.33913982e+00 7.78177321e-01 -1.51587337e-01 -2.29651257e-01
-4.65747952e-01 -4.41109806e-01 8.11091304e-01 1.21093176e-01
3.53546798e-01 4.63649482e-01 -1.15715361e+00 -4.31890994e-01
2.93417871e-01 -2.83471681e-02 1.24434078e+00 7.87610173e-01
9.42137778e-01 -6.69585586e-01 2.74865419e-01 9.15109098e-01
4.15316135e-01 4.12300408e-01 3.07104290e-01 -2.01555431e-01
-3.02088290e-01 6.34346247e-01 3.03949147e-01 -4.01962072e-01
-6.60193205e-01 2.08670929e-01 -6.41566038e-01 7.57467747e-02
7.11442232e-01 -2.50723034e-01 -6.77683532e-01 1.47291040e+00
5.77952683e-01 1.00996152e-01 1.46529287e-01 6.38974845e-01
9.28238809e-01 7.24728942e-01 1.93186194e-01 -4.47990596e-01
1.87665749e+00 -5.60866952e-01 -3.11797321e-01 2.89992750e-01
3.91120285e-01 -1.16892099e+00 9.49954629e-01 4.92559284e-01
-1.07916243e-01 -2.59241641e-01 -4.86217231e-01 6.34995997e-01
-3.66504073e-01 -1.40124992e-01 7.42422938e-01 9.32139814e-01
-7.31035888e-01 5.49360335e-01 -1.10131454e+00 -1.02888918e+00
-1.35958195e-01 2.50182807e-01 -3.69653523e-01 4.40847091e-02
-7.23153591e-01 9.02928710e-01 6.98103666e-01 -2.77839303e-01
-8.88090074e-01 -4.33874935e-01 -2.57899195e-01 -2.44361594e-01
1.98664054e-01 -7.47940123e-01 8.53494108e-01 -1.28022298e-01
-1.46089458e+00 8.01008344e-01 -2.38170102e-01 1.67220339e-01
1.23259887e-01 -2.88872451e-01 2.73494720e-02 -2.52046406e-01
4.12662774e-02 1.65423870e-01 6.53839558e-02 -1.22793067e+00
-7.57665634e-01 -1.06850833e-01 -4.20823634e-01 -1.53037190e-01
-1.17437476e-02 2.25865498e-01 -2.90817618e-01 -2.93333262e-01
2.01992202e-03 -1.33523405e+00 -4.50037658e-01 -4.84706879e-01
-4.83760148e-01 -5.15460968e-01 3.24192822e-01 -7.77314246e-01
5.67909181e-01 -1.90822554e+00 2.44168341e-01 -1.15945235e-01
-7.96669200e-02 4.28198159e-01 -1.56709939e-01 9.59361196e-01
2.71074951e-01 -5.94650693e-02 7.12320507e-02 4.84470844e-01
-2.14260623e-01 3.15059602e-01 2.80753267e-03 5.03472865e-01
-1.02892891e-02 4.26966190e-01 -1.01395094e+00 -1.81959853e-01
7.70644918e-02 6.01265371e-01 -5.78089356e-01 7.41249084e-01
-6.16919458e-01 1.06582654e+00 -3.99830997e-01 5.36340177e-01
8.46567214e-01 -1.97564065e-01 8.03910553e-01 -2.53698438e-01
-7.34900415e-01 3.98303270e-01 -8.21302354e-01 1.59062231e+00
-7.89211988e-02 3.90948296e-01 -7.76333511e-02 -7.06137300e-01
7.28308558e-01 4.01648432e-01 4.59772259e-01 4.58979383e-02
-7.90994316e-02 3.27978343e-01 8.36337581e-02 -1.13583171e+00
4.99010980e-02 -1.09208710e-01 2.91044444e-01 3.34774882e-01
-2.38038257e-01 5.04235506e-01 1.71663657e-01 -3.58125329e-01
1.12360930e+00 6.61718488e-01 6.09613121e-01 -6.92790985e-01
3.11617374e-01 6.03786707e-01 9.80846167e-01 5.77653646e-01
1.26511246e-01 4.14621323e-01 6.05099380e-01 -6.04360223e-01
-1.16176450e+00 -1.10574281e+00 -4.85908687e-01 1.68964744e+00
3.47455293e-01 -2.75815368e-01 -6.22593403e-01 -2.57774502e-01
-1.55226648e-01 1.14755765e-01 -3.60730469e-01 5.50110519e-01
-5.16860425e-01 -1.60518360e+00 7.40137041e-01 1.80253625e-01
-1.80133060e-01 -1.06735623e+00 -5.29048026e-01 5.74841261e-01
-5.02321362e-01 -5.21881640e-01 2.20279628e-03 6.44259989e-01
-2.51649916e-01 -1.73882258e+00 -5.14068782e-01 -1.14967322e+00
2.43748650e-01 3.30468088e-01 7.73144960e-01 -9.77309793e-02
-5.32663167e-01 -5.43535240e-02 -6.33661091e-01 -7.14011908e-01
-9.03304815e-01 3.25506888e-02 1.92292914e-01 -5.91264069e-01
4.46785003e-01 -4.90986496e-01 -7.23523676e-01 7.75563538e-01
-4.76596892e-01 3.73719841e-01 6.13829195e-01 1.18109846e+00
5.01086056e-01 -1.06881052e-01 8.02031457e-01 -1.19212580e+00
3.01274538e-01 -8.87532294e-01 -5.76136172e-01 5.86476505e-01
-6.30061030e-01 7.77312517e-02 3.76147270e-01 -4.96184498e-01
-1.50747502e+00 2.37700701e-01 -7.69327223e-01 9.84029889e-01
-4.04564351e-01 5.76996744e-01 -1.14429064e-01 7.71194622e-02
7.49222398e-01 4.24788266e-01 4.14059758e-01 -1.09001112e+00
-8.28427449e-02 8.94521475e-01 2.33352259e-01 -6.51736557e-01
1.57003596e-01 2.85577238e-01 -3.02574694e-01 -1.03363550e+00
-3.01892042e-01 -1.09222412e+00 -4.32010412e-01 -6.52502328e-02
1.24931669e+00 -8.80057335e-01 -1.37890851e+00 8.78660142e-01
-1.34768903e+00 -3.36346716e-01 3.94915611e-01 6.10141575e-01
-4.93404329e-01 6.57585621e-01 -1.10410154e+00 -6.38719618e-01
-3.78553659e-01 -1.33697343e+00 8.16257119e-01 4.22831364e-02
-1.70627534e-01 -9.39362288e-01 8.40981543e-01 3.39601599e-02
3.81813586e-01 3.06047380e-01 1.23118949e+00 -7.05555677e-01
-5.81176221e-01 3.51625681e-01 -3.02674860e-01 9.51428711e-02
3.73233795e-01 -1.35702893e-01 -8.73213470e-01 -2.94303268e-01
-8.78526196e-02 -2.05458775e-01 8.10941994e-01 6.51171878e-02
5.69988728e-01 -5.17050289e-02 -7.18912065e-01 6.38922334e-01
1.62406027e+00 5.37894368e-01 5.64692974e-01 7.35400736e-01
7.75454223e-01 1.04383612e+00 9.38719451e-01 4.42813843e-01
3.41899544e-01 5.92157483e-01 4.36576813e-01 2.94908047e-01
7.09595915e-04 1.89328790e-01 1.53748676e-01 9.56792295e-01
-6.55287385e-01 -2.44621098e-01 -1.12407351e+00 1.95249304e-01
-2.01584435e+00 -7.33166039e-01 -7.29290187e-01 2.13316894e+00
1.17056847e+00 -4.34685558e-01 2.00103715e-01 -3.96444410e-01
6.88418627e-01 -5.95934331e-01 -4.42587972e-01 -4.16436523e-01
-3.12881380e-01 -3.04046962e-02 3.32835257e-01 1.30959991e-02
-1.15016830e+00 5.14039755e-01 6.34542894e+00 4.30913150e-01
-9.15750146e-01 2.21042946e-01 -1.66738704e-01 6.90031409e-01
-9.52747166e-02 6.81721196e-02 -6.95581496e-01 5.55447221e-01
6.42459810e-01 2.05514014e-01 5.21485507e-01 9.87843156e-01
-3.97606427e-03 -1.74064890e-01 -9.19968843e-01 4.13469076e-01
-5.45777380e-01 -1.60220873e+00 -7.51047909e-01 1.71446651e-01
5.91136217e-01 6.53601587e-01 -7.53705978e-01 -4.21908163e-02
4.81061906e-01 -6.63977563e-01 3.03863972e-01 3.91476721e-01
7.59697497e-01 -3.10050160e-01 1.15346122e+00 5.99784851e-01
-1.14529061e+00 2.24285245e-01 -7.69713759e-01 -8.99142101e-02
2.69922674e-01 4.74102557e-01 -1.66138840e+00 4.60721433e-01
7.90799379e-01 4.28116709e-01 -2.61145204e-01 1.46399248e+00
-5.58031760e-02 6.06903374e-01 -1.53423861e-01 -6.15476906e-01
-4.33374822e-01 -2.40798637e-01 5.52822769e-01 1.84489763e+00
4.07402247e-01 -6.28677458e-02 3.43235523e-01 1.51103705e-01
6.83491528e-01 3.25304151e-01 -3.64310205e-01 -7.04345182e-02
4.02975023e-01 1.01160181e+00 -7.92652488e-01 -2.57366836e-01
-4.67206925e-01 5.47740996e-01 9.21934247e-02 1.36524931e-01
-4.73611832e-01 -9.50824246e-02 1.14492691e+00 -1.41124418e-02
1.80174202e-01 -9.17520523e-02 -2.43911251e-01 -8.57834399e-01
-6.47513688e-01 -8.08160067e-01 4.21908528e-01 -2.34135166e-01
-1.72608316e+00 6.71466351e-01 -2.50939637e-01 -1.00492215e+00
-3.53847742e-02 -9.13462996e-01 -4.89763916e-01 8.33455086e-01
-9.68341649e-01 -1.12513721e+00 -4.79030907e-01 -4.32210937e-02
5.59921741e-01 -2.58395672e-01 1.32430661e+00 1.72065601e-01
-3.00948292e-01 -2.09732905e-01 8.22182238e-01 -3.84522408e-01
8.09655428e-01 -1.33539879e+00 3.48256737e-01 2.32898042e-01
-7.68855453e-01 7.21933544e-01 1.13900399e+00 -1.08697867e+00
-1.69433820e+00 -1.15483332e+00 4.24836218e-01 -4.11778152e-01
7.19923496e-01 -3.22056949e-01 -1.05979490e+00 3.83413762e-01
3.73461038e-01 -1.05708206e+00 1.21302152e+00 3.43955606e-01
-2.80099064e-01 4.27663416e-01 -7.45107889e-01 3.03590626e-01
9.31504488e-01 -8.46278816e-02 -5.12573540e-01 6.46639585e-01
7.95368731e-01 -2.81950086e-01 -1.07349658e+00 5.44654012e-01
1.17147434e+00 -1.08375692e+00 1.06816876e+00 -6.45040631e-01
-1.20537698e-01 -4.73204643e-01 -1.48203850e-01 -1.34897327e+00
-4.31059659e-01 -2.06157938e-01 6.09783292e-01 1.02878225e+00
8.84311721e-02 -8.92865121e-01 6.50284290e-01 -2.32702613e-01
-6.43857956e-01 -4.78965938e-01 -6.16269648e-01 -7.74009645e-01
-3.36236298e-01 7.75877312e-02 5.65296471e-01 7.98680305e-01
3.05751115e-01 4.60321397e-01 -5.80660999e-01 1.80254176e-01
6.23763382e-01 3.18430841e-01 9.27794158e-01 -1.44137502e+00
-6.00294769e-01 2.90625170e-02 -2.10531384e-01 -5.98808706e-01
-2.68524528e-01 -7.50736952e-01 6.46156251e-01 -1.63575649e+00
6.17920637e-01 -1.12213564e+00 -1.12889469e-01 3.31358075e-01
-3.85334373e-01 1.75538242e-01 -4.87498313e-01 2.54247427e-01
-4.05469924e-01 -9.72392708e-02 8.37868810e-01 1.42085224e-01
-2.47103661e-01 4.15846199e-01 -3.82203728e-01 8.21017683e-01
8.16434443e-01 -5.55730999e-01 -1.26730233e-01 -6.41013682e-02
5.86343110e-01 -1.23366974e-01 9.17099640e-02 -5.10824323e-01
-8.95150751e-02 -5.69838166e-01 -7.33478889e-02 -7.48135030e-01
1.68847948e-01 -2.72939742e-01 1.07160366e+00 1.24242067e+00
3.17021906e-01 -1.56472594e-01 -2.81022396e-02 7.22925127e-01
1.51180640e-01 -1.83608934e-01 5.12452066e-01 -6.32972941e-02
-9.20582652e-01 -1.19293898e-01 -1.01132596e+00 -3.77663612e-01
9.84401941e-01 -2.45092176e-02 -8.59268010e-01 5.06830752e-01
-7.55512118e-01 -2.17377931e-01 9.58850086e-01 3.76399696e-01
3.70134801e-01 -7.16918647e-01 -7.40805507e-01 1.72196537e-01
3.83320451e-01 -2.25221455e-01 2.83607721e-01 1.09639454e+00
-1.33473063e+00 4.13526922e-01 -5.14205933e-01 -1.09680164e+00
-1.77267063e+00 8.99801493e-01 -4.38664407e-01 2.21954752e-02
-3.77910525e-01 8.08841288e-01 7.99528778e-01 -9.49724615e-01
1.78722531e-01 1.12113647e-01 -5.38995683e-01 -3.04773040e-02
4.18535203e-01 5.66403925e-01 1.97758619e-02 -2.53770649e-01
-5.98063231e-01 3.74965847e-01 3.30412775e-01 6.63154662e-01
1.65509856e+00 -9.64431018e-02 -9.03618217e-01 4.26149309e-01
9.20742989e-01 1.64243300e-02 -9.65456367e-01 -1.68432407e-02
3.99623841e-01 -4.24583912e-01 -1.14819396e+00 -7.85671294e-01
7.31303096e-02 5.25477469e-01 7.65675962e-01 3.39429289e-01
1.00598037e+00 2.02675298e-01 6.58277750e-01 5.29684901e-01
6.85334265e-01 -6.45609379e-01 6.05473178e-04 5.09477019e-01
4.28489864e-01 -1.15211308e+00 -4.15477425e-01 -1.00815713e+00
-4.12471771e-01 9.19502974e-01 3.54322761e-01 5.05876541e-01
4.12075222e-01 3.69738728e-01 1.77414685e-01 -2.31360942e-02
-1.01791573e+00 -2.65730973e-02 7.31883794e-02 1.04972005e+00
6.23821318e-01 5.20265937e-01 -7.45592296e-01 4.53455985e-01
1.07046045e-01 -2.12994993e-01 3.72601062e-01 9.76238132e-01
-1.04446042e+00 -1.61624384e+00 -5.89383662e-01 5.75900137e-01
-4.68034774e-01 -1.53794259e-01 -6.14384227e-02 3.35757911e-01
4.69965190e-01 8.17647755e-01 -4.49967414e-01 -5.41611910e-01
4.78580371e-02 1.09939240e-01 3.14540595e-01 -7.11394370e-01
-4.41382885e-01 5.74969053e-01 7.87079573e-01 -7.73771703e-02
-3.88970733e-01 -8.07950556e-01 -1.08051920e+00 -2.89414436e-01
-4.84014511e-01 5.09645820e-01 1.20212042e+00 5.87794065e-01
2.56689191e-01 3.80802184e-01 5.66713214e-01 -5.19302607e-01
-2.77395487e-01 -1.03302515e+00 -5.69784105e-01 3.53621691e-01
-2.75447160e-01 -5.73108971e-01 1.19628549e-01 3.86903882e-01] | [4.822412967681885, 5.352592468261719] |
c3d9644b-a4e8-4886-adfe-72aa5e8f7a67 | learning-a-representation-with-the-block | 1911.10301 | null | https://arxiv.org/abs/1911.10301v1 | https://arxiv.org/pdf/1911.10301v1.pdf | Learning a Representation with the Block-Diagonal Structure for Pattern Classification | Sparse-representation-based classification (SRC) has been widely studied and developed for various practical signal classification applications. However, the performance of a SRC-based method is degraded when both the training and test data are corrupted. To counteract this problem, we propose an approach that learns Representation with Block-Diagonal Structure (RBDS) for robust image recognition. To be more specific, we first introduce a regularization term that captures the block-diagonal structure of the target representation matrix of the training data. The resulting problem is then solved by an optimizer. Last, based on the learned representation, a simple yet effective linear classifier is used for the classification task. The experimental results obtained on several benchmarking datasets demonstrate the efficacy of the proposed RBDS method. | ['Zhen-Hua Feng', 'Xiao-Jun Wu', 'He-Feng Yin', 'Josef Kittler'] | 2019-11-23 | null | null | null | null | ['sparse-representation-based-classification'] | ['computer-vision'] | [ 4.51196939e-01 -3.88677359e-01 -1.11813404e-01 -3.09767157e-01
-9.30216014e-01 -5.40209897e-02 2.70290405e-01 -1.40824825e-01
-6.89293221e-02 6.50584579e-01 1.74236372e-01 1.41069323e-01
-2.77510613e-01 -3.78605634e-01 -5.28993547e-01 -1.07139552e+00
7.24570453e-02 -2.72170514e-01 -1.43068403e-01 4.03772332e-02
2.81298518e-01 2.81168580e-01 -1.38664138e+00 2.15331525e-01
8.55832756e-01 1.35994458e+00 1.18799262e-01 2.14950919e-01
3.37264210e-01 8.54672492e-01 -5.53590596e-01 1.60516784e-01
3.96222115e-01 -5.87735116e-01 -3.29031944e-01 4.39786077e-01
2.20772445e-01 2.21039116e-01 -4.82322037e-01 1.17843187e+00
5.02420068e-01 2.89427787e-01 6.51830614e-01 -8.64510417e-01
-3.45540583e-01 3.90984982e-01 -7.72173464e-01 3.00046414e-01
2.31407598e-01 -1.56366631e-01 8.66526246e-01 -1.08791065e+00
2.74302036e-01 9.99821126e-01 3.82563591e-01 2.00618878e-01
-1.30784738e+00 -7.37559080e-01 1.80637628e-01 3.19582641e-01
-1.59408998e+00 -3.94657165e-01 1.15505779e+00 -4.48049992e-01
3.50428969e-01 3.92817765e-01 5.32230496e-01 8.66909921e-01
5.42767942e-02 8.87749791e-01 1.19909024e+00 -3.40940922e-01
2.38732502e-01 2.00185537e-01 3.79348338e-01 4.59158391e-01
2.71283835e-01 2.86250114e-02 -4.71641988e-01 -2.64352560e-01
4.89047319e-01 7.45018423e-02 -6.93765104e-01 -5.89058161e-01
-8.46516967e-01 8.34815502e-01 4.32423323e-01 4.84625250e-01
-4.38514054e-01 -6.19217567e-02 5.14925361e-01 3.94227326e-01
4.70145732e-01 1.11193515e-01 9.12162475e-03 1.15358591e-01
-1.05557644e+00 -6.08235039e-02 5.76848984e-01 3.77600521e-01
5.42276323e-01 5.82021534e-01 -7.64480904e-02 1.27502739e+00
3.17596674e-01 5.14858305e-01 8.51362646e-01 -3.42409760e-01
5.57474077e-01 5.33828616e-01 -1.38092160e-01 -1.49781263e+00
-2.28382587e-01 -1.02788091e+00 -1.36107147e+00 -4.89376066e-03
1.43772811e-01 3.96954454e-02 -6.29371703e-01 1.36994302e+00
2.49929309e-01 8.14577222e-01 2.92195439e-01 1.23415112e+00
5.80740094e-01 8.65981758e-01 -2.31385902e-01 -3.96715075e-01
9.52380061e-01 -6.36815786e-01 -6.26429558e-01 -3.03185046e-01
4.32350099e-01 -5.15403271e-01 6.66206181e-01 5.52294791e-01
-6.75991654e-01 -6.01581514e-01 -1.26128864e+00 3.61568570e-01
2.12304935e-01 5.76643169e-01 2.32278079e-01 6.23425841e-01
-4.58925605e-01 1.90828189e-01 -5.89841306e-01 6.38761222e-02
2.86988467e-01 3.04966152e-01 -6.13168776e-01 -2.75633037e-01
-9.97477949e-01 3.97078544e-01 3.99448007e-01 5.01769841e-01
-7.35891521e-01 -6.16154552e-01 -8.67172718e-01 4.00882103e-02
2.63594925e-01 -1.73844650e-01 6.74932241e-01 -1.14066911e+00
-1.54525113e+00 6.34722233e-01 1.29650289e-03 -6.39658868e-01
2.17163160e-01 -3.54308188e-01 -3.88762355e-01 1.86930552e-01
-1.04394697e-01 -2.88095474e-01 1.57601953e+00 -1.25873637e+00
-3.41292411e-01 -3.74809265e-01 -4.29822594e-01 1.42440394e-01
-4.49647605e-01 -1.39249563e-01 -3.37018996e-01 -1.06405234e+00
4.76278305e-01 -8.48090589e-01 -2.89246380e-01 -3.10630232e-01
-4.43384349e-01 9.07718167e-02 9.12814915e-01 -7.29955435e-01
1.21313274e+00 -2.86678982e+00 4.38673526e-01 7.09291577e-01
-1.36655942e-01 4.62000608e-01 -1.46595597e-01 2.32423499e-01
-5.40796041e-01 -4.72080797e-01 -4.07404721e-01 -1.19781144e-01
-5.01282752e-01 1.49293942e-02 -4.56045985e-01 7.60666430e-01
5.79704121e-02 3.00872147e-01 -5.06587148e-01 -2.83609349e-02
2.94970214e-01 5.36622465e-01 -5.00400543e-01 4.29858088e-01
3.89729589e-01 7.08351672e-01 -5.42580485e-01 3.55560571e-01
7.25219846e-01 -2.65504837e-01 3.82498533e-01 -5.14615893e-01
9.87632573e-02 4.27593710e-03 -1.65544367e+00 1.36235654e+00
-5.41182637e-01 4.96214569e-01 1.54115051e-01 -1.73700845e+00
1.23417354e+00 1.98222280e-01 6.72612667e-01 -6.52599514e-01
1.29494309e-01 3.39725733e-01 7.78797194e-02 -2.85116225e-01
-3.13481539e-02 -9.13776383e-02 1.23530455e-01 1.47372320e-01
-1.79253042e-01 1.13310769e-01 -3.63144390e-02 9.45881754e-02
1.02586424e+00 -2.99595207e-01 4.98130232e-01 -2.75633603e-01
1.15860546e+00 -3.26143086e-01 8.01874220e-01 5.31558335e-01
1.27109019e-02 6.24173880e-01 2.94930309e-01 -4.93657202e-01
-5.50046921e-01 -4.86938030e-01 -3.02053630e-01 6.36833191e-01
8.47743750e-02 -3.04416776e-01 -5.17824888e-01 -5.93133807e-01
2.94767153e-02 4.23123360e-01 -3.62269968e-01 -4.93503243e-01
-7.00878143e-01 -9.26374853e-01 2.33203098e-01 3.11621040e-01
5.60971320e-01 -6.27739310e-01 -3.75962466e-01 3.28780383e-01
-9.33485478e-02 -1.16327572e+00 -3.27374190e-01 9.22339559e-02
-9.25100565e-01 -1.02580428e+00 -8.03750575e-01 -8.47075403e-01
7.89708197e-01 6.64841712e-01 4.58044350e-01 8.04967657e-02
-2.72023290e-01 3.43397439e-01 -4.90457505e-01 5.13001233e-02
-9.37229544e-02 -3.13611507e-01 1.23523474e-01 9.22129273e-01
-1.86286643e-01 -6.05425537e-01 -5.96480191e-01 4.25154924e-01
-8.33946288e-01 -2.17410818e-01 7.19975829e-01 1.22671688e+00
7.36380398e-01 2.22400546e-01 7.63260722e-01 -8.36642981e-01
5.63710868e-01 -4.94511843e-01 -7.07937777e-01 9.61492136e-02
-4.21153128e-01 -1.27218384e-02 8.44990730e-01 -4.88081276e-01
-9.12373543e-01 4.35305417e-01 6.92881469e-04 -6.06687844e-01
2.94240624e-01 9.00709569e-01 -9.52478647e-02 -3.11826587e-01
4.22096819e-01 6.61030412e-01 8.95906687e-02 -6.19427979e-01
6.69577494e-02 8.47486556e-01 4.38970447e-01 -3.48639190e-01
9.94434714e-01 3.55010599e-01 7.63310269e-02 -1.17314017e+00
-9.97245312e-01 -7.28284121e-01 -2.15985224e-01 -7.25630075e-02
3.33258510e-01 -1.17847109e+00 -3.32059324e-01 3.85195971e-01
-6.84609115e-01 3.63138951e-02 -1.99552745e-01 9.02762234e-01
-5.08734822e-01 5.81292212e-01 -2.40084440e-01 -9.42621171e-01
-4.95988041e-01 -9.99675214e-01 7.31909990e-01 5.25776967e-02
2.05953360e-01 -5.90956867e-01 1.58387169e-01 4.53288287e-01
1.89768657e-01 1.01291053e-01 8.46107423e-01 -6.29777730e-01
-3.66378933e-01 -6.48463130e-01 -2.65408933e-01 8.53288889e-01
1.79661930e-01 -3.88878614e-01 -8.07243168e-01 -5.95573902e-01
4.63458180e-01 -2.87436604e-01 8.39656234e-01 3.05426627e-01
1.34179091e+00 -2.84199744e-01 -1.85513884e-01 5.88882864e-01
1.48747349e+00 -4.94648516e-02 5.25943458e-01 9.88713950e-02
5.74265540e-01 2.98022330e-01 7.50917971e-01 6.30701065e-01
-9.61894020e-02 8.91287386e-01 1.39884591e-01 -2.91963220e-02
-7.37311468e-02 1.00557106e-02 4.84704465e-01 1.15381348e+00
-2.50351951e-02 1.58435449e-01 -6.44967616e-01 2.55511463e-01
-1.97550464e+00 -8.25417578e-01 -9.59732383e-02 2.38358331e+00
6.43912971e-01 8.42036586e-03 -5.25077097e-02 7.74100125e-01
6.44642413e-01 3.23623866e-01 -4.83919859e-01 -3.31250094e-02
-1.33636191e-01 2.19412968e-01 1.98119462e-01 5.24450913e-02
-1.17781079e+00 5.15844941e-01 5.97015619e+00 1.05011642e+00
-1.47607780e+00 -1.22585058e-01 4.77855265e-01 1.84854642e-01
2.83627480e-01 -2.09382698e-01 -4.81607288e-01 4.08675045e-01
6.84355199e-01 -1.47909403e-01 4.37217265e-01 8.56212974e-01
3.95813197e-01 2.03747693e-02 -8.16061914e-01 1.56298363e+00
3.52627993e-01 -9.67050910e-01 -5.18441238e-02 -1.67243019e-01
7.46845663e-01 -2.89640933e-01 2.78878212e-01 3.39649737e-01
-3.01875323e-01 -7.50932395e-01 6.03542209e-01 3.86469066e-01
6.32521689e-01 -7.02059388e-01 7.21169055e-01 4.07208741e-01
-1.26698470e+00 -4.57864463e-01 -4.93887782e-01 4.75645997e-02
-1.67259917e-01 1.07676244e+00 -5.16245425e-01 7.48631775e-01
4.43812937e-01 1.17246377e+00 -4.13995832e-01 1.22210455e+00
-2.54968643e-01 9.25389767e-01 -1.05308220e-01 3.27155352e-01
1.33817986e-01 -4.65707481e-01 8.47331762e-01 1.08818817e+00
2.77482986e-01 1.94553807e-01 3.44041824e-01 3.93151820e-01
-1.03960365e-01 4.81357396e-01 -6.41742349e-01 8.02532062e-02
8.53763893e-02 1.25702572e+00 -3.83688092e-01 -1.31732181e-01
-4.81120318e-01 8.78452122e-01 3.50461602e-01 4.44330692e-01
-6.06960654e-01 -1.91581219e-01 4.70694423e-01 -1.46121934e-01
5.57460666e-01 -3.34125072e-01 -1.00598328e-01 -1.47180343e+00
3.01971048e-01 -1.34179759e+00 4.07753736e-01 -2.29201615e-01
-1.24584007e+00 5.01994669e-01 -2.66142040e-01 -1.66102672e+00
-5.45554645e-02 -3.66211295e-01 -5.39239526e-01 6.34050250e-01
-1.62131631e+00 -8.96349549e-01 -1.72269836e-01 6.97282016e-01
6.00916684e-01 -4.84901786e-01 6.90382600e-01 5.49663901e-01
-1.02307129e+00 5.50328255e-01 5.34435451e-01 2.58427352e-01
4.77128386e-01 -8.53565037e-01 -3.45742285e-01 8.79357517e-01
3.47446978e-01 5.75780451e-01 5.22545815e-01 -3.79318297e-01
-1.70585179e+00 -1.11418521e+00 2.55251497e-01 3.76953244e-01
6.08143568e-01 -1.86990142e-01 -9.61930215e-01 3.79125923e-01
-1.52626723e-01 4.89970624e-01 7.91153431e-01 -7.74133354e-02
-5.58796763e-01 -6.24482453e-01 -9.30279672e-01 8.54491442e-02
5.41812301e-01 -5.78406215e-01 -4.24079537e-01 3.84204715e-01
9.11184177e-02 -3.23096395e-01 -7.47515976e-01 4.70952511e-01
4.45079565e-01 -6.82852149e-01 1.07426810e+00 -4.79072511e-01
4.31427620e-02 -5.06862700e-01 -3.22991461e-01 -1.46989024e+00
-3.33700061e-01 -3.95299166e-01 -1.69908911e-01 9.00207996e-01
1.43663809e-01 -5.58894575e-01 6.36599839e-01 1.62826419e-01
1.50749609e-01 -7.32390404e-01 -1.09124923e+00 -1.02535212e+00
-4.02165771e-01 -2.73606211e-01 9.49074607e-03 7.97961175e-01
5.51577173e-02 3.91297609e-01 -7.45982468e-01 2.11319283e-01
7.49563992e-01 4.63739961e-01 7.56229401e-01 -1.09525955e+00
-3.54096711e-01 5.86622879e-02 -9.03772593e-01 -1.04723537e+00
2.40877062e-01 -8.98983657e-01 1.38410121e-01 -8.82272899e-01
1.14778571e-01 -5.72716236e-01 -6.70555413e-01 1.83148369e-01
-1.98281571e-01 4.78020370e-01 2.13526741e-01 4.18869793e-01
-4.00223583e-01 8.29920769e-01 8.23953629e-01 -3.22811931e-01
-2.63726503e-01 3.13247353e-01 -6.25922859e-01 5.18489897e-01
7.35288143e-01 -5.87426722e-01 -3.94211709e-01 -9.09816548e-02
-2.47961536e-01 6.86329752e-02 2.45817065e-01 -1.21547830e+00
1.15170337e-01 1.85624249e-02 2.32818246e-01 -2.70368755e-01
4.48171109e-01 -1.04921782e+00 1.33952051e-01 6.48215711e-01
-3.80503267e-01 -5.75693786e-01 -6.65498674e-02 9.23927784e-01
-6.67498946e-01 -2.43356168e-01 1.14438891e+00 4.15193588e-01
-3.26425374e-01 1.08740963e-01 -3.13372076e-01 -1.62161738e-01
9.93171751e-01 -4.90526110e-02 2.40354627e-01 -3.33309263e-01
-6.36721551e-01 -9.79135111e-02 -2.68712252e-01 2.33798400e-01
9.34033632e-01 -1.44561136e+00 -9.25711453e-01 4.91913348e-01
2.33205587e-01 -3.91608745e-01 2.15644985e-01 1.01887095e+00
-9.94080901e-02 3.39438140e-01 -5.04452437e-02 -5.81401467e-01
-1.27639723e+00 4.75325614e-01 2.71297246e-01 -2.41614133e-01
-8.38661671e-01 5.71730793e-01 1.99441284e-01 -1.57826170e-01
2.49564186e-01 -2.13631839e-01 -4.86989647e-01 1.75545793e-02
6.93255365e-01 4.01631325e-01 1.81099921e-01 -8.69784713e-01
-2.82177091e-01 7.35625327e-01 -1.23647705e-01 1.36379898e-01
1.61246014e+00 -8.82644057e-02 -1.11720026e-01 3.93427163e-01
1.38432086e+00 2.32409704e-02 -8.67125154e-01 -5.00560224e-01
4.71442454e-02 -7.28564143e-01 3.77082586e-01 -2.10725114e-01
-1.36116302e+00 6.40748262e-01 7.22859383e-01 1.43393382e-01
1.29048002e+00 -6.10071242e-01 6.52915120e-01 4.86610055e-01
2.73564100e-01 -9.22430396e-01 1.77441403e-01 3.18730801e-01
1.20387721e+00 -1.19709146e+00 1.57878906e-01 -6.67892873e-01
-5.57707787e-01 1.10049987e+00 3.22861224e-01 -4.64485526e-01
9.11199689e-01 -9.29883420e-02 -2.05743294e-02 -1.17962593e-02
-4.19019997e-01 -1.66775025e-02 5.63243151e-01 3.97248149e-01
1.87066615e-01 -3.85387279e-02 -4.37374592e-01 5.55491149e-01
2.57304251e-01 -4.19777259e-02 1.91954538e-01 8.09714019e-01
-3.54909956e-01 -9.80907440e-01 -6.47646964e-01 4.35937732e-01
-3.45840096e-01 1.37695417e-01 -1.53449357e-01 1.82108432e-01
-2.91577548e-01 9.99010086e-01 -4.25374120e-01 -3.17355543e-01
3.66421074e-01 -1.70522541e-01 1.23456135e-01 -6.98040485e-01
-4.77195799e-01 2.85082966e-01 -2.80394703e-01 -5.33439577e-01
-4.52534348e-01 -5.69170594e-01 -9.61105227e-01 3.56758475e-01
-5.41960537e-01 4.04309452e-01 5.85993409e-01 7.80330539e-01
2.93175519e-01 2.63388008e-01 1.28277731e+00 -5.53987563e-01
-9.10911918e-01 -7.92550206e-01 -8.70216608e-01 4.26301271e-01
3.34550679e-01 -7.14669049e-01 -5.37666440e-01 -6.22225553e-02] | [12.44161319732666, 0.4167676270008087] |
83ded141-04d3-40a2-8023-5bf1359560b9 | deep-convolutional-forest-a-dynamic-deep | 2110.15718 | null | https://arxiv.org/abs/2110.15718v3 | https://arxiv.org/pdf/2110.15718v3.pdf | Deep convolutional forest: a dynamic deep ensemble approach for spam detection in text | The increase in people's use of mobile messaging services has led to the spread of social engineering attacks like phishing, considering that spam text is one of the main factors in the dissemination of phishing attacks to steal sensitive data such as credit cards and passwords. In addition, rumors and incorrect medical information regarding the COVID-19 pandemic are widely shared on social media leading to people's fear and confusion. Thus, filtering spam content is vital to reduce risks and threats. Previous studies relied on machine learning and deep learning approaches for spam classification, but these approaches have two limitations. Machine learning models require manual feature engineering, whereas deep neural networks require a high computational cost. This paper introduces a dynamic deep ensemble model for spam detection that adjusts its complexity and extracts features automatically. The proposed model utilizes convolutional and pooling layers for feature extraction along with base classifiers such as random forests and extremely randomized trees for classifying texts into spam or legitimate ones. Moreover, the model employs ensemble learning procedures like boosting and bagging. As a result, the model achieved high precision, recall, f1-score and accuracy of 98.38\%. | ['Shawkat K. Guirguis', 'Yasser F. Hassan', 'Mai A. Shaaban'] | 2021-10-10 | null | null | null | null | ['spam-detection'] | ['natural-language-processing'] | [-8.33577663e-02 -2.05139473e-01 -5.65766282e-02 -2.53979594e-01
-5.65834753e-02 -3.93285453e-01 6.21650815e-01 3.44967932e-01
-5.09518445e-01 7.41545677e-01 4.34079655e-02 -7.05254674e-01
2.09831893e-01 -1.16644681e+00 -3.26034278e-02 -6.40510619e-01
1.39148414e-01 1.25086457e-01 4.78542268e-01 -4.64911550e-01
7.13342130e-01 3.86948735e-01 -1.30388927e+00 6.10906899e-01
1.20514405e+00 1.06971693e+00 -2.78717935e-01 5.98732889e-01
-4.90688533e-01 7.33122706e-01 -7.54013360e-01 -7.83444583e-01
-1.19490974e-01 -1.68505952e-01 -5.77197731e-01 -5.27768552e-01
-8.98439214e-02 -5.72872281e-01 -3.86122406e-01 1.13522434e+00
4.70161080e-01 -2.55722612e-01 6.60470605e-01 -1.27140892e+00
-7.42415667e-01 2.35772654e-01 -5.57475209e-01 4.01465297e-01
2.12118223e-01 6.66304752e-02 4.23149526e-01 -7.73682415e-01
1.10380575e-01 1.09239900e+00 1.03696167e+00 5.71876884e-01
-8.63671720e-01 -1.19607711e+00 -2.35759124e-01 3.77415150e-01
-8.57589781e-01 -3.17581177e-01 4.52917635e-01 -6.32673144e-01
7.09175467e-01 2.01593667e-01 6.04864359e-01 1.21392822e+00
8.31414104e-01 5.27115047e-01 1.23960042e+00 -1.40917897e-01
1.77207902e-01 7.22229123e-01 5.77252865e-01 5.37051797e-01
7.60362506e-01 6.63464516e-02 -3.71415228e-01 -7.36311674e-01
2.14478254e-01 4.67932314e-01 1.95790995e-02 7.20040500e-02
-5.72514117e-01 1.36699367e+00 5.50230861e-01 4.38590378e-01
-4.08451527e-01 -2.54260391e-01 7.09803641e-01 2.68816084e-01
6.94985509e-01 2.29196504e-01 -3.71133506e-01 1.53243318e-01
-7.06864893e-01 2.77763098e-01 9.53576982e-01 1.76372126e-01
4.56786841e-01 -2.03484371e-01 2.78203785e-01 6.22451067e-01
4.95025545e-01 6.17852092e-01 8.04655254e-01 -1.19528398e-01
3.79136831e-01 9.54660773e-01 1.26674280e-01 -1.82816565e+00
-4.53928888e-01 -4.54576612e-01 -1.11145782e+00 6.01868108e-02
3.76150489e-01 -1.93597436e-01 -9.02509034e-01 1.26271355e+00
2.62774706e-01 1.23131290e-01 -2.41724864e-01 5.01370847e-01
8.43200028e-01 4.54902589e-01 3.72843325e-01 3.49375680e-02
1.54122257e+00 -4.66956943e-01 -7.30982363e-01 -2.45246500e-01
5.99142790e-01 -6.29848123e-01 6.40391231e-01 4.14966434e-01
-5.88365436e-01 -1.78531781e-01 -9.94256735e-01 3.17695588e-01
-9.62869883e-01 -3.38751018e-01 6.90213442e-01 1.40997064e+00
-6.39636099e-01 6.72369242e-01 -5.29199362e-01 -3.81055713e-01
8.48854184e-01 4.88980621e-01 -7.06993565e-02 1.23440675e-01
-1.73489475e+00 1.22153986e+00 3.23583215e-01 9.68303084e-02
-2.19532087e-01 -1.64827526e-01 -4.68322426e-01 7.47632012e-02
-8.50076824e-02 -5.21557927e-01 9.29594994e-01 -9.71583426e-01
-1.14280367e+00 7.02275276e-01 -8.02870244e-02 -6.14544153e-01
3.84389997e-01 -3.58841598e-01 -7.47250021e-01 -1.89673230e-02
6.00864273e-03 -1.50523171e-01 1.09400368e+00 -8.39646220e-01
-6.21440589e-01 -7.91138589e-01 -3.80751967e-01 -1.39403939e-01
-6.21636629e-01 3.18898499e-01 4.95030046e-01 -5.50062299e-01
2.91660607e-01 -6.58175707e-01 -2.33323321e-01 -6.96250021e-01
-1.92581728e-01 -1.86869055e-01 1.15450680e+00 -1.16473472e+00
1.66401958e+00 -1.65101957e+00 -6.60440207e-01 4.40905124e-01
6.30272627e-01 8.75223160e-01 3.88329625e-01 4.33502406e-01
2.24224791e-01 2.82432914e-01 -1.99889038e-02 3.19483280e-01
-3.61419797e-01 -3.32739979e-01 -5.90213776e-01 4.75160897e-01
-5.76925799e-02 7.19197571e-01 -7.45820403e-01 -2.97885150e-01
2.89542437e-01 4.25334871e-01 -3.29380870e-01 9.73307714e-02
1.11543618e-01 2.49213561e-01 -6.40388250e-01 5.54990053e-01
1.06592453e+00 -3.60886723e-01 1.74579605e-01 1.33433446e-01
2.00141296e-01 5.72080016e-01 -7.58626759e-01 5.10142565e-01
-4.63923633e-01 3.72296542e-01 6.97939890e-03 -9.69389796e-01
1.03170359e+00 2.48200059e-01 1.20779224e-01 -2.83403486e-01
7.74632335e-01 4.97309476e-01 -9.42171142e-02 -6.56051099e-01
3.53221208e-01 -1.14379980e-01 -4.07265350e-02 6.49911761e-01
-3.31757307e-01 4.26038593e-01 -3.38476330e-01 3.20174217e-01
1.05348086e+00 -4.35195655e-01 5.81166565e-01 -1.67423189e-01
9.15397286e-01 -1.45535961e-01 1.55197442e-01 7.36232340e-01
-3.83736998e-01 8.45298823e-03 3.70224208e-01 -6.39268875e-01
-7.11804390e-01 -8.47133100e-01 -1.55021310e-01 1.19911218e+00
2.74466008e-01 -8.38964209e-02 -8.91429961e-01 -1.15274405e+00
2.50792474e-01 7.15942919e-01 -5.67305982e-01 -3.67689461e-01
-5.50157785e-01 -1.18230593e+00 6.46624506e-01 2.24110290e-01
9.87463295e-01 -1.20822501e+00 -4.40565139e-01 3.54613394e-01
-1.95381030e-01 -4.93591309e-01 7.74689903e-03 -1.45580759e-02
-8.66254926e-01 -1.46485186e+00 -4.07372773e-01 -5.94517767e-01
5.06326675e-01 3.84325415e-01 6.89834476e-01 5.17327130e-01
-1.72692478e-01 -3.82836074e-01 -3.80803853e-01 -5.10197341e-01
-6.35637641e-01 4.13554639e-01 1.10269688e-01 5.33996783e-02
1.09431779e+00 -3.72985184e-01 -7.03454912e-01 3.81471634e-01
-7.01492846e-01 -2.29415074e-01 5.27478158e-01 9.12731647e-01
-7.24478424e-01 -2.90345717e-02 1.36745131e+00 -1.22489071e+00
1.08717787e+00 -9.78028476e-01 -2.13214770e-01 -4.39769402e-03
-8.61568153e-01 -3.30035269e-01 5.36111653e-01 -2.70621389e-01
-1.21470129e+00 -3.34553182e-01 -2.83842236e-01 5.07442892e-01
-2.14508742e-01 2.79069275e-01 1.00167030e-02 -1.61050886e-01
9.98389661e-01 3.54650497e-01 4.18353945e-01 -3.45948786e-01
5.13429604e-02 1.56970322e+00 -2.02975258e-01 2.37150759e-01
7.32467890e-01 3.39589417e-01 -4.51786101e-01 -8.84223640e-01
-7.16306031e-01 -6.22389257e-01 -3.71149033e-01 -1.31465003e-01
7.99937606e-01 -5.09534299e-01 -1.13210297e+00 1.20316303e+00
-1.28451514e+00 5.39522648e-01 7.76625752e-01 2.34488621e-01
1.06593028e-01 7.17348397e-01 -9.12515998e-01 -1.03630614e+00
-8.03182483e-01 -6.80966198e-01 2.62317121e-01 2.28247821e-01
-5.37632406e-01 -9.86467719e-01 -4.05916609e-02 5.85362256e-01
9.53913629e-01 1.53823188e-02 7.22413063e-01 -1.26174653e+00
-1.24420404e-01 -7.92302251e-01 -6.27509296e-01 4.95567679e-01
1.80394739e-01 -4.11483914e-01 -1.06758237e+00 -2.98914790e-01
3.61129224e-01 -2.33780235e-01 9.79700208e-01 2.40681425e-01
1.03951240e+00 -6.84449136e-01 -5.81081152e-01 1.99868277e-01
9.91737485e-01 4.85421449e-01 7.52720654e-01 6.82064712e-01
4.71505553e-01 7.71885574e-01 1.70882717e-01 2.84628749e-01
2.47097865e-01 1.42033607e-01 4.47267979e-01 2.87055552e-01
5.87724507e-01 -1.41074941e-01 2.22018242e-01 4.49408859e-01
1.51036412e-01 2.09675841e-02 -8.83143961e-01 1.86638340e-01
-1.79112196e+00 -1.11625421e+00 -4.12579536e-01 2.20652986e+00
6.49541140e-01 9.70216990e-02 1.62432358e-01 3.64622176e-01
1.10581195e+00 1.58383518e-01 -4.25215989e-01 -4.71820474e-01
2.84265667e-01 1.65560395e-01 5.78313112e-01 3.68377090e-01
-1.47539449e+00 8.05620313e-01 5.55429649e+00 7.14672863e-01
-1.15500915e+00 9.62735862e-02 7.21908569e-01 3.84188503e-01
5.11253346e-03 -3.56429279e-01 -8.63295019e-01 1.01880705e+00
9.51448977e-01 -5.23953559e-03 1.81896687e-01 8.36685479e-01
1.68263316e-01 -2.70096838e-01 -1.61172599e-01 7.30624914e-01
1.48539364e-01 -1.34474540e+00 -1.24765649e-01 1.63992226e-01
4.54308480e-01 -1.77341089e-01 -1.86280962e-02 4.24836218e-01
4.02088344e-01 -1.09616816e+00 5.06020617e-03 3.20007205e-01
2.29674071e-01 -7.96205580e-01 1.28601205e+00 5.70129216e-01
-5.20639658e-01 -3.73234600e-01 -1.66233540e-01 -2.18559101e-01
2.19349533e-01 8.08063447e-01 -1.15784895e+00 1.11534998e-01
7.64517963e-01 3.97465974e-01 -7.15547144e-01 8.40530455e-01
-1.25957383e-02 8.14466834e-01 -1.90865606e-01 -6.36283100e-01
5.49531467e-02 -6.66401610e-02 2.79496938e-01 1.18640053e+00
1.97834149e-01 -9.95961651e-02 -1.59416318e-01 4.27029580e-01
2.16911867e-01 2.72066176e-01 -6.76700473e-01 -5.79206944e-02
3.53388786e-01 1.13162935e+00 -7.59203851e-01 -3.26004058e-01
-1.96017280e-01 7.30201423e-01 -8.86203796e-02 7.65539855e-02
-8.73594642e-01 -7.43065774e-01 5.12094200e-01 5.27100146e-01
-1.36537254e-01 9.40406173e-02 -9.36324894e-01 -1.16609538e+00
-3.37630838e-01 -9.32364225e-01 3.96837950e-01 -7.78395534e-02
-1.50478804e+00 5.33227742e-01 -5.44490874e-01 -9.56412435e-01
-8.26173201e-02 -6.61743343e-01 -6.76228762e-01 1.07185996e+00
-1.02884090e+00 -9.37120378e-01 -3.31671655e-01 3.60231251e-01
2.00293958e-01 -6.04133308e-01 7.77551770e-01 2.04848886e-01
-3.97145152e-01 2.75614083e-01 1.62942067e-01 3.38535756e-01
5.84830403e-01 -7.77763665e-01 6.12234473e-01 4.91058439e-01
-7.07080066e-01 9.15027022e-01 5.40275991e-01 -1.10109019e+00
-5.83130121e-01 -8.58925223e-01 1.32808614e+00 -5.64310074e-01
8.15669298e-01 -2.93784469e-01 -1.10446477e+00 2.78882265e-01
-2.89597865e-02 -5.44315398e-01 8.18482578e-01 3.83374467e-02
-3.84931386e-01 1.32311270e-01 -1.57900023e+00 4.57358539e-01
4.23761845e-01 -3.31409186e-01 -6.01546466e-01 1.77802101e-01
2.45739520e-01 1.44793943e-01 -2.51654446e-01 1.90212891e-01
7.69573689e-01 -1.34725404e+00 7.90916145e-01 -9.63444829e-01
4.17473584e-01 8.78812820e-02 3.25790286e-01 -1.05968308e+00
-3.80403578e-01 -1.36554778e-01 -2.67922044e-01 8.41143191e-01
2.66000628e-01 -1.20604002e+00 1.16022551e+00 4.23892945e-01
4.76254642e-01 -7.70762682e-01 -6.99528039e-01 -2.63402432e-01
1.39374599e-01 -2.09679827e-01 4.84676331e-01 1.14621055e+00
1.94570795e-01 4.51830357e-01 -5.12278914e-01 -9.86730158e-02
7.50056148e-01 -4.23634171e-01 6.48449957e-01 -1.52259517e+00
1.36163533e-01 -4.94401038e-01 -3.29900235e-01 -6.55233026e-01
-7.64136240e-02 -6.52613401e-01 -4.67121184e-01 -1.21701980e+00
1.29294038e-01 -5.59671938e-01 -8.86289701e-02 1.62731707e-01
-2.59090036e-01 2.70034194e-01 -8.00214782e-02 3.73924792e-01
-2.45917395e-01 2.54332483e-01 8.20151150e-01 -7.98780769e-02
-3.02046061e-01 7.62158036e-01 -8.83239329e-01 1.02757418e+00
1.36401153e+00 -4.68385309e-01 -8.68739337e-02 1.96071386e-01
4.68789279e-01 -1.95243642e-01 3.91488254e-01 -5.81722200e-01
9.91350934e-02 -1.15606502e-01 6.14440143e-01 -5.67172885e-01
4.58915532e-03 -7.60911405e-01 -2.39094943e-01 1.05582666e+00
-1.78546742e-01 -5.10226302e-02 -8.97498131e-02 7.51590729e-01
6.27141818e-02 -3.85830402e-01 9.55295920e-01 -1.68400139e-01
-3.53237689e-01 -1.67618394e-02 -8.32184017e-01 -2.77757674e-01
1.24900377e+00 -2.85284847e-01 -7.16374636e-01 -5.28730690e-01
-6.29639089e-01 7.65910596e-02 7.84440786e-02 5.91325581e-01
6.77003384e-01 -1.01579332e+00 -6.31711721e-01 4.35490608e-01
-2.67517626e-01 -5.49015641e-01 3.52779895e-01 9.24261570e-01
-8.88322234e-01 3.75194937e-01 -3.33562106e-01 -1.49385110e-01
-1.42152035e+00 3.76322627e-01 2.67605692e-01 -3.76875222e-01
-2.42368713e-01 5.67297816e-01 -1.96135297e-01 -6.88589990e-01
1.45901397e-01 4.78104830e-01 -6.60405397e-01 2.35496491e-01
8.78839254e-01 9.62735295e-01 2.10742310e-01 -5.88970900e-01
-5.46033919e-01 -1.35963216e-01 -6.04084849e-01 2.95348138e-01
9.89718974e-01 2.95077842e-02 -4.87376660e-01 -4.64062244e-02
1.05774105e+00 -5.99188656e-02 -2.59971321e-01 -9.60974470e-02
3.49061370e-01 -6.10320807e-01 -1.05128922e-01 -9.71970975e-01
-6.52319491e-01 9.37638044e-01 5.65583050e-01 7.87463725e-01
6.32831514e-01 -3.64864767e-01 1.16532719e+00 4.23922122e-01
3.03634763e-01 -1.00660658e+00 -2.94083729e-02 6.22751236e-01
4.20426786e-01 -1.44505191e+00 -6.66885376e-02 -4.39334720e-01
-5.75495839e-01 1.14501214e+00 6.56648636e-01 -1.58062845e-01
1.02899194e+00 -3.63010056e-02 2.38993108e-01 -8.90081078e-02
-3.59280229e-01 3.56483996e-01 9.20910295e-03 7.40954459e-01
5.61306655e-01 2.79600807e-02 -7.28796184e-01 8.79585147e-01
-1.30419880e-01 7.45742172e-02 4.09586430e-01 7.04262495e-01
-1.08099127e+00 -7.89353251e-01 -5.89921772e-01 1.03321803e+00
-8.49195302e-01 -1.69317350e-01 -4.38948989e-01 3.73232305e-01
1.34472251e-01 1.15672076e+00 -1.68684989e-01 -7.94662952e-01
-1.18185394e-01 3.23630393e-01 -1.96547344e-01 -4.82878923e-01
-1.01783776e+00 -5.76501787e-01 7.82059580e-02 -5.61909750e-02
3.88384461e-02 -3.61097246e-01 -8.22195530e-01 -9.79423285e-01
-7.07889497e-01 2.03989580e-01 8.14744592e-01 9.85534787e-01
3.43325257e-01 -1.26092816e-02 8.70364547e-01 -5.32302618e-01
-1.11389816e+00 -1.18672955e+00 -4.30218756e-01 4.26468670e-01
1.79461598e-01 -5.55406153e-01 -6.00733936e-01 -3.20220530e-01] | [7.823044300079346, 9.997925758361816] |
bf8a7c39-f3a1-437d-a3cc-3ec801945070 | causal-lifting-and-link-prediction | 2302.01198 | null | https://arxiv.org/abs/2302.01198v1 | https://arxiv.org/pdf/2302.01198v1.pdf | Causal Lifting and Link Prediction | Current state-of-the-art causal models for link prediction assume an underlying set of inherent node factors -- an innate characteristic defined at the node's birth -- that governs the causal evolution of links in the graph. In some causal tasks, however, link formation is path-dependent, i.e., the outcome of link interventions depends on existing links. For instance, in the customer-product graph of an online retailer, the effect of an 85-inch TV ad (treatment) likely depends on whether the costumer already has an 85-inch TV. Unfortunately, existing causal methods are impractical in these scenarios. The cascading functional dependencies between links (due to path dependence) are either unidentifiable or require an impractical number of control variables. In order to remedy this shortcoming, this work develops the first causal model capable of dealing with path dependencies in link prediction. It introduces the concept of causal lifting, an invariance in causal models that, when satisfied, allows the identification of causal link prediction queries using limited interventional data. On the estimation side, we show how structural pairwise embeddings -- a type of symmetry-based joint representation of node pairs in a graph -- exhibit lower bias and correctly represent the causal structure of the task, as opposed to existing node embedding methods, e.g., GNNs and matrix factorization. Finally, we validate our theoretical findings on four datasets under three different scenarios for causal link prediction tasks: knowledge base completion, covariance matrix estimation and consumer-product recommendations. | ['Bruno Ribeiro', 'Nesreen Ahmed', 'Beatrice Bevilacqua', 'Leonardo Cotta'] | 2023-02-02 | null | null | null | null | ['knowledge-base-completion', 'knowledge-base-completion'] | ['graphs', 'knowledge-base'] | [ 2.21493974e-01 6.08910203e-01 -1.02586257e+00 -1.71902508e-01
1.38792634e-01 -6.75061285e-01 6.92443728e-01 4.94402379e-01
2.18120053e-01 6.42670691e-01 6.91188514e-01 -8.41460407e-01
-1.03750658e+00 -1.21593297e+00 -1.06476855e+00 -4.29797947e-01
-5.66486061e-01 5.45184255e-01 -1.70636550e-01 -2.98107266e-01
-1.38698727e-01 1.57214776e-01 -9.58017468e-01 -2.85313334e-02
6.82167292e-01 5.59311330e-01 -5.05059697e-02 3.11234742e-01
1.26918748e-01 6.09571815e-01 5.93203455e-02 -9.20768023e-01
1.07860260e-01 -1.99826047e-01 -7.06170976e-01 -4.10326988e-01
2.70674169e-01 -3.22175562e-01 -8.45259726e-01 6.77633286e-01
4.62527908e-02 -4.20696765e-01 9.80292261e-01 -1.76535547e+00
-1.10711670e+00 1.35971081e+00 -5.33249021e-01 2.13067040e-01
5.76970637e-01 -2.10725307e-01 2.05254269e+00 -3.63876373e-01
8.82610500e-01 1.44758403e+00 8.96618247e-01 8.58973265e-02
-1.88889933e+00 -6.25889719e-01 3.68300170e-01 3.03542703e-01
-9.26610708e-01 8.32848400e-02 9.87474680e-01 -6.52253330e-01
5.10676563e-01 2.08376706e-01 7.32879519e-01 1.58647299e+00
6.82481706e-01 4.70140755e-01 7.62808323e-01 -9.83478799e-02
-2.63309509e-01 -1.31130159e-01 2.91191310e-01 5.36778390e-01
5.53661048e-01 5.19261122e-01 -5.77300191e-01 -6.72126055e-01
8.37948501e-01 7.06473961e-02 -2.33896583e-01 -5.76949358e-01
-1.04937601e+00 1.31491876e+00 7.46308565e-01 2.23406538e-01
-6.33998394e-01 4.23798382e-01 1.72888502e-01 4.01444674e-01
4.44550186e-01 3.35403591e-01 -6.79126203e-01 3.58472854e-01
-4.79443222e-01 1.93817481e-01 1.04520750e+00 8.35622668e-01
5.74260592e-01 -4.99985516e-01 -8.48262534e-02 3.53677660e-01
6.50483847e-01 1.18577749e-01 -2.60407180e-01 -3.66957188e-01
4.00060743e-01 5.49716473e-01 -1.15326591e-01 -1.52693200e+00
-5.55529296e-01 -5.07533193e-01 -9.07718778e-01 -5.85926056e-01
3.53216588e-01 -8.71406570e-02 -4.55048978e-01 2.14144278e+00
3.50180089e-01 5.30307472e-01 -5.13193071e-01 6.63455307e-01
5.91901660e-01 4.04410660e-01 4.05596137e-01 -6.03738129e-01
1.25425363e+00 -3.42976868e-01 -9.25340414e-01 1.83593199e-01
5.12745738e-01 -5.82108021e-01 7.42881060e-01 2.90807411e-02
-6.75939918e-01 -1.94790319e-01 -7.75208533e-01 1.88975260e-01
-2.39680082e-01 -4.11271602e-01 1.40150976e+00 4.32708651e-01
-8.65804374e-01 6.59923851e-01 -3.47947150e-01 -5.38528323e-01
1.99616715e-01 4.11045432e-01 -3.97559255e-01 -3.18320841e-01
-1.76754272e+00 6.64047837e-01 5.00384457e-02 -1.06745884e-02
-9.20261323e-01 -1.48545659e+00 -5.65485775e-01 2.61766940e-01
6.43003821e-01 -1.16292608e+00 7.69219518e-01 -6.06980920e-01
-7.35210121e-01 2.83516705e-01 1.12156779e-01 -4.01503175e-01
3.92108738e-01 -4.43198346e-02 -6.61800861e-01 -2.31638059e-01
1.53835446e-01 2.65341759e-01 9.34089422e-01 -1.27931225e+00
-4.64373678e-01 -4.86783981e-01 3.94381166e-01 -5.12844734e-02
-3.86628032e-01 -3.97294194e-01 -2.52845407e-01 -5.03678441e-01
-1.05098106e-01 -1.02210975e+00 -3.00597936e-01 -1.11305654e-01
-8.71759415e-01 -5.31712592e-01 6.42822206e-01 -4.22021002e-01
1.57060385e+00 -1.94730604e+00 5.20269752e-01 4.17336851e-01
6.13505006e-01 -3.26687336e-01 -3.37509155e-01 9.20808554e-01
-5.98364353e-01 6.47607446e-01 2.13359118e-01 4.90624271e-02
9.52009261e-02 4.52019274e-01 -3.33625406e-01 5.62537313e-01
8.30629319e-02 1.07876754e+00 -9.38267350e-01 -3.78533006e-01
-2.79618781e-02 5.36374807e-01 -7.20306575e-01 -4.07826342e-02
-2.76668876e-01 2.20090970e-01 -6.04983747e-01 2.39863321e-01
3.81630570e-01 -2.89261490e-01 7.31722057e-01 -6.10938907e-01
8.33123550e-02 5.84538102e-01 -1.07214594e+00 1.48868251e+00
-4.04087842e-01 3.49371910e-01 -3.45206819e-02 -1.01708961e+00
3.69516373e-01 3.22758228e-01 9.49895442e-01 -4.29853231e-01
-1.02180213e-01 -1.78480968e-01 4.26041156e-01 -4.38911438e-01
1.46216333e-01 4.05845344e-02 -2.26730341e-03 3.48332614e-01
2.25119721e-02 5.03746927e-01 2.12584585e-01 6.43667519e-01
1.55457580e+00 -2.05283046e-01 2.85044700e-01 -3.01359951e-01
-1.79917738e-02 -8.11773762e-02 6.41481340e-01 6.91736698e-01
1.60954624e-01 -1.03933692e-01 1.17440486e+00 -2.26331353e-01
-9.28579330e-01 -1.24395323e+00 -4.31104392e-01 8.94771576e-01
-6.23640418e-02 -5.47207415e-01 -2.17431068e-01 -7.97784686e-01
8.58864486e-01 8.58676255e-01 -1.14811540e+00 -4.56539929e-01
-3.35882545e-01 -9.06840086e-01 3.42062682e-01 3.05166245e-01
-3.31583351e-01 -4.60640788e-01 2.90594816e-01 3.70597243e-01
-7.83307999e-02 -7.34386861e-01 -3.15000772e-01 9.43901464e-02
-8.94185424e-01 -1.42554700e+00 1.83364913e-01 -2.27752045e-01
4.32181925e-01 1.14037886e-01 1.43428195e+00 1.43863648e-01
-3.71128201e-01 6.46566868e-01 -2.59622574e-01 -3.29885744e-02
-2.43294999e-01 -1.77107174e-02 2.42871732e-01 1.99431494e-01
1.55050933e-01 -1.00243533e+00 -8.04755449e-01 3.25969577e-01
-7.51385987e-01 -2.50506520e-01 7.24700212e-01 7.71356285e-01
3.98138732e-01 3.77162993e-01 6.39795542e-01 -1.16589773e+00
6.90155685e-01 -1.17875326e+00 -3.53286684e-01 3.09126288e-01
-1.18358612e+00 6.91803172e-02 3.18385810e-01 -6.25885725e-01
-8.77245784e-01 -2.36816064e-01 2.40995765e-01 -1.94916517e-01
2.85655051e-01 1.02138042e+00 -1.91347346e-01 3.97400528e-01
4.81937051e-01 -4.00966167e-01 -5.23663536e-02 -5.63833535e-01
1.01457238e+00 2.55216490e-02 7.74049237e-02 -4.34522659e-01
1.04745209e+00 3.72564316e-01 4.63968843e-01 -4.38837796e-01
-5.32898366e-01 -3.36910188e-01 -6.38021052e-01 -6.69401214e-02
9.34258044e-01 -8.63383055e-01 -1.13589334e+00 -4.26926047e-01
-1.19074273e+00 -7.58462846e-02 -2.13161092e-02 6.06879354e-01
-1.54444933e-01 -9.13949460e-02 -6.95744097e-01 -6.92989469e-01
1.98987067e-01 -5.96179008e-01 5.59138596e-01 -3.40640068e-01
-4.12801445e-01 -1.46740901e+00 1.94211796e-01 1.95118889e-01
4.07581665e-02 1.77328125e-01 1.67335713e+00 -4.25518304e-01
-7.37847686e-01 -1.51450008e-01 -2.66918182e-01 -4.51230645e-01
2.75673449e-01 3.26522052e-01 -3.08276623e-01 -5.78372702e-02
-6.26568377e-01 3.15652907e-01 8.20792019e-01 7.56496727e-01
7.22334564e-01 -5.56140542e-01 -9.20064688e-01 2.14388296e-01
1.38234508e+00 -9.02925208e-02 3.91994894e-01 -2.61850417e-01
1.02525842e+00 9.25429702e-01 4.51370239e-01 4.02384520e-01
6.15797460e-01 9.15054739e-01 8.61102104e-01 -1.70855731e-01
-6.50650486e-02 -1.00781786e+00 1.34330213e-01 3.81513089e-01
-3.54388580e-02 -5.51824331e-01 -5.61286509e-01 5.45313120e-01
-1.93474829e+00 -9.40460384e-01 -1.07001078e+00 2.03143907e+00
7.99692094e-01 -7.83534050e-02 2.59779721e-01 -7.73139000e-02
5.53735793e-01 1.20515488e-01 -5.10430932e-01 -1.53131247e-01
7.00810105e-02 -2.38467291e-01 8.41925144e-01 5.48502088e-01
-8.19002569e-01 5.22796810e-01 6.24661350e+00 5.33530951e-01
-4.70014900e-01 3.70299786e-01 4.69103366e-01 -1.25284344e-02
-1.12021124e+00 2.94863552e-01 -6.88679755e-01 4.11923230e-01
9.44663763e-01 -1.22999080e-01 3.71158540e-01 5.04158318e-01
4.89948899e-01 3.04348141e-01 -1.53856146e+00 4.97786939e-01
-3.65195692e-01 -1.42504013e+00 1.06772967e-01 6.15672648e-01
7.04331398e-01 -3.16410750e-01 2.89629072e-01 5.27627207e-02
8.36747885e-01 -1.00598741e+00 3.27101886e-01 4.60767448e-01
6.30817950e-01 -6.49353027e-01 4.07899231e-01 2.40455512e-02
-1.02403378e+00 -4.84296352e-01 -2.39706278e-01 -1.07477300e-01
5.73526382e-01 1.32864904e+00 -8.52720201e-01 8.82486284e-01
4.82419610e-01 8.58013451e-01 -3.45945537e-01 5.00238657e-01
-4.28649396e-01 9.44766462e-01 -6.07413575e-02 1.69249505e-01
-6.12699129e-02 -3.91776294e-01 5.49933076e-01 7.10166097e-01
1.24753915e-01 1.36501238e-01 -1.16488688e-01 1.06484747e+00
-2.63234943e-01 8.48215669e-02 -1.01622975e+00 -2.68106818e-01
5.01666963e-01 1.24853480e+00 -2.66238391e-01 2.96207786e-01
-6.26671195e-01 6.50216877e-01 3.22008669e-01 4.57236260e-01
-1.02260697e+00 5.29671729e-01 1.02858639e+00 5.77760279e-01
9.93173793e-02 -7.51802027e-02 -1.71005398e-01 -7.61862338e-01
-2.61043817e-01 -4.92748231e-01 5.17595649e-01 -3.88966620e-01
-1.71177959e+00 -1.55272722e-01 1.39429430e-02 -7.26609528e-01
-1.70254320e-01 -4.29460883e-01 -4.96451199e-01 6.59847260e-01
-1.25220907e+00 -1.48896945e+00 3.67742389e-01 6.04077637e-01
-5.31906933e-02 1.91833451e-01 7.20989764e-01 5.94677329e-01
-7.20570147e-01 4.38792259e-01 -1.13142736e-01 -1.75176322e-01
6.74273610e-01 -1.24877727e+00 4.77302819e-02 5.11235058e-01
4.09470022e-01 9.17027235e-01 8.92032325e-01 -1.14418650e+00
-1.88737118e+00 -9.88427937e-01 1.21435416e+00 -7.69921839e-01
1.41685510e+00 -5.18746734e-01 -5.88762581e-01 9.66495752e-01
1.02771588e-01 -6.50270060e-02 8.44541669e-01 1.13966143e+00
-6.15973175e-01 -2.30565920e-01 -6.89551592e-01 6.55167401e-01
1.63241220e+00 -4.97656941e-01 -2.50101924e-01 5.35815895e-01
1.13005447e+00 4.28429991e-01 -1.29751146e+00 3.94935966e-01
6.07852697e-01 -7.54215837e-01 1.29572642e+00 -1.07282066e+00
9.64954197e-01 6.94470629e-02 -1.44952729e-01 -1.47702110e+00
-9.22307551e-01 -6.06640399e-01 -2.32623234e-01 1.37383044e+00
7.87073731e-01 -5.23188949e-01 4.91547763e-01 5.22822022e-01
1.67300597e-01 -6.66986883e-01 -9.16635036e-01 -5.44621527e-01
7.24296048e-02 -4.73454028e-01 6.34776413e-01 1.36894143e+00
-9.19487402e-02 7.96460271e-01 -5.30463278e-01 3.97505879e-01
7.36569583e-01 7.90408254e-02 5.02477586e-01 -1.80790424e+00
-5.97524583e-01 -3.78046095e-01 -2.78509974e-01 -7.86188722e-01
3.82046908e-01 -9.06320095e-01 -6.37001514e-01 -1.60393679e+00
3.59217763e-01 -7.84117281e-01 -3.34495187e-01 3.23529750e-01
-1.32024199e-01 -2.84505635e-01 3.00113000e-02 1.27272010e-01
2.13872138e-02 4.35202777e-01 1.26455867e+00 -2.74853706e-01
-1.40986651e-01 1.69697687e-01 -8.60411882e-01 4.28077489e-01
3.43668491e-01 -6.88027442e-01 -9.78274524e-01 -1.08329222e-01
9.98293757e-01 6.21470928e-01 7.05209255e-01 1.42726481e-01
4.86506149e-02 -2.04998046e-01 1.40567748e-02 -2.22362965e-01
3.54296193e-02 -1.21915603e+00 6.83082998e-01 5.70019841e-01
-6.26520991e-01 1.50880143e-01 -2.76360601e-01 1.36288500e+00
1.15274847e-01 1.24481790e-01 -2.46516719e-01 4.00911480e-01
-2.26957142e-01 6.06838107e-01 -1.22438394e-01 -3.54719877e-01
8.43721271e-01 3.23997647e-01 -4.69399422e-01 -5.38337588e-01
-6.82250977e-01 1.65383562e-01 -8.12297165e-02 7.29654789e-01
3.40934545e-01 -1.63753283e+00 -7.52450466e-01 -2.73096770e-01
9.87661816e-03 -7.75972545e-01 2.03486338e-01 1.29482722e+00
4.71180022e-01 4.87442702e-01 2.31738821e-01 -1.73100457e-01
-1.12455654e+00 1.02388752e+00 -1.73363045e-01 -5.91839373e-01
-2.43972674e-01 7.08529174e-01 4.60090190e-01 -2.72852927e-01
-1.30595744e-01 -1.70870245e-01 -3.25222433e-01 6.12000525e-01
-1.31925449e-01 3.13589424e-01 -2.81277776e-01 -3.95441204e-01
-3.32789779e-01 2.00045019e-01 1.83606539e-02 1.76487286e-02
1.42299604e+00 -1.34196371e-01 -4.21765238e-01 5.08196175e-01
1.31117320e+00 7.47392848e-02 -9.45869803e-01 -1.57896578e-01
-1.58044323e-01 -3.66955012e-01 1.98595256e-01 -8.43120158e-01
-1.08531153e+00 4.54965740e-01 3.70643407e-01 6.99968696e-01
6.83505654e-01 3.08493644e-01 3.35579902e-01 -1.96518257e-01
4.01314586e-01 -5.21149397e-01 -2.41553947e-01 -6.90054223e-02
9.59766388e-01 -1.07837462e+00 2.05830857e-01 -1.03160572e+00
-1.28144830e-01 7.18522727e-01 1.28542975e-01 1.45799130e-01
1.16682243e+00 -9.11223367e-02 -6.41862333e-01 -5.73908925e-01
-1.08223271e+00 -6.89183399e-02 5.78564823e-01 5.23795724e-01
6.16287947e-01 5.71436405e-01 -7.82174110e-01 5.66791892e-01
-6.53237030e-02 -4.68223035e-01 3.07118177e-01 1.04928873e-01
3.61278296e-01 -1.37536156e+00 -5.62297516e-02 7.89667368e-01
-3.50373656e-01 -3.85600626e-01 -5.87189615e-01 1.11133015e+00
1.61206469e-01 1.23490429e+00 2.03842014e-01 -5.98581731e-01
3.73257160e-01 -6.47173375e-02 2.95981437e-01 -4.00594711e-01
-2.43913218e-01 9.40079056e-03 3.88628662e-01 -7.34923065e-01
-4.43352282e-01 -9.66198802e-01 -8.10883939e-01 -8.06451321e-01
-4.60251391e-01 -1.30402178e-01 4.29278314e-01 7.65724421e-01
4.73668218e-01 7.50238895e-01 6.95245564e-01 -2.40575805e-01
-4.38860387e-01 -8.55466366e-01 -8.34231853e-01 6.51475489e-01
7.05024824e-02 -1.08388853e+00 -4.67995435e-01 6.96847076e-03] | [7.8410258293151855, 5.485200881958008] |
55ec2d87-00a2-4d9d-8007-4e46d1b6a052 | a-dynamic-feature-interaction-framework-for | 2306.05061 | null | https://arxiv.org/abs/2306.05061v1 | https://arxiv.org/pdf/2306.05061v1.pdf | A Dynamic Feature Interaction Framework for Multi-task Visual Perception | Multi-task visual perception has a wide range of applications in scene understanding such as autonomous driving. In this work, we devise an efficient unified framework to solve multiple common perception tasks, including instance segmentation, semantic segmentation, monocular 3D detection, and depth estimation. Simply sharing the same visual feature representations for these tasks impairs the performance of tasks, while independent task-specific feature extractors lead to parameter redundancy and latency. Thus, we design two feature-merge branches to learn feature basis, which can be useful to, and thus shared by, multiple perception tasks. Then, each task takes the corresponding feature basis as the input of the prediction task head to fulfill a specific task. In particular, one feature merge branch is designed for instance-level recognition the other for dense predictions. To enhance inter-branch communication, the instance branch passes pixel-wise spatial information of each instance to the dense branch using efficient dynamic convolution weighting. Moreover, a simple but effective dynamic routing mechanism is proposed to isolate task-specific features and leverage common properties among tasks. Our proposed framework, termed D2BNet, demonstrates a unique approach to parameter-efficient predictions for multi-task perception. In addition, as tasks benefit from co-training with each other, our solution achieves on par results on partially labeled settings on nuScenes and outperforms previous works for 3D detection and depth estimation on the Cityscapes dataset with full supervision. | ['Yifan Liu', 'Chunhua Shen', 'Yanning Zhang', 'Peng Wang', 'Ning Wang', 'Hao Chen', 'Yuling Xi'] | 2023-06-08 | null | null | null | null | ['scene-understanding'] | ['computer-vision'] | [ 3.60744834e-01 -4.97476719e-02 -2.43611202e-01 -5.48377812e-01
-7.22364962e-01 -5.78550339e-01 4.04678285e-01 7.82113075e-02
-5.00971317e-01 3.77888352e-01 -2.32631713e-01 -2.46602327e-01
9.76834912e-03 -5.69857001e-01 -8.01948905e-01 -7.75550961e-01
2.13005364e-01 2.85149217e-01 7.39347816e-01 4.39661182e-02
3.09911340e-01 4.61370349e-01 -1.95530498e+00 5.66464603e-01
7.94836164e-01 1.49745798e+00 9.39284027e-01 5.72972000e-01
-1.36469781e-01 4.88925487e-01 -3.34766358e-01 -8.51017796e-03
4.26276088e-01 3.25319082e-01 -6.76188946e-01 2.80871093e-01
6.49411380e-01 -4.52320993e-01 1.94419697e-02 7.56368995e-01
4.99150515e-01 2.76221961e-01 7.49198556e-01 -1.42344892e+00
-1.42224774e-01 7.12883174e-02 -8.91905308e-01 2.30070472e-01
8.02054629e-02 2.74688870e-01 9.21757758e-01 -8.91860902e-01
3.73887926e-01 1.32465398e+00 2.47881904e-01 4.43164647e-01
-9.55370843e-01 -5.85062861e-01 6.91898167e-01 1.92596465e-01
-1.11079335e+00 -3.97540301e-01 5.58456361e-01 -5.20192504e-01
1.23753607e+00 -2.34697647e-02 4.11673546e-01 8.99383962e-01
1.89654544e-01 1.05499506e+00 1.13096011e+00 4.82012257e-02
1.47030696e-01 1.18045963e-01 2.45503455e-01 8.10437083e-01
4.02446277e-02 7.28760958e-02 -7.09918916e-01 2.35951677e-01
6.56287968e-01 2.54595429e-02 -1.44590825e-01 -4.74189222e-01
-1.10373926e+00 6.96372986e-01 5.35170078e-01 -9.59686860e-02
-2.56806374e-01 2.54968822e-01 3.68804902e-01 1.97741479e-01
4.06979203e-01 1.24724293e-02 -8.18578362e-01 1.58859149e-01
-5.99317551e-01 2.48195589e-01 3.90514642e-01 1.01197875e+00
1.30370176e+00 -3.39568764e-01 -4.83325779e-01 9.09213185e-01
3.54862660e-01 5.38156986e-01 2.51843959e-01 -1.16794336e+00
8.22042823e-01 6.33287311e-01 7.37748714e-03 -7.92262077e-01
-7.28923440e-01 -3.33350599e-01 -6.91918015e-01 2.91818678e-01
4.15923089e-01 3.39748114e-02 -1.24648869e+00 1.68719053e+00
6.48989916e-01 8.41606781e-02 5.15915779e-03 1.06225455e+00
8.83094966e-01 5.90345681e-01 6.90782443e-02 2.12631300e-01
1.52366388e+00 -1.28149033e+00 -4.04483974e-02 -9.72901881e-01
5.99713445e-01 -6.11362278e-01 8.82266283e-01 3.81061465e-01
-8.55880618e-01 -9.57334518e-01 -9.46119487e-01 -5.00359297e-01
-3.88734072e-01 1.90327629e-01 8.34062278e-01 4.40476000e-01
-1.01510942e+00 1.92198396e-01 -6.23832226e-01 -1.80824593e-01
6.64908111e-01 3.71171087e-01 -3.03537220e-01 -3.68049383e-01
-7.21020162e-01 7.71322370e-01 4.72650051e-01 -4.72369492e-02
-1.18843472e+00 -6.65954173e-01 -1.01750112e+00 -3.51185873e-02
5.14220357e-01 -9.55883980e-01 1.23855639e+00 -6.31621063e-01
-1.19920540e+00 1.09196210e+00 -4.94049400e-01 -4.37634379e-01
3.93439382e-01 -2.88598984e-01 1.46983624e-01 1.63966626e-01
4.38432306e-01 1.33382046e+00 1.24786651e+00 -1.34870577e+00
-1.28041935e+00 -6.08984768e-01 2.14177370e-01 5.68101883e-01
-1.31510824e-01 -4.64373171e-01 -9.35248017e-01 -1.04938075e-01
4.60120678e-01 -7.04290390e-01 -4.23718125e-01 2.13443413e-01
-3.31345171e-01 -2.15728790e-01 6.96637213e-01 -2.16613531e-01
6.07545733e-01 -2.42388034e+00 3.27612221e-01 8.86362493e-02
5.35699189e-01 -8.98548737e-02 -1.50084674e-01 -1.56171605e-01
4.32556242e-01 -2.79771060e-01 -1.79357573e-01 -7.64950335e-01
-5.62353879e-02 4.37909812e-01 -1.72046348e-01 3.09475660e-01
2.43713543e-01 9.07802343e-01 -7.28399754e-01 -5.40311038e-01
4.79692996e-01 2.02386811e-01 -5.58703065e-01 1.16705701e-01
-3.31774354e-01 4.47066844e-01 -6.97420120e-01 6.96121454e-01
8.64818156e-01 -3.67548347e-01 -1.15833044e-01 -2.69775122e-01
-1.95413709e-01 1.87049940e-01 -1.10394490e+00 1.97855830e+00
-8.36758077e-01 4.46475714e-01 2.34980866e-01 -1.15443516e+00
1.05472636e+00 -1.38234362e-01 2.19674274e-01 -9.41824734e-01
1.17057905e-01 2.93720543e-01 -2.35127479e-01 -3.40651333e-01
5.79177320e-01 3.30550015e-01 -1.87591940e-01 2.28943676e-01
1.24705257e-02 -3.32192004e-01 5.98949678e-02 9.14267451e-03
8.91472340e-01 1.60763025e-01 2.49307171e-01 4.26217988e-02
4.52974439e-01 -4.57558259e-02 6.06977940e-01 8.81205857e-01
-4.39025253e-01 4.67404872e-01 2.62451500e-01 -3.42744023e-01
-7.22481251e-01 -1.03721833e+00 -9.56053436e-02 1.32270944e+00
6.23144805e-01 -1.00164391e-01 -3.72844875e-01 -7.07287252e-01
3.54195267e-01 3.06233585e-01 -5.65428853e-01 6.44178828e-03
-2.88327157e-01 -4.77431297e-01 2.23736972e-01 6.45939648e-01
7.92515695e-01 -9.49135065e-01 -1.16566312e+00 2.54387021e-01
-1.22056738e-01 -1.58500767e+00 -2.79105902e-01 7.79983759e-01
-7.79284716e-01 -9.85946774e-01 -5.98812640e-01 -7.77260602e-01
4.19898123e-01 1.03963959e+00 9.63444412e-01 -2.91821361e-01
-3.26711953e-01 4.28026050e-01 -1.27449378e-01 -3.82898510e-01
9.36103314e-02 2.84519285e-01 -2.74166405e-01 1.04785211e-01
2.78623402e-01 -4.92000699e-01 -7.34765470e-01 4.63129044e-01
-6.22785985e-01 3.43646288e-01 7.01901138e-01 5.49264967e-01
8.87842298e-01 -3.41108948e-01 3.78904760e-01 -6.19138181e-01
1.67127296e-01 -4.09691751e-01 -6.71082020e-01 8.28014985e-02
-2.66665131e-01 6.95167929e-02 3.94543648e-01 -1.37746632e-01
-1.01724470e+00 4.48522657e-01 6.26327842e-02 -6.92424417e-01
-4.10021216e-01 5.35672382e-02 -2.51833141e-01 -7.38330558e-02
4.67379868e-01 2.81485558e-01 -1.10519536e-01 -3.54589671e-01
3.86454374e-01 5.83021283e-01 3.85214269e-01 -4.66953397e-01
5.76767564e-01 6.77631855e-01 1.32081389e-01 -8.02922189e-01
-1.16617537e+00 -8.09279561e-01 -6.95962489e-01 -4.58095782e-02
1.15855157e+00 -1.30947661e+00 -7.91613579e-01 7.57128179e-01
-1.36330497e+00 -5.20795166e-01 -3.33091468e-02 3.01344186e-01
-6.05097711e-01 1.21864147e-01 -2.19141901e-01 -6.82990491e-01
-1.58306286e-01 -1.33694255e+00 1.55608571e+00 2.37435609e-01
2.21233487e-01 -7.80873358e-01 -3.27442765e-01 6.65683210e-01
8.49665254e-02 -1.11033462e-01 8.21563601e-01 -2.89198458e-01
-9.98226881e-01 3.52293491e-01 -8.53749394e-01 3.66836011e-01
-7.53590912e-02 -4.45473492e-01 -1.47616494e+00 -1.84237570e-01
-2.24606439e-01 -5.21213233e-01 1.42207670e+00 6.67806864e-01
1.44154334e+00 3.76381487e-01 -6.51007831e-01 9.55714941e-01
1.29028034e+00 5.03715649e-02 2.05162734e-01 3.12637657e-01
9.66788352e-01 7.27295339e-01 8.23087573e-01 4.89401907e-01
9.45574999e-01 6.49655700e-01 7.40705907e-01 -2.11361825e-01
-1.62160680e-01 8.58553685e-03 3.89196485e-01 2.42181763e-01
1.97128102e-01 -1.17068447e-01 -7.95919657e-01 5.03254950e-01
-1.91575122e+00 -5.62696993e-01 -1.38065135e-02 1.99229872e+00
4.37547147e-01 3.67513180e-01 1.05377298e-03 2.22302377e-02
3.45894188e-01 3.94415826e-01 -8.62751544e-01 -3.55190843e-01
-1.28206432e-01 5.14669493e-02 7.07985044e-01 3.97115558e-01
-1.25597405e+00 1.07868946e+00 5.14605618e+00 9.62682366e-01
-1.01635385e+00 2.60385007e-01 6.72865987e-01 -1.83588013e-01
-1.91762671e-01 -1.73153222e-01 -1.34930992e+00 1.51513487e-01
3.67348731e-01 1.87631473e-01 2.32868627e-01 9.56002593e-01
8.46888497e-02 -5.22093594e-01 -1.16719174e+00 1.22002232e+00
1.90849742e-03 -1.15774941e+00 6.21264800e-02 2.33585965e-02
6.57108545e-01 2.84824282e-01 2.16441348e-01 1.21807337e-01
1.58017695e-01 -8.64915192e-01 7.66734421e-01 1.42475858e-01
7.51388013e-01 -6.34700596e-01 3.12362790e-01 4.59169507e-01
-1.54041684e+00 -4.18019056e-01 -4.43408132e-01 -1.27999499e-01
1.24678381e-01 7.63262689e-01 -7.53971338e-01 5.77928245e-01
7.81955838e-01 9.03810322e-01 -5.19514918e-01 1.04171729e+00
-1.50686949e-01 -2.42647901e-02 -3.55195105e-01 1.21876620e-01
5.61793864e-01 7.27924630e-02 3.19994658e-01 1.07107306e+00
2.25311816e-01 -2.36526176e-01 5.97283781e-01 6.31187201e-01
1.64906889e-01 3.03569864e-02 -6.46410227e-01 5.56645155e-01
4.61832762e-01 1.33141768e+00 -7.50751317e-01 -2.86350697e-01
-6.77670658e-01 1.15591800e+00 5.80452800e-01 4.13050354e-01
-8.74066234e-01 -1.39363244e-01 1.03618240e+00 -1.62228242e-01
6.13791823e-01 -4.48512584e-01 -6.53543830e-01 -8.82154465e-01
3.42712142e-02 -4.79805529e-01 2.99793065e-01 -6.88567162e-01
-9.50377166e-01 4.79477048e-01 -4.49002488e-03 -1.05726171e+00
-9.24785342e-03 -8.33131254e-01 -3.81145746e-01 1.00480247e+00
-2.17069459e+00 -1.26538801e+00 -6.87542558e-01 8.94757807e-01
8.11245143e-01 1.93593800e-02 4.55835044e-01 3.04117262e-01
-5.93937695e-01 3.99205893e-01 -3.75965118e-01 -4.17983621e-01
6.18772388e-01 -1.09606874e+00 4.05938238e-01 6.89261615e-01
7.20585585e-02 -1.19741321e-01 1.51513666e-01 -4.45161819e-01
-1.16851938e+00 -1.25819647e+00 5.55604696e-01 -2.15646192e-01
2.78794676e-01 -4.99436826e-01 -6.49810493e-01 4.32342410e-01
-3.35697621e-01 1.64324846e-02 3.29282045e-01 2.82020234e-02
-4.62897360e-01 -2.79221267e-01 -9.01186109e-01 4.09403384e-01
1.40328765e+00 -4.95136708e-01 -3.14480186e-01 4.01131749e-01
9.11650419e-01 -5.34702837e-01 -4.03053671e-01 2.60394603e-01
5.11500180e-01 -1.13920224e+00 1.22151458e+00 -1.94276392e-01
4.73688513e-01 -3.60330582e-01 -4.53001440e-01 -1.06530750e+00
-2.14134648e-01 -2.41001368e-01 -5.51372282e-02 8.34410429e-01
4.74848717e-01 -6.63774788e-01 6.93575799e-01 2.68543273e-01
-5.65330863e-01 -8.08313310e-01 -1.03047085e+00 -5.90894997e-01
-2.43757665e-01 -8.04719269e-01 4.71445471e-01 4.27221507e-01
-7.01762319e-01 4.16772485e-01 -1.65237680e-01 4.83121753e-01
6.91038668e-01 5.50362051e-01 9.85385299e-01 -1.25729012e+00
-4.09185290e-01 -3.53582382e-01 -2.50115007e-01 -1.86284745e+00
1.27567738e-01 -9.55042779e-01 1.60562143e-01 -1.62309277e+00
4.83680442e-02 -8.00274968e-01 -2.41598994e-01 7.13634789e-01
-2.57283360e-01 1.89412728e-01 3.55332702e-01 4.98332791e-02
-7.18269706e-01 4.57262039e-01 1.63578808e+00 -1.65354908e-01
-4.58675295e-01 1.60391003e-01 -6.15885973e-01 6.53325737e-01
6.73226178e-01 -2.12442666e-01 -7.21799135e-01 -7.95415521e-01
5.85989244e-02 3.84168699e-03 6.88215196e-01 -1.06939638e+00
3.12444627e-01 -2.46389940e-01 4.40411508e-01 -8.06640863e-01
7.26501882e-01 -6.85681343e-01 -3.96775454e-01 2.59890407e-01
3.69964354e-02 -2.84958780e-01 3.67626816e-01 5.57221949e-01
-9.92331728e-02 -7.51491114e-02 7.47173309e-01 -2.21229032e-01
-1.40568125e+00 6.60525978e-01 -1.40959293e-01 3.52852829e-02
1.15977979e+00 -6.43988252e-01 -4.15173292e-01 6.00535385e-02
-5.67400932e-01 6.83167040e-01 2.18972638e-01 5.58625519e-01
8.24736297e-01 -8.22639942e-01 -5.30916750e-01 4.00105715e-01
3.57150972e-01 6.04028106e-01 5.45975268e-01 8.17080319e-01
2.46170387e-02 3.68406981e-01 -4.33244884e-01 -1.09797549e+00
-1.10341573e+00 3.69657874e-01 2.89930731e-01 -1.73018023e-01
-5.78323841e-01 1.18183446e+00 9.40358937e-01 -2.77328104e-01
1.89713240e-01 -5.51640034e-01 -1.79305136e-01 3.02090079e-01
3.97295266e-01 1.37777895e-01 1.77673638e-01 -3.89287740e-01
-3.53026211e-01 8.82463694e-01 -1.81290627e-01 5.60461320e-02
1.15990341e+00 -5.23311913e-01 2.56484479e-01 4.50426102e-01
1.14500415e+00 -4.99889672e-01 -1.83318019e+00 -2.83035576e-01
-3.13443899e-01 -4.56046253e-01 3.24827522e-01 -6.81593239e-01
-1.16897416e+00 1.12066841e+00 5.70984721e-01 -1.35918647e-01
1.19314253e+00 1.94251880e-01 7.45659649e-01 4.53411341e-01
4.59702700e-01 -9.93410110e-01 1.20302975e-01 6.87161505e-01
6.00398004e-01 -1.53081059e+00 -3.19062620e-01 -7.00172365e-01
-8.17904711e-01 7.65002072e-01 9.92485821e-01 1.19651318e-01
5.35989106e-01 2.47312903e-01 -1.43230206e-03 -2.53862709e-01
-8.95407081e-01 -6.75019920e-01 4.11330044e-01 9.15413499e-01
-1.08611979e-01 4.42601405e-02 2.57135898e-01 4.33409333e-01
5.46481684e-02 -2.66929984e-01 -5.83624514e-03 7.06428826e-01
-8.18347037e-01 -8.78642201e-01 -1.71886742e-01 6.30671680e-01
1.42972201e-01 -1.27705008e-01 3.63081954e-02 5.89457035e-01
4.49439943e-01 9.05068874e-01 3.17223936e-01 -4.08146590e-01
2.88792431e-01 -3.44443880e-02 5.44739366e-01 -8.14979434e-01
-5.34640551e-01 -6.92175627e-02 -3.55905704e-02 -9.19687033e-01
-4.18453604e-01 -5.57710648e-01 -1.31854498e+00 1.18765086e-02
6.84941262e-02 -5.08857846e-01 7.34014630e-01 1.11698627e+00
5.16902030e-01 6.35915875e-01 7.10814059e-01 -1.22047746e+00
-2.72708505e-01 -5.79233766e-01 -5.31139910e-01 1.56794786e-01
4.66948122e-01 -1.09436595e+00 -2.43444756e-01 -1.90286055e-01] | [8.320440292358398, -2.0204203128814697] |
0ea88c03-6084-4920-bc52-be175a44580c | action-keypoint-network-for-efficient-video | 2201.06304 | null | https://arxiv.org/abs/2201.06304v1 | https://arxiv.org/pdf/2201.06304v1.pdf | Action Keypoint Network for Efficient Video Recognition | Reducing redundancy is crucial for improving the efficiency of video recognition models. An effective approach is to select informative content from the holistic video, yielding a popular family of dynamic video recognition methods. However, existing dynamic methods focus on either temporal or spatial selection independently while neglecting a reality that the redundancies are usually spatial and temporal, simultaneously. Moreover, their selected content is usually cropped with fixed shapes, while the realistic distribution of informative content can be much more diverse. With these two insights, this paper proposes to integrate temporal and spatial selection into an Action Keypoint Network (AK-Net). From different frames and positions, AK-Net selects some informative points scattered in arbitrary-shaped regions as a set of action keypoints and then transforms the video recognition into point cloud classification. AK-Net has two steps, i.e., the keypoint selection and the point cloud classification. First, it inputs the video into a baseline network and outputs a feature map from an intermediate layer. We view each pixel on this feature map as a spatial-temporal point and select some informative keypoints using self-attention. Second, AK-Net devises a ranking criterion to arrange the keypoints into an ordered 1D sequence. Consequentially, AK-Net brings two-fold benefits for efficiency: The keypoint selection step collects informative content within arbitrary shapes and increases the efficiency for modeling spatial-temporal dependencies, while the point cloud classification step further reduces the computational cost by compacting the convolutional kernels. Experimental results show that AK-Net can consistently improve the efficiency and performance of baseline methods on several video recognition benchmarks. | ['Yi Yang', 'Yifan Sun', 'Xiaohan Wang', 'Yahong Han', 'Xu Chen'] | 2022-01-17 | null | null | null | null | ['point-cloud-classification'] | ['computer-vision'] | [ 1.00767121e-01 -6.89110518e-01 -4.66837585e-01 -1.92062303e-01
-6.28183544e-01 -3.73581797e-01 5.20294964e-01 -2.08993495e-01
-3.63690436e-01 2.39412993e-01 2.38047034e-01 1.75163761e-01
-2.27435067e-01 -7.54884124e-01 -8.69061470e-01 -9.83148813e-01
4.09111306e-02 8.92653912e-02 6.65483415e-01 1.04355603e-01
3.55557412e-01 8.22362542e-01 -1.70839882e+00 3.31218004e-01
7.36983478e-01 1.36719167e+00 4.89869803e-01 4.30204779e-01
-1.57470897e-01 8.39967549e-01 -3.04012716e-01 -1.33025110e-01
4.25622433e-01 -3.29528898e-01 -5.44738233e-01 2.03720242e-01
5.69945633e-01 -4.80446041e-01 -7.01960921e-01 1.05594850e+00
2.22898886e-01 4.95159656e-01 3.01437527e-01 -1.31319261e+00
-5.04131138e-01 4.33790356e-01 -5.25304198e-01 2.77993798e-01
2.97794163e-01 2.43905738e-01 9.73095953e-01 -1.17533708e+00
6.08222783e-01 1.03377855e+00 5.04151642e-01 2.94153988e-01
-9.82581258e-01 -6.36586308e-01 5.85551620e-01 6.41139150e-01
-1.56503820e+00 -5.48658311e-01 1.10584068e+00 -3.80233943e-01
7.68172026e-01 4.53553289e-01 9.97387469e-01 9.34925377e-01
-1.45414755e-01 1.17789268e+00 4.32228059e-01 -8.72190818e-02
3.16863090e-01 -4.42749113e-01 8.51400718e-02 5.31544149e-01
-6.13098312e-03 -6.08396195e-02 -7.03433692e-01 -8.73558819e-02
1.07884693e+00 8.17625463e-01 -6.04526401e-01 -7.85695374e-01
-1.49921000e+00 5.10562837e-01 4.99739707e-01 3.80022377e-01
-6.74476564e-01 2.69947052e-01 4.34536010e-01 2.51065940e-01
1.76263601e-01 1.13403395e-01 -3.21029007e-01 -1.83501124e-01
-1.09084904e+00 8.03375840e-02 3.36860627e-01 9.49175477e-01
8.75719070e-01 -7.94209540e-02 -3.61362636e-01 8.17505658e-01
1.38160363e-01 4.64631081e-01 4.48315680e-01 -9.70256209e-01
6.23415530e-01 8.36554289e-01 1.03611216e-01 -1.49762714e+00
-1.16353288e-01 -1.59066454e-01 -8.47094834e-01 4.10179384e-02
8.59738886e-02 2.26668835e-01 -9.54474032e-01 1.45327270e+00
2.49640003e-01 4.62845474e-01 -2.60785967e-01 1.25140858e+00
6.52163088e-01 9.26168323e-01 -5.62190190e-02 -3.74468833e-01
1.08658326e+00 -1.17501879e+00 -3.34724635e-01 7.74180144e-02
2.55565852e-01 -5.13245285e-01 8.55009496e-01 2.85159051e-01
-1.02826762e+00 -6.45565212e-01 -8.34263742e-01 -3.21933031e-02
-1.28089175e-01 2.70773977e-01 4.66217428e-01 -2.94249915e-02
-1.08138168e+00 5.06158054e-01 -9.18126941e-01 -2.21108541e-01
3.86335284e-01 3.56265306e-01 -4.94591385e-01 -3.01547199e-01
-8.96258175e-01 3.11891168e-01 4.55761850e-01 5.77458367e-02
-6.96840405e-01 -6.40642166e-01 -8.37889314e-01 2.31038705e-01
5.99710286e-01 -5.54323554e-01 8.61652374e-01 -1.27136242e+00
-1.44482231e+00 4.51518536e-01 -3.67427349e-01 -1.94809958e-01
5.28350890e-01 -2.60109484e-01 -4.07783389e-01 5.17808259e-01
1.32473111e-01 6.36533916e-01 1.04798615e+00 -1.07033372e+00
-9.96279478e-01 -1.68151826e-01 6.39265999e-02 3.87897164e-01
-2.94526130e-01 -9.99474432e-03 -1.29189634e+00 -9.21990454e-01
4.65672344e-01 -8.58664036e-01 -1.34832978e-01 1.82371601e-01
-1.91969022e-01 -2.85773218e-01 1.08489525e+00 -4.15542185e-01
1.27119243e+00 -2.32566690e+00 4.15094018e-01 2.55626529e-01
1.49461612e-01 2.30015785e-01 -2.36842394e-01 2.85193384e-01
-2.53504425e-01 -1.53040901e-01 1.16124740e-02 -4.62298132e-02
-2.40187600e-01 6.70768768e-02 -4.52758878e-01 3.86000365e-01
1.72886670e-01 1.02007866e+00 -1.05486107e+00 -3.63502324e-01
6.14341676e-01 5.46124756e-01 -6.40920937e-01 3.72773111e-02
-1.47664681e-01 1.62920788e-01 -6.69924974e-01 7.54501522e-01
6.26248419e-01 -2.92487234e-01 -9.33707952e-02 -4.95624930e-01
-3.20791841e-01 2.46880464e-02 -1.25845230e+00 1.69733560e+00
-5.90783991e-02 5.87807357e-01 -2.17413411e-01 -1.07302284e+00
7.95425296e-01 4.86897491e-02 1.03734875e+00 -6.59719348e-01
-8.38259384e-02 1.38224930e-01 -3.03957522e-01 -4.85399127e-01
5.55535555e-01 3.60140979e-01 7.27206022e-02 1.06661826e-01
-4.24029306e-03 4.33475047e-01 8.25437531e-02 -8.38619284e-03
1.01425576e+00 2.11649343e-01 1.81491986e-01 -2.40564775e-02
6.16779327e-01 -7.35665783e-02 8.56402338e-01 5.91476560e-01
-1.88238174e-01 8.95210922e-01 3.79458249e-01 -8.22754681e-01
-8.38321924e-01 -8.51198912e-01 1.90461665e-01 1.09121156e+00
6.91124022e-01 -5.20125449e-01 -3.45936954e-01 -7.78918743e-01
-2.91056167e-02 2.99881190e-01 -6.01983726e-01 -2.59357661e-01
-8.94239485e-01 -2.34483555e-01 -8.29212426e-04 6.14559770e-01
6.17084920e-01 -1.15142071e+00 -7.21793175e-01 1.22884236e-01
-2.92899728e-01 -8.28943610e-01 -8.37987542e-01 8.99826586e-02
-7.59598792e-01 -1.00849235e+00 -7.66066432e-01 -7.68003106e-01
6.90984845e-01 8.93589854e-01 8.55832338e-01 1.84860647e-01
6.72301799e-02 4.71405685e-01 -7.01267004e-01 1.99367329e-01
3.14526975e-01 -2.21845377e-02 7.47232288e-02 4.57918614e-01
5.34134448e-01 -6.08600020e-01 -9.66949582e-01 5.31161666e-01
-9.00335073e-01 1.53526068e-01 6.46358192e-01 8.69793594e-01
9.04536188e-01 8.14394429e-02 1.61936373e-01 -2.25758031e-01
-4.46393229e-02 -4.11568940e-01 -4.43399578e-01 2.82710910e-01
-3.63479368e-02 -1.06817193e-01 7.59355128e-01 -5.27081192e-01
-5.89221597e-01 3.65278721e-01 9.11063403e-02 -1.11237824e+00
2.66678222e-02 4.31615055e-01 -4.23394591e-01 -2.16336459e-01
8.67398009e-02 7.07015693e-01 -1.17631495e-01 -3.40771168e-01
1.78663060e-01 3.08948874e-01 4.09637749e-01 -4.60228145e-01
7.80477226e-01 4.97164994e-01 -2.65135556e-01 -9.21048284e-01
-4.94253725e-01 -6.76808476e-01 -8.23211670e-01 -3.63171130e-01
7.07281172e-01 -9.36042249e-01 -5.63327253e-01 4.58923429e-01
-9.12113369e-01 -1.46003008e-01 -3.06624651e-01 4.54577476e-01
-4.45226341e-01 4.03676033e-01 -4.27349091e-01 -4.57682312e-01
-1.57113791e-01 -1.32994211e+00 1.32932329e+00 3.29781085e-01
2.90602986e-02 -5.31661212e-01 -2.65417099e-01 -4.31497544e-02
9.53260511e-02 -7.17884600e-02 5.29974282e-01 -4.63583022e-01
-1.22968066e+00 -2.42230177e-01 -2.87716210e-01 8.34485888e-03
1.56676173e-01 2.32879400e-01 -6.25538051e-01 -2.51195192e-01
-1.62119910e-01 -6.18567802e-02 1.09898150e+00 4.75377321e-01
1.45593214e+00 -3.30041647e-01 -4.87376302e-01 9.32679653e-01
1.40495098e+00 4.86642927e-01 6.18832767e-01 4.71033335e-01
1.03351307e+00 2.40528211e-01 7.72243083e-01 3.62864345e-01
2.38279596e-01 9.16051447e-01 4.80937183e-01 1.50491059e-01
5.83870709e-02 -2.86028326e-01 5.19223928e-01 8.53364050e-01
-1.77190825e-01 6.28938675e-02 -7.21729577e-01 4.60349083e-01
-2.20930052e+00 -1.43240166e+00 3.78824472e-01 2.23315859e+00
5.81125736e-01 -1.09322138e-01 1.79416001e-01 7.77683547e-03
8.42475832e-01 5.03441215e-01 -7.37318456e-01 3.81484509e-01
-2.25468919e-01 -3.52537066e-01 4.49743032e-01 1.44976735e-01
-1.33607829e+00 8.98332059e-01 5.39730120e+00 9.69642162e-01
-1.47097409e+00 -2.17709854e-01 6.27779782e-01 -4.52113777e-01
-1.08490027e-01 -1.57983735e-01 -7.65080690e-01 7.01304913e-01
2.87442833e-01 1.11442842e-02 3.51065844e-01 1.05593657e+00
1.80719778e-01 8.17183703e-02 -1.06851912e+00 1.47616136e+00
1.82305992e-01 -1.65767026e+00 3.75728875e-01 -1.02011971e-01
6.37220562e-01 1.48502156e-01 -4.88226227e-02 2.21633270e-01
3.95340919e-02 -7.31517315e-01 1.01402283e+00 8.09862077e-01
7.68887460e-01 -8.46868277e-01 4.54749763e-01 2.33086839e-01
-1.53226507e+00 -2.73791462e-01 -5.09476304e-01 2.35584915e-01
2.08289195e-02 3.71867210e-01 3.72378156e-02 5.69104373e-01
9.96229589e-01 1.14022195e+00 -4.46632117e-01 1.35187256e+00
1.73440576e-01 2.93803930e-01 -3.83028448e-01 9.20661092e-02
4.01088387e-01 -4.97789413e-01 6.27094686e-01 1.07950735e+00
5.30724227e-01 3.08494329e-01 5.53660512e-01 4.83598918e-01
5.73658012e-02 -4.04730402e-02 -5.51979423e-01 1.56607136e-01
7.18726218e-01 1.16761649e+00 -9.32892621e-01 -3.81731629e-01
-5.42277277e-01 1.18810296e+00 2.13502377e-01 5.15324056e-01
-8.22319090e-01 -3.58689219e-01 8.97996843e-01 4.38805707e-02
8.30783248e-01 -2.12651804e-01 7.94521049e-02 -1.42066717e+00
3.66776317e-01 -8.72828901e-01 4.04149592e-01 -7.32173860e-01
-9.30095494e-01 5.83240628e-01 -1.02657013e-01 -1.73373628e+00
-3.89664248e-02 -5.16859055e-01 -5.12360394e-01 5.64741790e-01
-1.33344805e+00 -1.10011971e+00 -5.60834825e-01 7.32219160e-01
9.39597130e-01 -1.64139226e-01 2.84846723e-01 3.72598469e-01
-6.84218764e-01 4.95088547e-01 1.66942462e-01 3.30184519e-01
4.56461936e-01 -9.17527318e-01 3.36405694e-01 1.03771770e+00
2.43273526e-01 6.86856270e-01 1.54612988e-01 -5.69169164e-01
-1.69271874e+00 -1.35635519e+00 5.14121234e-01 -3.41980785e-01
5.87078810e-01 -2.29447618e-01 -1.00676346e+00 4.89326745e-01
-1.96930051e-01 4.32453394e-01 3.40509623e-01 -3.36777329e-01
-3.66549551e-01 -4.45144713e-01 -5.97284317e-01 9.00137305e-01
1.32079482e+00 -5.47158182e-01 -3.95419210e-01 1.73239097e-01
7.66833842e-01 -3.56308013e-01 -6.55220509e-01 4.15481240e-01
5.11975050e-01 -1.05967760e+00 1.20760632e+00 -3.45550030e-01
4.66353983e-01 -7.16894507e-01 -2.88189322e-01 -1.15339875e+00
-7.37271786e-01 -5.25145650e-01 -3.56210619e-01 1.10049295e+00
-9.49323699e-02 -3.18397343e-01 8.73448730e-01 4.64613318e-01
-1.46955654e-01 -9.86737430e-01 -9.47344065e-01 -6.63717806e-01
-4.87028033e-01 -4.89085615e-01 6.97062612e-01 9.19028282e-01
-2.44024932e-01 3.17584574e-02 -4.42171544e-01 1.91066206e-01
3.32188666e-01 5.23967326e-01 6.98893726e-01 -1.13978195e+00
1.66297629e-02 -7.23193765e-01 -6.84015572e-01 -1.59190834e+00
-9.36508179e-03 -6.35297060e-01 1.06039889e-01 -1.23547292e+00
2.16927424e-01 -5.43850899e-01 -5.99904060e-01 4.88078058e-01
-3.20657134e-01 1.25530735e-01 3.86766791e-01 7.11330473e-01
-8.98946404e-01 6.95516884e-01 1.02683151e+00 -1.91358224e-01
-4.15010303e-01 -5.27323373e-02 -4.13949221e-01 7.21261382e-01
4.62665468e-01 -6.52789325e-02 -5.57862103e-01 -5.03680766e-01
-2.91280299e-02 1.31054401e-01 5.32932580e-01 -1.21340883e+00
4.31591779e-01 -5.33631206e-01 6.77767992e-01 -9.10348833e-01
2.88542777e-01 -1.15785038e+00 2.91305751e-01 1.64674163e-01
-1.10812135e-01 1.21954665e-01 -1.75427478e-02 7.11827159e-01
-4.01144981e-01 9.69081651e-03 6.03555620e-01 -9.17877629e-02
-1.30080509e+00 8.87286663e-01 1.82857197e-02 -3.45009685e-01
1.20614636e+00 -7.36431718e-01 -5.98496124e-02 -8.72360095e-02
-4.61735785e-01 1.61702260e-01 8.47242594e-01 7.36657977e-01
9.12626386e-01 -1.65563262e+00 -2.70139486e-01 3.84896755e-01
1.44780800e-01 2.15429589e-01 4.39549863e-01 9.88031447e-01
-5.66094756e-01 4.18603182e-01 -1.72254965e-01 -9.95825052e-01
-1.23651505e+00 6.66805923e-01 2.97370613e-01 1.88222423e-01
-1.05204499e+00 7.78386414e-01 3.81358445e-01 8.70373249e-02
3.96095395e-01 -3.12028438e-01 -3.72253984e-01 2.45105371e-01
7.72191644e-01 1.43279269e-01 -1.34310186e-01 -9.15631473e-01
-4.62581187e-01 9.66993213e-01 -6.34912997e-02 2.51829028e-01
1.47643971e+00 -1.25748545e-01 -1.79466054e-01 5.02755582e-01
1.30692387e+00 -5.35910167e-02 -1.72217369e+00 -4.16239887e-01
-1.73474085e-02 -8.57664287e-01 -2.92856414e-02 -2.46530354e-01
-1.42270827e+00 4.66254801e-01 2.35656291e-01 7.51539767e-02
1.43441761e+00 -1.51417151e-01 7.22462773e-01 3.88761014e-01
4.50842738e-01 -9.59263504e-01 1.31021097e-01 6.25729561e-01
9.91936922e-01 -1.12341762e+00 -8.19533020e-02 -4.02002662e-01
-7.41251826e-01 1.20061851e+00 6.67086303e-01 -1.63927421e-01
5.23073196e-01 -7.89139569e-02 -8.61155540e-02 -1.42363176e-01
-7.03386426e-01 -8.39513764e-02 5.98654807e-01 4.44752663e-01
6.00073580e-03 -1.20002463e-01 6.40852675e-02 5.13775349e-01
1.72113523e-01 -1.95438135e-03 6.87709972e-02 7.89346814e-01
-5.09396553e-01 -6.60271585e-01 -3.09248120e-01 5.31234264e-01
-1.28126889e-01 -3.26596689e-03 -2.15858489e-01 6.23364687e-01
1.44226789e-01 4.23951328e-01 2.80194312e-01 -6.63363695e-01
2.96252877e-01 -9.70340893e-02 5.76459877e-02 -3.91381472e-01
-2.80432224e-01 2.06996083e-01 -4.84821528e-01 -9.36643779e-01
-4.62416977e-01 -8.65289152e-01 -1.13113618e+00 -1.16313368e-01
-2.48832896e-01 1.58233315e-01 1.78974301e-01 8.15667570e-01
6.47122383e-01 4.51178014e-01 7.95084178e-01 -1.18303311e+00
-2.09931701e-01 -5.20483196e-01 -2.93220907e-01 4.96487230e-01
5.32488644e-01 -7.98768044e-01 -1.46236748e-01 2.23097578e-01] | [9.03457260131836, 0.3352111876010895] |
d3a7bd25-6e8f-4ada-9b9d-05e6cc34b164 | about-explicit-variance-minimization-training | 2105.14117 | null | https://arxiv.org/abs/2105.14117v4 | https://arxiv.org/pdf/2105.14117v4.pdf | About Explicit Variance Minimization: Training Neural Networks for Medical Imaging With Limited Data Annotations | Self-supervised learning methods for computer vision have demonstrated the effectiveness of pre-training feature representations, resulting in well-generalizing Deep Neural Networks, even if the annotated data are limited. However, representation learning techniques require a significant amount of time for model training, with most of the time spent on precise hyper-parameter optimization and selection of augmentation techniques. We hypothesized that if the annotated dataset has enough morphological diversity to capture the diversity of the general population, as is common in medical imaging due to conserved similarities of tissue morphology, the variance error of the trained model is the dominant component of the Bias-Variance Trade-off. Therefore, we proposed the Variance Aware Training (VAT) method that exploits this data property by introducing the variance error into the model loss function, thereby, explicitly regularizing the model. Additionally, we provided a theoretical formulation and proof of the proposed method to aid interpreting the approach. Our method requires selecting only one hyper-parameter and matching or improving the performance of state-of-the-art self-supervised methods while achieving an order of magnitude reduction in the GPU training time. We validated VAT on three medical imaging datasets from diverse domains and for various learning objectives. These included a Magnetic Resonance Imaging (MRI) dataset for the heart semantic segmentation (MICCAI 2017 ACDC challenge), fundus photography dataset for ordinary regression of diabetic retinopathy progression (Kaggle 2019 APTOS Blindness Detection challenge), and classification of histopathologic scans of lymph node sections (PatchCamelyon dataset). Our code is available at https://github.com/DmitriiShubin/Variance-Aware-Training. | ['Sebastian D. Goodfellow', 'Danny Eytan', 'Dmitrii Shubin'] | 2021-05-28 | null | null | null | null | ['small-data'] | ['computer-vision'] | [ 4.99577105e-01 2.10341513e-01 -2.78530866e-01 -4.94209617e-01
-7.59067774e-01 -3.30498576e-01 3.02099645e-01 3.00003141e-01
-6.09642684e-01 6.13676250e-01 -1.44061923e-01 -3.74732822e-01
-2.64149547e-01 -5.60706198e-01 -7.99100161e-01 -9.84331489e-01
-1.39158309e-01 6.25596344e-01 1.48470374e-02 2.07438126e-01
5.62578030e-02 5.10134459e-01 -1.37503350e+00 6.43373951e-02
1.09000158e+00 1.20875835e+00 2.22028121e-01 6.73601627e-01
8.45627040e-02 6.64970994e-01 -1.66957527e-01 -9.72092673e-02
4.38474327e-01 -3.91100079e-01 -8.12314153e-01 2.43047014e-01
9.22926784e-01 -1.46458700e-01 -1.96975499e-01 9.97202694e-01
8.20205688e-01 2.46448461e-02 8.01535904e-01 -9.44312155e-01
-2.37295315e-01 3.02047670e-01 -5.19421160e-01 4.64981675e-01
-4.95413512e-01 1.93265036e-01 6.59190536e-01 -3.90220404e-01
6.79779470e-01 5.72136700e-01 7.07008481e-01 6.72207475e-01
-1.27572322e+00 -5.26386440e-01 -2.24269152e-01 1.44939691e-01
-1.22912669e+00 -2.42425933e-01 6.08423710e-01 -6.79325998e-01
7.18363762e-01 2.28432074e-01 8.35759997e-01 9.89098072e-01
1.31637707e-01 6.03433371e-01 1.31213844e+00 -4.38263446e-01
3.05403501e-01 1.45566747e-01 4.15744722e-01 1.05009413e+00
4.39477980e-01 1.43229544e-01 -1.14713930e-01 -3.62922579e-01
9.63407755e-01 -8.00290927e-02 -3.82971883e-01 -5.75683117e-01
-9.39342499e-01 7.68345475e-01 4.44670349e-01 8.98186117e-02
-3.18994224e-01 2.45714381e-01 6.31801963e-01 -5.14571257e-02
4.67854291e-01 3.44444096e-01 -5.07067382e-01 8.20577666e-02
-8.93961489e-01 -1.00340046e-01 5.25890887e-01 5.05517244e-01
6.42035842e-01 -1.04354672e-01 1.33503815e-02 8.88367951e-01
1.01606280e-01 2.48823792e-01 7.59540617e-01 -1.02526367e+00
-8.17929022e-03 7.96762586e-01 -3.53194207e-01 -6.48916662e-01
-7.66472161e-01 -9.20423508e-01 -1.22709203e+00 3.10992658e-01
6.79334700e-01 -2.48788804e-01 -1.15765190e+00 1.67442334e+00
5.29572845e-01 1.78790867e-01 -1.04200497e-01 9.61076677e-01
7.20490575e-01 1.68093726e-01 8.37677270e-02 -2.38859117e-01
1.38723075e+00 -7.17133820e-01 -1.84578255e-01 -1.72333851e-01
9.43673670e-01 -4.33533549e-01 1.07116663e+00 3.60759288e-01
-8.64636421e-01 -2.97110319e-01 -9.60174382e-01 -2.36822795e-02
6.64437655e-03 3.24080586e-01 1.02317655e+00 6.62320077e-01
-9.27582324e-01 6.59069300e-01 -1.13209784e+00 -4.67127919e-01
1.00941527e+00 4.90279227e-01 -2.73962051e-01 1.55020459e-02
-6.69426918e-01 6.65174186e-01 1.80527985e-01 -5.72078377e-02
-7.15917110e-01 -1.27582896e+00 -5.84692419e-01 -9.86827016e-02
1.38409540e-01 -1.09504092e+00 8.42575371e-01 -1.31343722e+00
-1.29765558e+00 1.13299215e+00 3.16286013e-02 -7.60879219e-01
7.33228326e-01 -9.90041420e-02 1.06034204e-01 4.72946525e-01
-1.31273821e-01 8.01026106e-01 8.28636825e-01 -1.05385792e+00
-2.45261550e-01 -5.76048315e-01 -1.96891576e-01 2.63567194e-02
-2.80612290e-01 -3.25660020e-01 -3.06724787e-01 -5.97295702e-01
5.75513430e-02 -1.13099086e+00 -5.10880411e-01 3.61613512e-01
-2.55170524e-01 1.49590388e-01 4.03678656e-01 -6.13885581e-01
8.32518995e-01 -2.15502572e+00 -1.11375935e-01 3.28204721e-01
3.94823045e-01 4.49223042e-01 -9.43708941e-02 -2.19243854e-01
-2.71936417e-01 3.63073796e-02 -5.47236741e-01 -1.92970932e-01
-5.63727200e-01 2.05896482e-01 -1.22914873e-02 8.08890700e-01
1.85876906e-01 7.18477607e-01 -6.34529054e-01 -6.79319501e-01
1.24375373e-01 6.94150984e-01 -6.38571918e-01 -2.77730282e-02
-1.24788418e-01 7.51774132e-01 -3.93932313e-01 5.35780370e-01
5.02640367e-01 -6.87899470e-01 2.64214903e-01 -4.97540236e-01
2.69372582e-01 9.41987783e-02 -9.20686305e-01 1.73602641e+00
-5.20602882e-01 5.32108247e-01 -1.52830571e-01 -1.23694289e+00
6.42039657e-01 -5.33771142e-03 8.39704216e-01 -7.22261906e-01
4.33703542e-01 2.59799868e-01 1.97266236e-01 -4.58480746e-01
-2.54043311e-01 -2.14982033e-02 5.70306957e-01 2.71606207e-01
1.22452088e-01 1.58662960e-01 1.62578091e-01 1.98027454e-02
1.01982093e+00 -2.70775836e-02 2.70955890e-01 -4.66007680e-01
2.81775892e-01 2.61113644e-01 5.76072216e-01 6.80835724e-01
-2.76189744e-01 7.80800343e-01 5.19771099e-01 -6.50364041e-01
-1.01244450e+00 -7.69922674e-01 -6.19598091e-01 5.91971278e-01
-1.20176487e-01 -1.28103331e-01 -8.98957968e-01 -7.34331191e-01
-9.41428691e-02 3.43712091e-01 -7.94360340e-01 -1.71340525e-01
-6.00848377e-01 -1.27205431e+00 6.05053365e-01 5.94186008e-01
4.79387939e-01 -6.00745082e-01 -1.09447801e+00 -5.30292094e-02
2.52104372e-01 -9.93344784e-01 -2.75602281e-01 3.51227432e-01
-1.33119071e+00 -1.32549322e+00 -7.69789279e-01 -6.90765083e-01
1.06403005e+00 -1.46046743e-01 1.02365792e+00 3.20881069e-01
-9.68627810e-01 3.63166571e-01 1.71078369e-02 -3.62486660e-01
-2.53978938e-01 3.85235846e-02 -6.81058094e-02 -3.91293615e-02
-9.08544436e-02 -6.50260389e-01 -9.20284808e-01 1.67742372e-01
-7.56222486e-01 2.57596076e-01 7.37992883e-01 1.22526038e+00
9.51463163e-01 -2.49680459e-01 3.66867006e-01 -1.10688937e+00
3.28421406e-02 -2.10222512e-01 -7.06469834e-01 1.69973239e-01
-8.12888980e-01 -2.50594374e-02 5.26956439e-01 -6.46403491e-01
-7.06941962e-01 3.26755077e-01 4.39481325e-02 -4.58500832e-01
-6.78145587e-02 3.97854418e-01 2.06028014e-01 -2.83302218e-01
9.11165237e-01 1.66779637e-01 4.76878732e-01 -2.29479536e-01
1.33079305e-01 3.18904132e-01 4.25266832e-01 -6.32405460e-01
3.10598582e-01 7.71011233e-01 4.35312599e-01 -8.88515294e-01
-7.59130061e-01 -4.06058401e-01 -4.92936552e-01 -4.24673185e-02
6.62692845e-01 -7.18804359e-01 -6.85805380e-01 5.39181828e-01
-8.06224883e-01 -5.87455451e-01 -5.25260985e-01 6.55927896e-01
-6.48231447e-01 3.66703033e-01 -3.61440033e-01 -5.32782912e-01
-7.20846832e-01 -1.20401573e+00 6.93723559e-01 2.27416441e-01
-4.22192477e-02 -1.09505057e+00 6.70594126e-02 5.83149195e-01
4.53371704e-01 3.52165669e-01 1.29274762e+00 -7.10137665e-01
-3.68378699e-01 -3.29250805e-02 -4.24424231e-01 6.31529033e-01
-1.20146200e-02 -9.48360786e-02 -9.07118201e-01 -2.71910310e-01
-6.19670376e-02 -3.76353145e-01 1.03619528e+00 9.23169076e-01
1.48444605e+00 -2.14306131e-01 -2.06906170e-01 1.14437234e+00
1.53752804e+00 -7.20234290e-02 5.76268435e-01 2.81459868e-01
7.61446953e-01 4.91901428e-01 1.34830445e-01 3.67766410e-01
2.08030567e-01 5.28351545e-01 4.07844037e-01 -5.01831591e-01
-4.20715660e-01 2.11735159e-01 -1.22942679e-01 4.88394827e-01
-1.69676617e-01 1.24094315e-01 -1.09821677e+00 5.99453211e-01
-1.66657889e+00 -4.95677412e-01 -1.05572179e-01 2.33256698e+00
9.74545062e-01 3.67784570e-03 1.20310552e-01 -2.69430578e-02
4.82405752e-01 -2.05831245e-01 -9.10604060e-01 -7.59709626e-02
3.04071028e-02 3.83175522e-01 8.37851405e-01 2.20084116e-01
-1.09801054e+00 6.81647897e-01 5.41873026e+00 8.06188345e-01
-1.38346505e+00 1.41673401e-01 9.63775337e-01 -1.91688955e-01
7.29598179e-02 -8.01280141e-03 -5.79881251e-01 2.56106645e-01
9.01072621e-01 4.97089103e-02 2.86322027e-01 8.95211875e-01
5.46784773e-02 -1.82620198e-01 -1.02826023e+00 1.00170755e+00
-3.41660460e-03 -1.63354361e+00 5.91455102e-02 2.52951439e-02
6.33142114e-01 2.93450415e-01 1.55555248e-01 -5.40843308e-02
-1.57944843e-01 -1.12879491e+00 1.18355542e-01 4.33712214e-01
9.93841648e-01 -4.72607970e-01 7.16367781e-01 4.74298596e-02
-5.98103881e-01 6.53451011e-02 -3.49848449e-01 2.96234965e-01
-2.53571123e-01 8.37437332e-01 -9.74289834e-01 2.92540371e-01
6.24429166e-01 6.37662351e-01 -7.61559129e-01 1.12552655e+00
2.02657714e-01 1.00699413e+00 -4.44729090e-01 2.67118067e-01
4.44579544e-03 -2.58842140e-01 4.44604009e-01 1.07119906e+00
1.13967143e-01 -4.27232683e-02 3.40621406e-03 7.33045459e-01
-9.98346284e-02 2.71817654e-01 -1.11694910e-01 -3.78907546e-02
5.82910627e-02 1.33193088e+00 -8.08098316e-01 -2.70666659e-01
-1.70259595e-01 6.74751878e-01 2.52224177e-01 2.25212917e-01
-7.72889555e-01 -3.39879058e-02 4.43244070e-01 3.64718854e-01
1.28421888e-01 3.03036980e-02 -6.67839050e-01 -1.07182181e+00
-5.08480705e-02 -8.08096886e-01 4.40681040e-01 -3.48863810e-01
-1.29290986e+00 5.41630030e-01 -6.52659014e-02 -1.13384771e+00
-8.02861992e-03 -7.06603169e-01 -3.79635185e-01 7.27566957e-01
-1.65240967e+00 -1.13077188e+00 -5.03182113e-01 4.38929975e-01
2.97539085e-01 -3.67155284e-01 7.98846960e-01 4.29838866e-01
-6.92962646e-01 7.07163095e-01 5.79628348e-02 1.28992364e-01
6.47082031e-01 -1.12080598e+00 -5.11862710e-02 4.75400358e-01
-4.58569586e-04 7.44266570e-01 4.54381108e-01 -4.48879033e-01
-1.31638980e+00 -1.17020845e+00 3.05349946e-01 -3.00060026e-02
6.50418758e-01 4.85790335e-02 -9.41057384e-01 4.46455717e-01
-1.97432056e-01 6.54486299e-01 7.91078150e-01 -1.39968678e-01
-2.77971655e-01 -1.63985640e-01 -1.22159874e+00 4.19021249e-01
9.62464452e-01 -2.28565186e-01 -1.60768345e-01 5.72022200e-01
3.34019125e-01 -5.87396204e-01 -8.51817608e-01 5.75026751e-01
6.19757533e-01 -7.61896968e-01 9.03956413e-01 -6.45416021e-01
5.30247927e-01 -3.95740643e-02 -1.56847373e-01 -9.08730865e-01
-1.84505414e-02 -4.06716704e-01 -1.67364955e-01 8.79073262e-01
3.04093152e-01 -6.74352467e-01 9.74255681e-01 4.48109686e-01
-7.45518133e-02 -1.30522823e+00 -1.00163794e+00 -6.75530553e-01
2.16252089e-01 -2.22608462e-01 -6.03852235e-02 8.90733898e-01
-5.24151862e-01 7.98714682e-02 4.41541001e-02 2.18079254e-01
9.06195819e-01 -1.25895396e-01 6.33158267e-01 -1.32698238e+00
-5.26125491e-01 -4.62291688e-01 -6.55835807e-01 -5.82875788e-01
1.34561419e-01 -1.24037731e+00 -4.32086796e-01 -1.40396726e+00
4.35048074e-01 -6.68610454e-01 -3.01793903e-01 5.50419748e-01
-9.31584612e-02 2.88039923e-01 -3.04330643e-02 2.99777180e-01
-2.01646566e-01 2.31816784e-01 1.21927035e+00 -1.61959231e-01
-2.11970001e-01 -1.27776396e-02 -6.83935225e-01 8.21018338e-01
8.96238387e-01 -5.59190094e-01 -5.44233739e-01 -3.66415352e-01
1.86510533e-02 -1.53771132e-01 8.70132327e-01 -1.17767143e+00
1.24499545e-01 7.07382038e-02 4.93239254e-01 -3.65971327e-02
6.59060851e-02 -7.25078464e-01 4.48796013e-03 6.97040737e-01
-4.90394741e-01 -2.93306917e-01 3.82049143e-01 3.67476344e-01
6.86593875e-02 -2.03209579e-01 1.23645473e+00 1.10230818e-02
-4.36243474e-01 4.44533199e-01 -1.43255591e-01 3.15554053e-01
8.01697314e-01 -2.03398794e-01 -4.75754887e-01 1.29101813e-01
-6.02325618e-01 -2.34327912e-02 4.30992067e-01 -6.39129803e-02
4.28700715e-01 -7.58554995e-01 -7.05307364e-01 2.22209841e-01
6.59957342e-03 2.14538351e-01 5.09117603e-01 1.23035157e+00
-7.55563498e-01 6.73890859e-02 -3.97423804e-01 -8.59234095e-01
-1.18950737e+00 2.19936296e-01 6.83424473e-01 -4.10002828e-01
-9.19844091e-01 7.97372580e-01 2.13176295e-01 -2.80161202e-01
2.20797420e-01 -3.07662725e-01 -1.11635858e-02 -6.90610707e-02
3.34225684e-01 4.13247019e-01 2.82189518e-01 -2.50622094e-01
-4.68747258e-01 7.22590148e-01 -2.96956420e-01 2.93274730e-01
1.36705959e+00 1.87993258e-01 -7.61605799e-02 1.51772782e-01
1.36988950e+00 -3.62030119e-01 -1.26304579e+00 -1.94421828e-01
-1.27664685e-01 -1.53232440e-01 5.90676248e-01 -9.80029523e-01
-1.39961958e+00 8.37670267e-01 1.27379251e+00 -3.92204195e-01
1.20237744e+00 -1.21273793e-01 7.27876604e-01 2.97818810e-01
3.86294246e-01 -8.92432690e-01 -1.00873262e-01 1.02050520e-01
6.53187573e-01 -1.43533635e+00 2.94663668e-01 -5.15794694e-01
-7.47976422e-01 1.01648986e+00 5.02677143e-01 -2.82971412e-01
6.54647171e-01 2.27052107e-01 1.97306171e-01 -1.45231143e-01
-5.66387951e-01 -9.09978822e-02 4.57214087e-01 6.45410299e-01
4.23734874e-01 -6.06046133e-02 -4.20916617e-01 5.78848302e-01
-7.32147787e-03 1.71249911e-01 3.89461070e-01 7.81330585e-01
-1.05674505e-01 -9.32003081e-01 7.83452615e-02 6.82923555e-01
-4.92384642e-01 -1.76446378e-01 -6.96130544e-02 7.04144716e-01
1.15169756e-01 3.67912054e-01 1.35690793e-01 2.07326021e-02
3.35055366e-02 -3.21542807e-02 7.20193326e-01 -4.81005460e-01
-5.78165114e-01 9.03182372e-04 -9.37830359e-02 -5.69241345e-01
-4.21435237e-01 -5.14368594e-01 -1.48259342e+00 1.70460910e-01
-1.26515508e-01 -1.78824961e-01 8.77685547e-01 8.32686484e-01
5.54046750e-01 6.26808465e-01 4.66510803e-01 -7.42517352e-01
-5.90610802e-01 -7.19490349e-01 -4.48711097e-01 3.36568087e-01
2.25085646e-01 -6.55950785e-01 -3.66375357e-01 1.40523538e-01] | [14.601045608520508, -2.396942615509033] |
64ff6a0b-9c36-4463-a72c-def4bdc37617 | single-model-attribution-via-final-layer | 2306.06210 | null | https://arxiv.org/abs/2306.06210v2 | https://arxiv.org/pdf/2306.06210v2.pdf | Single-Model Attribution of Generative Models Through Final-Layer Inversion | Recent groundbreaking developments on generative modeling have sparked interest in practical single-model attribution. Such methods predict whether a sample was generated by a specific generator or not, for instance, to prove intellectual property theft. However, previous works are either limited to the closed-world setting or require undesirable changes of the generative model. We address these shortcomings by proposing FLIPAD, a new approach for single-model attribution in the open-world setting based on final-layer inversion and anomaly detection. We show that the utilized final-layer inversion can be reduced to a convex lasso optimization problem, making our approach theoretically sound and computationally efficient. The theoretical findings are accompanied by an experimental study demonstrating the effectiveness of our approach, outperforming the existing methods. | ['Asja Fischer', 'Johannes Lederer', 'Jonas Ricker', 'Mike Laszkiewicz'] | 2023-05-26 | null | null | null | null | ['anomaly-detection'] | ['methodology'] | [ 3.95560771e-01 2.64488131e-01 -4.47649598e-01 -2.35672653e-01
-7.51354933e-01 -5.24349809e-01 7.58101702e-01 -8.70293081e-02
-3.19295451e-02 8.38543832e-01 -4.97287624e-02 -3.88646781e-01
-2.05830500e-01 -6.37645543e-01 -8.71466994e-01 -6.53114498e-01
1.63798794e-01 6.56104445e-01 -5.39873242e-01 3.83365929e-01
1.88525259e-01 2.45677695e-01 -9.18306470e-01 -2.40892887e-01
1.01395845e+00 8.80412459e-01 -3.63614857e-01 3.19317162e-01
1.42414302e-01 7.33656883e-01 -4.30886775e-01 -7.15419173e-01
4.86906946e-01 -5.30861259e-01 -5.07249892e-01 1.70348242e-01
3.67727757e-01 -2.20786497e-01 -9.58489403e-02 1.21182621e+00
3.49861205e-01 -6.81066513e-02 5.11711895e-01 -1.77227879e+00
-8.41702938e-01 6.78326428e-01 -5.61182916e-01 -1.05356090e-01
1.36561990e-01 -4.67260517e-02 1.01117373e+00 -7.76120663e-01
5.49334526e-01 1.15446067e+00 7.82992184e-01 3.70382428e-01
-1.58135867e+00 -9.96514261e-01 2.92723566e-01 -9.34872031e-02
-1.39541376e+00 -5.64273119e-01 8.52576613e-01 -6.88019335e-01
3.05876493e-01 3.04113001e-01 5.01012027e-01 1.42427206e+00
-4.38796058e-02 5.84686399e-01 1.32444799e+00 -2.06844166e-01
2.66130090e-01 2.80446649e-01 7.56693212e-03 8.84038806e-01
6.90416157e-01 3.14959362e-02 -5.52118957e-01 -7.60803699e-01
6.77771747e-01 1.36235133e-01 -2.81517684e-01 -5.03225088e-01
-1.32493341e+00 1.06783712e+00 2.12274298e-01 -1.09547138e-01
-4.72982019e-01 2.78024286e-01 1.24144807e-01 2.08638862e-01
7.52930403e-01 4.45423722e-01 1.17280722e-01 8.91823694e-02
-1.28464806e+00 2.63794243e-01 8.91056538e-01 9.64908183e-01
6.04148746e-01 1.26666695e-01 -8.31201300e-02 3.47872108e-01
4.97896701e-01 6.15148723e-01 2.74343401e-01 -8.21145654e-01
3.18262041e-01 4.29811686e-01 2.25511745e-01 -1.18424964e+00
-9.12041683e-03 -8.27437341e-01 -1.00024569e+00 -1.10486507e-01
4.00155634e-01 -7.95248672e-02 -6.14172161e-01 1.81361997e+00
1.87512398e-01 6.30569816e-01 -2.16827139e-01 6.87414467e-01
1.26867771e-01 2.72708148e-01 -5.93589060e-02 -2.17496768e-01
9.56664920e-01 -1.08124816e+00 -6.95953846e-01 -1.84587926e-01
5.73307812e-01 -5.30887127e-01 8.81502032e-01 3.54080498e-01
-7.51129508e-01 1.38024781e-02 -8.42086673e-01 3.36533040e-01
4.65376116e-03 9.62522030e-02 8.14467967e-01 7.73113370e-01
-5.34379482e-01 4.37439322e-01 -6.96061611e-01 -1.11574210e-01
1.06796908e+00 1.44257531e-01 -2.51748264e-01 2.06462786e-01
-8.34699452e-01 4.00634110e-01 1.36175275e-01 1.29061893e-01
-1.08073652e+00 -8.96062851e-01 -5.92874110e-01 5.03891110e-02
6.27796948e-01 -9.31347311e-01 7.85283566e-01 -8.44937384e-01
-1.15012848e+00 9.98220861e-01 -4.36626643e-01 -7.79517174e-01
1.10542440e+00 -2.90988624e-01 -1.96082935e-01 -2.14732021e-01
3.25256288e-01 -1.34724662e-01 1.15871787e+00 -1.13434756e+00
-2.39411473e-01 -2.79537559e-01 -2.70193312e-02 -3.09193283e-01
-5.03841221e-01 -3.74492854e-02 -1.65995094e-03 -8.42070758e-01
1.01843975e-01 -1.15802622e+00 -4.11353409e-01 4.36437614e-02
-1.06426275e+00 1.35752976e-01 7.48154044e-01 -5.30555964e-01
1.30395854e+00 -1.99613512e+00 -4.28543650e-02 3.96558374e-01
4.64815229e-01 2.29125321e-01 1.96509793e-01 2.56807238e-01
-4.06108759e-02 4.99017149e-01 -4.36821342e-01 -6.88166678e-01
3.43819499e-01 -1.06849611e-01 -8.79630029e-01 8.43797863e-01
-6.50029210e-03 9.65752840e-01 -9.62648690e-01 -3.23251039e-01
-2.43595503e-02 2.21190229e-01 -6.22327328e-01 2.56484747e-01
3.89307141e-02 6.49554908e-01 -6.28020942e-01 7.29831219e-01
6.56118095e-01 -5.43436050e-01 -3.98062654e-02 2.28608355e-01
2.04509422e-01 1.51065499e-01 -1.09253466e+00 1.34386086e+00
-4.13523555e-01 5.73260605e-01 6.06662668e-02 -1.02507007e+00
8.55118394e-01 1.64793834e-01 3.56464028e-01 -2.16692224e-01
-1.92605667e-02 3.74702036e-01 -2.30748430e-01 -1.49239421e-01
2.47615620e-01 -4.73885983e-01 -1.29071280e-01 8.22706699e-01
-3.80410284e-01 1.64700046e-01 -1.36842370e-01 2.46413991e-01
8.79419625e-01 3.22131664e-01 3.13830227e-01 -1.34895980e-01
2.70472467e-01 -1.48299649e-01 7.28406787e-01 1.20201123e+00
-1.93159565e-01 6.59370720e-01 6.97547317e-01 -1.96929336e-01
-9.68085051e-01 -1.03615415e+00 -1.73749566e-01 6.44657135e-01
-7.68600926e-02 -3.39863181e-01 -7.67748713e-01 -8.91114473e-01
2.90322334e-01 7.70768344e-01 -6.15211070e-01 -1.54931977e-01
-3.91510189e-01 -1.02855146e+00 6.59685552e-01 1.89284816e-01
2.11807698e-01 -6.61918163e-01 -4.36871022e-01 1.19463623e-01
-2.20454171e-01 -1.01628077e+00 -3.61506939e-01 -2.00670093e-01
-7.97246814e-01 -1.01963508e+00 -6.14416122e-01 -2.21304312e-01
9.09202158e-01 1.65675033e-03 8.97666037e-01 7.51633011e-03
-3.82180884e-02 1.68904901e-01 9.42165107e-02 -5.10836542e-01
-4.91581529e-01 1.98990703e-01 2.47332126e-01 6.01148129e-01
4.75778803e-02 -6.47645473e-01 -3.60019535e-01 2.48932108e-01
-6.70743465e-01 7.99953863e-02 4.50869501e-01 9.00912225e-01
7.25535095e-01 -1.48657143e-01 8.57271910e-01 -1.39316213e+00
6.06507838e-01 -7.00354636e-01 -8.39618862e-01 2.93382853e-01
-9.85396624e-01 3.23226377e-02 5.61519027e-01 -3.53520334e-01
-8.83936405e-01 -1.18543625e-01 2.99238801e-01 -7.94306815e-01
2.17890486e-01 5.29071212e-01 -2.59278059e-01 -1.57936871e-01
2.32372090e-01 2.85798162e-01 1.37821110e-02 -5.98512173e-01
2.74080724e-01 4.83873367e-01 5.40517449e-01 -6.60876334e-01
1.28026044e+00 7.97876179e-01 1.88998550e-01 -4.18216616e-01
-1.15233135e+00 -2.82597784e-02 -4.10934776e-01 4.40234393e-02
3.92393202e-01 -9.95867074e-01 -6.33906841e-01 4.03412610e-01
-8.13547611e-01 -1.26430422e-01 -3.96181107e-01 4.18966025e-01
-6.11799359e-01 3.62748116e-01 -1.02759816e-01 -1.12687016e+00
-3.25926542e-01 -1.04282618e+00 8.88107121e-01 -5.27251847e-02
-3.64581674e-01 -1.18843055e+00 1.45660803e-01 5.85137784e-01
4.87251580e-01 3.74429941e-01 8.35942209e-01 -9.81647074e-01
-7.18341112e-01 -5.65929115e-01 -2.09789455e-01 2.22783312e-01
6.10728189e-03 -9.97693315e-02 -1.03753054e+00 -2.91360825e-01
1.55263543e-01 -1.14640690e-01 7.44789720e-01 2.20578596e-01
1.30628467e+00 -5.50945461e-01 -5.09204984e-01 9.38364327e-01
1.23606634e+00 -3.98653626e-01 3.94182801e-01 4.51402962e-01
9.22413886e-01 1.35124013e-01 4.46528345e-01 6.94113672e-01
2.84641832e-01 6.51777387e-01 4.25047129e-01 -1.88250214e-01
8.20541978e-02 -7.61350036e-01 2.09740445e-01 3.96797568e-01
-8.04318264e-02 -2.91595668e-01 -7.69176483e-01 5.66458344e-01
-2.02226949e+00 -1.06313896e+00 -1.14965186e-01 2.49373937e+00
4.78719503e-01 -1.98116768e-02 1.55086787e-02 -1.64879486e-01
8.13760579e-01 2.31942222e-01 -6.28830135e-01 -6.52818531e-02
-3.62935930e-01 1.52516337e-02 5.15417576e-01 2.59521991e-01
-1.17946565e+00 6.40789211e-01 7.04314852e+00 7.23298907e-01
-9.57464993e-01 4.28247243e-01 7.37241566e-01 -3.17990452e-01
-6.23991966e-01 3.87766451e-01 -5.99003196e-01 7.29123712e-01
7.39811957e-01 -6.89240277e-01 3.69833618e-01 8.98435593e-01
2.63254672e-01 2.27512941e-01 -1.14085805e+00 9.20260608e-01
2.85165995e-01 -1.25156140e+00 1.19467773e-01 5.52441418e-01
8.65938067e-01 -9.41534266e-02 2.75214285e-01 3.20151681e-03
3.91711295e-01 -1.10163653e+00 8.71814549e-01 7.68374860e-01
5.87814867e-01 -5.54244280e-01 5.81092358e-01 3.22244406e-01
-6.49495482e-01 -1.16271876e-01 -2.73393542e-01 -1.38152599e-01
4.33312774e-01 9.28725183e-01 -7.23585129e-01 6.23875260e-01
1.97245046e-01 8.53160381e-01 -3.59747112e-01 9.92039204e-01
-4.08527970e-01 1.08684468e+00 -3.87463182e-01 3.66767973e-01
1.66278660e-01 -4.10672903e-01 1.10739517e+00 8.71310711e-01
5.19954383e-01 -5.28294504e-01 1.95981950e-01 1.39357102e+00
-4.31154191e-01 -1.15063369e-01 -6.95474148e-01 -3.91211718e-01
3.83434385e-01 1.10624719e+00 -4.25091356e-01 -3.19574684e-01
-2.03605399e-01 8.99199843e-01 1.76433012e-01 3.00095439e-01
-1.11812913e+00 -1.00158267e-01 5.34840941e-01 4.01585937e-01
1.90948367e-01 1.52203247e-01 -4.34186935e-01 -1.60985804e+00
2.26439908e-01 -7.46527791e-01 3.82182956e-01 -4.84369367e-01
-1.48489761e+00 2.83963770e-01 -1.89111561e-01 -1.13766706e+00
-1.77111030e-01 -2.33920291e-01 -6.62839830e-01 8.80148053e-01
-1.60313570e+00 -1.25357783e+00 -7.03287795e-02 3.52565497e-01
7.20370784e-02 -4.25226212e-01 8.00005198e-01 2.93777406e-01
-9.94382977e-01 7.40688741e-01 3.19556803e-01 1.40722897e-02
6.70142829e-01 -1.03103602e+00 1.76828027e-01 1.22433436e+00
2.34746233e-01 8.43475342e-01 6.33154035e-01 -6.75536096e-01
-1.30191445e+00 -1.28320920e+00 1.09174824e+00 -4.47476625e-01
1.06700993e+00 -6.22084081e-01 -7.19196558e-01 1.09354436e+00
-2.25600749e-01 2.33547881e-01 8.30052078e-01 2.70246774e-01
-4.47731763e-01 7.12496564e-02 -1.25582445e+00 4.50674057e-01
1.11494911e+00 -4.79910046e-01 -4.40346122e-01 4.69841212e-01
2.79438376e-01 -1.86768487e-01 -7.75730908e-01 2.15208568e-02
5.08665383e-01 -7.24993527e-01 9.99513328e-01 -9.65703666e-01
4.50796127e-01 -2.53550619e-01 -9.89657044e-02 -1.11352789e+00
-2.17458546e-01 -1.19504631e+00 -5.05917192e-01 1.40937114e+00
3.62707704e-01 -1.10965729e+00 6.50544643e-01 6.78168952e-01
1.03824914e-01 -6.58770084e-01 -1.03264773e+00 -8.65285099e-01
2.97579169e-02 -3.51334929e-01 8.43280792e-01 1.12081349e+00
-2.86939919e-01 6.62537217e-02 -9.45378125e-01 1.09995514e-01
1.13361669e+00 6.19680703e-01 9.99728262e-01 -1.44778967e+00
-3.90815258e-01 -3.96896303e-01 -4.62043464e-01 -7.29566813e-01
5.91819584e-01 -1.11195815e+00 -2.48056442e-01 -1.09667706e+00
5.94641328e-01 -7.78482437e-01 -4.88167048e-01 5.71526647e-01
-3.97971749e-01 4.63088632e-01 3.14602882e-01 4.74130839e-01
-4.92878109e-01 5.29475272e-01 8.40566516e-01 -1.68827012e-01
1.42654121e-01 2.26391211e-01 -1.02739811e+00 9.33801949e-01
5.81097186e-01 -8.94690573e-01 -3.31236899e-01 -1.71626955e-01
4.50363070e-01 -3.14976990e-01 9.78038907e-01 -7.60209560e-01
-1.89502276e-02 -2.23284021e-01 7.99729452e-02 -2.29664400e-01
7.36167058e-02 -6.33114159e-01 4.81564969e-01 4.97651964e-01
-4.39031899e-01 -2.47655317e-01 -2.82088578e-01 8.90816033e-01
6.88897222e-02 -1.67731345e-01 6.10989928e-01 6.68759421e-02
-1.91090092e-01 4.89998072e-01 8.40739384e-02 2.97843039e-01
1.06582391e+00 4.49325703e-02 -2.58620232e-01 -3.95543218e-01
-7.05500960e-01 -1.35826264e-02 6.89034164e-01 2.71589637e-01
2.52774477e-01 -1.29716241e+00 -9.27377284e-01 2.68421024e-01
3.28307897e-02 -2.48409927e-01 -5.04125245e-02 1.20527875e+00
-2.47718543e-02 3.72025579e-01 1.45009905e-01 -4.13061380e-01
-8.50648999e-01 5.21323144e-01 2.43862823e-01 -4.45677161e-01
-8.56395304e-01 6.63135409e-01 4.07570273e-01 -3.14675242e-01
-2.62766369e-02 2.42432952e-01 3.69754046e-01 -2.62613986e-02
3.07216793e-01 4.70105827e-01 -1.75562799e-01 -8.56879056e-01
-3.94698232e-01 2.16401294e-01 9.54606235e-02 6.00230433e-02
1.41875899e+00 -1.59038723e-01 -2.52595961e-01 6.00162327e-01
1.01791930e+00 2.70832300e-01 -1.06326771e+00 -3.52412403e-01
-2.95429349e-01 -8.30964327e-01 7.70281826e-04 -6.40474081e-01
-1.13388073e+00 7.18518376e-01 1.59597874e-01 1.11239120e-01
7.97193944e-01 -9.00065303e-02 7.33004153e-01 2.58966714e-01
6.17604733e-01 -7.68732667e-01 -3.36476892e-01 -2.16723233e-02
7.09737778e-01 -1.17067909e+00 1.75901130e-01 -3.54716063e-01
-4.50587720e-01 6.47540390e-01 1.50149316e-01 -1.57767385e-01
5.31909645e-01 7.22804815e-02 -2.40807027e-01 -1.14117950e-01
-4.02924895e-01 3.62386286e-01 1.76885188e-01 4.29626524e-01
1.46771103e-01 3.70474160e-01 -1.02436326e-01 8.56966138e-01
-3.37054253e-01 2.00577080e-01 6.77922308e-01 6.62782073e-01
1.73296019e-01 -1.03086734e+00 -4.05456305e-01 7.90542841e-01
-6.50223970e-01 -9.28000510e-02 -3.88055563e-01 4.35940951e-01
3.91874425e-02 7.90337503e-01 -3.64141650e-02 8.34701583e-02
9.70248803e-02 2.42575347e-01 1.17159091e-01 -4.85210985e-01
-3.93294990e-01 -3.48594375e-02 -1.16720423e-01 -5.29169202e-01
-3.55039001e-01 -8.52900982e-01 -7.43182421e-01 -4.83161628e-01
-5.45305073e-01 1.13000847e-01 4.11373883e-01 8.99562359e-01
6.90216005e-01 3.24425131e-01 7.32867777e-01 -5.15930593e-01
-7.10663617e-01 -6.61189258e-01 -5.90877414e-01 4.26131994e-01
4.43605572e-01 -8.35617542e-01 -6.31093025e-01 6.36214092e-02] | [8.052557945251465, 4.455775260925293] |
91d54406-4e6b-49c6-a929-b24552766ae9 | diacorrect-end-to-end-error-correction-for | 2210.17189 | null | https://arxiv.org/abs/2210.17189v1 | https://arxiv.org/pdf/2210.17189v1.pdf | DiaCorrect: End-to-end error correction for speaker diarization | In recent years, speaker diarization has attracted widespread attention. To achieve better performance, some studies propose to diarize speech in multiple stages. Although these methods might bring additional benefits, most of them are quite complex. Motivated by spelling correction in automatic speech recognition (ASR), in this paper, we propose an end-to-end error correction framework, termed DiaCorrect, to refine the initial diarization results in a simple but efficient way. By exploiting the acoustic interactions between input mixture and its corresponding speaker activity, DiaCorrect could automatically adapt the initial speaker activity to minimize the diarization errors. Without bells and whistles, experiments on LibriSpeech based 2-speaker meeting-like data show that, the self-attentitive end-to-end neural diarization (SA-EEND) baseline with DiaCorrect could reduce its diarization error rate (DER) by over 62.4% from 12.31% to 4.63%. Our source code is available online at https://github.com/jyhan03/diacorrect. | ['Yanhua Long', 'Heng Lu', 'Yuhang Cao', 'Jiangyu Han'] | 2022-10-31 | null | null | null | null | ['spelling-correction'] | ['natural-language-processing'] | [-2.64017330e-03 1.07327580e-01 2.56763488e-01 -5.57917178e-01
-1.12796414e+00 -4.39788818e-01 4.01126951e-01 -1.91068277e-01
-4.40726101e-01 2.99361229e-01 4.12529767e-01 -2.66030997e-01
2.14480519e-01 -1.58303782e-01 -4.70697433e-01 -6.43371582e-01
2.16453686e-01 1.44684121e-01 9.14229751e-02 -9.33378786e-02
1.45680100e-01 2.39507467e-01 -1.34303546e+00 1.52388901e-01
1.13231349e+00 7.04850614e-01 3.33548725e-01 9.11846101e-01
1.66012779e-01 2.60867834e-01 -9.70245957e-01 -1.95695326e-01
1.33958966e-01 -5.67091405e-01 -3.94721597e-01 -1.32747861e-02
2.77853996e-01 -7.92050883e-02 -4.23205972e-01 1.14942861e+00
9.75339293e-01 2.58382291e-01 4.96140361e-01 -9.59735155e-01
-4.67026561e-01 9.92296636e-01 -4.12473857e-01 4.59148407e-01
1.15081951e-01 8.57320204e-02 7.11220920e-01 -1.21004498e+00
-9.12913308e-02 1.32569695e+00 4.94279891e-01 9.14993644e-01
-1.20402181e+00 -1.09310758e+00 3.37157160e-01 2.66309261e-01
-1.84936213e+00 -1.20201302e+00 7.92096376e-01 -2.14475051e-01
9.10156608e-01 5.78884900e-01 1.04053788e-01 9.13807094e-01
-2.51514018e-01 9.20777023e-01 6.61494434e-01 -5.11410356e-01
1.55058637e-01 1.50595143e-01 3.15085649e-01 1.00008860e-01
-1.90609202e-01 1.67523712e-01 -7.31372893e-01 5.18690869e-02
3.94850463e-01 -2.60816127e-01 -5.72502732e-01 4.29472476e-01
-9.86224890e-01 4.57444251e-01 2.11977839e-01 5.90969980e-01
-2.76570976e-01 -2.23634705e-01 2.00790882e-01 6.32427558e-02
5.89411438e-01 3.00894082e-01 -3.45475405e-01 -2.97063798e-01
-1.00135386e+00 4.79064994e-02 5.95369220e-01 8.67520273e-01
4.06013727e-01 5.38977563e-01 -3.61492664e-01 1.38458204e+00
5.32899559e-01 6.72087610e-01 8.69523466e-01 -6.98958218e-01
5.74639440e-01 4.16881070e-02 -7.65995681e-02 -6.45018816e-01
-2.91268349e-01 -4.71520811e-01 -1.02004647e+00 -1.60663470e-03
2.99127042e-01 -4.47515100e-01 -9.02059495e-01 1.73717129e+00
1.94753721e-01 1.98162436e-01 6.08172119e-02 9.60535645e-01
6.82334244e-01 1.04134798e+00 -1.04055338e-01 -2.38704398e-01
1.22557259e+00 -9.86448765e-01 -1.04017901e+00 -3.98411989e-01
3.47063869e-01 -9.84015048e-01 1.05870020e+00 6.26502156e-01
-1.16675067e+00 -7.01878905e-01 -1.13514769e+00 2.61030734e-01
1.26596272e-01 1.73256338e-01 -1.45663127e-01 7.44935632e-01
-1.16680217e+00 2.62056470e-01 -9.18424964e-01 -1.06110469e-01
4.98077646e-02 3.68025571e-01 1.13898128e-01 2.76576728e-01
-1.23554301e+00 5.53002119e-01 3.00565213e-01 2.60211140e-01
-7.21646369e-01 -6.28738165e-01 -6.30905509e-01 1.43302143e-01
4.03456748e-01 -1.91729188e-01 1.80915821e+00 -8.75100851e-01
-2.07677102e+00 4.87467378e-01 -5.23308277e-01 -3.79877806e-01
2.28676394e-01 -4.69670683e-01 -7.29783893e-01 -3.25755149e-01
-3.49391699e-01 4.42520887e-01 7.17427850e-01 -1.12220824e+00
-6.17247105e-01 -3.80954206e-01 -5.94745576e-01 4.66048181e-01
-3.43078554e-01 3.31982404e-01 -6.32944047e-01 -1.05365825e+00
2.51717120e-01 -9.96042013e-01 6.14898168e-02 -5.92571199e-01
-5.29789746e-01 -4.44335610e-01 5.76380253e-01 -9.89625692e-01
1.71411598e+00 -2.35865164e+00 1.04106680e-01 -6.32421952e-03
9.03571025e-02 7.65583754e-01 -4.16728966e-02 1.62652284e-01
-3.19210105e-02 1.33655071e-01 -2.54842609e-01 -6.90792799e-01
1.43667892e-01 -2.42173016e-01 -6.19948329e-03 3.40807945e-01
1.13739677e-01 4.47991967e-01 -6.25309527e-01 -4.19422649e-02
2.21601352e-01 7.34184444e-01 -8.00362587e-01 5.65673053e-01
2.09225714e-01 3.65743458e-01 4.06193882e-02 3.84552896e-01
7.33956575e-01 3.48710120e-01 -1.17391653e-01 -7.16412067e-03
-3.19883794e-01 6.16620243e-01 -1.28398407e+00 1.46903539e+00
-5.34845114e-01 9.21803832e-01 4.39460993e-01 -6.90915942e-01
9.90416825e-01 6.16626620e-01 4.90667671e-02 -6.28508568e-01
2.50442922e-01 2.65637010e-01 3.34612995e-01 -1.47028133e-01
4.07816827e-01 -3.24233994e-02 1.24337032e-01 1.08501934e-01
-7.97179788e-02 -1.01489738e-01 -1.86551362e-01 -4.18420590e-04
7.33081579e-01 -4.74077612e-01 2.27679014e-01 -1.44542083e-01
7.02725053e-01 -6.89742684e-01 7.41897523e-01 5.97715080e-01
-4.71878529e-01 8.97460103e-01 -3.02082121e-01 3.00786674e-01
-6.77440464e-01 -1.09414542e+00 -1.05593301e-01 1.03739965e+00
-1.29220992e-01 -4.55923766e-01 -1.18089640e+00 -3.24407667e-01
-3.27708215e-01 1.12007022e+00 -8.95181373e-02 -3.30182284e-01
-7.13657498e-01 -5.83551884e-01 8.10751557e-01 2.84139752e-01
4.93850261e-01 -9.00877893e-01 1.88040137e-01 3.29562634e-01
-4.97313350e-01 -8.13351393e-01 -1.31034040e+00 1.87136605e-01
-5.32806218e-01 -1.93211466e-01 -9.49862063e-01 -6.68634057e-01
4.49739337e-01 5.49767613e-01 6.92406774e-01 -2.31780455e-01
2.20998839e-01 -6.60998374e-02 -3.96291643e-01 -5.30543804e-01
-8.44399214e-01 1.04868390e-01 4.71337825e-01 2.15350315e-01
6.73881590e-01 -5.46519101e-01 -6.22842073e-01 6.06898010e-01
-6.98705614e-01 -2.55380161e-02 5.07491171e-01 5.36456168e-01
4.69491273e-01 -1.21470410e-02 7.53146708e-01 -3.70021373e-01
6.60255969e-01 -3.10046345e-01 -6.25360489e-01 -7.97371045e-02
-8.17510724e-01 6.36848435e-02 5.46148419e-01 -6.73137486e-01
-1.16193628e+00 -2.46258080e-01 -7.61130333e-01 -4.51223403e-01
-3.94303352e-01 2.53279179e-01 -7.16062069e-01 3.89050305e-01
7.12760627e-01 3.03349733e-01 -2.05554038e-01 -6.65023744e-01
1.14341550e-01 1.39080012e+00 5.87376952e-01 -7.36408681e-03
7.00418413e-01 -2.18493387e-01 -1.01126504e+00 -1.06142700e+00
-6.60478890e-01 -6.15504682e-01 -3.18339586e-01 -1.86044574e-01
6.39450848e-01 -1.07189572e+00 -5.77794611e-01 8.34159374e-01
-1.23061895e+00 -4.73086447e-01 1.22682765e-01 7.08366454e-01
-1.12098306e-01 3.80935073e-01 -4.66070920e-01 -9.65430200e-01
-6.29208088e-01 -1.12131476e+00 7.53356159e-01 3.81260931e-01
-4.05685484e-01 -4.54313010e-01 1.57704592e-01 4.42096740e-01
6.80978358e-01 -5.75040877e-01 1.82979926e-01 -8.13222110e-01
-1.56935766e-01 -7.65225217e-02 1.53085262e-01 6.66227698e-01
4.05965984e-01 -1.13281138e-01 -1.26765501e+00 -2.43495449e-01
1.78464904e-01 4.04958248e-01 4.94706482e-01 5.33138275e-01
1.13121426e+00 -6.70354307e-01 -1.43961430e-01 5.24167538e-01
7.77806520e-01 7.41217136e-01 6.18540704e-01 8.69210213e-02
6.60438299e-01 3.39694500e-01 3.64058107e-01 5.72439432e-01
3.51958126e-01 1.12951720e+00 -7.65279820e-03 6.60264641e-02
-5.76557338e-01 -7.33392537e-02 7.82221854e-01 1.63774860e+00
1.76989079e-01 -6.01041257e-01 -8.77094030e-01 5.12143433e-01
-1.48726416e+00 -9.70191300e-01 -5.64863198e-02 2.40244675e+00
1.30327547e+00 2.41622210e-01 2.54217833e-01 2.60610729e-01
1.07227743e+00 5.76513894e-02 -6.00822151e-01 -5.23979902e-01
-3.10842972e-02 -2.99672578e-02 2.71661192e-01 9.51073349e-01
-7.91729093e-01 9.21539068e-01 5.02128267e+00 9.77531672e-01
-1.20271111e+00 2.12352335e-01 4.98631299e-01 -1.39838755e-01
-1.42825574e-01 -4.25384521e-01 -1.14460468e+00 6.31540775e-01
1.46589124e+00 -3.86681378e-01 6.71854496e-01 7.04069495e-01
6.78910971e-01 1.57127231e-01 -1.02889454e+00 1.24586713e+00
3.24860543e-01 -7.31861413e-01 -1.99185669e-01 -1.54371828e-01
4.22736108e-01 1.19022377e-01 1.54914865e-02 4.52158034e-01
2.72627831e-01 -7.21801877e-01 9.00170743e-01 2.51314521e-01
5.65719366e-01 -8.42345357e-01 4.94386554e-01 3.24158102e-01
-1.03060699e+00 1.64313242e-01 3.10881692e-03 1.98947817e-01
2.39253134e-01 6.61558747e-01 -1.16208363e+00 2.73126334e-01
6.00593746e-01 1.15239069e-01 -2.96669006e-01 1.21865737e+00
-3.76924455e-01 1.18239582e+00 -2.91196883e-01 1.55193750e-02
-2.03231990e-01 1.50413498e-01 9.18986976e-01 1.58108056e+00
4.02202129e-01 2.78837264e-01 -2.39178672e-01 5.84464252e-01
-2.39925519e-01 -1.32004833e-02 -4.00617570e-02 1.51107786e-02
8.95225585e-01 8.80178571e-01 -1.49710372e-01 -2.63225228e-01
-7.30652586e-02 1.24701381e+00 9.14453119e-02 5.16137600e-01
-1.11163378e+00 -9.57567692e-01 9.00532186e-01 -2.01336481e-03
2.75299758e-01 -3.17734927e-01 -2.31867507e-01 -1.10093319e+00
-4.79329266e-02 -1.10999179e+00 -1.68627296e-02 -5.24466932e-01
-9.74066734e-01 9.91807759e-01 -2.08949208e-01 -1.11799729e+00
-3.22883546e-01 -1.38167188e-01 -5.78658223e-01 1.07893515e+00
-1.29711270e+00 -5.06638587e-01 -2.63816208e-01 3.65716994e-01
1.18457878e+00 -4.40565161e-02 6.51742399e-01 7.09376812e-01
-1.04806685e+00 1.32968915e+00 2.20779702e-01 6.09921142e-02
9.99663472e-01 -1.03436148e+00 5.88718534e-01 1.24697495e+00
1.26165032e-01 4.98971134e-01 9.65995669e-01 -3.10548961e-01
-9.80302393e-01 -1.20374382e+00 1.12189102e+00 -2.45792851e-01
3.52107644e-01 -5.84745884e-01 -1.21995556e+00 4.99768883e-01
3.65457892e-01 -4.77833331e-01 5.29826760e-01 8.79508555e-02
-1.89565033e-01 -3.12368900e-01 -7.18883693e-01 9.23292577e-01
9.08002913e-01 -3.95468563e-01 -6.19749188e-01 -9.20590162e-02
8.61317158e-01 -3.59722465e-01 -4.72894430e-01 2.70231545e-01
2.46986389e-01 -7.08839357e-01 7.45788932e-01 1.58062503e-01
-2.68072098e-01 -4.67282146e-01 -2.14755937e-01 -1.67754281e+00
-4.53772634e-01 -1.13865244e+00 -1.57606885e-01 1.72257149e+00
5.63323617e-01 -6.25027478e-01 3.27534556e-01 6.93917155e-01
-6.04748309e-01 -1.66173026e-01 -1.01252413e+00 -9.77027833e-01
2.43009836e-03 -5.39424360e-01 4.73027438e-01 6.43881917e-01
3.92346345e-02 5.05261481e-01 -3.76634419e-01 5.64136326e-01
4.06158477e-01 -4.32447791e-01 7.52421916e-01 -9.07954216e-01
-3.39679986e-01 -7.54431844e-01 3.87327708e-02 -1.41376805e+00
3.94444913e-03 -8.68872821e-01 7.29280591e-01 -1.22498035e+00
-3.42594922e-01 -2.66518593e-01 -4.59937930e-01 3.60424817e-01
-5.82085729e-01 -8.37569684e-02 1.85826704e-01 8.70067403e-02
-4.29982454e-01 8.81087482e-01 6.78963184e-01 -2.34164417e-01
-7.48201668e-01 3.20557922e-01 -6.88252032e-01 5.76888680e-01
1.14348197e+00 -6.27366841e-01 -2.23119810e-01 -6.15871847e-01
-5.83678722e-01 -9.37717594e-03 -9.55958441e-02 -1.13704789e+00
3.65777731e-01 8.15422758e-02 -1.63972288e-01 -5.75324118e-01
4.35729086e-01 -3.97331834e-01 1.75250411e-01 2.51489669e-01
-4.25654233e-01 -8.86230618e-02 5.71789563e-01 2.86807865e-01
-3.55183601e-01 -2.26523906e-01 9.09359872e-01 2.96719253e-01
-3.21861923e-01 1.63938865e-01 -8.54120851e-01 1.26755657e-02
6.13633811e-01 -1.51727900e-01 -5.70213608e-02 -4.46370602e-01
-4.88926649e-01 9.72813964e-02 2.73310835e-03 5.20117760e-01
4.90666091e-01 -1.26099503e+00 -1.02281547e+00 2.48246461e-01
3.07897609e-02 -2.08225343e-02 4.43407208e-01 6.57674670e-01
7.72944987e-02 4.74192470e-01 3.76630187e-01 -5.11233747e-01
-1.53397346e+00 2.29513258e-01 5.32306969e-01 2.01649129e-01
-3.69461805e-01 1.28303826e+00 4.87956405e-01 -3.17157388e-01
7.73041785e-01 -4.51702893e-01 -1.11391552e-01 -1.35740727e-01
8.34815860e-01 3.92035007e-01 3.44954759e-01 -5.80290616e-01
-5.78509927e-01 1.97463840e-01 -3.37067485e-01 -4.75301713e-01
9.86254513e-01 -6.34614468e-01 4.96886700e-01 4.91920531e-01
1.08130956e+00 3.95001650e-01 -1.33500373e+00 -3.64450157e-01
4.91487756e-02 -3.73402774e-01 3.46867144e-01 -9.25802946e-01
-9.30905640e-01 8.92789185e-01 6.53247714e-01 1.74453020e-01
1.20588160e+00 -5.54967970e-02 9.04915810e-01 1.21922903e-01
-8.53381455e-02 -1.05705488e+00 9.63893253e-03 7.20015526e-01
1.25360930e+00 -1.22369194e+00 -5.90190709e-01 -1.43103153e-01
-6.43631637e-01 8.22266698e-01 7.24483848e-01 1.78618401e-01
6.34224594e-01 2.03108892e-01 3.58128071e-01 2.94836313e-01
-6.10222816e-01 -5.03015891e-02 3.62740815e-01 4.09812838e-01
6.54349566e-01 1.63739160e-01 -7.90656134e-02 9.32375848e-01
-3.20000529e-01 -4.80408847e-01 4.43484455e-01 4.98776525e-01
-5.20447195e-01 -9.71706331e-01 -7.25414395e-01 -5.80263771e-02
-4.86153245e-01 -3.70200545e-01 -3.08873981e-01 1.50233924e-01
-2.59050220e-01 1.55871296e+00 4.98545766e-02 -5.86017251e-01
7.33475268e-01 7.16284290e-02 2.76240744e-02 -6.33169174e-01
-6.38611495e-01 6.08552158e-01 5.67333102e-02 -1.59260631e-01
-1.11576848e-01 -7.62260318e-01 -1.31293309e+00 -2.63735563e-01
-6.40432298e-01 2.35389248e-01 7.49423921e-01 7.58951187e-01
7.36220837e-01 9.20688152e-01 8.70342195e-01 -8.94990683e-01
-4.91151363e-01 -1.45025826e+00 -3.54013890e-01 4.57927622e-02
4.78037417e-01 -2.98584491e-01 -7.04621255e-01 1.56836614e-01] | [14.714470863342285, 6.226296424865723] |
b24d1a66-afb4-453f-850e-4237659bbcf8 | why-using-either-aggregated-features-or | 2306.08274 | null | https://arxiv.org/abs/2306.08274v1 | https://arxiv.org/pdf/2306.08274v1.pdf | Why Using Either Aggregated Features or Adjacency Lists in Directed or Undirected Graph? Empirical Study and Simple Classification Method | Node classification is one of the hottest tasks in graph analysis. In this paper, we focus on the choices of node representations (aggregated features vs. adjacency lists) and the edge direction of an input graph (directed vs. undirected), which have a large influence on classification results. We address the first empirical study to benchmark the performance of various GNNs that use either combination of node representations and edge directions. Our experiments demonstrate that no single combination stably achieves state-of-the-art results across datasets, which indicates that we need to select appropriate combinations depending on the characteristics of datasets. In response, we propose a simple yet holistic classification method A2DUG which leverages all combinations of node representation variants in directed and undirected graphs. We demonstrate that A2DUG stably performs well on various datasets. Surprisingly, it largely outperforms the current state-of-the-art methods in several datasets. This result validates the importance of the adaptive effect control on the combinations of node representations and edge directions. | ['Makoto Onizuka', 'Yuya Sasaki', 'Seiji Maekawa'] | 2023-06-14 | null | null | null | null | ['classification-1'] | ['methodology'] | [ 1.40646100e-01 -2.02304885e-01 -5.41367531e-01 -1.18079871e-01
1.02437902e-02 -8.14178765e-01 9.94707763e-01 4.87267464e-01
-1.58899695e-01 4.69195545e-01 7.01970384e-02 -6.71171665e-01
-6.42034173e-01 -1.14581037e+00 -1.39190450e-01 -5.98127127e-01
-4.42455530e-01 3.11156511e-01 2.78545409e-01 -4.27360684e-01
3.38351369e-01 6.26260996e-01 -1.46213901e+00 -1.83147803e-01
5.91753125e-01 7.86406100e-01 -1.89424291e-01 5.75630426e-01
-1.75477892e-01 5.47326684e-01 -4.54613715e-01 -3.44781846e-01
3.50382000e-01 -2.57026851e-01 -7.05782354e-01 -1.12319276e-01
3.59996885e-01 1.83917731e-01 -4.09096777e-01 8.50137591e-01
4.39878494e-01 -1.21534139e-01 9.88178551e-01 -1.51601589e+00
-3.90752226e-01 7.60615468e-01 -7.16509700e-01 4.60477918e-01
3.72465432e-01 1.66828498e-01 1.61149693e+00 -4.26633209e-01
8.45787823e-01 1.05616605e+00 8.75005960e-01 1.07817590e-01
-1.60467720e+00 -4.52147812e-01 5.51842034e-01 -9.13906172e-02
-1.53295970e+00 -3.03527951e-01 7.65811563e-01 -4.87198263e-01
8.35829854e-01 4.50320929e-01 5.41751504e-01 9.53580737e-01
2.77238756e-01 2.53548115e-01 1.00876451e+00 -5.00723362e-01
2.51297206e-01 -8.74645412e-02 6.36175573e-01 8.03823292e-01
8.90371144e-01 -2.22090885e-01 -2.74138391e-01 -3.80926460e-01
3.95281017e-01 -1.40323788e-01 -3.37103814e-01 -6.67084336e-01
-1.12299740e+00 9.49930847e-01 5.63083470e-01 4.50394571e-01
-2.20806316e-01 2.81625599e-01 6.41061842e-01 4.74364012e-01
3.09811831e-01 8.22713614e-01 -2.65836626e-01 3.54595408e-02
-6.01880610e-01 1.03581101e-01 8.92686605e-01 6.63055241e-01
6.57843053e-01 -1.60969377e-01 -4.33125824e-01 7.20343173e-01
1.24005415e-01 5.14423251e-02 8.66037011e-02 -3.21026236e-01
4.20822531e-01 9.47064579e-01 -4.61470813e-01 -1.31997776e+00
-6.93300426e-01 -8.85979056e-01 -6.83745861e-01 8.06562975e-02
4.74689484e-01 -2.02986881e-01 -1.00932717e+00 1.78614330e+00
2.58302987e-01 -1.76878706e-01 -1.45136371e-01 6.26570761e-01
1.01358271e+00 2.09238201e-01 1.65837780e-01 1.55441254e-01
1.26858163e+00 -8.17515969e-01 -4.41765279e-01 -3.43052059e-01
1.07065237e+00 -4.31514561e-01 1.01127970e+00 1.42467484e-01
-6.85163260e-01 -2.31429100e-01 -1.22431934e+00 2.60630637e-01
-6.78788900e-01 -9.97062400e-02 9.85946655e-01 8.81160498e-01
-1.07327926e+00 7.64639974e-01 -5.14769137e-01 -7.21527576e-01
2.14102909e-01 4.14323241e-01 -5.36030471e-01 -1.39998913e-01
-1.02633882e+00 7.31995761e-01 3.60994667e-01 -2.40559310e-01
-3.68573427e-01 -5.40843368e-01 -7.17794538e-01 2.63529658e-01
4.12658840e-01 -6.26979828e-01 7.42083013e-01 -6.77452862e-01
-7.90477812e-01 8.16937447e-01 2.36379996e-01 -3.39218825e-01
3.78638327e-01 3.81852329e-01 -4.48316485e-01 6.92368299e-02
-6.53763115e-02 4.94673818e-01 4.66584146e-01 -1.21336198e+00
-3.88062596e-01 -3.61356705e-01 3.41704011e-01 7.39548653e-02
-6.85756207e-01 -2.40593642e-01 -4.94984895e-01 -6.42270386e-01
1.98342040e-01 -1.08772492e+00 -2.02782616e-01 -2.03052729e-01
-6.28045797e-01 -4.08582717e-01 6.46579921e-01 -8.70722812e-03
1.74846661e+00 -2.08569884e+00 1.03812926e-01 7.14603901e-01
5.73761046e-01 1.95054263e-02 -2.79789120e-01 9.29132640e-01
-2.08104417e-01 6.27424359e-01 1.70739785e-01 -1.89029500e-01
1.95974652e-02 1.08421594e-01 1.03443854e-01 6.49758756e-01
7.61138201e-02 7.72153974e-01 -9.25347149e-01 -3.74198318e-01
8.07699049e-04 3.67497772e-01 -4.00946975e-01 -1.74238250e-01
4.30577397e-02 8.94507393e-02 -6.90624058e-01 5.73623538e-01
6.09543502e-01 -6.88850224e-01 7.06271589e-01 -2.95967191e-01
1.52556494e-01 5.26181340e-01 -1.03699899e+00 1.18061113e+00
-2.76289910e-01 5.88513374e-01 -8.47228765e-02 -8.14692318e-01
9.99913990e-01 -7.80855939e-02 3.46704751e-01 -4.57966745e-01
1.21442869e-01 1.90460205e-01 5.80807626e-01 2.00848728e-02
3.79320115e-01 3.90945911e-01 -8.86319578e-02 4.54809904e-01
-3.44705991e-02 1.70808494e-01 5.50620675e-01 5.68243265e-01
1.69238591e+00 -4.03902203e-01 6.56469285e-01 -8.79137218e-01
2.12582231e-01 -1.29980728e-01 3.25424939e-01 1.01003754e+00
-2.12808093e-03 6.49410605e-01 1.13087738e+00 -3.70572269e-01
-5.25002241e-01 -7.82926083e-01 -8.99024084e-02 1.27439129e+00
1.68205708e-01 -9.94158149e-01 -4.24681008e-01 -9.15809989e-01
2.92946339e-01 4.92500007e-01 -9.99262035e-01 -2.61864334e-01
-2.57801324e-01 -8.12283099e-01 5.23755610e-01 4.57762033e-01
7.31070340e-02 -6.21678412e-01 -3.12295616e-01 -5.56441434e-02
4.35261846e-01 -1.14862573e+00 -1.98987067e-01 5.37240744e-01
-8.31519186e-01 -1.23130357e+00 -2.09410310e-01 -5.94554067e-01
7.56121755e-01 5.74948192e-01 1.50371003e+00 5.48743844e-01
8.33754614e-02 4.71131951e-01 -7.73073852e-01 8.81461874e-02
-2.00303093e-01 8.31595778e-01 -1.67158186e-01 -8.44079182e-02
3.91244777e-02 -9.21408415e-01 -5.77188969e-01 2.14584127e-01
-7.74727464e-01 -2.93552969e-02 5.97947538e-01 5.82540452e-01
1.45797297e-01 2.00667247e-01 3.90372157e-01 -1.54567993e+00
1.00242424e+00 -7.14312673e-01 -3.27312917e-01 4.25841749e-01
-9.37978685e-01 2.20697448e-01 7.08842397e-01 -1.83748290e-01
-1.83422953e-01 -2.91066408e-01 3.91682126e-02 -1.66660726e-01
2.23043814e-01 8.19351137e-01 -4.43520360e-02 -3.17596853e-01
6.06991529e-01 -2.18363479e-01 -1.45192593e-01 -3.69174391e-01
1.59074977e-01 3.08851987e-01 7.64165446e-03 -4.89715040e-01
6.32261753e-01 1.82384431e-01 4.35927749e-01 -7.94131577e-01
-1.85171247e-01 -2.14403063e-01 -5.74748576e-01 -7.63962492e-02
5.45965135e-01 -6.23002946e-01 -5.15871108e-01 2.14336753e-01
-6.84759557e-01 -2.75684297e-01 1.99264824e-01 6.57808334e-02
1.20696217e-01 3.13999712e-01 -4.60628420e-01 -5.20276248e-01
-3.20657194e-01 -1.30236006e+00 8.37030292e-01 -1.44480020e-01
-3.54862124e-01 -1.22304487e+00 -1.71724096e-01 -1.52981743e-01
4.89312947e-01 5.57079494e-01 1.24744880e+00 -1.02080584e+00
-4.03751343e-01 -2.67373294e-01 -3.23364556e-01 -2.50478834e-01
3.85459632e-01 3.04274529e-01 -6.24801219e-01 -5.95239520e-01
-9.47261989e-01 -5.06832451e-02 1.25731790e+00 4.88179587e-02
1.14900458e+00 -1.47054911e-01 -7.70938933e-01 6.17966175e-01
1.43870330e+00 -1.88529700e-01 4.79209602e-01 3.01459879e-01
8.54032636e-01 5.40453732e-01 3.17112386e-01 3.23727638e-01
4.37828451e-01 7.18536377e-01 6.63832188e-01 -5.65056913e-02
-2.08942145e-01 -1.85621530e-01 6.43865690e-02 4.48630393e-01
5.38939685e-02 -7.45405316e-01 -1.20131278e+00 4.33098912e-01
-1.64363456e+00 -5.70313394e-01 -3.45450461e-01 2.17085981e+00
2.85316229e-01 4.20972794e-01 3.22561085e-01 3.74968141e-01
7.30560780e-01 6.72234714e-01 -3.38037610e-01 -3.87715727e-01
6.21105358e-02 1.46597579e-01 7.06363618e-01 1.88146725e-01
-1.09287453e+00 6.54154897e-01 6.56613350e+00 8.41019750e-01
-1.19665432e+00 -2.37463683e-01 6.71877086e-01 3.01864147e-01
-5.78869522e-01 -3.79320309e-02 -7.10922003e-01 4.40031052e-01
7.74614990e-01 -1.65756747e-01 3.90454173e-01 5.86303353e-01
-1.82446018e-01 1.57780960e-01 -1.25158751e+00 7.22305536e-01
-1.53744906e-01 -1.27788723e+00 8.38674828e-02 3.93733740e-01
6.04997277e-01 2.24940956e-01 -1.23088099e-01 1.95248991e-01
4.58896726e-01 -1.16885459e+00 4.20104504e-01 3.57363932e-02
6.86245143e-01 -4.58533853e-01 6.47388816e-01 1.34377815e-02
-1.47008681e+00 -1.66399464e-01 1.01422705e-01 -7.72229061e-02
-2.79744774e-01 6.75718427e-01 -7.84083486e-01 6.87082708e-01
5.62629879e-01 6.85376644e-01 -1.09013557e+00 8.29234421e-01
-1.09222434e-01 7.46170998e-01 -3.54044557e-01 -3.16949517e-01
2.33631656e-01 -8.68760329e-03 5.32476664e-01 1.37954712e+00
2.11668670e-01 -2.03219295e-01 3.41567963e-01 5.08498907e-01
-1.48365289e-01 1.44667894e-01 -9.19714630e-01 -2.49909401e-01
8.75836372e-01 1.48877108e+00 -1.28805387e+00 1.17803989e-02
-4.41918314e-01 2.90876716e-01 6.04263365e-01 2.84406662e-01
-5.64018786e-01 -3.14071387e-01 6.97213948e-01 4.17584926e-01
5.08510292e-01 -3.39706182e-01 -2.51103431e-01 -8.95365238e-01
-7.61417001e-02 -8.58052850e-01 8.11456919e-01 -3.31898659e-01
-1.58765697e+00 7.49258459e-01 9.62749645e-02 -1.15696776e+00
1.55742308e-02 -7.10702658e-01 -9.23484027e-01 4.55785334e-01
-1.27848601e+00 -8.97872150e-01 -5.08943200e-01 3.16488862e-01
-9.50461254e-02 1.45606408e-02 6.81647182e-01 9.06361863e-02
-6.80997014e-01 7.23855495e-01 -4.14013863e-02 2.29825929e-01
3.96476060e-01 -1.25455880e+00 6.42825246e-01 7.87476242e-01
1.27940893e-01 8.21697414e-01 7.52946079e-01 -5.59438884e-01
-1.64455903e+00 -9.11958158e-01 5.30146837e-01 -1.11908644e-01
7.35619485e-01 -5.57593226e-01 -7.30788052e-01 7.00254500e-01
1.78654402e-01 4.47943151e-01 6.23360932e-01 6.10455811e-01
-5.19669652e-01 -2.86856741e-01 -1.01629317e+00 7.26926684e-01
1.49917305e+00 -2.81466067e-01 8.73309597e-02 1.75505504e-01
6.51680887e-01 -7.91231021e-02 -9.98076320e-01 8.02491367e-01
5.16513467e-01 -1.08658051e+00 8.65864336e-01 -6.39445722e-01
2.84414560e-01 -1.72148839e-01 -2.28712440e-01 -1.53345263e+00
-6.82560503e-01 -3.27474028e-01 -1.12238951e-01 1.46657526e+00
7.97958374e-01 -1.21709108e+00 7.47451782e-01 3.94717336e-01
2.45820537e-01 -1.10230350e+00 -6.22440755e-01 -6.61037147e-01
-2.23105755e-02 1.55731319e-02 7.33239949e-01 1.15358448e+00
-5.13895862e-02 6.90816700e-01 1.77733332e-01 -7.39031881e-02
3.21899325e-01 3.68858874e-01 9.20792818e-01 -1.51183331e+00
-6.67124763e-02 -8.26700568e-01 -8.59551966e-01 -7.05078721e-01
1.20866604e-01 -1.21118760e+00 -6.01390839e-01 -1.67169642e+00
1.59395099e-01 -9.16924119e-01 -4.14477110e-01 6.26232624e-01
-4.06077117e-01 2.27403522e-01 1.50987938e-01 5.56060001e-02
-5.87300003e-01 2.90153742e-01 8.72327447e-01 -1.21699929e-01
-2.09279045e-01 -1.34656295e-01 -1.06823564e+00 3.06462139e-01
1.04173517e+00 -2.82189846e-01 -5.90076804e-01 -4.39063638e-01
3.26936156e-01 -2.38838241e-01 1.61667943e-01 -9.83556330e-01
1.85889661e-01 -1.49271637e-01 2.08298996e-01 -1.95071310e-01
9.86583531e-02 -7.13378251e-01 2.32644171e-01 4.39605623e-01
-4.68414545e-01 5.36193669e-01 -2.85323784e-02 5.99302530e-01
5.61354943e-02 -3.82872224e-02 5.21875560e-01 -1.98094733e-02
-7.06679404e-01 2.56212085e-01 -5.93370125e-02 2.45973364e-01
1.00732172e+00 -4.00192350e-01 -6.84337378e-01 -3.71554941e-01
-3.46862644e-01 3.24690580e-01 6.09669328e-01 5.15653074e-01
1.57223612e-01 -1.25481951e+00 -7.06078351e-01 1.27784967e-01
4.39872622e-01 -5.74322045e-01 -2.49862432e-01 7.85985589e-01
-5.12304902e-01 2.88919926e-01 -5.79215139e-02 -6.26304984e-01
-1.48029399e+00 5.09996414e-01 2.80721575e-01 -6.46589100e-01
-4.87534583e-01 5.18552601e-01 6.00944124e-02 -5.49878895e-01
7.82496780e-02 -6.29817620e-02 -4.03519571e-01 2.05239877e-01
-3.19566876e-02 4.14015979e-01 3.80022973e-01 -5.36178946e-01
-6.21082664e-01 4.30442780e-01 -2.38293692e-01 4.36073542e-01
1.30870831e+00 3.54115479e-02 -1.16142780e-01 2.55953461e-01
1.21270418e+00 1.95977300e-01 -6.51213229e-01 5.50014824e-02
5.49050681e-02 -6.33196533e-01 7.92510286e-02 -6.19012892e-01
-1.42403483e+00 4.80764717e-01 9.31748599e-02 8.12825322e-01
1.03637218e+00 -6.86205598e-03 2.98520774e-01 2.67186224e-01
6.08922482e-01 -4.85295236e-01 -1.81393310e-01 3.16674173e-01
6.29239321e-01 -9.89221215e-01 5.42153418e-01 -9.23541248e-01
-4.89234835e-01 1.07293606e+00 6.26419723e-01 -1.92924455e-01
7.86348104e-01 1.65686920e-01 -1.33710146e-01 -4.38894838e-01
-9.82380450e-01 -4.12254721e-01 3.49695981e-01 4.57842886e-01
6.80803776e-01 3.11592489e-01 -5.77393055e-01 8.96308720e-02
-2.85953790e-01 -5.81103563e-01 3.95926297e-01 8.43234479e-01
-1.32038951e-01 -1.34372747e+00 -3.63730527e-02 9.67227578e-01
-3.88241798e-01 -1.06283188e-01 -9.24598396e-01 1.21936095e+00
-2.08237648e-01 1.00536478e+00 -4.56309654e-02 -6.46115899e-01
2.43195117e-01 -2.97548831e-01 4.30558145e-01 -6.51238978e-01
-9.30900574e-01 -4.86158848e-01 5.19070446e-01 -3.39751184e-01
-1.61133334e-01 -4.35546905e-01 -1.00820470e+00 -6.57277822e-01
-6.02701247e-01 2.24553347e-01 3.90900850e-01 5.77601910e-01
6.74590826e-01 6.06414616e-01 5.95939934e-01 -7.29467154e-01
-5.18507540e-01 -8.29659283e-01 -7.54590869e-01 5.59084713e-01
1.82090044e-01 -9.75133836e-01 -5.53371966e-01 -9.04967666e-01] | [6.968825340270996, 6.099739074707031] |
252eeb7f-45e7-445e-a025-d92d457c819a | templates-for-3d-object-pose-estimation | 2203.17234 | null | https://arxiv.org/abs/2203.17234v1 | https://arxiv.org/pdf/2203.17234v1.pdf | Templates for 3D Object Pose Estimation Revisited: Generalization to New Objects and Robustness to Occlusions | We present a method that can recognize new objects and estimate their 3D pose in RGB images even under partial occlusions. Our method requires neither a training phase on these objects nor real images depicting them, only their CAD models. It relies on a small set of training objects to learn local object representations, which allow us to locally match the input image to a set of "templates", rendered images of the CAD models for the new objects. In contrast with the state-of-the-art methods, the new objects on which our method is applied can be very different from the training objects. As a result, we are the first to show generalization without retraining on the LINEMOD and Occlusion-LINEMOD datasets. Our analysis of the failure modes of previous template-based approaches further confirms the benefits of local features for template matching. We outperform the state-of-the-art template matching methods on the LINEMOD, Occlusion-LINEMOD and T-LESS datasets. Our source code and data are publicly available at https://github.com/nv-nguyen/template-pose | ['Vincent Lepetit', 'Mathieu Salzmann', 'Yang Xiao', 'Yinlin Hu', 'Van Nguyen Nguyen'] | 2022-03-31 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Nguyen_Templates_for_3D_Object_Pose_Estimation_Revisited_Generalization_to_New_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Nguyen_Templates_for_3D_Object_Pose_Estimation_Revisited_Generalization_to_New_CVPR_2022_paper.pdf | cvpr-2022-1 | ['template-matching', '6d-pose-estimation-1', '6d-pose-estimation'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 3.76765579e-01 -3.17073194e-03 -5.68535067e-02 -5.33313096e-01
-8.27637792e-01 -8.08618844e-01 6.06862605e-01 -2.68233925e-01
-1.50414005e-01 3.45274448e-01 -3.54991615e-01 1.08153142e-01
-6.50584698e-02 -6.94272876e-01 -9.57284927e-01 -5.22302747e-01
1.65533677e-01 9.60878491e-01 6.46261215e-01 -1.52743712e-01
3.26583922e-01 1.05550015e+00 -1.93589365e+00 3.44976515e-01
4.83955115e-01 1.29461098e+00 2.38194898e-01 4.56167251e-01
2.81617250e-02 6.13700151e-02 -3.43300283e-01 -3.61327022e-01
7.69067287e-01 -6.37053773e-02 -6.98765635e-01 4.02256697e-01
1.09894550e+00 -3.28112364e-01 -6.47958338e-01 6.69293284e-01
4.66417640e-01 7.18757287e-02 5.17050743e-01 -1.21219862e+00
-4.41223294e-01 -1.53183147e-01 -3.79164159e-01 -2.42395252e-01
7.58058488e-01 1.70466825e-01 6.76348031e-01 -1.30683589e+00
1.08485878e+00 1.09080696e+00 8.86404455e-01 5.62315822e-01
-1.27476823e+00 -5.76376379e-01 -1.39977476e-02 3.52735251e-01
-1.56987405e+00 -5.41784167e-01 8.37598443e-01 -3.78469199e-01
7.93967783e-01 5.22533596e-01 8.07629049e-01 1.00150621e+00
-3.65074202e-02 5.77928066e-01 1.11692607e+00 -5.85488200e-01
1.32846877e-01 1.03931628e-01 -6.59416765e-02 7.29817808e-01
1.61508188e-01 4.25590098e-01 -5.85710227e-01 -1.11025587e-01
1.00869095e+00 1.07412614e-01 -3.57024461e-01 -1.18331146e+00
-1.22111356e+00 3.97156864e-01 5.98905444e-01 1.75120562e-01
-2.86444128e-01 2.09811687e-01 -3.48817743e-02 1.21469721e-01
3.26960713e-01 2.58822471e-01 -6.09885216e-01 -4.57838438e-02
-7.67474473e-01 2.34076470e-01 7.60665536e-01 1.22327065e+00
1.29262841e+00 -2.85111159e-01 2.89893389e-01 3.98128718e-01
5.55527471e-02 6.43343031e-01 3.14369023e-01 -9.07913983e-01
4.74604458e-01 7.06615865e-01 1.14520535e-01 -8.96974146e-01
-3.58807355e-01 -3.96485955e-01 -3.44852597e-01 5.00370920e-01
3.93780857e-01 4.80359495e-01 -1.19075286e+00 1.18740845e+00
5.19408882e-01 3.57490093e-01 -2.07419395e-01 7.49327958e-01
7.74203897e-01 2.99288988e-01 -6.16837084e-01 1.57664269e-01
1.00429261e+00 -8.83064687e-01 -3.44464153e-01 -4.27979022e-01
4.69724923e-01 -1.04563177e+00 9.55086052e-01 4.31700975e-01
-9.11607802e-01 -8.21263731e-01 -1.15314066e+00 -1.06341518e-01
-6.63979828e-01 7.24081397e-02 6.87291980e-01 6.50907576e-01
-8.25781167e-01 7.99530864e-01 -6.96451604e-01 -5.44156671e-01
4.11525100e-01 6.37422562e-01 -7.18541026e-01 -3.09462398e-01
-4.25652236e-01 1.17099369e+00 2.44102940e-01 2.66087443e-01
-6.29864037e-01 -6.41083717e-01 -8.60654771e-01 -4.53363270e-01
4.63372618e-01 -5.38095176e-01 1.30767632e+00 -9.71706390e-01
-1.31686378e+00 1.32502961e+00 -3.19038659e-01 5.35924621e-02
7.65726507e-01 -2.53275365e-01 -1.46641493e-01 8.45030174e-02
-1.11062489e-01 6.60215676e-01 9.22453940e-01 -1.55661702e+00
-4.80279386e-01 -4.47301775e-01 2.77641993e-02 1.52588159e-01
2.74045348e-01 -2.11998433e-01 -7.62055933e-01 -3.58576566e-01
7.27655053e-01 -1.10586095e+00 -7.36814365e-02 5.99171519e-01
-1.68322191e-01 -4.89142016e-02 1.15612686e+00 -2.19806880e-01
5.21633089e-01 -2.04820371e+00 1.10776797e-01 3.82570654e-01
-1.03756357e-02 1.59202546e-01 -2.76648402e-01 3.82523268e-01
-2.12711096e-01 -1.68885067e-01 -2.44384375e-03 -4.98320848e-01
2.22175848e-02 4.26203310e-01 -2.67140776e-01 7.44791567e-01
1.01974361e-01 9.17787075e-01 -6.90153718e-01 -3.94343317e-01
5.14827311e-01 5.39481401e-01 -1.02912173e-01 2.72798032e-01
-9.13882107e-02 5.32969892e-01 -1.87292680e-01 7.22083747e-01
9.84496474e-01 -9.75497812e-02 3.15191671e-02 -7.04856873e-01
-1.00983925e-01 2.92015165e-01 -1.46562254e+00 2.02443790e+00
-2.17981622e-01 3.81998628e-01 -3.33195060e-01 -8.30777824e-01
1.09898841e+00 1.06679276e-01 6.17319286e-01 -8.16913068e-01
6.43685535e-02 4.74902153e-01 -1.94801807e-01 -4.17138517e-01
2.51121193e-01 1.66954845e-01 3.74297388e-02 4.93320435e-01
5.31729460e-02 -5.64439178e-01 3.33316438e-02 -1.44885615e-01
8.37574601e-01 7.56162524e-01 2.31412321e-01 -2.50108987e-02
2.90663183e-01 4.50845361e-02 4.69281465e-01 7.21970856e-01
1.47413179e-01 1.14775789e+00 -7.16060400e-02 -9.12712157e-01
-1.17480981e+00 -1.29413986e+00 -4.17714745e-01 6.79085433e-01
3.75073463e-01 -3.76496971e-01 -4.70011979e-01 -5.73354185e-01
2.86869466e-01 2.29953378e-01 -8.49871993e-01 1.07146362e-02
-8.15211415e-01 -9.68384966e-02 2.09856465e-01 6.20969236e-01
3.64227355e-01 -8.00370514e-01 -8.80969703e-01 1.45766497e-01
1.61503866e-01 -1.17984176e+00 -4.15385157e-01 2.76016384e-01
-1.06931961e+00 -1.30379713e+00 -4.62192148e-01 -7.33744800e-01
1.01170194e+00 1.43106058e-01 1.14842880e+00 3.09674531e-01
-6.07070863e-01 7.15850949e-01 -2.65819907e-01 -4.36960995e-01
-8.30319300e-02 -4.57490012e-02 3.35976854e-02 4.39801328e-02
1.69566020e-01 -6.63510799e-01 -6.10536456e-01 7.83562124e-01
-7.84159064e-01 2.36891866e-01 5.64683259e-01 5.57367742e-01
8.98465157e-01 -4.24458236e-01 -2.55884409e-01 -6.14003658e-01
-1.07233405e-01 7.99056217e-02 -8.54428470e-01 4.89367276e-01
-2.87179947e-01 1.19109742e-01 6.24113753e-02 -6.88204348e-01
-7.04553246e-01 8.08925331e-01 2.60391444e-01 -5.84202766e-01
-2.06399083e-01 7.64326230e-02 -1.50436312e-01 -5.36369801e-01
6.90441608e-01 7.15985671e-02 -7.81425238e-02 -7.97108591e-01
2.86813259e-01 3.23726475e-01 7.69088149e-01 -7.13808656e-01
1.31919611e+00 6.44256413e-01 2.64395595e-01 -4.46752816e-01
-6.64522946e-01 -4.07181412e-01 -1.54911101e+00 -2.59807140e-01
4.58167404e-01 -5.61343074e-01 -5.56936204e-01 6.02441669e-01
-1.23292077e+00 -2.87283748e-01 -4.74312246e-01 2.89893717e-01
-7.92700469e-01 3.32397223e-01 4.22850475e-02 -5.59548438e-01
-1.00433454e-02 -1.09885716e+00 1.35817611e+00 1.50219813e-01
-8.22961330e-02 -7.70768583e-01 -3.40791382e-02 1.67345732e-01
3.33913743e-01 5.22569299e-01 4.77699310e-01 -3.26953739e-01
-1.03225863e+00 -5.30743480e-01 -9.85194594e-02 1.05573513e-01
3.64920288e-01 1.24244228e-01 -1.10182488e+00 -3.07071149e-01
-9.22737569e-02 -2.49414071e-01 4.66742098e-01 -1.25956625e-01
9.30048823e-01 -3.61650623e-02 -6.05458796e-01 8.21961045e-01
1.41003680e+00 1.92449063e-01 7.33066559e-01 5.57711065e-01
6.53076112e-01 4.86934364e-01 8.34299862e-01 2.93862343e-01
2.28542969e-01 1.15714455e+00 5.88344753e-01 -2.65492231e-01
-2.77120233e-01 -1.22654282e-01 8.01080465e-03 5.73255718e-01
-1.55162096e-01 2.18727097e-01 -1.09617352e+00 2.74268478e-01
-1.92687595e+00 -5.21056235e-01 -1.20574839e-01 2.47009158e+00
6.86805427e-01 -4.98936251e-02 -9.39048082e-02 1.09602727e-01
6.23771369e-01 -6.12586923e-02 -7.58641660e-01 -1.25710443e-01
-1.86988249e-01 6.32927895e-01 4.33807611e-01 4.19660538e-01
-1.14239442e+00 9.21256840e-01 6.08372021e+00 4.74424750e-01
-1.17691910e+00 -2.93575060e-02 1.69459552e-01 3.47736478e-03
-2.88579222e-02 2.95939118e-01 -6.58575714e-01 -4.48647370e-05
4.60466325e-01 -8.24212059e-02 4.91589069e-01 8.06456983e-01
-2.74753124e-01 -2.40624994e-01 -1.65265310e+00 1.22142768e+00
2.61075497e-01 -1.30943954e+00 -1.53115198e-01 5.17002754e-02
6.95907891e-01 4.70399372e-02 5.31240031e-02 -3.56561169e-02
-1.80697873e-01 -1.01142919e+00 8.87650371e-01 8.75146449e-01
9.13836002e-01 -1.96001276e-01 5.80177009e-01 1.21324100e-01
-1.33522809e+00 3.03342611e-01 -6.24324501e-01 7.79453889e-02
-1.56000033e-01 3.29457581e-01 -8.22538674e-01 7.51499534e-01
7.37532079e-01 5.69460750e-01 -1.10387897e+00 1.29339051e+00
-2.15534523e-01 -3.05761788e-02 -7.71799088e-01 2.70233244e-01
-1.34404317e-01 -9.67679694e-02 1.58225030e-01 7.85441279e-01
4.31278139e-01 1.01550139e-01 1.98765442e-01 8.30342829e-01
3.34830470e-02 -2.70851254e-02 -6.84955895e-01 3.43600810e-01
4.68706369e-01 1.11388576e+00 -6.13322854e-01 -2.75840342e-01
-3.32666337e-01 1.03584254e+00 2.93768108e-01 2.53551990e-01
-6.63658082e-01 -2.63079345e-01 4.18400615e-01 2.22290009e-01
5.96307695e-01 -4.75648403e-01 -1.78353220e-01 -1.15914536e+00
5.08672118e-01 -8.88216197e-01 1.02122806e-01 -1.14018047e+00
-1.17018402e+00 6.86220288e-01 1.42182171e-01 -1.57746041e+00
-3.17910969e-01 -8.98077309e-01 -4.18053627e-01 7.79395640e-01
-1.14934659e+00 -1.37799942e+00 -7.62146533e-01 7.35606849e-01
2.06051186e-01 1.69438630e-01 1.00137174e+00 1.67741686e-01
-4.37447652e-02 4.37658340e-01 1.76831618e-01 9.65536684e-02
6.69920087e-01 -8.88159454e-01 5.31581581e-01 5.76248944e-01
3.87316287e-01 4.15482312e-01 6.10222757e-01 -4.99059975e-01
-1.72647119e+00 -6.83655322e-01 5.74534714e-01 -7.99999833e-01
3.21012706e-01 -7.93198705e-01 -8.69479060e-01 8.59144032e-01
-1.31869540e-01 4.49579865e-01 1.76255196e-01 -1.52676657e-01
-6.03772819e-01 -4.26216125e-01 -1.23447335e+00 4.01184052e-01
1.45266652e+00 -4.56202239e-01 -6.14319265e-01 4.46507871e-01
3.85771811e-01 -1.05170035e+00 -9.50006485e-01 7.34293520e-01
9.24540579e-01 -1.03329349e+00 1.19481468e+00 -4.69072491e-01
-2.52743453e-01 -4.61729676e-01 -3.10891926e-01 -9.30734217e-01
-1.26695320e-01 -4.47642952e-01 1.47113372e-02 1.04927492e+00
2.18943119e-01 -7.52729297e-01 1.01603067e+00 8.03273141e-01
-5.60801961e-02 -7.24034429e-01 -1.30916691e+00 -1.11721015e+00
-2.10115477e-01 -5.48535228e-01 8.52428317e-01 7.05714345e-01
-5.26839674e-01 -2.93400973e-01 -1.80703491e-01 2.74769574e-01
6.88511014e-01 5.43752730e-01 1.27644098e+00 -1.36265743e+00
-1.23193324e-01 -9.61868912e-02 -8.04714739e-01 -1.00129545e+00
-1.45813283e-02 -7.69620299e-01 3.59831713e-02 -1.42050040e+00
-2.93211099e-02 -6.90937996e-01 -1.72781590e-02 9.10685658e-01
2.39190310e-01 5.46784580e-01 2.35277086e-01 3.83194029e-01
-4.18910891e-01 4.33164835e-01 1.15715849e+00 -1.28807053e-01
-3.09242271e-02 -2.39472929e-02 -7.34331608e-02 5.54479063e-01
9.06805634e-01 -7.96855271e-01 -8.91229659e-02 -3.96781772e-01
-3.14429402e-04 -3.31352025e-01 4.98632818e-01 -1.24770916e+00
2.29327321e-01 -1.78045779e-01 6.32147789e-01 -9.27537501e-01
8.28069746e-01 -1.15791428e+00 7.42704690e-01 4.81065691e-01
1.89003229e-01 2.73748308e-01 3.70953023e-01 1.42664507e-01
1.84687063e-01 -3.81065935e-01 4.96952862e-01 -2.44485348e-01
-7.26034284e-01 5.44649780e-01 4.21556309e-02 -3.82867724e-01
1.02183568e+00 -8.02771568e-01 -4.19580936e-01 -2.39912882e-01
-7.70239532e-01 -1.92794770e-01 9.79363620e-01 6.22959256e-01
8.42874110e-01 -1.45046067e+00 -5.80190599e-01 5.54149985e-01
3.75337392e-01 2.96076685e-01 -1.50199041e-01 6.62594378e-01
-6.79202020e-01 3.74194115e-01 -3.69195640e-01 -9.41115916e-01
-1.30258596e+00 6.93597972e-01 4.90376860e-01 1.65181026e-01
-5.85663497e-01 5.59404194e-01 2.53215343e-01 -7.56480753e-01
2.88850099e-01 -3.53804678e-01 6.18911266e-01 -2.53706962e-01
1.01563469e-01 2.84708500e-01 4.06145036e-01 -7.50972629e-01
-5.81123233e-01 1.29363322e+00 8.03992152e-04 -7.89298490e-02
1.40293300e+00 4.93378900e-02 -8.25839564e-02 4.21020091e-01
1.29490745e+00 6.50580646e-03 -1.34508061e+00 -5.11790693e-01
1.17838368e-01 -9.73133624e-01 -3.36590350e-01 -7.55511701e-01
-1.05767715e+00 8.88836384e-01 1.02454150e+00 -4.02947366e-01
1.01506174e+00 2.98774093e-01 3.95177156e-01 6.44931793e-01
9.16072428e-01 -8.15661907e-01 2.22231299e-01 3.68791670e-01
1.36964977e+00 -1.05061984e+00 2.83481568e-01 -6.78233445e-01
-6.70520142e-02 1.43911183e+00 6.28384531e-01 -1.87464818e-01
4.45431292e-01 3.02916706e-01 3.61929268e-01 -2.68625647e-01
-3.74424130e-01 -2.51767188e-01 6.28872097e-01 9.59404528e-01
-2.96499338e-02 -2.03671947e-01 3.28785121e-01 -4.61897440e-02
-3.67236763e-01 -1.11929469e-01 2.83035547e-01 1.20995462e+00
-2.64332205e-01 -1.56159079e+00 -7.02480614e-01 1.62015125e-01
1.11328825e-01 6.76086172e-02 -6.68394208e-01 1.19447803e+00
2.59590805e-01 5.01983523e-01 8.41273516e-02 -4.36088353e-01
7.53921628e-01 9.84526053e-02 1.11269021e+00 -6.75096869e-01
-4.25052732e-01 -6.86452389e-02 -9.99732167e-02 -9.28417206e-01
-5.55349469e-01 -8.67252350e-01 -1.04327035e+00 -1.00050196e-01
-4.67989534e-01 -2.87858874e-01 7.99742341e-01 6.23214304e-01
3.95236731e-01 2.48601828e-02 6.09183609e-01 -1.58200645e+00
-2.90639937e-01 -6.94251835e-01 -3.21436495e-01 6.72935545e-01
1.77220002e-01 -9.93995667e-01 -2.52802402e-01 9.18705463e-02] | [7.578765392303467, -2.64186429977417] |
7af42749-1be4-4b75-9ae8-87373ec67381 | easy-adaptation-to-mitigate-gender-bias-in | 2204.05459 | null | https://arxiv.org/abs/2204.05459v1 | https://arxiv.org/pdf/2204.05459v1.pdf | Easy Adaptation to Mitigate Gender Bias in Multilingual Text Classification | Existing approaches to mitigate demographic biases evaluate on monolingual data, however, multilingual data has not been examined. In this work, we treat the gender as domains (e.g., male vs. female) and present a standard domain adaptation model to reduce the gender bias and improve performance of text classifiers under multilingual settings. We evaluate our approach on two text classification tasks, hate speech detection and rating prediction, and demonstrate the effectiveness of our approach with three fair-aware baselines. | ['Xiaolei Huang'] | 2022-04-12 | null | https://aclanthology.org/2022.naacl-main.52 | https://aclanthology.org/2022.naacl-main.52.pdf | naacl-2022-7 | ['multilingual-text-classification'] | ['miscellaneous'] | [-2.85733283e-01 -1.36875406e-01 -4.64682549e-01 -6.45230830e-01
-7.36609876e-01 -6.33047819e-01 8.46096098e-01 2.32589647e-01
-7.76233017e-01 9.50327516e-01 5.29496968e-01 -4.19858158e-01
6.46802068e-01 -4.10995036e-01 -3.91700596e-01 -2.96821713e-01
5.85299253e-01 3.69635552e-01 1.95734948e-02 -3.33705485e-01
4.05443400e-01 -5.00608087e-02 -1.24000061e+00 2.99920499e-01
1.44085479e+00 3.81622076e-01 -4.25765753e-01 4.15199131e-01
3.84970009e-02 6.30490184e-01 -1.22733748e+00 -1.28744411e+00
-2.29273699e-02 -8.59026089e-02 -6.73232138e-01 -2.88851231e-01
9.08604980e-01 -6.39899611e-01 -3.89730811e-01 1.04620504e+00
8.73849869e-01 -2.79180229e-01 1.09635890e+00 -1.26907849e+00
-1.26182520e+00 9.87416327e-01 -9.69413221e-01 3.40765595e-01
2.87705094e-01 -2.77220219e-01 8.18929851e-01 -1.03117812e+00
5.96103430e-01 1.77082813e+00 7.06610680e-01 7.41972566e-01
-1.43251479e+00 -1.16295612e+00 2.67037868e-01 6.41291067e-02
-1.30984282e+00 -5.58027506e-01 7.11461306e-01 -6.59353077e-01
6.62466228e-01 -1.64158985e-01 1.24038167e-01 2.11572456e+00
1.78619385e-01 9.61817145e-01 1.62869132e+00 -4.73167539e-01
-9.32410955e-02 7.64869213e-01 4.35794353e-01 2.49701455e-01
5.70304096e-01 -1.97258219e-01 -9.06342030e-01 -3.64543110e-01
7.19168643e-03 -3.50452840e-01 1.69142917e-01 -1.44470513e-01
-7.91669667e-01 1.28737605e+00 -9.49002132e-02 9.34446380e-02
2.13183284e-01 -4.40771431e-01 8.91802669e-01 1.37671873e-01
1.08645117e+00 1.83913171e-01 -3.73946011e-01 -6.50821030e-02
-9.85789061e-01 5.63905895e-01 8.25963378e-01 9.51327264e-01
3.05454731e-01 9.47169289e-02 -3.91428351e-01 1.50205219e+00
9.97791737e-02 1.12077057e+00 5.26336133e-01 -4.77379292e-01
7.64868498e-01 1.85686588e-01 1.27586080e-02 -8.91786456e-01
-2.87351131e-01 -8.23834464e-02 -4.64934886e-01 -2.81920373e-01
5.29620051e-01 -4.69450772e-01 -7.34055996e-01 1.68667662e+00
1.18025459e-01 -4.24311042e-01 -1.02877114e-02 5.52508116e-01
8.21179509e-01 4.31178957e-01 7.05770791e-01 7.94182047e-02
1.48187923e+00 -8.24859917e-01 -1.01781929e+00 -4.46653455e-01
8.90733898e-01 -8.98935080e-01 1.13643241e+00 4.13359940e-01
-8.30439806e-01 -1.65560186e-01 -9.64384556e-01 -3.77536029e-01
-7.55953372e-01 3.41763407e-01 4.40009832e-01 1.33660793e+00
-5.86776376e-01 1.76360741e-01 -3.57605070e-01 -6.90598011e-01
3.94837379e-01 6.59404248e-02 -3.45699996e-01 5.44326976e-02
-1.67675328e+00 1.15350735e+00 -1.21457791e-02 -5.74514270e-01
-5.89569330e-01 -7.71043897e-01 -8.44729424e-01 -3.52159113e-01
-2.58800685e-01 -3.05149972e-01 1.50775361e+00 -8.44973743e-01
-1.10136497e+00 1.38590217e+00 -3.26173306e-01 -4.99861002e-01
4.44991320e-01 -6.20848179e-01 -6.01016879e-01 -2.22298712e-01
4.79727894e-01 6.84695244e-01 1.00131845e+00 -1.20203984e+00
-6.70078397e-01 -6.92229331e-01 -1.85472935e-01 3.40422004e-01
-1.17382765e+00 5.51888049e-01 1.61379695e-01 -9.73394513e-01
-6.39526546e-01 -1.05533099e+00 4.14508492e-01 -4.58517164e-01
-4.87056464e-01 -3.52929980e-01 1.06044734e+00 -1.33545434e+00
1.58662295e+00 -2.12018728e+00 -2.61879176e-01 -1.37982503e-01
1.88459843e-01 3.43952894e-01 2.45806158e-01 2.65042484e-01
3.51703465e-02 4.38800335e-01 3.68889868e-02 -7.11393535e-01
2.17678115e-01 9.28182155e-02 -3.85880560e-01 4.21114326e-01
1.31973714e-01 4.47157145e-01 -5.24026811e-01 -8.34869921e-01
-1.81866705e-01 5.24216831e-01 -7.10992038e-01 2.01712444e-01
3.76883417e-01 2.19539866e-01 -5.08217625e-02 7.65193343e-01
1.02758110e+00 5.65269768e-01 2.25172445e-01 1.75874569e-02
-2.32059717e-01 6.55973852e-01 -5.62097371e-01 1.19296789e+00
-4.82644558e-01 8.45890403e-01 1.95411161e-01 -4.57299232e-01
1.04798532e+00 3.83144356e-02 -4.19080555e-02 -7.89496839e-01
2.10853845e-01 2.85804302e-01 5.41595742e-02 -4.01334524e-01
7.60874450e-01 -4.93836254e-02 -5.55175424e-01 3.45166624e-01
8.97016823e-02 2.53699511e-01 2.80334234e-01 4.14778084e-01
6.16187871e-01 1.47475287e-01 2.12976098e-01 -4.45760906e-01
3.24579537e-01 -7.28649870e-02 4.50258464e-01 7.89388061e-01
-8.21665645e-01 4.88286108e-01 6.66748643e-01 -3.60646956e-02
-1.21606743e+00 -1.02850914e+00 -6.33349597e-01 2.11234808e+00
-1.56039849e-01 -3.81154716e-01 -9.90950465e-01 -1.25770414e+00
4.01990145e-01 8.90527487e-01 -7.17341185e-01 -3.34116071e-01
-5.56295395e-01 -1.00542784e+00 1.09973013e+00 6.23590052e-01
2.87643671e-01 -5.80021977e-01 2.69003008e-02 -2.04631537e-01
-3.34237546e-01 -1.10326290e+00 -6.40402019e-01 2.00999305e-01
-4.79727894e-01 -5.81553876e-01 -7.69652128e-01 -8.78200173e-01
1.96547106e-01 1.62057176e-01 1.42811692e+00 -2.30498746e-01
4.96068858e-02 3.16760279e-02 -3.24192524e-01 -8.19027543e-01
-4.50565040e-01 5.72692096e-01 2.39919603e-01 -3.06184798e-01
1.09910011e+00 -2.91621864e-01 -4.40537393e-01 2.34806553e-01
-6.51040316e-01 -2.68284291e-01 4.52500910e-01 7.74704874e-01
-3.20036262e-01 -4.28562284e-01 8.28864753e-01 -1.47494972e+00
9.91122246e-01 -5.48994958e-01 -7.72310868e-02 9.86746699e-02
-6.93115294e-01 -1.54823706e-01 5.58266103e-01 -6.88952625e-01
-1.38171041e+00 -2.32685447e-01 -1.79599702e-01 -9.53296851e-03
-2.58626193e-01 1.87167212e-01 -2.39581019e-01 1.95046961e-01
8.35534751e-01 -2.92217910e-01 -3.92606765e-01 -6.04853272e-01
2.02769384e-01 1.42241228e+00 3.45690966e-01 -8.84769678e-01
8.16959202e-01 1.43319041e-01 -6.70383394e-01 -7.69701362e-01
-1.02202332e+00 -3.38353097e-01 -9.63659286e-01 1.39390985e-02
8.05692136e-01 -1.29200053e+00 -2.21274838e-01 5.70853949e-01
-1.14932024e+00 2.05804616e-01 6.05369747e-01 3.99586737e-01
1.96583197e-01 3.08924526e-01 -9.37540054e-01 -9.99083102e-01
-4.11841124e-01 -8.21721315e-01 1.19550455e+00 -8.22489485e-02
-5.64838886e-01 -1.13553464e+00 2.74861664e-01 5.38496971e-01
2.60350376e-01 9.22313184e-02 1.06571221e+00 -1.22058666e+00
5.91326177e-01 3.89784165e-02 -3.07826966e-01 3.06883752e-01
-2.76334941e-01 2.67033905e-01 -1.29131138e+00 -3.46765876e-01
-4.04230505e-01 -7.24746525e-01 8.65697324e-01 1.08528644e-01
9.83176410e-01 -3.76555115e-01 -3.88716340e-01 3.57589811e-01
9.04009044e-01 -3.94996023e-03 4.21001911e-01 3.79112810e-01
8.23956847e-01 9.23823059e-01 6.74845099e-01 7.19577730e-01
8.31482232e-01 5.45567989e-01 -1.32861778e-01 -1.01302855e-01
-4.86065336e-02 -7.17767239e-01 5.96779704e-01 6.97237968e-01
3.00976336e-01 -1.43694475e-01 -1.12370646e+00 5.49185336e-01
-1.56913745e+00 -8.74549806e-01 -2.27258444e-01 1.98267245e+00
1.10264695e+00 8.46628621e-02 5.54609597e-01 -1.74047455e-01
1.15105987e+00 1.97175130e-01 -4.72071856e-01 -9.13980603e-01
-2.89184570e-01 1.16015919e-01 6.93576396e-01 2.68500805e-01
-1.59029460e+00 1.18143678e+00 7.24230003e+00 7.46895790e-01
-9.91183281e-01 2.94867247e-01 6.88663304e-01 -1.79187894e-01
-5.86906774e-03 -3.94893378e-01 -1.26000094e+00 4.91781980e-01
1.09619772e+00 -1.18635252e-01 1.33140370e-01 1.08352852e+00
-1.07184626e-01 1.13842741e-01 -9.85179424e-01 1.00323784e+00
5.06746590e-01 -2.25866735e-01 2.66812630e-02 2.09808573e-01
7.22248018e-01 -3.14098537e-01 5.98320901e-01 9.35562730e-01
5.93811095e-01 -9.37987208e-01 8.96055043e-01 -2.65905261e-01
6.41633272e-01 -1.07594168e+00 6.41079128e-01 1.64897993e-01
-4.74722713e-01 -2.68927187e-01 -4.35500830e-01 6.75305501e-02
-1.45250931e-01 3.70604724e-01 -7.72695422e-01 8.20346698e-02
1.00781727e+00 8.45310032e-01 -1.09288073e+00 3.30548048e-01
-2.12266222e-01 8.84917736e-01 1.35610074e-01 -3.09459632e-03
-2.61323094e-01 5.85372001e-02 3.10734838e-01 1.68179440e+00
1.52427390e-01 -5.16252100e-01 1.19186483e-01 5.34826577e-01
-4.06376362e-01 5.51071942e-01 -7.66870201e-01 -1.59264937e-01
9.43210483e-01 1.23110938e+00 -6.86104372e-02 -1.49479866e-01
-8.58381569e-01 8.56413901e-01 6.15204871e-01 2.73797750e-01
-9.66132164e-01 -5.68406940e-01 6.87889278e-01 2.28301242e-01
-3.35601903e-02 -1.16858952e-01 -7.43537068e-01 -1.36269236e+00
-2.89102674e-01 -1.43005586e+00 6.44507885e-01 -3.92722428e-01
-1.77667415e+00 2.18801260e-01 9.44708064e-02 -7.95910120e-01
-2.47747287e-01 -8.24014366e-01 -2.87965834e-01 9.44082856e-01
-1.35950732e+00 -1.48469639e+00 9.20507163e-02 6.24684513e-01
2.54320323e-01 -3.16239119e-01 6.10923111e-01 5.68453968e-01
-8.86638701e-01 1.24199557e+00 2.53536701e-01 6.09216034e-01
1.91342509e+00 -1.56159186e+00 2.71312624e-01 6.42247856e-01
-3.76131415e-01 9.80258882e-01 8.02582622e-01 -8.38029444e-01
-7.21340418e-01 -9.59785819e-01 1.30777478e+00 -8.88254881e-01
8.95473123e-01 -1.04890251e+00 -8.39761257e-01 7.73683608e-01
7.31334209e-01 -6.43567860e-01 1.13598621e+00 8.13060403e-01
-1.09889436e+00 5.71560338e-02 -1.25603414e+00 6.39743745e-01
8.68522406e-01 -6.70766532e-01 -6.57080293e-01 8.79979655e-02
4.82422620e-01 -2.64577508e-01 -9.83643472e-01 2.51123667e-01
7.60196507e-01 -8.96261811e-01 7.80018866e-01 -6.28292561e-01
8.25523078e-01 3.71806264e-01 3.97282243e-02 -1.57192314e+00
-2.34170109e-01 -1.60992280e-01 -7.83141032e-02 1.87255824e+00
4.92884696e-01 -5.31284571e-01 5.54368019e-01 6.15031958e-01
1.14035860e-01 -1.44142866e-01 -6.64017320e-01 -5.56223094e-01
1.05853450e+00 2.79272854e-01 5.18685699e-01 1.31641829e+00
2.20027924e-01 9.08908486e-01 -7.34901726e-01 -2.23846678e-02
5.69557965e-01 -2.52722651e-01 8.04619491e-01 -1.46825254e+00
1.60252377e-01 -3.45434874e-01 1.32127836e-01 -6.26103818e-01
8.85059595e-01 -7.97055006e-01 -2.11189061e-01 -9.91835415e-01
6.69539213e-01 -5.53168245e-02 6.12728558e-02 4.13978040e-01
-6.21578455e-01 3.05362105e-01 4.76954058e-02 1.46031529e-01
-4.85825121e-01 4.80469972e-01 6.40719354e-01 -2.84492493e-01
1.11003183e-01 -3.01644176e-01 -1.04456210e+00 5.93803883e-01
9.16306376e-01 -6.23735130e-01 -1.34532362e-01 -4.71355528e-01
1.63716804e-02 -5.13103366e-01 -2.97310725e-02 -6.49970055e-01
-2.91692555e-01 2.69026812e-02 7.35540867e-01 -3.89508903e-01
2.06904531e-01 -5.24256349e-01 -8.87878895e-01 1.98717549e-01
-6.81928456e-01 3.63317281e-01 1.02575772e-01 2.23264456e-01
-1.21773638e-01 -2.52023578e-01 9.57127213e-01 3.85484368e-01
-1.20787807e-01 8.97293687e-02 -5.87933898e-01 5.53148448e-01
6.95988655e-01 1.93085089e-01 -1.01003993e+00 -4.36240911e-01
-4.32415336e-01 1.30926803e-01 5.72111607e-01 8.18818450e-01
1.63471445e-01 -1.42358184e+00 -9.50395048e-01 -3.10696047e-02
5.40514112e-01 -1.02893865e+00 -1.06166005e-01 6.09885633e-01
-1.30959600e-01 5.40727675e-01 -2.72289395e-01 -2.69975930e-01
-1.77788222e+00 5.74898124e-01 1.46439433e-01 -3.39912653e-01
1.35124415e-01 5.14427841e-01 3.13515544e-01 -8.12309921e-01
4.03771579e-01 3.39212716e-01 -6.78270161e-01 3.95436764e-01
5.51667273e-01 5.55628002e-01 -3.62503119e-02 -9.39956605e-01
-3.86657536e-01 7.80699495e-03 -5.87764382e-01 -2.75427610e-01
8.39356244e-01 -2.67257780e-01 -1.45345867e-01 5.46632051e-01
1.15012288e+00 5.52749038e-01 -7.27849066e-01 -2.26603344e-01
2.08879814e-01 -6.72720730e-01 -1.33100465e-01 -9.13393378e-01
-5.91952205e-01 1.08004665e+00 6.12197220e-01 9.87117440e-02
7.14600682e-01 -1.92607954e-01 8.50192130e-01 4.77939174e-02
4.95811068e-02 -1.70953882e+00 1.52176488e-02 9.01035249e-01
4.31463420e-01 -1.53982592e+00 1.39980838e-02 -4.92274284e-01
-1.11680865e+00 6.78537250e-01 1.23077393e+00 2.69714832e-01
5.16112983e-01 2.67492205e-01 4.62276608e-01 3.64231676e-01
-7.04088330e-01 -1.03941806e-01 2.36317545e-01 1.09683108e+00
1.25655210e+00 9.91472378e-02 -8.23473811e-01 8.61218870e-01
-5.15898168e-01 -6.05808854e-01 4.52929586e-01 8.13215435e-01
-2.07522288e-01 -1.27244806e+00 -7.08925128e-01 4.59827274e-01
-1.18958426e+00 -4.31827128e-01 -1.00033939e+00 8.65245581e-01
1.61102369e-01 1.07778311e+00 -2.35190317e-02 -5.13008773e-01
3.88038516e-01 3.49503428e-01 3.38298768e-01 -6.14388466e-01
-9.27745461e-01 -1.51055353e-02 4.46883708e-01 9.76623520e-02
-2.59092271e-01 -9.67325628e-01 -5.92897415e-01 -7.18559325e-01
2.11005490e-02 5.09136841e-02 5.73459387e-01 5.95372260e-01
1.84810475e-01 -9.09029916e-02 7.20367610e-01 -4.28374529e-01
-7.68892646e-01 -1.26284790e+00 -7.14826226e-01 9.34247136e-01
1.20102435e-01 -6.89833701e-01 -4.31438893e-01 2.59204339e-02] | [9.080235481262207, 10.471344947814941] |
d252a221-01a1-40ac-8e3a-682197ae7438 | guided-filter-based-edge-preserving-image-non | 1609.01839 | null | http://arxiv.org/abs/1609.01839v1 | http://arxiv.org/pdf/1609.01839v1.pdf | Guided Filter based Edge-preserving Image Non-blind Deconvolution | In this work, we propose a new approach for efficient edge-preserving image
deconvolution. Our algorithm is based on a novel type of explicit image filter
- guided filter. The guided filter can be used as an edge-preserving smoothing
operator like the popular bilateral filter, but has better behaviors near
edges. We propose an efficient iterative algorithm with the decouple of
deblurring and denoising steps in the restoration process. In deblurring step,
we proposed two cost function which could be computed with fast Fourier
transform efficiently. The solution of the first one is used as the guidance
image, and another solution will be filtered in next step. In the denoising
step, the guided filter is used with the two obtained images for efficient
edge-preserving filtering. Furthermore, we derive a simple and effective method
to automatically adjust the regularization parameter at each iteration. We
compare our deconvolution algorithm with many competitive deconvolution
techniques in terms of ISNR and visual quality. | ['He-Yan Huang', 'Hang Yang', 'Ming Zhu', 'Zhongbo Zhang'] | 2016-09-07 | null | null | null | null | ['image-deconvolution'] | ['computer-vision'] | [ 1.85385287e-01 -3.42617542e-01 5.08739889e-01 -1.45205125e-01
-2.85973489e-01 -3.87023687e-01 2.62442201e-01 -4.56372976e-01
-6.56233549e-01 6.16324365e-01 4.90023673e-01 -2.55310237e-02
-1.21586874e-01 -7.15476036e-01 -4.68152642e-01 -9.99404073e-01
3.27358902e-01 -1.12812474e-01 5.73474705e-01 -1.29030511e-01
4.98310924e-01 4.00717437e-01 -1.31924152e+00 1.14497408e-01
1.23905444e+00 8.17755342e-01 6.43170655e-01 5.04958153e-01
-1.82784170e-01 5.47655523e-01 1.73469319e-03 -5.71476258e-02
3.21375072e-01 -7.87154913e-01 -6.96825445e-01 2.97556490e-01
4.51978594e-02 -6.14699125e-01 -3.93947572e-01 1.35562134e+00
7.06349373e-01 3.02327245e-01 5.89010060e-01 -6.76650107e-01
-5.60941696e-01 2.81286478e-01 -7.89702058e-01 2.80132771e-01
1.33793011e-01 -9.86425132e-02 3.30619395e-01 -1.08264053e+00
6.91648304e-01 1.12695563e+00 7.89163589e-01 4.40672934e-01
-1.05388188e+00 -4.43430662e-01 -2.15213791e-01 4.78932291e-01
-1.28457999e+00 -5.85643291e-01 8.72969747e-01 -3.64490688e-01
2.05924705e-01 2.43996575e-01 5.65552413e-01 3.40661585e-01
9.84491631e-02 6.73217893e-01 1.27924502e+00 -5.60876250e-01
8.93562958e-02 -2.27218971e-01 3.54591250e-01 5.85490465e-01
1.47133604e-01 1.94396526e-01 -2.62404680e-01 -2.66080856e-01
1.15022469e+00 6.74841180e-02 -1.03607142e+00 2.96521094e-02
-1.14288557e+00 5.82751572e-01 3.40161324e-01 4.47519690e-01
-5.42186499e-01 2.22789645e-01 2.16730133e-01 3.51624101e-01
6.66091323e-01 -2.39523244e-03 -1.46841019e-01 2.03917712e-01
-1.09956348e+00 1.00511074e-01 5.55798173e-01 3.55294734e-01
8.52350116e-01 -2.22959563e-01 -4.53638017e-01 9.75148022e-01
4.03048396e-01 2.03152627e-01 4.89875883e-01 -1.08144140e+00
-5.26198186e-02 1.12422660e-01 5.45450330e-01 -1.04199231e+00
-2.01150566e-01 -3.72732013e-01 -1.08087564e+00 4.08199936e-01
4.62907493e-01 -1.33965030e-01 -7.89546251e-01 1.30260170e+00
3.34607273e-01 7.29409754e-01 -2.17839435e-01 1.30246282e+00
6.93253756e-01 7.32655942e-01 -3.59844863e-01 -5.41357160e-01
1.47678983e+00 -1.10823047e+00 -1.25267184e+00 7.51380846e-02
2.51564205e-01 -1.14693570e+00 5.59034169e-01 4.24883962e-01
-1.44006443e+00 -3.68990242e-01 -8.64987910e-01 -2.25393474e-01
1.49794653e-01 3.96309346e-01 3.28239918e-01 3.89473379e-01
-1.27685559e+00 9.59490716e-01 -6.34056270e-01 -2.29044095e-01
7.22852051e-02 3.47207976e-03 -3.40988308e-01 -8.93307105e-02
-9.51416731e-01 8.71225834e-01 2.65455365e-01 3.89250726e-01
-6.09013975e-01 -4.96024966e-01 -4.43763942e-01 1.15537107e-01
1.06852457e-01 -8.26637089e-01 8.59198272e-01 -9.25655723e-01
-1.60242605e+00 8.00399363e-01 -4.81029451e-01 -1.41579524e-01
8.14999759e-01 -3.69880140e-01 -2.59490818e-01 2.86992729e-01
-1.07713975e-01 1.15973458e-01 1.29772341e+00 -1.34254503e+00
-4.01001751e-01 -1.20441020e-01 -3.96638423e-01 2.17853338e-01
-2.06972599e-01 3.15780751e-02 -6.74491286e-01 -1.06388664e+00
6.44988537e-01 -5.43005407e-01 -1.89075768e-01 4.28342462e-01
-1.58796296e-01 1.81874424e-01 7.93803036e-01 -1.28808963e+00
1.35370195e+00 -2.46201682e+00 2.87770212e-01 7.90082812e-02
3.28525484e-01 2.05332473e-01 -7.03619272e-02 2.35592648e-01
-1.47662312e-01 -1.09619945e-01 -5.94569266e-01 -2.40135208e-01
-4.60053056e-01 -6.96049258e-02 -1.77369088e-01 7.93457210e-01
-2.46867269e-01 5.71967483e-01 -7.79243350e-01 -2.85672128e-01
8.03178996e-02 7.13513553e-01 -4.79586154e-01 4.15544122e-01
3.77121657e-01 7.06676602e-01 -3.10305148e-01 1.56024218e-01
1.40329194e+00 -3.20890881e-02 -1.86508507e-01 -7.24456370e-01
-5.50762177e-01 -2.46403649e-01 -1.51822722e+00 1.64913619e+00
-2.35962331e-01 4.93883401e-01 8.45670640e-01 -9.53435481e-01
9.21365082e-01 5.17300785e-01 3.37314844e-01 -2.57333428e-01
2.77983725e-01 4.93365586e-01 -3.42602313e-01 -7.39356160e-01
4.87808138e-01 -3.21871549e-01 9.38999891e-01 5.21196783e-01
-1.68495014e-01 3.92359681e-02 1.21598423e-01 -1.05965734e-01
8.24468851e-01 1.56232908e-01 8.16357285e-02 -7.41443276e-01
1.04563069e+00 -2.53526390e-01 5.34911156e-01 5.69782197e-01
-6.67462051e-02 9.94111001e-01 -5.76146552e-03 -4.11496371e-01
-9.36703205e-01 -7.29373038e-01 -3.31172347e-01 5.81474721e-01
6.28259540e-01 -1.15405813e-01 -8.70412529e-01 -2.56833911e-01
-2.07994282e-01 3.45670402e-01 -3.23828906e-01 -1.41135799e-02
-5.78034520e-01 -8.31943393e-01 1.41134277e-01 1.83446050e-01
9.74462092e-01 -6.61417723e-01 -1.22414090e-01 3.25469285e-01
-4.99130249e-01 -6.92636251e-01 -1.04677844e+00 -1.50038868e-01
-1.21952784e+00 -1.07878256e+00 -1.42011690e+00 -1.03571784e+00
9.66307700e-01 8.04960191e-01 5.75202405e-01 4.68667895e-01
-2.94113457e-02 1.27456650e-01 -3.47188145e-01 3.22149515e-01
-3.32401395e-01 -6.74173892e-01 -2.37047687e-01 5.35490274e-01
-1.82071418e-01 -7.30200648e-01 -1.14438236e+00 5.94182193e-01
-9.57089245e-01 1.55728549e-01 4.83854473e-01 9.38520789e-01
5.25761008e-01 3.21344644e-01 1.66105524e-01 -5.77788234e-01
9.59082961e-01 -9.27396715e-02 -5.58887601e-01 7.97839910e-02
-4.35554624e-01 2.00920939e-01 5.77923179e-01 -3.57643783e-01
-1.42154264e+00 -2.52408050e-02 -2.56527990e-01 -3.40153456e-01
1.78897470e-01 4.23279047e-01 1.23694807e-01 -5.53538799e-01
5.81689358e-01 5.79612195e-01 2.41384685e-01 -1.09567976e+00
3.33085001e-01 6.37958467e-01 8.11236918e-01 -3.64849508e-01
8.18155825e-01 7.52282977e-01 -1.56784192e-01 -8.28293145e-01
-2.89779782e-01 -6.43881083e-01 -4.48494494e-01 -1.99777588e-01
8.23006988e-01 -8.02037001e-01 -7.45395243e-01 9.86589074e-01
-1.49560809e+00 -2.87348665e-02 1.14217363e-01 7.12051392e-01
-3.84908885e-01 1.02384126e+00 -8.59213710e-01 -7.05347121e-01
-4.45460290e-01 -1.11687732e+00 5.92213392e-01 2.97944576e-01
3.89477283e-01 -1.01772082e+00 1.10331684e-01 6.65339381e-02
7.04595327e-01 -1.29493386e-01 4.75144356e-01 1.77646184e-03
-5.21287203e-01 1.65676791e-02 -6.66977465e-01 5.62608480e-01
1.76012993e-01 -3.75611395e-01 -8.39781821e-01 -4.96889383e-01
5.92995524e-01 2.99900383e-01 1.23568058e+00 7.05363333e-01
1.06238532e+00 -1.88263446e-01 -2.80178666e-01 9.84353542e-01
1.58532763e+00 -3.78514966e-03 1.07468021e+00 3.75505805e-01
4.40234393e-01 3.37848395e-01 3.72480690e-01 2.54617512e-01
-2.39200182e-02 5.45152962e-01 1.08784899e-01 -3.03878099e-01
-4.33082610e-01 1.45735711e-01 2.41874978e-01 1.06454027e+00
-5.30046225e-01 1.34280222e-02 -4.50882852e-01 5.71896195e-01
-2.07562280e+00 -9.31688011e-01 -6.35211051e-01 2.41924763e+00
8.11655343e-01 -3.57866794e-01 -3.10158312e-01 6.11133389e-02
1.09674001e+00 2.20787264e-02 -1.10175520e-01 6.39385879e-02
-1.56140238e-01 3.78653244e-03 5.74594975e-01 9.64072108e-01
-9.86519277e-01 5.11964321e-01 6.74268246e+00 9.18575227e-01
-1.01522768e+00 3.31395924e-01 2.68921584e-01 4.35000151e-01
-3.55926841e-01 1.25228867e-01 -3.41177404e-01 5.42656362e-01
2.05625951e-01 -2.16580421e-01 6.83245480e-01 3.58057827e-01
8.05027068e-01 -2.83303529e-01 -5.94541430e-01 1.18383229e+00
-2.03130528e-01 -1.30078387e+00 -1.59562528e-01 -1.83176234e-01
5.99021912e-01 -4.06740189e-01 -3.36053282e-01 -5.48967540e-01
-1.33426800e-01 -7.23324418e-01 5.34497082e-01 8.48390639e-01
4.51546848e-01 -5.53979576e-01 7.77168870e-01 3.53785694e-01
-1.19639301e+00 1.75892159e-01 -4.09210294e-01 4.80534183e-03
4.42823529e-01 1.05670083e+00 1.67651281e-01 7.26022363e-01
6.24944866e-01 8.53169620e-01 5.07831462e-02 1.62134802e+00
-3.50161076e-01 4.18183744e-01 -1.39870495e-01 6.08732343e-01
5.45969512e-03 -9.23002720e-01 8.63075197e-01 1.21432745e+00
7.08716571e-01 5.78416169e-01 -4.02599461e-02 8.25674891e-01
1.10954002e-01 2.05747575e-01 -1.45684198e-01 3.16437781e-01
2.59137869e-01 1.31033182e+00 -6.67450726e-01 -3.32356960e-01
-4.20162231e-01 1.57834101e+00 -2.08579734e-01 5.80589473e-01
-6.23874545e-01 -4.63569522e-01 3.20819438e-01 2.20181063e-01
3.57213169e-01 -2.00678989e-01 -2.21939102e-01 -1.46263838e+00
-7.56558357e-03 -4.88934308e-01 1.53706208e-01 -9.88194466e-01
-1.28648090e+00 4.58792150e-01 -3.90400022e-01 -1.29350150e+00
4.42700773e-01 -4.98062044e-01 -8.90289605e-01 1.33715510e+00
-1.85026610e+00 -7.49322653e-01 -6.11641049e-01 8.07015657e-01
2.96260178e-01 2.43091926e-01 3.76176536e-01 6.66293979e-01
-3.29744160e-01 1.22053429e-01 5.42425036e-01 -6.60079494e-02
9.71224844e-01 -8.88992727e-01 -5.75889088e-02 1.09621310e+00
-6.22283936e-01 6.38468862e-01 8.96891356e-01 -7.69287348e-01
-1.05866516e+00 -7.23756552e-01 7.68060744e-01 5.37384510e-01
4.79169160e-01 2.25059390e-01 -1.22262812e+00 4.63920563e-01
3.06975335e-01 4.25183438e-02 2.54070640e-01 -5.25247991e-01
4.83714193e-02 3.86671051e-02 -1.34797084e+00 4.61829543e-01
8.37374389e-01 -2.65388906e-01 -5.66902757e-01 2.34455734e-01
4.61202800e-01 -3.83245677e-01 -6.93177700e-01 2.09986448e-01
4.74497974e-01 -1.09974504e+00 1.11691415e+00 1.15012050e-01
3.92169476e-01 -8.24256361e-01 2.76663572e-01 -1.38887858e+00
-9.03656840e-01 -9.68397975e-01 1.00395694e-01 1.05035818e+00
-9.35258642e-02 -8.76080453e-01 4.20228779e-01 1.22147612e-01
-1.72026262e-01 -3.00883353e-01 -7.36876726e-01 -6.58636749e-01
-3.10240358e-01 1.01409167e-01 1.50979742e-01 8.85504186e-01
-1.80533439e-01 -1.34204403e-01 -7.20019460e-01 2.98904419e-01
1.04273796e+00 -2.41657514e-02 2.56131858e-01 -1.16643143e+00
-2.70267069e-01 -3.83298397e-01 -9.07733366e-02 -1.53150749e+00
-2.86153316e-01 -7.48759747e-01 1.73846245e-01 -1.72691357e+00
3.89804572e-01 -7.22081959e-02 -1.98496282e-01 1.37625605e-01
-3.52917880e-01 2.77075231e-01 -3.22998092e-02 5.12570977e-01
1.81208134e-01 5.35427690e-01 1.76116216e+00 1.15032410e-02
-3.14003021e-01 -1.07611299e-01 -4.17650044e-01 7.87520051e-01
4.28612024e-01 -3.07696700e-01 -1.87735498e-01 -5.80484152e-01
-1.85521871e-01 3.00313551e-02 4.58853185e-01 -7.08245039e-01
5.07609785e-01 1.70209762e-02 1.51716501e-01 -4.55612034e-01
1.45749912e-01 -8.72886240e-01 3.44910830e-01 5.75949073e-01
5.00520542e-02 -4.02034223e-01 -8.42027541e-04 5.33569157e-01
-3.46362859e-01 -5.90249836e-01 1.13804638e+00 -3.04978073e-01
-6.69148982e-01 1.30968109e-01 -4.00077701e-01 -3.83675337e-01
6.66180491e-01 -3.73357624e-01 -2.74306059e-01 -4.23977733e-01
-9.85468566e-01 4.83054966e-02 6.06936395e-01 -7.79160261e-02
7.39877701e-01 -1.31702185e+00 -1.01946735e+00 5.07569276e-02
-4.62681204e-01 -3.24500680e-01 4.80061352e-01 1.36121190e+00
-9.55820501e-01 -1.45841390e-01 -3.46569538e-01 -3.22801858e-01
-1.14408231e+00 4.78481829e-01 5.48459649e-01 -8.73425603e-02
-9.11228001e-01 7.78013706e-01 3.62988502e-01 5.29383030e-03
1.23002985e-02 3.29434685e-02 -2.01457843e-01 -5.48561327e-02
9.56076503e-01 8.20896029e-01 -8.91905352e-02 -6.28233373e-01
-2.26857558e-01 1.02186775e+00 1.60787582e-01 -2.28390843e-01
1.45575690e+00 -6.57885551e-01 -9.13503408e-01 -7.07292110e-02
1.20090866e+00 2.19774321e-01 -1.22227597e+00 -1.48047447e-01
-2.63042659e-01 -7.86683738e-01 6.68712556e-01 -5.96333683e-01
-1.32885826e+00 6.55939937e-01 8.21271122e-01 1.14248760e-01
1.43941629e+00 -4.70141202e-01 1.00601065e+00 -1.50972381e-01
1.45664722e-01 -8.46584082e-01 -2.44704559e-01 3.81213307e-01
1.02958786e+00 -9.10728514e-01 1.48452118e-01 -7.87434995e-01
3.26928906e-02 1.39970338e+00 2.22244766e-02 -3.83168250e-01
7.48691857e-01 2.22949073e-01 1.79616008e-02 -5.69964200e-02
-8.91435817e-02 -2.09637612e-01 4.53398854e-01 3.70589495e-01
5.50413072e-01 -3.30111146e-01 -1.31591403e+00 4.32605952e-01
5.18791616e-01 4.45684761e-01 5.90938926e-01 5.81451595e-01
-9.22557354e-01 -1.04364443e+00 -8.72892857e-01 3.41178700e-02
-3.98844481e-01 -2.22558558e-01 1.23260960e-01 9.49115232e-02
6.74722791e-02 9.93962526e-01 -1.62929207e-01 -6.33843169e-02
1.69098094e-01 -1.39964759e-01 4.89116609e-01 -7.83250853e-02
-4.07027632e-01 5.19804716e-01 -6.75599724e-02 -5.82640648e-01
-5.66369355e-01 -3.26929420e-01 -1.21842086e+00 -3.87711138e-01
-5.22133291e-01 2.45620385e-01 5.04487634e-01 8.68528605e-01
2.68678784e-01 2.99558550e-01 6.29196644e-01 -1.09056425e+00
-2.44175091e-01 -8.27383399e-01 -9.47766244e-01 5.62150121e-01
5.31848311e-01 -5.18904090e-01 -7.50297964e-01 3.56775582e-01] | [11.56205940246582, -2.6783578395843506] |
643adc8e-c8bb-4120-8094-495c886550f2 | go-explore-a-new-approach-for-hard | 1901.10995 | null | https://arxiv.org/abs/1901.10995v4 | https://arxiv.org/pdf/1901.10995v4.pdf | Go-Explore: a New Approach for Hard-Exploration Problems | A grand challenge in reinforcement learning is intelligent exploration, especially when rewards are sparse or deceptive. Two Atari games serve as benchmarks for such hard-exploration domains: Montezuma's Revenge and Pitfall. On both games, current RL algorithms perform poorly, even those with intrinsic motivation, which is the dominant method to improve performance on hard-exploration domains. To address this shortfall, we introduce a new algorithm called Go-Explore. It exploits the following principles: (1) remember previously visited states, (2) first return to a promising state (without exploration), then explore from it, and (3) solve simulated environments through any available means (including by introducing determinism), then robustify via imitation learning. The combined effect of these principles is a dramatic performance improvement on hard-exploration problems. On Montezuma's Revenge, Go-Explore scores a mean of over 43k points, almost 4 times the previous state of the art. Go-Explore can also harness human-provided domain knowledge and, when augmented with it, scores a mean of over 650k points on Montezuma's Revenge. Its max performance of nearly 18 million surpasses the human world record, meeting even the strictest definition of "superhuman" performance. On Pitfall, Go-Explore with domain knowledge is the first algorithm to score above zero. Its mean score of almost 60k points exceeds expert human performance. Because Go-Explore produces high-performing demonstrations automatically and cheaply, it also outperforms imitation learning work where humans provide solution demonstrations. Go-Explore opens up many new research directions into improving it and weaving its insights into current RL algorithms. It may also enable progress on previously unsolvable hard-exploration problems in many domains, especially those that harness a simulator during training (e.g. robotics). | ['Kenneth O. Stanley', 'Joel Lehman', 'Adrien Ecoffet', 'Joost Huizinga', 'Jeff Clune'] | 2019-01-30 | null | null | null | null | ['montezumas-revenge'] | ['playing-games'] | [-1.52377337e-01 4.48025674e-01 -4.32701856e-01 2.44003296e-01
-7.69936442e-01 -9.14040387e-01 7.57972240e-01 -4.48497593e-01
-7.51246333e-01 1.35480654e+00 -7.59364441e-02 -4.75983977e-01
-2.80708641e-01 -4.93557423e-01 -9.16363180e-01 -6.50090575e-01
-5.09214759e-01 6.06764853e-01 -3.32181789e-02 -6.82674348e-01
4.24663305e-01 1.19077608e-01 -1.45138156e+00 -4.25876856e-01
1.02518475e+00 5.22887707e-01 3.17819417e-01 7.85197794e-01
4.12540317e-01 8.30053151e-01 -5.50403535e-01 4.74240072e-02
4.92324799e-01 -5.65705478e-01 -1.15608430e+00 -4.04271871e-01
-2.39845738e-01 -7.92738199e-01 -3.44469160e-01 9.32015002e-01
4.50090945e-01 3.61699224e-01 2.39877835e-01 -1.52428746e+00
-4.65021342e-01 8.68707538e-01 -4.95671839e-01 2.13212557e-02
6.64322019e-01 9.18274224e-01 8.68752539e-01 -1.85351059e-01
9.00875032e-01 1.21176827e+00 3.26631188e-01 1.04130280e+00
-1.21768856e+00 -8.59684050e-01 5.09620532e-02 1.62883282e-01
-9.13733721e-01 -3.01450431e-01 4.03723121e-01 -1.87841326e-01
1.15098548e+00 1.50034636e-01 8.45096529e-01 1.50151157e+00
1.72797039e-01 1.34486401e+00 1.53125894e+00 -9.97838005e-02
7.13530600e-01 -2.06438527e-01 -4.13774192e-01 7.07995474e-01
9.91102234e-02 1.11731017e+00 -7.57746041e-01 -1.53289989e-01
7.49387562e-01 -2.99276650e-01 -2.61468232e-01 -7.47613788e-01
-1.52619708e+00 8.44862998e-01 4.46187407e-01 4.32020519e-03
-4.19052720e-01 6.34558439e-01 3.08003753e-01 7.83702314e-01
-4.48582411e-01 1.38234901e+00 -3.95590067e-01 -9.93574381e-01
-5.74419260e-01 7.76101887e-01 1.02496254e+00 8.59558105e-01
7.96862185e-01 1.88268885e-01 3.91222268e-01 2.19854206e-01
-6.67926744e-02 5.17779589e-01 7.43932366e-01 -1.54099226e+00
3.86660844e-01 1.77516222e-01 6.15413249e-01 -4.70017791e-01
-5.85658848e-01 -3.40971649e-01 -2.80547976e-01 1.09336019e+00
4.28250104e-01 -5.09943485e-01 -7.03285694e-01 2.07463908e+00
2.53941655e-01 -1.28332311e-02 4.28884119e-01 1.14886069e+00
2.87756026e-01 5.42723000e-01 -1.62223920e-01 -3.97123629e-03
9.25556660e-01 -1.21113658e+00 -2.35057756e-01 -3.96927565e-01
8.35090876e-01 -4.52590548e-03 1.31770015e+00 8.45334351e-01
-1.15578091e+00 -1.92969620e-01 -1.28048301e+00 3.79350513e-01
-2.16626391e-01 -5.19676924e-01 1.00558984e+00 5.55459797e-01
-1.00458801e+00 9.92288411e-01 -9.94976044e-01 -3.27132940e-01
3.63811493e-01 4.47005302e-01 -4.28658664e-01 6.25545159e-02
-1.09280944e+00 1.17788386e+00 4.90242869e-01 -2.73544282e-01
-1.85569596e+00 -2.95949191e-01 -8.12099040e-01 -1.41226903e-01
9.78946865e-01 -5.51780224e-01 1.55797827e+00 -8.81875515e-01
-1.99272573e+00 5.68670034e-01 3.97984475e-01 -7.67062843e-01
8.80532086e-01 -4.73595798e-01 -4.50941660e-02 5.72697725e-04
3.07757497e-01 1.02926993e+00 7.12481558e-01 -1.13986421e+00
-5.76813042e-01 -1.71272140e-02 3.70565265e-01 5.03307760e-01
1.23371355e-01 -4.44163173e-01 9.15941447e-02 -2.60754257e-01
-2.82163233e-01 -1.26041222e+00 -4.70671594e-01 -2.11548328e-01
-2.54310906e-01 -3.26605320e-01 6.33321226e-01 -1.94528535e-01
6.15611494e-01 -1.96227252e+00 4.80918974e-01 -1.24730557e-01
2.83404887e-01 2.13244706e-01 -4.78656709e-01 5.12588322e-01
2.02257767e-01 2.39270274e-02 -2.00659677e-01 8.64337906e-02
4.23123300e-01 4.80723858e-01 -4.19574648e-01 3.22208881e-01
-9.01229009e-02 1.34805167e+00 -1.61963046e+00 -8.34205821e-02
1.37161762e-01 -2.50204265e-01 -6.65536880e-01 2.46858478e-01
-4.41309929e-01 7.02320814e-01 -5.44143617e-01 5.88169158e-01
2.18873978e-01 -1.81024686e-01 1.57763630e-01 8.94848764e-01
-9.42261666e-02 2.47516498e-01 -1.12484384e+00 2.06457424e+00
-2.78377891e-01 5.59435844e-01 1.20742351e-01 -7.50798762e-01
6.81363344e-01 5.66269495e-02 2.30671555e-01 -8.66283536e-01
-4.59334925e-02 4.02489483e-01 1.06450751e-01 -4.09727514e-01
6.34341955e-01 -8.11872706e-02 -3.37298870e-01 8.71381521e-01
-5.58196381e-02 -6.58749104e-01 7.77730793e-02 2.93632120e-01
1.58855867e+00 7.52444208e-01 3.33043635e-01 -2.43005022e-01
1.72037594e-02 3.41867894e-01 5.38978338e-01 1.33743262e+00
-5.83060026e-01 1.29760485e-02 7.66869426e-01 -5.42510211e-01
-9.61230516e-01 -1.20878279e+00 3.52477789e-01 1.22377026e+00
6.01662159e-01 -2.53711045e-01 -7.79247522e-01 -9.44867015e-01
1.90616831e-01 1.02456617e+00 -8.10927808e-01 -3.35965365e-01
-4.35611784e-01 -1.77222788e-01 8.54927838e-01 4.06512171e-01
7.19943166e-01 -1.64266884e+00 -1.44821203e+00 2.27391362e-01
-1.08641505e-01 -4.53852266e-01 -7.13555664e-02 5.80919504e-01
-6.94395542e-01 -9.82908726e-01 -7.86328793e-01 -3.52701873e-01
1.43728554e-01 1.27344221e-01 1.10040545e+00 1.30090311e-01
-2.06789792e-01 5.21854520e-01 -3.98992330e-01 -1.92233413e-01
-4.24087971e-01 1.28788278e-01 4.93651301e-01 -1.04700029e+00
9.53979716e-02 -7.21018255e-01 -5.11888504e-01 4.76119846e-01
-4.96478796e-01 -9.37273949e-02 6.83316231e-01 1.22060418e+00
1.54347494e-01 -1.54943392e-02 7.67986119e-01 -2.87436992e-01
8.76079559e-01 -5.58611512e-01 -8.74295115e-01 2.19539404e-02
-7.12564290e-01 4.31690842e-01 4.59396124e-01 -6.13465548e-01
-8.59998345e-01 -7.15420097e-02 2.59405017e-01 -3.78393531e-01
-1.39213771e-01 1.89765453e-01 2.92313635e-01 -2.02178121e-01
9.97283936e-01 4.16079700e-01 2.51917988e-01 -2.59884506e-01
4.90748942e-01 2.12258592e-01 6.76172793e-01 -1.04035640e+00
8.82983744e-01 1.52445719e-01 -1.08754478e-01 -5.18715978e-01
-3.54889691e-01 5.10377102e-02 -9.01954547e-02 -4.04205956e-02
5.99373162e-01 -6.88073158e-01 -1.43468535e+00 3.43497515e-01
-5.63418746e-01 -1.11905777e+00 -6.54450238e-01 5.06856859e-01
-1.15039563e+00 3.24924469e-01 -5.23327112e-01 -1.00159681e+00
2.01016329e-02 -1.34396064e+00 7.84182429e-01 4.10527259e-01
-3.98347795e-01 -5.34409642e-01 2.85295218e-01 1.54446438e-01
4.53276545e-01 2.26128340e-01 5.28382242e-01 -5.22846282e-01
-7.90316522e-01 2.79861450e-01 6.24986067e-02 -6.97257556e-03
-2.00702280e-01 -6.07890785e-01 -5.65840840e-01 -5.40781736e-01
-2.01559514e-01 -1.22443402e+00 6.49228156e-01 4.60506715e-02
8.52464914e-01 -2.10285261e-01 -4.25975323e-01 4.73283589e-01
9.31165159e-01 3.18740159e-01 5.50529778e-01 9.65740025e-01
3.35396528e-02 4.40725058e-01 9.95817065e-01 6.51932061e-01
4.19832289e-01 4.87326384e-01 9.25691843e-01 2.98379600e-01
3.27826023e-01 -6.86408997e-01 6.90309882e-01 1.90879434e-01
-1.64048806e-01 -1.66857056e-02 -8.11127603e-01 4.08375829e-01
-2.03482056e+00 -1.10365915e+00 4.45395887e-01 2.13425851e+00
8.52309287e-01 3.70221525e-01 3.68867725e-01 -2.18199715e-01
2.16564164e-01 4.10793442e-03 -1.27600420e+00 -6.80894673e-01
6.15581796e-02 1.61926657e-01 4.68767136e-01 4.91673380e-01
-7.67203927e-01 1.36980677e+00 7.15105724e+00 8.51560831e-01
-7.00920820e-01 -1.71933487e-01 2.87783980e-01 -3.78279716e-01
-1.49152145e-01 2.02841818e-01 -4.52944219e-01 3.36530000e-01
7.03179240e-01 -3.80046457e-01 1.27437186e+00 1.33221245e+00
-4.51347567e-02 -6.69014931e-01 -1.29113901e+00 8.64497244e-01
-2.37887740e-01 -1.13749135e+00 -5.76116681e-01 4.07188743e-01
7.76588976e-01 2.45800987e-01 2.91840494e-01 9.91979718e-01
1.33211470e+00 -1.43090808e+00 7.28857338e-01 1.32799849e-01
6.27263367e-01 -9.10017371e-01 4.00213331e-01 8.82242382e-01
-5.30770600e-01 -2.48504072e-01 -3.05682272e-01 -5.27189434e-01
1.11750744e-01 -2.12951243e-01 -8.69844377e-01 2.01151446e-01
6.55171931e-01 4.54174727e-01 -9.07251760e-02 8.27135444e-01
-7.03200459e-01 3.43811929e-01 -4.40192193e-01 -5.29356360e-01
7.92259872e-01 -1.63047500e-02 8.02631676e-01 7.13656843e-01
1.71249464e-01 2.64689833e-01 2.41140887e-01 1.11543226e+00
2.50249151e-02 -4.97748494e-01 -8.32854986e-01 -2.32623488e-01
4.84847993e-01 9.62421715e-01 -3.95952672e-01 -2.47721821e-01
3.46160650e-01 1.10203075e+00 5.00399470e-01 2.48878479e-01
-9.67728198e-01 -4.94341880e-01 8.26276481e-01 -4.94970292e-01
1.50796011e-01 -3.85776222e-01 -1.60629377e-02 -9.90855396e-01
-2.74926335e-01 -1.32816660e+00 2.42099434e-01 -9.35029209e-01
-7.86874712e-01 4.59319502e-01 -6.56947568e-02 -1.06062150e+00
-7.44203031e-01 -4.72445041e-01 -4.13317591e-01 5.06116331e-01
-1.49206555e+00 -6.75352395e-01 -4.26433496e-02 6.20490611e-01
5.79527617e-01 -3.18589926e-01 7.53211021e-01 -5.46418905e-01
-2.86181599e-01 5.82619190e-01 1.47124276e-01 -3.61251384e-01
4.18187469e-01 -1.40805089e+00 6.61333919e-01 4.03325051e-01
-8.04315507e-03 5.44049799e-01 8.45380008e-01 -7.55477130e-01
-1.70594716e+00 -2.92655915e-01 -3.24819237e-02 -7.25976408e-01
7.43143737e-01 -2.36749947e-01 -5.36082566e-01 7.66746521e-01
1.86278418e-01 -3.67261499e-01 9.99473631e-02 3.91322911e-01
-3.49601477e-01 4.63272452e-01 -1.24960494e+00 9.41625059e-01
1.21146476e+00 -3.23821634e-01 -1.07859027e+00 2.14116231e-01
8.00275683e-01 -7.00852871e-01 -4.71834123e-01 2.79732924e-02
7.43022203e-01 -1.13021398e+00 7.63597369e-01 -7.89438426e-01
6.69519782e-01 -1.84225649e-01 2.43220571e-02 -1.72645950e+00
-2.02546731e-01 -1.47718167e+00 -1.51893795e-01 4.70463663e-01
3.57842863e-01 -8.16907525e-01 1.06195378e+00 5.41089177e-01
6.69686869e-02 -6.91404402e-01 -9.27249551e-01 -1.41914713e+00
3.77218336e-01 -3.42141300e-01 7.04649985e-01 8.62690926e-01
7.43645370e-01 1.14704920e-02 -7.05171525e-01 -2.65341848e-01
7.60555148e-01 1.26327708e-01 1.24500453e+00 -6.72678292e-01
-7.19128072e-01 -5.76383471e-01 3.14210318e-02 -1.31511140e+00
2.70045966e-01 -7.14874506e-01 3.31949621e-01 -1.08233547e+00
8.65940601e-02 -5.65563142e-01 -1.44728959e-01 7.30865538e-01
1.07760720e-01 -1.54788375e-01 4.16402310e-01 1.91123426e-01
-1.05205369e+00 7.41247296e-01 1.49635172e+00 1.43925339e-01
-5.20942092e-01 -3.09060216e-01 -7.86004186e-01 7.40689635e-01
8.51796567e-01 -4.01809305e-01 -3.76365989e-01 -1.92224175e-01
2.73921907e-01 3.74309003e-01 4.59964156e-01 -1.17457640e+00
2.22685561e-01 -6.58436954e-01 1.17619604e-01 -1.22741558e-01
4.40661967e-01 -4.82805938e-01 -1.09836962e-02 9.96844530e-01
-4.60816443e-01 1.95447858e-02 2.30007380e-01 7.51067102e-01
3.04698944e-01 -3.29199851e-01 6.22973561e-01 -5.27829170e-01
-1.06538630e+00 -1.54385179e-01 -7.40044832e-01 3.47591847e-01
1.30783582e+00 -3.83908182e-01 -4.91426975e-01 -6.84172213e-01
-6.34390712e-01 8.38634431e-01 7.22876370e-01 4.15352732e-01
5.65439999e-01 -1.01223350e+00 -4.01916176e-01 7.50751272e-02
1.08945072e-02 -1.17931798e-01 7.60998949e-02 5.81777692e-01
-3.03434730e-01 1.35069534e-01 -5.77091515e-01 -3.20464283e-01
-8.43271673e-01 7.13901639e-01 3.44631732e-01 -4.30860430e-01
-8.42769861e-01 8.86196971e-01 2.00960308e-01 -6.76893413e-01
2.76198030e-01 -8.40425640e-02 2.07604840e-01 -5.57010591e-01
4.87999260e-01 5.84276617e-01 -6.99638009e-01 3.66090126e-02
-3.84532452e-01 2.02746257e-01 -3.94330956e-02 -6.30084038e-01
1.46018970e+00 9.25113335e-02 3.26416612e-01 1.01829000e-01
6.07317984e-01 -2.75733232e-01 -1.76177561e+00 1.52505055e-01
2.65008528e-02 -5.56301713e-01 -2.14161232e-01 -1.34418523e+00
-5.08812845e-01 7.16844618e-01 3.14471990e-01 2.25851350e-02
7.43305266e-01 -8.53446499e-02 7.92346358e-01 1.08824003e+00
1.16321218e+00 -1.39738405e+00 5.56122005e-01 5.76827049e-01
9.48627651e-01 -1.38823378e+00 -1.81266982e-02 5.77583969e-01
-1.07874131e+00 7.89601028e-01 8.82652462e-01 -3.26530159e-01
-1.64093763e-01 1.53515652e-01 -2.43768081e-01 -1.66414842e-01
-9.49635088e-01 -1.77148670e-01 -4.43973958e-01 8.94273341e-01
-3.86657625e-01 2.91310757e-01 4.24494594e-02 4.32636023e-01
-5.89513719e-01 1.74914777e-01 6.30797327e-01 1.17707872e+00
-7.81451821e-01 -9.66123402e-01 -3.90233636e-01 1.09728917e-01
3.16264816e-02 3.55741411e-01 -4.82667953e-01 1.05033898e+00
-2.48699754e-01 1.02643538e+00 -4.30667818e-01 -4.82297868e-01
9.30805579e-02 -2.46079907e-01 6.02660060e-01 -1.96705520e-01
-6.23862982e-01 -3.49493086e-01 1.23796478e-01 -1.25029719e+00
1.20498486e-01 -6.51922584e-01 -1.39512348e+00 -7.05016613e-01
-1.31713971e-01 3.89422894e-01 5.49048722e-01 7.87322760e-01
2.61863500e-01 1.33990645e-01 4.19952482e-01 -1.16252649e+00
-1.12483358e+00 -4.60358381e-01 -6.18630707e-01 1.60008013e-01
5.18764734e-01 -8.91221642e-01 -5.23302495e-01 -6.96477950e-01] | [4.009409427642822, 1.5951828956604004] |
3fa0c009-77d1-474c-9ecc-e3a44ca67544 | dual-adaptive-representation-alignment-for | 2306.10511 | null | https://arxiv.org/abs/2306.10511v1 | https://arxiv.org/pdf/2306.10511v1.pdf | Dual Adaptive Representation Alignment for Cross-domain Few-shot Learning | Few-shot learning aims to recognize novel queries with limited support samples by learning from base knowledge. Recent progress in this setting assumes that the base knowledge and novel query samples are distributed in the same domains, which are usually infeasible for realistic applications. Toward this issue, we propose to address the cross-domain few-shot learning problem where only extremely few samples are available in target domains. Under this realistic setting, we focus on the fast adaptation capability of meta-learners by proposing an effective dual adaptive representation alignment approach. In our approach, a prototypical feature alignment is first proposed to recalibrate support instances as prototypes and reproject these prototypes with a differentiable closed-form solution. Therefore feature spaces of learned knowledge can be adaptively transformed to query spaces by the cross-instance and cross-prototype relations. Besides the feature alignment, we further present a normalized distribution alignment module, which exploits prior statistics of query samples for solving the covariant shifts among the support and query samples. With these two modules, a progressive meta-learning framework is constructed to perform the fast adaptation with extremely few-shot samples while maintaining its generalization capabilities. Experimental evidence demonstrates our approach achieves new state-of-the-art results on 4 CDFSL benchmarks and 4 fine-grained cross-domain benchmarks. | ['Yonghong Tian', 'Jia Li', 'Tong Zhang', 'Yifan Zhao'] | 2023-06-18 | null | null | null | null | ['cross-domain-few-shot', 'cross-domain-few-shot-learning', 'meta-learning', 'few-shot-learning'] | ['computer-vision', 'computer-vision', 'methodology', 'methodology'] | [ 6.60541356e-02 -3.04410487e-01 -6.44111097e-01 -6.42067373e-01
-1.15041053e+00 -3.16732198e-01 6.30828977e-01 1.33649901e-01
-3.39751840e-01 7.77565300e-01 -1.77659497e-01 3.30403715e-01
-4.67862010e-01 -9.90492761e-01 -7.83907413e-01 -5.99288225e-01
8.28186199e-02 6.44404233e-01 5.41827381e-01 -3.61909986e-01
1.93028629e-01 2.18409687e-01 -1.82181180e+00 3.74638647e-01
1.14916301e+00 1.11989260e+00 1.57330826e-01 4.11864430e-01
-6.12117469e-01 4.40596879e-01 -5.20428777e-01 -3.31806362e-01
3.51832300e-01 -3.21783632e-01 -4.87588584e-01 1.56266510e-01
5.72651803e-01 -7.28835762e-02 -2.23702550e-01 1.13409936e+00
4.84499574e-01 6.13391936e-01 7.12741137e-01 -1.26991308e+00
-8.14630032e-01 4.96901959e-01 -7.53751636e-01 5.39314628e-01
3.08877766e-01 6.26901388e-02 1.03513467e+00 -1.30476785e+00
5.75498700e-01 1.09056807e+00 4.97336000e-01 3.32572073e-01
-1.21386409e+00 -6.82233751e-01 3.35236609e-01 6.88385904e-01
-1.63133645e+00 -5.33330739e-01 9.23426628e-01 -1.40965343e-01
8.19244504e-01 1.01526268e-01 3.74023497e-01 1.22025287e+00
-1.87737525e-01 7.75580406e-01 7.93124795e-01 -4.89060938e-01
6.14874363e-01 5.31292260e-01 2.96461493e-01 3.95192653e-01
1.89541727e-01 -8.19811970e-02 -5.81064105e-01 -4.47939783e-01
3.75649363e-01 4.02287334e-01 -1.41697839e-01 -1.01729524e+00
-9.76308644e-01 9.91187334e-01 4.08959478e-01 4.88467813e-01
-3.45805854e-01 -2.77851284e-01 5.26709557e-01 5.32779932e-01
4.27735090e-01 2.11398229e-01 -3.98185700e-01 1.37794793e-01
-7.77746856e-01 2.15366378e-01 7.72388875e-01 1.42638266e+00
1.10280490e+00 7.74670858e-03 -4.24073994e-01 1.20514667e+00
1.51601841e-03 4.81939167e-01 7.90881574e-01 -4.80447441e-01
5.74384272e-01 5.73920548e-01 1.77522153e-01 -8.06260169e-01
-2.52810325e-02 -7.58005738e-01 -5.91452897e-01 -2.22338408e-01
1.71467319e-01 6.41199201e-02 -8.42658997e-01 1.65617871e+00
6.49654388e-01 7.94512630e-01 3.11631650e-01 6.71841919e-01
5.32914221e-01 6.13970220e-01 1.11302167e-01 -4.80047911e-01
1.18693173e+00 -8.56450319e-01 -4.44179773e-01 -1.67805422e-02
5.08076012e-01 -4.95803297e-01 1.41168725e+00 2.28216290e-01
-8.06677938e-01 -7.49213696e-01 -1.21071875e+00 2.34577700e-01
-6.45285308e-01 -2.94419408e-01 3.93468708e-01 5.63373864e-01
-2.41918728e-01 4.34315950e-01 -5.20071566e-01 -3.15090060e-01
4.57443476e-01 1.21736571e-01 -5.07526696e-02 -2.98030138e-01
-1.28558278e+00 6.11248195e-01 7.14420915e-01 -4.94325846e-01
-8.33281219e-01 -1.04552162e+00 -8.52648556e-01 2.34825388e-01
7.31441736e-01 -6.54796302e-01 1.22773075e+00 -9.35679674e-01
-1.38299227e+00 6.52222276e-01 -1.21820442e-01 -4.35952991e-01
3.14972162e-01 -9.74554569e-02 -8.23844850e-01 4.08290625e-02
2.62874097e-01 6.53248578e-02 1.08900428e+00 -1.13020623e+00
-1.08305621e+00 -3.68272543e-01 -1.09488621e-01 2.88004547e-01
-6.71482325e-01 -3.38228464e-01 -4.82439101e-01 -6.39852941e-01
-4.06742617e-02 -5.75400412e-01 -2.66565429e-03 -1.93536550e-01
6.38949648e-02 -4.04951423e-01 8.67988110e-01 5.13295233e-02
1.18818235e+00 -2.22065735e+00 -1.15918135e-02 4.14669394e-01
3.73111814e-02 3.44545960e-01 -1.40090108e-01 2.61615753e-01
-3.60150337e-02 -5.27637661e-01 -3.37551594e-01 1.09444410e-02
9.41153467e-02 2.60985434e-01 -7.49821782e-01 3.91675651e-01
6.11356720e-02 7.82112122e-01 -1.19589913e+00 -5.65821767e-01
1.82416081e-01 1.34112045e-01 -4.90991384e-01 2.83971876e-01
-4.33044076e-01 7.30478689e-02 -6.04574203e-01 8.30618024e-01
8.10766697e-01 -2.50913322e-01 1.47541165e-01 -2.40656346e-01
1.98835373e-01 -3.67468745e-01 -1.37263513e+00 2.01911235e+00
-6.01481438e-01 -1.49328321e-01 -2.83086687e-01 -1.49327648e+00
1.12767172e+00 1.56612210e-02 5.39382815e-01 -7.43270814e-01
-8.31665918e-02 2.95226336e-01 -1.46991387e-01 -5.17323196e-01
3.59028697e-01 -3.60050082e-01 -2.12150589e-01 2.70241380e-01
4.62600589e-01 1.68049529e-01 2.31889695e-01 2.62928270e-02
7.93815851e-01 -5.55249862e-02 5.58519304e-01 -6.61107525e-02
6.59694135e-01 -2.65514962e-02 7.45169103e-01 9.58560765e-01
-2.21598715e-01 3.36211264e-01 -8.21259432e-03 -3.51672441e-01
-8.70156467e-01 -1.52010357e+00 -2.14089155e-01 1.76512265e+00
3.09071034e-01 -3.34027052e-01 -4.02180701e-01 -8.88254583e-01
2.76800960e-01 9.40413356e-01 -6.47854328e-01 -4.77760643e-01
-4.85924453e-01 -6.79945529e-01 1.95282131e-01 5.43326676e-01
4.27705765e-01 -7.96791375e-01 -6.27840579e-01 3.85224193e-01
1.75715908e-01 -8.30142558e-01 -5.60682833e-01 3.17884743e-01
-7.03141272e-01 -1.14225519e+00 -9.42840457e-01 -8.81341696e-01
4.28428859e-01 5.13858199e-01 8.63043964e-01 -3.63836974e-01
-5.22343636e-01 6.80724621e-01 -4.86297220e-01 -2.63086945e-01
-8.00704211e-02 5.95611818e-02 3.11961710e-01 3.68004501e-01
7.32693672e-01 -9.13856268e-01 -3.66219997e-01 3.24769139e-01
-9.86721754e-01 -5.08880973e-01 7.19043314e-01 1.31428897e+00
9.63967860e-01 -1.47645965e-01 9.58524466e-01 -1.17650461e+00
5.80568552e-01 -1.01042402e+00 -6.08628511e-01 8.34675670e-01
-6.87281907e-01 1.46337777e-01 9.03934956e-01 -8.29526603e-01
-1.28825569e+00 -3.09823435e-02 4.22273934e-01 -8.60805988e-01
-8.61978605e-02 4.07249838e-01 -3.40457410e-01 -4.34433594e-02
9.82256651e-01 4.92322057e-01 -3.11722457e-01 -4.00635093e-01
7.66177177e-01 7.26251006e-01 6.85286522e-01 -1.00086010e+00
9.05427098e-01 4.93477136e-01 -2.77837813e-01 -9.48596776e-01
-1.03560507e+00 -9.30674732e-01 -7.12209523e-01 2.15769395e-01
3.51621807e-01 -8.20052326e-01 -2.17547506e-01 9.17166620e-02
-6.74435735e-01 1.22851938e-01 -7.91589975e-01 4.74119842e-01
-9.09464717e-01 4.31018293e-01 -1.50245652e-02 -5.61959803e-01
-2.93236434e-01 -7.56823659e-01 8.49547625e-01 3.44927609e-01
1.30253032e-01 -9.18363810e-01 3.61058861e-01 -1.17635831e-01
4.22585636e-01 6.90198620e-04 1.00835586e+00 -1.26494431e+00
-3.65196258e-01 -1.07263610e-01 -1.43523544e-01 1.38776809e-01
3.23740512e-01 -4.96008664e-01 -1.00172698e+00 -4.58976567e-01
2.71465361e-01 -5.39078891e-01 8.03991914e-01 -7.01026153e-03
1.18923378e+00 -1.92847324e-03 -5.53167403e-01 8.03081632e-01
1.39670122e+00 1.99164718e-01 2.29367360e-01 1.48160845e-01
3.19054484e-01 3.38553727e-01 1.13331687e+00 7.95894206e-01
1.10227354e-01 7.29457378e-01 -5.63451648e-02 5.33840358e-01
2.02365294e-02 -3.19922268e-01 1.75492689e-02 8.05821061e-01
3.36706460e-01 1.38606697e-01 -8.21614861e-01 5.87204218e-01
-1.97717464e+00 -1.16124892e+00 6.49585783e-01 2.30548596e+00
8.35027695e-01 1.68999702e-01 1.04460366e-01 -2.53178567e-01
8.74523938e-01 1.05354199e-02 -1.09573781e+00 1.06477611e-01
-2.74059828e-02 3.21310788e-01 1.68338701e-01 1.63998351e-01
-1.07276726e+00 8.55761111e-01 5.47612572e+00 1.15784359e+00
-9.87018406e-01 2.57942498e-01 2.23385721e-01 -1.94779366e-01
-2.86126226e-01 1.20641561e-02 -1.04725289e+00 4.38543528e-01
8.26983869e-01 -7.22463369e-01 3.48539084e-01 1.25723159e+00
-4.38568294e-01 1.52499482e-01 -1.18048632e+00 1.19881332e+00
2.47234896e-01 -1.28881717e+00 1.24745823e-01 -3.01217169e-01
7.52778888e-01 -8.10232908e-02 1.83401734e-01 1.06048536e+00
6.02398962e-02 -4.76845056e-01 2.87337750e-01 7.89247930e-01
8.96697283e-01 -8.78901124e-01 3.23445529e-01 5.05447626e-01
-1.31822491e+00 -5.04974365e-01 -9.58802223e-01 4.10897523e-01
-1.44954160e-01 5.01098454e-01 -1.01925957e+00 7.31496394e-01
5.45638263e-01 8.45160961e-01 -4.96286035e-01 1.19872165e+00
3.09906006e-01 2.53093153e-01 -2.20528379e-01 2.95146387e-02
-1.46295186e-02 -1.62502471e-02 5.73327720e-01 1.09936726e+00
4.35200244e-01 1.75311089e-01 4.26575661e-01 8.27060342e-01
-4.20707949e-02 3.06068271e-01 -6.99995577e-01 2.43560761e-01
8.26963484e-01 1.14399517e+00 -5.10289550e-01 -7.69535124e-01
-6.52872205e-01 8.71898711e-01 6.31044745e-01 4.99151796e-01
-9.14425850e-01 -6.60860479e-01 6.83946073e-01 -2.16384325e-02
7.13485658e-01 2.44231358e-01 1.45239264e-01 -1.55456781e+00
1.28191143e-01 -8.32931340e-01 9.81242597e-01 -2.57126302e-01
-1.88793492e+00 1.92834646e-01 4.30747837e-01 -1.52484465e+00
-4.29773510e-01 -3.39056879e-01 -7.28459656e-01 6.63228631e-01
-1.65004694e+00 -1.03285825e+00 -3.58509809e-01 1.07963979e+00
8.46736312e-01 -6.72353387e-01 9.39798653e-01 3.08807284e-01
-2.46415794e-01 1.00707698e+00 5.00589252e-01 -1.99261069e-01
9.92620409e-01 -1.06016755e+00 1.74154416e-01 4.09155279e-01
3.50320041e-01 7.89982438e-01 4.68262017e-01 -5.94994366e-01
-1.61273575e+00 -1.20637667e+00 1.24153100e-01 -1.59668982e-01
7.65040100e-01 -2.08294258e-01 -1.37744999e+00 5.79303384e-01
-1.68411851e-01 6.68369949e-01 9.80002344e-01 2.86202550e-01
-6.85148895e-01 -5.16463697e-01 -1.16474652e+00 2.81306356e-01
1.02807808e+00 -6.26042366e-01 -1.00944197e+00 3.78082931e-01
6.82830453e-01 -1.76690429e-01 -9.12364662e-01 4.33385044e-01
3.04860145e-01 -7.84934223e-01 1.12636924e+00 -1.03797770e+00
-2.56154209e-01 -2.96757817e-01 -6.17440999e-01 -1.30956221e+00
-3.65410835e-01 -2.89065748e-01 -5.95222712e-01 1.28273618e+00
5.37742972e-02 -5.61910808e-01 7.90370703e-01 3.19661170e-01
-1.68581009e-01 -7.51927316e-01 -1.08021522e+00 -1.17150903e+00
-6.53821137e-03 -3.43890607e-01 6.01669967e-01 1.11171663e+00
7.19178021e-02 4.01696831e-01 -2.65274704e-01 2.61239298e-02
8.83031368e-01 6.19866788e-01 8.83606255e-01 -1.24836278e+00
-6.31836355e-01 -2.35680044e-01 -3.78989488e-01 -9.26235497e-01
4.24900740e-01 -1.07302058e+00 -3.85062806e-02 -8.46267283e-01
3.23388875e-01 -5.29353499e-01 -7.43639767e-01 2.14746818e-01
-1.57769710e-01 -2.93185115e-01 1.94771364e-02 2.66063064e-01
-1.06032896e+00 9.39037979e-01 8.24301898e-01 -3.15425992e-01
-3.49536777e-01 2.42571428e-01 -6.91681385e-01 5.71206093e-01
5.62413573e-01 -4.35142606e-01 -8.23082268e-01 -2.53866334e-03
-1.96402937e-01 -7.65544102e-02 1.42571270e-01 -1.12916744e+00
4.93550003e-01 -4.02998924e-01 5.37389398e-01 -5.25689363e-01
4.24813718e-01 -8.12958121e-01 -1.70514598e-01 1.67024240e-01
-3.48252535e-01 -3.86658370e-01 -3.85941230e-02 1.08623183e+00
-2.13901103e-01 -4.12238687e-01 9.74572361e-01 -1.64499521e-01
-1.13021791e+00 4.27333653e-01 3.24798524e-01 5.21746457e-01
1.43130302e+00 -2.16562331e-01 -1.93662137e-01 2.33326433e-03
-8.55262280e-01 3.61968726e-01 2.96431959e-01 5.02762377e-01
8.58795285e-01 -1.54411244e+00 -5.38618088e-01 4.22861010e-01
7.12352931e-01 -9.54644009e-02 4.63705868e-01 6.42464459e-01
1.97940350e-01 2.62171566e-01 -2.55313039e-01 -6.49519563e-01
-7.67593622e-01 1.15748715e+00 2.63793081e-01 -1.14742637e-01
-6.18347108e-01 7.09878623e-01 2.65855193e-01 -3.55548620e-01
3.71320307e-01 -1.03271976e-01 -9.94578302e-02 3.07492822e-01
8.11884642e-01 4.80355442e-01 -4.89590615e-02 -3.24885607e-01
-2.10646555e-01 5.42114615e-01 -4.70178366e-01 1.10581204e-01
1.35584474e+00 -1.76950589e-01 3.65308285e-01 9.28925097e-01
1.29127932e+00 -7.33193308e-02 -1.20239925e+00 -1.02973771e+00
2.57381856e-01 -7.10725486e-01 -4.91055131e-01 -4.36613232e-01
-6.34232223e-01 7.26668656e-01 8.21376920e-01 -2.11994071e-02
1.00382650e+00 1.56328425e-01 7.23825157e-01 8.11686695e-01
5.69564581e-01 -1.32262373e+00 3.64492744e-01 4.94257003e-01
5.57133198e-01 -1.34849203e+00 -7.14396685e-02 -2.06593171e-01
-5.07220507e-01 1.10516608e+00 8.61562431e-01 -2.47595817e-01
7.00369298e-01 -2.06236750e-01 -3.97443205e-01 -1.20006159e-01
-7.68691003e-01 -1.84689298e-01 4.69489425e-01 7.06242144e-01
5.72343171e-02 -3.24735083e-02 -1.66054182e-02 9.80534673e-01
1.70873180e-01 -8.04752484e-02 1.16009369e-01 1.02047908e+00
-8.43472183e-01 -1.02327991e+00 -2.95297772e-01 5.93128264e-01
3.62057276e-02 6.97470456e-02 9.55659375e-02 7.68203616e-01
7.22611845e-02 5.18577337e-01 -5.27386256e-02 -3.12690347e-01
5.64374149e-01 3.58392954e-01 5.20018697e-01 -1.05897009e+00
-5.34315109e-02 -9.97300372e-02 -3.93610507e-01 -4.17175263e-01
-1.92079768e-01 -4.53589022e-01 -1.20732999e+00 2.58562893e-01
-2.72084743e-01 3.64379644e-01 1.89682171e-01 8.85286808e-01
4.32321757e-01 2.97696054e-01 8.11412454e-01 -5.11234164e-01
-1.28350675e+00 -8.21989715e-01 -8.14857602e-01 5.30610979e-01
1.85631156e-01 -1.04003668e+00 -1.89077497e-01 -2.49347016e-01] | [10.038135528564453, 3.0571975708007812] |
bcb45e6e-4f10-44ae-9ba2-13aeb5c0b641 | consistency-of-spectral-clustering-for | 2109.10319 | null | https://arxiv.org/abs/2109.10319v4 | https://arxiv.org/pdf/2109.10319v4.pdf | Community detection for weighted bipartite networks | The bipartite network appears in various areas, such as biology, sociology, physiology, and computer science. \cite{rohe2016co} proposed Stochastic co-Blockmodel (ScBM) as a tool for detecting community structure of binary bipartite graph data in network studies. However, ScBM completely ignores edge weight and is unable to explain the block structure of a weighted bipartite network. Here, to model a weighted bipartite network, we introduce a Bipartite Distribution-Free model by releasing ScBM's distribution restriction. We also build an extension of the proposed model by considering the variation of node degree. Our models do not require a specific distribution on generating elements of the adjacency matrix but only a block structure on the expected adjacency matrix. Spectral algorithms with theoretical guarantees on the consistent estimation of node labels are presented to identify communities. Our proposed methods are illustrated by simulated and empirical examples. | ['Jingli Wang', 'Huan Qing'] | 2021-09-21 | null | null | null | null | ['stochastic-block-model'] | ['graphs'] | [ 1.31519184e-01 3.50222766e-01 -2.41407588e-01 2.13964097e-02
3.68900329e-01 -6.26718223e-01 1.58221796e-01 6.46762401e-02
7.94264898e-02 8.44289184e-01 -1.19010046e-01 -6.79980218e-01
-7.36090839e-01 -9.43185270e-01 -4.50574040e-01 -7.19472647e-01
-5.01622558e-01 2.87394881e-01 5.02088010e-01 -1.62409797e-01
2.07466576e-02 2.40409300e-01 -8.26728284e-01 -2.12599725e-01
7.38269091e-01 1.74875259e-01 7.74879977e-02 6.31075919e-01
-1.92206740e-01 3.66180420e-01 -2.50379145e-01 -4.64018583e-01
4.44121420e-01 -6.79574072e-01 -6.23253465e-01 3.28296155e-01
-3.69091511e-01 3.16289365e-01 -7.79579401e-01 1.40234530e+00
1.93976998e-01 -3.38147134e-01 8.75096500e-01 -1.66741908e+00
-4.88130689e-01 1.10472858e+00 -1.33766055e+00 2.80401170e-01
3.19347084e-01 -2.35029370e-01 1.24348247e+00 -4.06753629e-01
6.81357086e-01 1.06344759e+00 7.23914504e-01 2.05262169e-01
-1.75236905e+00 -7.99925566e-01 1.79001272e-01 1.80929080e-01
-1.68550026e+00 1.30292028e-01 1.05896258e+00 -6.10833049e-01
1.39550820e-01 4.60718721e-01 8.24194431e-01 6.80868268e-01
5.85711561e-02 4.00255352e-01 1.30489600e+00 -4.10557181e-01
5.19190468e-02 -4.30885442e-02 4.92616713e-01 6.61071002e-01
9.83371019e-01 -1.50128543e-01 -1.03285074e-01 -5.55119514e-01
6.41122460e-01 1.38131052e-01 -3.82095605e-01 -9.05886531e-01
-1.03449690e+00 7.78733492e-01 3.48146409e-01 4.80525941e-01
-1.58719957e-01 2.43176326e-01 2.54202306e-01 4.03865337e-01
2.73375511e-01 -3.22019011e-01 6.03458472e-03 4.12993699e-01
-9.29018676e-01 -1.84385478e-01 1.02519703e+00 1.23046720e+00
8.80161345e-01 -1.20595612e-01 1.38806209e-01 4.96394038e-01
6.12647295e-01 6.12917006e-01 -1.97916731e-01 -6.26675963e-01
4.18295920e-01 6.69056654e-01 -4.53862436e-02 -1.60646415e+00
-6.78915739e-01 -4.98624176e-01 -1.37523437e+00 -3.01963091e-01
3.68086010e-01 -1.22458331e-01 -4.69087064e-01 1.78275120e+00
3.25899065e-01 3.56554985e-01 -3.26629996e-01 6.23147309e-01
6.14110351e-01 3.90505344e-01 -2.82914490e-01 -4.96382564e-01
9.21280682e-01 -5.33897996e-01 -5.68572938e-01 1.60754681e-01
3.77491266e-01 -4.18586731e-01 4.01739299e-01 2.06824228e-01
-8.55545998e-01 4.11209231e-03 -8.80293489e-01 7.21754968e-01
1.65803343e-01 -2.55186617e-01 6.61905110e-01 1.01316094e+00
-1.31712532e+00 2.65010923e-01 -5.91398418e-01 -7.24401772e-01
-2.88559459e-02 4.17156428e-01 -3.48711789e-01 -1.02592096e-01
-1.17216432e+00 4.06996906e-01 3.81136358e-01 3.43359232e-01
-5.61973929e-01 -2.50691742e-01 -6.44574344e-01 2.53008515e-01
3.01582992e-01 -6.87324703e-01 4.05503601e-01 -1.10414267e+00
-9.40824509e-01 7.44205356e-01 1.07047390e-02 -4.68071669e-01
5.35054088e-01 8.79437029e-01 -4.15208101e-01 2.44546220e-01
-2.42637843e-02 -1.48129091e-02 5.25507569e-01 -1.36231208e+00
-5.37123531e-02 -6.64227977e-02 8.33880007e-02 -2.39380866e-01
-2.36169368e-01 -1.93186492e-01 -1.55718789e-01 -4.09384191e-01
4.96984661e-01 -1.00369000e+00 -5.23424447e-01 -3.22198689e-01
-6.56993151e-01 -1.77742038e-02 3.42131853e-01 -3.06429356e-01
1.55488813e+00 -2.08919406e+00 7.74789974e-02 1.00914705e+00
6.92014396e-01 -2.10967198e-01 -2.19546407e-01 1.06216145e+00
-1.64159447e-01 4.09381300e-01 -3.82410675e-01 1.92923710e-01
7.66022205e-02 5.95282130e-02 4.16723862e-02 1.13281357e+00
-2.02005282e-01 3.57612878e-01 -1.02551925e+00 -5.42068064e-01
-1.11724064e-01 1.30839750e-01 -6.17499948e-01 -3.49560082e-01
3.31036836e-01 2.59521008e-01 -3.09958875e-01 2.76716739e-01
1.14517832e+00 -7.06217229e-01 1.03882599e+00 2.36298908e-02
1.30099609e-01 -3.50045055e-01 -1.66123605e+00 1.08021379e+00
-8.87872577e-02 3.83245200e-01 5.35382628e-01 -1.38039708e+00
9.12308097e-01 2.16796666e-01 7.66887963e-01 6.14244901e-02
2.28108317e-01 3.32652554e-02 8.09059858e-01 -2.88110197e-01
-1.07151248e-01 -7.58678978e-03 6.34279326e-02 6.36587083e-01
-1.33718431e-01 3.39677185e-01 5.69629312e-01 7.79439867e-01
1.50061238e+00 -3.78028959e-01 4.49322402e-01 -8.28718901e-01
6.32670224e-01 -3.56101155e-01 7.65796781e-01 7.75876343e-01
-3.56653184e-01 2.71062821e-01 9.94190335e-01 1.09604388e-01
-1.15422356e+00 -9.84137058e-01 -1.73201278e-01 4.53685880e-01
5.91826856e-01 -3.13369453e-01 -5.31404972e-01 -4.22380477e-01
1.88518271e-01 -8.30154642e-02 -6.26242876e-01 -1.88785959e-02
2.11821217e-02 -1.01894426e+00 4.41323966e-01 -9.97242332e-02
2.15910435e-01 -5.52535951e-01 1.51430011e-01 -6.42093346e-02
-2.61198193e-01 -9.10520911e-01 -6.27631366e-01 8.22039973e-03
-7.89438128e-01 -1.43604255e+00 -4.86585796e-01 -8.16677392e-01
1.01968098e+00 5.37912190e-01 9.49197829e-01 5.32312214e-01
-2.86516339e-01 4.20707703e-01 -3.73866767e-01 1.19192474e-01
-3.58424783e-01 6.38413727e-02 2.25839257e-01 3.44122142e-01
2.21810669e-01 -1.03331304e+00 -7.48109639e-01 5.63499093e-01
-8.74975801e-01 1.19611323e-01 4.86433983e-01 8.54420841e-01
6.87104911e-02 2.12989762e-01 6.94632709e-01 -1.03929698e+00
6.45074785e-01 -1.02801526e+00 -7.27405667e-01 2.33375028e-01
-8.06679189e-01 -1.75342076e-02 3.20546508e-01 -2.73855358e-01
-6.37340009e-01 -1.10417735e-02 3.44883412e-01 -8.89962241e-02
3.20007473e-01 7.33863890e-01 -1.58612534e-01 -1.50578752e-01
2.77732551e-01 1.72791213e-01 2.07375027e-02 -2.48294532e-01
2.46609956e-01 6.30193293e-01 -6.87536970e-02 -4.05102164e-01
1.00183225e+00 6.45984530e-01 6.36111856e-01 -7.14528084e-01
-9.85712633e-02 -7.54363060e-01 -7.44869113e-01 -3.13155621e-01
2.82675326e-01 -6.82575166e-01 -1.24470937e+00 7.92865232e-02
-1.07176113e+00 1.02926210e-01 2.97369123e-01 4.65324134e-01
-2.78623044e-01 9.01787758e-01 -6.66601717e-01 -1.26333642e+00
1.39888331e-01 -6.17955089e-01 3.01285654e-01 -1.40847087e-01
3.00461072e-02 -1.24046397e+00 4.46216792e-01 -1.68257244e-02
2.30559092e-02 3.51321220e-01 9.27297950e-01 -5.45664966e-01
-5.67060769e-01 -1.20244786e-01 -6.00013375e-01 -3.42520960e-02
1.24216035e-01 2.44814113e-01 -3.72145504e-01 -4.47109252e-01
-3.68110955e-01 2.31004268e-01 7.33691931e-01 4.32680190e-01
8.02993774e-01 -3.12800407e-01 -7.32708633e-01 3.49414974e-01
1.54664326e+00 -7.75300860e-02 3.92140299e-01 -1.10141732e-01
5.85946560e-01 7.62160420e-01 3.22836824e-02 5.86423755e-01
2.67013758e-01 3.52083832e-01 5.89978337e-01 7.24400766e-03
2.04012021e-01 -2.93522537e-01 2.68458366e-01 1.31527317e+00
-8.93451944e-02 -4.94257927e-01 -1.03103602e+00 7.33188927e-01
-1.95682549e+00 -1.13167477e+00 -9.74544406e-01 2.11521077e+00
6.97846472e-01 8.03295299e-02 4.65619415e-01 2.21111476e-01
1.44746482e+00 -9.78902541e-03 -2.13021591e-01 -1.54154763e-01
-2.91999280e-01 -1.19948380e-01 9.43286419e-01 7.25670815e-01
-6.73993587e-01 3.94508451e-01 6.81930971e+00 7.12615728e-01
-4.71719027e-01 1.12433404e-01 2.83438832e-01 3.01401973e-01
-7.20026493e-01 4.62481111e-01 -4.02290344e-01 6.25807166e-01
6.68222725e-01 -5.79811513e-01 3.52553934e-01 5.19770920e-01
3.21226865e-01 -9.89120454e-02 -6.22902870e-01 8.70356202e-01
-2.41750970e-01 -8.80647123e-01 -2.42466360e-01 5.80956936e-01
8.84131074e-01 -2.47347891e-01 -3.65062237e-01 -1.27527252e-01
7.84021020e-01 -8.11803937e-01 1.07924834e-01 6.08935297e-01
5.75894713e-01 -6.26343369e-01 7.39829600e-01 4.32234317e-01
-1.28250945e+00 -2.87506357e-02 -4.26685542e-01 -2.26941139e-01
7.78559744e-02 9.33946490e-01 -7.05168962e-01 8.21660757e-01
4.29657549e-01 7.64787495e-01 -3.64731252e-01 1.35301661e+00
8.28339458e-02 9.54885781e-01 -3.75938714e-01 -1.85250074e-01
-8.50141980e-03 -7.83569157e-01 7.51795590e-01 1.10355628e+00
2.92774171e-01 -8.83257948e-03 3.85931045e-01 8.18299055e-01
-4.93059792e-02 3.04729968e-01 -7.71973193e-01 -1.56304762e-01
5.32359540e-01 1.33310080e+00 -1.34032297e+00 9.64690745e-02
-5.06191492e-01 7.19182372e-01 1.24284819e-01 5.13182998e-01
-6.67562604e-01 -4.53770667e-01 2.94309735e-01 4.24731731e-01
3.32811534e-01 -2.92774022e-01 -1.37936875e-01 -1.21552777e+00
-2.21223041e-01 -4.88744318e-01 2.73194760e-01 -2.95030922e-01
-1.56521809e+00 3.26682150e-01 1.50621742e-01 -1.19727004e+00
1.20153472e-01 -3.61715764e-01 -6.15329683e-01 7.00868964e-01
-1.05277538e+00 -9.08101737e-01 -1.61840081e-01 5.77983201e-01
-2.28606626e-01 2.15748902e-02 2.44296193e-01 6.05576396e-01
-5.55247247e-01 2.05310822e-01 4.86065000e-01 1.78362519e-01
3.61690223e-01 -1.35342789e+00 6.76363856e-02 8.00261676e-01
1.12015925e-01 7.83828378e-01 1.04267073e+00 -8.72453570e-01
-1.17608654e+00 -7.13876188e-01 6.93635166e-01 -1.12179220e-01
1.23783612e+00 -7.04142570e-01 -5.27190983e-01 5.80597639e-01
2.88966030e-01 1.60971999e-01 6.63362920e-01 2.12069973e-01
-2.13104978e-01 -2.00052962e-01 -1.09262037e+00 5.33490300e-01
1.48121011e+00 -2.43985832e-01 7.88432360e-02 4.32056218e-01
4.73007977e-01 2.87203103e-01 -7.43993104e-01 3.29590708e-01
5.29331744e-01 -1.03300750e+00 6.83539689e-01 -5.06559193e-01
-2.20610574e-02 -4.85489160e-01 -1.23509141e-02 -1.14362752e+00
-7.56044090e-01 -7.66370296e-01 9.04439464e-02 1.24903309e+00
3.72872561e-01 -8.13885391e-01 8.36211324e-01 -1.64939184e-02
5.46493888e-01 -4.30110037e-01 -9.88687992e-01 -9.68886554e-01
-1.62407651e-03 -1.82371978e-02 3.11425537e-01 1.04971421e+00
4.11149800e-01 2.99318880e-01 -6.04896605e-01 1.96156442e-01
9.78412688e-01 7.47644901e-02 6.72654569e-01 -1.59410691e+00
-4.59273607e-01 -5.65944850e-01 -7.96449125e-01 -7.45225608e-01
2.70709693e-01 -1.00438035e+00 -2.23322272e-01 -1.56959891e+00
7.93437898e-01 -5.81973076e-01 -3.11257511e-01 -1.55156121e-01
1.13227352e-01 3.03216130e-01 1.75995454e-02 2.31722921e-01
-5.88427603e-01 3.55732828e-01 1.00754654e+00 -1.91473931e-01
3.82480845e-02 2.45219409e-01 -5.19251287e-01 5.61678827e-01
6.14289045e-01 -6.28161430e-01 -6.75270617e-01 2.73347914e-01
7.55184054e-01 3.58555943e-01 3.59282196e-01 -6.00263298e-01
2.44183317e-01 -2.64008522e-01 -2.01882988e-01 -4.64248747e-01
-1.98582128e-01 -1.20652270e+00 7.67302930e-01 7.96488643e-01
-9.41030532e-02 1.86337292e-01 -3.30923706e-01 1.24942160e+00
8.23358074e-02 -3.90772969e-01 6.07839286e-01 8.94634575e-02
2.00817287e-02 2.95350939e-01 -4.57635701e-01 -7.04077631e-02
1.25178480e+00 -2.14564651e-01 -3.03296149e-01 -6.58728540e-01
-1.02009249e+00 4.85067099e-01 4.48842585e-01 1.06916912e-02
2.83541530e-01 -1.34224844e+00 -7.34898329e-01 -1.97729513e-01
-2.22403929e-02 -6.16101742e-01 3.83074373e-01 1.37862539e+00
-5.68415284e-01 3.58553648e-01 -5.49961987e-04 -5.83369493e-01
-1.41584086e+00 8.05390954e-01 1.70529723e-01 -3.52840424e-01
-2.63310671e-01 3.23766530e-01 2.34900564e-01 -4.92783755e-01
-4.82581183e-03 1.69063181e-01 -9.56685171e-02 -1.33965909e-01
-7.22569525e-02 5.62412024e-01 -4.78430599e-01 -6.59043968e-01
-5.10349035e-01 3.68801504e-01 2.48600602e-01 -7.79173449e-02
1.01317835e+00 -7.07731783e-01 -6.49462521e-01 6.32533371e-01
8.68282020e-01 3.43971908e-01 -7.06887364e-01 -2.65309662e-01
3.00574332e-01 -4.01101559e-01 -3.18533450e-01 -1.49791121e-01
-1.05225790e+00 6.15739644e-01 1.76256239e-01 9.76341844e-01
8.20051908e-01 -2.72012490e-04 2.55303502e-01 1.12731218e-01
6.40906870e-01 -7.18567252e-01 -2.16102555e-01 2.52933711e-01
4.62445945e-01 -9.76496816e-01 1.34231329e-01 -7.60322154e-01
-3.07718843e-01 8.57093632e-01 4.44766909e-01 -3.24622452e-01
1.36541152e+00 2.85026655e-02 -4.86260921e-01 -2.58119255e-01
-5.91690004e-01 -4.45045471e-01 7.21250623e-02 6.12468362e-01
3.78498793e-01 1.78600684e-01 -1.04353440e+00 5.23025870e-01
2.33686537e-01 -2.42878139e-01 9.14531767e-01 5.26158333e-01
-3.97553086e-01 -1.13893175e+00 -2.44454205e-01 3.90840858e-01
-3.44102740e-01 -3.07237655e-01 -6.81583762e-01 7.31420994e-01
-7.51465335e-02 9.87401903e-01 -7.31588379e-02 -3.88372064e-01
-1.54104903e-01 -2.30589613e-01 4.48064059e-01 -6.12479150e-01
-2.32398495e-01 1.23880953e-01 -4.64323647e-02 2.66113281e-02
-6.76178753e-01 -5.79098582e-01 -7.50898719e-01 -8.12627137e-01
-7.37560153e-01 6.73222363e-01 2.46536270e-01 6.54564619e-01
3.16717476e-01 3.64600092e-01 8.95490587e-01 -1.71262562e-01
-2.07523301e-01 -1.01344192e+00 -1.26970136e+00 1.85388222e-01
1.68397576e-01 -6.41576946e-01 -6.82609320e-01 -1.62191600e-01] | [6.922760486602783, 5.187610626220703] |
fee9332c-5783-49b3-9dde-aa347ff51e8f | distributed-representations-of-atoms-and | 2107.14664 | null | https://arxiv.org/abs/2107.14664v1 | https://arxiv.org/pdf/2107.14664v1.pdf | Distributed Representations of Atoms and Materials for Machine Learning | The use of machine learning is becoming increasingly common in computational materials science. To build effective models of the chemistry of materials, useful machine-based representations of atoms and their compounds are required. We derive distributed representations of compounds from their chemical formulas only, via pooling operations of distributed representations of atoms. These compound representations are evaluated on ten different tasks, such as the prediction of formation energy and band gap, and are found to be competitive with existing benchmarks that make use of structure, and even superior in cases where only composition is available. Finally, we introduce a new approach for learning distributed representations of atoms, named SkipAtom, which makes use of the growing information in materials structure databases. | ['Keith T. Butler', 'Ricardo Grau-Crespo', 'Luis M. Antunes'] | 2021-07-30 | null | null | null | null | ['formation-energy'] | ['miscellaneous'] | [ 1.63646743e-01 -3.85779887e-01 -4.44092453e-01 -1.46414638e-01
-8.11480045e-01 -4.09949183e-01 7.92548537e-01 7.57276893e-01
-2.44140565e-01 1.22403061e+00 2.23645791e-01 -2.97395408e-01
3.36234719e-02 -1.27972984e+00 -9.09797370e-01 -1.27047896e+00
-3.52044106e-02 6.03994012e-01 2.20991626e-01 -1.37518138e-01
6.90131366e-01 8.50581050e-01 -1.89924145e+00 8.71929049e-01
9.01846349e-01 1.42983758e+00 3.16385061e-01 2.79166728e-01
-5.23708761e-01 8.59890282e-01 -5.79505324e-01 -7.90917575e-02
7.75178000e-02 -4.42267597e-01 -8.53208244e-01 -4.49443877e-01
3.73585552e-01 6.29133359e-02 -2.10175723e-01 9.44073617e-01
5.48788786e-01 4.46908385e-01 1.41934216e+00 -5.41071415e-01
-9.77327228e-01 8.06146026e-01 -4.70127732e-01 -6.26807809e-02
3.60194743e-01 -9.05620158e-02 1.29484403e+00 -1.02000380e+00
8.26855719e-01 1.03954315e+00 2.95531541e-01 4.37380284e-01
-1.28775823e+00 -5.23218572e-01 5.75060374e-04 5.34695268e-01
-1.49249160e+00 -3.07159603e-01 8.13813031e-01 -3.69650036e-01
2.01303363e+00 2.21250225e-02 2.91874111e-01 7.00346649e-01
5.03998220e-01 1.52177542e-01 8.59449625e-01 -6.57952309e-01
5.83289981e-01 -1.23214468e-01 1.82557434e-01 5.48759758e-01
8.33839834e-01 8.84864554e-02 -8.35190058e-01 -5.10142624e-01
2.29967430e-01 3.14580292e-01 2.10992098e-02 -3.67166579e-01
-8.69078696e-01 1.00026488e+00 9.34849620e-01 5.51197052e-01
-6.84823215e-01 4.17500913e-01 4.17944312e-01 1.74096867e-01
7.61144936e-01 7.33632803e-01 -6.37510538e-01 1.84879005e-01
-4.83845502e-01 4.49524194e-01 5.63520312e-01 7.65239894e-01
1.07082260e+00 5.05687743e-02 5.08133247e-02 8.20363820e-01
4.15023975e-02 3.43806982e-01 6.81408823e-01 -6.03835762e-01
4.44906116e-01 5.40482819e-01 3.92127819e-02 -5.09284556e-01
-2.36868337e-01 1.70877293e-01 -7.71537960e-01 2.15513140e-01
5.72753996e-02 1.44252658e-01 -7.34378338e-01 1.17504239e+00
2.79114753e-01 -3.55105847e-01 9.80248079e-02 2.89680958e-01
1.10002327e+00 1.28885448e+00 1.26631856e-01 -3.14672649e-01
6.60155237e-01 -9.04925108e-01 -3.29547375e-01 3.85968834e-01
7.25091457e-01 -4.83141154e-01 3.11320573e-01 5.58640301e-01
-1.37725806e+00 -5.17908156e-01 -9.74021494e-01 -3.63934487e-01
-1.13058138e+00 -2.37691894e-01 1.29915798e+00 6.96020663e-01
-7.36470163e-01 1.33192968e+00 -5.92725098e-01 1.31340563e-01
6.65297031e-01 7.91359603e-01 -3.57035607e-01 -3.05291623e-01
-6.65109992e-01 8.47037673e-01 4.39640969e-01 -2.59962648e-01
-7.54160523e-01 -9.91077483e-01 -6.55727386e-01 -1.81524251e-02
4.66567054e-02 -6.16181254e-01 9.43645060e-01 -7.37408876e-01
-1.52413237e+00 4.50806230e-01 -1.90305650e-01 -3.91556680e-01
-2.83342034e-01 7.40610808e-02 -1.50692567e-01 1.73313648e-01
-2.76852638e-01 5.23678720e-01 7.03340709e-01 -9.44061935e-01
-8.70559514e-02 -2.46927634e-01 -2.54866898e-01 -2.19538033e-01
-5.45230567e-01 5.81395887e-02 5.09120107e-01 -2.57160008e-01
-2.06280760e-02 -4.62961435e-01 -5.89306891e-01 -1.05110899e-01
-4.50283527e-01 -7.14523077e-01 3.61594468e-01 -3.63238990e-01
6.88201547e-01 -1.67381990e+00 4.99568403e-01 3.94400507e-01
3.49820167e-01 1.80338556e-03 -9.78102535e-02 7.88712382e-01
-4.19010133e-01 1.56702518e-01 -1.40739620e-01 -1.03163444e-01
-7.98498839e-02 -2.51203209e-01 -1.34210467e-01 5.27261555e-01
2.16250986e-01 8.66681099e-01 -5.63077927e-01 -1.77638307e-01
2.01074049e-01 4.65605229e-01 -8.58442783e-01 -4.22036313e-02
-8.34535241e-01 5.09795034e-03 -5.04054189e-01 7.49849439e-01
7.78395712e-01 -4.55872446e-01 1.70333087e-01 -2.06036180e-01
-4.45301980e-01 9.39894319e-01 -8.05106819e-01 1.69989836e+00
-3.64931256e-01 3.98174562e-02 -2.84484684e-01 -1.06830680e+00
9.58292723e-01 2.50594676e-01 6.53385699e-01 -7.07823277e-01
7.45876506e-02 4.95068073e-01 3.78391922e-01 -1.09834531e-02
6.10820115e-01 -3.55328918e-01 1.67586908e-01 6.24948204e-01
2.89943010e-01 -5.22665203e-01 1.52270705e-01 1.08215041e-01
1.07290852e+00 -5.44418907e-03 3.83959740e-01 -4.34029698e-01
3.65794629e-01 -2.70413131e-01 -9.37291458e-02 5.70943296e-01
5.10012865e-01 4.80469316e-01 1.59079835e-01 -7.28713989e-01
-1.10186601e+00 -9.46582854e-01 -2.95955241e-01 1.15723097e+00
-2.28656605e-01 -8.26003611e-01 -4.12700891e-01 -3.93225551e-01
2.18511626e-01 3.40036869e-01 -5.29546142e-01 -3.64406735e-01
-2.41467938e-01 -1.03460491e+00 -2.73988992e-01 6.83091164e-01
9.21172788e-04 -1.33026373e+00 -1.75724134e-01 2.22455829e-01
5.11992574e-01 -5.54198444e-01 5.19277900e-02 7.19744325e-01
-8.74582231e-01 -8.70861173e-01 -7.93349683e-01 -6.03243411e-01
5.37731588e-01 3.02331477e-01 1.11642897e+00 1.48070589e-01
-8.52719605e-01 -1.52148204e-02 -6.77826926e-02 -5.93965769e-01
-3.56641710e-01 4.64294523e-01 -4.74399552e-02 -2.04738140e-01
1.61347628e-01 -7.19157636e-01 -4.76768166e-01 -4.43266749e-01
-6.46520793e-01 -4.02629673e-02 5.57870567e-01 6.10108733e-01
7.51827240e-01 1.72289550e-01 7.16001928e-01 -1.05261421e+00
5.17204702e-01 -8.13437045e-01 -3.70177805e-01 2.17590779e-01
-3.68960410e-01 4.44555849e-01 9.27465439e-01 -1.37624636e-01
-1.00696516e+00 -8.39745477e-02 -2.31814891e-01 -1.06224477e-01
-4.92550470e-02 5.21661103e-01 -2.58638799e-01 -3.46879244e-01
9.08379614e-01 -1.09962642e-01 -4.06363249e-01 -6.17649138e-01
3.43082458e-01 7.15398431e-01 -5.63502423e-02 -1.26514101e+00
1.51552692e-01 1.39518067e-01 4.18143123e-01 -1.19255698e+00
-3.91250938e-01 -3.83702457e-01 -4.85428065e-01 4.32743102e-01
5.83449900e-01 -7.57484853e-01 -4.55679357e-01 3.30692321e-01
-1.48810077e+00 1.62390359e-02 -4.65169340e-01 2.91333675e-01
-4.89226073e-01 1.04056731e-01 -6.68240249e-01 -7.29020953e-01
-5.98362327e-01 -9.74889934e-01 9.17043149e-01 4.55707721e-02
-1.15116373e-01 -8.60424936e-01 6.21605180e-02 1.02397231e-02
6.24156296e-01 2.33089909e-01 1.73276806e+00 -6.64349973e-01
-8.33607256e-01 -2.88241327e-01 -8.59609917e-02 2.55066544e-01
6.80785596e-01 1.63990632e-01 -1.09439039e+00 -3.30315046e-02
-4.32983220e-01 -5.77727377e-01 1.40046275e+00 4.29643571e-01
1.83416009e+00 6.69692382e-02 -4.95333135e-01 2.43887797e-01
1.55243278e+00 1.19385980e-01 8.47486496e-01 -2.89041162e-01
9.08275127e-01 7.59091437e-01 -1.98810443e-01 7.12363243e-01
-1.17641725e-01 3.94397259e-01 3.61234963e-01 3.50796878e-01
-2.65563309e-01 -1.19433008e-01 1.98379278e-01 8.95219624e-01
-8.28860402e-01 -2.34748840e-01 -6.92595541e-01 2.65436381e-01
-1.27161837e+00 -1.02414203e+00 -1.05404750e-01 2.29228497e+00
8.80516768e-01 8.13627057e-03 9.60554704e-02 1.90220490e-01
6.06819570e-01 2.20352963e-01 -8.75465572e-01 -6.53137863e-01
1.27967782e-02 1.23943758e+00 6.21076941e-01 1.74042165e-01
-7.39177346e-01 9.15803790e-01 8.00475883e+00 9.67733681e-01
-1.05355895e+00 6.80653155e-02 8.52981925e-01 -3.11492980e-01
-6.38151824e-01 -2.91778862e-01 -8.74706626e-01 4.18245584e-01
1.51148021e+00 3.24846953e-02 5.40099800e-01 7.71560311e-01
-3.41154277e-01 4.54140420e-04 -1.34738827e+00 1.05496860e+00
-1.64775074e-01 -2.20028687e+00 5.07841110e-01 1.07415691e-01
1.28477645e+00 2.87985146e-01 1.04349107e-01 -2.12022215e-02
1.87901080e-01 -1.36545932e+00 4.96082336e-01 5.89765251e-01
8.35352004e-01 -8.78315389e-01 3.31717581e-01 1.65801588e-02
-1.12356830e+00 -1.37856267e-02 -7.57280648e-01 -3.81443053e-01
-5.07285118e-01 5.39256394e-01 -5.07934690e-01 6.80320799e-01
6.22647285e-01 9.02868986e-01 -2.39176035e-01 9.99455035e-01
7.11855367e-02 3.74864340e-01 -3.62480074e-01 -6.70413494e-01
-8.54203328e-02 -2.21849740e-01 -3.65385085e-01 7.07789183e-01
4.22329396e-01 -1.42054617e-01 1.38357118e-01 1.27823794e+00
-6.38795435e-01 3.18393618e-01 -9.33172762e-01 -5.38519084e-01
4.15516764e-01 9.39718306e-01 -6.37422979e-01 -3.80922019e-01
-5.98847389e-01 6.45176053e-01 4.36252594e-01 4.10740376e-02
-3.25419068e-01 -4.36653882e-01 7.45209634e-01 2.53625423e-01
7.60479033e-01 -3.02684426e-01 -1.19251363e-01 -6.28162444e-01
-1.87215492e-01 -6.35814548e-01 -1.22701883e-01 -6.70678675e-01
-1.60607433e+00 3.70470047e-01 -8.51987489e-03 -5.88761330e-01
-2.28352800e-01 -1.34141028e+00 -7.89123535e-01 9.63786125e-01
-1.60649097e+00 -9.95445549e-01 2.58314878e-01 4.78925377e-01
1.64553985e-01 -2.08018914e-01 1.16578794e+00 9.23610628e-02
-3.99047405e-01 2.77813375e-01 6.22778416e-01 -5.94283104e-01
5.16889155e-01 -1.16512859e+00 3.29406917e-01 -2.15164095e-01
2.48289824e-01 7.08865523e-01 4.23994094e-01 -5.82748592e-01
-1.74465764e+00 -9.03606117e-01 7.17404783e-01 -1.98390409e-01
6.33974314e-01 -4.94500607e-01 -7.85583615e-01 4.79911789e-02
1.39350057e-01 4.52427953e-01 1.17059863e+00 3.98183703e-01
-5.49779356e-01 -1.05317697e-01 -1.24262142e+00 7.74671659e-02
1.12861001e+00 -7.71853805e-01 -1.66808784e-01 9.51436520e-01
8.37316751e-01 1.05033092e-01 -9.88092124e-01 -1.11693271e-01
3.12693864e-01 -7.68226445e-01 1.26547289e+00 -8.10244918e-01
7.22398520e-01 7.40295351e-02 -6.06968045e-01 -1.20550382e+00
-3.29896450e-01 -5.30776739e-01 -3.21887672e-01 8.86803746e-01
7.90815949e-01 -6.97785079e-01 9.35491383e-01 5.56480348e-01
-4.08533305e-01 -7.61995733e-01 -1.01684225e+00 -7.84412265e-01
7.53425419e-01 -9.07269195e-02 1.03141880e+00 4.61060494e-01
-2.95596998e-02 3.08001488e-01 -3.70162725e-02 -2.66371191e-01
1.63734213e-01 5.82434118e-01 2.02269480e-01 -1.53426838e+00
-5.25928855e-01 -5.01975834e-01 -4.41436231e-01 -7.16815472e-01
4.40198630e-01 -1.28565943e+00 -1.68978572e-01 -1.45497239e+00
3.40163976e-01 -3.75931829e-01 -7.46372342e-01 2.66138196e-01
4.30759877e-01 1.12499766e-01 -6.62780702e-02 1.21051565e-01
-6.84979856e-01 1.03492057e+00 1.15957344e+00 -5.73893666e-01
3.71775292e-02 -4.24139977e-01 -8.60257864e-01 5.13878822e-01
7.81981766e-01 -3.90404075e-01 -2.62882654e-02 -7.88551643e-02
5.46377063e-01 -4.57441032e-01 -4.51697595e-02 -1.28459418e+00
1.22557551e-01 -2.73301661e-01 5.44416010e-01 -6.60815835e-01
7.20390022e-01 -4.51499522e-01 -2.36974031e-01 3.55887741e-01
-3.11981201e-01 -2.13442668e-01 4.06726412e-02 3.90300125e-01
-1.11851178e-01 -6.36157870e-01 4.88748103e-01 -5.66005290e-01
-4.53373700e-01 5.76878667e-01 -3.13858241e-01 -5.37029088e-01
1.02809393e+00 1.17379360e-01 -3.65076095e-01 2.38128439e-01
-2.02333689e-01 -6.94961786e-01 5.18388450e-01 -1.31299809e-01
6.62139714e-01 -1.31342256e+00 -4.09639269e-01 2.48493582e-01
1.23327941e-01 -2.64081713e-02 1.76461875e-01 8.36646333e-02
-6.92620575e-01 6.55689120e-01 -7.75393918e-02 -2.09407844e-02
-8.22363257e-01 9.66005087e-01 2.00444683e-01 -6.33197129e-02
-4.52830166e-01 8.86613488e-01 2.36133158e-01 -1.51599526e-01
-1.57978788e-01 -7.78452277e-01 1.23770736e-01 -6.29267693e-02
8.20301712e-01 3.08007002e-01 6.52348876e-01 -3.53209585e-01
-3.65952253e-01 5.11907339e-01 -4.99632120e-01 4.59312648e-01
1.70443213e+00 7.70221353e-01 -5.98043203e-01 6.87360644e-01
1.26809645e+00 2.74339784e-02 -9.18453157e-01 -1.56899914e-01
-6.81386217e-02 -7.77869672e-02 2.24613138e-02 -6.36621714e-01
-9.24914479e-01 1.14676642e+00 3.80597889e-01 2.14892700e-01
6.06049716e-01 3.35840702e-01 7.69922018e-01 1.01628411e+00
6.84469759e-01 -9.85085607e-01 1.74063846e-01 4.09791201e-01
8.64972293e-01 -1.11837149e+00 4.53070939e-01 -5.19445837e-01
8.31138566e-02 1.21205664e+00 1.94051087e-01 -4.02628630e-01
8.45259368e-01 9.08133090e-02 -9.66698170e-01 -2.68244892e-01
-9.63621736e-01 2.57600453e-02 2.15080187e-01 7.13580251e-01
8.24769974e-01 2.93804973e-01 -1.23238906e-01 6.27725422e-01
1.33779016e-03 -3.32259536e-01 1.15425117e-01 1.26341760e+00
-7.76962817e-01 -1.74793589e+00 -2.88824916e-01 7.65544415e-01
-3.63242865e-01 -4.95502234e-01 -7.93646455e-01 1.54702902e-01
4.33215290e-01 6.68716311e-01 8.07583556e-02 -8.67924541e-02
-2.82451093e-01 2.23605335e-01 1.06092751e+00 -1.07685637e+00
-5.57799280e-01 -4.96386290e-01 1.42534748e-01 -4.60150272e-01
-6.19557679e-01 -5.14378548e-01 -1.41124845e+00 -5.83205521e-01
-5.13569832e-01 3.72951716e-01 8.71023595e-01 9.06656504e-01
4.70530212e-01 6.72248363e-01 6.19101048e-01 -1.40785241e+00
-3.94442052e-01 -8.77022862e-01 -1.00067568e+00 1.30863801e-01
1.54436186e-01 -8.25302482e-01 -9.57508385e-02 -3.50385725e-01] | [5.183877944946289, 5.509886741638184] |
3a0bb82f-38b2-4932-8a89-a0df683bcabe | deep-rotation-equivariant-network | 1705.08623 | null | http://arxiv.org/abs/1705.08623v2 | http://arxiv.org/pdf/1705.08623v2.pdf | Deep Rotation Equivariant Network | Recently, learning equivariant representations has attracted considerable
research attention. Dieleman et al. introduce four operations which can be
inserted into convolutional neural network to learn deep representations
equivariant to rotation. However, feature maps should be copied and rotated
four times in each layer in their approach, which causes much running time and
memory overhead. In order to address this problem, we propose Deep Rotation
Equivariant Network consisting of cycle layers, isotonic layers and decycle
layers. Our proposed layers apply rotation transformation on filters rather
than feature maps, achieving a speed up of more than 2 times with even less
memory overhead. We evaluate DRENs on Rotated MNIST and CIFAR-10 datasets and
demonstrate that it can improve the performance of state-of-the-art
architectures. | ['Haifeng Liu', 'Junying Li', 'Deng Cai', 'Zichen Yang'] | 2017-05-24 | null | null | null | null | ['rotated-mnist'] | ['computer-vision'] | [-1.49238333e-01 6.57444820e-02 4.66141440e-02 -5.15687346e-01
-1.54053316e-01 -6.26957953e-01 7.02127516e-01 -5.18182158e-01
-7.22264171e-01 3.34447056e-01 2.15051651e-01 -3.47791702e-01
2.42581993e-01 -8.49348366e-01 -8.82479489e-01 -4.39985096e-01
2.02509865e-01 -9.09061059e-02 2.95200586e-01 -3.78710717e-01
5.60486056e-02 7.38319695e-01 -1.05405247e+00 2.85730034e-01
3.37905526e-01 4.32791561e-01 -3.28855887e-02 6.07398808e-01
1.91154301e-01 7.27939963e-01 -3.43930930e-01 -3.63326520e-01
4.41521972e-01 -1.48868501e-01 -8.70106280e-01 -2.23691463e-01
5.00983417e-01 -5.26203513e-01 -8.92841697e-01 1.17501819e+00
3.65664929e-01 2.25300774e-01 6.66115403e-01 -6.98351145e-01
-1.02289438e+00 9.63567674e-01 -6.66838706e-01 4.04807299e-01
-2.05659240e-01 -9.11881328e-02 9.31423068e-01 -9.78957713e-01
4.33005691e-01 1.51835871e+00 3.72786522e-01 4.98763353e-01
-9.92262423e-01 -9.05850887e-01 4.28349584e-01 2.02012137e-01
-1.30628121e+00 -3.94142210e-01 5.81821501e-01 -2.59824414e-02
1.09926140e+00 1.96764633e-01 5.93375266e-01 1.20130324e+00
3.07708412e-01 6.15639329e-01 7.46080399e-01 -1.83040306e-01
-1.38184458e-01 -3.70548397e-01 1.85388982e-01 6.97343171e-01
5.21233916e-01 -1.59150198e-01 1.28610069e-02 3.77774507e-01
1.27830887e+00 3.36408228e-01 -2.53833830e-02 -2.54630148e-01
-1.25737572e+00 8.84305179e-01 1.13311350e+00 -8.61222669e-03
-1.70591757e-01 6.81142926e-01 4.80852753e-01 4.62533444e-01
1.51632681e-01 4.96562898e-01 -3.75311077e-01 2.03565702e-01
-2.81459510e-01 1.48496404e-01 3.00373882e-01 8.98286283e-01
7.27125406e-01 5.17429113e-01 1.21092126e-01 6.55621409e-01
3.24064583e-01 3.24632317e-01 7.78927326e-01 -6.37991548e-01
3.62372458e-01 5.78685820e-01 -2.90803581e-01 -9.33296323e-01
-6.86260104e-01 -7.71797597e-01 -1.08123040e+00 2.96361968e-02
6.87426254e-02 -1.69763297e-01 -1.23081791e+00 1.75392151e+00
7.42416531e-02 3.73361379e-01 2.25687847e-01 1.07601309e+00
6.58637941e-01 5.98299325e-01 -4.69553359e-02 2.54928797e-01
1.54176927e+00 -1.14124620e+00 -5.20361304e-01 -2.07742289e-01
6.24659538e-01 -9.92264688e-01 1.05346465e+00 1.35061324e-01
-9.84602690e-01 -6.69993222e-01 -1.57022607e+00 -3.60202760e-01
-3.41233164e-01 3.56946886e-01 9.87792492e-01 5.38445532e-01
-8.60202909e-01 6.74198091e-01 -1.24102485e+00 -2.30872810e-01
2.54959911e-01 5.11855185e-01 -4.39198583e-01 1.46978632e-01
-1.09566605e+00 6.49703681e-01 4.36350733e-01 3.23997408e-01
-8.74599874e-01 -4.86760139e-01 -9.37993050e-01 3.70161891e-01
-5.00465743e-02 -5.88860452e-01 1.32552600e+00 -1.11452401e+00
-1.71796405e+00 4.78772759e-01 6.15062825e-02 -5.56271434e-01
5.66456258e-01 -5.62175393e-01 -6.84522033e-01 9.36102401e-03
-2.69785374e-01 7.17980504e-01 9.40073192e-01 -6.00155950e-01
-4.14102077e-01 -3.20843935e-01 3.86522353e-01 1.38829634e-01
-4.34171408e-01 2.24468764e-02 -5.14073670e-01 -9.96318758e-01
6.14108264e-01 -1.38308275e+00 -2.78311998e-01 -2.74690747e-01
-2.69577384e-01 -2.46414140e-01 9.65509176e-01 -3.63196492e-01
7.42203057e-01 -2.29043627e+00 3.81303839e-02 2.23583691e-02
2.83156544e-01 4.15457129e-01 -4.61525947e-01 -7.87358359e-02
-5.27317464e-01 1.60848439e-01 1.88316032e-01 1.19214058e-02
-1.30923659e-01 1.48364410e-01 -6.76980972e-01 7.38759756e-01
5.24647772e-01 8.91632199e-01 -6.40030384e-01 1.23733722e-01
2.95176029e-01 8.13837886e-01 -8.71502578e-01 -9.01967734e-02
1.38862938e-01 1.87768534e-01 -3.44703466e-01 2.67783552e-01
9.35820758e-01 -2.23336503e-01 2.99150586e-01 -4.17765319e-01
-1.34266064e-01 4.56297547e-01 -1.23772919e+00 1.64228499e+00
-5.32647192e-01 7.52168059e-01 -5.84441423e-01 -1.11116993e+00
9.77769256e-01 1.52805179e-01 9.84037593e-02 -6.48720622e-01
4.17466909e-01 -5.31941317e-02 2.42117152e-01 8.73351172e-02
7.10453749e-01 1.52072206e-01 -6.08355664e-02 3.41579944e-01
1.02664463e-01 3.24166298e-01 -9.57510024e-02 5.47697283e-02
7.69575238e-01 1.79012418e-01 1.48477450e-01 -2.67265320e-01
5.76902866e-01 -4.89016742e-01 6.57171547e-01 4.24345523e-01
-1.39341382e-02 7.19818115e-01 7.55467772e-01 -9.58751619e-01
-1.02133679e+00 -1.04120708e+00 -1.76355526e-01 1.06426632e+00
1.69564039e-01 -5.50548613e-01 -7.78930843e-01 -5.73955476e-01
-3.39715481e-01 3.11903358e-01 -8.02608669e-01 -5.18385768e-01
-1.11324692e+00 -9.89178121e-01 6.49551690e-01 9.16675150e-01
9.68832612e-01 -7.08536446e-01 -5.49966693e-01 1.51623562e-01
2.64421821e-01 -1.18739128e+00 -3.97757590e-01 1.30402669e-01
-1.06519413e+00 -9.92635548e-01 -5.29938161e-01 -8.63809645e-01
9.80301321e-01 6.18650198e-01 6.35902405e-01 -2.36625150e-01
-2.00932637e-01 -2.43032709e-01 -3.45577985e-01 -1.55049786e-01
2.10114121e-02 5.05726874e-01 2.66941786e-01 -1.18534856e-01
2.02966988e-01 -6.10660017e-01 -8.59469235e-01 4.93312478e-01
-9.71312165e-01 -1.36837712e-03 7.59624064e-01 6.92426205e-01
5.65474093e-01 -4.12142158e-01 5.00199139e-01 -8.91926527e-01
4.17162001e-01 -1.55301496e-01 -8.00260901e-01 1.52919414e-02
-3.04872364e-01 5.64394534e-01 9.21817541e-01 -4.98697251e-01
-8.87165129e-01 2.90851355e-01 -6.31127730e-02 -5.96439183e-01
2.79216528e-01 2.41172239e-01 4.70047742e-02 -1.23678379e-01
8.38524461e-01 -8.50271210e-02 -3.56274843e-01 -5.59344411e-01
7.19434321e-01 4.18678373e-01 7.45733559e-01 -3.51238281e-01
1.12635720e+00 5.62836468e-01 6.91322610e-02 -6.13335550e-01
-6.28282130e-01 5.02320342e-02 -7.38817215e-01 3.06502640e-01
8.23268831e-01 -1.39006305e+00 -5.58795691e-01 2.95136243e-01
-1.05445015e+00 6.57224208e-02 -4.85867970e-02 9.15704608e-01
-1.86562091e-01 1.19099580e-01 -6.65684462e-01 5.24148457e-02
-3.92442375e-01 -1.49262333e+00 5.36003590e-01 3.01718265e-01
-1.13592215e-03 -5.31947434e-01 1.11726709e-02 -3.19374293e-01
5.94528675e-01 1.67216167e-01 7.78159380e-01 -4.95241344e-01
-7.58825183e-01 -3.71957660e-01 -5.66507339e-01 2.98141569e-01
1.08568341e-01 1.23107933e-01 -1.12238848e+00 -6.77382767e-01
-3.08576554e-01 -2.52490610e-01 1.22709835e+00 1.22390203e-01
1.47570825e+00 -4.00032490e-01 -1.24888755e-01 1.21025050e+00
1.21288013e+00 -5.19486852e-02 7.26280153e-01 5.08493721e-01
8.28460872e-01 -3.29100564e-02 1.02529764e-01 2.00799793e-01
2.68472701e-01 4.94550496e-01 5.84816873e-01 -1.33339778e-01
-1.97089091e-01 -2.05803066e-01 4.81991023e-01 1.04267907e+00
-5.08492589e-01 9.18106884e-02 -4.53876138e-01 2.39747524e-01
-1.69939494e+00 -6.66446805e-01 5.23187704e-02 1.95593381e+00
3.68008375e-01 1.93295613e-01 -1.90654263e-01 5.58071770e-02
5.55761039e-01 4.84156609e-01 -5.74934423e-01 -7.21820414e-01
2.86717601e-02 3.68902296e-01 6.87970817e-01 2.84449279e-01
-1.34881532e+00 1.29346311e+00 6.36325598e+00 2.80543298e-01
-1.62860465e+00 -1.90454453e-01 3.34077775e-01 3.12090907e-02
-1.66594774e-01 9.46629345e-02 -7.13108182e-01 -3.74575704e-02
8.40486705e-01 -1.69337448e-02 5.31015694e-01 1.16030037e+00
-2.38318726e-01 7.14461982e-01 -1.08255446e+00 9.75887418e-01
-3.92509513e-02 -1.21566176e+00 4.99145895e-01 -5.99018559e-02
7.47149467e-01 4.14025098e-01 3.73640686e-01 6.14789069e-01
5.59084713e-01 -1.16307855e+00 5.02564549e-01 2.03139067e-01
7.63268650e-01 -1.14188111e+00 5.60424387e-01 -2.25372195e-01
-1.44531393e+00 4.36373129e-02 -1.07913840e+00 -4.97290492e-02
-2.86465913e-01 1.07206985e-01 -7.03805208e-01 4.94117141e-01
7.92404771e-01 8.98855150e-01 -6.64179862e-01 6.43829226e-01
-4.42968249e-01 4.66400146e-01 -1.30533010e-01 1.48213968e-01
5.64846992e-01 -7.44252279e-02 2.33883157e-01 1.22408867e+00
3.84137332e-01 -1.51021499e-02 -1.76367000e-01 6.08038962e-01
-6.79045975e-01 -9.80288014e-02 -5.15246868e-01 1.14774570e-01
2.95609474e-01 1.44833875e+00 -7.70183384e-01 -2.41935611e-01
-4.94584024e-01 1.06810927e+00 5.27808845e-01 3.93002689e-01
-1.09764457e+00 -8.84356797e-01 9.11045432e-01 -1.34368107e-01
3.93978477e-01 -4.82822180e-01 3.92933860e-02 -1.43114448e+00
-1.90533116e-01 -7.11165071e-01 2.48855218e-01 -4.39145774e-01
-6.59623682e-01 7.91548491e-01 -2.52055466e-01 -1.28514755e+00
-1.78267419e-01 -9.70112383e-01 -6.28066421e-01 6.73938632e-01
-1.60771823e+00 -9.78151798e-01 -4.76821542e-01 6.35851979e-01
4.58582580e-01 -2.70738065e-01 7.18989730e-01 3.51666391e-01
-9.44806397e-01 9.16373909e-01 1.31344348e-01 6.01951480e-01
8.10257077e-01 -1.07885385e+00 1.06046224e+00 1.02038574e+00
3.22479010e-01 1.13054287e+00 3.59024793e-01 -7.25393891e-02
-1.61266565e+00 -1.50979650e+00 2.99821198e-01 -1.63942516e-01
5.05502105e-01 -3.77664506e-01 -9.09950733e-01 1.31369567e+00
3.62126708e-01 4.18256879e-01 2.30868652e-01 3.14712040e-02
-9.88537192e-01 -3.46843541e-01 -6.18870437e-01 9.58209753e-01
1.07338011e+00 -5.66243887e-01 -6.01502120e-01 2.63121247e-01
9.96987462e-01 -5.91714203e-01 -5.70807993e-01 5.34180999e-01
7.09810078e-01 -6.35668755e-01 1.03595412e+00 -9.29107845e-01
2.81127095e-01 -5.08450091e-01 -1.93000749e-01 -1.41952813e+00
-6.17623448e-01 -6.29115701e-01 1.42640874e-01 7.15225458e-01
3.17053407e-01 -9.01331544e-01 4.15578723e-01 3.70554775e-02
-2.80772209e-01 -3.58403951e-01 -7.45234907e-01 -7.58212149e-01
3.07814002e-01 -1.95398360e-01 8.08366716e-01 1.11314797e+00
-2.87677109e-01 5.77585280e-01 -2.93934852e-01 3.15917760e-01
2.01194093e-01 -1.41374663e-01 9.41934586e-01 -1.06286287e+00
-7.17970505e-02 -3.41233820e-01 -6.09345853e-01 -1.15510905e+00
3.46181601e-01 -1.25691009e+00 -4.80622083e-01 -1.09077632e+00
1.11110739e-01 -1.00716382e-01 -6.02321386e-01 5.98615170e-01
6.78538978e-02 4.43751395e-01 2.27205694e-01 7.31728300e-02
-4.21072125e-01 5.96803844e-01 1.22819638e+00 -9.31896642e-02
-6.20635077e-02 -3.44242632e-01 -8.06420386e-01 1.05468464e+00
1.04705191e+00 -3.36160243e-01 -5.46685934e-01 -1.00101340e+00
3.51371080e-01 -4.67026263e-01 2.84883142e-01 -1.18161798e+00
3.85012217e-02 8.50073993e-02 9.16998446e-01 -3.88296902e-01
1.54654115e-01 -6.20376527e-01 -1.11712813e-02 7.43418336e-01
-3.08701158e-01 6.92309797e-01 2.30344743e-01 5.89635909e-01
-1.51697308e-01 6.84327707e-02 1.06290627e+00 -4.58747484e-02
-5.40069580e-01 4.15946066e-01 -2.35551506e-01 -3.53382051e-01
8.81167352e-01 2.68047333e-01 -6.57457113e-01 -1.00339111e-02
-5.32485545e-01 -6.81891963e-02 2.00384602e-01 7.73673534e-01
6.96600556e-01 -1.38570964e+00 -6.55755997e-01 4.52188998e-01
-1.11182004e-01 2.14171574e-01 1.43566251e-01 4.72535282e-01
-1.05843914e+00 5.26045322e-01 -5.93653202e-01 -4.37966049e-01
-8.81796479e-01 5.02647281e-01 4.39731061e-01 -1.08768068e-01
-9.59715486e-01 5.89787960e-01 4.26311582e-01 -4.87856984e-01
1.32782413e-02 -6.48237169e-01 -3.45255554e-01 -1.64375007e-01
5.39304852e-01 3.71901453e-01 1.00254409e-01 -6.04470491e-01
-3.94029409e-01 8.54946673e-01 -6.49874568e-01 -2.14110371e-02
1.24708378e+00 3.29281241e-01 5.10020256e-02 -1.05743542e-01
1.43116081e+00 -1.72195017e-01 -1.12732053e+00 -2.12280750e-01
-1.55006289e-01 -1.85604870e-01 1.09622523e-01 -1.95066676e-01
-1.41994393e+00 1.05023122e+00 8.26057553e-01 -2.42823958e-01
8.92750502e-01 -3.57046336e-01 5.79406738e-01 1.04906082e+00
-9.73010734e-02 -8.47405553e-01 1.81768715e-01 7.56321311e-01
9.51849997e-01 -9.91587281e-01 1.04223922e-01 -1.24017701e-01
-4.80603009e-01 1.31847155e+00 7.74800181e-01 -9.11693692e-01
6.44350827e-01 3.55631225e-02 7.03053549e-02 1.47762567e-01
-6.31598055e-01 1.28192857e-01 2.82214403e-01 2.49099687e-01
6.27772808e-01 2.13839442e-01 -4.20812041e-01 4.68036890e-01
-5.24927616e-01 -4.13300008e-01 5.73065639e-01 6.32595420e-01
-4.19107229e-01 -9.21357989e-01 -9.10098180e-02 1.29437432e-01
-6.62507534e-01 3.40028256e-02 -8.67130831e-02 9.58591819e-01
-3.84142727e-01 4.43081886e-01 3.19661379e-01 -3.98472458e-01
5.06529391e-01 -2.02098846e-01 3.34018558e-01 -3.47556114e-01
-3.68936628e-01 1.29720837e-01 -2.71549553e-01 -6.64422274e-01
-2.92516202e-01 -3.33628744e-01 -1.54366994e+00 -2.54052192e-01
-6.29703933e-03 -1.54296726e-01 6.92443311e-01 5.90378046e-01
3.39592487e-01 1.00089836e+00 8.74631405e-01 -6.25513792e-01
-7.77155519e-01 -1.13891900e+00 -2.35749036e-01 1.01114869e-01
4.16272610e-01 -4.45151687e-01 -1.80167750e-01 -1.00521989e-01] | [8.943256378173828, 2.3149805068969727] |
aebbfdd4-9c54-4006-b9ad-bd04bc2b11cb | a-corpus-to-learn-refer-to-as-relations-for | null | null | https://aclanthology.org/L18-1062 | https://aclanthology.org/L18-1062.pdf | A Corpus to Learn Refer-to-as Relations for Nominals | null | ['Kai-Wei Chang', 'Wasi Ahmad'] | 2018-05-01 | a-corpus-to-learn-refer-to-as-relations-for-1 | https://aclanthology.org/L18-1062 | https://aclanthology.org/L18-1062.pdf | lrec-2018-5 | ['learning-semantic-representations'] | ['methodology'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.436799049377441, 3.6749534606933594] |
81637978-a9db-4d5d-bee2-aff6c89a7f77 | an-empirical-study-of-training-self | 2104.02057 | null | https://arxiv.org/abs/2104.02057v4 | https://arxiv.org/pdf/2104.02057v4.pdf | An Empirical Study of Training Self-Supervised Vision Transformers | This paper does not describe a novel method. Instead, it studies a straightforward, incremental, yet must-know baseline given the recent progress in computer vision: self-supervised learning for Vision Transformers (ViT). While the training recipes for standard convolutional networks have been highly mature and robust, the recipes for ViT are yet to be built, especially in the self-supervised scenarios where training becomes more challenging. In this work, we go back to basics and investigate the effects of several fundamental components for training self-supervised ViT. We observe that instability is a major issue that degrades accuracy, and it can be hidden by apparently good results. We reveal that these results are indeed partial failure, and they can be improved when training is made more stable. We benchmark ViT results in MoCo v3 and several other self-supervised frameworks, with ablations in various aspects. We discuss the currently positive evidence as well as challenges and open questions. We hope that this work will provide useful data points and experience for future research. | ['Kaiming He', 'Saining Xie', 'Xinlei Chen'] | 2021-04-05 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Chen_An_Empirical_Study_of_Training_Self-Supervised_Vision_Transformers_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Chen_An_Empirical_Study_of_Training_Self-Supervised_Vision_Transformers_ICCV_2021_paper.pdf | iccv-2021-1 | ['self-supervised-image-classification'] | ['computer-vision'] | [ 2.46994808e-01 8.47470164e-02 -2.24396974e-01 -4.90290940e-01
-4.22594547e-01 -7.01066315e-01 8.18301141e-01 -2.72353649e-01
-3.44248831e-01 5.14410496e-01 6.07930981e-02 -2.27031946e-01
-5.56621738e-02 -2.59349495e-01 -6.87653184e-01 -8.68663907e-01
-4.37538736e-02 1.43209606e-01 5.24380207e-01 -4.14266974e-01
1.24338321e-01 2.72975981e-01 -1.73981965e+00 1.37910202e-01
7.25873291e-01 6.65314734e-01 9.64805484e-02 8.10646415e-01
1.02465257e-01 1.14738798e+00 -4.53162581e-01 -4.81811672e-01
2.23127902e-01 -4.64466989e-01 -9.04617906e-01 2.59628117e-01
9.70617056e-01 -4.49004136e-02 -3.09841216e-01 9.55575168e-01
3.78456056e-01 -1.77462220e-01 6.95807576e-01 -1.31871796e+00
-6.23926997e-01 4.21335399e-01 -4.25483495e-01 3.19532633e-01
-2.66701430e-01 5.29400170e-01 1.00513458e+00 -8.66310537e-01
7.32465208e-01 9.34945226e-01 9.24907207e-01 7.13979065e-01
-1.06904471e+00 -3.60096157e-01 1.96995094e-01 4.01845813e-01
-9.05475557e-01 -7.89853275e-01 5.61617374e-01 -6.57251060e-01
1.03692770e+00 -1.63332950e-02 5.47354817e-01 1.27028787e+00
-3.63378152e-02 9.35565829e-01 1.14580357e+00 -5.18603384e-01
1.41368628e-01 3.31511468e-01 3.70601982e-01 7.84198344e-01
3.06339085e-01 5.12326658e-01 -4.69468415e-01 1.80479616e-01
6.06166661e-01 -3.21852207e-01 -2.06595495e-01 -6.74111187e-01
-1.20353806e+00 8.39716792e-01 6.25816166e-01 4.51617122e-01
1.32051157e-02 1.81966290e-01 3.72251600e-01 4.88780737e-01
3.43421102e-01 6.34481668e-01 -4.20541286e-01 -2.38985077e-01
-1.14470112e+00 1.18538864e-01 7.63131797e-01 8.90308499e-01
8.94447505e-01 3.38012248e-01 -4.09736559e-02 7.67609060e-01
7.44334310e-02 2.14221179e-01 2.51930714e-01 -1.27139759e+00
-2.47432664e-02 3.28689516e-01 -6.95850924e-02 -6.66165829e-01
-4.86529648e-01 -6.92456961e-01 -7.00074852e-01 5.79074264e-01
6.67280912e-01 -1.96854234e-01 -1.10788023e+00 1.56731260e+00
-5.50825968e-02 2.30864048e-01 1.14089251e-01 7.84702301e-01
1.04423809e+00 1.98120356e-01 -1.31604046e-01 -1.03880502e-01
1.02385163e+00 -1.54396236e+00 -5.82615614e-01 -6.20604098e-01
4.68867868e-01 -8.55488181e-01 9.96760726e-01 4.79132056e-01
-9.31832135e-01 -5.19182503e-01 -1.25508773e+00 -2.05848762e-03
-2.42827818e-01 1.75701290e-01 9.16701138e-01 8.21320593e-01
-1.29746282e+00 8.68755162e-01 -9.74577665e-01 -8.46750915e-01
5.20014703e-01 1.59950227e-01 -2.99317777e-01 9.50374752e-02
-8.34718764e-01 1.05947542e+00 2.56113589e-01 1.73163582e-02
-9.27268863e-01 -4.56837445e-01 -7.22869575e-01 -2.30255648e-01
5.62695980e-01 -6.55223191e-01 1.59152770e+00 -1.15797722e+00
-1.34495568e+00 1.09570801e+00 -3.44924510e-01 -6.70712411e-01
5.23796797e-01 8.38665068e-02 -2.59669036e-01 1.88238978e-01
2.22483594e-02 8.42473269e-01 1.06101394e+00 -1.38123941e+00
-5.72765231e-01 -1.08390547e-01 1.29889905e-01 4.37905255e-04
-2.13406146e-01 6.37751445e-02 -5.83187222e-01 -4.48208213e-01
3.11539154e-02 -1.03671741e+00 -1.89169735e-01 2.47718245e-02
-1.36203200e-01 -1.72766224e-01 5.34081697e-01 -2.65773326e-01
8.67321968e-01 -2.11750245e+00 -8.93123746e-02 -2.03879759e-01
4.45859194e-01 4.54236299e-01 -1.59948379e-01 3.39283973e-01
-2.07115382e-01 1.25162661e-01 -4.58048195e-01 -4.82510298e-01
-1.68637127e-01 3.12403142e-01 -1.16526589e-01 4.37002242e-01
4.50848520e-01 1.08237457e+00 -9.53426957e-01 -3.65495622e-01
3.61766160e-01 3.34813476e-01 -3.11493218e-01 -5.51727302e-02
-1.12881392e-01 3.52737367e-01 -6.70799427e-03 8.23641181e-01
5.60628951e-01 -4.17642117e-01 -1.69665948e-01 -3.08106065e-01
-3.99770141e-01 2.83395439e-01 -9.24444973e-01 1.54517186e+00
4.12060358e-02 1.09806192e+00 1.03617564e-01 -1.30863380e+00
7.78917670e-01 1.12789497e-02 2.86266118e-01 -6.22151971e-01
-9.83656347e-02 9.40255076e-02 1.56428486e-01 -6.66332006e-01
4.89271730e-01 5.99789657e-02 4.96415377e-01 2.98546612e-01
3.88109207e-01 -3.99770647e-01 4.05419737e-01 3.48664105e-01
1.11361611e+00 2.78612614e-01 1.91437945e-01 -3.75594914e-01
2.37511665e-01 4.85709101e-01 3.13784361e-01 1.10212541e+00
-6.57532156e-01 9.38156486e-01 5.08292675e-01 -3.82023305e-01
-1.11105883e+00 -9.48982716e-01 -2.52382070e-01 9.62125897e-01
-5.66875525e-02 -5.67058682e-01 -6.75109446e-01 -7.46973395e-01
-2.89894909e-01 3.47526670e-01 -6.94332123e-01 -1.40824467e-01
-3.14967304e-01 -9.67533648e-01 5.54539323e-01 6.10840499e-01
4.84788895e-01 -1.10168219e+00 -4.45342720e-01 -3.39626782e-02
1.41234502e-01 -1.33037949e+00 1.04582252e-03 5.19301653e-01
-9.25296962e-01 -1.24010229e+00 -7.11949646e-01 -9.16639030e-01
4.49537784e-01 7.00072289e-01 1.34644246e+00 3.37528974e-01
4.53395909e-03 6.07510149e-01 -3.66378605e-01 -5.02158701e-01
-4.77843136e-01 2.32718691e-01 6.58452362e-02 -3.04899365e-01
4.07614201e-01 -3.87344062e-01 -6.40855253e-01 3.54378879e-01
-7.01268733e-01 -2.21155733e-02 4.52344000e-01 1.00620997e+00
2.56746262e-01 -6.87500164e-02 5.09876132e-01 -9.92993474e-01
3.83686244e-01 -3.60882372e-01 -2.53222525e-01 1.49428055e-01
-9.38979745e-01 6.31139278e-02 4.96504188e-01 -2.17776716e-01
-9.18290317e-01 2.56642848e-01 -1.61081731e-01 -2.62222975e-01
-3.83111477e-01 3.64665091e-01 3.37233543e-01 -3.81892711e-01
1.13555276e+00 9.31374058e-02 5.13691492e-02 -3.24395955e-01
4.66842800e-01 5.96283376e-01 8.00819814e-01 -5.15404284e-01
1.02887130e+00 6.78060472e-01 -3.05887938e-01 -1.01800275e+00
-9.65851724e-01 -5.05004585e-01 -5.76159298e-01 -2.20572904e-01
6.75744534e-01 -9.91088748e-01 -3.77181411e-01 5.15058398e-01
-9.72149909e-01 -5.71515381e-01 -3.97947758e-01 2.37376064e-01
-5.56145489e-01 5.95680058e-01 -5.32609940e-01 -5.72597265e-01
-1.02248497e-01 -1.28253877e+00 6.91936314e-01 4.65804547e-01
-1.06023602e-01 -1.07766676e+00 1.97881415e-01 4.08791840e-01
5.90976059e-01 -6.38050884e-02 3.26826602e-01 -4.51057881e-01
-5.56769907e-01 1.42582923e-01 -3.51660073e-01 6.77432537e-01
1.22425050e-01 4.65066820e-01 -1.34301209e+00 -4.64529574e-01
-7.02756271e-02 -7.36212552e-01 1.37646806e+00 3.20984244e-01
8.03938150e-01 1.42441660e-01 -4.84626479e-02 8.45921457e-01
1.17334282e+00 -1.27394199e-01 6.17291391e-01 7.22650230e-01
6.77400887e-01 6.89951718e-01 3.84479433e-01 -1.20877221e-01
6.62299097e-01 4.14857715e-01 5.71607888e-01 -5.08325696e-01
-4.52038378e-01 8.58448595e-02 4.06392187e-01 7.65715718e-01
-4.16277528e-01 1.04198106e-01 -1.06564701e+00 6.80142343e-01
-2.01560545e+00 -1.04655004e+00 -4.58443016e-01 1.97313571e+00
6.50487721e-01 3.55504423e-01 2.85962522e-01 1.37651995e-01
4.75157499e-01 2.76659191e-01 -7.05658317e-01 -2.22428143e-01
-4.58766520e-01 2.00061411e-01 4.16216582e-01 4.31122959e-01
-1.18890703e+00 1.23024654e+00 7.85686541e+00 4.39869612e-01
-1.31575465e+00 6.87779486e-02 4.16792691e-01 4.13216725e-02
-6.19463474e-02 3.85250539e-01 -5.92438042e-01 1.24610625e-01
6.94195151e-01 1.62129939e-01 3.59385699e-01 8.41301501e-01
-5.01986519e-02 -2.77285188e-01 -1.13381958e+00 1.06840682e+00
1.70254797e-01 -1.44308364e+00 -4.77900922e-01 -2.79086620e-01
8.28500807e-01 7.06576526e-01 1.10938512e-01 3.15793216e-01
3.70631039e-01 -1.02788699e+00 5.96856654e-01 2.55187631e-01
6.05177224e-01 -3.38028133e-01 5.78148782e-01 1.79273561e-01
-9.12946761e-01 1.24446012e-01 -6.08955562e-01 -2.64635175e-01
-3.17859530e-01 5.21853268e-01 -6.22340083e-01 5.44709384e-01
1.00221241e+00 1.20737576e+00 -1.22000825e+00 1.23162520e+00
-3.59594434e-01 1.01423168e+00 -2.39112124e-01 6.31232485e-02
3.82869303e-01 -2.22644415e-02 4.41478789e-01 1.35735691e+00
-1.40176788e-01 -4.05647248e-01 -1.64369628e-01 8.60475898e-01
2.43193984e-01 -1.76372424e-01 -6.78155303e-01 -1.33266896e-01
5.88891879e-02 1.60010445e+00 -8.24555635e-01 -2.72403151e-01
-6.57049656e-01 9.80379045e-01 5.99699378e-01 5.99232376e-01
-5.11602819e-01 -8.24503321e-03 5.15802145e-01 -1.11816019e-01
4.12804127e-01 -4.05158818e-01 -4.76213306e-01 -1.66030824e+00
3.87132987e-02 -9.90935862e-01 2.75226086e-01 -7.91686893e-01
-1.46301508e+00 5.22541523e-01 -1.92887112e-01 -1.11919785e+00
-3.15189809e-01 -8.04978967e-01 -7.42528498e-01 2.83929378e-01
-1.88594651e+00 -1.08035111e+00 -3.94116759e-01 5.44457018e-01
5.50235987e-01 -3.31012756e-01 6.72647059e-01 -1.59562100e-02
-6.81821942e-01 6.62192822e-01 1.86183795e-01 4.94996421e-02
1.03858936e+00 -1.47615588e+00 7.40067542e-01 1.00119591e+00
6.06117964e-01 5.53029358e-01 8.94008517e-01 -2.62316376e-01
-1.16107666e+00 -7.09603190e-01 6.78297162e-01 -8.05912018e-01
8.00601184e-01 -3.67393911e-01 -1.04987514e+00 7.75219023e-01
5.26812851e-01 1.05606943e-01 3.92032295e-01 3.71761829e-01
-5.51785648e-01 -5.75037710e-02 -7.16631591e-01 6.20595753e-01
1.19229114e+00 -5.74136078e-01 -6.77970052e-01 3.19281876e-01
5.29389024e-01 -2.69588679e-01 -6.37410462e-01 4.35184240e-01
4.45869774e-01 -1.42755103e+00 9.75165665e-01 -3.93342257e-01
3.74423116e-01 -2.28494540e-01 1.54958621e-01 -1.38936424e+00
-3.93536985e-01 -6.69095397e-01 -1.07365116e-01 1.17312825e+00
4.56043184e-01 -5.48068285e-01 9.88869131e-01 3.43446076e-01
-2.35889599e-01 -5.77780902e-01 -6.54338002e-01 -9.01591301e-01
3.25807333e-01 -6.41881287e-01 1.00012280e-01 1.13697445e+00
-1.47460982e-01 6.38135850e-01 -4.14314210e-01 -1.15429260e-01
7.36007333e-01 -1.84896290e-01 9.45980072e-01 -1.16247201e+00
-2.52498955e-01 -5.24870336e-01 -3.82787257e-01 -9.53843713e-01
-7.05072135e-02 -8.37008357e-01 -6.40291721e-02 -1.43542337e+00
2.42325231e-01 -3.23322326e-01 -2.38762572e-01 6.79761946e-01
-2.35884547e-01 4.18702453e-01 3.74339074e-01 4.47467983e-01
-7.27115154e-01 3.47964287e-01 1.22499287e+00 -2.35045478e-01
-1.51353925e-01 -7.54408985e-02 -9.24007177e-01 8.10495257e-01
1.09340620e+00 -2.15488657e-01 -4.49232161e-01 -4.96982157e-01
8.72776657e-02 -5.31263351e-01 5.54503024e-01 -1.26145983e+00
4.59856927e-01 -4.62469906e-02 3.00155252e-01 -5.30154586e-01
1.45632133e-01 -5.65920055e-01 -3.43571454e-01 2.74464667e-01
-9.77889374e-02 -2.56866217e-02 2.13316634e-01 2.73241520e-01
-1.71546385e-01 -3.11205387e-01 9.25309360e-01 -2.06412882e-01
-1.07272053e+00 1.33917347e-01 -2.14802772e-01 1.76771786e-02
7.67090440e-01 -2.87904263e-01 -5.90017676e-01 -4.93315846e-01
-6.92756712e-01 2.80108124e-01 7.18827546e-01 5.04669368e-01
4.24721450e-01 -1.02248728e+00 -7.55789340e-01 1.27024695e-01
4.02391702e-01 -8.65610167e-02 4.03501913e-02 8.56496692e-01
-4.04241323e-01 2.85486966e-01 -1.64548531e-01 -1.07975698e+00
-1.05941057e+00 4.43924606e-01 5.31590581e-01 4.93522398e-02
-6.43100619e-01 7.15719759e-01 3.73773798e-02 -4.09600496e-01
5.19627094e-01 -2.85571188e-01 -2.97536790e-01 -3.70352045e-02
5.30491412e-01 1.68623835e-01 2.38084361e-01 -4.10772949e-01
-3.55587214e-01 6.62900150e-01 -3.77924144e-01 -2.52996292e-02
1.40725267e+00 -2.06053838e-01 1.13656120e-02 5.62170208e-01
7.86706150e-01 -2.10275486e-01 -1.65044916e+00 -3.40560764e-01
5.29890284e-02 -8.16210136e-02 7.02664852e-02 -7.43327320e-01
-1.13240659e+00 9.97615039e-01 7.20067620e-01 2.90299326e-01
1.03334272e+00 2.84454286e-01 3.14623773e-01 5.05418539e-01
1.17482409e-01 -1.09776354e+00 1.73223600e-01 7.80493379e-01
6.64647222e-01 -1.71366274e+00 1.40899837e-01 -3.05237859e-01
-7.68947601e-01 1.14353025e+00 8.45766842e-01 -1.85204819e-01
5.65199018e-01 3.08830768e-01 4.30104733e-01 -4.63543475e-01
-8.45407069e-01 -7.50927925e-01 1.79249749e-01 9.55273330e-01
5.47557533e-01 -3.02284867e-01 -7.58602843e-02 9.77018774e-02
-3.48008633e-01 9.55014154e-02 5.98249733e-01 8.66632700e-01
-3.82466406e-01 -1.12196052e+00 -2.54537135e-01 3.04137439e-01
-2.33722061e-01 -1.43330321e-01 -6.34347081e-01 8.16357613e-01
1.47529811e-01 1.11758912e+00 -1.41910076e-01 -5.41894138e-01
2.74826616e-01 6.67242557e-02 6.17287278e-01 -5.06985366e-01
-5.49923241e-01 -1.18727289e-01 2.12136686e-01 -4.31751043e-01
-9.49432909e-01 -6.52256012e-01 -8.99878442e-01 -2.98924297e-01
-3.80565822e-01 -1.60169408e-01 6.05584323e-01 9.91779327e-01
1.35084569e-01 4.82775629e-01 5.16404748e-01 -8.55236650e-01
-4.30526137e-01 -8.05511773e-01 -1.58627033e-01 3.06278497e-01
7.24576473e-01 -5.60061455e-01 -5.90678573e-01 2.27896571e-01] | [9.660088539123535, 2.1822853088378906] |
531ef108-8941-433e-ba23-df0dc76ca2fd | hybrid-symbiotic-organisms-search-feedforward | 1906.10121 | null | https://arxiv.org/abs/1906.10121v3 | https://arxiv.org/pdf/1906.10121v3.pdf | Metaheuristics optimized feedforward neural networks for efficient stock price prediction | The prediction of stock prices is an important task in economics, investment and making financial decisions. This has, for decades, spurred the interest of many researchers to make focused contributions to the design of accurate stock price predictive models; of which some have been utilized to predict the next day opening and closing prices of the stock indices. This paper proposes the design and implementation of a hybrid symbiotic organisms search trained feedforward neural network model for effective and accurate stock price prediction. The symbiotic organisms search algorithm is used as an efficient optimization technique to train the feedforward neural networks, while the resulting training process is used to build a better stock price prediction model. Furthermore, the study also presents a comparative performance evaluation of three different stock price forecasting models; namely, the particle swarm optimization trained feedforward neural network model, the genetic algorithm trained feedforward neural network model and the well-known ARIMA model. The system developed in support of this study utilizes sixteen stock indices as time series datasets for training and testing purpose. Three statistical evaluation measures are used to compare the results of the implemented models, namely the root mean squared error, the mean absolute percentage error and the mean absolution deviation. The computational results obtained reveal that the symbiotic organisms search trained feedforward neural network model exhibits outstanding predictive performance compared to the other models. However, the performance study shows that the three metaheuristics trained feedforward neural network models have promising predictive competence for solving problems of high dimensional nonlinear time series data, which are difficult to capture by traditional models. | ['Bradley J. Pillay', 'Absalom E. Ezugwu'] | 2019-06-23 | null | null | null | null | ['stock-price-prediction'] | ['time-series'] | [-3.04973423e-01 -6.08235061e-01 -1.24301098e-01 8.72033685e-02
3.61401170e-01 -3.65671247e-01 5.43037534e-01 -7.88419098e-02
-3.69586676e-01 1.07253563e+00 -2.59046406e-01 -3.37148666e-01
-7.33195603e-01 -1.08036554e+00 -2.75117487e-01 -1.10856128e+00
-4.17787582e-01 6.08108938e-01 2.97515579e-02 -5.56410432e-01
9.88571763e-01 6.30372286e-01 -2.22391772e+00 -5.40812373e-01
9.49163020e-01 1.39426553e+00 2.52068490e-01 4.21120256e-01
-2.63025016e-01 9.25202847e-01 -8.28090131e-01 4.88183461e-03
5.93779981e-01 -3.83171380e-01 1.05541805e-02 -3.82943124e-01
-5.84040940e-01 1.05802126e-01 3.71142030e-01 9.43808317e-01
3.80096287e-01 4.41991776e-01 4.40538734e-01 -1.30436182e+00
-7.87708223e-01 5.17963648e-01 -3.12655956e-01 8.49787831e-01
-3.76481205e-01 3.89773995e-02 6.24091148e-01 -6.50355518e-01
-2.97450032e-02 7.99507678e-01 7.18511462e-01 -1.05083864e-02
-7.18529880e-01 -8.45501959e-01 -3.77542973e-01 2.31147543e-01
-1.04015470e+00 1.89851180e-01 7.41518676e-01 -2.95135498e-01
1.28783357e+00 5.33592343e-01 1.26745343e+00 2.32910980e-02
4.82458591e-01 2.15492040e-01 1.47597504e+00 -5.77599227e-01
5.26580155e-01 3.15058917e-01 3.16114783e-01 2.49134570e-01
8.39874446e-01 7.58101404e-01 -1.84693888e-01 -3.98735821e-01
6.39608979e-01 4.52362522e-02 -1.34938315e-01 4.11568999e-01
-5.39018750e-01 1.16178834e+00 2.16325432e-01 1.03704596e+00
-9.98902917e-01 -1.87147796e-01 1.87501252e-01 3.13841999e-01
6.96272612e-01 8.05903196e-01 -6.54049397e-01 -3.92746389e-01
-1.11629474e+00 4.08225924e-01 1.00798118e+00 3.98528203e-02
2.08898574e-01 9.38061833e-01 4.88485068e-01 5.24832547e-01
3.38928133e-01 8.90310645e-01 1.15618038e+00 -4.03293043e-01
-8.02003369e-02 7.38937974e-01 2.07187355e-01 -1.47627020e+00
-5.23483396e-01 -7.99973726e-01 -5.27818322e-01 5.39847195e-01
5.47047108e-02 -3.69681776e-01 -4.91136461e-01 9.12855625e-01
8.49763453e-02 4.35312003e-01 4.83637184e-01 5.14973581e-01
2.36829758e-01 1.16244876e+00 -2.14560196e-01 -7.80168176e-01
8.75674427e-01 -8.28285813e-01 -8.76134634e-01 2.51255542e-01
1.23337552e-01 -7.20496953e-01 7.22635686e-02 4.21912193e-01
-1.15458262e+00 -5.15617430e-01 -1.03202736e+00 1.11573923e+00
-8.61787438e-01 2.12751925e-02 5.18206179e-01 7.90154159e-01
-1.01726925e+00 9.20732200e-01 -7.24186063e-01 1.82239145e-01
-2.86678880e-01 5.79809666e-01 4.98420149e-01 8.05532813e-01
-1.24308491e+00 1.21748006e+00 9.34467554e-01 4.09819543e-01
4.69736785e-01 -6.18671954e-01 -2.50534207e-01 2.73634702e-01
-1.06036328e-01 -3.02523166e-01 6.96599603e-01 -1.13320112e+00
-1.45338333e+00 2.24795654e-01 2.83151239e-01 -9.02581573e-01
7.14504868e-02 2.84957796e-01 -6.48896396e-01 -2.35559508e-01
-2.69301564e-01 -1.36318803e-01 6.20041311e-01 -8.86756182e-01
-9.82094646e-01 -2.78816223e-01 -5.38836896e-01 3.41982115e-03
-4.19865608e-01 1.26265824e-01 5.89884698e-01 -1.12106359e+00
-6.01619855e-02 -8.16839397e-01 -4.85912114e-02 -8.40470791e-01
4.01900530e-01 -3.52116853e-01 8.95344436e-01 -8.02712917e-01
1.35356367e+00 -1.59250057e+00 -1.76720574e-01 8.99479568e-01
-5.37113786e-01 5.86752653e-01 2.87691206e-01 4.97047722e-01
-2.81687945e-01 8.76298025e-02 -1.62765995e-01 3.50076437e-01
-5.13055883e-02 4.58677024e-01 -3.06786090e-01 3.05995587e-02
8.94310400e-02 5.29837549e-01 -3.62655044e-01 1.71647109e-02
2.32177228e-01 3.12502891e-01 -1.63411144e-02 7.17304200e-02
-1.15942344e-01 -3.31430972e-01 -4.83494371e-01 6.73524857e-01
7.04924762e-01 -1.35684147e-01 -2.59943694e-01 5.60059659e-02
-8.94297183e-01 -3.03352863e-01 -1.30751503e+00 2.32058570e-01
-7.09351227e-02 3.66529912e-01 -4.05474603e-01 -1.24817705e+00
1.58529341e+00 3.73602509e-01 6.90723360e-01 -8.07796836e-01
4.74449515e-01 8.62559915e-01 2.21492380e-01 -3.96997154e-01
5.85442960e-01 -2.79987901e-01 5.53183496e-01 6.65252864e-01
-5.02833903e-01 1.11959696e-01 5.74138641e-01 -8.01253796e-01
2.61742741e-01 -6.05925098e-02 2.62053668e-01 -4.22810823e-01
7.44870126e-01 3.12448382e-01 5.50885439e-01 1.57397375e-01
6.33362830e-02 -1.27199665e-01 2.27233116e-02 -9.63947117e-01
-7.22343206e-01 -2.64441997e-01 -2.49727577e-01 7.28750408e-01
-1.53232008e-01 5.23710728e-01 -4.20187771e-01 1.92724451e-01
2.49368995e-01 1.10126197e+00 -5.76186001e-01 -5.00658229e-02
-5.03196895e-01 -1.47382295e+00 2.21213937e-01 2.22220689e-01
8.27407122e-01 -1.32955170e+00 -1.11789596e+00 5.63255489e-01
4.27180976e-01 -3.29267830e-01 3.72197330e-01 3.89941096e-01
-1.05505180e+00 -1.06316876e+00 -9.46758032e-01 -6.90190613e-01
3.29358965e-01 -3.52881178e-02 1.00209546e+00 5.81612170e-01
2.07824409e-02 1.40294194e-01 -5.35339773e-01 -1.09400392e+00
-3.37402791e-01 -2.07164556e-01 6.05747402e-02 -2.39001960e-02
4.86257583e-01 -5.35275042e-01 -2.91273206e-01 1.23036996e-01
-6.64604187e-01 -4.80379224e-01 5.48332810e-01 8.87569785e-01
4.76858824e-01 1.02735519e+00 8.45689058e-01 -1.43943205e-01
1.05164611e+00 -6.84743226e-01 -1.30826807e+00 4.26293254e-01
-1.37519038e+00 -2.20401995e-02 3.72270107e-01 -2.65957177e-01
-8.91439795e-01 -5.80984116e-01 1.12549476e-01 -2.74814874e-01
3.40292424e-01 1.02467453e+00 8.16066980e-01 -4.12675142e-01
2.16854095e-01 7.63415754e-01 5.16386807e-01 -4.12214875e-01
-7.94575453e-01 4.10429060e-01 2.38863170e-01 -1.54495195e-01
7.99257994e-01 -1.83987007e-01 2.26387411e-01 -8.73877645e-01
-1.81683496e-01 -3.21630567e-01 -1.07112110e-01 -4.63440955e-01
6.00192487e-01 -3.08859825e-01 -7.58612275e-01 9.65671897e-01
-7.58261442e-01 8.37882087e-02 -2.82099158e-01 5.44163525e-01
-2.89572001e-01 -4.10612553e-01 -3.05627644e-01 -1.49869275e+00
-8.00072610e-01 -9.05168653e-01 1.96197331e-01 7.34671414e-01
-2.22115349e-02 -1.45056713e+00 3.26230317e-01 -1.64101347e-02
9.20916259e-01 6.09184444e-01 8.95173728e-01 -1.01843345e+00
-3.21599483e-01 -4.50005233e-01 2.80017614e-01 4.86070514e-01
1.94945559e-01 5.13722599e-01 -4.36484158e-01 -4.02136222e-02
5.15858650e-01 1.59877732e-01 2.28260875e-01 8.04787159e-01
3.37031960e-01 -5.17292380e-01 8.55000224e-03 4.22854006e-01
1.66981936e+00 1.40856886e+00 5.54605484e-01 1.23429239e+00
-1.48056403e-01 5.43179274e-01 7.57692754e-01 6.57839179e-01
2.44317111e-02 8.85130465e-02 3.16794574e-01 2.84916371e-01
8.84388030e-01 4.35150355e-01 1.04358003e-01 1.03324234e+00
-7.60706484e-01 -1.46709219e-01 -9.30830479e-01 3.85804288e-02
-1.62827885e+00 -1.47708166e+00 4.66297157e-02 1.90225542e+00
4.04853106e-01 9.39515382e-02 2.94984221e-01 8.45153689e-01
6.59119546e-01 2.89272182e-02 -4.36089277e-01 -6.81041896e-01
-5.67562282e-01 4.54756171e-01 6.20275557e-01 4.04790193e-02
-9.38126147e-01 1.98221281e-01 5.61884928e+00 4.85914409e-01
-1.70125735e+00 -5.27663529e-01 6.01842523e-01 7.59174675e-02
-7.43493512e-02 -4.37659770e-01 -7.01646447e-01 9.88877952e-01
1.34196997e+00 -7.93958306e-01 4.01675880e-01 7.34278858e-01
5.83100080e-01 -3.65053415e-01 6.12665415e-02 8.41116309e-01
-1.22392043e-01 -1.62318242e+00 -3.67461778e-02 2.37876177e-01
9.52844620e-01 -1.64055631e-01 2.06787258e-01 6.72779381e-02
6.68404112e-03 -7.39468336e-01 6.25268042e-01 1.03245533e+00
-5.52470088e-01 -1.30133235e+00 1.31656229e+00 5.44264436e-01
-1.08536863e+00 -5.42182565e-01 -3.90750408e-01 -3.18751276e-01
2.74015358e-03 3.04304659e-01 -3.87587339e-01 5.68678796e-01
8.73451114e-01 4.82652187e-01 -2.21115649e-01 1.41669536e+00
6.63251638e-01 5.53214610e-01 -4.89635408e-01 -8.29838037e-01
4.36775774e-01 -8.53627443e-01 4.28368568e-01 3.86001796e-01
9.60125327e-01 2.50468075e-01 -2.38498479e-01 7.47694433e-01
7.97091126e-01 4.64253902e-01 -2.54140973e-01 -4.65862721e-01
6.45820200e-01 6.04455888e-01 -8.61378789e-01 -2.90198833e-01
-1.03136152e-01 7.62023032e-02 -5.52550614e-01 -6.22380106e-03
-6.29800141e-01 -4.07573760e-01 4.94288445e-01 -1.34961262e-01
3.67528498e-01 6.41073957e-02 -6.47864401e-01 -6.13107502e-01
-1.54362753e-01 -7.21764624e-01 3.15070748e-01 -8.29708278e-01
-1.14391255e+00 7.40522027e-01 1.80023327e-01 -1.24164224e+00
-8.39643538e-01 -7.80265212e-01 -9.70965147e-01 1.27015233e+00
-1.55109751e+00 -3.65295768e-01 9.12437662e-02 2.43884221e-01
3.12724918e-01 -1.13542628e+00 8.09959650e-01 -3.05156037e-02
-7.35149384e-01 -9.48607624e-02 6.69949651e-01 -1.31216481e-01
-2.89665937e-01 -9.63200152e-01 4.07938994e-02 7.43945897e-01
-3.81786466e-01 4.22913015e-01 9.34371352e-01 -9.93512988e-01
-1.23074341e+00 -5.98201692e-01 8.83056819e-01 6.75864220e-01
8.76857102e-01 8.45940351e-01 -1.07163823e+00 7.92234391e-02
3.41337472e-01 -3.55560958e-01 5.64849973e-01 -5.42151213e-01
4.81025487e-01 -2.09643424e-01 -1.34237123e+00 1.26918793e-01
-2.29763001e-01 3.55262846e-01 -7.88176060e-01 -2.59837657e-02
1.95535436e-01 1.56737007e-02 -1.08586776e+00 3.21174622e-01
6.22602463e-01 -1.20374727e+00 9.47765529e-01 -1.84430346e-01
-8.81840102e-03 -4.81196195e-02 1.70462862e-01 -1.41274214e+00
-3.44788641e-01 -4.01560545e-01 -4.61179316e-02 1.13969791e+00
5.92307329e-01 -1.53546607e+00 9.36471045e-01 9.78681207e-01
1.34083152e-01 -9.02873099e-01 -7.94186652e-01 -1.05310118e+00
-5.82523532e-02 1.56319708e-01 1.13928199e+00 9.84936833e-01
-3.45687866e-01 -3.45809013e-01 -7.82047436e-02 1.16806431e-03
5.70860326e-01 3.64874363e-01 2.49962151e-01 -1.64283514e+00
-1.88277140e-01 -9.98528242e-01 -3.91229540e-01 4.30616587e-02
3.53541365e-03 -2.43400961e-01 -3.91862541e-01 -1.15457261e+00
-6.50522053e-01 -6.28099561e-01 -5.19471526e-01 1.94872126e-01
-4.86337207e-02 -1.25640869e-01 4.36602682e-01 5.28335154e-01
7.09442556e-01 4.96780902e-01 9.22318816e-01 2.02341497e-01
-4.88463908e-01 6.18814826e-01 -3.37931305e-01 6.66383862e-01
1.21814859e+00 -2.85622299e-01 -4.94163722e-01 1.12393901e-01
3.75395596e-01 2.67964870e-01 -3.97795737e-02 -1.23730326e+00
2.63102978e-01 -5.29766738e-01 3.76240313e-01 -9.05300081e-01
1.97444662e-01 -1.13544357e+00 8.09436262e-01 1.15693557e+00
1.74466953e-01 1.02642095e+00 3.53765041e-01 1.37442157e-01
-6.97444677e-01 -8.07662964e-01 5.02385855e-01 -1.33883700e-01
-9.67201829e-01 -4.01206277e-02 -4.37830001e-01 -3.23747367e-01
1.42856264e+00 -9.09541905e-01 -2.27178171e-01 -7.43519887e-02
-3.94021571e-01 8.22302923e-02 -4.70709167e-02 2.62383431e-01
5.05744517e-01 -1.18677473e+00 -4.99540925e-01 1.18491627e-01
-7.03877568e-01 -5.84656477e-01 -1.64308041e-01 7.20765710e-01
-1.16340768e+00 6.03611290e-01 -7.10465074e-01 -1.02752537e-01
-1.20137572e+00 3.47576797e-01 7.89795935e-01 -3.00181448e-01
2.56539811e-03 6.33869171e-01 -8.61322641e-01 9.04426426e-02
-2.89983433e-02 -3.20981681e-01 -8.27536166e-01 5.93971252e-01
6.81631148e-01 9.61435437e-01 9.76130962e-02 -7.07004786e-01
-1.07904151e-01 8.12705874e-01 4.73037779e-01 7.88353682e-02
2.08926630e+00 2.09148288e-01 -5.46519578e-01 5.77503979e-01
7.77586460e-01 -4.43938911e-01 -7.64361620e-01 1.02008313e-01
5.53823829e-01 -2.51850277e-01 3.33067060e-01 -9.61202264e-01
-1.24698067e+00 1.52991608e-01 7.32731402e-01 1.04936290e+00
1.48096693e+00 -1.12327778e+00 8.20927560e-01 4.26274747e-01
3.26032162e-01 -1.35025108e+00 -5.16518950e-01 5.86458743e-01
9.60979998e-01 -1.09843409e+00 -1.03121344e-02 -9.65684205e-02
-3.41602653e-01 1.54515660e+00 1.89548284e-01 -4.58764553e-01
1.10141563e+00 6.12084448e-01 3.23539190e-02 -7.92191643e-03
-8.19783926e-01 -4.53650318e-02 2.33679071e-01 3.65250856e-01
1.96004927e-01 -2.59131622e-02 -6.38115466e-01 5.43178737e-01
-6.13438308e-01 4.52090144e-01 3.39353472e-01 1.27754664e+00
-7.30731726e-01 -8.83599222e-01 -9.86653030e-01 6.04759037e-01
-6.17730975e-01 -7.09600002e-02 -1.74212649e-01 1.17201412e+00
-1.35682821e-02 7.24889398e-01 5.71497917e-01 -3.17747295e-01
3.17331940e-01 1.56911984e-01 -3.99412736e-02 1.57448575e-01
-1.13682485e+00 -6.64935708e-02 -1.82206973e-01 2.98870534e-01
-8.38020444e-01 -9.27370012e-01 -1.22917116e+00 -6.88847363e-01
-4.30288792e-01 8.82000923e-01 1.03129303e+00 1.01782489e+00
-1.69652831e-02 3.94758344e-01 1.00135016e+00 -1.14391112e+00
-7.99534380e-01 -8.19741726e-01 -8.47401559e-01 -2.69631714e-01
1.10000685e-01 -7.84635305e-01 -5.68121970e-01 -1.11877888e-01] | [4.810681343078613, 4.015768527984619] |
3a0ff4bc-e7f8-4fa6-83aa-924c7484c010 | scalenet-guiding-object-proposal-generation | 1704.06752 | null | http://arxiv.org/abs/1704.06752v1 | http://arxiv.org/pdf/1704.06752v1.pdf | ScaleNet: Guiding Object Proposal Generation in Supermarkets and Beyond | Motivated by product detection in supermarkets, this paper studies the
problem of object proposal generation in supermarket images and other natural
images. We argue that estimation of object scales in images is helpful for
generating object proposals, especially for supermarket images where object
scales are usually within a small range. Therefore, we propose to estimate
object scales of images before generating object proposals. The proposed method
for predicting object scales is called ScaleNet. To validate the effectiveness
of ScaleNet, we build three supermarket datasets, two of which are real-world
datasets used for testing and the other one is a synthetic dataset used for
training. In short, we extend the previous state-of-the-art object proposal
methods by adding a scale prediction phase. The resulted method outperforms the
previous state-of-the-art on the supermarket datasets by a large margin. We
also show that the approach works for object proposal on other natural images
and it outperforms the previous state-of-the-art object proposal methods on the
MS COCO dataset. The supermarket datasets, the virtual supermarkets, and the
tools for creating more synthetic datasets will be made public. | ['Wei Shen', 'Weichao Qiu', 'Alan Yuille', 'Siyuan Qiao', 'Chenxi Liu'] | 2017-04-22 | scalenet-guiding-object-proposal-generation-1 | http://openaccess.thecvf.com/content_iccv_2017/html/Qiao_ScaleNet_Guiding_Object_ICCV_2017_paper.html | http://openaccess.thecvf.com/content_ICCV_2017/papers/Qiao_ScaleNet_Guiding_Object_ICCV_2017_paper.pdf | iccv-2017-10 | ['object-proposal-generation'] | ['computer-vision'] | [-3.17662060e-02 1.57435000e-01 -1.28005460e-01 -4.11952317e-01
-5.77052474e-01 -4.75074083e-01 4.57228422e-01 1.79776624e-01
-3.54461670e-01 3.40261638e-01 -3.35578889e-01 2.16873363e-01
7.13084787e-02 -9.81895983e-01 -9.22672153e-01 -3.88198853e-01
-2.08951727e-01 6.63722217e-01 1.07906938e+00 -4.43612665e-01
3.78526926e-01 4.06140625e-01 -1.67864156e+00 4.02446508e-01
6.47150755e-01 1.36441839e+00 4.07317042e-01 5.37301838e-01
1.64253488e-02 3.68251532e-01 -4.57108259e-01 -6.26352370e-01
7.47930169e-01 -1.11085825e-01 -5.88555872e-01 5.80175340e-01
6.96414173e-01 -4.38747406e-01 7.25887716e-02 1.19992042e+00
3.32040250e-01 -1.33296162e-01 4.61633593e-01 -1.59488606e+00
-7.03022778e-01 7.32951880e-01 -7.10081756e-01 1.73866585e-01
3.26602943e-02 1.85653176e-02 9.86549139e-01 -9.58441556e-01
8.30204368e-01 1.49759936e+00 3.84915382e-01 1.07930161e-01
-1.09143162e+00 -7.93204844e-01 1.68338045e-01 2.07329020e-01
-1.50875139e+00 1.19082600e-01 8.26673985e-01 -2.95027822e-01
3.47376466e-01 2.50387728e-01 8.21695447e-01 6.42561972e-01
-3.71809397e-03 1.32413197e+00 1.17160082e+00 -3.24933618e-01
3.50007027e-01 5.58852732e-01 -1.25172243e-01 4.09539312e-01
2.50452697e-01 1.66456059e-01 -3.32914531e-01 -1.10173136e-01
1.17031312e+00 -4.10347953e-02 1.80491909e-01 -8.81688893e-01
-1.60152972e+00 1.00052392e+00 8.77956390e-01 2.29108319e-01
-3.73543710e-01 1.09891467e-01 2.91298240e-01 2.36693844e-01
6.56054676e-01 3.54305089e-01 -5.54216444e-01 4.92300570e-01
-7.93074071e-01 8.34268510e-01 5.78921497e-01 1.41364515e+00
5.06869018e-01 -3.16519171e-01 -5.56179322e-02 6.45946026e-01
4.13328111e-01 3.65452528e-01 3.73161107e-01 -7.56247044e-01
5.42401075e-01 8.47859502e-01 6.46043837e-01 -8.62290621e-01
-4.44037855e-01 -2.85483629e-01 -4.07516152e-01 4.21185046e-02
6.32830262e-01 2.80583084e-01 -1.01524830e+00 1.32630360e+00
6.38064146e-01 7.14762649e-03 -1.83372647e-01 1.08971822e+00
1.00588751e+00 6.11473322e-01 1.51584717e-02 -1.57155842e-01
1.54268670e+00 -1.16259134e+00 -5.02261460e-01 -4.94486019e-02
3.31141382e-01 -1.34693098e+00 1.05833328e+00 6.01798654e-01
-1.10459793e+00 -1.04752970e+00 -9.14303064e-01 2.01692626e-01
-3.79434139e-01 4.83144432e-01 8.82433534e-01 5.01386285e-01
-7.02256680e-01 4.96262908e-01 -6.45104647e-01 -7.11136937e-01
3.04481775e-01 2.35641509e-01 -3.86948250e-02 3.72823358e-01
-8.54902506e-01 7.97669113e-01 9.60838735e-01 -1.67059079e-01
-9.04462457e-01 -3.73171210e-01 -5.77118456e-01 -1.42778695e-01
5.58758378e-01 -3.10778648e-01 1.36611378e+00 -1.22731864e+00
-1.11998463e+00 9.85471129e-01 5.32219410e-01 -4.97145385e-01
7.67876089e-01 -9.00580958e-02 -4.98612970e-01 1.75357148e-01
2.69073159e-01 1.42379355e+00 1.04176748e+00 -1.41870070e+00
-1.08624363e+00 -3.22945088e-01 4.21925783e-01 -7.59120146e-03
-3.95787470e-02 1.36016458e-01 -6.23728931e-01 -9.88323569e-01
5.73995113e-01 -1.21675777e+00 -3.68282706e-01 1.23324506e-01
-4.30223525e-01 -5.79861343e-01 6.59485757e-01 -2.63741910e-01
8.50280106e-01 -2.00572944e+00 -1.95281640e-01 1.17611147e-01
-1.72818914e-01 4.96594831e-02 -3.56531382e-01 2.03453124e-01
5.13861254e-02 1.56248501e-02 3.15928727e-01 -2.38951594e-02
2.64160316e-02 -4.78654206e-02 -1.05750710e-01 3.08005095e-01
2.69302189e-01 8.32405865e-01 -7.39800036e-01 -8.99838984e-01
2.43027329e-01 2.50917096e-02 -4.60015059e-01 1.87436104e-01
-3.58755499e-01 1.96801648e-01 -6.58216119e-01 8.27065408e-01
1.13278496e+00 -2.22777203e-01 1.55302426e-02 -5.61059773e-01
-1.20028168e-01 -1.52415708e-01 -1.83313334e+00 1.63493252e+00
1.52564868e-02 1.14552610e-01 -2.29235664e-01 -7.30580926e-01
1.09139228e+00 1.11584283e-01 2.88232535e-01 -6.00681186e-01
1.48312047e-01 2.95399040e-01 3.38845402e-02 -4.22134072e-01
9.36797321e-01 -3.64062190e-03 -5.36681041e-02 2.90464938e-01
2.63060983e-02 -7.25036144e-01 9.24506068e-01 2.38588437e-01
3.36660594e-01 3.59276593e-01 1.69522971e-01 -3.14394087e-01
2.78116435e-01 1.24504790e-01 4.73070264e-01 5.88099658e-01
-3.87051612e-01 8.63980651e-01 4.93699640e-01 -8.32819819e-01
-1.66291678e+00 -1.21058202e+00 -3.21228653e-01 1.26634800e+00
7.19548583e-01 -2.45859087e-01 -9.25856709e-01 -8.89591575e-01
1.36830106e-01 4.99182880e-01 -7.43719578e-01 5.23396730e-01
-5.81095397e-01 -6.66428149e-01 -2.30002761e-01 7.37144351e-01
6.20252371e-01 -1.38941562e+00 -4.90025103e-01 2.22100556e-01
-1.88955411e-01 -1.25873542e+00 -5.32368898e-01 -3.38448942e-01
-1.11636078e+00 -1.28414106e+00 -8.62922251e-01 -9.74871337e-01
7.42394269e-01 4.44382638e-01 1.25259995e+00 -1.13242693e-01
-3.03623527e-01 1.49725964e-02 -6.15071476e-01 -8.61137211e-01
-5.64197302e-01 3.04324776e-01 -1.02197677e-01 1.07890323e-01
2.08632112e-01 -1.58538952e-01 -9.81899917e-01 9.95755255e-01
-1.21730757e+00 2.47022182e-01 7.22259283e-01 5.21047831e-01
7.73549914e-01 1.90137967e-01 5.93235791e-01 -7.42271423e-01
3.05121541e-01 -9.18531865e-02 -1.02199423e+00 2.88392603e-01
-3.56791437e-01 -3.95892598e-02 1.65294513e-01 -8.39059651e-01
-8.52518141e-01 3.15404773e-01 1.27687588e-01 -5.09911403e-02
-6.74831197e-02 -2.14715853e-01 8.31628684e-04 -3.59718055e-02
7.58334994e-01 -1.73235476e-01 -2.33286396e-01 -6.46031439e-01
4.95147258e-01 6.01254284e-01 -5.31303138e-02 -6.02335930e-02
9.11898851e-01 5.88300169e-01 -2.44272113e-01 -6.35045767e-01
-7.90567636e-01 -6.76022053e-01 -7.94436514e-01 -1.36301816e-01
6.11236155e-01 -8.86309326e-01 -6.15937948e-01 2.55950540e-01
-1.12310445e+00 -6.15089461e-02 -5.50629795e-01 5.76955736e-01
-6.66934848e-01 8.79611224e-02 -5.85755467e-01 -5.77266872e-01
-2.23655835e-01 -1.28703594e+00 1.45616066e+00 3.05355132e-01
1.01897582e-01 -4.64308828e-01 -2.85066664e-01 4.37500238e-01
3.96507561e-01 1.68813437e-01 4.43185806e-01 -6.78568363e-01
-9.09850180e-01 -2.70262837e-01 -5.44079661e-01 3.44848573e-01
-2.46468961e-01 -8.18111971e-02 -6.76168382e-01 -3.34284514e-01
-9.00741965e-02 -4.75277543e-01 7.51607358e-01 4.09748465e-01
9.53028917e-01 -5.55519313e-02 -3.41506183e-01 3.14825252e-02
1.41186762e+00 -5.69704287e-02 5.69766641e-01 5.86629391e-01
2.32218415e-01 6.04118347e-01 1.33743560e+00 3.61946851e-01
4.10055578e-01 8.53284299e-01 8.47343504e-01 -2.96201587e-01
7.72624183e-03 -2.65017301e-01 -1.19555980e-01 3.70283723e-01
-1.16383687e-01 -7.90219083e-02 -2.69020468e-01 6.68800533e-01
-1.99134147e+00 -7.31258154e-01 -2.35322163e-01 2.03045487e+00
5.35614967e-01 4.14431989e-01 6.88121557e-01 1.11369856e-01
8.44913900e-01 5.71187101e-02 -5.37257791e-01 -4.44062613e-02
-1.51361059e-03 -6.84936121e-02 7.48890162e-01 -1.48105830e-01
-1.53605497e+00 1.04093778e+00 6.47735262e+00 9.09336030e-01
-9.22332227e-01 3.80101949e-01 7.02209651e-01 4.79191914e-02
-2.92045493e-02 2.29551420e-02 -1.01237833e+00 2.02999607e-01
3.02265704e-01 8.06705728e-02 6.78465664e-02 1.40963602e+00
1.49928913e-01 -2.20824838e-01 -9.24267650e-01 8.42944443e-01
8.37306008e-02 -1.06531584e+00 -5.09687662e-02 -1.37774199e-01
9.10117328e-01 -5.83184175e-02 -1.62630573e-01 2.55003393e-01
1.06711820e-01 -4.02779996e-01 1.25584769e+00 1.36562198e-01
1.94526374e-01 -6.72123015e-01 8.75423074e-01 3.61418426e-01
-1.42234671e+00 -1.42184556e-01 -8.83932829e-01 3.31858963e-01
3.25554013e-01 3.51567477e-01 -9.08981860e-01 2.41854340e-01
8.14367652e-01 3.19553643e-01 -1.10042226e+00 1.49423015e+00
-2.97191828e-01 3.74502152e-01 -4.92117792e-01 -2.43156999e-01
2.22477287e-01 -6.69836700e-02 2.51143515e-01 8.89649987e-01
3.72888029e-01 -8.54830593e-02 5.32957673e-01 1.02792788e+00
-9.74837504e-03 5.23134947e-01 -2.46612161e-01 1.47697017e-01
7.97969401e-02 1.65691078e+00 -1.36207080e+00 -5.57235122e-01
-3.58635932e-01 7.36554027e-01 -4.32490520e-02 8.88207704e-02
-9.80318725e-01 -2.13704839e-01 -1.16340965e-01 5.12216508e-01
6.78391039e-01 -6.70862347e-02 1.94297895e-01 -1.04697168e+00
1.98794529e-01 -7.60685384e-01 1.85075387e-01 -1.03606617e+00
-1.23008180e+00 6.27095819e-01 4.03078198e-01 -1.62819338e+00
1.47872753e-02 -5.81321299e-01 -1.96210533e-01 2.01090440e-01
-1.34730232e+00 -1.69438434e+00 -4.35670078e-01 2.04805866e-01
9.61033463e-01 -1.99524444e-02 2.68655956e-01 1.22437410e-01
-5.57033904e-02 1.85859814e-01 -2.44427547e-02 1.16658635e-01
7.26060688e-01 -1.16308641e+00 6.82506442e-01 4.65532511e-01
3.43391210e-01 3.16107661e-01 8.27475309e-01 -6.18516088e-01
-1.19599831e+00 -1.00423801e+00 5.81666470e-01 -3.90900195e-01
7.67336547e-01 -5.45698225e-01 -5.45527160e-01 5.49480498e-01
2.49145955e-01 3.14673871e-01 6.30785003e-02 -2.24749431e-01
3.42303365e-02 -2.78112084e-01 -1.23074818e+00 2.16299653e-01
9.21623886e-01 3.16956013e-01 -7.00412750e-01 8.17443728e-01
7.53188908e-01 -4.58513856e-01 -8.97196472e-01 5.66765606e-01
6.43062592e-01 -1.07204270e+00 9.45638299e-01 -2.75712281e-01
2.17270732e-01 -4.13355500e-01 -7.82089382e-02 -1.06280375e+00
-1.45189062e-01 -1.42802119e-01 3.16665262e-01 1.13537908e+00
5.12338519e-01 -3.64721626e-01 1.09746253e+00 3.10422510e-01
2.40649283e-01 -4.65839118e-01 -6.63521469e-01 -9.91371572e-01
-3.69616538e-01 -3.21547419e-01 7.26571143e-01 5.31131029e-01
-5.12554526e-01 2.91550398e-01 -9.79294032e-02 1.20522920e-02
7.75206447e-01 4.78893548e-01 1.16095185e+00 -1.18189454e+00
-2.27984264e-01 -1.96200967e-01 -6.75887525e-01 -1.00375557e+00
-4.46718901e-01 -4.84617889e-01 -7.48836175e-02 -1.49485505e+00
4.63977605e-01 -6.15796626e-01 -1.03915937e-01 1.09035425e-01
-1.25048295e-01 8.19034576e-01 6.54939055e-01 3.12892199e-01
-8.72441113e-01 2.73780167e-01 1.81983900e+00 -1.64140910e-01
-4.25675601e-01 3.67164284e-01 -4.20157522e-01 9.12381768e-01
7.32469201e-01 -4.98086721e-01 -3.86391670e-01 9.80825946e-02
3.47901911e-01 -2.09761873e-01 3.33617240e-01 -9.66111958e-01
-2.33979702e-01 -1.72865927e-01 4.17768210e-01 -1.09467804e+00
2.37198800e-01 -7.19783008e-01 2.81364396e-02 5.55281520e-01
-3.31395686e-01 2.80332081e-02 -1.89020280e-02 3.83789629e-01
-1.82686225e-01 -3.39980990e-01 9.10264790e-01 -4.16035771e-01
-8.24154437e-01 2.93338746e-01 2.02214465e-01 -7.81546906e-02
1.39772451e+00 -2.15912446e-01 -1.57177269e-01 -1.37108847e-01
-8.64281297e-01 1.21494658e-01 4.61874753e-01 8.22470546e-01
5.13397872e-01 -1.45460331e+00 -8.50492954e-01 2.25825250e-01
4.75517869e-01 -3.14273611e-02 1.26583064e-02 7.90252030e-01
-8.28547180e-01 2.71673024e-01 -2.44043946e-01 -8.78648520e-01
-1.25667393e+00 9.76415575e-01 -5.79261519e-02 -1.68279201e-01
-3.67960542e-01 8.19744229e-01 4.82474208e-01 -4.77912784e-01
1.05650753e-01 -8.56996059e-01 -2.52364278e-01 6.50257617e-02
4.39638406e-01 2.48460129e-01 5.32366261e-02 -6.61325395e-01
-1.06108911e-01 6.25012994e-01 -2.77978659e-01 2.90287868e-03
1.38705552e+00 -3.83816250e-02 1.56284556e-01 3.35597724e-01
7.81529725e-01 -2.42350608e-01 -1.32102883e+00 -2.99948514e-01
-6.23682849e-02 -8.11805308e-01 -2.63000041e-01 -5.51577389e-01
-1.12678421e+00 5.26713550e-01 1.23228598e+00 7.60737598e-01
1.05199361e+00 3.36013287e-01 6.62145257e-01 9.53390151e-02
7.06121504e-01 -1.46335578e+00 5.65642774e-01 -1.37425944e-01
1.19780934e+00 -1.65518212e+00 3.06123286e-01 -8.58663797e-01
-9.10697401e-01 9.90170062e-01 6.68895662e-01 -4.01343226e-01
5.36007643e-01 -3.66342887e-02 -1.14359455e-02 -1.05371244e-01
-3.61331612e-01 -3.24023873e-01 4.99186158e-01 3.83253723e-01
3.55182022e-01 2.17955619e-01 -6.56927824e-01 2.38578856e-01
-2.51778215e-01 1.12541460e-01 4.57423449e-01 9.49744999e-01
-6.77990377e-01 -1.25012660e+00 -7.08864152e-01 2.97639579e-01
-3.00867409e-01 1.28656700e-01 -1.92334861e-01 1.31867242e+00
4.75712061e-01 8.39397311e-01 5.63905016e-02 2.73368079e-02
5.63396215e-01 -4.29518878e-01 6.17189825e-01 -4.28088814e-01
-5.96431732e-01 5.27780354e-01 3.25833447e-02 -5.94171464e-01
-7.07114398e-01 -4.50021982e-01 -9.28511500e-01 1.15752993e-02
-9.56888258e-01 5.12079783e-02 9.67882037e-01 5.16162634e-01
2.91948002e-02 1.54293746e-01 7.42046118e-01 -1.17950523e+00
-6.21804297e-01 -1.32304144e+00 -1.00743902e+00 8.84477437e-01
-3.41164887e-01 -6.92551553e-01 -1.31925121e-01 1.78419933e-01] | [9.192612648010254, 0.8143323063850403] |
9506af0e-33df-4edf-810a-324341107934 | hopeedi-a-multilingual-hope-speech-detection | null | null | https://aclanthology.org/2020.peoples-1.5 | https://aclanthology.org/2020.peoples-1.5.pdf | HopeEDI: A Multilingual Hope Speech Detection Dataset for Equality, Diversity, and Inclusion | Over the past few years, systems have been developed to control online content and eliminate abusive, offensive or hate speech content. However, people in power sometimes misuse this form of censorship to obstruct the democratic right of freedom of speech. Therefore, it is imperative that research should take a positive reinforcement approach towards online content that is encouraging, positive and supportive contents. Until now, most studies have focused on solving this problem of negativity in the English language, though the problem is much more than just harmful content. Furthermore, it is multilingual as well. Thus, we have constructed a Hope Speech dataset for Equality, Diversity and Inclusion (HopeEDI) containing user-generated comments from the social media platform YouTube with 28,451, 20,198 and 10,705 comments in English, Tamil and Malayalam, respectively, manually labelled as containing hope speech or not. To our knowledge, this is the first research of its kind to annotate hope speech for equality, diversity and inclusion in a multilingual setting. We determined that the inter-annotator agreement of our dataset using Krippendorff’s alpha. Further, we created several baselines to benchmark the resulting dataset and the results have been expressed using precision, recall and F1-score. The dataset is publicly available for the research community. We hope that this resource will spur further research on encouraging inclusive and responsive speech that reinforces positiveness. | ['Bharathi Raja Chakravarthi'] | 2020-12-01 | null | null | null | null | ['hope-speech-detection-for-tamil', 'hope-speech-detection-for-malayalam', 'hope-speech-detection', 'hope-speech-detection-for-english'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [-1.91200063e-01 5.86053491e-01 -4.43204969e-01 -3.35544407e-01
-5.89077771e-01 -8.95794928e-01 1.09631288e+00 2.00696081e-01
-6.20117188e-01 8.32789302e-01 1.08869565e+00 -4.55072016e-01
7.65103623e-02 -3.20325881e-01 -2.39862561e-01 -2.24752977e-01
5.47075987e-01 2.92985104e-02 -3.77036817e-02 -6.06810868e-01
4.77273941e-01 2.08129570e-01 -1.12326980e+00 2.30018467e-01
1.16021788e+00 2.48802811e-01 -3.56807351e-01 6.30669117e-01
-2.40309052e-02 1.37825274e+00 -8.57776999e-01 -1.05791068e+00
-6.04000278e-02 -4.64164674e-01 -9.79686320e-01 -1.12587750e-01
5.70686698e-01 -2.23621324e-01 -3.28234732e-01 1.14395833e+00
7.63438523e-01 7.28353783e-02 3.61848891e-01 -1.02564037e+00
-9.93198097e-01 1.04670811e+00 -3.04099351e-01 2.76591003e-01
5.46980798e-01 1.31967396e-01 1.18990326e+00 -7.15493798e-01
1.07383430e+00 1.24341440e+00 4.98080313e-01 6.93473577e-01
-1.05111384e+00 -6.77243888e-01 -2.22920552e-01 -8.77244025e-02
-1.18739653e+00 -8.64244401e-01 5.15161514e-01 -5.79995573e-01
7.18957782e-01 7.78266609e-01 7.59748876e-01 1.55202174e+00
-1.12233117e-01 7.04184890e-01 1.47126317e+00 -5.01964927e-01
-7.03056753e-02 7.33047009e-01 -1.33523256e-01 3.88794839e-01
-2.48359665e-02 -2.85648257e-01 -5.92239499e-01 -2.05548495e-01
-9.20405090e-02 -5.04103363e-01 -4.29809630e-01 2.68331975e-01
-1.04219627e+00 1.19717658e+00 9.92027894e-02 6.43488348e-01
-6.56032264e-02 -5.71841836e-01 6.82127893e-01 4.07463133e-01
8.21616054e-01 4.76497620e-01 -1.61952689e-01 -7.42969871e-01
-8.55108023e-01 1.02289125e-01 1.31508982e+00 6.55702829e-01
4.07279842e-02 -1.24237515e-01 -8.70779753e-02 1.35546267e+00
3.59507561e-01 7.77831912e-01 3.35909337e-01 -9.48457181e-01
3.59952778e-01 3.97987634e-01 -3.53111215e-02 -1.28535998e+00
-2.48313546e-01 -3.44673514e-01 -4.29708034e-01 -8.32817703e-02
3.89060050e-01 -4.02802974e-01 -4.33158755e-01 1.71544313e+00
2.90790349e-01 -5.09588957e-01 1.83982983e-01 9.34407115e-01
1.07467473e+00 7.33664870e-01 1.26741037e-01 -4.02980208e-01
1.33284318e+00 -6.78783715e-01 -1.19789898e+00 -4.78750356e-02
8.11254203e-01 -1.37207413e+00 1.34085417e+00 3.57190102e-01
-9.17231560e-01 2.53917605e-01 -8.24483871e-01 -2.66053826e-01
-4.91717786e-01 -2.83212334e-01 3.51845801e-01 1.18695521e+00
-7.42684722e-01 2.99266994e-01 -1.55280605e-01 -8.01278234e-01
3.42565030e-01 -1.72816396e-01 -5.40407300e-01 2.80178696e-01
-1.52217448e+00 1.15620303e+00 3.36001366e-02 -1.62617445e-01
-4.51471865e-01 -4.09097642e-01 -5.80329537e-01 -5.08615732e-01
3.24664295e-01 1.91075668e-01 1.20377576e+00 -1.30012703e+00
-1.37418020e+00 1.18692076e+00 2.64650464e-01 -2.47132540e-01
6.25130713e-01 -3.77142191e-01 -7.77004600e-01 -1.17566414e-01
3.36696386e-01 3.84207368e-01 5.67607641e-01 -1.00521994e+00
-5.90271294e-01 -2.26283044e-01 1.97809145e-01 4.71649282e-02
-7.88815081e-01 8.57895494e-01 2.37479225e-01 -7.84288764e-01
-4.29624170e-01 -1.06147194e+00 5.05714238e-01 -3.09229374e-01
-6.00557268e-01 -2.26866588e-01 9.19711590e-01 -1.21185613e+00
1.57495463e+00 -2.28739834e+00 -7.37089887e-02 -1.96015146e-02
2.58735865e-01 5.53625584e-01 2.50420809e-01 9.17225301e-01
2.31440723e-01 8.36282372e-01 4.88439463e-02 -6.60390174e-03
7.29864612e-02 2.93706477e-01 -2.35638186e-01 7.65388370e-01
-3.17753136e-01 4.98818219e-01 -1.02039242e+00 -5.99518657e-01
-4.52236012e-02 6.93097055e-01 -5.48366010e-01 -1.37999728e-01
2.20216453e-01 3.70144844e-01 -7.60948658e-02 5.14459252e-01
4.16243613e-01 1.17570914e-01 2.60872751e-01 2.03954443e-01
-7.11478174e-01 5.07441103e-01 -5.02490997e-01 1.12107265e+00
-4.14861262e-01 1.22939146e+00 3.64023477e-01 -2.47600794e-01
7.93408632e-01 4.83044565e-01 2.95652926e-01 -7.33473897e-01
3.99560779e-01 4.88743633e-01 2.49447212e-01 -6.98257446e-01
8.07463467e-01 -1.60482794e-01 -1.90414384e-01 3.34061146e-01
-3.13304573e-01 -2.22930565e-01 1.78668782e-01 6.34274125e-01
8.46862972e-01 -2.08635285e-01 3.36356133e-01 -2.27959439e-01
5.99912763e-01 -1.11529410e-01 3.35687250e-01 5.53247333e-01
-7.93670952e-01 3.08925569e-01 7.41753101e-01 2.35577077e-02
-1.16102278e+00 -7.27321208e-01 -4.16953772e-01 1.40950525e+00
-3.74718815e-01 -6.31064773e-01 -8.82187188e-01 -7.33183980e-01
-3.17596823e-01 1.15854228e+00 -2.94475824e-01 -7.07095414e-02
-3.64072829e-01 -2.61757672e-01 9.80948746e-01 -3.21290582e-01
5.87996185e-01 -1.03122842e+00 -2.01635480e-01 1.69349313e-02
-6.07740879e-01 -1.18460953e+00 -4.88973498e-01 -2.77435094e-01
-1.67041332e-01 -1.04603505e+00 -5.32687843e-01 -4.42517638e-01
1.18474841e-01 1.30065292e-01 7.97200263e-01 1.47920698e-01
2.43535936e-01 2.97488272e-01 -7.51572371e-01 -5.00955224e-01
-8.37076366e-01 2.59181440e-01 -1.45805672e-01 -1.30027622e-01
4.45416093e-01 -3.12704295e-01 -2.56401539e-01 1.25698432e-01
-8.10600877e-01 -1.13068148e-01 9.87269357e-02 6.66175663e-01
-3.64639372e-01 -3.39955539e-01 6.51221275e-01 -1.03227687e+00
1.10086322e+00 -7.48451412e-01 1.85155515e-02 -7.74443001e-02
-4.58964914e-01 -6.07522249e-01 6.47682726e-01 -3.17758262e-01
-1.15968120e+00 -4.31393802e-01 -4.33171719e-01 2.17563316e-01
-3.45138423e-02 4.84358817e-01 2.13670239e-01 7.52997324e-02
9.26727712e-01 -1.10465616e-01 7.67866820e-02 -3.11189920e-01
4.82683241e-01 1.52174866e+00 3.54259253e-01 -3.43267679e-01
5.76798856e-01 3.82189989e-01 -7.62816548e-01 -1.40815604e+00
-1.02331591e+00 -5.03608406e-01 -2.27595732e-01 -6.86068058e-01
8.58346045e-01 -8.73463929e-01 -6.32350087e-01 3.14011961e-01
-1.07756901e+00 -1.49277583e-01 -1.83634404e-02 4.27597344e-01
-5.17855063e-02 5.46890855e-01 -6.41118765e-01 -1.17994535e+00
-4.88951296e-01 -5.58293462e-01 5.65188348e-01 1.25346616e-01
-8.31051171e-01 -1.00705814e+00 1.44225433e-01 9.13272560e-01
4.70500916e-01 4.07018691e-01 5.07349491e-01 -1.08610380e+00
3.12774509e-01 -2.89489776e-01 5.07337265e-02 5.64012051e-01
1.09791428e-01 5.53329170e-01 -8.57635796e-01 -4.74203974e-02
-3.21980342e-02 -6.17215395e-01 2.42417142e-01 -2.17482269e-01
3.26052368e-01 -1.13018143e+00 2.82097012e-01 -2.65603215e-01
8.38662386e-01 -3.45187150e-02 8.89748931e-01 6.51125550e-01
5.12897551e-01 8.65771115e-01 5.95910668e-01 7.01752543e-01
5.40138841e-01 5.15935361e-01 2.36405402e-01 1.48939848e-01
-9.98141244e-02 -4.69309241e-01 6.70512319e-01 1.15254343e+00
7.11540952e-02 -3.74367774e-01 -9.64678288e-01 6.67904615e-01
-1.56210947e+00 -1.48760235e+00 -5.98794818e-01 1.94357824e+00
1.12120068e+00 7.87534788e-02 3.74343246e-01 2.79502511e-01
7.57018864e-01 5.17582715e-01 2.74276257e-01 -1.00494492e+00
-2.37868100e-01 -1.81617379e-01 3.57688844e-01 8.12511384e-01
-1.08435059e+00 1.04841995e+00 5.47910643e+00 8.34945977e-01
-1.24111259e+00 4.18392628e-01 4.95377153e-01 -1.82498589e-01
-5.33558249e-01 7.36358315e-02 -5.79193115e-01 7.31776655e-01
1.06375837e+00 -1.70984700e-01 4.64731216e-01 8.07543099e-01
7.32869804e-01 3.37602645e-02 -3.73439431e-01 7.12134540e-01
2.99595058e-01 -8.88718605e-01 -5.63943803e-01 1.29193753e-01
7.42668211e-01 2.08940253e-01 1.83303937e-01 3.38304818e-01
3.64246070e-01 -9.15284634e-01 1.09863162e+00 2.45501637e-01
4.86718833e-01 -6.51039779e-01 6.81082249e-01 5.76332688e-01
-1.71339810e-01 2.91011669e-03 4.29635681e-02 -2.32521519e-01
3.16050261e-01 5.22712886e-01 -7.51974225e-01 -4.21866477e-02
6.06072366e-01 7.08610058e-01 -6.13035977e-01 6.52538240e-01
-6.52940035e-01 9.12214637e-01 -1.85605139e-01 -5.05093813e-01
4.67248529e-01 -5.38513884e-02 8.67119908e-01 1.44365609e+00
-1.13542959e-01 4.56106886e-02 -1.43981591e-01 3.50419044e-01
-2.75446147e-01 6.39180183e-01 -8.52723181e-01 -6.48195565e-01
7.62812078e-01 1.39090192e+00 -4.87082601e-01 -1.71080112e-01
-6.54009998e-01 6.12341404e-01 2.95888364e-01 5.83875291e-02
-9.05363321e-01 -2.44693130e-01 3.13898265e-01 3.67070824e-01
-3.07472110e-01 -1.30980074e-01 -3.81560773e-01 -1.03442705e+00
-1.10783659e-01 -1.38108635e+00 3.63051325e-01 -5.82586050e-01
-1.37102485e+00 3.46917033e-01 -2.66788185e-01 -7.11226583e-01
-1.95534788e-02 -1.68755725e-01 -1.59624577e-01 5.54938495e-01
-1.07149649e+00 -1.00140858e+00 -1.95869580e-02 2.99024910e-01
4.14915890e-01 -1.09293591e-02 6.15685105e-01 5.02655208e-01
-6.01324499e-01 4.94870961e-01 1.09777734e-01 2.49076277e-01
1.04220462e+00 -9.51828957e-01 -7.19776601e-02 6.80371940e-01
2.56217755e-02 6.89363956e-01 1.10484266e+00 -7.99686551e-01
-9.45717514e-01 -6.12095714e-01 1.62405837e+00 -6.50851309e-01
1.20944524e+00 -4.27796215e-01 -6.98827028e-01 6.16420686e-01
7.87532687e-01 -5.30613184e-01 9.86527205e-01 1.55352950e-01
-4.58673328e-01 2.53506154e-01 -1.33658123e+00 8.12418818e-01
9.64171767e-01 -8.20290208e-01 -5.85097075e-01 7.69413531e-01
4.79315162e-01 -2.82948911e-01 -1.11152327e+00 -2.02020526e-01
7.35365331e-01 -1.05295885e+00 3.88808280e-01 -5.82515359e-01
7.35087097e-01 6.63070679e-02 -2.50166059e-01 -1.13553250e+00
-1.55161589e-01 -8.95907223e-01 2.26049677e-01 1.87857246e+00
6.72674656e-01 -6.44227207e-01 4.17115092e-01 8.26432228e-01
-3.01138937e-01 -3.99409950e-01 -9.93032217e-01 -5.20411074e-01
3.64355356e-01 -2.29589894e-01 7.64121860e-02 1.48846078e+00
3.37896258e-01 6.33561134e-01 -9.24863875e-01 -2.57321894e-01
1.22996040e-01 -7.08318591e-01 8.93355489e-01 -9.74881411e-01
1.81658506e-01 -5.07540047e-01 -1.66118771e-01 -3.71914178e-01
3.83946002e-01 -8.64086926e-01 -3.53632629e-01 -1.33780468e+00
2.68495351e-01 -1.15157329e-01 6.18424654e-01 4.27185029e-01
1.78659201e-01 2.86853731e-01 3.46313208e-01 3.34133267e-01
-5.68487048e-01 4.89547253e-01 1.27737701e+00 -1.32206222e-02
-1.81668669e-01 -4.03649896e-01 -1.15464854e+00 6.21003747e-01
9.74223435e-01 -6.19539440e-01 -3.13425303e-01 -9.93533731e-02
5.02671838e-01 -3.86561185e-01 2.49331579e-01 -4.61523443e-01
-3.18861678e-02 -3.34263563e-01 -1.03593551e-01 -6.62587211e-02
1.85295954e-01 -6.74312055e-01 6.75711930e-02 3.88020456e-01
-5.55982411e-01 -2.08069295e-01 -2.09030703e-01 1.47450849e-01
-2.09849834e-01 -4.14325804e-01 9.25015628e-01 -1.25314891e-02
-2.41446897e-01 -1.88804626e-01 -7.80370057e-01 6.19780898e-01
1.00888586e+00 -1.25594988e-01 -8.26335490e-01 -9.61526871e-01
-5.44200897e-01 1.15437573e-02 6.79671407e-01 7.02430904e-01
2.08015174e-01 -1.25021899e+00 -1.13820660e+00 -3.38420242e-01
1.39145255e-01 -9.36407983e-01 3.89545709e-02 1.11068034e+00
-5.91238558e-01 3.05798322e-01 1.06524816e-02 -1.15358546e-01
-1.45657706e+00 2.79633194e-01 4.98515181e-03 1.71184957e-01
-4.74005461e-01 5.13501823e-01 -5.43097019e-01 -6.17253006e-01
7.72549212e-02 5.62904835e-01 -5.51841259e-01 4.54187393e-01
5.41166008e-01 6.36074245e-01 -1.89595312e-01 -1.49736619e+00
-3.90287757e-01 -9.56062973e-02 -1.28051266e-01 -4.60579962e-01
1.14094996e+00 -3.72387320e-01 -4.90579009e-01 4.45103645e-01
1.28400075e+00 8.97045910e-01 -4.48647529e-01 2.64714032e-01
9.55234095e-02 -9.55259681e-01 -1.61757320e-01 -9.80658889e-01
-6.74780011e-01 3.79004240e-01 2.10341156e-01 9.33963597e-01
6.07862294e-01 -2.48667877e-02 5.90866268e-01 4.50632768e-03
-1.98317364e-01 -1.64784908e+00 3.63639090e-03 7.15212822e-01
1.32982469e+00 -1.15889525e+00 -3.51586305e-02 -4.59073007e-01
-9.27894652e-01 9.29148674e-01 4.05047864e-01 3.28347683e-01
4.91113365e-01 -1.61139981e-03 4.14938062e-01 -1.93278536e-01
-6.07549131e-01 3.04028131e-02 -5.78402057e-02 5.86285770e-01
1.28997946e+00 2.58255690e-01 -1.29602861e+00 3.02587450e-01
-7.78406441e-01 -3.13746601e-01 9.22035277e-01 8.10709536e-01
-5.01756012e-01 -1.05501735e+00 -3.75857592e-01 3.04139853e-01
-1.24784899e+00 -1.76293209e-01 -1.17220390e+00 1.05185270e+00
-7.60012791e-02 1.54622889e+00 -3.57855231e-01 -4.25857902e-01
3.10980976e-01 -8.11334699e-02 -1.46941915e-01 -4.16613430e-01
-9.90498424e-01 2.59629171e-02 1.16093493e+00 -6.92930445e-02
-5.22080421e-01 -7.88442016e-01 -9.96386707e-01 -1.10218120e+00
-3.36316943e-01 1.90616429e-01 8.31485331e-01 7.67097235e-01
2.46705294e-01 -9.24151838e-02 6.83768749e-01 -1.40676528e-01
-5.07885277e-01 -1.12856972e+00 -3.67226481e-01 5.64115524e-01
1.08881526e-01 -2.59956062e-01 -6.55950308e-01 -3.13931257e-01] | [8.891242027282715, 10.581218719482422] |
6b9c3f6a-edb9-4647-9623-8a577204146b | amritacen-at-semeval-2016-task-11-complex | null | null | https://aclanthology.org/S16-1159 | https://aclanthology.org/S16-1159.pdf | AmritaCEN at SemEval-2016 Task 11: Complex Word Identification using Word Embedding | null | ['Soman K. P', '', 'Sanjay S.P', 'An Kumar M'] | 2016-06-01 | null | null | null | semeval-2016-6 | ['complex-word-identification'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.406706809997559, 3.7817747592926025] |
6841aee4-6716-4985-a6f7-1c1ba2689c01 | who-are-you-referring-to-weakly-supervised | 2211.14563 | null | https://arxiv.org/abs/2211.14563v2 | https://arxiv.org/pdf/2211.14563v2.pdf | Who are you referring to? Coreference resolution in image narrations | Coreference resolution aims to identify words and phrases which refer to same entity in a text, a core task in natural language processing. In this paper, we extend this task to resolving coreferences in long-form narrations of visual scenes. First we introduce a new dataset with annotated coreference chains and their bounding boxes, as most existing image-text datasets only contain short sentences without coreferring expressions or labeled chains. We propose a new technique that learns to identify coreference chains using weak supervision, only from image-text pairs and a regularization using prior linguistic knowledge. Our model yields large performance gains over several strong baselines in resolving coreferences. We also show that coreference resolution helps improving grounding narratives in images. | ['Hakan Bilen', 'Frank Keller', 'Basura Fernando', 'Arushi Goel'] | 2022-11-26 | null | null | null | null | ['coreference-resolution'] | ['natural-language-processing'] | [ 5.73264420e-01 4.87677395e-01 -5.34250259e-01 -5.61055183e-01
-1.21414971e+00 -8.79237771e-01 8.08512688e-01 5.90937361e-02
-5.43242157e-01 6.98106587e-01 9.75786269e-01 7.50530064e-02
4.42100614e-02 -2.41186857e-01 -7.09926784e-01 -5.32903016e-01
2.21942663e-01 8.26370120e-01 2.47879028e-01 -2.98298448e-01
-4.05716710e-03 5.31574667e-01 -1.17399907e+00 1.03771412e+00
3.77919197e-01 4.20981854e-01 3.48079085e-01 5.55218577e-01
-1.19051628e-01 1.35259640e+00 -3.16407949e-01 -9.29368436e-01
-5.70077635e-02 -4.51880306e-01 -1.35439456e+00 -7.57169947e-02
1.15950453e+00 -1.12284862e-01 -5.19611716e-01 1.23388696e+00
3.54437172e-01 1.43359035e-01 4.55347657e-01 -1.26170218e+00
-1.94846556e-01 1.07852805e+00 -8.41578603e-01 3.65382642e-01
7.36126959e-01 -4.62439775e-01 1.61123884e+00 -6.82285190e-01
1.26312733e+00 1.82344520e+00 5.09840906e-01 7.71907270e-01
-1.53804219e+00 -1.01943886e+00 1.34201810e-01 4.73458320e-01
-1.20812142e+00 -5.57808697e-01 7.26265728e-01 -3.54979903e-01
1.12390578e+00 3.38323325e-01 1.00193182e-02 1.49967599e+00
-4.53537256e-02 9.09923732e-01 8.35263252e-01 -4.71762508e-01
-4.56994921e-01 -1.38221942e-02 3.26710761e-01 5.58807075e-01
-1.63829207e-01 1.46749556e-01 -6.50857270e-01 -1.18954107e-01
4.76518691e-01 -3.47112715e-01 -5.69455385e-01 -5.15674174e-01
-1.35173893e+00 9.30729747e-01 5.66390336e-01 4.31961417e-01
-4.05266089e-03 1.09580524e-01 5.39408803e-01 9.86249149e-02
-1.38928786e-01 5.99471629e-01 4.06774655e-02 3.24618697e-01
-6.79740429e-01 3.01018268e-01 6.96681023e-01 1.32858300e+00
5.28435946e-01 -8.57882738e-01 -4.04216886e-01 6.73646867e-01
-1.30833119e-01 6.07833326e-01 -1.55813411e-01 -1.75226581e+00
7.19061613e-01 3.72627437e-01 1.82996124e-01 -1.39269257e+00
-5.70876539e-01 -1.14248447e-01 -8.38543117e-01 -2.54611522e-01
7.63844624e-02 3.96665819e-02 -3.27498674e-01 2.08076048e+00
-6.33619446e-03 2.28560448e-01 3.93610299e-01 1.06150460e+00
1.20738792e+00 4.14020419e-01 2.75785238e-01 -6.82931066e-01
1.65951180e+00 -1.17011261e+00 -1.06187081e+00 -2.97093153e-01
5.39385617e-01 -9.08455908e-01 8.18357468e-01 2.88514465e-01
-1.13647401e+00 -3.60034645e-01 -6.65216923e-01 -4.21886504e-01
2.27168694e-01 -2.66324610e-01 4.64023501e-01 -2.56482065e-01
-9.72686887e-01 2.63563097e-01 -3.34260494e-01 -5.09757340e-01
2.74339497e-01 2.68304914e-01 -7.06972301e-01 -1.02627479e-01
-1.30086672e+00 1.04695201e+00 4.86543626e-01 -3.21525276e-01
-7.67073572e-01 -5.60275495e-01 -1.17499530e+00 2.95203663e-02
7.56401539e-01 -8.10442746e-01 1.60838199e+00 -1.14798987e+00
-7.56484389e-01 1.63843381e+00 -5.23154438e-01 -6.43065155e-01
2.69884020e-01 -5.55499554e-01 -5.32808244e-01 7.19503105e-01
5.36708713e-01 1.12117243e+00 5.17967045e-01 -1.67522311e+00
-8.31000924e-01 -2.28389561e-01 2.83920705e-01 3.49715203e-01
1.65103182e-01 6.28288925e-01 -6.98312163e-01 -6.45648062e-01
-2.56977025e-02 -9.75874901e-01 -6.87657744e-02 -2.17953488e-01
-3.48063469e-01 -3.79071087e-01 8.86189342e-01 -1.61858201e-01
1.13082027e+00 -1.98479390e+00 4.48298037e-01 -1.62743449e-01
1.23175524e-01 -1.08168669e-01 -4.49416608e-01 2.03857541e-01
-5.14607310e-01 -1.57859251e-01 -9.81233343e-02 -5.32020390e-01
-1.02689564e-01 6.97287321e-01 -8.23164225e-01 3.28983486e-01
3.81311886e-02 8.82238507e-01 -1.31913888e+00 -1.30637085e+00
-3.89583074e-02 2.23355502e-01 -4.54848468e-01 2.36731127e-01
-3.55423570e-01 3.61821175e-01 -1.99910626e-01 1.95899606e-01
5.69901764e-01 -3.67487431e-01 4.84243333e-01 -8.35375011e-01
-1.24095917e-01 9.66791138e-02 -8.34428191e-01 2.18041849e+00
-3.09007704e-01 9.00936604e-01 4.30826783e-01 -1.00101411e+00
6.15723014e-01 5.37649035e-01 3.40475857e-01 -7.10240066e-01
2.46829495e-01 -2.24531010e-01 -2.34288409e-01 -7.81102180e-01
4.19651836e-01 -3.65033180e-01 -5.40368915e-01 4.79480147e-01
1.81683108e-01 -3.46123874e-01 3.83450836e-01 7.81783879e-01
8.33568454e-01 -1.08632654e-01 3.35253298e-01 -1.27912611e-01
7.81250179e-01 3.94670755e-01 7.62361288e-01 8.62807870e-01
-2.93888420e-01 6.76320612e-01 7.41094291e-01 -2.02557564e-01
-8.49582672e-01 -1.11343920e+00 -1.40058473e-01 1.20961607e+00
4.99241114e-01 -6.95922315e-01 -6.32958174e-01 -9.19209778e-01
-4.12310392e-01 8.67421091e-01 -5.83135068e-01 2.24615246e-01
-1.04714668e+00 -8.33646655e-02 6.39597058e-01 6.90045834e-01
3.79951686e-01 -1.18824244e+00 -6.76898539e-01 -2.88123161e-01
-9.33534563e-01 -1.72648442e+00 -8.36691618e-01 1.51045516e-01
-5.05565107e-01 -1.42022586e+00 -2.51386672e-01 -1.25658870e+00
5.20844400e-01 3.11097026e-01 1.70506120e+00 -3.88898514e-02
-1.06427372e-01 4.53338414e-01 -2.53363490e-01 -1.07773654e-01
-4.14677888e-01 -2.55245924e-01 -1.47961155e-01 -4.03584957e-01
6.35917485e-01 -2.53025234e-01 -1.59506202e-01 2.10669219e-01
-5.27383626e-01 2.84722477e-01 4.31580096e-01 9.60039437e-01
6.00561023e-01 -4.31589901e-01 -2.36595109e-01 -1.04771185e+00
4.00105983e-01 -1.42636478e-01 -3.46470863e-01 4.92975652e-01
-6.37229756e-02 1.51427552e-01 1.08246617e-01 -1.69452906e-01
-1.55923736e+00 1.78249776e-01 1.08655907e-01 -6.71818972e-01
-3.88159633e-01 1.48840591e-01 -2.81394035e-01 3.68300110e-01
5.87803960e-01 -4.56610918e-01 -3.79679978e-01 -1.02095298e-01
6.95204020e-01 4.37625915e-01 1.57649004e+00 -8.60026062e-01
4.75947887e-01 8.89080226e-01 1.11429639e-01 -6.20584488e-01
-1.68584239e+00 -1.00035059e+00 -1.12684667e+00 1.19663281e-02
1.13627183e+00 -1.10147643e+00 -9.64218676e-01 -3.59847873e-01
-1.84267938e+00 -2.72915643e-02 -1.12573586e-01 4.61243927e-01
-9.12892222e-01 6.97303355e-01 -6.71439469e-01 -3.96647662e-01
-4.29778993e-01 -8.62246454e-01 1.14868414e+00 1.92554310e-01
-8.65716040e-01 -7.17814982e-01 5.43745816e-01 7.52387166e-01
-5.21905780e-01 4.01307493e-01 6.90688372e-01 -4.03549105e-01
-1.92833200e-01 4.36988354e-01 -6.05129659e-01 -2.18453437e-01
-9.49665755e-02 -2.09845275e-01 -8.62500668e-01 -2.64022350e-01
-6.16337247e-02 -6.92777395e-01 1.06768525e+00 1.88006371e-01
6.87980235e-01 -2.93106019e-01 -7.86774099e-01 6.70462668e-01
1.15824521e+00 6.92483559e-02 4.95733291e-01 3.86449635e-01
5.80245674e-01 1.05675673e+00 9.37264681e-01 1.52655570e-02
1.92485780e-01 8.04843843e-01 2.82094598e-01 -1.95256412e-01
-2.35923976e-01 -3.60699564e-01 2.15280071e-01 3.79941881e-01
1.43773019e-01 -2.14325786e-01 -1.05818427e+00 8.20710659e-01
-2.26767206e+00 -1.57252049e+00 -3.65585238e-01 1.32744539e+00
1.15364516e+00 -1.04968706e-02 -2.64715791e-01 -4.10277516e-01
1.04040432e+00 1.92617968e-01 -2.86620378e-01 -2.20285162e-01
-5.98411739e-01 2.52775192e-01 2.15219289e-01 8.86720002e-01
-1.46338248e+00 1.21549547e+00 6.64974117e+00 2.03957558e-01
-5.78265786e-01 1.42497465e-01 7.07447380e-02 -2.69884914e-01
-3.85872722e-01 1.21984832e-01 -7.36075938e-01 -2.85456240e-01
4.46055859e-01 2.82724537e-02 -7.28001073e-02 5.29866517e-01
1.38463929e-01 -1.05093926e-01 -1.42840636e+00 1.40662336e+00
6.93535924e-01 -1.59192741e+00 1.25840470e-01 -5.84821939e-01
8.00321043e-01 5.60939871e-02 -2.99786270e-01 -1.18438125e-01
8.40817869e-01 -1.16841722e+00 5.69897711e-01 3.03439915e-01
7.82175422e-01 -7.24414527e-01 8.23519886e-01 1.50474235e-01
-1.14898157e+00 2.66327541e-02 -1.57029435e-01 1.55424476e-01
6.94564462e-01 -9.76055413e-02 -4.21940774e-01 4.51849073e-01
1.02604890e+00 9.06371653e-01 -1.48679540e-01 4.34671670e-01
-8.09633136e-01 4.26703021e-02 -4.26313980e-03 5.60417235e-01
3.29904228e-01 6.44621924e-02 7.66503692e-01 1.75847220e+00
-1.41391307e-01 7.87060380e-01 2.08018199e-01 8.21229577e-01
-2.39374220e-01 -2.27337435e-01 -6.87796414e-01 3.75399947e-01
5.94317913e-01 1.40493321e+00 -4.12830859e-01 -3.66048694e-01
-6.96514189e-01 1.10182071e+00 5.57235897e-01 2.87428051e-01
-6.61345482e-01 -1.66802749e-01 6.93883061e-01 -1.30420119e-01
2.30731443e-01 7.74524510e-02 1.31673411e-01 -1.22043288e+00
-5.25461912e-01 -1.11050141e+00 1.18118703e+00 -1.11078763e+00
-1.26231277e+00 4.96321440e-01 3.52458060e-01 -8.22132766e-01
-2.43000969e-01 -4.56116617e-01 -6.15039825e-01 4.18329865e-01
-1.55031526e+00 -1.27489030e+00 -4.73402232e-01 1.05503297e+00
5.30587256e-01 1.85493067e-01 8.28572154e-01 -3.33647691e-02
-3.32738906e-01 3.96055043e-01 -5.05994201e-01 6.78393066e-01
1.35680938e+00 -1.23935044e+00 -6.72558546e-02 8.37007165e-01
5.10620415e-01 7.40027606e-01 1.12860215e+00 -5.60019553e-01
-7.95579314e-01 -7.83608019e-01 1.29231465e+00 -5.53357065e-01
8.63286853e-01 -3.80926579e-01 -7.46745944e-01 1.07056868e+00
1.01787853e+00 -1.93911672e-01 7.52389610e-01 2.01702386e-01
-7.13267088e-01 1.22736588e-01 -8.55609000e-01 6.25884891e-01
1.39747512e+00 -8.57985854e-01 -1.54982543e+00 2.29975104e-01
8.89315128e-01 -5.39842069e-01 -6.51755691e-01 6.48671985e-01
2.91511118e-01 -8.89947176e-01 1.10003865e+00 -8.78057182e-01
5.66662431e-01 -1.93801254e-01 -3.33160669e-01 -7.74345219e-01
-4.54691559e-01 -7.67732143e-01 9.37405378e-02 1.38260782e+00
2.31040850e-01 3.12427819e-01 7.79869437e-01 6.08980119e-01
1.66187491e-02 2.25574046e-01 -7.76035070e-01 -2.91944295e-01
9.15733352e-02 -2.56155610e-01 7.74542615e-02 1.32048309e+00
7.61426687e-01 9.66135859e-01 -3.31643879e-01 2.00736135e-01
1.02120936e+00 7.12285757e-01 7.26136386e-01 -1.23189437e+00
-1.02580748e-02 -2.92723894e-01 1.24929443e-01 -1.10135257e+00
1.13589680e+00 -8.59292209e-01 2.46951669e-01 -1.17622101e+00
1.01839399e+00 5.36273196e-02 2.83456109e-02 4.79039818e-01
8.55554193e-02 2.26124167e-01 3.94975096e-01 2.79156953e-01
-1.09014666e+00 1.57640576e-01 1.13388824e+00 -5.18067658e-01
1.57107294e-01 -6.85221851e-01 -6.61288619e-01 1.10559416e+00
2.88627714e-01 -6.22801065e-01 -4.06811714e-01 -5.83876431e-01
3.26339155e-01 3.60982388e-01 4.69058603e-01 -3.89964461e-01
5.91817677e-01 -3.72731209e-01 2.30039228e-02 -9.13036227e-01
1.36939481e-01 -8.06901455e-01 -3.11745238e-02 2.50493467e-01
-8.06970417e-01 1.03927486e-01 2.62146235e-01 4.33598101e-01
-5.37954211e-01 -3.37469220e-01 7.88362265e-01 -4.03360575e-01
-1.09302938e+00 -6.75225258e-02 1.47706876e-02 5.57047367e-01
1.02927387e+00 3.06802303e-01 -5.58134556e-01 -3.83062959e-01
-1.09589458e+00 6.89580083e-01 5.65495491e-01 5.62108994e-01
6.15752101e-01 -1.32041478e+00 -8.07338655e-01 -4.50400680e-01
2.80803770e-01 8.92324224e-02 1.17009312e-01 8.14493120e-01
-1.59016252e-01 7.62926280e-01 -2.11817428e-01 -7.76493490e-01
-2.06221104e+00 9.45276141e-01 4.34974313e-01 -2.21829608e-01
-5.94401121e-01 9.71151769e-01 7.41707027e-01 -1.19892612e-01
6.26489937e-01 -1.22328117e-01 -5.92887104e-01 3.70345503e-01
6.66117966e-01 -1.60549656e-01 -4.74387079e-01 -1.32603264e+00
-5.95209360e-01 8.89423609e-01 -2.21824795e-01 -2.32038751e-01
1.07311797e+00 -5.37955642e-01 -3.50586206e-01 3.17879170e-01
1.17733657e+00 -2.40767114e-02 -1.10536587e+00 -6.27355278e-01
3.47700924e-01 -3.22965026e-01 -4.30558883e-02 -4.38660383e-01
-8.29274774e-01 7.94657767e-01 1.35190323e-01 -6.02430522e-01
1.09176075e+00 9.71401393e-01 6.74312532e-01 6.18721604e-01
1.07069135e-01 -9.31455672e-01 4.05490398e-01 5.72469652e-01
1.19367325e+00 -1.43091178e+00 -8.49514976e-02 -5.32911241e-01
-9.77878928e-01 1.03609776e+00 6.98690712e-01 9.11452100e-02
2.32388139e-01 7.62041390e-01 3.79934400e-01 -5.20580351e-01
-1.05185342e+00 -4.59291011e-01 1.50933117e-01 6.44428372e-01
5.35300195e-01 -2.89018422e-01 -1.74327254e-01 7.92206109e-01
-2.19858572e-01 -2.46718168e-01 4.00140375e-01 5.31134248e-01
-2.89191335e-01 -1.03013206e+00 -5.30314028e-01 -2.74359196e-01
-5.15554070e-01 -2.31777012e-01 -7.95201540e-01 7.40478635e-01
1.98054507e-01 1.10547376e+00 6.02366865e-01 3.98259275e-02
5.25963783e-01 -3.05799276e-01 7.27907479e-01 -5.99362433e-01
-4.64883000e-01 9.91126373e-02 4.19716954e-01 -8.62778127e-01
-1.26498640e+00 -9.27939296e-01 -1.57911098e+00 -1.34531230e-01
-2.92881668e-01 4.92201507e-01 -4.62765545e-01 9.72911596e-01
-1.22641884e-01 1.29559830e-01 1.41412690e-01 -4.92362499e-01
-8.20815489e-02 -5.56105316e-01 -2.24281386e-01 1.27038693e+00
5.24465263e-01 -6.29087031e-01 -2.17758983e-01 5.63096225e-01] | [9.291836738586426, 9.507184028625488] |
071246cb-e41a-40a4-b303-c147742e401a | sievenet-a-unified-framework-for-robust-image | 2001.06265 | null | https://arxiv.org/abs/2001.06265v1 | https://arxiv.org/pdf/2001.06265v1.pdf | SieveNet: A Unified Framework for Robust Image-Based Virtual Try-On | Image-based virtual try-on for fashion has gained considerable attention recently. The task requires trying on a clothing item on a target model image. An efficient framework for this is composed of two stages: (1) warping (transforming) the try-on cloth to align with the pose and shape of the target model, and (2) a texture transfer module to seamlessly integrate the warped try-on cloth onto the target model image. Existing methods suffer from artifacts and distortions in their try-on output. In this work, we present SieveNet, a framework for robust image-based virtual try-on. Firstly, we introduce a multi-stage coarse-to-fine warping network to better model fine-grained intricacies (while transforming the try-on cloth) and train it with a novel perceptual geometric matching loss. Next, we introduce a try-on cloth conditioned segmentation mask prior to improve the texture transfer network. Finally, we also introduce a dueling triplet loss strategy for training the texture translation network which further improves the quality of the generated try-on results. We present extensive qualitative and quantitative evaluations of each component of the proposed pipeline and show significant performance improvements against the current state-of-the-art method. | ['Kumar Ayush', 'Abhijeet Kumar', 'Balaji Krishnamurthy', 'Surgan Jandial', 'Mayur Hemani', 'Ayush Chopra'] | 2020-01-17 | null | null | null | null | ['geometric-matching'] | ['computer-vision'] | [ 6.12867594e-01 -2.35060453e-02 1.03659347e-01 -4.16875422e-01
-8.65667939e-01 -6.08979881e-01 2.65162706e-01 -2.88109541e-01
5.06103039e-03 1.42626375e-01 -1.29454225e-01 4.57129516e-02
2.49175578e-01 -7.44966388e-01 -1.10217166e+00 -5.51726878e-01
5.14659643e-01 4.80664432e-01 4.87002283e-01 -3.44042063e-01
-1.04859862e-02 3.95398378e-01 -1.08246708e+00 6.14465296e-01
7.24686682e-01 1.31450534e+00 2.86472350e-01 6.12046003e-01
3.11377943e-01 9.63617414e-02 -2.14081064e-01 -8.04832101e-01
6.09663069e-01 -6.81528449e-01 -4.68101054e-01 6.41781449e-01
9.43522394e-01 -2.85614848e-01 -5.66962734e-02 1.05976367e+00
2.31162250e-01 1.75695837e-01 4.55365837e-01 -9.20227885e-01
-8.45259845e-01 3.18451881e-01 -8.22190464e-01 -4.20405805e-01
2.83907235e-01 3.26662064e-01 5.80215812e-01 -9.54438865e-01
1.02242434e+00 1.42163348e+00 8.95882010e-01 3.55747402e-01
-1.59584880e+00 -4.07135487e-01 3.25128697e-02 -1.07150286e-01
-1.13489926e+00 -1.67786375e-01 1.25781178e+00 -2.92754054e-01
3.95847023e-01 4.11848545e-01 9.83582199e-01 9.72047448e-01
6.09264851e-01 7.19926894e-01 1.44785810e+00 -5.56195199e-01
6.01401068e-02 -4.05769199e-02 -4.13534224e-01 8.38102579e-01
-3.38194698e-01 2.54679620e-01 -1.45107001e-01 3.55740100e-01
1.17775679e+00 -1.36265501e-01 6.05718493e-02 -6.07680678e-01
-1.01313281e+00 4.40457582e-01 7.87221074e-01 -4.26437706e-02
-5.66061497e-01 1.84078768e-01 3.71882260e-01 2.25310028e-01
7.94759631e-01 2.37377316e-01 -3.69597614e-01 4.32541847e-01
-1.25526047e+00 3.20555031e-01 5.03607094e-01 9.42398190e-01
5.07000625e-01 -4.20825183e-02 -3.03341478e-01 9.79165912e-01
3.49605918e-01 6.51664257e-01 -3.18953544e-01 -9.83248830e-01
3.45112205e-01 3.34085047e-01 -1.33839399e-02 -1.15371311e+00
-9.12341401e-02 -1.77397579e-01 -7.39295602e-01 2.89285183e-01
5.64322948e-01 2.33153686e-01 -1.43928146e+00 1.30728555e+00
7.28114009e-01 2.61497870e-02 -5.74774742e-01 1.28463089e+00
8.42057884e-01 3.99688482e-01 1.17405526e-01 1.55710250e-01
1.53634953e+00 -1.22171330e+00 -7.37579584e-01 -1.08851478e-01
-3.83141264e-02 -1.50128603e+00 1.27417624e+00 3.64218652e-01
-1.45988441e+00 -8.36193800e-01 -1.00998998e+00 -3.47308457e-01
-1.57526627e-01 3.07030320e-01 2.90014386e-01 5.63274026e-01
-8.96751404e-01 6.97203755e-01 -8.54005158e-01 -4.74671721e-01
4.34495628e-01 2.04873711e-01 -2.80349314e-01 -3.69772553e-01
-8.01046848e-01 7.90000856e-01 1.57131970e-01 4.75790590e-01
-8.50154400e-01 -9.09982860e-01 -1.08270288e+00 -4.28074449e-01
6.24804020e-01 -9.41057682e-01 8.56630564e-01 -1.26298857e+00
-1.92661524e+00 1.25393820e+00 9.61388350e-02 -1.48151033e-02
8.53154063e-01 -1.44429013e-01 -2.04741210e-01 1.07167913e-02
1.59848243e-01 7.64880896e-01 1.11367774e+00 -1.69617260e+00
-2.45458275e-01 -2.37672836e-01 -9.04927701e-02 6.27944395e-02
4.80349809e-01 1.35403305e-01 -1.12127161e+00 -1.34378028e+00
1.69661492e-01 -1.21076798e+00 -4.71051745e-02 5.63019335e-01
-8.61184239e-01 3.22925925e-01 8.99845362e-01 -1.31747317e+00
8.52115035e-01 -1.98720193e+00 5.18727958e-01 3.46881330e-01
1.67483184e-02 2.86781371e-01 -4.48519975e-01 9.35166627e-02
-2.06199754e-02 -2.59196669e-01 -3.76533866e-01 -7.80738235e-01
-3.06392126e-02 2.19742417e-01 -1.03618041e-01 4.49723542e-01
2.30484962e-01 1.22015882e+00 -5.46752155e-01 -3.55244637e-01
7.81818271e-01 6.43085957e-01 -5.38710177e-01 2.99778253e-01
-5.30879915e-01 7.73407340e-01 -7.01784119e-02 8.90425324e-01
1.03367555e+00 1.89991087e-01 2.34515414e-01 -9.87918139e-01
-2.20087953e-02 -2.20702931e-01 -1.04058468e+00 1.87087607e+00
-5.48747361e-01 1.89214170e-01 4.55802828e-01 -4.15583968e-01
8.34732890e-01 1.98757891e-02 4.17458713e-01 -9.30598974e-01
4.35040265e-01 2.17623100e-01 -3.60818863e-01 -3.41370404e-01
4.75390613e-01 -2.29920059e-01 -2.30477490e-02 1.93661258e-01
-2.28289943e-02 -6.05627894e-01 4.86983955e-02 -3.07755291e-01
1.10299163e-01 9.12325382e-01 -3.31639767e-01 -2.45589435e-01
3.58590662e-01 1.15304857e-01 4.28771555e-01 1.34550557e-01
6.25919551e-02 1.15069199e+00 1.93216890e-01 -5.39186954e-01
-1.58239198e+00 -1.00538397e+00 7.59691596e-02 8.97770882e-01
3.75139892e-01 -1.41543075e-01 -1.27500391e+00 -7.61303484e-01
-7.94327408e-02 6.61366820e-01 -9.13810372e-01 1.54140312e-02
-7.89584339e-01 -4.29983199e-01 1.87961608e-01 5.83284974e-01
8.13132882e-01 -9.47280645e-01 -2.08087325e-01 9.19689536e-02
-3.31043065e-01 -1.23511815e+00 -1.10782325e+00 -1.25459030e-01
-6.54115915e-01 -1.00476491e+00 -9.47557330e-01 -9.79357898e-01
8.37582469e-01 7.88157433e-02 1.17840219e+00 -3.23913880e-02
-3.10364306e-01 1.95854828e-01 -1.89135462e-01 -1.90642238e-01
-6.17664516e-01 -1.78821117e-01 -2.27046609e-01 3.63885850e-01
-1.98888183e-01 -1.63689345e-01 -8.61874104e-01 7.35781789e-01
-1.09788489e+00 5.86991549e-01 4.31604207e-01 6.59230888e-01
1.03913021e+00 3.93456519e-02 -3.23501438e-01 -5.60691237e-01
2.81093597e-01 1.64582118e-01 -4.21355218e-01 4.79396850e-01
-9.16859657e-02 -2.79645801e-01 4.42987591e-01 -6.40534699e-01
-1.06461215e+00 2.00385183e-01 -3.70661914e-01 -8.27475369e-01
-1.56338494e-02 -4.11832556e-02 -4.05947417e-01 -4.16128993e-01
2.45834827e-01 -8.23973119e-03 2.06453592e-01 -6.29407048e-01
6.29981339e-01 1.17488131e-01 5.06859243e-01 -6.52401924e-01
9.25876141e-01 5.69152832e-01 -9.94074717e-03 -6.09514654e-01
-9.93749976e-01 2.84969993e-02 -1.04148948e+00 -5.39783597e-01
1.15728259e+00 -5.74781060e-01 -5.34120440e-01 8.06077659e-01
-1.06270409e+00 -1.01435781e+00 -3.88551772e-01 1.03166074e-01
-6.90553188e-01 2.23536134e-01 -8.64822686e-01 -2.36826867e-01
-3.49358141e-01 -1.39865136e+00 1.48039079e+00 3.87126990e-02
-2.44030491e-01 -1.02797008e+00 -1.07710838e-01 7.94963002e-01
4.41644311e-01 6.23009562e-01 7.14798450e-01 3.01158607e-01
-5.73108196e-01 -2.45109685e-02 -4.25508142e-01 6.56682849e-01
1.37099430e-01 4.69132215e-02 -7.29032636e-01 -3.14602464e-01
-1.61858499e-01 -1.32417232e-01 7.11559534e-01 6.90710545e-01
9.17790890e-01 -1.38460025e-01 -5.89934252e-02 9.93593454e-01
1.44858384e+00 -1.19907476e-01 7.79637098e-01 1.81738436e-01
1.19399846e+00 7.11477935e-01 9.90935206e-01 -1.48715496e-01
3.17037582e-01 1.28394878e+00 2.43691564e-01 -8.71143341e-01
-1.00781751e+00 -4.73092914e-01 3.05409968e-01 7.21509397e-01
-3.91521990e-01 -1.76136941e-02 -3.16300958e-01 2.15358257e-01
-1.59236670e+00 -4.84918475e-01 -2.99190223e-01 2.19795370e+00
7.97015190e-01 7.66355125e-03 2.41438925e-01 -2.15059996e-01
6.03162646e-01 -4.07523438e-02 -5.62426209e-01 -5.38715184e-01
-1.19473897e-01 3.30610096e-01 3.72155696e-01 7.65196502e-01
-1.32683647e+00 1.39363766e+00 5.71424103e+00 1.05367112e+00
-1.43275905e+00 2.35420167e-01 8.78302932e-01 1.84848085e-01
-2.04116285e-01 -2.19855487e-01 -4.26737547e-01 3.15220416e-01
9.63061973e-02 5.28446436e-01 5.56942165e-01 5.41525006e-01
4.01583433e-01 2.98875161e-02 -9.41520333e-01 7.65628815e-01
3.47571969e-01 -1.02791429e+00 7.96970427e-02 -3.12341154e-02
1.00266325e+00 -5.66646099e-01 2.44550213e-01 -8.84679183e-02
6.23931829e-03 -8.36481869e-01 1.34987485e+00 6.30450308e-01
1.31063342e+00 -5.80765486e-01 5.48909187e-01 -3.41624111e-01
-1.38108969e+00 4.31472719e-01 -6.79441765e-02 4.25443053e-01
4.13882583e-01 6.19503736e-01 -3.42774451e-01 7.65571117e-01
6.24456525e-01 6.08142495e-01 -5.71498513e-01 7.59156346e-01
-3.10907096e-01 2.57881224e-01 -3.47750485e-01 5.51847517e-01
1.31428286e-01 -4.59781110e-01 4.38456446e-01 1.16854739e+00
1.65099129e-01 -6.77768439e-02 5.14763415e-01 1.21616995e+00
-2.00437084e-02 6.40883148e-02 -2.21673965e-01 3.28546941e-01
-2.10653305e-01 1.36783838e+00 -1.15059948e+00 -3.93419117e-01
-4.40201722e-02 1.71346259e+00 -1.27665043e-01 4.69402075e-01
-1.04386246e+00 8.08962137e-02 3.06009054e-01 5.66511035e-01
5.59537649e-01 -2.22962976e-01 -3.84055763e-01 -1.01710892e+00
1.86499357e-01 -9.99138594e-01 -2.21404433e-02 -9.74273086e-01
-1.44534910e+00 6.44824505e-01 -1.02784596e-01 -1.21664512e+00
2.88329124e-01 -5.15187860e-01 -3.64555299e-01 9.63647366e-01
-1.19125700e+00 -2.22341633e+00 -3.88625145e-01 5.78643322e-01
8.81147623e-01 5.12089252e-01 5.36488414e-01 3.98261994e-01
-5.64313412e-01 6.96374297e-01 -1.56224802e-01 -4.14401665e-03
8.57128382e-01 -9.91440117e-01 5.57908773e-01 7.98447967e-01
-3.68957430e-01 4.63748574e-01 7.41173804e-01 -1.12515318e+00
-1.56696606e+00 -1.45370150e+00 4.78726387e-01 -4.27482069e-01
2.81839311e-01 -5.77276111e-01 -6.74566984e-01 6.80923522e-01
1.91843718e-01 7.30507970e-02 1.45626813e-01 -3.13492537e-01
-2.67731726e-01 -1.61194906e-01 -1.49212217e+00 7.17177749e-01
7.93865442e-01 -3.48078251e-01 -2.59565771e-01 1.98282197e-01
3.17675889e-01 -8.58402073e-01 -1.07893753e+00 4.29005325e-01
9.97007489e-01 -8.80544782e-01 1.17158520e+00 -1.66059062e-01
7.04940557e-01 -5.42358279e-01 -1.71072960e-01 -1.29390454e+00
-3.98678601e-01 -7.78322935e-01 2.63410747e-01 1.18288648e+00
2.34672874e-02 -6.74505234e-02 7.22252369e-01 6.79680347e-01
-1.24306738e-01 -8.43793511e-01 -6.43813848e-01 -4.19491827e-01
6.64795749e-03 -3.67302537e-01 1.64467573e-01 9.77479696e-01
-7.52472162e-01 9.17518511e-04 -6.11619115e-01 3.17886621e-02
1.04997742e+00 2.76397824e-01 8.29221904e-01 -6.16915107e-01
-3.51676464e-01 -2.11939335e-01 -9.22745019e-02 -1.15028250e+00
-4.48315799e-01 -7.47179389e-01 2.92273223e-01 -1.52194500e+00
1.71652168e-01 -6.09653294e-01 1.72471702e-01 4.48126346e-01
-2.03054860e-01 1.02876616e+00 4.49802727e-01 6.83268253e-03
-3.49091440e-01 4.52487260e-01 1.93965578e+00 -1.14299163e-01
-3.37281466e-01 -2.18632638e-01 -3.02755445e-01 7.65285075e-01
4.24085736e-01 -2.39882827e-01 -4.97564077e-02 -3.76464725e-01
-1.85431913e-01 6.99337572e-04 8.13168526e-01 -7.19276786e-01
-2.24516287e-01 -1.13651559e-01 6.64753079e-01 -4.21059579e-01
4.81182933e-01 -9.09637451e-01 4.62229371e-01 3.42868775e-01
-2.81332731e-01 -1.41339853e-01 4.08816069e-01 2.89347529e-01
2.01318532e-01 3.33872795e-01 1.23890662e+00 -2.40313858e-02
-3.48593235e-01 2.97455847e-01 8.16056281e-02 -3.59710187e-01
1.10239649e+00 -4.17954981e-01 3.38859856e-01 2.03329436e-02
-1.07463324e+00 -3.89363356e-02 9.69147563e-01 6.06115222e-01
7.24118292e-01 -1.42804515e+00 -6.66269660e-01 4.60727990e-01
-1.87447846e-01 -1.04477972e-01 6.10480189e-01 1.08645284e+00
-8.24822307e-01 -7.49441385e-02 -3.83712471e-01 -6.04431748e-01
-1.62500191e+00 6.79264426e-01 5.06776452e-01 -2.67992914e-01
-7.40655184e-01 8.24414611e-01 4.95081246e-01 -7.23697484e-01
5.50050847e-03 -5.10747552e-01 3.80309612e-01 -3.30313563e-01
9.73915383e-02 2.63032287e-01 1.14495857e-02 -8.57905269e-01
-1.16315156e-01 1.33926570e+00 -7.53442347e-02 -1.76671654e-01
1.24262929e+00 -2.81902552e-01 -1.50714889e-01 1.74909323e-01
9.16942060e-01 -5.48133850e-02 -1.57071674e+00 -1.23971716e-01
-5.95031261e-01 -6.94514453e-01 8.34000111e-02 -1.09158766e+00
-1.36922705e+00 7.54324794e-01 8.70939255e-01 -1.58640862e-01
1.22171783e+00 -1.65856332e-01 1.37745142e+00 -5.78174591e-01
3.79397392e-01 -1.11476505e+00 1.42958283e-01 7.14753494e-02
1.27113628e+00 -9.89247739e-01 4.57072072e-02 -1.13682079e+00
-7.22381234e-01 1.01953876e+00 4.09447432e-01 -4.53251600e-01
6.37465477e-01 2.56908983e-01 3.37130338e-01 -4.27905858e-01
-4.91604246e-02 7.14940876e-02 9.17760491e-01 6.06262922e-01
2.02951342e-01 1.10738128e-01 -8.06763917e-02 2.32220650e-01
-4.16036546e-02 2.51191199e-01 -1.04488291e-01 7.36108065e-01
1.71848193e-01 -1.27343953e+00 -7.01372862e-01 3.05350143e-02
-3.49013597e-01 -3.07590719e-02 -3.30255002e-01 6.09095097e-01
4.73256856e-01 7.49852657e-01 -1.18351489e-01 -4.77119833e-01
7.69717038e-01 -3.82703006e-01 1.03647828e+00 -3.87854576e-01
-8.95877481e-01 6.26996756e-01 1.14184007e-01 -9.36710477e-01
-2.85069406e-01 -3.58141929e-01 -6.42196655e-01 -4.57536578e-01
-1.49273425e-01 -6.51357412e-01 7.32330084e-01 7.67587483e-01
2.90256828e-01 6.80378616e-01 4.63126421e-01 -1.48948121e+00
-7.65515566e-02 -6.84638023e-01 -4.79227453e-01 8.48502338e-01
-7.07177892e-02 -6.34554505e-01 9.58500504e-02 4.69288617e-01] | [11.93808364868164, -0.882952094078064] |
84beab00-0a1c-43a1-a2f8-a393c2d239e2 | ielm-an-open-information-extraction-benchmark | 2210.14128 | null | https://arxiv.org/abs/2210.14128v1 | https://arxiv.org/pdf/2210.14128v1.pdf | IELM: An Open Information Extraction Benchmark for Pre-Trained Language Models | We introduce a new open information extraction (OIE) benchmark for pre-trained language models (LM). Recent studies have demonstrated that pre-trained LMs, such as BERT and GPT, may store linguistic and relational knowledge. In particular, LMs are able to answer ``fill-in-the-blank'' questions when given a pre-defined relation category. Instead of focusing on pre-defined relations, we create an OIE benchmark aiming to fully examine the open relational information present in the pre-trained LMs. We accomplish this by turning pre-trained LMs into zero-shot OIE systems. Surprisingly, pre-trained LMs are able to obtain competitive performance on both standard OIE datasets (CaRB and Re-OIE2016) and two new large-scale factual OIE datasets (TAC KBP-OIE and Wikidata-OIE) that we establish via distant supervision. For instance, the zero-shot pre-trained LMs outperform the F1 score of the state-of-the-art supervised OIE methods on our factual OIE datasets without needing to use any training sets. Our code and datasets are available at https://github.com/cgraywang/IELM | ['Dawn Song', 'Xiao Liu', 'Chenguang Wang'] | 2022-10-25 | null | null | null | null | ['open-information-extraction'] | ['natural-language-processing'] | [-1.98557362e-01 9.06569123e-01 -5.38808048e-01 -7.95873180e-02
-8.68315339e-01 -4.40527380e-01 8.61472666e-01 4.37077075e-01
-6.36589885e-01 8.24929178e-01 1.30834833e-01 -4.05915052e-01
-4.60844755e-01 -1.08192670e+00 -9.27672744e-01 1.69328123e-01
-1.17014222e-01 9.23356712e-01 4.72803652e-01 -5.55576265e-01
-2.83738256e-01 -9.80022773e-02 -1.45584488e+00 7.09579825e-01
8.60909045e-01 1.04367709e+00 -1.14925161e-01 3.74406993e-01
-4.47226048e-01 1.20419610e+00 -3.08098584e-01 -1.04086339e+00
-4.88802604e-02 -6.71171769e-02 -1.39606237e+00 -4.87590104e-01
3.30375493e-01 3.10432702e-01 -5.60550153e-01 7.51284003e-01
1.29816040e-01 -1.15271593e-02 7.53326714e-01 -1.06849849e+00
-1.16218412e+00 1.27835059e+00 1.14258891e-02 2.29346529e-01
5.08770049e-01 -3.23496282e-01 1.68292534e+00 -1.28575444e+00
1.22905779e+00 9.63387191e-01 6.83916688e-01 3.92497092e-01
-1.16205347e+00 -4.41436052e-01 -5.34044169e-02 5.75756311e-01
-1.62937975e+00 -5.63340247e-01 4.22728062e-01 -2.98508942e-01
1.49047613e+00 3.28902483e-01 1.46371350e-01 1.17417812e+00
-8.73302110e-03 8.89551520e-01 9.18131232e-01 -7.16083407e-01
-1.29574910e-01 3.64779025e-01 4.04825598e-01 8.23219299e-01
4.03907776e-01 -1.17651328e-01 -8.07159781e-01 1.04033805e-01
2.45020807e-01 -3.18725348e-01 -3.12664568e-01 -1.83317736e-01
-1.29272985e+00 5.01304388e-01 6.17390811e-01 6.21102750e-01
-1.37092292e-01 -2.79837936e-01 5.42699814e-01 4.95787263e-01
6.63719952e-01 8.60032082e-01 -9.56723332e-01 -8.33167695e-03
-5.43320417e-01 -1.14750080e-01 1.42772424e+00 1.16313958e+00
8.06297958e-01 -9.02923584e-01 -2.24557564e-01 1.02075374e+00
5.06832637e-02 -3.37029509e-02 3.73323321e-01 -7.45017707e-01
9.63347852e-01 7.90402293e-01 -1.27514929e-01 -8.18496108e-01
-3.48877430e-01 -5.55159032e-01 -7.37495244e-01 -4.81585145e-01
3.79267991e-01 -1.10320903e-01 -6.22061133e-01 1.39289832e+00
2.26705685e-01 6.10367069e-03 5.66348314e-01 3.62209558e-01
1.54685700e+00 6.09237194e-01 -4.90792356e-02 -1.43079832e-01
1.34336221e+00 -1.22512615e+00 -9.05921757e-01 -4.17750865e-01
9.68341649e-01 -5.45753181e-01 1.26370335e+00 1.95866108e-01
-7.09663332e-01 -3.75087827e-01 -1.11697102e+00 -5.42245805e-01
-1.06815350e+00 2.32955754e-01 6.62722111e-01 -6.27622847e-03
-7.20066965e-01 5.89074790e-01 -5.30785382e-01 -5.54469883e-01
5.50719142e-01 8.40808526e-02 -6.73718691e-01 -1.08787015e-01
-1.81201947e+00 1.17274129e+00 8.13719392e-01 6.94660991e-02
-7.13169873e-01 -8.47124577e-01 -9.70982611e-01 1.70988023e-01
1.32157183e+00 -6.02681220e-01 1.07130766e+00 -9.75907668e-02
-1.23995304e+00 1.24226522e+00 -1.41505420e-01 -6.24324143e-01
2.27497995e-01 -4.81605768e-01 -6.12139583e-01 1.31779253e-01
2.10408494e-01 5.67214727e-01 5.70726991e-02 -1.24929917e+00
-4.54563230e-01 2.52899379e-02 4.40310687e-01 -1.48576275e-01
-4.39188629e-01 1.40573710e-01 -5.70887983e-01 -3.80205482e-01
-4.86246943e-02 -7.12583482e-01 1.36549979e-01 -3.67697746e-01
-8.43923390e-01 -8.49606872e-01 4.59253818e-01 -5.49138546e-01
1.62644601e+00 -1.71735334e+00 1.12382486e-01 -4.67136577e-02
5.14387369e-01 5.87974131e-01 -1.18732810e-01 6.50233030e-01
-5.51034510e-02 4.43267971e-01 -1.41857564e-01 -2.99658388e-01
2.72036284e-01 5.25723219e-01 -2.89034277e-01 -1.61319867e-01
3.03753376e-01 1.31759405e+00 -1.09088945e+00 -8.28090072e-01
-1.10840537e-01 1.14949167e-01 -3.05199385e-01 2.71545112e-01
-6.26103818e-01 4.43763770e-02 -1.73929766e-01 6.88511789e-01
-1.10534199e-01 -5.38216412e-01 1.35227129e-01 -1.35373056e-01
2.14755267e-01 1.00136888e+00 -8.33872318e-01 1.68439853e+00
-6.18004262e-01 5.51549435e-01 -4.55331504e-01 -7.94099152e-01
8.06749105e-01 4.99535918e-01 3.45982939e-01 -5.23334801e-01
1.15024872e-01 4.84795362e-01 -1.22629210e-01 -6.30496144e-01
5.79658926e-01 3.40869352e-02 -3.07333797e-01 2.21876696e-01
8.07833254e-01 -3.51937003e-02 6.43306673e-01 6.43361628e-01
1.36966634e+00 1.94336593e-01 5.54098785e-01 -1.21878311e-01
5.41485786e-01 6.84245974e-02 5.83037913e-01 6.92927778e-01
1.74083978e-01 2.27565750e-01 7.49569535e-01 -6.78675473e-02
-6.06863081e-01 -1.01167762e+00 -4.91420746e-01 1.13165963e+00
-8.62149820e-02 -1.19693971e+00 -3.41053516e-01 -1.09812629e+00
-9.15516093e-02 8.95058274e-01 -7.59675682e-01 2.13883854e-02
-3.29865664e-01 -4.20834035e-01 7.07518578e-01 1.88518018e-01
4.87013519e-01 -1.04914045e+00 -7.00117871e-02 2.05399007e-01
-5.17252624e-01 -1.47832847e+00 -4.44132835e-02 5.33305466e-01
-3.00538689e-01 -1.13977623e+00 -1.85702369e-01 -7.71443546e-01
4.23005641e-01 -2.79248834e-01 1.64605284e+00 -6.37601912e-02
1.96947642e-02 2.18128324e-01 -8.15136194e-01 -3.44740987e-01
-3.34417403e-01 7.82035530e-01 -9.22802612e-02 -2.80944049e-01
8.16503465e-01 -5.26294351e-01 1.04087824e-02 3.44371647e-01
-6.10594273e-01 2.32272685e-01 4.53969389e-01 8.37962627e-01
6.86742842e-01 3.41289453e-02 6.76935971e-01 -1.45959091e+00
5.26050091e-01 -7.88902581e-01 -2.92879134e-01 8.35942268e-01
-6.56453967e-01 2.64999717e-01 6.18759811e-01 -2.86512673e-01
-1.01026475e+00 -4.75810736e-01 -1.33872464e-01 -1.31319612e-01
2.34674662e-02 1.22195387e+00 -2.57489860e-01 2.43296862e-01
8.36179256e-01 -4.19610560e-01 -7.27401257e-01 -6.80556059e-01
7.45257497e-01 8.10415447e-01 1.02294612e+00 -8.53525937e-01
7.90504754e-01 1.37444869e-01 -7.82457963e-02 -4.15123641e-01
-1.69358325e+00 -5.68868518e-01 -9.91528392e-01 1.10663496e-01
7.52816260e-01 -8.31753373e-01 -4.95524436e-01 -3.68093587e-02
-1.21885252e+00 -3.36364776e-01 -6.09486639e-01 6.58169687e-02
-3.20617497e-01 1.41568948e-02 -8.91394675e-01 -4.34723556e-01
-2.28484780e-01 -4.75532353e-01 7.36947894e-01 5.59493387e-03
-3.50705326e-01 -1.25817466e+00 1.89660981e-01 4.20968533e-01
1.00696549e-01 1.30988106e-01 9.57531095e-01 -1.03750908e+00
-4.83936399e-01 -1.90839931e-01 -2.28647187e-01 1.99685708e-01
8.67812857e-02 -1.35203481e-01 -9.21312451e-01 2.25277245e-01
-3.26121479e-01 -8.73114109e-01 1.11118329e+00 -4.61048931e-01
8.93229306e-01 -4.82441157e-01 -5.05727232e-01 4.75077748e-01
1.27989697e+00 -3.02767545e-01 5.07307291e-01 4.26094711e-01
7.49545991e-01 9.09169912e-01 7.34195590e-01 2.61063688e-02
9.07884121e-01 6.38810158e-01 5.39344735e-02 1.97131634e-01
-3.71776760e-01 -8.17120254e-01 2.25087717e-01 1.18446982e+00
-3.15489136e-02 -2.77847290e-01 -1.28656495e+00 7.34671772e-01
-2.01410675e+00 -6.18854761e-01 -1.62143052e-01 1.86778307e+00
1.57967520e+00 4.00970817e-01 -4.72444355e-01 2.24689506e-02
3.19128692e-01 1.44393295e-01 -2.43838578e-01 -1.60148963e-01
-4.24657613e-01 4.53241497e-01 3.00427169e-01 5.74815929e-01
-1.18871593e+00 1.38438034e+00 4.97695065e+00 1.07894206e+00
-5.48470140e-01 4.61298138e-01 2.87542760e-01 4.92837057e-02
-4.46939558e-01 3.58431429e-01 -1.17508829e+00 8.01322833e-02
1.30554938e+00 -2.50752538e-01 1.70361742e-01 6.10682607e-01
-6.18555129e-01 -1.79876462e-01 -1.53934968e+00 7.13861883e-01
5.78774884e-02 -1.61710608e+00 -9.77348834e-02 -1.16553688e-02
6.45530641e-01 4.02407557e-01 -3.21244389e-01 1.01003301e+00
6.18484557e-01 -1.11014426e+00 5.63732922e-01 7.06758022e-01
9.50136065e-01 -1.57743797e-01 8.96323025e-01 5.03297567e-01
-1.17879450e+00 1.13055795e-01 -4.99918729e-01 -3.96303460e-02
7.98735693e-02 7.33443737e-01 -7.65211642e-01 1.10955441e+00
6.67451382e-01 9.68074441e-01 -1.07131183e+00 6.01302028e-01
-9.10589278e-01 6.31024122e-01 -4.95932847e-01 1.93248875e-02
1.36045158e-01 4.90351170e-02 4.43459332e-01 1.16853154e+00
-3.44641432e-02 1.80101350e-01 -2.00819522e-02 9.79945242e-01
-4.70115185e-01 7.53984451e-02 -4.64648426e-01 -3.64531785e-01
4.83418196e-01 1.10725427e+00 -2.55828321e-01 -5.99101961e-01
-7.26350784e-01 6.68511450e-01 1.07137978e+00 2.34669536e-01
-5.84668100e-01 -7.02146769e-01 9.93022546e-02 3.72696631e-02
3.85391206e-01 -3.65467854e-02 -2.51818188e-02 -1.69235945e+00
6.67439625e-02 -5.93949318e-01 8.24600697e-01 -7.15880632e-01
-1.62821710e+00 7.84355760e-01 2.11565882e-01 -9.04878974e-01
-4.40271944e-01 -6.39970422e-01 -1.17479458e-01 5.57899177e-01
-1.66166484e+00 -1.20541382e+00 4.03947532e-02 4.43237722e-01
2.80155182e-01 -1.05430089e-01 1.02165806e+00 5.89968085e-01
-7.92355120e-01 6.26505852e-01 -1.25552893e-01 6.51355207e-01
8.21861327e-01 -1.23173225e+00 3.61593902e-01 8.77355278e-01
8.82118642e-01 7.99907923e-01 5.25060952e-01 -8.06638002e-01
-9.94861484e-01 -9.53392565e-01 1.74019265e+00 -1.01639950e+00
1.18901873e+00 -5.90766251e-01 -1.09589875e+00 1.18021786e+00
2.40257591e-01 4.80963230e-01 6.98027194e-01 7.86071420e-01
-6.79806471e-01 3.68913892e-03 -6.64075613e-01 3.74327421e-01
1.49667764e+00 -9.38882530e-01 -1.32482994e+00 4.80922371e-01
1.06104958e+00 -5.19843638e-01 -1.36967480e+00 7.00055778e-01
1.80504993e-01 -6.17012084e-01 9.42443967e-01 -9.16356564e-01
8.05137098e-01 -8.27101022e-02 -7.21391365e-02 -1.18248045e+00
9.03203413e-02 -6.56295180e-01 -7.51966596e-01 1.71845686e+00
1.13007116e+00 -5.77311218e-01 4.13611472e-01 5.16654789e-01
-1.42641932e-01 -1.20538926e+00 -9.05383170e-01 -1.01347876e+00
7.38008618e-02 -5.81091106e-01 2.83023030e-01 1.07002175e+00
5.70638478e-01 8.18123996e-01 -1.44659951e-01 9.74563286e-02
3.32829386e-01 1.14467226e-01 6.43567264e-01 -1.20809102e+00
-4.22657371e-01 -1.71020776e-01 -6.26305789e-02 -9.31205750e-01
4.83822912e-01 -1.43428922e+00 -1.91093504e-01 -1.82536876e+00
1.45077437e-01 -8.86784196e-01 -4.56481159e-01 9.66428638e-01
-3.33096057e-01 5.02963737e-02 1.76039636e-01 3.62673461e-01
-1.13719904e+00 6.42847538e-01 1.04339290e+00 -8.64637643e-02
-1.70652680e-02 -2.29552045e-01 -6.33561254e-01 6.31299555e-01
5.41518927e-01 -7.30198085e-01 -2.10814163e-01 -3.28924984e-01
8.12520623e-01 -9.01807100e-02 2.06203252e-01 -7.92203128e-01
5.95293343e-01 1.63451225e-01 -2.85559773e-01 -4.17447120e-01
1.26226902e-01 -5.24225712e-01 -1.02983534e-01 -2.98939738e-03
-5.53496718e-01 -5.06493270e-01 6.42456859e-02 2.66616434e-01
-5.83549559e-01 -5.12054205e-01 9.61600766e-02 -1.30692795e-01
-8.63862872e-01 1.85306251e-01 6.17667809e-02 7.57481694e-01
5.76545417e-01 4.13241774e-01 -7.73436546e-01 -1.50004506e-01
-7.48548806e-01 4.58369195e-01 -4.47382852e-02 6.14620566e-01
3.31385463e-01 -1.18447876e+00 -5.97457469e-01 -2.92261511e-01
6.51494563e-01 3.86110812e-01 -1.80064157e-01 8.83021772e-01
-2.53090560e-01 9.05660629e-01 2.47210935e-01 -1.11030899e-01
-9.20636415e-01 7.67379880e-01 2.16122761e-01 -1.00188804e+00
-6.25659883e-01 1.11312556e+00 -2.59884417e-01 -9.13493097e-01
2.98038363e-01 -5.89478195e-01 -3.37062240e-01 1.73109010e-01
2.13141501e-01 2.39632204e-02 2.46415347e-01 -3.69691193e-01
-3.34962785e-01 2.33379811e-01 -4.94329864e-03 5.77886589e-03
1.37789297e+00 -1.66257769e-01 -4.43138659e-01 8.19818854e-01
1.21356869e+00 2.21107364e-01 -6.96806192e-01 -9.60347950e-01
8.13466251e-01 -1.74746856e-01 -1.09620467e-01 -1.10609698e+00
-5.57652652e-01 6.84573054e-01 -3.98968011e-01 2.20641509e-01
6.36766672e-01 8.09799075e-01 9.37442660e-01 8.69130433e-01
5.96134365e-01 -9.12831724e-01 -6.93769157e-02 9.65448141e-01
1.00854802e+00 -1.35095656e+00 -4.32298407e-02 -8.26139212e-01
-5.67275345e-01 7.73683727e-01 7.07831085e-01 2.13236839e-01
8.31422091e-01 2.93697715e-01 -1.34627551e-01 -4.91830349e-01
-1.22749412e+00 -7.41755903e-01 7.25950658e-01 3.80059481e-01
5.24898708e-01 -5.52960970e-02 -3.03057760e-01 1.16381323e+00
-6.21707797e-01 1.44642293e-01 3.21905375e-01 7.24550843e-01
-7.23529905e-02 -1.24345171e+00 2.27509201e-01 5.53208649e-01
-3.21575046e-01 -4.75619882e-01 -6.22714043e-01 8.76567602e-01
3.39672893e-01 9.65200007e-01 -2.24460065e-01 -4.94228363e-01
3.89231235e-01 3.61582130e-01 3.43253732e-01 -1.16670847e+00
-5.80118120e-01 -5.69615960e-01 9.54428375e-01 -4.71553028e-01
-4.64318395e-01 -3.20204258e-01 -1.20283735e+00 -7.06487074e-02
-4.11090165e-01 3.35983902e-01 1.09249443e-01 1.18616879e+00
3.23888332e-01 6.54205859e-01 -2.88253482e-02 -6.95245638e-02
-1.41176477e-01 -1.21517539e+00 -2.32827619e-01 4.67524976e-01
-4.26011384e-02 -8.03154051e-01 -3.71541619e-01 -3.08372229e-01] | [9.65295696258545, 8.550941467285156] |
53f301f3-efdf-46ac-a79a-8291857bfb94 | improving-language-plasticity-via-pretraining | 2307.01163 | null | https://arxiv.org/abs/2307.01163v2 | https://arxiv.org/pdf/2307.01163v2.pdf | Improving Language Plasticity via Pretraining with Active Forgetting | Pretrained language models (PLMs) are today the primary model for natural language processing. Despite their impressive downstream performance, it can be difficult to apply PLMs to new languages, a barrier to making their capabilities universally accessible. While prior work has shown it possible to address this issue by learning a new embedding layer for the new language, doing so is both data and compute inefficient. We propose to use an active forgetting mechanism during pretraining, as a simple way of creating PLMs that can quickly adapt to new languages. Concretely, by resetting the embedding layer every K updates during pretraining, we encourage the PLM to improve its ability of learning new embeddings within a limited number of updates, similar to a meta-learning effect. Experiments with RoBERTa show that models pretrained with our forgetting mechanism not only demonstrate faster convergence during language adaptation but also outperform standard ones in a low-data regime, particularly for languages that are distant from English. | ['Mikel Artetxe', 'Pontus Stenetorp', 'Sebastian Riedel', 'David Ifeoluwa Adelani', 'Roberta Raileanu', 'Kelly Marchisio', 'Yihong Chen'] | 2023-07-03 | null | null | null | null | ['meta-learning'] | ['methodology'] | [-4.68662046e-02 3.88345532e-02 -2.37059563e-01 -4.40761536e-01
-3.65582526e-01 -6.36025250e-01 7.18054473e-01 4.01366323e-01
-1.33223796e+00 5.52777171e-01 3.71749550e-01 -4.66493756e-01
3.58780861e-01 -7.57664084e-01 -7.93109119e-01 -2.56076366e-01
-7.28541687e-02 5.72292566e-01 4.36244190e-01 -4.12493140e-01
6.14858158e-02 6.65161788e-01 -1.24393916e+00 1.55528635e-01
8.74543130e-01 -3.90422605e-02 5.10724008e-01 6.60086334e-01
-5.68907440e-01 4.46122795e-01 -2.65585154e-01 -4.66697484e-01
2.60965019e-01 -2.03104869e-01 -8.36547136e-01 -2.48168960e-01
1.74614564e-01 -3.04123461e-01 -2.56259650e-01 6.86908841e-01
4.12328899e-01 4.49981779e-01 3.73324215e-01 -6.76281095e-01
-9.91085887e-01 1.03329682e+00 -2.73788244e-01 5.70373654e-01
1.16646126e-01 3.18875849e-01 9.23571587e-01 -1.18174779e+00
6.93708599e-01 1.24137473e+00 7.40184486e-01 1.01743734e+00
-1.33741677e+00 -5.26160240e-01 6.56097591e-01 -1.92681029e-01
-1.21939349e+00 -7.15040863e-01 3.93581659e-01 -1.74484879e-01
1.37323213e+00 -1.11612752e-01 6.01398826e-01 9.25657272e-01
2.61169404e-01 9.17578638e-01 7.35755920e-01 -7.08409131e-01
1.64112285e-01 4.89277035e-01 1.24638647e-01 6.59294188e-01
4.02472019e-01 -6.12037666e-02 -7.09324718e-01 -1.21627234e-01
5.61410487e-01 3.99840996e-02 -4.20514122e-02 -3.05799425e-01
-1.23720300e+00 8.07246745e-01 4.95836169e-01 5.96402466e-01
-3.46246064e-01 5.62154129e-02 5.09110689e-01 7.44338214e-01
5.98454297e-01 8.34158897e-01 -7.09483981e-01 -9.78508219e-02
-8.48275781e-01 -2.07247790e-02 7.39415288e-01 6.29625082e-01
9.82704580e-01 1.84797943e-01 2.21507728e-01 9.54872072e-01
2.16856301e-01 2.33641833e-01 9.66979206e-01 -4.47659612e-01
4.02031332e-01 5.48452973e-01 -1.21885121e-01 -5.37083387e-01
-3.86344433e-01 -3.34798127e-01 -7.48808742e-01 7.69516006e-02
3.02547812e-01 -2.90220708e-01 -9.67356861e-01 2.00783467e+00
1.70105919e-01 1.29285648e-01 1.81368992e-01 4.55346495e-01
2.51369953e-01 8.78380060e-01 4.68292743e-01 -1.50688648e-01
9.59164798e-01 -1.07475865e+00 -3.33006948e-01 -8.68595779e-01
8.76683652e-01 -6.49295986e-01 1.60696793e+00 2.51899630e-01
-1.09298241e+00 -6.20477319e-01 -9.75898325e-01 -2.73243934e-01
-5.77048182e-01 -3.26975584e-01 8.76414299e-01 5.56772411e-01
-1.29155815e+00 4.98657614e-01 -1.12922633e+00 -7.80370474e-01
1.08198635e-01 3.55824530e-01 -5.16456485e-01 -1.46890834e-01
-1.24231219e+00 1.21551371e+00 7.39325583e-01 -5.03478087e-02
-6.38054609e-01 -8.24824214e-01 -8.19901288e-01 -7.44912680e-03
1.30164534e-01 -8.28952253e-01 1.18765759e+00 -1.26948857e+00
-1.69004619e+00 7.61932850e-01 -3.02531183e-01 -6.43928468e-01
3.72100681e-01 -3.98737520e-01 -4.10313874e-01 -2.94719279e-01
-2.60074764e-01 8.48401010e-01 1.01181662e+00 -1.09564018e+00
-4.78958726e-01 -1.51478481e-02 5.93872517e-02 3.31773788e-01
-8.47666025e-01 -2.36454816e-03 -4.61183846e-01 -6.29867494e-01
-2.67172381e-02 -7.97759473e-01 -4.94731694e-01 -3.72089595e-02
2.58688062e-01 -2.37489775e-01 3.23368996e-01 -5.77691734e-01
1.46545172e+00 -2.19428182e+00 2.22806945e-01 6.17980286e-02
9.84308273e-02 6.72067046e-01 -6.21370614e-01 5.31914294e-01
1.72910020e-01 3.18330497e-01 -3.10038656e-01 -6.79319561e-01
-8.28291476e-02 5.87740183e-01 -4.97950673e-01 2.17510741e-02
4.28873986e-01 9.97683525e-01 -1.10386086e+00 -3.09198815e-02
-5.13503365e-02 4.28413153e-01 -9.59909320e-01 1.03829160e-01
-1.66641980e-01 1.19641080e-01 -5.80625385e-02 1.39389783e-02
4.67966467e-01 -9.91326869e-02 2.15303093e-01 3.68523389e-01
-2.94433683e-01 5.23793638e-01 -1.10027158e+00 1.77318633e+00
-1.04945755e+00 4.36287165e-01 -1.08008005e-01 -6.87267005e-01
8.09182703e-01 1.69809878e-01 -7.79031590e-02 -5.11817396e-01
-1.86167300e-01 1.89714432e-01 2.49571577e-01 -1.55829087e-01
8.17657650e-01 -4.32127088e-01 2.38995943e-02 8.40996802e-01
3.04058194e-01 -1.34556204e-01 2.61673510e-01 2.77179867e-01
1.01531374e+00 -1.13927372e-01 3.05992365e-01 -2.01925814e-01
4.48499233e-01 -2.37865612e-01 5.06143332e-01 9.46226001e-01
-7.17125684e-02 3.05163950e-01 -1.39999598e-01 -6.57176733e-01
-1.16830587e+00 -1.27076697e+00 1.85970619e-01 1.61922300e+00
-3.60618740e-01 -6.03322864e-01 -4.01784241e-01 -6.39034390e-01
7.44052306e-02 8.21822464e-01 -4.63589311e-01 -5.65345764e-01
-8.97665918e-01 -8.67222369e-01 5.23661911e-01 5.76472282e-01
3.05482417e-01 -1.12485385e+00 -4.86888975e-01 6.05661690e-01
2.56492078e-01 -7.44998276e-01 -6.25543475e-01 4.36172843e-01
-1.16216755e+00 -3.36450130e-01 -6.71692073e-01 -9.76343691e-01
9.42787826e-01 1.50108486e-01 1.19941759e+00 4.09284860e-01
-4.19990420e-02 4.14759070e-01 -2.01956153e-01 -2.85390049e-01
-7.65265346e-01 7.52053857e-01 4.31340635e-01 -1.41334936e-01
4.89266306e-01 -5.62395573e-01 -2.24028304e-01 -1.89562261e-01
-1.12800109e+00 7.47456551e-02 6.26534104e-01 8.94668758e-01
2.57495403e-01 -1.64012611e-01 6.31487250e-01 -1.17972660e+00
9.26081419e-01 -4.64916945e-01 -4.98916090e-01 2.48597428e-01
-8.13974619e-01 4.83151615e-01 9.15494144e-01 -7.64327347e-01
-1.04076314e+00 8.29544067e-02 -3.82742018e-01 1.37911573e-01
5.43392412e-02 6.60896540e-01 2.75068253e-01 1.97729077e-02
8.56739819e-01 2.75357991e-01 -1.08087167e-01 -6.26609862e-01
6.45559311e-01 3.39971244e-01 2.83390939e-01 -6.41793668e-01
1.19901919e+00 2.26583704e-01 -8.09031248e-01 -9.35304761e-01
-7.29697287e-01 -3.48478049e-01 -9.69989419e-01 2.40352079e-01
4.90671933e-01 -9.38897610e-01 -1.03250146e-01 2.61003166e-01
-1.08599472e+00 -7.58293569e-01 -5.51086009e-01 6.06921136e-01
-9.45757255e-02 1.84681699e-01 -6.94501340e-01 -4.79758590e-01
-2.62571007e-01 -7.28001773e-01 4.48935926e-01 2.80456752e-01
-4.54497039e-01 -1.59808552e+00 4.13841337e-01 -1.97278321e-01
9.78914082e-01 -5.31562865e-01 1.15381598e+00 -7.14172184e-01
-1.82248071e-01 -8.04558173e-02 1.51024655e-01 5.23297906e-01
1.70662642e-01 -1.97479259e-02 -8.87689114e-01 -6.86183989e-01
-1.96162924e-01 -2.83782393e-01 1.05686021e+00 -1.59094408e-01
6.99165761e-01 -2.77127385e-01 -2.87412852e-01 5.73725879e-01
1.29379427e+00 -1.21989720e-01 3.11168700e-01 4.89637703e-01
4.19068515e-01 1.35071427e-01 -3.46457995e-02 1.58308327e-01
5.72357476e-01 2.18344346e-01 -2.98884839e-01 -2.03462854e-01
-2.33870044e-01 -5.55478990e-01 7.22676098e-01 1.64100289e+00
1.62860021e-01 -1.10252619e-01 -1.26433718e+00 8.08493376e-01
-1.60034215e+00 -6.97719872e-01 5.16468346e-01 2.40567040e+00
1.27461398e+00 5.35311818e-01 -1.08649477e-01 -2.86994040e-01
2.03787267e-01 1.28622040e-01 -6.45062566e-01 -7.93733239e-01
-2.26008222e-01 5.25652409e-01 4.34117943e-01 9.70054686e-01
-7.17334390e-01 1.52150881e+00 7.24370909e+00 2.31985524e-01
-1.41210985e+00 3.80242735e-01 3.09890360e-01 -1.33212969e-01
-5.98109007e-01 1.19069755e-01 -9.69996095e-01 2.32423738e-01
1.32720244e+00 -5.69662988e-01 7.11794436e-01 6.35334849e-01
-6.97467625e-02 -8.30904692e-02 -1.07466185e+00 6.19977355e-01
1.85024768e-01 -1.16069162e+00 4.99226809e-01 -4.74734634e-01
7.09435225e-01 5.54703534e-01 8.63301605e-02 7.10080266e-01
6.99890375e-01 -9.02609587e-01 5.81092060e-01 3.62767369e-01
4.41224843e-01 -6.67060137e-01 5.20579398e-01 6.37570798e-01
-8.70359123e-01 -1.32666007e-01 -6.73062563e-01 -3.07337523e-01
1.60885170e-01 2.80421317e-01 -9.57346082e-01 -5.07468879e-02
5.34480453e-01 5.39820671e-01 -9.77854609e-01 8.65999222e-01
-4.64679360e-01 8.40380728e-01 -3.77480805e-01 2.83254627e-02
1.80892020e-01 9.19179246e-02 3.15720856e-01 1.33472180e+00
2.20055878e-01 -1.08127125e-01 1.29479945e-01 5.63788652e-01
-3.01730722e-01 2.95855105e-01 -8.00856113e-01 -2.81896919e-01
5.68034351e-01 9.66917157e-01 -5.49848735e-01 -4.38077003e-01
-5.50101995e-01 1.22053218e+00 8.41907442e-01 4.42693323e-01
-2.77759790e-01 -3.78709018e-01 6.53037608e-01 3.01048905e-01
1.93961605e-01 -8.29370558e-01 -9.14092287e-02 -1.47315240e+00
-1.83024377e-01 -7.33897805e-01 4.38742816e-01 -5.05966306e-01
-1.35127521e+00 7.64768720e-01 -8.40392858e-02 -6.08420193e-01
-2.70766616e-01 -6.22785509e-01 -4.87227410e-01 8.86656642e-01
-1.67038500e+00 -9.19078231e-01 2.92987376e-01 3.84940207e-01
6.38588309e-01 -2.95190424e-01 1.06693077e+00 2.02451259e-01
-3.47957730e-01 7.82884002e-01 2.02616364e-01 4.67668986e-03
9.05179858e-01 -1.08781981e+00 8.40067267e-01 1.11179972e+00
5.36331236e-01 1.34351289e+00 6.70178592e-01 -3.97726625e-01
-1.44817197e+00 -1.00027382e+00 1.32392681e+00 -6.21678472e-01
8.33538651e-01 -8.07099819e-01 -1.37807345e+00 1.06014466e+00
3.56065899e-01 -1.68803930e-01 7.01510847e-01 5.07751584e-01
-6.20955348e-01 -1.84320897e-01 -7.74099588e-01 1.01340282e+00
1.02499735e+00 -7.09794521e-01 -9.94448304e-01 2.59762913e-01
8.74475360e-01 -1.16876900e-01 -5.96297801e-01 -2.00540107e-02
2.72293627e-01 -4.02659416e-01 8.01139534e-01 -8.63868058e-01
-1.35454074e-01 -2.79401213e-01 1.66127563e-01 -1.75928736e+00
-5.17308354e-01 -8.97842109e-01 6.36783149e-03 1.09920394e+00
8.14710796e-01 -9.67893958e-01 4.60442930e-01 5.42166531e-01
7.41494596e-02 -2.90297478e-01 -8.60110044e-01 -9.10554409e-01
6.43774927e-01 -5.24577796e-01 4.58171368e-01 1.03345573e+00
7.73828551e-02 5.26989043e-01 -1.82663471e-01 1.10583387e-01
2.20515709e-02 -3.67509991e-01 7.89117396e-01 -1.16462445e+00
-3.95774484e-01 -4.32759196e-01 -4.57554534e-02 -1.18123162e+00
4.01962906e-01 -1.25785232e+00 1.13073878e-01 -1.43979621e+00
1.21422961e-01 -6.51188731e-01 -6.71282589e-01 7.43015110e-01
-3.89420509e-01 1.12742543e-01 3.16998750e-01 1.44313589e-01
-3.98827821e-01 3.74701828e-01 9.00222063e-01 -4.06969562e-02
-4.72099304e-01 -2.37138554e-01 -7.90678859e-01 7.27669716e-01
8.21197748e-01 -6.47657275e-01 -4.80693787e-01 -1.20931375e+00
5.99150300e-01 -7.87344992e-01 -1.52991027e-01 -1.04069352e+00
4.35148418e-01 -1.77837148e-01 2.54834414e-01 4.20465134e-02
8.49451870e-02 -6.19526207e-01 -1.85148910e-01 6.12917066e-01
-5.65182984e-01 6.13587081e-01 6.86386347e-01 3.65173787e-01
-4.12232168e-02 -4.06147599e-01 8.13775897e-01 -3.11170638e-01
-9.45553422e-01 3.35478961e-01 -7.62294590e-01 1.30275443e-01
6.05400026e-01 7.17846304e-02 6.25089779e-02 -1.94215149e-01
-7.47270882e-01 1.86991781e-01 6.94445789e-01 8.44812155e-01
4.99029338e-01 -1.18462038e+00 -6.55118346e-01 5.78543663e-01
2.28167310e-01 -1.54025614e-01 -1.08119369e-01 5.12841821e-01
-4.58409816e-01 2.33638808e-01 -1.26716927e-01 -1.63984835e-01
-8.95696044e-01 6.42592192e-01 2.48942658e-01 -4.00426269e-01
-6.28131807e-01 1.07433677e+00 -4.63431291e-02 -1.03845692e+00
1.30728826e-01 -5.06946087e-01 4.89415452e-02 -3.56200039e-02
7.26142883e-01 -2.81542659e-01 1.22538604e-01 -2.52300739e-01
-2.82361507e-01 3.40330303e-01 -6.87487721e-01 -3.30273449e-01
1.55796897e+00 -2.08613098e-01 -1.26025006e-01 9.32715654e-01
8.75300407e-01 1.95289984e-01 -1.11766779e+00 -7.49704838e-01
4.05506015e-01 -2.35072657e-01 -6.43121004e-02 -7.20382690e-01
-5.93267500e-01 1.09994650e+00 5.11599600e-01 -5.35778962e-02
8.88893843e-01 -2.08971128e-01 8.14385653e-01 7.92616546e-01
2.86923081e-01 -1.28829277e+00 1.64377019e-01 1.16246915e+00
6.50539935e-01 -1.17379606e+00 -2.66971719e-02 4.26960021e-01
-5.69482923e-01 1.14891648e+00 5.88014543e-01 -1.19386479e-01
7.30554223e-01 3.72687161e-01 3.38023245e-01 3.34810406e-01
-1.21092582e+00 -2.64347764e-03 7.48561695e-02 4.78327453e-01
6.87032342e-01 -7.46223405e-02 -1.50881037e-01 3.58213514e-01
-2.45985910e-01 -1.39795825e-01 4.92273062e-01 1.17571592e+00
-6.43573105e-01 -1.54125595e+00 8.91480595e-02 2.28380039e-01
-7.48790577e-02 -4.70471740e-01 -1.66098669e-01 8.42963934e-01
2.26737149e-02 4.68080461e-01 2.06869230e-01 -6.97088763e-02
3.34419370e-01 5.89008152e-01 3.48909527e-01 -1.21105456e+00
-7.73142457e-01 -4.61016178e-01 -2.72556275e-01 -2.66839713e-01
-8.53613019e-02 -6.67045295e-01 -1.26889706e+00 -3.51143628e-01
-1.47341251e-01 1.17750935e-01 4.78431791e-01 9.35910046e-01
4.24332976e-01 3.39270204e-01 3.35183799e-01 -5.55168390e-01
-7.37652004e-01 -8.65645409e-01 -1.62455484e-01 3.45258564e-01
4.27902371e-01 -1.99379012e-01 -2.71596640e-01 2.21542686e-01] | [10.59633731842041, 8.5772123336792] |
20d02ba6-dc78-4306-966b-2a15d53fbdd8 | exposing-flaws-of-generative-model-evaluation | 2306.04675 | null | https://arxiv.org/abs/2306.04675v1 | https://arxiv.org/pdf/2306.04675v1.pdf | Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | We systematically study a wide variety of image-based generative models spanning semantically-diverse datasets to understand and improve the feature extractors and metrics used to evaluate them. Using best practices in psychophysics, we measure human perception of image realism for generated samples by conducting the largest experiment evaluating generative models to date, and find that no existing metric strongly correlates with human evaluations. Comparing to 16 modern metrics for evaluating the overall performance, fidelity, diversity, and memorization of generative models, we find that the state-of-the-art perceptual realism of diffusion models as judged by humans is not reflected in commonly reported metrics such as FID. This discrepancy is not explained by diversity in generated samples, though one cause is over-reliance on Inception-V3. We address these flaws through a study of alternative self-supervised feature extractors, find that the semantic information encoded by individual networks strongly depends on their training procedure, and show that DINOv2-ViT-L/14 allows for much richer evaluation of generative models. Next, we investigate data memorization, and find that generative models do memorize training examples on simple, smaller datasets like CIFAR10, but not necessarily on more complex datasets like ImageNet. However, our experiments show that current metrics do not properly detect memorization; none in the literature is able to separate memorization from other phenomena such as underfitting or mode shrinkage. To facilitate further development of generative models and their evaluation we release all generated image datasets, human evaluation data, and a modular library to compute 16 common metrics for 8 different encoders at https://github.com/layer6ai-labs/dgm-eval. | ['Gabriel Loaiza-Ganem', 'J. Eric T. Taylor', 'Anthony L. Caterini', 'Zhaoyan Liu', 'Valentin Villecroze', 'Brendan Leigh Ross', 'Yi Sui', 'Rasa Hosseinzadeh', 'Jesse C. Cresswell', 'George Stein'] | 2023-06-07 | null | null | null | null | ['memorization'] | ['natural-language-processing'] | [ 5.61723858e-02 -4.67352644e-02 2.47966334e-01 -5.44821441e-01
-6.02391720e-01 -6.72936380e-01 9.01010394e-01 -8.14969987e-02
-5.09215355e-01 5.98271668e-01 4.11101490e-01 -1.81286409e-01
-1.91319734e-01 -6.94183826e-01 -7.48752296e-01 -5.87315261e-01
6.05424345e-02 3.86158884e-01 -1.39416918e-01 -1.15838371e-01
3.83020312e-01 2.32316405e-01 -1.51236880e+00 3.53386104e-01
6.57698512e-01 7.98965752e-01 8.35831910e-02 8.35835099e-01
4.07563865e-01 8.04773927e-01 -8.11602294e-01 -5.42226315e-01
3.02675784e-01 -8.78178477e-01 -8.93628180e-01 1.02879569e-01
8.58112216e-01 -4.68880385e-01 -4.36554849e-01 8.41616988e-01
7.51176238e-01 1.26809239e-01 8.91360104e-01 -1.26815355e+00
-1.42712331e+00 8.74215662e-01 -1.77339494e-01 5.88933289e-01
1.36376947e-01 8.30974162e-01 8.40769291e-01 -9.28788781e-01
7.62615204e-01 1.48151851e+00 7.04122245e-01 4.71310735e-01
-1.40928674e+00 -8.17226171e-01 -1.57730415e-01 2.00438146e-02
-1.29852748e+00 -6.15851164e-01 3.59988719e-01 -5.20476997e-01
1.07701075e+00 1.18274186e-02 7.57573664e-01 1.55195153e+00
3.26103717e-01 4.70481902e-01 1.38884473e+00 -9.94213223e-02
2.10616991e-01 1.52159393e-01 1.42094657e-01 6.24311268e-01
5.03987372e-01 3.99707377e-01 -5.69004774e-01 1.02965996e-01
9.62737501e-01 -3.18862230e-01 -3.63223970e-01 -7.95907080e-02
-1.17661142e+00 9.13798690e-01 6.91412807e-01 3.08957994e-01
-2.83849359e-01 4.53761131e-01 2.28234321e-01 4.43511069e-01
3.69887799e-01 7.70854592e-01 -8.67112726e-02 -1.28378481e-01
-1.13106787e+00 3.70535523e-01 5.81305265e-01 8.01847577e-01
7.81200945e-01 4.27362800e-01 -1.72185570e-01 5.58741152e-01
2.73667693e-01 5.84604144e-01 7.36531496e-01 -1.32723486e+00
3.57871912e-02 1.78197101e-01 -1.93450034e-01 -1.02596903e+00
-4.80306774e-01 -8.53129923e-01 -6.21824741e-01 3.32464635e-01
3.91046613e-01 -1.92285106e-02 -1.07848036e+00 1.97387195e+00
-3.51810724e-01 -2.42890388e-01 -7.41410553e-02 9.59443033e-01
1.07461023e+00 2.77350128e-01 2.52433479e-01 2.01824620e-01
1.12839413e+00 -7.39039719e-01 -1.94944501e-01 -3.96102965e-01
4.00856256e-01 -6.86353207e-01 1.33753788e+00 5.68706870e-01
-1.35760021e+00 -9.49166238e-01 -1.24073648e+00 -1.71390653e-01
-2.55344599e-01 -2.44230613e-01 8.50349665e-01 7.71732926e-01
-1.35934651e+00 8.20271969e-01 -6.92610383e-01 -5.23686588e-01
7.00006843e-01 1.58909082e-01 -2.17885301e-01 -4.84379642e-02
-1.02753329e+00 1.22063315e+00 2.38526329e-01 -1.99038237e-01
-1.36790776e+00 -7.82389462e-01 -8.54291797e-01 2.59123389e-02
-2.88388163e-01 -1.18227637e+00 1.15017700e+00 -1.31846213e+00
-1.03092182e+00 9.25224245e-01 1.89673409e-01 -5.65512538e-01
3.63858968e-01 -2.20171884e-02 -4.26216274e-01 -4.75169383e-02
-7.01121166e-02 1.18391848e+00 8.56414616e-01 -1.51602840e+00
-4.44708392e-02 -1.82416752e-01 4.72160056e-02 1.01858594e-01
-1.98510230e-01 -3.50930691e-01 5.01732156e-02 -5.54443538e-01
-1.97426856e-01 -7.96450913e-01 -5.95679022e-02 -2.95190096e-01
-3.33906204e-01 1.42344862e-01 2.36946031e-01 -4.52769428e-01
9.18348730e-01 -2.11273956e+00 -1.44206583e-01 1.76307291e-01
4.71987009e-01 1.29996955e-01 -6.30368769e-01 3.47935379e-01
-2.18694061e-01 4.59222198e-01 -1.91071451e-01 -5.14361501e-01
1.58129364e-01 1.77510679e-01 -3.34981620e-01 4.95849460e-01
2.73807138e-01 1.23816490e+00 -7.36109018e-01 -2.30345786e-01
3.02277893e-01 7.71108687e-01 -7.66715288e-01 -2.12362140e-01
2.13992987e-02 2.83237666e-01 1.50639981e-01 3.57691705e-01
6.70531869e-01 -4.61626083e-01 -1.32603317e-01 -2.34959736e-01
1.58071160e-01 4.50354934e-01 -8.73521924e-01 1.84604931e+00
-3.28517228e-01 9.10106480e-01 -5.34496546e-01 -6.53077781e-01
7.04605222e-01 8.38391203e-03 1.21728133e-03 -1.12364686e+00
3.26048374e-01 7.39823747e-03 3.82665992e-01 -2.49199569e-01
6.74138248e-01 -9.02807713e-02 1.03470288e-01 6.97652996e-01
6.16762042e-01 -2.07109779e-01 2.98019946e-01 3.62507522e-01
1.12351668e+00 5.96629344e-02 9.60916504e-02 -4.33074057e-01
-3.63436997e-01 -3.79901789e-02 1.08236879e-01 1.13530588e+00
-1.32042110e-01 8.32182825e-01 1.58297166e-01 -2.50458807e-01
-1.16084731e+00 -1.45841789e+00 -2.67456532e-01 1.02374363e+00
-5.88065945e-02 -4.91928607e-01 -9.52830553e-01 -2.28252739e-01
-3.49658169e-02 9.41267967e-01 -8.83080244e-01 -4.57128614e-01
4.15115356e-02 -9.93454576e-01 8.25467348e-01 5.28104544e-01
4.19802457e-01 -1.20502532e+00 -8.63666713e-01 6.75736740e-02
3.65245529e-02 -8.70999157e-01 -1.72316432e-01 1.36130139e-01
-8.31075788e-01 -8.67967725e-01 -4.47193086e-01 -3.30871254e-01
6.09501004e-01 1.39728934e-01 1.54767513e+00 1.79505646e-01
-3.68242651e-01 5.63605487e-01 -1.79104120e-01 -2.60888994e-01
-6.32510722e-01 4.65118811e-02 -9.05689821e-02 -5.96958995e-01
2.75761187e-01 -7.06833661e-01 -9.54794943e-01 2.66774744e-01
-1.08981681e+00 -3.77108604e-02 7.50313699e-01 6.41033113e-01
2.84385741e-01 -9.65205878e-02 4.78366882e-01 -8.02997828e-01
1.00802684e+00 -5.50400257e-01 -1.32744715e-01 -5.40894195e-02
-9.79618728e-01 2.06946626e-01 2.62029558e-01 -5.45762062e-01
-8.98298800e-01 -4.83813465e-01 -7.73124695e-02 -2.98581183e-01
-2.67795920e-01 5.23970366e-01 1.89657554e-01 -6.56157434e-02
1.31335700e+00 1.50782377e-01 -1.20995417e-01 -5.15326224e-02
5.77172041e-01 2.56742954e-01 6.91968381e-01 -5.17208934e-01
7.02905536e-01 2.95177728e-01 -4.04959232e-01 -5.75006068e-01
-6.82791531e-01 1.93579704e-01 -2.49077186e-01 -5.82505800e-02
6.66297972e-01 -1.11285138e+00 -6.90577865e-01 3.45554471e-01
-1.00745595e+00 -5.99367380e-01 -5.38503826e-01 5.39265215e-01
-7.11453438e-01 7.18201324e-02 -7.59293437e-01 -3.99198204e-01
-3.64219278e-01 -1.20143843e+00 7.61951268e-01 2.27795914e-01
-6.70511663e-01 -1.06506038e+00 2.13301450e-01 3.44777077e-01
9.95349348e-01 1.02076821e-01 8.18223000e-01 -5.67218781e-01
-5.44017553e-01 2.31231332e-01 -2.47432783e-01 4.62543398e-01
-5.84518574e-02 1.67247564e-01 -1.34333301e+00 -2.83710569e-01
-6.44020885e-02 -7.05139816e-01 1.21263480e+00 5.67254245e-01
9.94536519e-01 -2.18299344e-01 1.58708811e-01 7.78718054e-01
1.40771055e+00 -2.64792480e-02 1.07536924e+00 1.93609267e-01
4.25254494e-01 3.13108057e-01 -3.61376777e-02 3.62636209e-01
5.12501538e-01 2.26798490e-01 3.28629464e-01 -1.46446481e-01
-4.54595923e-01 -3.60071331e-01 5.43532133e-01 6.56521618e-01
-1.64602444e-01 -5.63376307e-01 -7.19012380e-01 4.20866847e-01
-1.34894478e+00 -1.25867391e+00 9.88715962e-02 1.98717558e+00
8.51172149e-01 3.85865331e-01 1.14253223e-01 -6.19482398e-02
2.66724795e-01 7.53301084e-02 -5.46305120e-01 -5.17281294e-01
-2.77993202e-01 6.10544980e-01 3.98078442e-01 5.08890569e-01
-7.17051744e-01 9.33130860e-01 7.64782190e+00 6.93849742e-01
-1.13989639e+00 2.13733703e-01 8.80101204e-01 -3.25208545e-01
-6.30687296e-01 2.80304346e-02 -5.25255859e-01 3.77003998e-01
1.16856420e+00 -2.23019555e-01 6.59606636e-01 6.45975828e-01
-2.13000048e-02 -2.19389394e-01 -1.34805679e+00 9.66493964e-01
2.78258771e-01 -1.24624407e+00 2.59444714e-01 9.60911959e-02
8.11680853e-01 2.47536495e-01 6.49408340e-01 4.27531302e-01
4.62004811e-01 -1.56581807e+00 9.20783877e-01 6.97348416e-01
6.68230176e-01 -5.28589129e-01 3.58138442e-01 1.16465436e-02
-5.03903568e-01 1.50744066e-01 -6.22072339e-01 -1.18411131e-01
-1.96731482e-02 7.44195938e-01 -6.58776343e-01 5.22105694e-02
6.23981357e-01 5.53952336e-01 -1.11042440e+00 8.70282233e-01
-1.68805897e-01 7.93041587e-01 -1.48690462e-01 2.24211976e-01
4.45903055e-02 3.26764852e-01 1.95602074e-01 1.40504646e+00
2.93331295e-01 -2.00630113e-01 -3.49593431e-01 1.47980309e+00
-1.86205178e-01 -1.62929788e-01 -5.93254268e-01 -2.40068942e-01
3.92765671e-01 1.25577629e+00 -7.58817792e-01 -3.21610570e-01
-1.46402657e-01 9.69592690e-01 2.36679479e-01 5.04561543e-01
-1.02029502e+00 -2.14204919e-02 4.85675365e-01 1.47866249e-01
5.87997437e-02 -4.72719610e-01 -4.49067771e-01 -1.02656913e+00
-2.58685172e-01 -1.11556768e+00 9.89060625e-02 -1.21981001e+00
-1.36779141e+00 8.15473020e-01 1.57598183e-01 -6.79847896e-01
-4.23689038e-01 -5.39076567e-01 -5.78404248e-01 8.73882771e-01
-1.02572334e+00 -8.89216721e-01 -6.47259474e-01 6.59766972e-01
3.69073361e-01 -4.59821634e-02 7.28404224e-01 9.46184322e-02
-2.14832604e-01 7.35951066e-01 -4.40552831e-01 -9.70415547e-02
8.87096941e-01 -1.05989325e+00 7.34360576e-01 8.32516015e-01
5.54773688e-01 1.02040219e+00 8.43486905e-01 -4.88125533e-01
-1.01492906e+00 -7.03299940e-01 4.47937459e-01 -7.89378583e-01
4.04167205e-01 -3.41271460e-01 -7.57995963e-01 8.36119950e-01
6.56334937e-01 -1.51047856e-01 6.56184554e-01 1.14002936e-01
-6.12318814e-01 1.75382659e-01 -9.91360307e-01 4.95527595e-01
1.45101511e+00 -5.11068940e-01 -4.49124455e-01 6.77570254e-02
5.32029092e-01 -2.05854401e-01 -9.12293196e-01 2.61092722e-01
6.28246784e-01 -1.37000144e+00 9.26356673e-01 -5.69609582e-01
7.57070422e-01 5.40974084e-03 -2.32262865e-01 -1.62471330e+00
-6.97803855e-01 -2.92113513e-01 1.69878677e-01 1.21093512e+00
6.19557619e-01 -6.04122996e-01 4.89190072e-01 5.08397996e-01
-1.68742165e-01 -4.73283172e-01 -4.26453710e-01 -6.69414759e-01
3.07790279e-01 -5.83586574e-01 3.73419434e-01 9.70906973e-01
-3.68819773e-01 4.68377262e-01 -8.94928426e-02 -3.08564723e-01
6.90554082e-01 -1.71537280e-01 7.86264539e-01 -9.55262661e-01
-5.31457305e-01 -7.01800108e-01 -6.38178766e-01 -6.59039915e-01
-2.73373742e-02 -1.12980747e+00 -1.75419018e-01 -1.46943200e+00
4.65944260e-01 -3.36789876e-01 -2.15054870e-01 5.51140726e-01
3.73061523e-02 5.37563622e-01 2.79641747e-01 2.35332698e-01
-5.00690222e-01 4.76287693e-01 1.31720793e+00 -2.16196254e-02
1.39496148e-01 -5.56225121e-01 -1.18661225e+00 5.63165545e-01
8.19864392e-01 -3.94998223e-01 -8.37643981e-01 -6.88439310e-01
3.08086008e-01 -3.86900425e-01 9.36980844e-01 -1.41356432e+00
1.28097646e-02 3.88243385e-02 1.04892051e+00 -3.63473296e-02
1.99808136e-01 -2.69486248e-01 4.66797918e-01 3.03975582e-01
-5.08425534e-01 3.00671577e-01 3.87220770e-01 7.16467798e-02
5.00388443e-03 -1.21566884e-01 8.24389577e-01 -2.52424449e-01
-7.79049516e-01 2.04714894e-01 -3.69680554e-01 4.88279641e-01
5.36438882e-01 -2.29319111e-01 -9.01203871e-01 -6.16902709e-01
-6.53727889e-01 -2.16979489e-01 7.56682336e-01 4.39166516e-01
6.57174647e-01 -1.20475399e+00 -8.59709442e-01 2.17358410e-01
5.72330952e-02 -4.82782900e-01 3.38341087e-01 6.83032870e-01
-3.16289753e-01 6.18407242e-02 -5.50373673e-01 -6.06206894e-01
-7.15939403e-01 4.08003360e-01 4.99764949e-01 6.97477758e-02
-3.15658629e-01 8.41016829e-01 3.72156829e-01 -1.81337502e-02
-1.22596651e-01 -4.41330016e-01 3.09031576e-01 -4.95510027e-02
3.72059047e-01 1.21195011e-01 1.05843976e-01 -4.55180734e-01
-2.38796920e-01 2.82848150e-01 -9.73915029e-03 -3.58173817e-01
1.19105065e+00 -3.09613859e-03 3.17662209e-01 2.97316223e-01
1.13528514e+00 -3.21162194e-01 -1.32433259e+00 5.76455891e-02
-4.52857941e-01 -2.74220169e-01 -5.73489815e-03 -1.02095044e+00
-1.21089184e+00 8.67790043e-01 7.58721769e-01 -1.05959289e-02
1.03537071e+00 4.88875173e-02 4.31173205e-01 2.84964353e-01
3.05846989e-01 -9.60366547e-01 4.02050644e-01 4.28112298e-01
1.09433520e+00 -1.07483292e+00 8.04817006e-02 1.41536698e-01
-8.74153137e-01 7.05113947e-01 6.77510977e-01 -4.02349442e-01
4.72267926e-01 4.09010977e-01 6.41217157e-02 -4.81251329e-01
-9.11008656e-01 -1.57440573e-01 3.27220768e-01 6.16540790e-01
6.25175655e-01 -5.76138422e-02 -8.42091143e-02 5.07294178e-01
-9.22198176e-01 7.54264221e-02 5.41136265e-01 6.28954768e-01
-4.25140828e-01 -7.52904713e-01 -5.38367108e-02 4.67094034e-01
-2.63706058e-01 -3.89873147e-01 -3.79890561e-01 9.79072392e-01
2.61643142e-01 8.94572675e-01 2.31549755e-01 -6.26254201e-01
1.18716151e-01 9.14807767e-02 1.08119667e+00 -5.11751592e-01
-7.45159030e-01 -2.83587992e-01 -7.64657110e-02 -6.56592369e-01
-5.63228667e-01 -5.28593421e-01 -1.12999713e+00 -7.04860032e-01
-1.21140525e-01 -1.76002324e-01 7.40004420e-01 7.62441337e-01
4.53517079e-01 5.81434488e-01 2.29655564e-01 -1.01494586e+00
-5.59094310e-01 -1.13917577e+00 -5.27343869e-01 7.48052001e-01
2.89873164e-02 -6.81285739e-01 -5.60455203e-01 1.91597253e-01] | [10.33255386352539, 2.105342149734497] |
85e2e5d8-f65c-46f0-be69-14697df33fa1 | tensor-completion-via-tensor-networks-with-a | 2010.15819 | null | https://arxiv.org/abs/2010.15819v1 | https://arxiv.org/pdf/2010.15819v1.pdf | Tensor Completion via Tensor Networks with a Tucker Wrapper | In recent years, low-rank tensor completion (LRTC) has received considerable attention due to its applications in image/video inpainting, hyperspectral data recovery, etc. With different notions of tensor rank (e.g., CP, Tucker, tensor train/ring, etc.), various optimization based numerical methods are proposed to LRTC. However, tensor network based methods have not been proposed yet. In this paper, we propose to solve LRTC via tensor networks with a Tucker wrapper. Here by "Tucker wrapper" we mean that the outermost factor matrices of the tensor network are all orthonormal. We formulate LRTC as a problem of solving a system of nonlinear equations, rather than a constrained optimization problem. A two-level alternative least square method is then employed to update the unknown factors. The computation of the method is dominated by tensor matrix multiplications and can be efficiently performed. Also, under proper assumptions, it is shown that with high probability, the method converges to the exact solution at a linear rate. Numerical simulations show that the proposed algorithm is comparable with state-of-the-art methods. | ['Ping Li', 'Yunfeng Cai'] | 2020-10-29 | null | null | null | null | ['video-inpainting'] | ['computer-vision'] | [ 2.49252155e-01 -3.21532995e-01 2.70472020e-02 1.79350480e-01
-5.68895757e-01 -4.60212708e-01 1.54108152e-01 -2.48449877e-01
-3.26829970e-01 6.37235999e-01 5.71906157e-02 -2.87990242e-01
-6.38212144e-01 -1.86401114e-01 -6.82894051e-01 -1.06408668e+00
-1.13340162e-01 1.59911841e-01 -3.44603896e-01 -2.23032832e-01
1.61906749e-01 4.52731878e-01 -9.70805287e-01 -4.54249419e-02
1.00300503e+00 1.16743910e+00 4.50990051e-02 5.85529268e-01
1.86934367e-01 8.98825407e-01 -7.90488347e-02 -2.67821908e-01
5.66922605e-01 -1.46635637e-01 -9.18299198e-01 6.43736124e-01
4.93626028e-01 -4.58895534e-01 -6.13336980e-01 1.18587840e+00
2.61774212e-01 4.56669748e-01 3.53897780e-01 -1.14162135e+00
-5.79329193e-01 4.33111817e-01 -8.82813692e-01 -1.66689754e-01
2.71808237e-01 -3.46967250e-01 1.00474346e+00 -1.36743975e+00
4.64371234e-01 1.04934633e+00 7.13608086e-01 -5.07033132e-02
-1.49200320e+00 -4.21533674e-01 -9.29681435e-02 2.55749404e-01
-1.60172081e+00 -3.07029128e-01 8.37964773e-01 -5.80334306e-01
2.84031808e-01 4.71542031e-01 5.33714175e-01 4.86891389e-01
-1.29827261e-01 7.87610114e-01 1.32670021e+00 -3.72583807e-01
1.13167971e-01 -2.24986121e-01 1.31190270e-01 8.51868451e-01
5.04226804e-01 -2.46915072e-01 -6.76292241e-01 -4.65445191e-01
6.26989841e-01 2.73931921e-01 -5.46946883e-01 -2.89603502e-01
-1.58951676e+00 9.77409899e-01 3.02793652e-01 2.91605026e-01
-5.68623960e-01 1.01011559e-01 2.05905855e-01 1.04932897e-01
6.53387785e-01 2.37214237e-01 -1.66091561e-01 2.19718710e-01
-9.80657697e-01 3.82413089e-01 8.43618512e-01 8.51721704e-01
8.08830261e-01 2.92959690e-01 5.40898591e-02 9.75400388e-01
1.78327292e-01 6.98889017e-01 4.01303694e-02 -1.03880203e+00
6.52404785e-01 1.21344075e-01 3.30674350e-01 -1.46235085e+00
-2.99739003e-01 -4.80646998e-01 -1.46830857e+00 -2.65369922e-01
3.69787186e-01 -1.17632747e-01 -4.07845080e-01 1.24786568e+00
4.14701998e-01 5.75589180e-01 3.89318541e-02 1.04698420e+00
3.44704509e-01 9.57181513e-01 -5.72046220e-01 -4.45730686e-01
1.25760353e+00 -8.50419343e-01 -9.23104823e-01 1.05342023e-01
5.65327406e-01 -1.14453828e+00 4.51386392e-01 8.94682229e-01
-9.18810070e-01 -3.55469994e-02 -9.78308022e-01 -1.55341983e-01
1.09676585e-01 5.17011404e-01 8.11212897e-01 3.40531796e-01
-8.42893422e-01 8.19965839e-01 -8.11387122e-01 -7.54166767e-02
6.38843477e-02 2.40761191e-01 -5.06766498e-01 -5.38692296e-01
-8.32887530e-01 6.27792418e-01 2.17435256e-01 7.84165144e-01
-7.75168717e-01 -5.99513531e-01 -6.44391894e-01 -3.87231499e-01
6.33951843e-01 -5.87224126e-01 9.64730918e-01 -5.00713468e-01
-1.52178514e+00 3.29671025e-01 -3.00378740e-01 -1.79548323e-01
2.75094122e-01 -3.30488890e-01 -1.93539485e-01 4.14364398e-01
1.81750089e-01 -1.28630266e-01 1.38996649e+00 -1.13205552e+00
-3.43696594e-01 -2.74979711e-01 7.13403448e-02 6.48053437e-02
-4.45038110e-01 -5.09897538e-04 -3.25938880e-01 -9.07118320e-01
7.42826045e-01 -1.36610365e+00 -4.92043972e-01 1.42415509e-01
-6.18829608e-01 -4.76465262e-02 7.03215837e-01 -1.01102722e+00
1.09002995e+00 -2.05727005e+00 6.46674156e-01 3.21154535e-01
4.86692131e-01 1.27191454e-01 -1.02196999e-01 6.90634668e-01
-3.52818429e-01 -9.01481211e-02 -5.45388639e-01 -4.16484952e-01
-2.37659320e-01 2.97325253e-01 -3.68877620e-01 1.03710651e+00
7.11609647e-02 2.87261724e-01 -8.94146085e-01 -3.38838369e-01
1.02848541e-02 4.74213302e-01 -5.32525182e-01 1.14157870e-01
2.03388080e-01 4.79311079e-01 -5.17498136e-01 7.32980132e-01
1.08077669e+00 -2.83998728e-01 9.33444127e-02 -8.40264797e-01
-5.14295459e-01 -3.23366940e-01 -1.61767054e+00 1.71490836e+00
-3.28613997e-01 5.45388043e-01 6.24512732e-01 -1.27098823e+00
5.46065867e-01 3.78466785e-01 7.31885254e-01 -2.92516407e-02
-5.18897101e-02 4.72439021e-01 -1.61203831e-01 -5.50537705e-01
8.92235816e-01 -1.70309424e-01 3.02308142e-01 3.17352802e-01
-1.92617655e-01 -5.00457957e-02 6.04057729e-01 3.78295779e-01
1.03307867e+00 4.91580032e-02 3.80459875e-02 -1.69613451e-01
6.16970122e-01 -2.61401832e-02 6.70772731e-01 4.79899675e-01
2.68604696e-01 4.65522170e-01 4.41445559e-01 -2.09431723e-01
-1.12066507e+00 -4.31342006e-01 -2.68594116e-01 6.33964777e-01
-1.55907616e-01 -5.41422963e-01 -5.83803773e-01 -3.92594375e-02
-1.42989784e-01 7.13716149e-02 -3.56268615e-01 2.02116430e-01
-5.56973219e-01 -8.93400848e-01 3.96291316e-01 -1.07956417e-01
6.25651062e-01 -1.24556951e-01 6.34747446e-02 3.39405894e-01
-4.43772078e-01 -1.53148401e+00 -5.98617315e-01 -2.13875160e-01
-1.31327403e+00 -9.99896407e-01 -1.14112806e+00 -4.13401544e-01
9.84641910e-01 9.81568933e-01 6.00634694e-01 5.43225333e-02
-5.23965470e-02 2.06517190e-01 -4.75998700e-01 4.14548844e-01
2.98327971e-02 -2.50838637e-01 3.93938363e-01 7.79525816e-01
-5.86800516e-01 -6.13565505e-01 -5.30120194e-01 3.93757492e-01
-1.43375921e+00 1.98843658e-01 7.21177459e-01 1.00157118e+00
7.39558876e-01 4.27798539e-01 -1.48905039e-01 -6.90225005e-01
6.19476438e-01 -3.43215674e-01 -8.24743927e-01 1.32471830e-01
-4.58485574e-01 3.35136861e-01 6.91791594e-01 -5.16886711e-01
-7.04689264e-01 2.43830696e-01 3.35260600e-01 -8.13601196e-01
5.65027773e-01 1.12557125e+00 6.10974766e-02 -6.67540312e-01
3.28842402e-01 3.26677561e-01 4.46154773e-02 -7.63498247e-01
4.92185950e-01 3.14346015e-01 4.08559769e-01 -8.97169709e-01
1.42332661e+00 6.11143410e-01 5.70755422e-01 -1.24025822e+00
-1.03708279e+00 -6.79416418e-01 -4.06929225e-01 -2.85176635e-01
4.98413056e-01 -8.18181932e-01 -9.29455757e-01 4.98222798e-01
-1.21067858e+00 1.17483996e-01 5.95695674e-02 8.29926491e-01
-2.86016077e-01 9.37565148e-01 -6.45438910e-01 -8.53427947e-01
-2.31161863e-01 -1.07087541e+00 6.72772288e-01 -1.59556851e-01
4.32072788e-01 -7.89678395e-01 1.22416079e-01 5.24524391e-01
2.99923629e-01 4.35399592e-01 6.02129698e-01 5.70739098e-02
-8.02789330e-01 -3.18278223e-01 -4.56577092e-01 6.75729096e-01
-9.49524492e-02 -5.93028702e-02 -5.56348920e-01 -4.27574158e-01
2.60852158e-01 -2.78941356e-02 6.32746935e-01 2.44098917e-01
1.13381004e+00 -8.90433729e-01 4.12497036e-02 9.32989061e-01
1.53530025e+00 -2.27784425e-01 4.66322362e-01 -2.17860788e-02
1.19337773e+00 5.17014861e-01 5.15159369e-01 6.06753111e-01
1.81727529e-01 6.50014043e-01 4.46134418e-01 2.61376705e-02
3.75249267e-01 2.47940496e-02 4.37706590e-01 1.46556604e+00
-4.60228413e-01 1.45206913e-01 -6.84395850e-01 3.92752171e-01
-1.95506775e+00 -7.62213349e-01 -6.80990398e-01 2.47830868e+00
6.78629339e-01 -4.97292280e-01 -8.44222680e-02 3.56017530e-01
6.60828292e-01 3.53257120e-01 -4.54994291e-01 -2.53497902e-02
-3.86185914e-01 8.98822695e-02 9.64240432e-01 5.86140275e-01
-1.04369652e+00 6.16480947e-01 5.69132853e+00 8.82807851e-01
-9.42482889e-01 2.91000575e-01 1.98997483e-01 3.40233088e-01
-1.28108323e-01 4.33150977e-01 -3.89373839e-01 4.72534709e-02
4.64400798e-01 6.30680695e-02 1.05377185e+00 4.81179059e-01
4.40978169e-01 -4.01618183e-02 -6.59082055e-01 1.26521325e+00
1.55405939e-01 -1.16454744e+00 -1.00901291e-01 3.09230745e-01
9.61767018e-01 -2.27442503e-01 3.94283608e-02 -1.16493881e-01
2.13393159e-02 -7.05476820e-01 5.12402892e-01 7.22874582e-01
6.85635567e-01 -7.01731920e-01 4.70979065e-01 1.44936159e-01
-1.35295260e+00 5.52655086e-02 -4.91311997e-01 -1.06716482e-02
2.42011994e-01 1.30221891e+00 -4.19321001e-01 1.10592842e+00
3.93462092e-01 9.66807783e-01 -2.08068132e-01 1.17213643e+00
-8.62502381e-02 6.81356668e-01 -5.30135036e-01 1.89192921e-01
4.08082694e-01 -9.55797851e-01 8.11587393e-01 7.34861255e-01
6.14053786e-01 2.95692384e-01 3.83986145e-01 5.90035021e-01
-1.31753541e-03 3.16669792e-01 -4.59720343e-01 -3.82469594e-01
-2.29591429e-02 1.59247065e+00 -4.76022363e-01 -2.20192000e-01
-2.70556420e-01 1.04089653e+00 -1.46428617e-02 7.08117962e-01
-5.16938508e-01 -3.61348808e-01 6.37393653e-01 -1.71627626e-01
2.01208487e-01 -7.39227772e-01 1.27740175e-01 -1.62057412e+00
2.72226125e-01 -1.02375102e+00 1.42871961e-01 -8.17903399e-01
-1.19926298e+00 5.06484032e-01 -3.66094038e-02 -1.52935827e+00
1.76639244e-01 -8.25780988e-01 -2.32579440e-01 7.88604021e-01
-1.31022227e+00 -9.47899401e-01 -2.19958037e-01 7.78774977e-01
4.59912419e-02 1.43434480e-01 5.36222756e-01 5.83047330e-01
-1.02240789e+00 1.28055468e-01 5.95297933e-01 8.69068429e-02
3.53653759e-01 -1.01400077e+00 -3.49949270e-01 1.32345140e+00
-2.54021972e-01 6.54156268e-01 8.36178601e-01 -4.93577987e-01
-2.23941231e+00 -1.04374599e+00 7.68926442e-01 3.67271841e-01
1.07944548e+00 -4.09559309e-02 -7.42580593e-01 4.84343112e-01
4.88169566e-02 2.89351493e-01 4.66439515e-01 -9.00091082e-02
-4.48785096e-01 -4.94588494e-01 -8.31639946e-01 4.72196728e-01
6.11829221e-01 -6.00539505e-01 -2.22931840e-02 9.95608807e-01
6.24895036e-01 -4.83233035e-01 -1.23047173e+00 7.97924995e-02
4.27671075e-01 -5.16938865e-01 1.01320863e+00 -5.35892010e-01
4.02257293e-01 -8.64571393e-01 -4.40614998e-01 -1.23326325e+00
-3.28377455e-01 -1.24149609e+00 -2.92215496e-01 9.22704756e-01
1.48696125e-01 -6.44577265e-01 5.31015635e-01 3.92881155e-01
-2.11706325e-01 -5.75813174e-01 -1.02671075e+00 -9.35537398e-01
-4.02146161e-01 -4.62180197e-01 2.79201061e-01 1.11445582e+00
-1.56975612e-01 3.10064942e-01 -1.03210366e+00 4.94424194e-01
1.08714819e+00 7.19436705e-02 6.34516418e-01 -1.00842166e+00
-4.17223841e-01 -1.04129791e-01 -1.98800609e-01 -1.25731754e+00
6.82238266e-02 -1.06940269e+00 2.12514885e-02 -1.38589346e+00
6.30265623e-02 -5.43345213e-01 4.58600186e-02 2.98570842e-01
4.14480530e-02 2.28844300e-01 2.98073977e-01 4.78339553e-01
-3.01416695e-01 7.18596399e-01 1.37464809e+00 -2.20881313e-01
6.64423266e-03 -9.57462639e-02 -4.98642623e-01 5.76053381e-01
5.76115310e-01 -6.07159138e-01 -1.56671822e-01 -4.17635262e-01
6.77601874e-01 4.93213266e-01 3.01872164e-01 -9.10754263e-01
3.45636398e-01 -2.28328809e-01 -2.34974489e-01 -6.11640871e-01
5.90006948e-01 -6.77113295e-01 3.69070023e-01 3.31253439e-01
-9.42161009e-02 1.53071478e-01 -2.49686554e-01 6.37139559e-01
-2.38351613e-01 -5.97663820e-01 5.05504310e-01 1.20741636e-01
-5.50909862e-02 7.07005262e-01 -4.36007619e-01 -2.03015625e-01
3.14051479e-01 1.85994163e-01 -1.36185840e-01 -3.65767777e-01
-7.07576811e-01 -8.85072574e-02 3.74910720e-02 -1.63148686e-01
7.56219387e-01 -1.50893235e+00 -9.25312638e-01 -2.36429170e-01
7.46724848e-03 1.58802569e-01 3.96368235e-01 1.45100892e+00
-6.72573149e-01 3.99348676e-01 3.78209680e-01 -6.70496643e-01
-9.10509527e-01 4.24660236e-01 1.03139199e-01 -4.14037287e-01
-4.31042433e-01 5.61671734e-01 -1.62442699e-01 -3.27262878e-01
-1.73820797e-02 -3.63155365e-01 4.98802625e-02 -1.43895075e-02
5.63024580e-01 7.34532654e-01 1.35484233e-01 -9.11700130e-01
-2.33821906e-02 8.83763194e-01 1.12062201e-01 -1.83076724e-01
1.44190395e+00 -6.42242357e-02 -9.38637972e-01 2.45102718e-01
1.40184689e+00 1.17073826e-01 -8.82151008e-01 -5.59161067e-01
-1.72078654e-01 -6.28605783e-01 5.40129542e-01 -2.72148177e-02
-1.34952760e+00 6.00672841e-01 4.00679946e-01 3.03144157e-01
1.21174014e+00 -6.30125284e-01 8.02928329e-01 8.33624482e-01
2.88923234e-01 -8.97964239e-01 -7.15722665e-02 6.03202999e-01
1.09524739e+00 -8.73855293e-01 5.60662329e-01 -6.68132126e-01
-2.41215318e-01 1.22101736e+00 -1.07963525e-01 -2.70087540e-01
6.35681927e-01 -2.38878116e-01 -4.79029596e-01 1.07457899e-01
-4.32311773e-01 -1.92074820e-01 4.39847589e-01 1.54653579e-01
2.72380024e-01 2.10372284e-01 -4.84930158e-01 -7.55799040e-02
-1.95914507e-02 2.06000600e-02 7.73114324e-01 7.91578710e-01
-1.43462896e-01 -1.15592515e+00 -1.01939631e+00 4.98246074e-01
-4.40560549e-01 -1.95369825e-01 8.76483768e-02 3.69594246e-01
-1.15493111e-01 1.15141463e+00 -5.91516197e-01 -3.89031500e-01
1.44809231e-01 -2.75967717e-01 4.28305745e-01 -1.98832914e-01
-1.62925571e-01 3.10317546e-01 1.78541407e-01 -5.49996316e-01
-6.95916295e-01 -7.49194205e-01 -8.07756066e-01 -5.24416387e-01
-5.12865365e-01 2.89629430e-01 8.95950079e-01 9.25786376e-01
1.10103630e-01 2.56431755e-02 1.01368868e+00 -9.72379267e-01
-4.90884244e-01 -9.34713125e-01 -8.88744295e-01 6.40817210e-02
4.34923589e-01 -5.87753832e-01 -6.55369103e-01 -1.94489751e-02] | [7.3836846351623535, 4.460415363311768] |
8f3a8d60-bfcd-4bf9-8dca-2f3b80c28b95 | magnification-independent-histopathological | 2107.01063 | null | https://arxiv.org/abs/2107.01063v2 | https://arxiv.org/pdf/2107.01063v2.pdf | Magnification-independent Histopathological Image Classification with Similarity-based Multi-scale Embeddings | The classification of histopathological images is of great value in both cancer diagnosis and pathological studies. However, multiple reasons, such as variations caused by magnification factors and class imbalance, make it a challenging task where conventional methods that learn from image-label datasets perform unsatisfactorily in many cases. We observe that tumours of the same class often share common morphological patterns. To exploit this fact, we propose an approach that learns similarity-based multi-scale embeddings (SMSE) for magnification-independent histopathological image classification. In particular, a pair loss and a triplet loss are leveraged to learn similarity-based embeddings from image pairs or image triplets. The learned embeddings provide accurate measurements of similarities between images, which are regarded as a more effective form of representation for histopathological morphology than normal image features. Furthermore, in order to ensure the generated models are magnification-independent, images acquired at different magnification factors are simultaneously fed to networks during training for learning multi-scale embeddings. In addition to the SMSE, to eliminate the impact of class imbalance, instead of using the hard sample mining strategy that intuitively discards some easy samples, we introduce a new reinforced focal loss to simultaneously punish hard misclassified samples while suppressing easy well-classified samples. Experimental results show that the SMSE improves the performance for histopathological image classification tasks for both breast and liver cancers by a large margin compared to previous methods. In particular, the SMSE achieves the best performance on the BreakHis benchmark with an improvement ranging from 5% to 18% compared to previous methods using traditional features. | ['Qianni Zhang', 'Huiyu Zhou', 'Yaqi Wang', 'Xingru Huang', 'Yibao Sun'] | 2021-07-02 | null | null | null | null | ['histopathological-image-classification'] | ['medical'] | [ 3.32182080e-01 2.87721828e-02 -2.84360588e-01 -4.79978651e-01
-6.68978155e-01 -2.41589949e-01 4.70064312e-01 7.97812343e-01
-7.32283235e-01 3.56672674e-01 -9.19837505e-02 -1.01836696e-01
-4.40839201e-01 -6.67707086e-01 -6.28116667e-01 -1.20962930e+00
-8.08296949e-02 2.30790496e-01 -5.33520579e-02 -7.48771206e-02
1.55300915e-01 6.82383478e-01 -1.33328223e+00 2.28284404e-01
7.41654873e-01 1.07459009e+00 2.48041853e-01 3.97260189e-01
-7.50891864e-02 5.39572954e-01 -3.31880540e-01 -3.52545649e-01
3.51363122e-01 -7.80499950e-02 -6.60333514e-01 2.97428459e-01
3.56850505e-01 -1.71547681e-01 -3.19541603e-01 1.18267262e+00
4.97731954e-01 -2.16749191e-01 8.71676624e-01 -1.12614691e+00
-5.33596218e-01 2.02389866e-01 -8.59030247e-01 3.06341588e-01
-1.80765003e-01 1.95789516e-01 1.03931403e+00 -6.87545300e-01
4.36606348e-01 8.72181475e-01 6.39740109e-01 4.67060804e-01
-1.39536583e+00 -6.57185495e-01 -7.22510889e-02 3.47827137e-01
-1.35073698e+00 -2.07622290e-01 7.91985989e-01 -2.79026836e-01
3.20831418e-01 5.05217552e-01 6.38513327e-01 7.42324948e-01
5.53745449e-01 7.56356716e-01 1.10300505e+00 -4.58068788e-01
1.03010321e-02 5.29450476e-01 3.54561955e-04 7.44915724e-01
3.26550961e-01 -1.08821832e-01 -1.30706385e-01 -9.94597599e-02
6.57517195e-01 5.69903135e-01 -5.29469132e-01 -5.27091146e-01
-1.34307671e+00 8.58493805e-01 7.97423363e-01 4.36786771e-01
-1.94403812e-01 -8.90003219e-02 5.01600683e-01 4.06568348e-01
3.87399882e-01 5.90326786e-01 -2.52915561e-01 4.59460765e-01
-5.82025111e-01 -4.39091139e-02 2.37716094e-01 2.21491665e-01
7.62118876e-01 -3.78206402e-01 -1.03504933e-01 1.10068095e+00
1.00091197e-01 1.78480953e-01 8.68903100e-01 -4.55868840e-01
1.45045400e-01 9.71211672e-01 -4.06335473e-01 -1.24193561e+00
-5.39971292e-01 -7.52390862e-01 -1.37531710e+00 3.58565211e-01
5.37169397e-01 5.10068476e-01 -7.76520133e-01 1.64713705e+00
4.28102523e-01 1.16476335e-01 -2.11218987e-02 8.06441009e-01
5.83065450e-01 2.16554612e-01 -9.64822061e-03 -2.07895443e-01
1.64235544e+00 -7.28609204e-01 -5.27735889e-01 -5.55465370e-02
8.54278862e-01 -5.56814909e-01 1.07519841e+00 1.53485045e-01
-6.80948317e-01 -3.11550677e-01 -1.14468670e+00 -8.67235195e-03
-3.76233131e-01 1.88452706e-01 6.11720383e-01 3.22683066e-01
-7.90201664e-01 6.01256669e-01 -8.28541934e-01 -1.26096785e-01
6.80963635e-01 4.49324101e-01 -7.77629614e-01 -2.24573448e-01
-8.56988072e-01 6.61165655e-01 3.11840594e-01 8.21199920e-03
-4.50548261e-01 -1.05295515e+00 -8.57990205e-01 1.12724423e-01
8.74131918e-02 -4.18121278e-01 7.49050796e-01 -1.03113627e+00
-9.64011192e-01 1.09153700e+00 4.32816520e-02 -3.63459289e-01
4.96150494e-01 4.20058876e-01 -3.39276463e-01 4.45834905e-01
3.70437205e-02 7.91076005e-01 7.53181994e-01 -1.21559703e+00
-6.37724102e-01 -4.30405110e-01 -1.52811101e-02 1.25458106e-01
-9.58404243e-01 -4.01141524e-01 -1.87865108e-01 -7.30016649e-01
8.17489401e-02 -8.67864132e-01 -3.54309112e-01 5.98504364e-01
-2.88082927e-01 -1.45842463e-01 6.98765874e-01 -3.05967838e-01
8.99050653e-01 -2.35879850e+00 4.04197052e-02 2.35939339e-01
5.39435267e-01 1.78915158e-01 -3.67035955e-01 3.64382192e-02
-3.21331769e-01 2.01249272e-01 -1.94768280e-01 -3.73367190e-01
-2.50724107e-01 2.28066593e-01 2.01050580e-01 7.98089862e-01
5.59000373e-01 8.52247953e-01 -9.77390289e-01 -6.72647953e-01
2.56390840e-01 3.52669269e-01 -3.35640967e-01 1.38951495e-01
2.62409627e-01 3.90222937e-01 -2.42963403e-01 5.07918954e-01
7.00261056e-01 -5.81638336e-01 2.66841173e-01 -4.16283250e-01
2.44851992e-01 -2.15663731e-01 -8.64476383e-01 1.38618422e+00
-6.13856018e-01 4.56644177e-01 -1.04943663e-01 -1.28334868e+00
5.02420366e-01 2.93856978e-01 6.34795368e-01 -8.09648812e-01
2.23214239e-01 2.67477006e-01 2.04020113e-01 -5.50004780e-01
9.01296083e-03 -3.58326495e-01 5.37992045e-02 3.87705445e-01
4.85576764e-02 2.33656690e-02 8.73611569e-02 9.72059220e-02
1.14019692e+00 -6.80888295e-01 4.23462689e-01 -3.57426107e-01
7.43886948e-01 -2.40821272e-01 7.98912704e-01 3.64844620e-01
-3.35035712e-01 5.99175155e-01 5.74020505e-01 -5.98403037e-01
-9.94528949e-01 -9.37793076e-01 -5.53651869e-01 6.28057480e-01
2.97685534e-01 -4.77079535e-03 -2.76980817e-01 -1.00238121e+00
1.62906736e-01 4.61161025e-02 -9.19272542e-01 -4.06993985e-01
-4.66298997e-01 -1.19634950e+00 3.16209912e-01 5.13206303e-01
2.25758225e-01 -6.99655473e-01 -3.23793024e-01 -3.58059630e-02
-4.59974185e-02 -9.76383507e-01 -4.95477557e-01 4.08218563e-01
-9.83386457e-01 -1.38813829e+00 -7.54343808e-01 -9.41458523e-01
1.30692446e+00 7.16112018e-01 7.45786548e-01 5.50580978e-01
-7.06942618e-01 -3.10866199e-02 -3.51786107e-01 -3.30702186e-01
-6.06777191e-01 -3.53074539e-03 -7.62452781e-02 3.55418384e-01
2.34977990e-01 -5.20814657e-01 -8.61106634e-01 3.03002745e-01
-1.46591175e+00 4.96780546e-03 1.07907331e+00 1.36521924e+00
6.37621045e-01 2.65497923e-01 6.50096476e-01 -8.35031331e-01
2.81594634e-01 -4.27838117e-01 -2.66616583e-01 2.34650001e-01
-6.66152000e-01 5.72639443e-02 9.34667647e-01 -5.23028374e-01
-5.32749295e-01 -5.40578179e-02 -6.75631762e-02 -4.59361404e-01
-1.09602466e-01 3.55180770e-01 9.18116346e-02 -3.89009565e-01
5.03723443e-01 1.91446841e-01 5.83073616e-01 -1.11247323e-01
-2.67801553e-01 5.93731523e-01 2.61825860e-01 -6.27533421e-02
8.87618899e-01 8.44035268e-01 2.50818223e-01 -6.51665509e-01
-6.80687487e-01 -7.64572799e-01 -3.60812187e-01 -3.02421525e-02
7.04169869e-01 -7.76256144e-01 -5.56596696e-01 4.59190637e-01
-6.39915109e-01 -2.71668565e-02 -2.44325683e-01 5.57632327e-01
-1.73949093e-01 5.11399984e-01 -7.83188462e-01 -3.03736031e-01
-2.84668714e-01 -1.32136118e+00 9.81617033e-01 3.03358287e-01
-1.21629033e-02 -1.17557073e+00 -1.95002630e-01 2.48690888e-01
2.90675789e-01 3.25494587e-01 1.46738756e+00 -7.55707741e-01
-3.42584044e-01 -4.22315687e-01 -4.28566307e-01 5.13000846e-01
6.66060090e-01 -6.52826354e-02 -8.52537870e-01 -6.02292299e-01
6.08321242e-02 -3.89154345e-01 9.16178823e-01 1.56502068e-01
1.38014960e+00 -1.79313570e-01 -5.00486255e-01 6.59261644e-01
1.62099016e+00 -1.18037924e-01 4.88940805e-01 6.27536118e-01
4.28513587e-01 6.83980465e-01 3.62119764e-01 2.56618589e-01
2.23264724e-01 5.20954728e-01 6.60465419e-01 -5.47447145e-01
-9.93029177e-02 1.56246156e-01 -1.39870271e-01 7.43077457e-01
2.79984146e-01 -3.58424634e-02 -6.25920117e-01 7.94052839e-01
-1.42599499e+00 -8.01598132e-01 -3.11073512e-02 2.37976933e+00
1.02000260e+00 8.71755406e-02 -2.64976233e-01 5.24946570e-01
6.79118931e-01 1.75995946e-01 -4.59892988e-01 -2.05464195e-02
-1.42447114e-01 4.97115254e-02 4.49715793e-01 -6.12228923e-03
-1.10824358e+00 1.63853735e-01 5.08968925e+00 9.50982749e-01
-1.45304251e+00 4.32466157e-02 1.02328694e+00 -1.29035309e-01
-2.92133719e-01 -3.17536265e-01 -5.15390694e-01 5.12099862e-01
4.41891164e-01 4.90605794e-02 -7.65084922e-02 4.47385609e-01
2.74957329e-01 -1.12209149e-01 -1.11129498e+00 8.84739876e-01
6.18863441e-02 -1.33565617e+00 8.91153961e-02 3.69979918e-01
6.94257617e-01 -1.82085946e-01 2.73509264e-01 7.12073818e-02
-1.67919144e-01 -1.03785563e+00 3.15455854e-01 1.49119943e-01
8.69025171e-01 -7.68340766e-01 1.14391041e+00 2.03499943e-01
-8.58117044e-01 -1.90424576e-01 -4.62857664e-01 3.79193485e-01
-1.69783160e-01 1.00015664e+00 -9.40617085e-01 6.69991970e-01
5.09260416e-01 7.20842779e-01 -7.73224533e-01 1.08983493e+00
-2.06326339e-02 2.72693664e-01 -8.25800002e-02 1.45481214e-01
2.56240636e-01 8.38698298e-02 3.55061740e-01 1.05585039e+00
1.71434835e-01 -3.34982932e-01 5.25288703e-03 4.41214383e-01
-2.45355532e-01 2.24285141e-01 -3.79548788e-01 2.05874026e-01
2.62595743e-01 1.75162745e+00 -8.89570534e-01 -1.56387210e-01
-2.65807480e-01 8.32609415e-01 3.13582629e-01 -1.01922452e-02
-7.48486221e-01 -2.63056099e-01 6.34758711e-01 2.22388804e-01
2.29489803e-01 1.97850019e-01 -1.68321311e-01 -1.05314207e+00
1.98662981e-01 -8.70315611e-01 3.58591378e-01 -1.37745962e-01
-1.50960886e+00 5.58151364e-01 -4.15281981e-01 -1.59369564e+00
1.37058616e-01 -6.59042716e-01 -7.60534227e-01 3.59611779e-01
-2.11961579e+00 -1.05767369e+00 -4.02640015e-01 2.62900412e-01
4.09687966e-01 7.58939385e-02 7.95028806e-01 5.05622864e-01
-4.78638381e-01 9.36874032e-01 4.19472694e-01 2.59165227e-01
8.58554065e-01 -1.35380578e+00 -2.62134701e-01 4.06834304e-01
4.29778472e-02 4.81434822e-01 4.36324924e-01 -7.64260218e-02
-1.17299879e+00 -1.33206630e+00 6.43697143e-01 -3.22636068e-02
5.77143669e-01 -2.77611673e-01 -1.11542213e+00 2.18874797e-01
-7.99628198e-02 5.72453439e-01 1.05320323e+00 -2.93098092e-01
-4.38600153e-01 -4.84375149e-01 -1.23760784e+00 5.96921325e-01
5.94387591e-01 -3.76417518e-01 -2.89767951e-01 4.39773649e-01
2.71142811e-01 -1.61495104e-01 -1.09278822e+00 5.38523436e-01
5.85271060e-01 -8.71321142e-01 9.43687856e-01 -2.92233944e-01
6.15965962e-01 -3.36425453e-01 4.68726493e-02 -1.52284491e+00
-4.21889812e-01 1.19315885e-01 5.19822598e-01 9.17348444e-01
3.06972414e-01 -8.15635979e-01 7.34926045e-01 6.84970841e-02
-2.96839997e-02 -1.17551219e+00 -1.17683566e+00 -8.81748199e-01
1.15649141e-01 3.92127186e-02 4.31987584e-01 1.11467135e+00
-1.55397877e-01 -5.59103377e-02 7.29959458e-02 1.69702724e-01
6.25643134e-01 1.37422457e-01 5.11335254e-01 -1.18212712e+00
-1.78363249e-01 -6.84265733e-01 -9.21477616e-01 -5.71666956e-01
1.66421667e-01 -1.16803920e+00 -1.77120138e-02 -1.19533253e+00
6.12267554e-01 -7.66866446e-01 -8.09615493e-01 3.69654447e-01
-3.53266865e-01 7.45039463e-01 7.44321500e-04 3.52375239e-01
-4.16181117e-01 4.91363674e-01 1.36588037e+00 -5.85027397e-01
1.69296190e-01 -1.04868695e-01 -8.36902142e-01 5.90758502e-01
6.44163966e-01 -6.39644563e-01 -2.53577799e-01 -2.95681804e-01
1.16262019e-01 -2.27415532e-01 6.60579622e-01 -1.02930641e+00
1.27166376e-01 -8.80885571e-02 5.72821259e-01 -1.12525322e-01
1.44121706e-01 -8.23037922e-01 -8.48108232e-02 7.37472713e-01
-3.46005291e-01 -1.19464718e-01 5.16279489e-02 7.27985144e-01
-5.31503022e-01 -2.99499720e-01 9.32700574e-01 1.63548738e-02
-3.20233434e-01 3.92348915e-01 -2.32878447e-01 -3.45708549e-01
1.14236617e+00 -3.50316495e-01 -2.56312728e-01 1.77551564e-02
-2.25350335e-01 1.95956007e-01 5.70287049e-01 4.06222343e-01
6.38573289e-01 -1.35711241e+00 -7.13036001e-01 2.97519863e-01
5.16978860e-01 2.18584195e-01 4.92967099e-01 1.34469867e+00
-3.90040547e-01 2.63559878e-01 -2.59109586e-01 -8.06871593e-01
-1.33271980e+00 3.68050933e-01 3.50984097e-01 -7.34607160e-01
-5.11862397e-01 7.36438036e-01 6.00057662e-01 -4.21989948e-01
2.64338162e-02 -3.68414462e-01 -2.41320580e-01 1.01674162e-01
6.38722062e-01 -1.78540442e-02 5.01742840e-01 -4.19919670e-01
-2.86151797e-01 5.00357449e-01 -6.33208454e-01 5.04858315e-01
1.48223209e+00 6.71389922e-02 -1.81365490e-01 3.38934064e-01
1.67793989e+00 1.95760410e-02 -1.02722132e+00 -2.85113066e-01
-1.33024797e-01 -5.22575200e-01 1.58210248e-01 -5.47230124e-01
-1.38985085e+00 8.15893590e-01 7.86897302e-01 2.35165283e-01
1.30526865e+00 -1.00532323e-01 7.99654543e-01 6.78735599e-02
7.59960636e-02 -7.69577384e-01 3.34747672e-01 -7.24292770e-02
6.63568616e-01 -1.49571788e+00 8.16846639e-02 -3.10212195e-01
-3.65589917e-01 1.24463427e+00 4.34560865e-01 -1.17800362e-01
3.27404410e-01 1.58520803e-01 1.68631211e-01 -1.66890457e-01
-7.58609712e-01 2.01598078e-01 1.09450437e-01 3.46949369e-01
2.99442381e-01 -5.48704229e-02 -3.64780873e-01 3.45307142e-01
2.61223644e-01 -7.64566213e-02 5.79189122e-01 8.22742403e-01
-2.85239816e-01 -1.21274710e+00 -2.13999391e-01 8.97971034e-01
-7.08814621e-01 1.49708658e-01 5.91867343e-02 8.32285345e-01
7.62545913e-02 6.22188389e-01 2.29769841e-01 -4.03100669e-01
8.63916576e-02 -3.55709881e-01 3.52151603e-01 -4.89929944e-01
-6.76949143e-01 -1.92354858e-01 -4.52120602e-01 -2.45266020e-01
-3.34688425e-01 -3.21781218e-01 -1.21628964e+00 -1.30851060e-01
-6.00811183e-01 9.83516723e-02 5.00490367e-01 8.39873970e-01
2.66051948e-01 6.27119780e-01 1.14503276e+00 -5.80789089e-01
-8.47955465e-01 -7.44645536e-01 -6.53509855e-01 7.62032509e-01
5.62421083e-01 -6.74779177e-01 -7.30440080e-01 -1.10495865e-01] | [15.03271198272705, -2.683673620223999] |
c3a31e87-9458-4698-b752-0eabd7f41e65 | from-visual-to-acoustic-question-answering | 1902.11280 | null | http://arxiv.org/abs/1902.11280v1 | http://arxiv.org/pdf/1902.11280v1.pdf | From Visual to Acoustic Question Answering | We introduce the new task of Acoustic Question Answering (AQA) to promote
research in acoustic reasoning. The AQA task consists of analyzing an acoustic
scene composed by a combination of elementary sounds and answering questions
that relate the position and properties of these sounds. The kind of relational
questions asked, require that the models perform non-trivial reasoning in order
to answer correctly. Although similar problems have been extensively studied in
the domain of visual reasoning, we are not aware of any previous studies
addressing the problem in the acoustic domain. We propose a method for
generating the acoustic scenes from elementary sounds and a number of relevant
questions for each scene using templates. We also present preliminary results
obtained with two models (FiLM and MAC) that have been shown to work for visual
reasoning. | ['Jerome Abdelnour', 'Jean Rouat', 'Giampiero Salvi'] | 2019-02-28 | null | null | null | null | ['acoustic-question-answering'] | ['speech'] | [ 4.74350631e-01 2.03165457e-01 1.20130742e+00 -4.43269670e-01
-1.00438106e+00 -6.06860399e-01 5.73677838e-01 2.19622836e-01
-1.47418067e-01 8.08495879e-02 1.92928582e-01 -5.44467330e-01
-4.50338930e-01 -9.49788749e-01 -5.88127077e-01 -2.82196671e-01
2.96168290e-02 5.77057898e-01 9.56235766e-01 -5.30484438e-01
3.35336745e-01 2.32614815e-01 -2.03930974e+00 6.30527139e-01
2.12595969e-01 8.56376052e-01 6.82063758e-01 1.48189044e+00
-6.66076720e-01 1.21113586e+00 -1.03972018e+00 -4.75878656e-01
-2.48147994e-01 -9.79939520e-01 -1.44141591e+00 -1.86338618e-01
6.73374593e-01 4.66304757e-02 7.06397668e-02 1.06395805e+00
4.51348871e-01 4.57232922e-01 7.18861938e-01 -1.07466626e+00
-6.53423488e-01 5.67940414e-01 3.39529186e-01 4.74482387e-01
1.01881123e+00 -2.28432938e-01 1.37580109e+00 -6.96856081e-01
2.61254698e-01 1.45483828e+00 6.69875681e-01 7.23466396e-01
-1.01596165e+00 -2.96707243e-01 -1.05781607e-01 6.22287273e-01
-1.17935109e+00 -3.05429667e-01 9.47438717e-01 -2.96556801e-01
9.56111491e-01 8.63543868e-01 6.63752079e-01 7.20212817e-01
-6.44772202e-02 2.07098082e-01 1.11767876e+00 -7.27107525e-01
4.01766479e-01 2.57459074e-01 3.03529531e-01 6.64675236e-01
-2.93012679e-01 -3.43998760e-01 -5.53716481e-01 -9.80459005e-02
5.59751391e-01 -9.84852850e-01 9.14865434e-02 -4.12652455e-02
-1.04151726e+00 6.86436355e-01 -3.83560918e-02 6.95959389e-01
-2.52272457e-01 4.54495639e-01 1.78147018e-01 3.36730152e-01
-1.57903314e-01 7.88238823e-01 -1.17397405e-01 -6.11372069e-02
-2.77714938e-01 6.56323314e-01 1.18589735e+00 7.50844598e-01
5.49123764e-01 1.75684214e-01 -3.59147936e-02 9.21256900e-01
6.51447594e-01 7.64505208e-01 1.31995931e-01 -1.43552125e+00
2.01069444e-01 2.93378141e-02 -1.67849082e-02 -1.41126847e+00
-4.72255111e-01 1.55784354e-01 9.24606398e-02 6.12105429e-02
8.08205128e-01 3.24474335e-01 -3.01403940e-01 1.68495679e+00
3.92085403e-01 1.36037692e-01 2.52795011e-01 8.09148788e-01
1.53903723e+00 9.68357861e-01 1.23167865e-01 -1.38841450e-01
1.94005871e+00 -4.42068011e-01 -9.04295444e-01 -7.57077485e-02
1.95050523e-01 -1.23509157e+00 1.59216022e+00 4.75729376e-01
-1.35195875e+00 -9.87791479e-01 -8.43965650e-01 -3.51529270e-01
-3.20661396e-01 -3.26123834e-01 2.83258557e-01 7.51926601e-01
-1.10864103e+00 -1.48304245e-02 -2.96852171e-01 -4.72721487e-01
-2.98961937e-01 1.01007670e-01 1.77791789e-01 3.42558503e-01
-1.16904771e+00 9.79208291e-01 -1.82289645e-01 1.17142297e-01
-9.34210002e-01 -3.53457749e-01 -7.65684903e-01 7.83423185e-02
5.27713895e-01 -6.68861568e-01 1.91540611e+00 -5.31286895e-01
-1.55969656e+00 8.81852806e-01 -1.79285541e-01 -1.51727378e-01
-2.33207256e-01 -8.08266997e-02 -6.39631152e-01 5.61707854e-01
-6.69432431e-02 4.97565508e-01 6.19809091e-01 -1.51402259e+00
-4.46057975e-01 -2.38298669e-01 6.78978145e-01 1.29349157e-01
1.38869947e-02 3.24937075e-01 -3.04247648e-01 -6.25659823e-01
3.08097631e-01 -6.06426835e-01 -4.65650111e-02 -1.83532506e-01
-2.81216800e-01 -4.57507282e-01 7.99900830e-01 -7.23822236e-01
1.12985194e+00 -2.10342336e+00 -8.67201313e-02 3.23387235e-01
-8.39819163e-02 -7.76651651e-02 -3.94452006e-01 4.84293193e-01
-4.59370539e-02 2.46422943e-02 -1.86538145e-01 -5.09053022e-02
4.54737663e-01 4.42278296e-01 -6.78238034e-01 -2.59378612e-01
4.29265872e-02 6.59741819e-01 -9.47677314e-01 -7.90229261e-01
2.26814225e-01 3.25302690e-01 -6.93479419e-01 6.91290379e-01
-6.20578766e-01 2.39267752e-01 -5.11705041e-01 1.21353239e-01
2.22866908e-01 2.18459785e-01 1.68110095e-02 -2.10399225e-01
1.13594234e-01 8.03162634e-01 -1.36879170e+00 1.29081798e+00
-6.11558139e-01 8.17515016e-01 1.27351239e-01 -9.66162860e-01
1.11590624e+00 4.08668160e-01 7.07119778e-02 -8.42076361e-01
-6.44373242e-04 2.49462072e-02 1.37778714e-01 -1.29207730e+00
6.26524746e-01 -7.12032199e-01 -3.26865435e-01 4.76180226e-01
5.13584306e-03 -1.30274034e+00 1.46645769e-01 1.70915768e-01
9.72923160e-01 1.69841722e-01 1.52562648e-01 -2.21739039e-01
8.57555568e-01 4.13567647e-02 -9.75327641e-02 9.35530484e-01
-1.16018474e-01 6.11825168e-01 2.89419413e-01 -1.88490331e-01
-8.76005292e-01 -1.29034126e+00 1.13053836e-01 1.38459706e+00
3.08044821e-01 -6.39275074e-01 -1.10336626e+00 -2.42271926e-02
-5.35459399e-01 1.12923908e+00 -3.17835450e-01 2.27093577e-01
-5.90790808e-01 -2.98169017e-01 5.70241511e-01 3.13335806e-01
-3.85898650e-02 -1.64028776e+00 -8.31939995e-01 1.80403203e-01
-4.17938918e-01 -1.38793600e+00 -6.03285842e-02 -4.08309791e-03
-3.96345496e-01 -1.11572814e+00 -1.34159297e-01 -1.03111827e+00
2.04581931e-01 9.55780521e-02 1.25768375e+00 1.63366273e-01
-3.58554572e-01 1.23593891e+00 -4.36704129e-01 -6.90249383e-01
-9.16804075e-01 -5.45861065e-01 -3.47784340e-01 -6.54021725e-02
2.16863155e-01 -5.40496349e-01 -3.71533372e-02 3.15955013e-01
-1.08864439e+00 -1.21665642e-01 -7.97502175e-02 3.14020477e-02
4.84223604e-01 1.61205664e-01 4.42590624e-01 -6.77387238e-01
9.14878964e-01 -9.87265706e-02 -2.17768595e-01 3.16553295e-01
4.27371264e-01 1.37671962e-01 6.86447859e-01 -2.63395309e-01
-1.07971299e+00 -2.22110793e-01 -7.81434000e-01 2.38565341e-01
-6.97532058e-01 1.95917398e-01 -1.41443163e-01 -2.80930251e-02
8.22771847e-01 9.02541652e-02 -3.28663379e-01 -2.18396142e-01
4.57020015e-01 3.87914419e-01 9.18751776e-01 -8.83799314e-01
8.61236393e-01 3.08974415e-01 2.58583128e-01 -1.28651845e+00
-7.87028790e-01 -3.79996508e-01 -3.99030328e-01 -6.26856387e-01
1.31666660e+00 -1.46357268e-01 -1.00907743e+00 5.91405435e-03
-1.42906475e+00 -1.12267092e-01 -6.38574600e-01 4.95096356e-01
-1.03552246e+00 6.64080262e-01 -2.99550205e-01 -1.28260982e+00
2.46902421e-01 -1.11500454e+00 8.86221051e-01 1.26229808e-01
-6.59919322e-01 -8.24162006e-01 3.16517919e-01 5.92962146e-01
3.46101373e-01 -2.09964707e-01 1.42490637e+00 -6.14353120e-01
-6.88378036e-01 2.67068654e-01 1.77258447e-01 2.84259230e-01
-4.70728949e-02 2.92469952e-02 -1.43191302e+00 5.34313262e-01
3.91270041e-01 -5.04013777e-01 5.18925190e-01 2.37825274e-01
1.14636445e+00 1.25396371e-01 3.99331480e-01 -2.22308204e-01
1.08157194e+00 6.92740500e-01 7.24816620e-01 -9.67788696e-02
4.18345928e-01 1.16852665e+00 3.20545197e-01 1.18741184e-01
7.88376093e-01 8.06233406e-01 4.27096218e-01 2.79598624e-01
-4.87968206e-01 -2.36532509e-01 1.73047110e-01 1.24054253e+00
-2.73051262e-02 -1.05213940e-01 -9.70451236e-01 5.81223667e-01
-1.46565914e+00 -1.23857129e+00 -6.49037004e-01 1.76235521e+00
5.67036688e-01 -6.25252873e-02 1.58732459e-01 6.41167939e-01
3.41957003e-01 1.55837178e-01 1.21856913e-01 -1.01135719e+00
1.08698353e-01 7.49436557e-01 -7.02316463e-01 9.49013650e-01
-8.36015522e-01 7.37327933e-01 7.68768167e+00 6.08982205e-01
-2.89974213e-01 -5.63634373e-02 1.60085857e-01 4.64901030e-01
-9.79137242e-01 1.99464187e-01 -3.68584722e-01 -2.79736489e-01
1.14516890e+00 3.45491886e-01 5.42469025e-01 3.35448682e-01
2.67949384e-02 -5.51642418e-01 -1.09763956e+00 7.99705684e-01
3.88494253e-01 -8.74456823e-01 3.09301049e-01 -6.82425320e-01
3.78814177e-04 -8.81950796e-01 -7.46102035e-02 2.33974636e-01
2.42775247e-01 -8.03589344e-01 1.07019532e+00 9.86995280e-01
5.76663315e-02 -6.15119755e-01 3.12230170e-01 9.79464278e-02
-1.48102093e+00 1.17823808e-02 -4.28242236e-01 -2.39178911e-01
2.65629619e-01 2.22178906e-01 -8.32285404e-01 3.83623153e-01
1.04480040e+00 -3.56886268e-01 -7.41792500e-01 9.87186432e-01
-2.58255392e-01 7.37906933e-01 -3.05307209e-01 -5.56751788e-01
-4.84423935e-02 -2.43080631e-02 6.24879062e-01 1.08564115e+00
3.37877452e-01 4.48458582e-01 -1.43023804e-01 1.01158702e+00
5.55932701e-01 4.47662055e-01 -5.70727527e-01 -5.34462230e-03
1.61819950e-01 9.22666848e-01 -7.34351099e-01 -2.02069163e-01
-5.02008379e-01 5.31415284e-01 -1.50176391e-01 2.18118489e-01
-6.26285791e-01 -4.65011805e-01 1.68890908e-01 -6.76641688e-02
2.29279712e-01 -4.11238253e-01 -4.76728231e-02 -4.70934927e-01
-1.80705134e-02 -8.81852984e-01 5.05770147e-01 -1.54613435e+00
-1.26557076e+00 6.40987039e-01 3.03837597e-01 -7.19404638e-01
-4.10246491e-01 -6.76787853e-01 -6.62143588e-01 7.31565773e-01
-1.19787371e+00 -7.80538142e-01 -3.31636995e-01 6.71186626e-01
5.61069965e-01 5.25726564e-02 1.17228472e+00 -4.25782017e-02
1.89074397e-01 -8.52532014e-02 -8.23707461e-01 -1.49221942e-01
2.98877150e-01 -1.45553470e+00 3.53758216e-01 4.69143122e-01
7.60752439e-01 3.94458026e-01 1.21707606e+00 -2.03955069e-01
-1.39437401e+00 -2.75222540e-01 1.25233662e+00 -7.45782733e-01
7.82261372e-01 -5.28656423e-01 -1.13805127e+00 2.00195536e-01
2.89579570e-01 -3.36699098e-01 8.99253368e-01 -2.08272133e-02
-5.24275243e-01 -2.00138807e-01 -7.77920902e-01 4.48405325e-01
8.83951306e-01 -1.20671558e+00 -1.35200846e+00 2.21711934e-01
8.09996486e-01 -2.26248994e-01 -5.47682822e-01 2.21248880e-01
4.70536888e-01 -1.13728631e+00 1.18104231e+00 -5.15739381e-01
2.49424011e-01 -6.68566585e-01 -6.43715620e-01 -8.08815002e-01
-4.43022735e-02 -3.73789340e-01 4.04201597e-01 1.12558389e+00
4.27186608e-01 -2.43929565e-01 3.38058323e-01 2.50747383e-01
-2.35484123e-01 -1.93093255e-01 -9.64070380e-01 -5.26214361e-01
-1.79864526e-01 -1.36402333e+00 3.82584393e-01 4.57444102e-01
6.30642027e-02 6.56412065e-01 -2.17361048e-01 5.77412963e-01
2.12512165e-01 2.95986563e-01 7.69638062e-01 -1.24722528e+00
-7.49349594e-01 -3.45681489e-01 -3.97591442e-01 -8.18972528e-01
1.11554421e-01 -7.18831360e-01 4.81005818e-01 -1.73489583e+00
-8.81316662e-02 -6.73896894e-02 2.54721075e-01 1.00347616e-01
-1.19491540e-01 2.84998327e-01 5.04431248e-01 -1.98904127e-01
-5.76142609e-01 1.12624310e-01 1.33316183e+00 5.90276113e-03
1.32840380e-01 6.29434437e-02 -5.51134884e-01 8.90572727e-01
6.03114367e-01 -3.76361609e-01 -7.28313923e-01 -2.77395487e-01
6.81408763e-01 3.72657239e-01 7.30710626e-01 -1.22021794e+00
3.21775198e-01 -9.49368775e-02 -1.01625070e-01 -7.96643198e-01
6.93343759e-01 -8.33010733e-01 1.05345309e-01 2.09700316e-01
-6.48468614e-01 3.93835247e-01 2.50005633e-01 3.27981710e-01
-3.76106709e-01 -8.86926353e-01 6.09842420e-01 -1.41769096e-01
-9.15844977e-01 -4.66492444e-01 -8.37459207e-01 2.96119779e-01
7.57174432e-01 -8.18010271e-02 -5.38216121e-02 -9.37736094e-01
-1.31080568e+00 -1.90313578e-01 -4.01525557e-01 3.35280836e-01
8.93950820e-01 -1.20185101e+00 -5.23355007e-01 -4.48671319e-02
1.05016388e-01 -3.88392687e-01 4.10923034e-01 3.32523495e-01
-7.65190482e-01 3.64484638e-01 9.05223489e-02 -5.49230337e-01
-1.39685464e+00 5.82371533e-01 5.17315924e-01 2.49828190e-01
-4.13114101e-01 8.32981288e-01 2.99864024e-01 -4.61881995e-01
3.25527906e-01 -6.06074214e-01 -5.84186077e-01 -8.80608428e-03
6.28688753e-01 1.20422825e-01 -9.26303566e-02 -8.15637350e-01
-2.54566908e-01 9.28821146e-01 7.13680089e-01 -6.90617979e-01
9.04006064e-01 -9.49747786e-02 -2.10273445e-01 1.02109563e+00
9.31547642e-01 3.43357891e-01 -3.95785213e-01 -8.37307572e-02
4.69878502e-02 -1.39475942e-01 -4.46027458e-01 -5.54552794e-01
-3.63044590e-01 1.20298564e+00 4.66166437e-01 1.05221438e+00
1.16437757e+00 6.41910970e-01 5.52338123e-01 6.32996738e-01
2.72390068e-01 -1.07947576e+00 6.70672655e-01 7.15332925e-01
1.25719845e+00 -4.52647954e-01 -4.48810488e-01 -6.94580674e-01
-5.31805873e-01 1.19476426e+00 4.89119589e-01 4.43298230e-03
5.92653692e-01 4.32069629e-01 4.31126148e-01 -6.02650821e-01
-6.08671308e-01 -5.31293690e-01 4.72542703e-01 9.01746154e-01
2.55140543e-01 -3.89710851e-02 1.27643570e-02 5.35236955e-01
-9.12691593e-01 -5.64844191e-01 5.50852656e-01 7.83137321e-01
-9.33736920e-01 -9.15015161e-01 -8.96644592e-01 -6.20896183e-02
-3.32990438e-01 -7.18536526e-02 -7.35480428e-01 6.00784302e-01
1.29347995e-01 1.39769804e+00 6.78676069e-02 -4.11023080e-01
8.61218512e-01 3.58808309e-01 7.58706808e-01 -7.46447563e-01
-5.88350475e-01 -4.29069884e-02 5.31447470e-01 -3.93714458e-01
-7.18443990e-01 -6.53358757e-01 -1.60527909e+00 1.74276128e-01
-8.98360610e-02 3.96492571e-01 7.09668040e-01 1.08408070e+00
-4.43608224e-01 7.85337627e-01 3.90074223e-01 -3.34354579e-01
-2.48593509e-01 -6.95405424e-01 -5.49580038e-01 4.66780722e-01
2.88035035e-01 -2.02852234e-01 -3.21410328e-01 2.43480638e-01] | [15.296625137329102, 5.17949104309082] |
f4b29bb9-2f09-4248-8c35-23dc9ebc0462 | detection-of-rem-sleep-behaviour-disorder-by | 1811.04662 | null | http://arxiv.org/abs/1811.04662v1 | http://arxiv.org/pdf/1811.04662v1.pdf | Detection of REM Sleep Behaviour Disorder by Automated Polysomnography Analysis | Evidence suggests Rapid-Eye-Movement (REM) Sleep Behaviour Disorder (RBD) is
an early predictor of Parkinson's disease. This study proposes a
fully-automated framework for RBD detection consisting of automated sleep
staging followed by RBD identification. Analysis was assessed using a limited
polysomnography montage from 53 participants with RBD and 53 age-matched
healthy controls. Sleep stage classification was achieved using a Random Forest
(RF) classifier and 156 features extracted from electroencephalogram (EEG),
electrooculogram (EOG) and electromyogram (EMG) channels. For RBD detection, a
RF classifier was trained combining established techniques to quantify muscle
atonia with additional features that incorporate sleep architecture and the EMG
fractal exponent. Automated multi-state sleep staging achieved a 0.62 Cohen's
Kappa score. RBD detection accuracy improved by 10% to 96% (compared to
individual established metrics) when using manually annotated sleep staging.
Accuracy remained high (92%) when using automated sleep staging. This study
outperforms established metrics and demonstrates that incorporating sleep
architecture and sleep stage transitions can benefit RBD detection. This study
also achieved automated sleep staging with a level of accuracy comparable to
manual annotation. This study validates a tractable, fully-automated, and
sensitive pipeline for RBD identification that could be translated to wearable
take-home technology. | ['Mkael Symmonds', 'Navin Cooray', 'Fernando Andreotti', 'Michele T. M. Hu', 'Christine Lo', 'Maarten De Vos'] | 2018-11-12 | null | null | null | null | ['sleep-staging'] | ['medical'] | [ 2.86903292e-01 -2.70168573e-01 -3.94318521e-01 -2.92113364e-01
-6.58357739e-01 -2.54183680e-01 2.01998115e-01 -1.41143173e-01
-8.33494604e-01 9.48422730e-01 4.57943976e-01 -3.03687513e-01
-3.36656332e-01 2.62885150e-02 5.63853264e-01 -6.14572346e-01
-3.28103483e-01 4.83791292e-01 1.22332700e-01 6.08195141e-02
1.86922029e-01 3.57416183e-01 -1.23407316e+00 7.75519945e-03
8.86019111e-01 1.01655459e+00 6.53822243e-01 7.93240786e-01
4.70560998e-01 4.56393540e-01 -7.88134575e-01 2.00604126e-01
3.35048363e-02 -5.07252634e-01 -6.52750075e-01 -9.38349366e-02
-3.16093452e-02 -3.74059379e-01 -6.07737638e-02 5.78526378e-01
7.99471021e-01 -3.64843197e-02 3.39044333e-01 -8.27283323e-01
-3.96659076e-01 -1.86636895e-01 -3.69606972e-01 1.18200946e+00
5.14583468e-01 6.94859266e-01 8.79251599e-01 -1.56085327e-01
3.51942092e-01 4.66691136e-01 1.13932753e+00 1.03171694e+00
-1.67733800e+00 -8.92822385e-01 -6.83789551e-01 6.56156600e-01
-1.12554181e+00 -6.77996397e-01 1.96721241e-01 -4.40948874e-01
1.54901636e+00 4.72516149e-01 1.67888045e+00 9.27961946e-01
6.82165504e-01 1.19082093e-01 1.63805187e+00 6.17205128e-02
5.27353346e-01 -2.06586301e-01 6.65401697e-01 2.35975772e-01
4.31152344e-01 -9.74668264e-02 -3.94651234e-01 -2.23179162e-01
3.19839805e-01 2.02789545e-01 -2.28729874e-01 4.73386973e-01
-1.00201273e+00 5.89816272e-01 -1.38893545e-01 5.64758539e-01
-5.35553515e-01 -7.36638904e-02 4.11914229e-01 4.14898008e-01
4.83724684e-01 5.36196351e-01 -2.84109861e-01 -8.40518475e-01
-1.67440248e+00 -1.28590927e-01 4.74317551e-01 3.12560111e-01
2.81951845e-01 -2.79310703e-01 -2.94733316e-01 8.67969811e-01
6.10990405e-01 6.82290196e-01 1.20138156e+00 -1.24488926e+00
1.56478062e-02 9.04673159e-01 1.98604435e-01 -2.67074674e-01
-1.30099797e+00 -1.54998690e-01 -6.18636131e-01 4.09954607e-01
5.54401353e-02 1.30762443e-01 -6.87640786e-01 1.12392271e+00
-9.70215574e-02 -3.49287868e-01 -2.83793330e-01 9.47006524e-01
4.86822248e-01 -2.81144798e-01 1.34210318e-01 -6.15157664e-01
1.72785234e+00 -4.04889494e-01 -5.88978589e-01 -4.50043142e-01
2.42237672e-01 -1.87489912e-01 1.33105612e+00 6.24417663e-01
-9.51235831e-01 -1.91501990e-01 -1.18142009e+00 -2.94579361e-02
1.61708221e-01 2.60010391e-01 3.34600747e-01 1.14979744e+00
-1.35234129e+00 6.65910780e-01 -1.63075221e+00 -8.64804387e-01
9.08124685e-01 1.10187292e+00 -4.16995496e-01 5.07194698e-01
-6.54346287e-01 1.39386332e+00 -3.80531609e-01 -7.13669360e-02
-6.76727712e-01 -5.05149305e-01 -2.57937610e-01 -5.01252770e-01
-7.07850933e-01 -1.16893649e+00 8.83377373e-01 -4.28422928e-01
-1.51942670e+00 1.16445100e+00 -4.53878880e-01 -7.26633906e-01
7.25811347e-02 -1.94199905e-01 -6.46964908e-01 6.02651238e-01
4.05312806e-01 1.72183216e-01 8.27264965e-01 -1.71542749e-01
-4.81594235e-01 -1.10872698e+00 -3.88979405e-01 5.15869968e-02
-1.11529276e-01 3.68560016e-01 4.85199213e-01 -3.47194493e-01
-1.86148919e-02 -1.16309476e+00 -5.95643148e-02 -1.46833748e-01
-1.52450949e-01 -1.25452265e-01 4.89435464e-01 -1.05908179e+00
1.43221521e+00 -1.93866265e+00 -9.60921198e-02 -1.26681909e-01
9.27144885e-01 1.00626625e-01 5.06467223e-01 3.20523158e-02
4.40792888e-02 -6.22146875e-02 -2.62124717e-01 -6.66888058e-01
-2.67434921e-02 2.52299190e-01 3.96749407e-01 1.19375372e+00
-1.22185454e-01 9.28616226e-01 -6.59997225e-01 -2.26815253e-01
3.14296812e-01 4.08119798e-01 -1.92899778e-01 -2.89653409e-02
5.23186684e-01 6.58074200e-01 1.67805433e-01 6.82808757e-01
3.30477506e-02 -2.60890990e-01 -1.11708250e-02 -6.10417873e-02
-2.27362797e-01 8.61039400e-01 -3.06995213e-01 1.64566886e+00
-3.15453500e-01 8.76695275e-01 1.05653389e-03 -3.29497606e-01
8.28769445e-01 9.72058550e-02 5.61927378e-01 -6.86570525e-01
1.37763530e-01 2.13307649e-01 2.97862589e-01 -5.93705118e-01
-9.86520350e-02 -6.57319605e-01 3.22208583e-01 7.49162376e-01
-2.72378735e-02 7.84650221e-02 1.78889826e-01 -2.64507830e-01
2.02012801e+00 -8.99155959e-02 7.23895371e-01 -4.38586175e-01
3.65352243e-01 -5.09870127e-02 3.28327864e-01 4.29310918e-01
-6.59789920e-01 5.24920106e-01 1.26014292e-01 -1.15225859e-01
-5.56468785e-01 -1.01956093e+00 -3.99278700e-01 5.44150591e-01
-7.68078491e-02 -8.76163244e-01 -5.27289391e-01 -4.66912180e-01
-1.33431247e-02 7.05590606e-01 -8.12197208e-01 -3.86342227e-01
-1.44983545e-01 -1.20027161e+00 6.84623897e-01 5.82555830e-01
3.63430917e-01 -8.99918139e-01 -1.06383789e+00 2.05154672e-01
4.01669741e-02 -6.32680237e-01 -2.64613062e-01 6.29978895e-01
-1.37764609e+00 -1.18104506e+00 -7.90525794e-01 -3.64348233e-01
3.30323398e-01 1.19102644e-02 6.90140367e-01 -2.53139585e-01
-6.53479457e-01 4.96041119e-01 -1.26490623e-01 8.84366035e-02
-2.61449933e-01 1.97525080e-02 6.32245600e-01 -4.82673347e-01
1.17730737e+00 -1.30967057e+00 -1.25117898e+00 4.06174213e-01
-4.39771898e-02 -3.15790057e-01 9.00077343e-01 2.95278102e-01
7.73399711e-01 -4.56422508e-01 6.53444767e-01 3.69526520e-02
8.21261108e-01 -4.48747486e-01 -1.47535736e-02 -4.38826561e-01
-1.44950271e+00 -2.02925846e-01 3.42123121e-01 -1.49938300e-01
-4.17563349e-01 -2.29011729e-01 -2.06608415e-01 -1.17875367e-01
-3.20970446e-01 -3.47005129e-01 4.47851345e-02 5.88110350e-02
7.10811496e-01 4.58693624e-01 6.99097872e-01 -5.65528512e-01
-2.06085041e-01 1.22879171e+00 4.85231459e-01 3.53455275e-01
2.38644585e-01 7.37866640e-01 -5.73394336e-02 -1.17652369e+00
-4.29764479e-01 -1.09245682e+00 -8.37194562e-01 1.14629492e-01
1.20750022e+00 -8.95549476e-01 -7.78865814e-01 3.16193908e-01
-4.11875367e-01 -7.04576015e-01 -5.62090576e-01 8.80938768e-01
-7.56414890e-01 3.97437006e-01 -3.64502221e-01 -7.92132914e-01
-1.10629785e+00 -8.15747023e-01 1.04704344e+00 -4.45168465e-02
-1.42415750e+00 -7.63227940e-01 7.75022626e-01 6.18129909e-01
4.46427941e-01 -2.27684811e-01 5.43055415e-01 -7.29596496e-01
2.52414048e-01 1.03376433e-02 5.73367886e-02 4.77474302e-01
6.63866282e-01 -6.81698620e-01 -5.94404638e-01 -5.97180314e-02
8.00154626e-01 1.70417175e-01 4.96274143e-01 4.33941156e-01
1.05333194e-01 -1.92150906e-01 -3.26756001e-01 3.77176255e-01
1.05095279e+00 3.23281705e-01 8.05381835e-01 8.85160446e-01
1.92782044e-01 -1.45134553e-01 -3.94432768e-02 2.63328135e-01
4.79687899e-01 6.86206043e-01 1.14953242e-01 5.41253090e-01
-4.87531155e-01 5.87217867e-01 9.27452326e-01 4.28291887e-01
-6.10746562e-01 5.76481044e-01 -7.58595169e-01 3.03274632e-01
-1.14267480e+00 -1.08555973e+00 -4.32585746e-01 1.93647051e+00
7.86246061e-01 2.94643492e-01 8.92728686e-01 3.82801890e-01
4.51094955e-01 -1.83443278e-01 -8.62407148e-01 -3.16890985e-01
-8.39267392e-03 6.34841740e-01 5.57516336e-01 -5.20450473e-02
-4.14072692e-01 1.64694920e-01 7.02718067e+00 1.26672149e-01
-9.24429595e-01 7.75765657e-01 -1.78031892e-01 -1.02172875e+00
2.97603786e-01 -3.16442281e-01 -8.08667421e-01 1.00318301e+00
1.63498843e+00 6.69030249e-02 7.54577458e-01 5.41750610e-01
1.02036142e+00 -6.74302459e-01 -7.72535086e-01 1.13841534e+00
9.09285992e-02 -1.02925050e+00 -8.32933128e-01 5.16143322e-01
5.26291840e-02 8.35280657e-01 -3.29486012e-01 -8.72013047e-02
-4.04210210e-01 -8.66483510e-01 2.86069751e-01 9.52213049e-01
1.23845005e+00 -2.30313316e-01 5.99112809e-01 -1.12446226e-01
-9.51044381e-01 -2.96982318e-01 1.26928627e-01 -2.77152061e-01
9.52040777e-02 3.29737365e-01 -9.64241266e-01 -2.88003296e-01
1.03967178e+00 1.25828433e+00 -8.58993471e-01 1.14946628e+00
-3.79175901e-01 9.30591047e-01 -5.60060382e-01 -1.63160786e-01
-2.56892234e-01 -1.82842344e-01 8.75413716e-01 7.96761751e-01
1.89079523e-01 -1.81799605e-02 -8.08461905e-01 1.02680731e+00
5.07584929e-01 -3.35651398e-01 -8.00334811e-02 -1.38164297e-01
1.01112179e-01 1.36995387e+00 -9.35805976e-01 2.63839513e-01
-4.60703671e-01 1.11117709e+00 -2.68350929e-01 -2.69621044e-01
-3.39854211e-01 -2.62831360e-01 7.96426475e-01 4.52022821e-01
2.30266396e-02 -2.36238092e-01 -6.49760127e-01 -8.14628422e-01
-3.91257629e-02 -6.63777292e-01 2.80102521e-01 -8.08744550e-01
-1.19852078e+00 3.57180119e-01 -2.32512504e-01 -1.20055747e+00
8.49861354e-02 -3.20367724e-01 -8.30756605e-01 8.58785927e-01
-7.86863685e-01 -8.44092906e-01 -3.09528768e-01 5.76429188e-01
4.86568540e-01 -2.19610736e-01 9.55350637e-01 2.54398227e-01
-6.69215381e-01 6.12535365e-02 5.42160980e-02 -4.60217804e-01
5.26612103e-01 -1.64153337e+00 2.64828447e-02 3.89918268e-01
-3.34860861e-01 1.00799310e+00 5.00176728e-01 -9.21044111e-01
-1.22961390e+00 -8.76901448e-01 8.08611631e-01 -7.57176161e-01
8.85636806e-01 -3.12320814e-02 -2.34829783e-01 4.71292496e-01
-1.72085151e-01 -6.05097830e-01 1.45133114e+00 -3.55680734e-02
5.09487689e-01 -2.75317430e-01 -1.46579492e+00 2.29298458e-01
1.09963775e+00 -7.18495548e-01 -1.35676062e+00 1.20878816e-01
-8.01011473e-02 1.53073356e-01 -1.35141683e+00 -6.77612871e-02
1.08212626e+00 -1.15370607e+00 5.58744073e-01 -1.45375967e-01
-1.53378949e-01 -2.58350670e-01 4.97557223e-02 -8.29822540e-01
-3.37986559e-01 -1.03913939e+00 -5.09078503e-01 8.11943054e-01
-3.35659046e-04 -6.60447478e-01 8.02287519e-01 7.25501835e-01
-4.39627826e-01 -8.72510493e-01 -1.21082509e+00 -8.65040839e-01
-4.20213133e-01 -5.67457795e-01 1.79507568e-01 -5.97561419e-04
5.21119058e-01 6.32989228e-01 1.49643183e-01 -2.30171338e-01
2.75689751e-01 -1.40098989e-01 1.12327509e-01 -1.41761613e+00
-7.15260357e-02 -4.43582386e-01 -1.05293977e+00 -4.10331756e-01
5.94389364e-02 -1.21308160e+00 -4.13533449e-01 -2.14481401e+00
3.53168905e-01 -1.41508341e-01 -4.31026667e-01 6.25762224e-01
1.82714343e-01 9.31378484e-01 -3.02352667e-01 8.05549920e-01
-5.94641566e-01 2.42331609e-01 5.94039381e-01 -7.31305080e-03
-7.97629476e-01 4.03177232e-01 -1.02544129e+00 7.32057035e-01
1.17503083e+00 -6.89454377e-01 -4.41876203e-01 3.81676912e-01
2.51497746e-01 -3.67456824e-01 5.91309965e-01 -1.36647582e+00
-7.35305157e-03 4.25306976e-01 5.97012877e-01 -4.58768994e-01
5.90502322e-01 -5.85177004e-01 2.61100233e-01 8.57780576e-01
1.92246124e-01 -3.94160450e-02 1.34376258e-01 5.64460993e-01
6.38824403e-01 -9.80554447e-02 7.48757899e-01 -1.70282170e-01
-3.78291517e-01 -9.93524343e-02 -1.11060095e+00 1.81312859e-02
7.19707847e-01 -1.00486791e+00 -4.61430311e-01 -2.40786802e-02
-1.32017708e+00 -8.64762887e-02 7.39147663e-01 -5.49307130e-02
5.27403116e-01 -8.78495097e-01 -2.18498245e-01 2.86941797e-01
-1.16567910e-01 -5.66917956e-01 5.83521612e-02 2.09347439e+00
-3.24240595e-01 5.78705311e-01 -7.34386146e-01 -7.26535022e-01
-1.39259362e+00 3.78489494e-02 4.00861740e-01 -2.20049787e-02
-1.20314860e+00 5.27281821e-01 -7.64209569e-01 4.49894965e-01
-4.92324419e-02 -5.77657819e-01 -3.49377781e-01 4.16711032e-01
8.20676506e-01 1.09808254e+00 2.09573135e-01 -6.74146831e-01
-7.83161938e-01 3.76598239e-01 2.63974100e-01 -6.75593019e-02
1.54948735e+00 -7.06308484e-01 -3.80541056e-01 9.57843959e-01
9.16153193e-01 9.08987671e-02 -8.04919660e-01 9.16338027e-01
1.06089704e-01 2.58657157e-01 2.35482216e-01 -1.05794585e+00
-6.41243398e-01 5.42334616e-01 1.40791702e+00 3.53766590e-01
1.15588367e+00 1.61014199e-01 1.18336999e+00 3.29878718e-01
4.08654183e-01 -9.83088315e-01 -3.63654196e-01 -3.97485763e-01
6.80310011e-01 -7.78015852e-01 4.76292133e-01 6.43708825e-01
-5.53692102e-01 1.08752310e+00 3.43933813e-02 -3.51306796e-01
5.96515059e-01 -1.49672013e-03 -1.77458197e-01 -5.46956360e-01
-3.61243218e-01 -3.26318443e-01 2.93743014e-01 9.19614792e-01
9.45339203e-02 -1.12046458e-01 -8.50617468e-01 1.30442154e+00
-6.21403873e-01 5.17113924e-01 3.24168801e-01 7.05101490e-01
-5.68065166e-01 -9.72213447e-01 -1.17643110e-01 1.55263078e+00
-7.57018387e-01 -1.17610522e-01 -3.49764973e-01 7.22447634e-01
2.52749056e-01 1.20436466e+00 1.44034490e-01 -4.77792382e-01
3.38490382e-02 3.00091594e-01 5.78911483e-01 -1.08380532e+00
-8.13075125e-01 -1.69981971e-01 3.51597309e-01 -7.69991159e-01
-9.00112092e-01 -1.15751791e+00 -1.43831885e+00 2.31893510e-01
-1.40666500e-01 -6.25125170e-02 5.71224511e-01 1.24418318e+00
5.50090611e-01 2.97399610e-01 1.02961086e-01 -7.78917134e-01
-1.57835275e-01 -1.26303411e+00 -1.06727397e+00 -1.96234256e-01
7.64191031e-01 -8.54395747e-01 -6.77470624e-01 8.03776830e-02] | [13.547517776489258, 3.390143394470215] |
6fcfbb64-2f3b-426d-b2d9-f1c81e861e1b | mydigitalfootprint-an-extensive-context | 2306.15990 | null | https://arxiv.org/abs/2306.15990v1 | https://arxiv.org/pdf/2306.15990v1.pdf | MyDigitalFootprint: an extensive context dataset for pervasive computing applications at the edge | The widespread diffusion of connected smart devices has contributed to the rapid expansion and evolution of the Internet at its edge. Personal mobile devices interact with other smart objects in their surroundings, adapting behavior based on rapidly changing user context. The ability of mobile devices to process this data locally is crucial for quick adaptation. This can be achieved through a single elaboration process integrated into user applications or a middleware platform for context processing. However, the lack of public datasets considering user context complexity in the mobile environment hinders research progress. We introduce MyDigitalFootprint, a large-scale dataset comprising smartphone sensor data, physical proximity information, and Online Social Networks interactions. This dataset supports multimodal context recognition and social relationship modeling. It spans two months of measurements from 31 volunteer users in their natural environment, allowing for unrestricted behavior. Existing public datasets focus on limited context data for specific applications, while ours offers comprehensive information on the user context in the mobile environment. To demonstrate the dataset's effectiveness, we present three context-aware applications utilizing various machine learning tasks: (i) a social link prediction algorithm based on physical proximity data, (ii) daily-life activity recognition using smartphone-embedded sensors data, and (iii) a pervasive context-aware recommender system. Our dataset, with its heterogeneity of information, serves as a valuable resource to validate new research in mobile and edge computing. | ['Franca Delmastro', 'Mattia Giovanni Campana'] | 2023-06-28 | null | null | null | null | ['activity-recognition', 'link-prediction', 'edge-computing'] | ['computer-vision', 'graphs', 'time-series'] | [ 3.84233564e-01 -4.75652069e-01 -5.90356588e-01 -3.23474050e-01
-3.78894478e-01 -5.03260672e-01 5.49046099e-01 1.88722491e-01
-1.69077232e-01 5.93808651e-01 6.03974044e-01 -4.44140196e-01
-2.81809390e-01 -9.17821646e-01 -2.82553643e-01 -3.29696655e-01
-1.18547946e-01 -5.53983562e-02 2.32091889e-01 -2.60614574e-01
3.97314690e-02 2.43205637e-01 -1.82134175e+00 4.21169937e-01
7.37719595e-01 1.22878158e+00 4.72041190e-01 1.01832175e+00
1.27474189e-01 5.30327320e-01 -2.71235645e-01 3.73136774e-02
-1.34546667e-01 -1.28579706e-01 -4.13738877e-01 -1.54089928e-01
5.84639162e-02 -3.91879566e-02 -2.93037802e-01 3.79263103e-01
7.95408905e-01 2.10890427e-01 9.25502554e-02 -1.39715791e+00
-4.30348694e-01 2.11457059e-01 -4.87018004e-03 5.17026544e-01
1.21806133e+00 -1.07721932e-01 8.67921710e-01 -6.54009521e-01
5.00751674e-01 4.27477658e-01 1.00769603e+00 2.08372846e-01
-8.73386860e-01 -2.05193028e-01 1.45398244e-01 3.97453040e-01
-1.61192417e+00 -6.91026449e-01 8.38105798e-01 -1.41855016e-01
1.20992172e+00 6.66447878e-01 1.14928150e+00 1.66690445e+00
-1.74787063e-02 3.92688274e-01 6.60899460e-01 -3.77593875e-01
4.65565473e-01 3.03340971e-01 1.28836691e-01 3.00309777e-01
1.79891199e-01 -2.76105344e-01 -8.93195927e-01 -6.56616211e-01
1.74741715e-01 6.50003076e-01 -1.85659468e-01 -5.43521382e-02
-1.31333816e+00 -3.38849910e-02 1.25413999e-01 6.20245755e-01
-4.73087072e-01 -1.94383191e-03 1.82866063e-02 2.45379448e-01
5.04222214e-01 -1.89968973e-01 -6.86263680e-01 -1.04724598e+00
-6.88350558e-01 -2.67717987e-01 1.32071483e+00 1.08340657e+00
6.83339238e-01 -5.21846235e-01 2.00133681e-01 9.62980151e-01
4.78573084e-01 6.10220969e-01 7.40964174e-01 -9.85638142e-01
4.79490399e-01 8.29643548e-01 1.86467737e-01 -1.44828749e+00
-5.96907496e-01 -2.75204539e-01 -4.34103459e-01 -7.58824766e-01
2.43291214e-01 -4.34824228e-01 2.39119288e-02 1.63907659e+00
5.30296206e-01 8.13067138e-01 -2.54878759e-01 3.25883061e-01
7.40376055e-01 3.57752740e-01 9.56763923e-02 -3.05361718e-01
1.29385328e+00 -6.28651798e-01 -6.06867015e-01 -1.99793279e-01
7.66290188e-01 -4.68115896e-01 1.40245163e+00 3.07252824e-01
-7.55711496e-01 -3.12165350e-01 -9.71962452e-01 1.77747920e-01
-8.39644253e-01 -2.65749931e-01 1.00497437e+00 1.13883698e+00
-1.15261710e+00 3.30717951e-01 -6.54548109e-01 -1.09437644e+00
5.11493027e-01 5.03960907e-01 -2.87643164e-01 8.65846202e-02
-9.83182788e-01 2.34289289e-01 -2.84662366e-01 -3.01441282e-01
1.34521991e-01 -5.77328622e-01 -5.75541437e-01 6.25859350e-02
4.97831613e-01 -8.68507624e-01 9.02256548e-01 -8.05410862e-01
-1.57198048e+00 6.79914594e-01 -4.01200235e-01 1.60410302e-03
2.30114624e-01 -1.43167362e-01 -1.19889784e+00 -6.38280064e-02
-2.76647508e-02 -1.06751554e-01 3.64170432e-01 -9.27647114e-01
-9.63825881e-01 -5.41520059e-01 -1.82600453e-01 2.47772604e-01
-8.23276162e-01 -1.30681202e-01 -7.56020725e-01 -4.29488957e-01
-1.40312105e-01 -1.09116066e+00 -7.83346966e-02 -3.57737154e-01
-5.27830496e-02 9.16866288e-02 1.09684730e+00 -4.22314882e-01
1.84496152e+00 -2.12506652e+00 -5.52077830e-01 5.14844000e-01
1.05928451e-01 -2.37800349e-02 8.39935318e-02 5.52286267e-01
3.25770557e-01 2.22018316e-01 1.36100680e-01 -6.31591320e-01
-7.35358074e-02 3.71190131e-01 1.36481598e-01 7.40327165e-02
-6.97578728e-01 1.06286192e+00 -1.23152041e+00 -4.56977725e-01
8.20636377e-02 5.80745578e-01 -6.97472513e-01 -4.73575778e-02
1.99097812e-01 4.72093195e-01 -3.76786798e-01 1.00115967e+00
1.06333323e-01 -5.10444820e-01 4.79613245e-01 -6.53302073e-02
5.96175678e-02 3.35132986e-01 -1.22601724e+00 1.79720938e+00
-8.28977227e-01 5.87438822e-01 1.50713682e-01 -7.32006729e-01
4.07117963e-01 3.33028495e-01 1.03533268e+00 -7.10473359e-01
5.38708046e-02 2.80884486e-02 -5.25008142e-01 -8.40296924e-01
7.12279022e-01 7.84306884e-01 -1.87468171e-01 6.57311141e-01
-4.37823594e-01 5.70191562e-01 -8.73659253e-02 2.20275775e-01
1.46549070e+00 -8.93760845e-02 5.04244864e-01 -8.76683742e-03
3.38184625e-01 -5.58494210e-01 5.63458145e-01 7.95422137e-01
-5.32326937e-01 3.94411087e-01 -1.28543705e-01 -2.53755003e-01
-1.71510816e-01 -1.07690251e+00 5.00604808e-02 1.54747891e+00
3.18110466e-01 -1.00975692e+00 -5.21715045e-01 -7.26962566e-01
1.97931938e-03 2.29467273e-01 -4.86410588e-01 1.43424962e-02
-2.87305683e-01 -8.39379132e-01 3.43065828e-01 3.99775267e-01
7.58314908e-01 -7.95583248e-01 -2.72508055e-01 8.54823142e-02
-6.70765519e-01 -1.33337629e+00 -7.53188670e-01 -3.10886383e-01
-8.01339328e-01 -1.17094922e+00 1.01108905e-02 -4.90582645e-01
3.84344727e-01 8.16058457e-01 1.25440824e+00 2.31906354e-01
-1.56567413e-02 1.36863458e+00 -3.12016249e-01 -8.08392838e-02
1.51153445e-01 3.79093707e-01 4.32046294e-01 3.88053656e-01
6.76798046e-01 -1.30179572e+00 -9.50115979e-01 8.98181558e-01
-4.74998295e-01 -3.63768846e-01 1.10864021e-01 1.83715448e-01
6.00691319e-01 4.22763601e-02 7.78851986e-01 -8.94613743e-01
6.73276424e-01 -1.34706783e+00 2.04976618e-01 1.81734934e-01
-7.88984656e-01 -9.69123185e-01 1.79734319e-01 -5.56797683e-01
-1.10281849e+00 -1.07376538e-02 -3.40162739e-02 4.50198025e-01
-3.29988301e-01 4.02590841e-01 -7.32480466e-01 -1.92699041e-02
8.67723584e-01 -2.99600195e-02 -4.92925704e-01 -3.32679510e-01
2.50117630e-01 1.42358315e+00 4.36431140e-01 -4.45344537e-01
2.11834013e-01 6.70721292e-01 -1.66132987e-01 -1.35494745e+00
-4.24608946e-01 -8.06127548e-01 -4.52395439e-01 -3.73895198e-01
3.86544466e-01 -9.44216549e-01 -9.12441075e-01 2.79333830e-01
-3.75936508e-01 -6.60670042e-01 -2.11545974e-01 3.33442181e-01
-3.44870389e-01 3.13451797e-01 -2.48312503e-01 -8.80879641e-01
-2.26961330e-01 -5.20212710e-01 9.16674733e-01 2.86245346e-01
-7.25635469e-01 -1.39498138e+00 1.20134577e-01 7.71716535e-01
7.23787665e-01 2.19245389e-01 7.57757947e-02 -4.32402164e-01
-4.40235376e-01 -5.53595185e-01 -2.18639922e-04 -1.81700811e-01
7.54214227e-01 -1.12847999e-01 -1.13774800e+00 -1.66216679e-02
-1.95104718e-01 3.44768018e-01 1.18342340e-01 1.88492730e-01
1.29183674e+00 -2.96251178e-01 -1.00820410e+00 6.16441727e-01
9.64729965e-01 1.88866198e-01 6.70214236e-01 1.88191056e-01
8.51898253e-01 9.76633802e-02 2.71970928e-01 6.92047775e-01
1.08441901e+00 9.05701756e-01 2.44866490e-01 2.15032384e-01
-4.58194651e-02 -2.88642704e-01 3.09989363e-01 8.21803749e-01
-3.32879454e-01 -4.09678459e-01 -8.63925040e-01 4.76165086e-01
-2.19931793e+00 -1.18421030e+00 -1.25757784e-01 2.28502393e+00
3.62693608e-01 3.36577259e-02 4.75295573e-01 2.92440355e-01
5.35329461e-01 3.09936889e-02 -5.11680722e-01 -7.84574822e-02
-1.01704895e-01 -4.06096578e-01 2.23678917e-01 3.25662374e-01
-8.80033374e-01 4.96502310e-01 6.37753916e+00 5.72744370e-01
-9.72888708e-01 4.73547101e-01 5.04215062e-01 -4.86246288e-01
-3.51397961e-01 -2.37700999e-01 -6.41386569e-01 1.03547978e+00
1.24706089e+00 1.57266259e-01 9.46893156e-01 1.02632928e+00
6.47373796e-01 -5.50232708e-01 -1.02532268e+00 1.51609218e+00
6.74552470e-02 -1.24508953e+00 -5.85111260e-01 4.33402568e-01
7.48694479e-01 5.36365330e-01 7.04971105e-02 2.54394919e-01
-1.74370706e-01 -4.44973379e-01 -1.56953540e-02 1.01197767e+00
7.19476700e-01 -4.06246603e-01 2.98422843e-01 3.97930861e-01
-1.42020965e+00 -4.01690364e-01 5.22330105e-01 -4.83838797e-01
1.96275234e-01 8.91326249e-01 -6.54210687e-01 1.95525959e-01
1.09319997e+00 1.10539734e+00 -7.05709100e-01 7.20028400e-01
3.52455527e-01 8.37537766e-01 -8.08229208e-01 -2.40894958e-01
-6.35206342e-01 -2.62895614e-01 5.51276386e-01 1.33123052e+00
6.22290075e-01 7.64205232e-02 1.43582895e-01 1.35014027e-01
-3.00380528e-01 1.48172647e-01 -1.03732395e+00 1.38546536e-02
9.40854669e-01 1.49287117e+00 -6.82746410e-01 -1.23699598e-01
-7.68027425e-01 1.04964280e+00 -7.73717761e-02 4.62902129e-01
-5.98993838e-01 -3.12521398e-01 9.85765100e-01 5.18664539e-01
-1.31282359e-01 -5.58767080e-01 -4.24365789e-01 -1.36960733e+00
1.91784188e-01 -5.52400053e-01 4.77420807e-01 -7.17045486e-01
-1.18628597e+00 1.29945800e-01 -3.99969071e-01 -1.13144016e+00
-2.70846188e-01 -1.38229698e-01 -9.24754202e-01 3.75804991e-01
-1.06649697e+00 -1.15926385e+00 -7.95566082e-01 9.80123878e-01
1.93703771e-01 -1.90831915e-01 9.13427055e-01 8.64468634e-01
-7.03481555e-01 7.11681485e-01 1.61262736e-01 -4.02833551e-01
6.93423152e-01 -9.93709087e-01 2.07474396e-01 4.68436003e-01
2.21887261e-01 8.51349354e-01 1.28709540e-01 -7.60550618e-01
-1.64200890e+00 -9.07393038e-01 1.13699508e+00 -1.07842052e+00
6.60258293e-01 -4.98338372e-01 -5.10957003e-01 7.31179953e-01
-1.52439892e-01 4.62529749e-01 1.55347466e+00 5.33157825e-01
1.84843183e-01 -4.20876771e-01 -1.36680651e+00 9.83146012e-01
1.88487196e+00 -8.51747870e-01 4.56523001e-02 1.28447756e-01
4.52506155e-01 -4.58012037e-02 -1.11490297e+00 1.19614772e-01
9.67367053e-01 -1.08458698e+00 1.02368009e+00 -6.35403320e-02
-4.01387513e-01 -2.28133932e-01 -5.97133636e-01 -9.36284363e-01
-8.97343680e-02 -1.25064981e+00 -9.99567151e-01 1.37770176e+00
3.55142266e-01 -8.57456982e-01 1.04080129e+00 1.12532640e+00
2.09348097e-01 -8.41612339e-01 -8.45148861e-01 -3.63737673e-01
-9.28440571e-01 -1.22135878e+00 9.98244762e-01 1.12271309e+00
4.73372459e-01 3.37817192e-01 -1.31987855e-01 7.27561265e-02
1.34349450e-01 -2.56596625e-01 1.01901150e+00 -1.43212795e+00
-4.88334417e-01 -2.41512939e-01 -3.45299870e-01 -1.12848330e+00
-3.27201456e-01 -6.38940394e-01 -6.69979870e-01 -1.30592299e+00
-9.34914649e-02 -8.73275638e-01 -2.70699263e-01 3.54490340e-01
1.83830019e-02 4.94267911e-01 -2.97804087e-01 1.92438066e-01
-1.18097699e+00 1.87586427e-01 5.21395683e-01 8.71833116e-02
-8.70455384e-01 5.90581954e-01 -9.79966998e-01 8.15244377e-01
1.00801718e+00 1.13466069e-01 -6.71990573e-01 -7.59217888e-02
9.90632534e-01 -2.74687916e-01 2.09755540e-01 -1.17750704e+00
4.59284276e-01 -1.57308936e-01 3.05631131e-01 -1.02843702e-01
5.11605620e-01 -1.29209900e+00 3.11122119e-01 -8.91764387e-02
2.19671160e-01 -8.56957883e-02 -1.29658625e-01 1.01755428e+00
3.93158585e-01 3.93896371e-01 -1.19653895e-01 1.70612872e-01
-6.84910595e-01 4.61771101e-01 -6.01523101e-01 -1.43745124e-01
9.99840558e-01 -7.34030247e-01 -1.77815095e-01 -7.43129730e-01
-1.04094410e+00 1.13772802e-01 6.59004986e-01 7.08432496e-01
2.49765962e-01 -1.53172493e+00 2.73439169e-01 2.37561956e-01
3.16016972e-01 -5.78143418e-01 2.28638247e-01 9.52629983e-01
4.33946624e-02 -7.10036280e-03 1.94941297e-01 -6.42252922e-01
-1.42673564e+00 2.26962715e-01 3.12743247e-01 1.00476429e-01
-3.80283743e-01 1.66699871e-01 -8.14511240e-01 -4.38746512e-01
2.59265512e-01 -3.53206366e-01 -3.22871238e-01 1.98804885e-01
8.04963708e-01 8.67228687e-01 2.51834244e-01 -5.92669308e-01
-4.44356352e-01 4.90703017e-01 6.10743582e-01 3.55043001e-02
1.15496922e+00 -1.01771331e+00 2.21641108e-01 7.41320133e-01
1.21582949e+00 5.86065769e-01 -9.27076697e-01 -4.06712681e-01
1.41782299e-01 -4.98749971e-01 -1.80629920e-02 -8.14671934e-01
-6.97363794e-01 1.21121995e-01 8.47377360e-01 8.32908928e-01
1.16022265e+00 -6.19639084e-02 1.15542197e+00 6.76803112e-01
8.31674099e-01 -1.43004739e+00 -2.47066915e-02 4.11235571e-01
3.32064569e-01 -1.59164810e+00 -1.35848865e-01 -4.46017623e-01
-4.07159179e-01 5.02860844e-01 2.23971352e-01 6.52851522e-01
1.47284257e+00 7.91372880e-02 -8.78958777e-02 4.36155833e-02
-6.42955422e-01 -3.48174483e-01 5.62885590e-02 1.07271218e+00
3.96408021e-01 1.07212953e-01 7.50536248e-02 1.18320739e+00
-1.22014061e-01 2.93106616e-01 1.92160487e-01 1.07722139e+00
-1.76311955e-01 -1.19341421e+00 -2.16287613e-01 7.04941094e-01
-4.80534375e-01 1.65007234e-01 -3.25780123e-01 2.12720707e-01
4.21875596e-01 1.65059698e+00 -4.92879450e-02 -5.92776656e-01
3.67329538e-01 8.25773105e-02 1.16003804e-01 -4.89620715e-01
-5.98585427e-01 -5.21173239e-01 3.64559054e-01 -1.06989336e+00
-3.77459288e-01 -9.38277066e-01 -1.03848696e+00 -5.21719992e-01
1.02855712e-01 -1.53798223e-01 9.71372068e-01 9.80268359e-01
1.20232868e+00 4.16799545e-01 8.26756001e-01 -1.12153900e+00
6.26622140e-01 -7.02214539e-01 -4.18145388e-01 4.79265720e-01
1.80009212e-02 -4.04076248e-01 -1.95349559e-01 1.15617029e-01] | [7.475320339202881, 1.2474594116210938] |
b5265bac-45ab-404c-926c-b0be3c2f3329 | finegan-unsupervised-hierarchical | 1811.11155 | null | http://arxiv.org/abs/1811.11155v2 | http://arxiv.org/pdf/1811.11155v2.pdf | FineGAN: Unsupervised Hierarchical Disentanglement for Fine-Grained Object Generation and Discovery | We propose FineGAN, a novel unsupervised GAN framework, which disentangles
the background, object shape, and object appearance to hierarchically generate
images of fine-grained object categories. To disentangle the factors without
supervision, our key idea is to use information theory to associate each factor
to a latent code, and to condition the relationships between the codes in a
specific way to induce the desired hierarchy. Through extensive experiments, we
show that FineGAN achieves the desired disentanglement to generate realistic
and diverse images belonging to fine-grained classes of birds, dogs, and cars.
Using FineGAN's automatically learned features, we also cluster real images as
a first attempt at solving the novel problem of unsupervised fine-grained
object category discovery. Our code/models/demo can be found at
https://github.com/kkanshul/finegan | ['Yong Jae Lee', 'Krishna Kumar Singh', 'Utkarsh Ojha'] | 2018-11-27 | finegan-unsupervised-hierarchical-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Singh_FineGAN_Unsupervised_Hierarchical_Disentanglement_for_Fine-Grained_Object_Generation_and_Discovery_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Singh_FineGAN_Unsupervised_Hierarchical_Disentanglement_for_Fine-Grained_Object_Generation_and_Discovery_CVPR_2019_paper.pdf | cvpr-2019-6 | ['fine-grained-visual-categorization'] | ['computer-vision'] | [ 1.77110985e-01 1.44550696e-01 -9.87499058e-02 -3.95189226e-01
-4.34898406e-01 -8.69087994e-01 7.90190816e-01 -4.20886308e-01
3.75165910e-01 4.46812302e-01 3.55715811e-01 7.19893053e-02
2.39787996e-02 -8.40919137e-01 -7.74459839e-01 -6.34052992e-01
1.71412438e-01 7.39656031e-01 -1.31311342e-01 1.18967809e-01
1.12239063e-01 4.08588707e-01 -1.61702251e+00 3.25468391e-01
8.57628286e-01 5.84451616e-01 1.27062127e-02 8.21296096e-01
1.66846856e-01 8.03537726e-01 -3.79745573e-01 -4.53029633e-01
5.03701448e-01 -7.39842176e-01 -6.68586969e-01 7.21906245e-01
6.18570864e-01 -3.05685818e-01 -2.77905315e-01 1.20654619e+00
4.96301651e-02 -1.73160896e-01 1.04131079e+00 -1.52588499e+00
-9.74516451e-01 5.43806672e-01 -6.91745639e-01 -6.70769662e-02
-1.80197682e-03 2.99574584e-01 1.15606391e+00 -6.70133770e-01
3.46190661e-01 1.38723898e+00 2.50652879e-01 7.01930463e-01
-1.59838593e+00 -9.26847339e-01 -7.93490484e-02 -3.24496217e-02
-1.52762556e+00 -4.32898164e-01 7.66535521e-01 -9.48984742e-01
1.28721967e-01 1.78585142e-01 8.11658323e-01 1.08009136e+00
-5.89951612e-02 6.87745452e-01 1.31226277e+00 -3.33211303e-01
2.33696774e-01 6.52167797e-02 -5.02446368e-02 9.72598910e-01
3.90755206e-01 4.09154922e-01 -3.58676195e-01 -1.17897704e-01
1.19825983e+00 1.62465751e-01 -2.15362031e-02 -6.13481939e-01
-1.33320844e+00 1.02212727e+00 6.08198822e-01 7.81007931e-02
-1.87624797e-01 4.01550800e-01 -2.61385590e-01 1.57283336e-01
1.56709373e-01 6.14539206e-01 -3.45088989e-01 3.49341810e-01
-6.54364705e-01 1.67154521e-01 6.13143742e-01 1.04112124e+00
9.56861556e-01 8.17276761e-02 -3.65756273e-01 6.81550205e-01
5.76327860e-01 5.50616443e-01 9.81790647e-02 -1.24384165e+00
-1.31688109e-02 5.89500904e-01 6.17468730e-04 -8.38579595e-01
2.05704451e-01 -3.76019895e-01 -1.07434082e+00 3.22470665e-01
2.36485556e-01 -1.30412757e-01 -1.37367487e+00 1.72176659e+00
2.23360077e-01 1.28246009e-01 -2.45181173e-01 8.05889249e-01
6.90671206e-01 5.48280299e-01 4.85502183e-02 2.91071087e-01
1.38862479e+00 -1.08647811e+00 -3.34918797e-01 -1.54743001e-01
-1.16612054e-01 -5.77963948e-01 1.04317153e+00 1.49337754e-01
-8.64375532e-01 -6.05413258e-01 -9.94257808e-01 1.96661919e-01
-1.58922032e-01 1.59194499e-01 8.74684453e-01 4.92050856e-01
-9.22092199e-01 2.06411362e-01 -7.46730447e-01 -3.57023925e-02
7.98761368e-01 2.12474614e-01 -2.31110811e-01 -7.15809241e-02
-7.83430994e-01 3.41852725e-01 3.69904578e-01 -1.88646674e-01
-1.42870116e+00 -5.78026891e-01 -8.81570995e-01 8.17583129e-02
1.37515396e-01 -1.21693397e+00 1.07158792e+00 -1.07883239e+00
-1.22733152e+00 1.08878124e+00 9.52052623e-02 -1.76293656e-01
2.91888446e-01 2.05974281e-01 -1.17542759e-01 2.21284926e-02
2.92203933e-01 9.37730730e-01 1.21829164e+00 -1.66278696e+00
-5.14672458e-01 -2.25788370e-01 6.40492961e-02 -1.74890384e-02
2.46929601e-02 -1.68152928e-01 -3.86446595e-01 -9.95805681e-01
-8.89675543e-02 -9.97132361e-01 -4.15516704e-01 -7.79790804e-02
-6.76136792e-01 -2.68156584e-02 4.47166204e-01 -3.26158762e-01
6.09637916e-01 -2.22161937e+00 3.58475327e-01 2.28495941e-01
7.68241286e-01 -2.49718547e-01 -3.43926340e-01 1.09124742e-01
-9.47204828e-02 2.07392141e-01 -3.09722990e-01 -1.52835578e-01
1.86542317e-01 1.63792074e-01 -1.96776465e-01 3.46280396e-01
3.17943692e-01 1.15360737e+00 -8.14794362e-01 -4.29525346e-01
2.25734711e-01 4.64706749e-01 -7.72769690e-01 4.22729164e-01
-4.34505701e-01 7.51110852e-01 -5.43554008e-01 5.41864634e-01
5.80085695e-01 -6.82078898e-01 1.55742690e-01 -2.91969538e-01
3.83615136e-01 -1.04884118e-01 -1.05690515e+00 1.31181407e+00
-1.86102435e-01 5.87950945e-01 -8.85584727e-02 -6.53988779e-01
6.98817670e-01 3.05408016e-02 2.37377077e-01 -3.62277180e-01
2.76071489e-01 -1.96285099e-01 -6.20723888e-02 -1.38703376e-01
1.07781798e-01 -2.67993331e-01 -4.15502012e-01 5.05136073e-01
2.68898875e-01 -4.26993638e-01 1.88148543e-01 3.46816421e-01
8.62641811e-01 -1.60627887e-01 2.86949962e-01 -3.52297366e-01
-2.26661898e-02 -7.42642730e-02 4.67007846e-01 7.69256234e-01
1.60566941e-01 8.79109800e-01 3.20985287e-01 -4.64123905e-01
-1.23464441e+00 -1.46340406e+00 1.08427592e-01 9.32030797e-01
1.41917005e-01 -2.59682626e-01 -7.82417536e-01 -7.45100260e-01
4.06341329e-02 4.97679651e-01 -9.16796327e-01 -5.01182936e-02
-1.42297551e-01 -5.57011843e-01 3.82866323e-01 3.41420949e-01
4.64499503e-01 -1.09094930e+00 -4.16043282e-01 -3.05478811e-01
-2.04593822e-01 -8.19388688e-01 -7.65158117e-01 1.33134022e-01
-4.58946615e-01 -1.13391697e+00 -5.38100183e-01 -6.82206035e-01
1.18646741e+00 1.81583747e-01 1.43193519e+00 2.14047238e-01
-3.92919868e-01 2.83769578e-01 -2.90879726e-01 -8.01945552e-02
-5.97951293e-01 -8.00896734e-02 -1.44137770e-01 2.51458764e-01
8.63190964e-02 -7.58729875e-01 -5.14846027e-01 3.63346726e-01
-1.02444577e+00 5.59609354e-01 6.86744630e-01 8.59931707e-01
5.82908690e-01 4.30718631e-01 7.70820901e-02 -1.10720754e+00
2.57394224e-01 -4.02665854e-01 -9.15620744e-01 2.54238307e-01
-3.54213178e-01 3.91055673e-01 3.18848819e-01 -4.02534664e-01
-8.46294701e-01 2.06783637e-01 1.12805836e-01 -6.01814449e-01
-6.01235807e-01 1.16683654e-01 -4.77870494e-01 1.72789514e-01
5.94391286e-01 1.92904592e-01 -2.82682031e-01 -3.70247304e-01
8.10233235e-01 3.68755162e-01 5.90718389e-01 -7.87132859e-01
1.24425626e+00 4.85500753e-01 -3.24615866e-01 -3.54170442e-01
-1.04069400e+00 -7.81109855e-02 -7.48382390e-01 1.67383105e-02
1.07608414e+00 -1.15670359e+00 -4.80971336e-01 4.39825535e-01
-8.84028077e-01 -6.97090149e-01 -5.96066415e-01 1.16077699e-01
-7.67371118e-01 -9.37640816e-02 -5.40977240e-01 -2.18302280e-01
-1.16988588e-02 -9.16363895e-01 1.18124843e+00 3.60658199e-01
-1.63903743e-01 -7.88495004e-01 4.70229611e-02 4.49687958e-01
2.07598612e-01 3.10465932e-01 8.88280988e-01 -1.77027121e-01
-1.05024123e+00 1.23815127e-01 -3.84819359e-01 1.55332908e-01
4.70473558e-01 1.86368674e-01 -6.76849723e-01 -3.36375356e-01
-1.52525902e-01 -3.51881593e-01 9.65285480e-01 3.32854986e-01
1.16743183e+00 -7.29628623e-01 -1.94778740e-01 8.66674662e-01
1.36148763e+00 1.28985330e-01 6.18117929e-01 -1.57772154e-01
1.00839555e+00 2.66849488e-01 2.28002742e-01 4.78169113e-01
4.55667377e-01 5.94998240e-01 3.72463971e-01 -1.67364851e-01
-4.71660942e-01 -6.09400928e-01 2.31652893e-02 7.87605524e-01
1.12375561e-02 -2.47394368e-01 -6.45778894e-01 5.70109248e-01
-1.61085463e+00 -1.03085375e+00 3.07529479e-01 1.68436170e+00
8.32699776e-01 -6.40187189e-02 2.19643310e-01 -3.01715225e-01
9.52324331e-01 3.22986506e-02 -6.02976680e-01 4.38341424e-02
3.72468866e-02 7.57450461e-02 4.41475123e-01 4.97255266e-01
-1.05427527e+00 1.06900859e+00 6.60756683e+00 8.02724063e-01
-9.48461533e-01 4.36815768e-02 7.47997761e-01 6.40200749e-02
-7.12940931e-01 1.94511980e-01 -7.06448317e-01 5.92540920e-01
4.21210557e-01 -1.74336031e-01 8.19784284e-01 6.77848816e-01
-2.11080864e-01 3.47132295e-01 -1.07520354e+00 8.53372574e-01
-2.98034698e-02 -1.47480905e+00 3.73358428e-01 2.90313601e-01
1.14759445e+00 -1.36428639e-01 1.96583048e-01 -7.72197684e-03
1.28002608e+00 -1.18175185e+00 8.80380750e-01 4.45822001e-01
1.02237153e+00 -4.55913424e-01 1.54022679e-01 2.49002934e-01
-1.08449936e+00 3.20225582e-02 -2.54833192e-01 2.22099163e-02
-9.40464959e-02 5.96335411e-01 -8.52402568e-01 2.10631445e-01
6.75736547e-01 7.83393800e-01 -7.40371346e-01 7.88789928e-01
-6.41336322e-01 6.83731854e-01 -5.01032099e-02 3.96683067e-01
-1.70460433e-01 -2.65169851e-02 2.80641168e-01 8.85084569e-01
3.31734955e-01 2.99555719e-01 3.13851833e-01 1.31797445e+00
-2.47787029e-01 -5.48542321e-01 -4.32771713e-01 -3.23024422e-01
5.30008614e-01 1.34797180e+00 -8.95203829e-01 -4.24962729e-01
-1.16542988e-01 1.01815784e+00 2.88678735e-01 3.28107983e-01
-8.35244000e-01 -1.61697179e-01 8.11292350e-01 1.71086192e-01
6.25745237e-01 -2.11285427e-01 -2.65076160e-01 -1.52741623e+00
-2.89701790e-01 -1.09122491e+00 2.89780647e-01 -8.94648433e-01
-1.49964547e+00 6.46764457e-01 -5.95057160e-02 -1.18099380e+00
-2.69311219e-01 -3.24455142e-01 -3.84079784e-01 5.29079974e-01
-9.65250373e-01 -1.51757443e+00 -5.52059829e-01 7.42279708e-01
5.53881705e-01 -1.92581445e-01 7.30724871e-01 7.88750201e-02
-2.96913743e-01 4.92749423e-01 -1.08918868e-01 3.27117354e-01
4.28664148e-01 -1.30630088e+00 5.65755963e-01 8.23454916e-01
4.91148233e-01 5.39778829e-01 7.82039762e-01 -6.22846603e-01
-1.01033461e+00 -1.31550634e+00 3.42228860e-01 -6.35819674e-01
5.41826069e-01 -6.62159204e-01 -3.72072816e-01 8.28947604e-01
2.81835198e-01 1.56993836e-01 6.34544492e-01 -1.09392107e-01
-8.40760171e-01 -8.34100600e-03 -1.09255314e+00 4.92290378e-01
1.10166168e+00 -3.95608515e-01 -4.87413168e-01 2.35025316e-01
9.42280173e-01 -1.67563871e-01 -7.85971105e-01 4.10092086e-01
5.39850771e-01 -9.64662015e-01 9.66680229e-01 -4.04667109e-01
8.32705319e-01 -6.87175393e-01 -3.16949069e-01 -1.53890002e+00
-8.08660984e-01 -2.75076121e-01 -4.92542759e-02 1.23168933e+00
2.60310382e-01 -6.13783598e-01 8.66134048e-01 2.53881812e-01
1.65156513e-01 -3.23270202e-01 -3.28468859e-01 -4.77682978e-01
-4.00049128e-02 6.83001727e-02 1.07078302e+00 9.80993152e-01
-6.18830085e-01 6.17678285e-01 -5.87605357e-01 2.78123289e-01
1.05582893e+00 6.78290904e-01 9.16879177e-01 -1.29611492e+00
-7.37305164e-01 -3.47616583e-01 -5.35560131e-01 -1.01544440e+00
1.05170660e-01 -8.38232458e-01 2.17520908e-01 -1.44668531e+00
7.47545123e-01 -7.02462614e-01 -2.29181960e-01 6.31488144e-01
-3.18460763e-02 7.76796639e-01 3.98515105e-01 3.10004264e-01
-6.63506567e-01 5.10464430e-01 1.56630182e+00 -4.51577753e-01
1.15708180e-01 -2.59341635e-02 -1.32489252e+00 6.45592093e-01
6.39131725e-01 -5.45404017e-01 -4.16835815e-01 -5.56284726e-01
5.62323257e-02 5.59491813e-02 6.82108343e-01 -7.61917114e-01
-1.96252707e-02 -4.29974228e-01 7.07268417e-01 -4.28931683e-01
2.17707753e-01 -7.97465920e-01 6.80008590e-01 2.91091025e-01
-4.13331866e-01 -2.71441042e-01 -1.17386013e-01 3.60006392e-01
-2.59799808e-01 2.40046725e-01 9.54472303e-01 -4.41317916e-01
-6.28103912e-01 6.40974820e-01 -2.79482573e-01 3.47159915e-02
9.63739097e-01 -3.70829590e-02 -3.77300888e-01 -4.84344155e-01
-8.94279420e-01 1.14573531e-01 9.44754481e-01 4.65064675e-01
5.04855812e-01 -1.51616991e+00 -9.03034925e-01 5.96977174e-01
1.61242574e-01 1.40814558e-01 2.22923577e-01 1.42659351e-01
-2.47385919e-01 1.94901690e-01 -5.73840320e-01 -6.51712537e-01
-1.12835860e+00 6.39303327e-01 1.37767017e-01 -2.44442329e-01
-2.86171347e-01 9.65085566e-01 9.55964923e-01 -5.77266574e-01
-1.81947485e-01 3.29017863e-02 -2.71988809e-02 -2.98783720e-01
2.76986390e-01 7.79973855e-03 -3.63263398e-01 -6.40207052e-01
-1.33338392e-01 6.89706326e-01 1.22597851e-02 3.26633044e-02
1.43092752e+00 -2.39529535e-01 -2.59225160e-01 1.01864487e-01
1.09446394e+00 1.93382815e-01 -1.55742490e+00 -9.08170119e-02
-3.11538786e-01 -7.10414827e-01 -3.09605539e-01 -7.20316470e-01
-1.29205847e+00 5.73996603e-01 5.73907197e-01 2.61982560e-01
1.29243004e+00 5.94625890e-01 3.74753177e-01 -1.42181188e-01
4.99852866e-01 -2.45326132e-01 4.24988389e-01 3.80744845e-01
8.21547568e-01 -1.09146166e+00 -8.84505659e-02 -5.31329155e-01
-6.34243071e-01 6.33610368e-01 7.03410625e-01 -3.59136552e-01
6.84627473e-01 3.77813280e-01 8.35798215e-03 -3.99894744e-01
-8.74218762e-01 -3.31728429e-01 4.27190751e-01 7.33306527e-01
1.86733752e-01 5.68838120e-01 3.75878721e-01 4.85041410e-01
-4.91060793e-01 2.42690407e-02 3.95595342e-01 4.16960001e-01
-2.97590226e-01 -1.14421368e+00 -3.45673710e-01 5.71579754e-01
-2.65603364e-01 -8.62780362e-02 -3.71138692e-01 3.53405893e-01
3.93645644e-01 6.86235964e-01 2.25344285e-01 -5.59832752e-01
-1.10802157e-02 -4.29792374e-01 7.35509157e-01 -8.92210066e-01
-4.70037619e-03 4.66692895e-02 -2.33352870e-01 -3.45913172e-01
-4.82715577e-01 -5.32754421e-01 -7.52926350e-01 -2.83173651e-01
-1.35743067e-01 1.79087579e-01 2.77222663e-01 8.27013195e-01
4.12492275e-01 5.36371827e-01 8.05430472e-01 -9.68061149e-01
-2.09188506e-01 -6.04835033e-01 -6.08254492e-01 5.68815112e-01
4.86057669e-01 -7.24251688e-01 -4.76579547e-01 5.65772772e-01] | [11.668497085571289, -0.3561124801635742] |
05c66244-6cc8-4c17-ab5c-8468ebdbb3e7 | 3d-reconstruction-of-multiple-objects-by | 2211.02150 | null | https://arxiv.org/abs/2211.02150v1 | https://arxiv.org/pdf/2211.02150v1.pdf | 3D Reconstruction of Multiple Objects by mmWave Radar on UAV | In this paper, we explore the feasibility of utilizing a mmWave radar sensor installed on a UAV to reconstruct the 3D shapes of multiple objects in a space. The UAV hovers at various locations in the space, and its onboard radar senor collects raw radar data via scanning the space with Synthetic Aperture Radar (SAR) operation. The radar data is sent to a deep neural network model, which outputs the point cloud reconstruction of the multiple objects in the space. We evaluate two different models. Model 1 is our recently proposed 3DRIMR/R2P model, and Model 2 is formed by adding a segmentation stage in the processing pipeline of Model 1. Our experiments have demonstrated that both models are promising in solving the multiple object reconstruction problem. We also show that Model 2, despite producing denser and smoother point clouds, can lead to higher reconstruction loss or even loss of objects. In addition, we find that both models are robust to the highly noisy radar data obtained by unstable SAR operation due to the instability or vibration of a small UAV hovering at its intended scanning point. Our exploratory study has shown a promising direction of applying mmWave radar sensing in 3D object reconstruction. | ['Xiaohui Liang', 'Honggang Zhang', 'Zhuoming Huang', 'Yue Sun'] | 2022-11-03 | null | null | null | null | ['point-cloud-reconstruction', '3d-object-reconstruction', 'object-reconstruction'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 4.00413215e-01 -1.35880768e-01 5.93559921e-01 -2.05255076e-01
-6.56052589e-01 -5.55199683e-01 3.77453804e-01 -5.43958247e-01
-1.02300934e-01 3.63012969e-01 -2.27813706e-01 -3.79114449e-01
-4.66658711e-01 -9.86566484e-01 -6.35274947e-01 -5.60922027e-01
-1.52411968e-01 7.92641878e-01 2.28276495e-02 -2.52754778e-01
1.93963826e-01 1.24988496e+00 -1.42111421e+00 -1.00884596e-02
5.59398293e-01 1.35482776e+00 4.27273899e-01 5.82860470e-01
4.32797313e-01 2.68434405e-01 -5.81513464e-01 1.28289416e-01
1.17050362e+00 9.60052833e-02 -1.80536792e-01 1.65661573e-01
7.19851136e-01 -7.06858695e-01 -2.93339431e-01 9.46052313e-01
3.13882768e-01 5.67213111e-02 4.43179220e-01 -7.62349010e-01
-2.65657455e-01 1.38831064e-01 -7.45915651e-01 5.34599870e-02
8.63066986e-02 2.98018187e-01 7.41127789e-01 -8.73595595e-01
3.07533115e-01 9.00717020e-01 9.49785709e-01 -4.84466963e-02
-1.02204335e+00 -5.70239067e-01 -3.40081662e-01 -1.93960220e-01
-1.62367415e+00 -5.53517342e-01 6.43104076e-01 -4.30381745e-01
9.61729050e-01 4.13343936e-01 7.42690206e-01 6.46995664e-01
6.03170753e-01 3.33326876e-01 7.42253244e-01 -2.99798608e-01
2.76645035e-01 -4.08147454e-01 8.91182646e-02 4.78604078e-01
5.97077429e-01 7.44134307e-01 -1.01829223e-01 -1.66851729e-01
9.90831852e-01 2.47506127e-01 -5.60885489e-01 -1.67430729e-01
-1.11821508e+00 6.88944042e-01 6.28375649e-01 1.71124056e-01
-8.35826278e-01 1.65669903e-01 -5.17341673e-01 2.73919344e-01
4.09825146e-01 8.41686666e-01 -3.49803656e-01 3.33887726e-01
-1.44967771e+00 5.21439314e-01 6.08036041e-01 7.42605448e-01
6.80954814e-01 4.80744183e-01 4.02006298e-01 2.78304398e-01
5.73585033e-01 8.85302544e-01 -8.73500928e-02 -8.60012889e-01
4.69591349e-01 2.71972209e-01 3.58981550e-01 -1.11709356e+00
-7.17968464e-01 -8.29143465e-01 -6.23694360e-01 6.55226409e-01
-8.10385719e-02 -4.06572819e-01 -1.37521279e+00 9.97463763e-01
-1.97228827e-02 2.44245872e-01 2.35989690e-01 1.36682892e+00
7.28152931e-01 6.83867574e-01 -4.87398803e-01 -1.34643555e-01
9.39912856e-01 -4.61701512e-01 -3.17477107e-01 -8.29662621e-01
1.59815416e-01 -7.13197231e-01 2.90880919e-01 4.55913454e-01
-9.81628537e-01 -5.51119089e-01 -1.32177997e+00 5.85604250e-01
1.36602065e-02 1.92444518e-01 5.94140291e-01 2.84060031e-01
-8.95838916e-01 5.98009527e-01 -1.00231194e+00 -1.68579429e-01
5.05568922e-01 2.48686254e-01 -4.48952392e-02 -2.69724250e-01
-6.53583884e-01 1.13151956e+00 -5.91999590e-02 5.92776358e-01
-7.33262181e-01 -8.28698337e-01 -7.38242745e-01 6.98568299e-02
-2.49502603e-02 -7.89860964e-01 1.03226840e+00 -6.11321807e-01
-8.43381107e-01 4.77764159e-01 2.08423316e-01 -5.74446440e-01
1.59380034e-01 -3.86592031e-01 -5.19196212e-01 3.15804064e-01
-3.24719911e-03 3.78105640e-01 1.09437668e+00 -1.57837498e+00
-3.23875517e-01 -8.21660161e-01 -3.30486983e-01 1.36009216e-01
7.04224169e-01 -1.87475458e-01 4.22552139e-01 -5.76879680e-01
9.04513001e-01 -8.02927673e-01 -4.90238905e-01 -1.73490256e-01
-1.13589890e-01 6.59794390e-01 1.00474834e+00 -7.32826769e-01
5.37995875e-01 -1.99952900e+00 2.00096816e-02 4.29636508e-01
9.91544873e-02 2.23426968e-01 -2.18579546e-01 4.17991936e-01
-5.89163639e-02 -1.62201688e-01 -6.21193528e-01 -3.66573669e-02
-4.51749474e-01 -4.89897467e-02 -7.77571321e-01 5.17050266e-01
9.82386842e-02 9.01479423e-01 -1.32843107e-01 3.79684001e-01
1.26955705e-02 2.86248237e-01 -1.06601268e-01 5.42923100e-02
-3.93042713e-01 3.56682360e-01 -4.59339738e-01 1.27394855e+00
1.35097218e+00 8.45561251e-02 -1.76374033e-01 -3.44011694e-01
-4.06477481e-01 -2.61261277e-02 -1.11466849e+00 1.37053704e+00
-1.39303714e-01 6.87294781e-01 7.19189405e-01 -9.90858793e-01
1.13753057e+00 8.05920735e-02 3.98826569e-01 -7.16231942e-01
1.05657235e-01 6.83867633e-02 5.62237538e-02 -5.65960824e-01
8.27589452e-01 -7.45938480e-01 1.50061235e-01 4.40251708e-01
-4.30206299e-01 -6.50261939e-01 -5.95831096e-01 -2.25146949e-01
1.26007915e+00 -1.46911209e-02 -1.55138159e-02 -7.90779851e-03
-2.67958522e-01 6.00763083e-01 4.27996784e-01 7.86060393e-01
9.04428363e-02 9.60231900e-01 -4.39322382e-01 -6.03539526e-01
-7.37960696e-01 -1.26612175e+00 -2.03251123e-01 1.92709684e-01
2.85709769e-01 1.62427291e-01 -3.29639427e-02 -1.91106692e-01
1.35785133e-01 9.09693480e-01 -3.18467952e-02 2.76814271e-02
-5.41351974e-01 -9.78506684e-01 3.10576171e-01 3.85761201e-01
5.30235887e-01 -8.43969643e-01 -1.40126622e+00 1.99194700e-01
-1.71363160e-01 -1.12420738e+00 3.38197678e-01 2.25694373e-01
-1.35274470e+00 -9.66620028e-01 -2.71153629e-01 -3.75833064e-01
5.81458807e-01 8.97636950e-01 9.77154374e-01 -5.12902699e-02
-4.24332172e-01 5.54652214e-01 -6.03942573e-01 -7.45309353e-01
3.93508822e-02 -5.20471454e-01 2.30399475e-01 -1.49235874e-01
2.29783460e-01 -7.51381695e-01 -3.43741059e-01 2.36525297e-01
-8.38806927e-01 -2.24301472e-01 9.13846493e-01 2.24796981e-01
4.63874608e-01 4.72335875e-01 1.74015135e-01 -3.30094904e-01
2.33341470e-01 -3.92579287e-01 -9.40637290e-01 -2.20355883e-01
-2.32200339e-01 -2.74276257e-01 3.07462782e-01 8.79848897e-02
-7.58388162e-01 2.94486791e-01 -4.07438315e-02 -7.39523470e-01
-4.73458588e-01 8.50803494e-01 -1.59734190e-02 -3.94566059e-01
6.21806264e-01 1.65577784e-01 2.16179788e-01 -5.39239347e-01
-9.68211666e-02 4.92799312e-01 5.00467002e-01 1.68053985e-01
1.37190270e+00 7.77275562e-01 2.14138970e-01 -1.33168125e+00
-8.19474995e-01 -2.87401855e-01 -7.79318213e-01 -3.12966108e-01
7.18548656e-01 -1.03934813e+00 8.14042811e-04 2.84345478e-01
-1.21999776e+00 -3.29784185e-01 -2.86644906e-01 8.61685097e-01
-4.14948553e-01 1.41839638e-01 -1.20610788e-01 -8.54662418e-01
-4.42911774e-01 -8.63978803e-01 1.22514498e+00 1.48190528e-01
-3.37539576e-02 -5.10611475e-01 8.74332264e-02 6.05529726e-01
5.02010405e-01 4.47404236e-01 5.87367237e-01 -2.29546726e-01
-1.23035276e+00 -4.36099976e-01 -1.35319829e-01 1.01156026e-01
-3.67215425e-02 -2.20741779e-01 -7.46961772e-01 -5.00897408e-01
6.07733965e-01 4.51315008e-03 1.03773046e+00 8.95312071e-01
6.99413478e-01 -1.20210849e-01 -3.78670037e-01 1.00616133e+00
1.85467422e+00 3.34202021e-01 7.04813302e-01 3.07725728e-01
2.99775690e-01 4.55206335e-01 1.06812751e+00 3.87837052e-01
-1.95213988e-01 4.36747462e-01 1.03129733e+00 -7.76300505e-02
1.31828442e-01 2.32216328e-01 3.92844170e-01 2.65144020e-01
-3.04977298e-01 -5.89546748e-02 -9.23915148e-01 2.86755264e-01
-1.43389308e+00 -1.05039644e+00 -2.58607268e-01 2.08163238e+00
-2.09128469e-01 -1.57845188e-02 -4.69395995e-01 -1.31708577e-01
2.95503110e-01 4.01465356e-01 -4.27113235e-01 -3.28227617e-02
-1.03749283e-01 4.43808287e-01 8.34591448e-01 5.37565708e-01
-9.68798399e-01 7.60210752e-01 6.74985886e+00 8.96522775e-02
-1.26531923e+00 -2.34129086e-01 -1.61056206e-01 -2.84666717e-01
-1.60113260e-01 1.01056010e-01 -6.20334685e-01 1.69591065e-02
8.93802047e-01 2.71337509e-01 4.19952214e-01 5.43459058e-01
3.29395384e-01 -4.17105079e-01 -7.29163527e-01 9.39122140e-01
2.49698177e-01 -1.68797219e+00 1.44919693e-01 2.29387805e-01
3.48388404e-01 4.70008373e-01 3.85705940e-02 -1.80526868e-01
1.59540966e-01 -9.97380674e-01 7.87941158e-01 1.04501247e+00
5.91630936e-01 -7.69194424e-01 7.73389578e-01 7.60310352e-01
-9.03964400e-01 -2.12632388e-01 -4.96587396e-01 -3.76091719e-01
4.38560456e-01 7.90926814e-01 -1.24830222e+00 7.76904881e-01
7.07346916e-01 3.59178811e-01 -3.25693309e-01 1.12368250e+00
6.23963960e-02 3.99283916e-01 -4.93739069e-01 4.94152695e-01
2.53764957e-01 -6.48489714e-01 1.11169827e+00 6.02962494e-01
8.73371959e-01 6.29028976e-01 1.63660511e-01 1.14132297e+00
4.43560034e-01 -7.37773836e-01 -1.12643969e+00 -3.07842921e-02
2.57871181e-01 1.30931520e+00 -5.77330887e-01 1.47551045e-01
-2.00786814e-01 5.11595666e-01 -3.50610554e-01 3.90610337e-01
-5.40022314e-01 -2.06700876e-01 7.09750116e-01 6.05856657e-01
6.47663832e-01 -8.99889767e-01 -7.46888876e-01 -7.84428179e-01
6.33668751e-02 -5.04079163e-01 -9.54683721e-02 -1.35721922e+00
-1.09450269e+00 7.13047564e-01 -7.86083192e-02 -1.55410802e+00
-9.00866464e-02 -5.12793601e-01 -7.35004902e-01 9.37198400e-01
-1.30479050e+00 -1.32623720e+00 -5.35065651e-01 8.08087736e-02
5.11989415e-01 -4.74482447e-01 7.33689666e-01 -1.58942640e-01
-2.67357975e-01 -2.66086936e-01 -1.83791116e-01 -2.44156152e-01
1.35643020e-01 -5.79549491e-01 4.70415115e-01 1.27426267e+00
7.61560053e-02 4.23533618e-01 9.39363599e-01 -1.18981385e+00
-1.87605226e+00 -1.31126702e+00 3.48356396e-01 -4.97727245e-01
3.56299102e-01 -8.91107023e-02 -6.72449350e-01 1.05344844e+00
-1.91374138e-01 -2.28299901e-01 5.15770376e-01 -2.88212866e-01
2.75413811e-01 -2.38334220e-02 -1.49273169e+00 2.31458470e-01
6.99156165e-01 -8.19849446e-02 -1.03507328e+00 3.65555733e-01
6.45365298e-01 -3.13430816e-01 -5.59413850e-01 8.91044319e-01
4.30145681e-01 -9.50849116e-01 1.08771086e+00 -3.47564131e-01
3.81179988e-01 -4.85083103e-01 -6.82841718e-01 -1.55215287e+00
-6.22852683e-01 -1.60811290e-01 -1.03346407e-01 4.80161220e-01
3.85107160e-01 -5.61199427e-01 9.70865667e-01 2.48097375e-01
-5.00745475e-01 -4.21384543e-01 -9.85632479e-01 -8.56690466e-01
-1.95882767e-01 -6.40105247e-01 4.71188188e-01 6.76558495e-01
-6.65730774e-01 2.61945367e-01 -4.91574556e-02 1.28861594e+00
9.28415298e-01 7.92000771e-01 6.96851075e-01 -1.71119368e+00
-1.81356251e-01 2.98680421e-02 -1.53616652e-01 -1.12573040e+00
-1.24658585e-01 -8.60540152e-01 3.16204756e-01 -1.76074159e+00
-4.27405536e-01 -6.86432183e-01 4.17030156e-01 9.70362872e-02
5.80509961e-01 1.01450957e-01 2.15308368e-01 5.45887709e-01
1.82250500e-01 5.71554780e-01 1.06846929e+00 -1.03716031e-01
-3.49747628e-01 5.92586458e-01 -5.68100989e-01 8.80997658e-01
7.04011798e-01 -3.11744392e-01 1.34128198e-01 -9.99462724e-01
1.01493061e-01 4.15025204e-01 7.95790017e-01 -1.41080904e+00
4.04817313e-01 -1.88682199e-01 7.51414597e-01 -1.38328624e+00
8.43949854e-01 -1.34887087e+00 4.69704002e-01 3.90286595e-01
4.90396678e-01 5.69969229e-02 3.28473419e-01 4.62490439e-01
-7.76859298e-02 -2.70919204e-01 1.02798486e+00 -4.38509196e-01
-6.59059227e-01 3.22563887e-01 -4.78390843e-01 -6.58300161e-01
8.37727129e-01 -5.49812376e-01 -9.99570042e-02 -2.94663310e-01
-6.40228152e-01 6.69346377e-02 4.34235066e-01 1.12001620e-01
1.19690561e+00 -9.62796628e-01 -8.83851886e-01 8.31519723e-01
-2.48836190e-01 3.69307995e-01 1.45767868e-01 6.97844267e-01
-8.76442552e-01 3.83290440e-01 -2.99862623e-01 -7.32271314e-01
-1.16281474e+00 1.34004399e-01 4.53025222e-01 2.48195216e-01
-9.80705619e-01 7.84480929e-01 -2.47670487e-01 -4.11484718e-01
-1.78741544e-01 -4.30529594e-01 1.46723881e-01 2.62343362e-02
5.39029300e-01 2.44598940e-01 3.58131438e-01 -5.75393617e-01
-2.94034272e-01 6.93825603e-01 1.86935067e-01 -2.82402784e-01
1.96498680e+00 8.81648138e-02 -2.07912147e-01 2.02197671e-01
5.26763976e-01 1.74874604e-01 -1.12995708e+00 -4.49401559e-03
-3.50484312e-01 -6.05048656e-01 6.13403141e-01 -8.58524323e-01
-1.20049226e+00 5.46880543e-01 5.29697537e-01 2.35695586e-01
1.20254016e+00 6.46077991e-02 9.37604666e-01 6.12043798e-01
6.47578120e-01 -6.81365669e-01 -4.03408438e-01 7.37112522e-01
1.17352653e+00 -8.55085671e-01 3.68545383e-01 -1.07361928e-01
-4.40267473e-01 1.22450721e+00 3.15460116e-01 -3.24119627e-01
7.05091536e-01 4.68181074e-01 9.74012241e-02 -8.04740310e-01
-4.39797074e-01 -1.23844929e-01 6.03122003e-02 6.69394135e-01
-3.60715955e-01 1.80002987e-01 3.97797137e-01 6.13501370e-01
-5.72010636e-01 -1.38536870e-01 7.23083496e-01 1.10444605e+00
-1.07264483e+00 -4.21643913e-01 -9.76853013e-01 7.16511071e-01
2.13077277e-01 -7.64757544e-02 -4.68116373e-01 6.86687171e-01
7.78492317e-02 8.34819376e-01 4.89160746e-01 -5.52764893e-01
5.89307606e-01 -2.47758359e-01 5.56706369e-01 -6.48989618e-01
-2.93568939e-01 -2.27245055e-02 1.82893705e-02 -5.17538071e-01
-1.52755052e-01 -7.98338592e-01 -1.19402397e+00 7.67887291e-03
-3.33112061e-01 -4.43263166e-02 8.82010043e-01 9.69943643e-01
3.67412806e-01 3.43257815e-01 7.03032017e-01 -1.27396870e+00
-7.32745051e-01 -9.03172493e-01 -9.37158644e-01 -5.78544796e-01
6.98965132e-01 -6.53770924e-01 -5.86569309e-01 -2.66685516e-01] | [9.799431800842285, -2.2252862453460693] |
9ef50cac-4173-462f-8fac-94566a45f370 | hierarchical-interactive-reconstruction | 2304.07473 | null | https://arxiv.org/abs/2304.07473v1 | https://arxiv.org/pdf/2304.07473v1.pdf | Hierarchical Interactive Reconstruction Network For Video Compressive Sensing | Deep network-based image and video Compressive Sensing(CS) has attracted increasing attentions in recent years. However, in the existing deep network-based CS methods, a simple stacked convolutional network is usually adopted, which not only weakens the perception of rich contextual prior knowledge, but also limits the exploration of the correlations between temporal video frames. In this paper, we propose a novel Hierarchical InTeractive Video CS Reconstruction Network(HIT-VCSNet), which can cooperatively exploit the deep priors in both spatial and temporal domains to improve the reconstruction quality. Specifically, in the spatial domain, a novel hierarchical structure is designed, which can hierarchically extract deep features from keyframes and non-keyframes. In the temporal domain, a novel hierarchical interaction mechanism is proposed, which can cooperatively learn the correlations among different frames in the multiscale space. Extensive experiments manifest that the proposed HIT-VCSNet outperforms the existing state-of-the-art video and image CS methods in a large margin. | ['Feng Jiang', 'Chen Hui', 'Wenxue Cui', 'Tong Zhang'] | 2023-04-15 | null | null | null | null | ['video-compressive-sensing', 'compressive-sensing'] | ['computer-vision', 'computer-vision'] | [ 3.84906754e-02 -4.58210111e-01 -4.96801026e-02 -1.05814651e-01
-3.15723330e-01 1.41016999e-03 2.66253620e-01 -2.77538240e-01
-2.09465191e-01 2.91065156e-01 5.40142715e-01 4.59317155e-02
-2.73325801e-01 -6.00064456e-01 -6.74928129e-01 -9.38421130e-01
-1.14768885e-01 -5.59008360e-01 6.54030859e-01 -1.38653502e-01
1.59649059e-01 2.98080035e-02 -1.14517665e+00 1.94757447e-01
7.37313509e-01 1.46851015e+00 7.73541212e-01 1.22444980e-01
1.03991151e-01 1.31309545e+00 -1.06935594e-02 1.05172701e-01
2.12186381e-01 -6.07172728e-01 -1.69642508e-01 1.08616352e-01
9.74632129e-02 -6.68673098e-01 -1.18326950e+00 1.10300410e+00
3.97363305e-01 1.94447204e-01 -3.34997312e-03 -8.40087652e-01
-6.40050173e-01 7.19913006e-01 -7.24184692e-01 4.36886072e-01
2.57496744e-01 1.50747046e-01 9.23221111e-01 -8.82627487e-01
3.68972659e-01 1.34008753e+00 5.47132671e-01 6.93685710e-02
-8.60530198e-01 -1.04196346e+00 5.03306627e-01 6.16743028e-01
-1.74589801e+00 -4.42612559e-01 1.11449814e+00 -2.87626266e-01
3.18155110e-01 -2.13332344e-02 9.91007805e-01 9.47278857e-01
2.67831162e-02 1.12150455e+00 9.58486676e-01 -2.75236905e-01
1.77363113e-01 -5.79470873e-01 -4.27247763e-01 7.07975924e-01
-4.97931056e-03 3.46611589e-01 -7.88933456e-01 6.30174056e-02
1.37852550e+00 4.96811897e-01 -6.57224119e-01 -2.80404896e-01
-1.43991768e+00 6.76014304e-01 6.96251988e-01 7.21266389e-01
-5.17842352e-01 3.86676431e-01 3.04224938e-01 -1.23032006e-02
3.62817198e-01 -1.04672633e-01 -1.54839277e-01 1.84256256e-01
-1.05237198e+00 3.57902050e-02 3.05439770e-01 8.82172406e-01
8.00712705e-01 3.29153806e-01 -3.45186144e-01 6.78238750e-01
4.34214264e-01 2.70502955e-01 3.70459944e-01 -1.23819602e+00
4.39110607e-01 3.03351402e-01 -9.86565575e-02 -1.59783614e+00
-8.22321773e-02 -6.26085818e-01 -1.42056835e+00 -4.36314821e-01
-2.29642570e-01 -1.84581950e-01 -5.73934317e-01 1.72903883e+00
2.29856119e-01 9.48120415e-01 -9.52162966e-02 1.11956573e+00
8.48046124e-01 7.83792973e-01 -1.70648932e-01 -5.32233655e-01
9.42582548e-01 -8.25385571e-01 -7.74476826e-01 -1.66159824e-01
4.67195846e-02 -6.93959177e-01 5.84192157e-01 2.90496856e-01
-1.03869998e+00 -8.78990889e-01 -1.13204169e+00 8.44655335e-02
4.04456228e-01 1.15884561e-02 7.11219609e-01 8.61096010e-03
-9.95434523e-01 5.19881845e-01 -1.02445805e+00 2.81330366e-02
5.74562430e-01 1.75004914e-01 -1.06329061e-01 -6.74484491e-01
-1.35514081e+00 1.36570439e-01 5.72792709e-01 3.85454774e-01
-1.04401171e+00 -4.03223515e-01 -8.66497040e-01 3.52334917e-01
6.74254179e-01 -5.34354448e-01 8.46331477e-01 -9.37486887e-01
-1.45983589e+00 4.20588218e-02 -2.21412614e-01 -1.48938000e-01
4.37365294e-01 -1.02776200e-01 -3.53587300e-01 5.46658933e-01
5.71692064e-02 5.39893925e-01 1.10889637e+00 -1.20589626e+00
-6.16432250e-01 -1.44445956e-01 1.88917175e-01 2.22276181e-01
-5.08989751e-01 -2.28327036e-01 -9.18435574e-01 -1.14627218e+00
5.55069745e-01 -7.52727985e-01 -4.97001946e-01 5.71647808e-02
-2.69299954e-01 1.12845683e-02 8.39811265e-01 -5.84672391e-01
1.39217758e+00 -2.49885225e+00 5.32408237e-01 9.07469690e-02
4.81688499e-01 9.52487215e-02 -7.26233190e-03 3.46751034e-01
1.41003683e-01 -1.67997628e-01 -1.74535498e-01 -3.18524867e-01
-3.06441873e-01 3.29044789e-01 -1.06258117e-01 4.34323788e-01
-3.14922065e-01 5.88281453e-01 -1.04520214e+00 -7.00704873e-01
3.41280669e-01 5.93258083e-01 -8.23016167e-01 3.75211567e-01
6.36150036e-03 9.08290446e-01 -7.55610347e-01 4.53953862e-01
9.39229786e-01 -5.67446709e-01 1.77293926e-01 -5.23136318e-01
-3.06819141e-01 -1.25077844e-01 -1.19277799e+00 2.17406535e+00
-2.93732077e-01 5.93186080e-01 2.27005780e-01 -1.23442793e+00
5.81445515e-01 3.79254818e-01 7.93639481e-01 -8.19092691e-01
1.10159270e-01 1.09801069e-01 -8.43488947e-02 -5.00765681e-01
3.61303717e-01 -1.62106335e-01 2.26293385e-01 -7.25094825e-02
3.48415747e-02 2.85368711e-01 -6.87899590e-02 3.35917085e-01
1.02352226e+00 -1.96033437e-02 1.56452239e-01 -2.84359902e-01
8.24544251e-01 -5.61214387e-01 1.16436315e+00 5.86217046e-01
-8.00800025e-02 5.61549842e-01 1.91700190e-01 -2.15094358e-01
-7.21862972e-01 -8.75766516e-01 1.54572725e-01 8.62004817e-01
7.28763402e-01 -6.53663516e-01 -4.24743503e-01 -2.28409991e-01
-3.06503087e-01 1.00522012e-01 -4.71882463e-01 -2.41849110e-01
-6.34124935e-01 -3.17656487e-01 9.59437191e-02 5.02813935e-01
1.16256440e+00 -7.98984647e-01 -3.71460885e-01 4.15824592e-01
-4.61048663e-01 -1.55058205e+00 -6.33950949e-01 -2.42929012e-01
-8.71198416e-01 -9.24964547e-01 -8.54136765e-01 -9.23535049e-01
3.88700157e-01 8.84456813e-01 6.54201508e-01 4.15073067e-01
8.74783546e-02 2.76260406e-01 -8.27619255e-01 1.54763088e-01
1.61224365e-01 -1.76973149e-01 -6.12313934e-02 4.50766027e-01
7.75431469e-02 -1.01139247e+00 -1.15737522e+00 3.06060404e-01
-1.22370124e+00 4.04157758e-01 7.75264919e-01 9.37222660e-01
4.84156430e-01 3.80709529e-01 2.96143800e-01 -4.36890930e-01
1.37618735e-01 -6.17342710e-01 -4.87325549e-01 4.70323451e-02
-1.20060764e-01 -2.29206145e-01 6.80228531e-01 -4.29387838e-01
-9.75626230e-01 3.59668303e-03 -8.70729536e-02 -9.97520387e-01
1.72508329e-01 9.00684178e-01 -2.18458280e-01 -3.04702759e-01
-4.17146832e-02 8.00804317e-01 -1.90189719e-01 -4.38505232e-01
1.49780698e-02 2.92932779e-01 4.92368788e-01 -5.53082645e-01
7.24162102e-01 6.57489836e-01 9.96364951e-02 -8.42843950e-01
-9.26607072e-01 -4.12363678e-01 -6.62639380e-01 -1.43666774e-01
1.00249434e+00 -1.44989920e+00 -6.44581497e-01 6.19660437e-01
-1.17268062e+00 -2.49198988e-01 2.11656794e-01 8.53279352e-01
-2.97866285e-01 8.03650737e-01 -9.55563605e-01 -4.97597992e-01
-5.23963012e-03 -1.36842430e+00 9.45881546e-01 2.43297577e-01
3.74951154e-01 -7.93108284e-01 -3.81755292e-01 2.18557402e-01
3.53495359e-01 8.82422775e-02 5.92129052e-01 -2.23091934e-02
-1.17565072e+00 1.83760092e-01 -4.41392541e-01 5.69774747e-01
5.43071032e-02 -3.67826968e-01 -6.16531551e-01 -4.07146484e-01
2.37398997e-01 -1.87184602e-01 9.89543080e-01 6.09815538e-01
1.49583054e+00 -3.22304726e-01 -6.94025159e-02 9.05145288e-01
1.41002917e+00 3.01604092e-01 6.94477320e-01 1.97451964e-01
1.01472902e+00 2.90655375e-01 4.56351966e-01 7.30738223e-01
5.33487737e-01 4.27053869e-01 6.11328900e-01 -1.67448401e-01
2.09043667e-01 -2.85860062e-01 2.34842360e-01 1.31464136e+00
-1.43459573e-01 -2.26828083e-02 -3.25310677e-01 4.04997766e-01
-2.09960079e+00 -9.87117350e-01 2.50899822e-01 1.79883695e+00
6.41918421e-01 5.68256201e-03 -3.28111380e-01 4.53472026e-02
7.18681693e-01 7.64880359e-01 -5.57473958e-01 7.39051163e-01
-1.97685391e-01 5.81592359e-02 3.26913565e-01 2.54312366e-01
-1.05881798e+00 8.54509652e-01 5.32574177e+00 1.07567191e+00
-1.15489542e+00 2.69636840e-01 5.56769550e-01 5.98725453e-02
-2.66117454e-01 1.96242034e-01 -2.57287741e-01 7.44559884e-01
1.28193736e-01 1.10652588e-01 4.80290651e-01 6.24441206e-01
2.45900720e-01 -6.43970259e-03 -8.09186399e-01 1.40719283e+00
-6.34869421e-03 -1.63669336e+00 -1.60289351e-02 8.92819613e-02
8.29882503e-01 -6.50326908e-03 7.28641525e-02 1.75243497e-01
2.22940549e-01 -7.85957992e-01 7.82656550e-01 5.53355753e-01
6.58829391e-01 -8.60594928e-01 6.38286710e-01 3.46318036e-01
-1.66661620e+00 -2.22893864e-01 -4.05667484e-01 -9.78420898e-02
1.25867471e-01 6.51785016e-01 3.09101731e-01 7.80884385e-01
1.00455976e+00 1.39168239e+00 -1.97518796e-01 9.00469184e-01
-1.76787153e-01 6.23759627e-01 -1.10787593e-01 3.33931237e-01
6.02667451e-01 -2.81731755e-01 4.19507921e-01 9.27265525e-01
6.46354318e-01 7.67212510e-01 7.45982289e-01 6.77854717e-01
-1.94144249e-03 -1.21591575e-01 -2.73183286e-01 9.80835557e-02
6.63418770e-01 1.05308783e+00 -4.08134967e-01 -5.02734721e-01
-6.58392549e-01 1.11233068e+00 -1.08363472e-01 5.53636491e-01
-5.64745784e-01 -1.16423309e-01 4.67930436e-01 4.68720868e-02
8.70313048e-01 -5.20856142e-01 2.64853448e-01 -1.59340644e+00
-8.05842923e-04 -8.25377047e-01 3.29002351e-01 -7.41165161e-01
-1.10815287e+00 4.30740237e-01 -1.29364192e-01 -1.53383172e+00
1.04462810e-01 -1.55809090e-01 -6.39451087e-01 6.19988561e-01
-1.64731646e+00 -1.09390330e+00 -6.81887627e-01 1.14153290e+00
4.73828256e-01 -1.68325394e-01 2.08780661e-01 5.54007649e-01
-5.92721641e-01 3.72932464e-01 2.22465932e-01 3.05024147e-01
4.26567227e-01 -5.69111943e-01 -1.68590263e-01 1.10634840e+00
-1.61565304e-01 7.61820853e-01 5.06000221e-01 -5.11703014e-01
-1.57759631e+00 -1.04499626e+00 1.30480394e-01 5.80331147e-01
7.18784630e-01 -1.57152116e-01 -9.48076785e-01 5.22091091e-01
2.49219954e-01 4.27961826e-01 5.93300819e-01 -1.98150426e-01
-4.89957958e-01 -4.06627983e-01 -7.08195567e-01 5.21804631e-01
1.12935913e+00 -7.36475110e-01 -1.82902515e-01 1.97047710e-01
1.07041860e+00 -3.31303388e-01 -8.57789159e-01 6.08774304e-01
4.60447848e-01 -1.43041325e+00 1.08824944e+00 5.38387187e-02
7.65073538e-01 -4.82198387e-01 -6.55949831e-01 -1.14855826e+00
-6.32409751e-01 -6.28884315e-01 -3.00479710e-01 9.81230497e-01
-4.42934245e-01 -4.43356276e-01 6.71792865e-01 -4.46178988e-02
-3.00315380e-01 -8.24631751e-01 -9.79050040e-01 -5.04326522e-01
-3.34182739e-01 -4.71506149e-01 5.12071669e-01 1.08657825e+00
-3.49420696e-01 2.52911687e-01 -7.55647123e-01 4.57483292e-01
7.79815674e-01 2.09578276e-01 4.32644755e-01 -1.09708345e+00
-6.29934728e-01 -3.07682484e-01 -5.51613569e-01 -1.83903182e+00
5.70422783e-02 -4.83258367e-01 -3.41848657e-03 -1.30182672e+00
2.92735010e-01 -4.24198896e-01 -6.95186019e-01 4.94133048e-02
-2.07856253e-01 -2.82904562e-02 2.96686530e-01 4.70187366e-01
-9.24945414e-01 9.62569058e-01 1.52081192e+00 -2.89143641e-02
2.08306327e-01 -3.74709159e-01 -6.61253393e-01 7.59831667e-01
3.33152652e-01 -2.18914613e-01 -5.05593240e-01 -6.65778220e-01
2.57772431e-02 6.36381090e-01 5.05111396e-01 -1.27923787e+00
5.81516981e-01 -2.62966514e-01 6.03594899e-01 -6.82409644e-01
4.51309502e-01 -8.85559261e-01 -8.93109571e-03 4.64710742e-01
-1.48648307e-01 -1.27599478e-01 -1.92361161e-01 1.09690225e+00
-6.60290599e-01 1.96051374e-01 7.50092328e-01 -4.93404269e-01
-8.79721999e-01 9.75569546e-01 -2.74232984e-01 -2.32460976e-01
6.58868372e-01 -4.19457927e-02 2.36331429e-02 -5.61923265e-01
-4.55039173e-01 3.25688571e-01 3.00066203e-01 1.95321009e-01
8.94862771e-01 -1.48983335e+00 -5.16210437e-01 2.47322589e-01
-1.55016128e-02 3.84360850e-01 9.72464800e-01 9.78663325e-01
-4.63675946e-01 2.23293424e-01 -1.10608302e-01 -8.40717196e-01
-9.08350110e-01 6.16863012e-01 -1.06996449e-03 -1.36658862e-01
-6.90551281e-01 9.46629822e-01 7.04878092e-01 1.41871214e-01
2.51327068e-01 -1.41647324e-01 -1.86933532e-01 -2.38216415e-01
6.35536253e-01 8.52042437e-02 -5.32219887e-01 -8.05646122e-01
-8.82550329e-02 6.63020670e-01 -9.66084599e-02 5.30687384e-02
1.43238366e+00 -4.16445076e-01 -3.22408348e-01 2.55595684e-01
1.35798669e+00 -1.12734117e-01 -1.61553657e+00 -8.08751345e-01
-2.17789829e-01 -8.87100875e-01 5.69049716e-01 -1.21097779e-03
-1.61130571e+00 9.09588277e-01 4.26219702e-01 4.83803339e-02
1.47021759e+00 -1.92022637e-01 1.21614122e+00 2.22328559e-01
4.98990029e-01 -7.68275797e-01 5.54669261e-01 5.71675777e-01
6.71997666e-01 -1.17215991e+00 1.30502582e-01 -5.50971389e-01
-3.93724769e-01 1.00394249e+00 5.82963109e-01 -3.36040020e-01
1.00180006e+00 -3.98274027e-02 -3.71881098e-01 -1.87006325e-01
-4.98645902e-01 -1.33337006e-01 1.32025301e-01 2.62692660e-01
3.68490428e-01 -1.35937557e-01 -7.73727894e-02 5.37931859e-01
3.45766664e-01 1.12669058e-01 2.08365709e-01 9.30611253e-01
-4.55020279e-01 -8.60527396e-01 -1.90943703e-01 2.28573889e-01
-4.27295506e-01 -3.85405988e-01 1.70466945e-01 5.47469199e-01
3.52046758e-01 9.75737393e-01 -4.01757509e-02 -7.95377195e-01
-2.76459120e-02 -8.32265377e-01 3.43272269e-01 -3.02092284e-01
-9.09134969e-02 7.86823869e-01 -3.66916835e-01 -7.54060507e-01
-8.30096126e-01 -6.20595694e-01 -1.09435248e+00 -3.00126433e-01
-2.30123609e-01 1.51469484e-01 8.62364322e-02 1.06481862e+00
2.73335516e-01 6.28669798e-01 9.14455414e-01 -1.16938913e+00
-8.88434723e-02 -9.09262300e-01 -8.12248588e-01 8.23856294e-02
5.75438678e-01 -7.32339144e-01 -3.50344837e-01 -5.42278960e-02] | [11.116991996765137, -1.8912874460220337] |
392f59d3-72ae-447c-ba38-a5441997dca1 | alphastock-a-buying-winners-and-selling | 1908.02646 | null | https://arxiv.org/abs/1908.02646v1 | https://arxiv.org/pdf/1908.02646v1.pdf | AlphaStock: A Buying-Winners-and-Selling-Losers Investment Strategy using Interpretable Deep Reinforcement Attention Networks | Recent years have witnessed the successful marriage of finance innovations and AI techniques in various finance applications including quantitative trading (QT). Despite great research efforts devoted to leveraging deep learning (DL) methods for building better QT strategies, existing studies still face serious challenges especially from the side of finance, such as the balance of risk and return, the resistance to extreme loss, and the interpretability of strategies, which limit the application of DL-based strategies in real-life financial markets. In this work, we propose AlphaStock, a novel reinforcement learning (RL) based investment strategy enhanced by interpretable deep attention networks, to address the above challenges. Our main contributions are summarized as follows: i) We integrate deep attention networks with a Sharpe ratio-oriented reinforcement learning framework to achieve a risk-return balanced investment strategy; ii) We suggest modeling interrelationships among assets to avoid selection bias and develop a cross-asset attention mechanism; iii) To our best knowledge, this work is among the first to offer an interpretable investment strategy using deep reinforcement learning models. The experiments on long-periodic U.S. and Chinese markets demonstrate the effectiveness and robustness of AlphaStock over diverse market states. It turns out that AlphaStock tends to select the stocks as winners with high long-term growth, low volatility, high intrinsic value, and being undervalued recently. | ['Yang Zhang', 'Ke Tang', 'Junjie Wu', 'Jingyuan Wang', 'Zhang Xiong'] | 2019-07-24 | null | null | null | null | ['deep-attention', 'deep-attention'] | ['computer-vision', 'natural-language-processing'] | [-9.21655536e-01 -1.13664746e-01 -2.81210423e-01 -1.92569986e-01
-2.88404197e-01 -4.52328116e-01 4.36687350e-01 -2.24920303e-01
-2.97325701e-01 8.62987339e-01 1.39989227e-01 -4.91342038e-01
-3.79222125e-01 -1.08899009e+00 -4.20741618e-01 -4.85530943e-01
-1.33140236e-01 6.17598355e-01 -4.63309325e-02 -5.89427531e-01
5.08092284e-01 4.57704335e-01 -9.28584874e-01 -9.74195600e-02
7.61459351e-01 1.50761294e+00 -4.15335894e-01 1.42719492e-01
-2.19531924e-01 1.32694435e+00 -6.81548417e-01 -8.61999929e-01
7.85951912e-01 -4.54130828e-01 -3.93616617e-01 -1.78288966e-01
-3.65101695e-01 -8.52369606e-01 -3.61516088e-01 9.68648195e-01
4.69207883e-01 -1.23460099e-01 5.61137617e-01 -1.25292075e+00
-1.33818579e+00 1.10545659e+00 -8.36889148e-01 6.05256021e-01
-4.49025542e-01 3.52637142e-01 1.54028893e+00 -7.43063927e-01
3.80781293e-02 1.24749660e+00 5.48793793e-01 4.55654144e-01
-7.65605152e-01 -8.00865352e-01 3.68181258e-01 4.42248918e-02
-5.59968889e-01 6.56141639e-02 8.33174229e-01 -3.06197107e-01
1.02491820e+00 -5.10736778e-02 9.18562651e-01 6.49137080e-01
5.20563543e-01 9.30094302e-01 1.02212656e+00 -9.09001008e-02
2.07754761e-01 -1.02376919e-02 4.57644165e-02 2.26060644e-01
4.26249146e-01 6.07872844e-01 -2.50783712e-01 1.55427039e-01
1.24880826e+00 2.95134485e-01 2.30460227e-01 -9.33267251e-02
-9.19757843e-01 1.37538052e+00 7.03479409e-01 3.84207666e-01
-6.60149872e-01 4.17342365e-01 3.54917139e-01 6.79552376e-01
6.02665246e-01 7.62951255e-01 -6.15498662e-01 -1.24171138e-01
-6.61910415e-01 5.16240835e-01 6.61907554e-01 6.15405321e-01
4.02626604e-01 7.89419770e-01 -2.05518812e-01 4.49643612e-01
4.08826172e-01 5.64677656e-01 7.57031381e-01 -1.05214095e+00
6.34217978e-01 4.88420904e-01 1.66376293e-01 -9.83103275e-01
-2.72459626e-01 -7.91770518e-01 -6.75759554e-01 4.80179936e-01
2.08637550e-01 -4.33471739e-01 -3.23057532e-01 1.55362046e+00
-1.70581877e-01 -1.13297895e-01 2.23619416e-01 7.70213127e-01
1.77460611e-01 6.30508661e-01 -5.45567125e-02 -2.14113697e-01
1.06101668e+00 -1.09617090e+00 -6.67404056e-01 -3.54943387e-02
6.85676932e-02 -2.55692303e-01 8.36249590e-01 2.72125781e-01
-1.31887043e+00 -4.86095399e-01 -1.19413650e+00 3.60129297e-01
-2.42437541e-01 -2.43478090e-01 9.86692429e-01 4.90366131e-01
-8.21216643e-01 8.48892450e-01 -7.27644682e-01 6.05027020e-01
7.45934069e-01 4.45249230e-01 3.99802834e-01 8.38792205e-01
-1.58130836e+00 7.99106419e-01 3.11762869e-01 1.96527690e-01
-8.17837596e-01 -6.75033927e-01 -3.94993871e-01 3.90711755e-01
6.07937872e-01 -4.92838293e-01 1.53683138e+00 -1.49575365e+00
-1.81930900e+00 3.41172546e-01 8.84144306e-01 -1.26100504e+00
8.03215742e-01 -6.13558888e-01 -1.96307540e-01 9.29467976e-02
2.73866626e-03 3.21550906e-01 7.20409513e-01 -6.46277785e-01
-6.46313429e-01 -2.50363916e-01 1.79160431e-01 2.55987465e-01
-5.05268097e-01 1.09195970e-02 2.09333643e-01 -1.32314765e+00
-3.89574051e-01 -5.72606444e-01 -1.05905809e-01 -2.18401194e-01
5.52199781e-02 -3.59268159e-01 2.97224492e-01 -6.09283388e-01
1.26322794e+00 -1.88328362e+00 -4.62229215e-02 -5.85112758e-02
2.93993056e-01 3.86310160e-01 6.60234168e-02 3.66054893e-01
-1.70590967e-01 3.44624639e-01 -1.49015561e-01 5.36555685e-02
3.78254980e-01 2.51954906e-02 -9.32162881e-01 1.41289413e-01
6.10200763e-01 1.45297980e+00 -7.98538923e-01 -5.98396808e-02
3.52817625e-02 -1.73544332e-01 -6.00618482e-01 4.51146871e-01
-4.30924833e-01 1.54485747e-01 -7.86324799e-01 7.74174333e-01
5.18296480e-01 -1.32045522e-01 -6.87706769e-02 2.64204592e-01
-2.39092916e-01 1.43769458e-01 -9.35184240e-01 5.84006131e-01
-2.42597118e-01 3.88376772e-01 -3.34225953e-01 -1.17580175e+00
1.24051821e+00 1.76195577e-01 4.57855850e-01 -1.21625912e+00
1.94684714e-01 5.64255118e-01 1.55069664e-01 -1.40639454e-01
3.53802115e-01 -7.31465518e-01 -6.47823513e-02 9.12393093e-01
-2.37173870e-01 -4.54956777e-02 4.33027335e-02 -1.28772542e-01
5.21695018e-01 1.96666270e-01 2.92938977e-01 -3.39816421e-01
3.08705777e-01 -3.96720260e-01 7.39200234e-01 6.66107059e-01
-3.84964257e-01 3.15892518e-01 9.28523540e-01 -8.36913943e-01
-1.03010476e+00 -1.10099280e+00 1.65929973e-01 9.72275794e-01
-1.44746721e-01 4.82936442e-01 -4.98340964e-01 -7.10857749e-01
5.20091474e-01 6.79054320e-01 -7.17143357e-01 -1.39546186e-01
-7.27025330e-01 -1.01297688e+00 3.80110413e-01 1.00071430e+00
9.21412587e-01 -1.58494246e+00 -7.66121566e-01 5.42286754e-01
4.28435951e-01 -4.36706722e-01 -3.76501769e-01 1.84112966e-01
-9.21346426e-01 -1.05655813e+00 -1.08557379e+00 -4.81845051e-01
1.72522962e-01 -1.47897899e-01 1.24492335e+00 7.79818296e-02
1.16019741e-01 1.68926626e-01 -1.44457087e-01 -9.01854873e-01
-1.16051592e-01 -7.53321797e-02 2.38333978e-02 -2.81249322e-02
3.22750062e-01 -3.28559607e-01 -7.47288167e-01 2.37042338e-01
-9.44388747e-01 -3.76911938e-01 5.68732917e-01 1.13987768e+00
4.18569565e-01 1.90212056e-01 1.39095485e+00 -7.90383458e-01
9.79713261e-01 -7.48494208e-01 -1.13977456e+00 3.20865542e-01
-1.13878071e+00 2.34584838e-01 7.61701703e-01 -2.77817398e-01
-1.26888657e+00 -7.80686200e-01 -1.81065015e-02 -3.13653171e-01
6.87322915e-01 6.02078378e-01 9.08739641e-02 2.27444872e-01
8.93659592e-02 2.30307758e-01 1.41406596e-01 -4.00672078e-01
2.21068978e-01 3.90067995e-01 2.72780210e-01 -5.30812144e-01
7.17201412e-01 2.54256874e-01 -1.76335573e-01 -1.29549503e-02
-8.93547893e-01 3.16157907e-01 -3.94394755e-01 2.43188515e-02
5.94543099e-01 -8.43195021e-01 -8.65493298e-01 9.34169710e-01
-6.23690128e-01 -3.53574485e-01 -5.63382983e-01 4.28063929e-01
-7.35330582e-01 2.73121625e-01 -1.05053377e+00 -1.00984144e+00
-7.37450361e-01 -8.28259110e-01 4.01259989e-01 5.74294984e-01
2.16024429e-01 -1.20164251e+00 2.29296386e-01 1.98797300e-01
6.88189268e-01 1.82689950e-01 8.51621091e-01 -8.03571522e-01
-6.71101809e-01 7.60999173e-02 -2.95871139e-01 6.83494031e-01
1.50515765e-01 -2.97104171e-03 -5.71110427e-01 -9.95226800e-02
4.18370694e-01 -5.23072064e-01 1.04160070e+00 4.05223429e-01
8.53035569e-01 -4.53752428e-01 4.54785168e-01 6.25717103e-01
1.29263806e+00 8.19641411e-01 5.33865154e-01 9.50128794e-01
2.51120001e-01 5.89603961e-01 8.04620206e-01 7.82254636e-01
2.58496821e-01 2.65845716e-01 4.85569805e-01 -3.48973237e-02
3.76614094e-01 -3.93595874e-01 5.52769840e-01 8.02090228e-01
-1.67053834e-01 -1.26379550e-01 -6.99560761e-01 3.77923310e-01
-1.94530046e+00 -1.13831127e+00 6.84389114e-01 1.81461179e+00
7.57728279e-01 7.05950320e-01 4.61523980e-01 -1.02663629e-01
4.81840044e-01 1.69625416e-01 -1.12183774e+00 -5.11622906e-01
-4.37583923e-01 1.99046507e-01 5.02325773e-01 2.02897996e-01
-1.09308314e+00 8.77902150e-01 6.15031052e+00 6.63170099e-01
-1.21607471e+00 -3.05676758e-01 1.31962609e+00 -1.15669049e-01
-6.36579275e-01 -2.14685008e-01 -8.30162644e-01 6.07260406e-01
8.78946126e-01 -6.38852000e-01 4.03455436e-01 8.88087273e-01
-3.70654427e-02 5.35227954e-01 -7.65941858e-01 5.89096487e-01
-3.51568013e-01 -1.42015076e+00 1.76733732e-01 -2.26948373e-02
6.51148438e-01 -1.50290251e-01 6.40355349e-01 6.31149888e-01
5.36076486e-01 -1.11090446e+00 1.14009917e+00 6.59017861e-01
1.87724844e-01 -1.11266422e+00 9.63832200e-01 3.82752344e-02
-1.03349984e+00 -5.68347633e-01 -5.61348677e-01 -2.76021928e-01
-4.46145982e-02 2.83303738e-01 -2.39687756e-01 5.93899608e-01
6.24601066e-01 1.08206487e+00 -3.20393234e-01 6.12265706e-01
-2.97371626e-01 5.65702200e-01 1.62712093e-02 -3.17613572e-01
6.70195997e-01 -5.71463227e-01 1.41219541e-01 6.18396282e-01
2.92323351e-01 8.67979378e-02 -1.67414457e-01 1.20487070e+00
-2.25268513e-01 -4.54386417e-03 -4.22881633e-01 -4.47569460e-01
2.19444171e-01 7.88262665e-01 -7.47244060e-01 -2.83007771e-01
-5.49471915e-01 4.28539544e-01 2.83112496e-01 2.11440429e-01
-1.01197696e+00 -3.58649582e-01 5.14232397e-01 5.02865436e-03
7.05946386e-01 -9.15732384e-02 -5.42097092e-01 -1.24641967e+00
-3.88300903e-02 -1.01727498e+00 5.47489405e-01 -3.91426325e-01
-1.70345640e+00 6.00688040e-01 -1.44404918e-01 -1.12314677e+00
-3.59446943e-01 -7.00893104e-01 -7.17532754e-01 7.38396525e-01
-1.92714119e+00 -6.78961694e-01 5.66330314e-01 3.43351722e-01
6.28131509e-01 -9.14186358e-01 3.15409988e-01 2.51669697e-02
-6.32583261e-01 5.77345908e-01 5.36946297e-01 4.19304729e-01
1.80953369e-01 -1.42342854e+00 7.00594068e-01 5.26124060e-01
2.22363234e-01 4.94005620e-01 1.61005974e-01 -6.82969689e-01
-1.04395068e+00 -7.77518809e-01 4.44151342e-01 -2.62216538e-01
1.24188519e+00 4.83598970e-02 -8.71094704e-01 6.83932126e-01
4.61863875e-01 -2.82406390e-01 3.56197655e-01 -2.31014997e-01
-2.54059106e-01 -5.75054049e-01 -9.57644939e-01 5.97258151e-01
4.08511132e-01 -2.05992386e-01 -9.15871441e-01 -8.01307186e-02
8.98833632e-01 2.87099797e-02 -8.87414992e-01 2.20868021e-01
4.94242430e-01 -1.25822234e+00 9.66265798e-01 -7.19540298e-01
4.81705099e-01 2.37345502e-01 2.47996926e-01 -1.12249982e+00
-3.84534866e-01 -8.36606979e-01 -3.48519534e-01 1.26061523e+00
2.75014549e-01 -1.10042286e+00 7.45494068e-01 4.94114220e-01
1.74874350e-01 -1.13293946e+00 -1.01874363e+00 -8.29497099e-01
7.18964398e-01 -2.65128333e-02 1.10375047e+00 6.40380323e-01
-2.54764497e-01 1.35013968e-01 -6.77086473e-01 -4.46039140e-01
6.40296161e-01 6.33508623e-01 2.03982025e-01 -1.01825154e+00
-5.77946126e-01 -9.41660881e-01 3.57578285e-02 -9.40962076e-01
8.44505578e-02 -4.13731247e-01 -5.02659738e-01 -8.63512158e-01
6.78656772e-02 -4.50785905e-01 -9.68710780e-01 4.45536196e-01
-1.49394289e-01 -2.11318642e-01 5.42000055e-01 4.20342267e-01
-4.24146980e-01 9.86320078e-01 1.26576531e+00 -2.37066343e-01
-8.69615376e-02 3.74326676e-01 -1.19216692e+00 6.22449875e-01
1.25394928e+00 -2.43559122e-01 -4.08573389e-01 -4.40461665e-01
5.56342661e-01 3.01249087e-01 1.71915051e-02 -6.13906860e-01
-2.58324891e-01 -4.83641624e-01 4.27254647e-01 -4.35552716e-01
-2.03896970e-01 -5.37062705e-01 -1.83229461e-01 9.42287385e-01
-3.85982990e-01 7.14325368e-01 8.52561370e-02 4.80244488e-01
-4.63320494e-01 -2.57622808e-01 7.97819674e-01 -4.56103086e-01
-5.60618162e-01 5.70190430e-01 -1.55152842e-01 4.36758548e-01
1.13523781e+00 9.79534164e-02 -2.47166425e-01 -5.35032451e-01
-2.26282179e-01 4.91137981e-01 6.80063516e-02 6.33040965e-01
6.88818395e-01 -1.31142640e+00 -8.24588835e-01 1.80233061e-01
-4.78953689e-01 -2.11215228e-01 -3.21372226e-02 2.77088791e-01
-9.38056886e-01 6.61179900e-01 -6.03785992e-01 2.01262370e-01
-3.38548809e-01 6.30854487e-01 7.54092455e-01 -7.53629923e-01
-4.76024538e-01 7.04335749e-01 4.85891074e-01 -3.83045338e-02
2.48586088e-01 -3.08878779e-01 -3.57171893e-01 3.09626669e-01
6.21729851e-01 4.51823562e-01 -2.87524968e-01 -1.21010818e-01
-4.56410609e-02 5.55606484e-01 -2.70297766e-01 -1.54962480e-01
1.85795355e+00 6.42726049e-02 6.90354928e-02 4.16355729e-01
6.40594661e-01 -2.99589992e-01 -1.76234365e+00 -8.59573558e-02
4.02526140e-01 -3.56188864e-01 -1.65490478e-01 -8.03419352e-01
-1.71154141e+00 1.07611680e+00 3.48345518e-01 4.98434812e-01
1.09912133e+00 -4.26770687e-01 1.05452335e+00 3.10718983e-01
2.96776742e-01 -1.34104276e+00 6.18908644e-01 5.07476509e-01
1.10298729e+00 -1.15332544e+00 -1.19542948e-03 5.78420401e-01
-1.07787418e+00 1.31412184e+00 6.54302001e-01 -6.16779745e-01
8.81094038e-01 1.82795167e-01 3.42211932e-01 -9.09554288e-02
-8.58336568e-01 1.29734576e-01 -2.25507431e-02 2.54325509e-01
1.85658082e-01 -1.10267811e-02 -2.08769158e-01 1.07813561e+00
-5.53723350e-02 4.43000067e-03 3.99307311e-01 7.91346252e-01
-5.06220043e-01 -1.06823575e+00 -1.50634050e-01 3.37561339e-01
-9.83219981e-01 -2.83284068e-01 -3.24773401e-01 9.63167727e-01
-2.30180755e-01 4.04303014e-01 2.80155033e-01 -2.61480417e-02
3.15269023e-01 -3.64479393e-01 1.01807371e-01 -3.45837265e-01
-1.16324103e+00 2.29454786e-01 -4.64283615e-01 -1.95055291e-01
-2.91353047e-01 -7.04404175e-01 -1.31148863e+00 -3.91714096e-01
-1.42570123e-01 2.56811827e-01 9.91693977e-03 6.58942342e-01
2.19967335e-01 4.60888833e-01 1.10640168e+00 -3.95966768e-01
-1.54440534e+00 -6.39087975e-01 -1.11093652e+00 3.43062758e-01
4.40064192e-01 -8.16582978e-01 -3.75941545e-01 -4.90332484e-01] | [4.462952613830566, 4.031271934509277] |
39c58029-50e8-47db-8fe2-7e6c86254b96 | understand-customer-behavior-and-complaints | null | null | http://web.tecnico.ulisboa.pt/~mcasquilho/CD_Casquilho/PRINT/qp0103goodman.pdf | http://web.tecnico.ulisboa.pt/~mcasquilho/CD_Casquilho/PRINT/qp0103goodman.pdf | Understand Customer Behavior And Complaints Eight areas of quantifiable data can be integrated into quality assurance decisions | USTOMER COMPLAINTS PROVIDE
valuable quality assurance, service and
marketing data. But the challenge is to
use the data to make decisions that
result in substantive action.
To use complaint data to solve problems in design, marketing, installation, distribution
and after sale use and maintenance, you
should have a basic understanding of
customer complaint and market
behavior.
This understanding will provide a framework for interpreting the data and extrapolating
it to the entire customer base.
The framework will allow
organizations not only to quantify the implications of the data
but also to set priorities and allocate scarce quality assurance resources to mitigate problems.
In fact, unsolicited complaints submitted at the
time a problem occurs are less costly than systematic sampling and inspection and provide more timely information than is typically available from
warranty data.
Eight factors about customer behavior are key to | ['Steve Newman', 'John Goodman'] | 2023-01-01 | null | null | null | qual-ity-progress-2023-1 | ['marketing'] | ['miscellaneous'] | [ 2.42493048e-01 -2.15279952e-01 -4.35804337e-01 -5.69745481e-01
-9.37034070e-01 -7.72120774e-01 -1.71543851e-01 9.51268077e-01
-3.81446272e-01 6.00342095e-01 4.64868486e-01 -1.01633239e+00
-5.16875386e-01 -8.04216385e-01 -2.06386015e-01 -2.68517256e-01
6.92820370e-01 4.18856263e-01 -2.32993156e-01 -3.20673436e-01
7.86951125e-01 9.94045734e-01 -1.25209641e+00 3.54176342e-01
5.54489195e-01 1.05355668e+00 -4.94248010e-02 3.22000474e-01
-3.50736439e-01 7.33201921e-01 -7.38474190e-01 -6.50082007e-02
1.98440745e-01 -2.62389630e-01 -9.30889726e-01 7.58796573e-01
-3.46691072e-01 -7.44771421e-01 5.66460490e-01 5.54217875e-01
2.46006101e-01 1.27755394e-02 3.38311732e-01 -1.04366958e+00
-6.40368819e-01 3.36821288e-01 -3.74154031e-01 4.61357921e-01
7.09123790e-01 3.15696478e-01 1.16098893e+00 -5.22721827e-01
2.95149505e-01 9.66957510e-01 2.31890380e-01 -2.42991194e-01
-1.41722786e+00 -3.34573716e-01 3.46606195e-01 -1.42709672e-01
-9.35336292e-01 -2.99390614e-01 5.48676848e-01 -5.91358542e-01
7.33770907e-01 5.05843639e-01 7.83150673e-01 2.95950025e-01
2.25624815e-01 5.09516180e-01 8.72264802e-01 -3.34929198e-01
2.97577202e-01 4.32845116e-01 4.92599159e-01 -3.80942166e-01
7.83140421e-01 -3.03021342e-01 -1.40071958e-01 -3.41437757e-01
3.09749663e-01 4.94583875e-01 -1.75033078e-01 7.53306895e-02
-3.00412953e-01 9.60477531e-01 2.31931526e-02 3.62793952e-01
-8.08633089e-01 -2.03397438e-01 2.53423691e-01 3.84464204e-01
1.03720672e-01 5.73172033e-01 -6.09462202e-01 -4.82797623e-01
-5.48051596e-01 4.26988214e-01 1.05888069e+00 6.93438888e-01
4.79777545e-01 7.40732253e-02 3.73173267e-01 5.64418137e-01
5.19450188e-01 6.56266034e-01 -2.79408932e-01 -9.71386254e-01
1.82712108e-01 6.70095861e-01 6.72574580e-01 -1.00387359e+00
-4.41682845e-01 -5.18799424e-01 3.84325206e-01 -3.02957073e-02
2.32933715e-01 -1.02970507e-02 -6.74145579e-01 7.48173952e-01
5.94748259e-02 -1.11437786e+00 -1.12222187e-01 8.35902631e-01
2.19360352e-01 4.37016696e-01 7.71643594e-02 -3.67009670e-01
1.11168075e+00 9.88937095e-02 -9.17967796e-01 -1.36974379e-01
7.23368943e-01 -1.24032438e+00 1.08442068e+00 6.92921877e-01
-1.06988406e+00 -5.00024222e-02 -7.04556882e-01 3.50105137e-01
-2.18563318e-01 -4.06960011e-01 6.92269862e-01 4.76423651e-01
-3.55787247e-01 5.42272270e-01 -5.09023130e-01 -3.68980020e-01
1.46889195e-01 2.37877443e-01 1.94363534e-01 -3.42295438e-01
-7.19946206e-01 1.02115571e+00 4.39725304e-03 5.11783361e-01
-3.19977462e-01 -4.32291716e-01 -6.82798505e-01 2.79496107e-02
4.77184504e-01 -6.72260150e-02 1.73987520e+00 -3.98293346e-01
-3.09165329e-01 1.49877578e-01 -7.82625005e-02 6.82937056e-02
4.15105447e-02 -1.45401850e-01 -6.57373488e-01 -8.56687501e-02
4.85012919e-01 -3.48941624e-01 -2.96586342e-02 -1.15775180e+00
-9.56847787e-01 -5.51382303e-01 -9.20463204e-02 -2.78985322e-01
3.98055077e-01 6.09508991e-01 3.62829715e-01 7.78320059e-02
5.65808713e-01 -4.65275496e-01 -2.61273772e-01 -5.22818983e-01
-1.65466115e-01 -1.73204709e-02 7.59948134e-01 -7.96203017e-01
1.11663282e+00 -1.94971466e+00 -8.27294111e-01 4.43494618e-01
-2.11684853e-01 -1.11902699e-01 3.28317404e-01 1.12051129e+00
1.15968123e-01 2.81907678e-01 5.09565949e-01 3.35126698e-01
1.73408270e-01 3.17263722e-01 -2.52182245e-01 5.88497162e-01
3.63971472e-01 4.28097308e-01 -6.14526808e-01 -8.78511295e-02
3.42685670e-01 -8.00725147e-02 -5.81574321e-01 -2.62720920e-02
6.58889189e-02 2.93320447e-01 -7.35553741e-01 1.22854447e+00
5.95463991e-01 -2.35852540e-01 1.54278308e-01 4.06081649e-03
-4.98904586e-01 8.32912445e-01 -1.06609690e+00 6.04255557e-01
-3.65778387e-01 2.97482938e-01 4.31512833e-01 -9.37802911e-01
9.74484622e-01 1.69002235e-01 6.68810070e-01 -9.92274880e-01
5.85703373e-01 4.74685907e-01 -1.43026803e-02 -6.34892583e-01
7.23509789e-01 -3.48695755e-01 -3.41980994e-01 9.65441108e-01
-6.51595891e-01 -5.80395870e-02 1.78190440e-01 -6.81055784e-02
9.58261609e-01 -5.23173451e-01 -1.68890789e-01 -1.59225434e-01
2.75974572e-02 5.80920577e-01 8.26382577e-01 3.04060549e-01
-5.86261339e-02 2.32939105e-02 4.53427464e-01 -3.11141014e-01
-1.07341087e+00 -6.95685744e-01 -3.32314134e-01 6.20900154e-01
-1.60505414e-01 2.24816967e-02 -1.10420562e-01 -1.10317148e-01
6.18479729e-01 1.20168030e+00 -7.94449896e-02 -1.70847237e-01
-1.23676769e-02 -4.44062412e-01 -4.08457398e-01 7.87755430e-01
3.67219411e-02 -5.88882565e-01 -6.21785104e-01 7.87499309e-01
1.32063717e-01 -8.57628107e-01 -3.75426590e-01 1.47910342e-01
-7.49693871e-01 -1.36753905e+00 -7.72851706e-02 -3.01829517e-01
9.52206612e-01 4.52510625e-01 7.90760994e-01 3.45333457e-01
-3.93519014e-01 7.06589103e-01 -4.52283382e-01 -6.09507203e-01
-5.91325939e-01 -3.84873241e-01 -1.23646624e-01 -3.51025403e-01
9.41146970e-01 -2.94885010e-01 -4.71179754e-01 6.16816342e-01
-8.95822644e-01 -8.51289570e-01 4.95072067e-01 5.65385342e-01
4.23337847e-01 4.51705933e-01 1.02200711e+00 -9.18907225e-01
1.19802785e+00 -3.37916166e-01 -6.43828392e-01 1.27852634e-01
-1.32586181e+00 -5.99151969e-01 1.04590990e-01 1.23074703e-01
-8.68756115e-01 -3.67813200e-01 -6.82178438e-02 4.56638366e-01
-4.31724787e-01 1.05671811e+00 -1.04373485e-01 3.77158493e-01
2.88311213e-01 -1.70634851e-01 2.94593394e-01 -8.49784315e-01
-8.02531391e-02 1.04597354e+00 2.14277819e-01 -5.74019551e-01
4.48081166e-01 2.87817478e-01 -2.90820777e-01 -4.60208535e-01
-7.38901079e-01 -8.61240625e-01 -4.57002997e-01 -2.41379783e-01
4.58255738e-01 -4.47599709e-01 -1.06436408e+00 -2.86253422e-01
-6.67772591e-01 1.77730590e-01 -6.03015602e-01 6.36376739e-01
-2.59783447e-01 -5.81923649e-02 -5.81758201e-01 -1.33565879e+00
1.56555906e-01 -1.17358208e+00 5.28907180e-01 1.95137098e-01
-8.81816328e-01 -9.45935547e-01 -5.96736670e-01 9.20217454e-01
6.70445144e-01 4.37306464e-01 1.00126648e+00 -9.44487214e-01
-8.32251251e-01 -1.00027502e+00 -5.85253388e-02 6.33251905e-01
7.22780049e-01 1.78640902e-01 -5.07885575e-01 -3.01544547e-01
3.98934960e-01 -1.89283788e-01 2.15521768e-01 6.34199798e-01
4.13853616e-01 -5.07128358e-01 -5.49569465e-02 -4.17734802e-01
1.32198286e+00 8.81980181e-01 4.22452241e-01 6.18512332e-01
5.81392273e-03 1.24459827e+00 1.31038797e+00 6.80912197e-01
4.00345981e-01 2.79952019e-01 1.90955281e-01 1.14636116e-01
6.16173148e-01 -1.55854374e-02 -1.48539945e-01 5.60619771e-01
3.57778132e-01 1.55654937e-01 -8.87455344e-01 7.35324323e-01
-1.79650271e+00 -8.32497358e-01 -3.27539533e-01 2.11531281e+00
3.68361682e-01 5.61873198e-01 3.18530053e-01 4.16718513e-01
3.59083980e-01 -4.54307407e-01 -5.76858878e-01 -9.72086072e-01
2.47767478e-01 -3.02163392e-01 9.43159878e-01 3.84456813e-01
-1.27261609e-01 9.51863900e-02 7.48611069e+00 3.38590667e-02
-6.46878779e-01 -3.75279456e-01 6.44995451e-01 -2.12322071e-01
-1.00604367e+00 3.68030518e-01 -9.07798648e-01 1.69172958e-01
1.08039808e+00 -5.40340304e-01 1.87587857e-01 9.53138053e-01
1.27413821e+00 -6.19309902e-01 -1.11296105e+00 3.55543613e-01
-4.34524626e-01 -1.41470838e+00 -4.21795964e-01 6.01470172e-01
3.16629648e-01 -7.07994759e-01 6.65631741e-02 -1.53417453e-01
4.23385918e-01 -7.50797868e-01 6.60404682e-01 4.77996826e-01
-3.72399911e-02 -1.17991233e+00 1.02330613e+00 1.95170701e-01
-6.46462023e-01 -5.30714869e-01 -2.73658603e-01 -6.54877841e-01
9.87443268e-01 6.74900889e-01 -1.13538873e+00 3.24387163e-01
4.95237321e-01 5.24083599e-02 8.68566036e-02 1.04170525e+00
1.61721408e-01 5.00787675e-01 -7.36551359e-02 -1.16484657e-01
2.73221493e-01 -3.13468218e-01 1.00349709e-01 5.60127318e-01
2.46743616e-02 1.72395602e-01 3.64300579e-01 1.03109515e+00
5.79202473e-01 1.16248123e-01 -3.38907123e-01 -7.36225367e-01
7.92578042e-01 7.38747776e-01 -6.32653773e-01 6.16953373e-02
-6.37300849e-01 1.14426903e-01 -6.96579635e-01 3.30471694e-01
-1.31397009e-01 -3.26420426e-01 7.04262376e-01 8.10299039e-01
2.03338906e-01 -4.10919368e-01 -3.97177130e-01 -9.61317196e-02
1.13073327e-01 -9.25261319e-01 1.41189232e-01 -4.58424628e-01
-1.26781094e+00 -1.33373111e-01 4.01299037e-02 -8.18261027e-01
-4.38277781e-01 -5.15168607e-01 -4.94124234e-01 1.14567173e+00
-1.28345966e+00 -6.98692441e-01 8.03946406e-02 -4.63892184e-02
2.40327254e-01 1.55724153e-01 4.52243119e-01 3.98825377e-01
-3.85254949e-01 7.60164931e-02 2.75581837e-01 -2.36404967e-02
2.23305270e-01 -9.10948873e-01 -4.59620431e-02 2.32166365e-01
-6.81540430e-01 8.58695149e-01 9.23764050e-01 -1.03469241e+00
-1.50278473e+00 -5.46622038e-01 9.07355428e-01 -3.49933833e-01
1.01263356e+00 1.13716401e-01 -7.97499835e-01 6.47781968e-01
-1.00360975e-01 -9.89572287e-01 1.31198704e+00 4.90927666e-01
3.62777293e-01 -4.19923961e-01 -1.41665578e+00 1.87901214e-01
2.35971764e-01 -7.04442441e-01 -5.94975412e-01 5.04646957e-01
7.36039817e-01 -1.59611493e-01 -1.28834355e+00 -2.44005427e-01
5.01359284e-01 -7.28393197e-01 2.79302478e-01 -8.88380468e-01
-8.07574484e-03 -1.95132121e-01 -4.10491854e-01 -8.97081077e-01
-2.38208130e-01 -3.78660768e-01 1.06716442e+00 1.46703279e+00
9.07132745e-01 -8.87000680e-01 6.80469751e-01 1.82246101e+00
-4.06789750e-01 -8.00821543e-01 -4.36306685e-01 -7.79902458e-01
-3.03491235e-01 -9.86536086e-01 9.78007793e-01 7.24276543e-01
2.52601206e-02 -4.22372818e-02 8.59409273e-02 2.15819031e-01
6.38626695e-01 4.59841013e-01 7.14991868e-01 -1.13685918e+00
6.42381310e-02 6.58040047e-02 -2.67007262e-01 -7.68638194e-01
-5.12946784e-01 -9.34088826e-02 -2.88839996e-01 -2.01989579e+00
-2.85383761e-02 -6.24548793e-01 -1.45398304e-01 4.24790800e-01
4.44768906e-01 -2.52848029e-01 1.17849171e-01 3.84102255e-01
-2.39105821e-02 -8.97297170e-03 1.10745072e+00 7.06601739e-02
-4.10773665e-01 4.47385520e-01 -1.70954072e+00 5.55818677e-01
7.70270288e-01 -3.29360485e-01 -2.43899181e-01 -2.08862737e-01
6.37287259e-01 6.16824478e-02 2.96938270e-01 -1.84341744e-01
2.58547634e-01 -8.74700010e-01 2.71805018e-01 -1.14551592e+00
1.53940227e-02 -1.29420590e+00 5.84600866e-01 4.81952935e-01
-3.36395562e-01 5.51202118e-01 6.92200139e-02 4.32878971e-01
-2.13005587e-01 -6.38151109e-01 2.53656894e-01 -2.34907749e-03
-2.99169421e-01 -1.59566224e-01 -6.06243551e-01 -3.52869660e-01
9.73115504e-01 -5.60977101e-01 -6.45284534e-01 -9.24131095e-01
-1.13222885e+00 5.61434925e-01 7.70458639e-01 2.03000799e-01
4.05039519e-01 -1.27756226e+00 -4.54892784e-01 1.38958141e-01
-3.84785496e-02 -1.97240457e-01 1.47668123e-01 9.93236363e-01
-4.79208618e-01 8.53020549e-01 1.15783244e-01 -3.56559455e-01
-7.84287095e-01 4.48085517e-01 -4.09704726e-03 3.25076461e-01
-4.50474620e-01 3.42983872e-01 -5.27559042e-01 2.17230052e-01
1.19778430e-02 -5.07005870e-01 1.47678824e-02 3.41165245e-01
7.05198705e-01 4.48963225e-01 2.38793433e-01 -6.48861051e-01
-4.02246714e-01 7.82445446e-02 -5.82841992e-01 -4.05515097e-02
1.71613264e+00 -5.41315019e-01 -5.51936403e-03 5.91828346e-01
9.68579113e-01 1.60390988e-01 -1.09119749e+00 2.91721430e-02
4.50779885e-01 -1.36024237e+00 9.25275534e-02 -6.53188467e-01
-9.46533024e-01 3.04265559e-01 2.32790738e-01 1.11337793e+00
9.31007445e-01 2.30226621e-01 8.24571192e-01 4.77565564e-02
3.09751511e-01 -1.67594469e+00 2.11927458e-03 -2.89540589e-01
9.21151280e-01 -8.51386786e-01 1.97911084e-01 -2.89304703e-01
-7.35363007e-01 8.68130684e-01 -6.19303668e-03 2.32961029e-01
6.46127284e-01 1.62645951e-01 1.57241374e-01 -5.86629033e-01
-5.42841375e-01 -1.11976601e-01 -3.25359374e-01 5.71012795e-01
5.57289243e-01 -2.02304423e-02 -9.16568518e-01 6.38838589e-01
2.28943434e-02 1.83919817e-01 1.00553083e+00 1.52187598e+00
-9.20759320e-01 -1.56475055e+00 -6.54869795e-01 1.00650764e+00
-6.17263973e-01 2.84572810e-01 -5.35905182e-01 9.63364303e-01
-3.30319762e-01 1.55907404e+00 2.01371536e-02 -4.04816940e-02
1.12942016e+00 -3.98326591e-02 -1.38196751e-01 -8.47537279e-01
-3.48612815e-01 6.64555192e-01 9.42300856e-01 -3.27028483e-01
-1.58280820e-01 -1.11671686e+00 -1.38694680e+00 -7.89972007e-01
-8.07415485e-01 5.38901031e-01 1.05111051e+00 9.07606065e-01
3.67758811e-01 7.06991851e-01 9.11503553e-01 -1.43102169e-01
-1.05282116e+00 -6.67044461e-01 -1.28750682e+00 4.44475412e-01
1.97573632e-01 -6.58026397e-01 -5.64399302e-01 -2.95874596e-01] | [9.259124755859375, 5.895352363586426] |
ea7a3af7-d7f6-4fec-8de1-f988f950b4fa | uthealth-at-semeval-2016-task-12-an-end-to | null | null | https://aclanthology.org/S16-1201 | https://aclanthology.org/S16-1201.pdf | UTHealth at SemEval-2016 Task 12: an End-to-End System for Temporal Information Extraction from Clinical Notes | null | ['Sungrim Moon', 'Hee-Jin Lee', 'Yaoyun Zhang', 'Jun Xu', 'Hua Xu', 'Yonghui Wu', 'Jingqi Wang'] | 2016-06-01 | null | null | null | semeval-2016-6 | ['temporal-information-extraction'] | ['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] | [-7.154355525970459, 3.720409631729126] |
06b5c400-198f-47c3-b07d-7b13247513c4 | document-classification-with-word-sense | null | null | https://openreview.net/forum?id=FnXMKuW3fx | https://openreview.net/pdf?id=FnXMKuW3fx | Document Classification with Word Sense Knowledge | The performance of Word Sense Disambiguation (WSD) on a standard evaluation framework has reached an estimated upper bound. However, there is limited research on the application of WSD to relevant NLP tasks due to the high computational cost of supervised systems. In this paper, we propose a partial WSD method with sense category information and incorporate the sense knowledge into a supervised document classification framework. Experimental results show that the proposed method can constantly boost the system’s performance on document classification datasets against strong baselines. | ['Anonymous'] | 2022-01-16 | null | null | null | acl-arr-january-2022-1 | ['word-sense-disambiguation'] | ['natural-language-processing'] | [ 2.06470266e-01 -1.18544795e-01 -3.91338140e-01 -3.59561801e-01
-9.47984040e-01 -6.05199635e-01 9.64900315e-01 4.89819199e-01
-8.65013957e-01 8.36666942e-01 4.06319708e-01 -3.02864105e-01
-2.39773672e-02 -5.99415720e-01 7.55172148e-02 -5.17947435e-01
3.16099793e-01 4.46359217e-01 4.36220556e-01 -6.22060180e-01
4.51667935e-01 1.47925854e-01 -1.42722952e+00 2.92203873e-01
1.05332053e+00 7.89890349e-01 4.46221054e-01 1.97741821e-01
-6.51935697e-01 3.59751165e-01 -9.26113904e-01 -3.30380738e-01
-2.87110428e-03 -1.51558518e-01 -1.07382679e+00 -3.00982565e-01
3.21748644e-01 2.62604445e-01 1.98393151e-01 1.20046031e+00
7.06279993e-01 4.27571565e-01 5.09174109e-01 -9.82165337e-01
-7.34679163e-01 9.07257199e-01 -4.61476237e-01 4.53974813e-01
6.76536441e-01 -4.22482610e-01 1.45149040e+00 -9.56062973e-01
7.16408789e-01 1.29480958e+00 3.87447774e-01 4.72745627e-01
-1.03561902e+00 -5.18569112e-01 4.99549985e-01 2.07711205e-01
-1.41009104e+00 -8.04648995e-02 7.56866992e-01 -4.93304245e-02
1.59978652e+00 9.02694091e-02 4.04762566e-01 1.25666368e+00
-1.20277755e-01 1.03094900e+00 1.20361042e+00 -9.68945205e-01
4.36845690e-01 -5.61595783e-02 9.73374009e-01 2.92570680e-01
6.13651812e-01 -2.28719860e-01 -6.49141610e-01 -9.20081288e-02
1.20693810e-01 -2.99028873e-01 -1.56832188e-01 -1.92944985e-02
-1.01920855e+00 7.76754379e-01 2.38350317e-01 8.67500961e-01
-2.06311241e-01 -4.69896674e-01 4.14751440e-01 2.77550727e-01
8.66814196e-01 1.14830494e+00 -6.33171618e-01 -1.42823532e-01
-9.95528936e-01 5.37695765e-01 8.84128630e-01 7.27177024e-01
3.93337727e-01 -4.31297660e-01 -3.35069984e-01 1.15906072e+00
1.07693419e-01 5.41979730e-01 8.35623741e-01 -4.19596225e-01
5.18616974e-01 8.96584809e-01 1.19903080e-01 -1.04256046e+00
-5.63681424e-01 -4.59452808e-01 -5.52630067e-01 -2.12200344e-01
4.78325151e-02 -1.29211187e-01 -9.13703263e-01 1.54764295e+00
2.55315959e-01 1.75206065e-01 4.56297427e-01 9.57496583e-01
7.98962116e-01 4.70314384e-01 5.21346509e-01 -2.34515846e-01
1.42468262e+00 -7.85627186e-01 -9.08572912e-01 -7.36999810e-01
7.80067682e-01 -1.01259542e+00 1.44971359e+00 3.13596189e-01
-6.05714202e-01 -3.15734565e-01 -1.31441426e+00 -2.77406722e-02
-7.05173135e-01 2.11437959e-02 8.22811306e-01 6.04229212e-01
-8.16004455e-01 2.17021942e-01 -7.71985710e-01 -8.40048432e-01
1.90742090e-01 -6.79891035e-02 -1.87365502e-01 -4.59202230e-01
-1.82936573e+00 1.25587034e+00 9.69099700e-01 -4.15640414e-01
-1.31807178e-01 -6.24325752e-01 -8.43318582e-01 -8.45535249e-02
4.96148854e-01 -7.06359565e-01 1.31062949e+00 -2.39446089e-01
-1.16898870e+00 8.88044477e-01 -2.98489630e-01 -4.61646736e-01
-1.25506356e-01 -4.60425436e-01 -7.02879548e-01 -7.67257204e-03
4.39164251e-01 4.09561187e-01 6.46920726e-02 -1.03784966e+00
-9.01106179e-01 -3.90527666e-01 2.54140913e-01 6.13792658e-01
-5.61591208e-01 7.93469325e-02 -4.89017993e-01 -8.69105935e-01
8.93785954e-02 -9.48518455e-01 -2.83077359e-01 -7.53987014e-01
-3.87668312e-01 -9.01454747e-01 6.60601139e-01 -1.99299574e-01
1.45207381e+00 -2.07035565e+00 -1.81211233e-01 1.19200595e-01
-2.92593032e-01 6.80244207e-01 -3.55856955e-01 5.21848440e-01
1.86431453e-01 1.45516902e-01 -3.30805868e-01 -3.10600579e-01
1.30775020e-01 4.64420915e-01 -4.79533136e-01 -3.33980918e-01
7.70588890e-02 7.48178661e-01 -1.31572998e+00 -4.66132015e-01
6.44965917e-02 3.73953134e-01 -2.93676436e-01 1.66677758e-02
-2.92838573e-01 -3.63358945e-01 -7.64306128e-01 4.38654333e-01
6.03896856e-01 -1.56827524e-01 4.89307284e-01 -9.18862224e-03
6.88277483e-02 1.01579165e+00 -1.16937387e+00 2.04606795e+00
-4.64898437e-01 2.62078375e-01 -4.24664885e-01 -9.46135819e-01
8.36065888e-01 1.75391525e-01 3.01980644e-01 -6.84540391e-01
-2.21976563e-01 1.52815998e-01 -1.57883227e-01 -2.64636576e-01
8.45273376e-01 -2.97419876e-01 -2.85662919e-01 4.63413984e-01
9.47920680e-02 -3.11788678e-01 6.03131771e-01 4.52900946e-01
1.13639653e+00 -1.75452217e-01 5.94421089e-01 -6.01274610e-01
4.24442708e-01 4.05647874e-01 3.23282331e-01 8.79237056e-01
-1.61291033e-01 2.63047040e-01 1.25234768e-01 -1.46704227e-01
-4.21172351e-01 -8.63573253e-01 -2.89754808e-01 1.32191181e+00
2.60022610e-01 -8.22503626e-01 -7.19615579e-01 -9.89772320e-01
5.84259480e-02 9.90994930e-01 -4.41544175e-01 -2.15379700e-01
-2.92045027e-01 -1.05926073e+00 7.52955973e-01 5.42222440e-01
7.42909849e-01 -9.91358936e-01 -1.30001307e-01 3.41535360e-01
-3.84998500e-01 -1.30108154e+00 -2.65971720e-01 2.02548385e-01
-5.87585568e-01 -9.95909154e-01 -3.67738903e-01 -9.02827978e-01
1.80876240e-01 6.26475036e-01 1.20119357e+00 -1.00672789e-01
3.69320996e-02 1.29808947e-01 -8.02360773e-01 -5.97010911e-01
-1.02200687e-01 6.20647192e-01 2.60385454e-01 -5.24678230e-01
1.14280343e+00 -2.08603650e-01 -4.05691117e-01 9.52933058e-02
-1.07580709e+00 -3.67385089e-01 4.63751495e-01 7.02811241e-01
4.89404917e-01 2.43134141e-01 7.13605285e-01 -1.09099352e+00
1.27489305e+00 -2.58017749e-01 -2.46645585e-01 4.23571795e-01
-1.28976500e+00 3.52239132e-01 1.22159660e-01 -1.94462642e-01
-1.47537458e+00 -2.19800189e-01 -1.49609312e-01 3.50594610e-01
-1.62572220e-01 9.28718865e-01 -2.19140023e-01 2.97836781e-01
8.17586243e-01 -7.21690012e-03 -5.73857069e-01 -6.82774365e-01
4.45319414e-01 8.59766483e-01 3.19315195e-01 -7.36947536e-01
4.95897919e-01 1.35812685e-01 -4.24653947e-01 -6.78194046e-01
-1.83196092e+00 -9.70910072e-01 -5.61089754e-01 3.82314503e-01
6.12797141e-01 -8.58721673e-01 6.33506477e-02 1.99114427e-01
-1.15516341e+00 1.83801651e-01 -7.50526190e-02 6.59140348e-01
1.27458140e-01 4.20105904e-01 -3.22640657e-01 -4.39979821e-01
-4.99639004e-01 -7.14683235e-01 1.18367934e+00 1.37265444e-01
-5.60980141e-01 -1.17554867e+00 4.75841343e-01 3.63284051e-01
4.41925019e-01 -1.30789235e-01 7.50367343e-01 -1.26108682e+00
1.06233887e-01 -2.91802704e-01 -8.85121673e-02 4.02644694e-01
3.67768347e-01 -4.71972972e-01 -1.02406108e+00 -1.15041606e-01
-1.45022899e-01 -3.49971563e-01 1.15624738e+00 1.73846155e-01
6.92795515e-01 -1.52360484e-01 -7.30401874e-01 -6.68680519e-02
1.37619591e+00 8.48681927e-02 2.44272381e-01 6.43791556e-01
3.45561385e-01 2.96377510e-01 1.02478933e+00 3.64489108e-01
3.69351208e-01 6.30016744e-01 -1.56421661e-01 9.91361961e-02
-3.21513414e-01 -2.14169905e-01 1.04980476e-01 7.84977436e-01
7.27152526e-02 -5.61117887e-01 -1.28878951e+00 5.42927623e-01
-1.71965659e+00 -7.79611826e-01 1.41286746e-01 1.80845749e+00
1.24131823e+00 4.25381422e-01 -3.18430483e-01 2.38006502e-01
6.98285937e-01 4.67858911e-01 -1.63996533e-01 -4.01492029e-01
-4.56720978e-01 4.29718912e-01 9.70980823e-02 5.99568784e-01
-1.22284496e+00 1.41721451e+00 6.76276636e+00 1.00676310e+00
-7.36274004e-01 8.19486007e-03 -1.48483133e-02 9.43905339e-02
-2.24447325e-01 -1.22263961e-01 -1.14394331e+00 1.20666802e-01
6.06170952e-01 -5.14164627e-01 3.30666662e-03 8.22458804e-01
-2.24063933e-01 -2.24508401e-02 -8.77347171e-01 9.37276721e-01
2.32477903e-01 -9.74623740e-01 3.32871288e-01 -1.71255827e-01
7.51872957e-01 3.07696372e-01 -3.14063728e-01 5.47626972e-01
8.39540005e-01 -5.65108061e-01 9.34456587e-02 -5.41488677e-02
4.34729934e-01 -7.50301778e-01 1.02699852e+00 2.99037665e-01
-9.34384465e-01 3.11258495e-01 -5.17317474e-01 -2.95763880e-01
-1.63327716e-02 1.01388466e+00 -1.31303489e+00 5.29111624e-01
4.29709524e-01 8.62814426e-01 -6.86386466e-01 6.97781444e-01
-6.69199765e-01 9.85315204e-01 -3.62850904e-01 -3.46462667e-01
4.36541855e-01 2.47352675e-01 6.21362627e-01 1.65604353e+00
1.12726621e-01 4.07826483e-01 5.44872403e-01 2.65818775e-01
-2.01768160e-01 3.34207654e-01 -5.53350925e-01 -1.84272081e-01
7.78493702e-01 1.09268975e+00 -7.00348198e-01 -5.23676157e-01
-2.91240454e-01 8.29350173e-01 4.00544316e-01 2.36633837e-01
-1.23980165e-01 -6.20710433e-01 9.21392322e-01 -3.24268311e-01
7.39096105e-02 -1.78636760e-01 -5.20342827e-01 -1.25635004e+00
1.89962491e-01 -5.92384875e-01 1.05074358e+00 -5.14049649e-01
-1.66349864e+00 7.06200898e-01 1.11164622e-01 -8.96809578e-01
-2.71271348e-01 -6.33298874e-01 -2.26074114e-01 6.56807899e-01
-1.96480262e+00 -9.44162786e-01 3.29946354e-02 3.78633022e-01
5.53423643e-01 -2.26745933e-01 1.27950501e+00 4.47561592e-02
-3.22817832e-01 6.34535491e-01 9.13341269e-02 3.07982981e-01
1.03786337e+00 -1.65273511e+00 5.65376401e-01 9.70420003e-01
4.46249694e-01 8.42230499e-01 7.88971066e-01 -6.34404898e-01
-8.60871375e-01 -9.10927355e-01 1.54424238e+00 -6.05751753e-01
7.17234492e-01 -1.21405482e-01 -1.08239365e+00 4.85638380e-01
5.42599499e-01 -5.46554923e-02 1.06502724e+00 6.83820128e-01
-8.23249161e-01 -1.91230755e-02 -1.06775045e+00 4.94922698e-01
1.14020741e+00 -4.46226060e-01 -1.51750171e+00 3.28180254e-01
8.81758451e-01 -3.03733587e-01 -6.40277088e-01 3.74649853e-01
2.37836614e-01 -2.13280797e-01 1.07849991e+00 -8.59922111e-01
2.33488008e-01 -3.13118249e-01 -3.36341798e-01 -1.84712231e+00
-2.41393983e-01 -1.37314558e-01 1.11341476e-01 1.57724440e+00
6.08193278e-01 -6.25229001e-01 1.64158344e-01 6.48866236e-01
-1.54738367e-01 -3.81326914e-01 -8.98556173e-01 -9.50292468e-01
2.65995383e-01 -5.99501729e-01 7.44849205e-01 1.37140131e+00
5.62779307e-01 8.34374428e-01 1.73084110e-01 2.94934571e-01
3.52883041e-01 3.29091817e-01 1.18557535e-01 -1.37557578e+00
5.23928553e-03 -4.90690559e-01 -4.31678563e-01 -8.46103549e-01
5.65373600e-01 -1.21525919e+00 1.02638371e-01 -1.75218177e+00
2.42270499e-01 -3.36902201e-01 -1.05289268e+00 7.80396938e-01
-8.81345510e-01 1.99433062e-02 4.15195711e-02 8.26834515e-03
-1.03820753e+00 3.55902761e-01 8.96002233e-01 -2.33564347e-01
-1.13703012e-01 -4.34886605e-01 -1.07990599e+00 7.12855220e-01
8.00217688e-01 -7.27825582e-01 -4.87138838e-01 -7.15053678e-01
2.51869082e-01 -3.48148972e-01 -3.30996603e-01 -8.67842138e-01
2.45948881e-01 -3.49803180e-01 4.62197438e-02 -3.17314208e-01
-3.38479616e-02 -5.47767699e-01 -6.90433443e-01 3.06485325e-01
-7.21678257e-01 6.99842125e-02 2.35218629e-01 5.30551195e-01
-5.43525338e-01 -5.10668717e-02 5.16919851e-01 -6.13716291e-03
-1.03138709e+00 -6.39697015e-02 -1.66889295e-01 5.58911681e-01
6.53997242e-01 3.18659544e-01 -4.01991636e-01 2.44935080e-01
-3.27925146e-01 3.36851031e-01 3.22070926e-01 1.01075161e+00
4.20568228e-01 -1.40882325e+00 -7.29810536e-01 -1.87444061e-01
6.71745360e-01 -1.28216952e-01 -2.63610035e-01 6.99016154e-02
6.35045394e-02 7.86588430e-01 6.58028498e-02 -3.77093911e-01
-1.18735588e+00 6.19817913e-01 -3.62447090e-02 -7.58927703e-01
-4.77165490e-01 8.41962993e-01 -2.31369480e-01 -5.12063682e-01
1.47571951e-01 -2.71656066e-01 -6.38897121e-01 2.53244609e-01
9.05867398e-01 1.59305975e-01 5.52895367e-01 -3.01144123e-01
-8.09306622e-01 3.90942067e-01 -3.66713196e-01 -1.27576888e-01
1.30501139e+00 -2.22831830e-01 7.31645059e-03 3.09828073e-01
9.70779657e-01 -2.04378832e-02 -4.27139342e-01 -6.82990074e-01
7.03802526e-01 -4.17243898e-01 3.75839084e-01 -1.29034626e+00
-6.44380152e-01 6.03128016e-01 4.60805267e-01 2.23057196e-01
1.08524334e+00 1.50016233e-01 9.31209087e-01 9.01203811e-01
4.37653542e-01 -1.24047995e+00 -2.83932567e-01 9.80067372e-01
6.58081651e-01 -1.36674011e+00 -1.34984300e-01 -3.79389286e-01
-7.33085275e-01 7.07325637e-01 5.25264680e-01 -3.66007350e-02
6.76554263e-01 3.29073220e-01 3.99182886e-01 -1.74129173e-01
-9.06665385e-01 -5.74605167e-01 5.86892009e-01 5.95862806e-01
7.64562190e-01 7.51316696e-02 -1.00092757e+00 8.99177969e-01
-3.34450185e-01 9.55223888e-02 1.74854055e-01 9.97229278e-01
-7.73392260e-01 -1.43777978e+00 -3.39281335e-02 2.18380645e-01
-6.46871924e-01 -5.42312860e-01 -7.46299207e-01 6.28026307e-01
-9.42715406e-02 1.26351750e+00 -2.40613148e-01 -2.20297709e-01
5.39213300e-01 4.98559386e-01 2.24872917e-01 -1.00916457e+00
-2.46458292e-01 -2.90555954e-01 3.40653241e-01 -4.76460546e-01
-8.08009446e-01 -6.29562080e-01 -1.36830485e+00 1.90317735e-01
-2.99212754e-01 5.76989233e-01 5.14279366e-01 1.33638430e+00
4.24573392e-01 2.65412450e-01 3.38650256e-01 -1.04998671e-01
-6.11243427e-01 -1.23243821e+00 -6.21981442e-01 7.37074256e-01
1.41874045e-01 -8.06737363e-01 -1.88022688e-01 -1.58881009e-01] | [10.26850700378418, 9.035025596618652] |
b364e6e6-07da-4cb5-8484-eca5e0ad535e | deep-reinforcement-learning-assisted-1 | 2206.11715 | null | https://arxiv.org/abs/2206.11715v1 | https://arxiv.org/pdf/2206.11715v1.pdf | Deep Reinforcement Learning-Assisted Federated Learning for Robust Short-term Utility Demand Forecasting in Electricity Wholesale Markets | Short-term load forecasting (STLF) plays a significant role in the operation of electricity trading markets. Considering the growing concern of data privacy, federated learning (FL) is increasingly adopted to train STLF models for utility companies (UCs) in recent research. Inspiringly, in wholesale markets, as it is not realistic for power plants (PPs) to access UCs' data directly, FL is definitely a feasible solution of obtaining an accurate STLF model for PPs. However, due to FL's distributed nature and intense competition among UCs, defects increasingly occur and lead to poor performance of the STLF model, indicating that simply adopting FL is not enough. In this paper, we propose a DRL-assisted FL approach, DEfect-AwaRe federated soft actor-critic (DearFSAC), to robustly train an accurate STLF model for PPs to forecast precise short-term utility electricity demand. Firstly. we design a STLF model based on long short-term memory (LSTM) using just historical load data and time data. Furthermore, considering the uncertainty of defects occurrence, a deep reinforcement learning (DRL) algorithm is adopted to assist FL by alleviating model degradation caused by defects. In addition, for faster convergence of FL training, an auto-encoder is designed for both dimension reduction and quality evaluation of uploaded models. In the simulations, we validate our approach on real data of Helsinki's UCs in 2019. The results show that DearFSAC outperforms all the other approaches no matter if defects occur or not. | ['Yanru Zhang', 'Yingjie Zhou', 'Changkun Jiang', 'Shengrong Bu', 'Yuxi Chen', 'Shunji Yang', 'Feng Hong', 'Xiaoyi Wang', 'Weilong Chen', 'Chenghao Huang'] | 2022-06-23 | null | null | null | null | ['load-forecasting'] | ['miscellaneous'] | [-4.26574469e-01 8.40211883e-02 1.68230534e-02 -3.63621622e-01
-5.19388080e-01 -4.21791017e-01 4.40040588e-01 1.22424819e-01
-1.62641227e-01 1.02202344e+00 -4.38820347e-02 -5.08178353e-01
-4.36013550e-01 -9.93208587e-01 -7.09545195e-01 -9.11229610e-01
-4.05097187e-01 4.80847359e-01 -5.46088874e-01 5.64218871e-02
-6.72170743e-02 5.51633179e-01 -1.14604974e+00 -2.52581593e-02
1.01838124e+00 1.42451084e+00 2.18327850e-01 1.09527826e-01
-1.99924558e-01 8.80360425e-01 -7.78690219e-01 -4.31002080e-01
5.94248652e-01 -5.72417267e-02 -6.61688030e-01 -1.24726176e-01
-6.64228082e-01 -6.02874577e-01 -6.27306253e-02 1.05944276e+00
5.63473582e-01 1.04759194e-01 3.70337903e-01 -1.43498921e+00
-5.33907712e-01 9.11846519e-01 -3.31310600e-01 -1.12821482e-01
-1.02700546e-01 2.01540560e-01 1.02417433e+00 -4.06278282e-01
8.74565765e-02 6.96297944e-01 5.09917557e-01 2.81952858e-01
-1.11516762e+00 -5.87384105e-01 2.55677789e-01 1.40321761e-01
-1.12727189e+00 -1.63852245e-01 9.22285378e-01 -1.00325763e-01
1.08221138e+00 5.42249680e-01 8.85481656e-01 8.94919038e-01
4.72716987e-01 1.03934765e+00 8.63810718e-01 -2.40215272e-01
8.14547896e-01 3.41283113e-01 -4.94372994e-01 -1.84927490e-02
1.45392969e-01 3.65866601e-01 3.87027115e-02 -3.97324115e-01
3.95725429e-01 1.32976979e-01 -2.55823076e-01 -3.55195522e-01
-7.26332545e-01 8.95996273e-01 3.82911384e-01 5.23397744e-01
-7.93371558e-01 -2.72823162e-02 7.50909925e-01 4.86364216e-01
7.58190155e-01 3.83079588e-01 -8.27743709e-01 -5.18289693e-02
-9.81222332e-01 -7.33540654e-02 7.78139234e-01 7.62643933e-01
2.02054411e-01 8.26591492e-01 5.01646399e-02 5.41544080e-01
1.98191494e-01 5.09951591e-01 9.26607311e-01 -7.76687384e-01
4.48621571e-01 4.67015922e-01 2.13313058e-01 -7.50680089e-01
-4.31665480e-01 -9.14667904e-01 -1.27867961e+00 3.86597574e-01
2.48561904e-01 -4.46412414e-01 -2.58175105e-01 1.46892345e+00
1.09133427e-03 1.41151860e-01 3.32731307e-01 6.23369098e-01
4.65476029e-02 7.68569231e-01 -4.04133312e-02 -5.62551022e-01
7.85915375e-01 -5.12404978e-01 -8.08715165e-01 2.04282448e-01
6.20217144e-01 -2.07120463e-01 5.09914696e-01 5.82043290e-01
-9.12445843e-01 -3.09168249e-01 -7.05518246e-01 6.13723755e-01
-6.32617772e-01 4.90499511e-02 4.23824966e-01 5.33435941e-01
-8.78718197e-01 8.56261075e-01 -6.07901812e-01 1.20847799e-01
5.45367718e-01 4.86411482e-01 -5.35683855e-02 3.93376499e-01
-1.71381342e+00 1.02458715e+00 5.23530483e-01 5.08856058e-01
-7.54193604e-01 -6.46284401e-01 -5.66805482e-01 4.49265391e-01
1.25916362e-01 -3.87747288e-01 1.10677838e+00 -1.07703400e+00
-1.53705347e+00 -1.12442933e-01 5.18603981e-01 -1.21012759e+00
7.95984089e-01 2.66141951e-01 -8.48346889e-01 -2.41595894e-01
-4.26169217e-01 1.00851282e-01 1.06154251e+00 -1.09767413e+00
-6.87900066e-01 -1.89757019e-01 -1.85239106e-01 3.68383341e-02
-5.59414685e-01 -5.09870350e-01 7.20301807e-01 -6.94221616e-01
-4.26294893e-01 -4.14757639e-01 -4.13672864e-01 -3.81834328e-01
-2.65855134e-01 -4.16065782e-01 1.06804538e+00 -8.85741830e-01
1.34116435e+00 -2.12628698e+00 -3.24123442e-01 5.38751960e-01
-1.46032348e-01 5.07591546e-01 7.16463998e-02 5.08315206e-01
-2.75897324e-01 -4.59037116e-03 -1.81723237e-01 -3.42855155e-01
5.90619981e-01 5.26763439e-01 -4.80686039e-01 5.32557905e-01
2.71051116e-02 1.04368186e+00 -8.50170493e-01 7.39260614e-02
5.77394366e-01 4.27554220e-01 6.19302019e-02 5.42759784e-02
-6.05662346e-01 4.07258034e-01 -5.92963040e-01 5.17169774e-01
7.48174489e-01 -1.37368366e-01 8.89060125e-02 -7.50554800e-02
-1.98057815e-01 -2.93608099e-01 -1.01222730e+00 1.30636144e+00
-8.25342178e-01 2.94386595e-01 5.37515339e-03 -1.37806571e+00
1.26709616e+00 7.75114655e-01 8.10845971e-01 -8.38805974e-01
1.49310619e-01 5.77036738e-01 -2.23042250e-01 -2.30041698e-01
2.63300598e-01 -3.80284674e-02 1.26950234e-01 6.54089153e-01
-1.15661003e-01 -1.00335218e-01 -2.81445682e-02 -2.18441963e-01
9.11526382e-01 4.71504852e-02 2.73562707e-02 -2.31958270e-01
7.07799435e-01 -5.09356976e-01 6.73610568e-01 3.25158924e-01
-6.35595694e-02 1.65432930e-01 2.98986137e-01 -8.63624573e-01
-9.29438293e-01 -5.46134651e-01 -1.89076588e-01 2.83380359e-01
-5.10824382e-01 8.19601044e-02 -4.18948233e-01 -9.13316548e-01
3.25719953e-01 1.49725151e+00 -3.18705946e-01 -1.97597504e-01
-1.51739240e-01 -8.52327824e-01 2.17392042e-01 3.28596443e-01
5.62022328e-01 -1.26845193e+00 -8.35928738e-01 7.24931777e-01
1.70141160e-01 -5.02784252e-01 -1.28401935e-01 6.01277769e-01
-5.77250600e-01 -7.02193260e-01 -9.28013980e-01 -1.11135326e-01
5.96466959e-01 -3.59157503e-01 9.60649848e-01 -3.39221716e-01
9.44031924e-02 1.18582368e-01 -1.83161348e-01 -6.03042603e-01
-5.11002004e-01 1.48857906e-01 -2.58424338e-02 1.83567658e-01
2.76216298e-01 -6.73839688e-01 -4.85077560e-01 1.29129827e-01
-8.43321025e-01 -4.27658498e-01 3.29415143e-01 9.18906391e-01
3.34753454e-01 9.12840486e-01 1.42855549e+00 -8.72973144e-01
7.09043384e-01 -6.77457809e-01 -1.28843367e+00 5.03849864e-01
-1.28232527e+00 -7.36423507e-02 1.31198382e+00 -1.07273623e-01
-1.23178649e+00 6.67349845e-02 -1.57878563e-01 -7.23788977e-01
5.18009663e-02 6.19792819e-01 -4.92188409e-02 9.76494700e-02
6.35179803e-02 3.83526176e-01 1.30039349e-01 -5.09608269e-01
-9.78093222e-03 8.07183623e-01 7.26477951e-02 -1.71504691e-01
7.41854548e-01 6.48907498e-02 -1.16565917e-02 -3.29238951e-01
-4.45253432e-01 1.23977281e-01 -5.83402403e-02 -1.86517224e-01
3.66127104e-01 -1.00282073e+00 -1.06926632e+00 5.08642197e-01
-8.60936344e-01 -2.11637467e-01 -9.32948411e-01 4.84277576e-01
-6.77643239e-01 -3.22804786e-02 -3.20056289e-01 -1.25234532e+00
-6.39970779e-01 -9.20924366e-01 4.10525560e-01 -2.67442549e-03
-4.13226038e-02 -1.32322276e+00 -3.45434211e-02 6.99872822e-02
9.21785474e-01 4.51946586e-01 9.14820790e-01 -8.63269150e-01
-2.77013123e-01 -5.14781117e-01 1.25648424e-01 9.03977275e-01
3.55957091e-01 -2.50871807e-01 -1.10279894e+00 -7.06929803e-01
4.06749398e-01 -1.54641479e-01 1.86039209e-01 3.55843633e-01
1.32636893e+00 -6.68558121e-01 1.82649866e-01 3.67819697e-01
1.52436852e+00 5.19679844e-01 4.58121955e-01 3.55258763e-01
2.38358945e-01 3.38885695e-01 5.09953141e-01 9.80566978e-01
3.81430209e-01 8.97633284e-02 8.15473557e-01 2.41587851e-02
5.75940490e-01 -2.39165545e-01 4.87583250e-01 7.31301963e-01
1.86168790e-01 -4.38622057e-01 -4.47818756e-01 5.00814259e-01
-1.85726774e+00 -9.73273098e-01 2.00997680e-01 2.34703565e+00
4.17936146e-01 1.98120952e-01 -8.37290436e-02 4.45612818e-01
5.47142625e-01 1.83292940e-01 -1.01089215e+00 -7.25728393e-01
-3.69456679e-01 -7.67784566e-02 7.74481475e-01 8.77083391e-02
-6.89032853e-01 -4.00897339e-02 4.62442398e+00 1.10720289e+00
-1.20623815e+00 9.85864364e-03 1.13993430e+00 -3.48933302e-02
-6.00317597e-01 -2.93664306e-01 -2.13800997e-01 8.98513198e-01
1.23830795e+00 -5.07103324e-01 7.79908657e-01 8.59999835e-01
4.85140979e-01 2.16022208e-02 -9.10402596e-01 7.53350556e-01
-5.32813072e-01 -1.17045212e+00 -3.33491743e-01 1.38308078e-01
9.13247883e-01 1.56551436e-01 5.63565902e-02 3.34993452e-01
4.95521158e-01 -8.29744935e-01 7.42964566e-01 7.62742817e-01
4.11647856e-01 -1.27275705e+00 1.15470922e+00 6.51961207e-01
-1.08031368e+00 -7.16636360e-01 -4.18802083e-01 2.31347561e-01
2.56462038e-01 1.03238499e+00 -5.01839757e-01 1.15182173e+00
6.97572827e-01 6.15357101e-01 -1.03130899e-01 7.36225486e-01
-4.42393124e-03 4.94758427e-01 -4.78186548e-01 5.46841621e-02
2.67917454e-01 -3.17373604e-01 2.90822446e-01 3.56883317e-01
6.45129800e-01 -2.51329273e-01 -1.02779426e-01 7.54223049e-01
-1.83216278e-02 -2.54966058e-02 -6.68622315e-01 -1.13987349e-01
4.87745941e-01 1.13727379e+00 -1.98482186e-01 -3.23499665e-02
-4.50121015e-01 7.21622586e-01 -5.75299263e-02 3.58054370e-01
-6.72887683e-01 -2.15035006e-01 2.38465995e-01 -1.34326905e-01
4.73018706e-01 2.36900762e-01 -1.13549724e-01 -9.79292214e-01
-1.99984815e-02 -7.39823937e-01 5.23848832e-01 -6.17112935e-01
-1.66351163e+00 7.98398972e-01 -4.28866684e-01 -1.47742677e+00
-7.14307189e-01 -4.92239781e-02 -6.73506975e-01 9.95720208e-01
-1.98522437e+00 -9.42045867e-01 5.63462615e-01 8.22267890e-01
4.04081523e-01 -3.88191402e-01 1.05277109e+00 3.51694643e-01
-6.57214105e-01 3.57494742e-01 8.61122727e-01 -1.34412184e-01
7.25860968e-02 -1.17912805e+00 1.54079050e-01 6.13894522e-01
-1.43685844e-02 -8.79780799e-02 4.94730383e-01 -5.43056011e-01
-1.35452592e+00 -1.30301309e+00 8.87417674e-01 3.23315084e-01
6.84895873e-01 -4.56278212e-02 -7.74470329e-01 5.85809708e-01
5.67248940e-01 3.01827520e-01 3.51653159e-01 -4.33385760e-01
3.35376769e-01 -6.64008498e-01 -1.81454241e+00 -7.90940076e-02
2.01744244e-01 -2.53620833e-01 -2.45994955e-01 3.96452516e-01
3.85063171e-01 -2.93644480e-02 -1.23338377e+00 1.68504238e-01
1.41548395e-01 -8.45479667e-01 5.62276542e-01 -1.05989739e-01
-2.84992039e-01 -1.14303872e-01 -6.07009307e-02 -1.87403810e+00
-8.15878958e-02 -8.29090595e-01 -4.77948248e-01 1.38152099e+00
3.29457760e-01 -1.41583395e+00 6.41709447e-01 1.01823950e+00
9.55935270e-02 -7.18417525e-01 -1.43135464e+00 -1.10276663e+00
7.65799657e-02 -3.17791313e-01 1.54487169e+00 1.02041912e+00
1.77682757e-01 -3.17982733e-01 -4.93227005e-01 9.71321315e-02
8.06828499e-01 4.29165542e-01 1.36517644e-01 -1.06871879e+00
-2.41703704e-01 -2.75549412e-01 -2.56826878e-02 -3.29753309e-01
3.33727002e-01 -6.96884155e-01 -2.98495442e-01 -1.26872015e+00
-6.06068134e-01 -7.12189078e-01 -8.71861100e-01 5.49795926e-01
6.36238098e-01 -2.27738827e-01 4.03271288e-01 -1.19239017e-01
-2.39207298e-01 9.97992873e-01 9.51996326e-01 -2.32897997e-01
5.95830977e-02 6.49792552e-01 -2.55218804e-01 4.06560570e-01
1.24766517e+00 -2.13428408e-01 -6.51500046e-01 -3.49229187e-01
3.72843921e-01 4.72846091e-01 1.53757423e-01 -8.27441812e-01
2.59868056e-01 1.25570148e-01 5.59486151e-01 -8.27658057e-01
7.78974295e-02 -1.66659606e+00 8.18992436e-01 7.71119416e-01
-1.98152378e-01 1.41727000e-01 -1.71206407e-02 4.58704382e-01
-3.26462269e-01 -3.70499492e-01 3.02361548e-01 -2.07957834e-01
-2.43479058e-01 4.91380543e-01 -3.82711291e-01 -4.16550726e-01
1.13945556e+00 1.78530812e-01 -7.46801049e-02 -5.58442950e-01
-7.54555106e-01 7.56493747e-01 -3.99278216e-02 2.58606851e-01
3.89310718e-01 -1.29645872e+00 -6.23602033e-01 4.99665588e-01
-5.18431664e-01 1.11582480e-01 3.28037947e-01 5.70556045e-01
6.54360652e-02 5.17674804e-01 2.03571752e-01 -1.83374763e-01
-3.85083377e-01 7.27299392e-01 3.07696521e-01 -5.38255155e-01
-5.14981866e-01 4.45056915e-01 -4.02878135e-01 -5.06599247e-01
3.33416313e-01 -4.32743937e-01 -1.09325573e-01 2.96671093e-01
2.51584113e-01 5.43285906e-01 5.05889595e-01 -3.17466736e-01
-6.89096153e-02 -1.46051705e-01 1.12150170e-01 3.38227868e-01
1.87254596e+00 -3.23524475e-01 -4.76705655e-02 4.73852158e-01
1.12061441e+00 -4.20355618e-01 -1.43683875e+00 -1.07886434e-01
1.65691823e-01 -3.72211754e-01 3.13855708e-01 -1.10336936e+00
-1.79648280e+00 5.49605548e-01 6.64865315e-01 8.54251266e-01
1.42113554e+00 -6.59289002e-01 1.00831032e+00 1.04584672e-01
7.17798710e-01 -1.23819613e+00 -8.21814358e-01 1.26731962e-01
8.36744666e-01 -9.84346747e-01 -1.11813188e-01 4.98394668e-01
-6.26141667e-01 1.24582267e+00 5.58227301e-02 1.91249065e-02
8.78928006e-01 4.84847248e-01 -1.08154550e-01 1.99405164e-01
-9.80288029e-01 3.63353491e-01 -2.61963338e-01 3.60526174e-01
-3.46230805e-01 2.41616488e-01 -1.27203092e-01 7.10009754e-01
-5.68718538e-02 2.80156314e-01 5.38984656e-01 8.40946972e-01
2.12118909e-01 -1.22020888e+00 -2.97747284e-01 7.12173283e-01
-4.65035021e-01 1.17592022e-01 4.05682623e-01 4.62687790e-01
8.92045051e-02 8.82514119e-01 5.86508885e-02 -7.03005642e-02
3.02102625e-01 4.81325574e-02 -1.18627094e-01 1.28450558e-01
-9.13927674e-01 1.03674814e-01 -1.76103026e-01 -3.64951998e-01
-1.73705697e-01 -8.44115853e-01 -1.19942236e+00 -4.83501345e-01
-6.85559452e-01 5.41863680e-01 8.93396735e-01 8.78102124e-01
3.32401454e-01 4.97064978e-01 1.60764945e+00 -4.69684124e-01
-1.24245739e+00 -6.48584902e-01 -1.12876618e+00 -4.18417193e-02
3.93162847e-01 -1.51493102e-01 -7.02435374e-01 -5.87429523e-01] | [5.872704982757568, 2.728205919265747] |
1400ab46-c47f-4bc6-aafd-fd4d711b9a23 | streaming-end-to-end-target-speaker-asr | 2209.04175 | null | https://arxiv.org/abs/2209.04175v2 | https://arxiv.org/pdf/2209.04175v2.pdf | Streaming Target-Speaker ASR with Neural Transducer | Although recent advances in deep learning technology have boosted automatic speech recognition (ASR) performance in the single-talker case, it remains difficult to recognize multi-talker speech in which many voices overlap. One conventional approach to tackle this problem is to use a cascade of a speech separation or target speech extraction front-end with an ASR back-end. However, the extra computation costs of the front-end module are a critical barrier to quick response, especially for streaming ASR. In this paper, we propose a target-speaker ASR (TS-ASR) system that implicitly integrates the target speech extraction functionality within a streaming end-to-end (E2E) ASR system, i.e. recurrent neural network-transducer (RNNT). Our system uses a similar idea as adopted for target speech extraction, but implements it directly at the level of the encoder of RNNT. This allows TS-ASR to be realized without placing extra computation costs on the front-end. Note that this study presents two major differences between prior studies on E2E TS-ASR; we investigate streaming models and base our study on Conformer models, whereas prior studies used RNN-based systems and considered only offline processing. We confirm in experiments that our TS-ASR achieves comparable recognition performance with conventional cascade systems in the offline setting, while reducing computation costs and realizing streaming TS-ASR. | ['Takahiro Shinozaki', 'Marc Delcroix', 'Tsubasa Ochiai', 'Hiroshi Sato', 'Takafumi Moriya'] | 2022-09-09 | null | null | null | null | ['speech-separation', 'speech-extraction'] | ['speech', 'speech'] | [ 2.82427251e-01 2.22490728e-01 2.20389441e-01 -2.69653052e-01
-1.36817801e+00 -7.15638161e-01 4.41077113e-01 -3.34278822e-01
-4.77126241e-01 7.64814615e-02 3.45575720e-01 -7.69235253e-01
1.18979432e-01 -2.32075438e-01 -4.45323169e-01 -5.42704642e-01
2.50839502e-01 2.56745666e-01 9.15201157e-02 -3.16042602e-01
-3.09535265e-01 6.99798763e-01 -1.41726899e+00 6.28714561e-01
3.54784697e-01 8.63590837e-01 3.41578990e-01 1.41419911e+00
-2.10335385e-02 1.01577044e+00 -1.14040899e+00 -3.59009087e-01
2.90759504e-01 -5.27717292e-01 -5.36879182e-01 -7.28433505e-02
2.57250309e-01 -3.86278152e-01 -6.51440024e-01 7.00619578e-01
1.05678940e+00 1.70116946e-01 3.56200725e-01 -8.17955971e-01
-1.64900064e-01 1.12589943e+00 -9.44197401e-02 3.60375524e-01
2.93045968e-01 1.03804432e-01 1.10618818e+00 -1.27056110e+00
3.61181125e-02 1.26796639e+00 5.80192268e-01 7.92317510e-01
-1.16537261e+00 -6.36447728e-01 -3.52599681e-03 -9.67026781e-03
-1.38610160e+00 -1.53113210e+00 7.11941898e-01 -4.66561578e-02
1.59620094e+00 5.09644628e-01 4.62968081e-01 1.24443316e+00
-3.28583658e-01 1.13693535e+00 7.31763661e-01 -4.84329909e-01
1.69832572e-01 1.34941444e-01 1.38019070e-01 1.58660144e-01
-3.87224495e-01 1.71407551e-01 -8.35722566e-01 2.12231994e-01
6.01113081e-01 -2.48756006e-01 -2.72599757e-01 6.28060877e-01
-9.24676716e-01 5.84024847e-01 -5.23637347e-02 7.40797341e-01
-6.12882555e-01 6.48336112e-02 6.83289468e-01 7.17582107e-01
6.79644167e-01 2.14617723e-03 -5.12433171e-01 -4.38901961e-01
-1.64065695e+00 -1.69779211e-01 9.72284436e-01 8.61899674e-01
1.23495348e-01 7.28698075e-01 -2.91583776e-01 1.26385701e+00
2.25083157e-01 6.48404896e-01 8.14278603e-01 -4.91222441e-01
7.59773672e-01 -1.47339627e-01 -3.32209051e-01 -3.83470267e-01
-6.58136681e-02 -8.33479941e-01 -8.10631573e-01 6.88424855e-02
1.40166834e-01 -3.19052100e-01 -9.88700151e-01 1.41477442e+00
7.18179792e-02 8.05120766e-02 4.31472272e-01 9.96101797e-01
8.68396461e-01 9.85895157e-01 -3.25789183e-01 -3.31853479e-01
1.32732689e+00 -1.23574758e+00 -9.16943431e-01 -3.37992698e-01
6.81762695e-01 -1.01048052e+00 8.85216534e-01 5.30459583e-01
-1.36234105e+00 -5.83581567e-01 -8.37002873e-01 -1.29428804e-01
-2.57054538e-01 7.23714113e-01 -8.50740448e-02 8.97206962e-01
-1.49835348e+00 2.98123270e-01 -7.24169552e-01 -1.78912207e-01
-1.35071173e-01 4.22044367e-01 -2.03504503e-01 3.47372741e-01
-1.26036417e+00 8.01042080e-01 5.82965538e-02 5.48544168e-01
-1.03980339e+00 -4.87529397e-01 -7.53718317e-01 3.46980155e-01
4.54443634e-01 -4.11914140e-01 1.91589057e+00 -1.26987290e+00
-2.34452868e+00 5.84186673e-01 -5.99755108e-01 -7.54767179e-01
4.97305155e-01 -3.82925928e-01 -6.10630631e-01 5.14571518e-02
-5.38979590e-01 1.67076111e-01 1.22488499e+00 -8.61756444e-01
-4.39068913e-01 -3.13863218e-01 -2.80161500e-01 2.51389831e-01
-2.23407224e-01 5.39379716e-01 -1.20529994e-01 -8.14456701e-01
-1.42733268e-02 -7.14763045e-01 1.10947028e-01 -7.29638338e-01
-4.37998116e-01 -2.91268885e-01 8.52005541e-01 -1.18369126e+00
1.42456830e+00 -2.34444046e+00 8.28694105e-02 4.77759913e-02
-2.05175057e-02 9.51773167e-01 -2.88958877e-01 6.98202908e-01
-4.34141010e-01 9.67493374e-03 -7.78292045e-02 -9.23125923e-01
-3.51048931e-02 -2.20736135e-02 -8.31854403e-01 3.14659923e-01
1.90453738e-01 7.49115527e-01 -6.32458866e-01 -1.33504003e-01
1.99111715e-01 7.46687889e-01 -1.76427960e-01 5.87319434e-01
1.87632740e-01 2.74413582e-02 1.45699158e-01 4.25004512e-01
2.35524744e-01 3.53230029e-01 1.50838137e-01 2.66653169e-02
-3.24535459e-01 1.26892579e+00 -1.09373689e+00 1.22320879e+00
-9.60243940e-01 9.09078836e-01 7.99878657e-01 -1.10338473e+00
1.00496626e+00 1.07839537e+00 9.56808850e-02 -4.91350114e-01
8.47696513e-02 5.65705717e-01 1.49003655e-01 -2.27765784e-01
5.32899737e-01 -3.73473108e-01 2.25077704e-01 3.41429979e-01
2.66470253e-01 -2.29001060e-01 -2.95066625e-01 1.22320659e-01
1.14500797e+00 -2.59655237e-01 2.47426122e-01 2.51018167e-01
6.17299318e-01 -3.94727498e-01 2.28036642e-01 6.70359254e-01
-2.01632604e-02 8.30891490e-01 1.60603672e-01 2.09392592e-01
-9.74183500e-01 -9.09406722e-01 4.89119858e-01 1.32161224e+00
-7.67366290e-01 -5.10140479e-01 -8.87445688e-01 -3.78318012e-01
-4.53760177e-01 8.37913990e-01 -1.37422318e-02 1.40656814e-01
-1.03085899e+00 -2.08548754e-01 1.16260040e+00 5.48996091e-01
2.52872426e-02 -1.04506218e+00 -1.82534561e-01 5.40523529e-01
-1.87103450e-01 -1.27173209e+00 -7.13888645e-01 4.09493506e-01
-7.12511539e-01 -2.51487672e-01 -1.01642144e+00 -5.82403183e-01
9.51740891e-03 4.89778727e-01 8.97354484e-01 -3.45903814e-01
1.09749243e-01 3.44758332e-01 -5.34114659e-01 -3.93190503e-01
-9.31869745e-01 3.88953716e-01 1.95865050e-01 2.85301954e-01
1.85371339e-01 -5.79465032e-01 -1.91395491e-01 1.16165876e-01
-7.57090509e-01 -7.57772028e-02 7.21255481e-01 6.53935850e-01
2.19875723e-01 -6.35797158e-02 8.71163547e-01 -5.26236832e-01
7.03616202e-01 -2.27407783e-01 -5.47148943e-01 1.44266248e-01
-3.03389490e-01 -1.95292324e-01 9.65325296e-01 -5.70640147e-01
-1.03306377e+00 1.45722821e-01 -8.51741612e-01 -7.28460491e-01
-2.16838289e-02 3.48410785e-01 -2.88707018e-01 4.29006547e-01
2.72059679e-01 6.72281682e-01 -3.47995535e-02 -7.00186908e-01
4.53981131e-01 1.57993102e+00 3.60102832e-01 -6.49563447e-02
6.86905563e-01 -5.53177595e-02 -6.19358480e-01 -1.47753072e+00
-5.67347229e-01 -7.09398448e-01 -4.01906669e-01 -1.55368060e-01
4.90529418e-01 -1.18061185e+00 -7.72859335e-01 5.77382684e-01
-1.51363575e+00 -4.09929067e-01 -3.58519226e-01 6.58842742e-01
-4.80987400e-01 1.80978537e-01 -8.92218888e-01 -1.50617456e+00
-8.40839028e-01 -1.06031287e+00 1.33680511e+00 -1.35666370e-01
-2.36850172e-01 -7.16622114e-01 -4.92302775e-02 4.50195938e-01
8.13149393e-01 -8.12148333e-01 2.64487654e-01 -1.13571680e+00
-2.63172925e-01 -6.76255673e-02 9.95256677e-02 8.84607077e-01
3.03901155e-02 -9.46255103e-02 -1.67736125e+00 -3.48764539e-01
4.40980911e-01 -2.36831494e-02 8.97784054e-01 3.96762103e-01
6.82741582e-01 -6.77166462e-01 5.19896708e-02 3.46517980e-01
8.83456826e-01 3.40967864e-01 5.73543906e-01 -3.75233889e-01
5.74644387e-01 8.61447334e-01 2.37516940e-01 -1.57808326e-02
-1.91117059e-02 8.94498110e-01 -2.11829618e-01 -1.54410154e-01
-5.52505076e-01 -1.79424673e-01 1.36368656e+00 1.72447360e+00
1.42968744e-01 -5.78576207e-01 -8.29650164e-01 6.38452947e-01
-1.57413280e+00 -9.88153279e-01 1.42362073e-01 2.27228904e+00
8.69882762e-01 8.36184435e-03 4.41987723e-01 4.96723533e-01
8.06791544e-01 2.96668410e-01 -2.58970588e-01 -6.93768561e-01
-4.05223519e-02 4.00067121e-01 2.82421857e-01 8.36596608e-01
-6.93180084e-01 9.89552438e-01 5.84393644e+00 1.14234626e+00
-1.59309804e+00 4.79309708e-01 3.30214143e-01 -5.64435303e-01
7.96357915e-02 -3.48522663e-01 -1.04043353e+00 1.09923750e-01
1.93081164e+00 7.65527412e-02 7.87662625e-01 7.15102792e-01
5.30094326e-01 4.60229158e-01 -1.32634401e+00 1.23035192e+00
9.51703414e-02 -7.56210089e-01 -1.28600076e-01 -1.37919500e-01
-1.32617906e-01 3.09366465e-01 7.51129398e-03 4.66299027e-01
9.26557481e-02 -1.01948059e+00 1.03302789e+00 6.50752708e-02
9.32689190e-01 -6.80100083e-01 5.94313860e-01 4.64368045e-01
-1.49276900e+00 -1.41385794e-01 9.63561609e-02 9.16953310e-02
3.64874512e-01 7.65476346e-01 -1.44704509e+00 5.36954224e-01
2.79903889e-01 2.67133981e-01 -9.77618247e-02 5.47155440e-01
-3.37035716e-01 1.29318571e+00 -3.93970102e-01 4.28402126e-02
2.27840140e-01 7.93659538e-02 9.76044655e-01 1.90671098e+00
3.13736528e-01 -7.64368773e-02 -3.10315102e-01 5.69325626e-01
-6.58706874e-02 1.41385123e-01 -5.63488066e-01 -3.80134523e-01
5.67317128e-01 1.17706275e+00 -5.31799316e-01 -5.10803521e-01
-3.52342784e-01 1.11743093e+00 1.96156040e-01 4.49130684e-01
-5.87684274e-01 -6.53388858e-01 5.55523753e-01 1.28832281e-01
4.82223153e-01 -4.43367451e-01 -1.62425991e-02 -1.16929317e+00
1.15498595e-01 -1.28925788e+00 -1.95782706e-02 -6.23415887e-01
-1.06527591e+00 9.93248940e-01 -4.17792380e-01 -8.85786593e-01
-6.37245655e-01 -4.56891865e-01 -4.27316636e-01 1.05639732e+00
-1.38722456e+00 -1.07425702e+00 3.93103302e-01 4.77653116e-01
1.18660879e+00 -1.31876573e-01 6.57418668e-01 5.43889225e-01
-6.87100351e-01 7.69519925e-01 -4.31132801e-02 4.73806143e-01
4.19214278e-01 -1.11015582e+00 7.88100660e-01 1.14795804e+00
4.97736216e-01 6.99773371e-01 4.54219103e-01 -3.79035234e-01
-1.56719172e+00 -1.02834058e+00 1.16490221e+00 -1.39736950e-01
5.77387214e-01 -1.03096008e+00 -8.95052075e-01 7.70608246e-01
2.89200813e-01 -2.36943498e-01 4.72595334e-01 8.44975933e-03
-3.64314944e-01 -3.97447795e-01 -5.75657308e-01 5.01430154e-01
9.24503505e-01 -1.22653699e+00 -6.67768478e-01 5.11406548e-03
1.11447132e+00 -2.44624734e-01 -5.55316627e-01 2.23403866e-03
5.36727428e-01 -7.25183904e-01 8.41701508e-01 -1.54983550e-01
-3.22909690e-02 -2.78941780e-01 -3.49729866e-01 -1.42180717e+00
1.10021017e-01 -1.24855208e+00 -3.08002055e-01 1.48823559e+00
6.17286563e-01 -8.07590067e-01 3.26611727e-01 2.49188855e-01
-4.23081309e-01 -4.63589847e-01 -1.19758809e+00 -1.05314159e+00
-2.95296907e-01 -6.99854791e-01 3.11745375e-01 4.74785715e-01
-2.14142144e-01 7.62971640e-01 -3.19359004e-01 3.50638062e-01
1.93401963e-01 -2.19017476e-01 6.05645716e-01 -7.29445875e-01
-7.16705024e-01 -5.45295835e-01 5.62455058e-02 -1.43335927e+00
2.52702326e-01 -9.24715996e-01 4.09706056e-01 -1.30402482e+00
-4.26755160e-01 -1.13173597e-01 -1.77692398e-01 4.57346350e-01
1.76935375e-01 -1.41405314e-01 3.12772185e-01 1.36848390e-01
-1.74894035e-01 6.09258115e-01 6.65442050e-01 1.61108505e-02
-4.61530328e-01 3.28241020e-01 -3.53860497e-01 4.79478985e-01
5.91282606e-01 -6.25127852e-01 -2.92890847e-01 -4.50833440e-01
-9.87608433e-02 4.62459564e-01 8.98690075e-02 -9.38938439e-01
5.91665566e-01 5.72545111e-01 -1.40928149e-01 -6.36983871e-01
6.73243701e-01 -6.61150396e-01 1.69038121e-02 2.67534554e-01
-4.95077699e-01 -1.06715031e-01 1.54517859e-01 1.14479303e-01
-5.02917588e-01 -2.70848334e-01 6.73642159e-01 9.73532423e-02
-1.73565950e-02 -1.71942025e-01 -9.24844384e-01 -1.96582362e-01
3.78946960e-01 -2.47334782e-02 4.31246571e-02 -7.05462396e-01
-7.63437569e-01 -4.31955159e-01 -2.63666540e-01 2.83877879e-01
7.48701036e-01 -7.93955684e-01 -9.77452278e-01 3.34401697e-01
-4.87469256e-01 -2.26017572e-02 8.35858807e-02 9.14519846e-01
4.59875464e-02 6.91198885e-01 6.27690554e-01 -3.89876157e-01
-1.62048280e+00 3.32033187e-01 5.83598793e-01 -2.07574651e-01
-6.29307449e-01 9.56237376e-01 2.99266167e-02 -4.27487016e-01
5.42604029e-01 -3.80413294e-01 3.76939960e-02 2.69696891e-01
6.18411601e-01 5.34052968e-01 5.59353948e-01 -7.65990436e-01
-3.61662865e-01 2.34040514e-01 -8.66058245e-02 -7.32340276e-01
1.38663507e+00 -1.31891489e-01 8.82187486e-03 8.82630110e-01
1.20158029e+00 4.34557438e-01 -6.17667615e-01 -2.83290714e-01
-4.70967516e-02 1.12967148e-01 4.04964089e-01 -6.57446861e-01
-9.72299099e-01 1.24245512e+00 2.85481870e-01 4.73866194e-01
1.33642614e+00 -2.08497662e-02 1.12345219e+00 4.76194501e-01
-5.52274985e-04 -9.95086372e-01 -1.42415777e-01 7.67281175e-01
1.02103877e+00 -7.41784871e-01 -5.85833490e-01 -3.00133228e-01
-4.82771099e-01 1.26685953e+00 1.27180919e-01 1.04534879e-01
5.71816742e-01 7.03836977e-01 2.85795212e-01 9.85404402e-02
-9.85067070e-01 -2.24786833e-01 1.91991001e-01 1.66181207e-01
6.64960086e-01 1.33451447e-01 1.69671610e-01 6.82333648e-01
-5.33758521e-01 -1.77036569e-01 4.16076332e-01 6.94428444e-01
-2.23395959e-01 -1.05260348e+00 -4.93770719e-01 3.86589207e-02
-7.82089412e-01 -6.02883756e-01 -5.71109891e-01 2.63187975e-01
-4.75827545e-01 1.42527056e+00 9.52582955e-02 -5.85885525e-01
4.69887495e-01 3.57090920e-01 2.29069054e-01 -9.18158591e-01
-1.23096776e+00 5.35670698e-01 3.73442858e-01 -5.10591388e-01
-2.49727935e-01 -5.18739462e-01 -9.94280457e-01 8.98978636e-02
-6.83679163e-01 2.50179946e-01 1.12154281e+00 9.92648721e-01
5.53924143e-01 7.54824817e-01 8.32653761e-01 -7.97127843e-01
-8.67804706e-01 -1.31481600e+00 -6.59203291e-01 -2.10655794e-01
9.87163603e-01 1.68722868e-01 -5.13347030e-01 7.05064833e-02] | [14.57767391204834, 6.4388275146484375] |
7323590c-2d46-481a-b743-3bca75d9e4c6 | learn-to-cluster-faces-via-pairwise-1 | 2205.13117 | null | https://arxiv.org/abs/2205.13117v1 | https://arxiv.org/pdf/2205.13117v1.pdf | Learn to Cluster Faces via Pairwise Classification | Face clustering plays an essential role in exploiting massive unlabeled face data. Recently, graph-based face clustering methods are getting popular for their satisfying performances. However, they usually suffer from excessive memory consumption especially on large-scale graphs, and rely on empirical thresholds to determine the connectivities between samples in inference, which restricts their applications in various real-world scenes. To address such problems, in this paper, we explore face clustering from the pairwise angle. Specifically, we formulate the face clustering task as a pairwise relationship classification task, avoiding the memory-consuming learning on large-scale graphs. The classifier can directly determine the relationship between samples and is enhanced by taking advantage of the contextual information. Moreover, to further facilitate the efficiency of our method, we propose a rank-weighted density to guide the selection of pairs sent to the classifier. Experimental results demonstrate that our method achieves state-of-the-art performances on several public clustering benchmarks at the fastest speed and shows a great advantage in comparison with graph-based clustering methods on memory consumption. | ['Xiaolin Wei', 'Pengfei Yan', 'Di Qiu', 'Junfu Liu'] | 2022-05-26 | learn-to-cluster-faces-via-pairwise | http://openaccess.thecvf.com//content/ICCV2021/html/Liu_Learn_To_Cluster_Faces_via_Pairwise_Classification_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Liu_Learn_To_Cluster_Faces_via_Pairwise_Classification_ICCV_2021_paper.pdf | iccv-2021-1 | ['face-clustering'] | ['computer-vision'] | [-1.22733675e-01 -3.26497853e-01 -1.95349947e-01 -5.51350713e-01
-4.76880759e-01 -2.59593785e-01 3.96248460e-01 -4.20262478e-03
-1.97777376e-01 3.92869055e-01 -1.80158406e-01 -1.31012291e-01
-4.26228166e-01 -8.00336599e-01 -3.00241202e-01 -1.06210756e+00
-3.91716436e-02 6.57705963e-01 1.99914247e-01 1.92449987e-01
1.25123158e-01 4.89640415e-01 -1.60827315e+00 -8.28754455e-02
9.29462612e-01 1.03558409e+00 -2.69428175e-03 -1.00000978e-01
-1.55330598e-01 3.88591558e-01 -1.81787997e-01 -6.60546541e-01
1.54238185e-02 -4.40699607e-01 -5.39743483e-01 4.40834433e-01
3.42670828e-01 8.17728639e-02 -2.49016792e-01 1.28314877e+00
4.38985586e-01 1.18342698e-01 7.12455928e-01 -1.48917019e+00
-2.78105557e-01 4.76874769e-01 -1.05757225e+00 2.76490711e-02
2.18252808e-01 -1.29896015e-01 9.56346154e-01 -9.49041903e-01
4.60769355e-01 1.37512314e+00 4.74788278e-01 3.65545243e-01
-1.14874518e+00 -9.85756516e-01 3.65447640e-01 5.16921639e-01
-1.89700425e+00 -5.87726772e-01 9.45502043e-01 -2.89549053e-01
3.05300772e-01 2.20819890e-01 4.91399765e-01 5.53229392e-01
-4.08438206e-01 4.77881074e-01 9.51773882e-01 -5.35037220e-02
2.98916429e-01 -1.11269660e-01 -7.79683962e-02 9.34525073e-01
2.02573195e-01 -5.22997737e-01 -3.99890810e-01 -2.68663257e-01
5.10045230e-01 2.02240318e-01 -4.30316538e-01 -4.89358574e-01
-7.68569827e-01 8.58575583e-01 5.45689464e-01 2.24940255e-01
-8.67798030e-02 -1.34371519e-01 2.87666649e-01 -2.70212125e-02
5.48500597e-01 -2.64231861e-01 -3.68362367e-02 2.08878860e-01
-9.19654369e-01 -2.79515475e-01 4.90756959e-01 8.79345596e-01
7.79405534e-01 -5.05187690e-01 1.41382366e-01 1.09328842e+00
5.07526517e-01 2.98072726e-01 1.60195783e-01 -7.49047041e-01
5.35310984e-01 9.13879454e-01 -4.41216290e-01 -1.51470363e+00
-3.46804172e-01 -2.09911183e-01 -1.28447068e+00 -2.91979581e-01
4.29811090e-01 1.15120962e-01 -6.90974176e-01 1.55494797e+00
7.40604222e-01 5.68248212e-01 -3.77933383e-01 1.05481267e+00
6.82766736e-01 4.53391492e-01 4.14077193e-03 -7.45506525e-01
1.38097191e+00 -6.96643472e-01 -5.52160084e-01 6.56933784e-02
4.92256135e-01 -6.27740145e-01 7.77605116e-01 2.99126118e-01
-6.34656787e-01 -5.12831807e-01 -7.49547720e-01 3.12605292e-01
-1.59767017e-01 1.30254373e-01 8.25882971e-01 7.52098918e-01
-9.77982521e-01 3.97878021e-01 -7.81714261e-01 -3.34783405e-01
7.17108011e-01 6.37601137e-01 -4.45379525e-01 -4.39197093e-01
-7.74848759e-01 6.33067936e-02 3.04778129e-01 3.27504933e-01
-3.19638699e-01 -3.35029215e-01 -7.73542762e-01 1.04400419e-01
7.80305266e-01 -2.59491593e-01 5.68471670e-01 -6.73453271e-01
-1.25082707e+00 8.05121601e-01 -3.15233141e-01 1.60561115e-01
4.08104867e-01 2.90862441e-01 -5.47134042e-01 3.59838814e-01
8.74848589e-02 5.32213032e-01 8.37228000e-01 -1.25071931e+00
-5.82091689e-01 -8.26640606e-01 -2.57852316e-01 4.24483716e-02
-7.35885262e-01 -9.86534730e-02 -1.17687130e+00 -4.00371641e-01
4.12772417e-01 -1.02026033e+00 -1.63768575e-01 6.94070235e-02
-4.97313738e-01 -6.99033022e-01 9.98543620e-01 -2.13095918e-01
1.31112158e+00 -2.38119793e+00 2.90198810e-02 6.37395978e-01
5.27900279e-01 2.19674453e-01 6.02053851e-02 1.75626844e-01
1.86098859e-01 5.08431420e-02 -3.36376429e-01 -3.59373093e-01
-2.48929992e-01 -1.58535764e-02 1.72673896e-01 7.13336110e-01
6.49230778e-02 4.86021280e-01 -8.33668470e-01 -1.02716184e+00
7.21040592e-02 4.81139868e-01 -5.90556204e-01 2.11898476e-01
-2.92476714e-02 5.28303623e-01 -6.82419121e-01 4.10215169e-01
1.01252663e+00 -6.11223996e-01 8.28464329e-01 -3.04472625e-01
3.82718951e-01 -2.46272564e-01 -1.30513203e+00 1.52105999e+00
-1.61597446e-01 2.07219079e-01 3.95300239e-01 -1.24200439e+00
8.56677532e-01 6.88949078e-02 6.37308955e-01 -2.72309273e-01
2.40899935e-01 6.81012198e-02 7.92712420e-02 -3.66997123e-01
-8.32350031e-02 7.73806423e-02 2.23965332e-01 3.74288231e-01
-1.92303449e-01 3.16841900e-01 4.59460378e-01 4.39880341e-01
8.41472566e-01 -1.85614586e-01 1.03248201e-01 -4.02178377e-01
9.34664190e-01 -4.74090427e-01 7.38765299e-01 4.99821492e-02
-1.00260153e-01 4.44955856e-01 6.45682454e-01 -2.76382625e-01
-3.25385064e-01 -8.07914019e-01 -7.45123923e-02 7.55330205e-01
4.51670527e-01 -7.67708957e-01 -9.97129619e-01 -1.06423712e+00
2.81778872e-02 -1.50525635e-02 -5.30433357e-01 -4.72415015e-02
-4.63949025e-01 -1.03473353e+00 1.29296333e-01 3.57147753e-01
5.51998615e-01 -6.85371995e-01 2.72300635e-02 -8.20273384e-02
-3.09354424e-01 -1.29690731e+00 -6.81728363e-01 -3.15901756e-01
-7.85264730e-01 -1.53894949e+00 -4.27701861e-01 -1.08576500e+00
1.12486899e+00 6.54272139e-01 8.68922353e-01 6.19050503e-01
-2.59987503e-01 2.33952865e-01 -2.61144996e-01 2.84720659e-01
2.62980968e-01 1.29918918e-01 1.41427711e-01 5.66913486e-01
6.06378138e-01 -7.18948126e-01 -7.52296388e-01 6.98006094e-01
-6.81509137e-01 -1.69403657e-01 5.20039618e-01 6.95269108e-01
6.32724643e-01 5.78622162e-01 7.00925171e-01 -1.10446191e+00
3.68561447e-01 -5.19045413e-01 -6.30402803e-01 3.91206950e-01
-6.33333802e-01 -6.60318360e-02 7.55606472e-01 -3.14700037e-01
-9.42796648e-01 2.77401537e-01 8.70829746e-02 -4.48665261e-01
-1.06315746e-03 4.11789477e-01 -7.78011799e-01 -1.39764205e-01
-2.50827894e-02 1.66594908e-01 5.58879785e-02 -3.85144681e-01
2.80975759e-01 7.37700105e-01 2.25715250e-01 -6.01420581e-01
8.48165631e-01 6.04604602e-01 2.68465340e-01 -8.60616028e-01
-5.39868772e-01 -6.03271723e-01 -7.89824426e-01 -3.03574234e-01
7.77634263e-01 -8.22664261e-01 -1.12091458e+00 2.27557704e-01
-8.24868083e-01 7.07807466e-02 4.86080468e-01 3.93409342e-01
-9.01512653e-02 6.93360925e-01 -5.65366685e-01 -8.03140759e-01
-3.98308933e-02 -1.18429220e+00 1.03921199e+00 3.07116508e-01
2.72016019e-01 -7.46604085e-01 -2.31750131e-01 6.21281087e-01
-5.40589429e-02 4.64586727e-02 9.96730506e-01 -4.27381665e-01
-6.62367404e-01 -4.35224138e-02 -5.62349141e-01 -9.73343179e-02
3.69029939e-01 4.69236961e-03 -7.08950996e-01 -4.10089880e-01
-2.48650834e-01 -2.55322784e-01 8.42920780e-01 7.81491399e-02
1.66134977e+00 -4.92616445e-02 -7.54830420e-01 5.43407977e-01
1.26405001e+00 2.26704217e-02 3.91169608e-01 -3.45788240e-01
1.04451096e+00 8.57138753e-01 6.10336959e-01 4.93690670e-01
3.47317308e-01 7.50951886e-01 3.76402050e-01 1.19374797e-01
1.10376580e-02 -2.08148420e-01 -2.08785869e-02 9.76890624e-01
-8.16730335e-02 -2.40019858e-01 -7.24498272e-01 2.48929456e-01
-2.01925015e+00 -7.24110067e-01 -1.95161223e-01 2.34385681e+00
7.13791788e-01 5.63278235e-02 2.96927273e-01 1.68384016e-01
1.20172250e+00 1.74334079e-01 -4.90705997e-01 4.00432527e-01
3.05353343e-01 -1.33667141e-01 5.51054664e-02 2.14561269e-01
-1.09600520e+00 9.09933448e-01 5.11547375e+00 1.28616869e+00
-9.09019053e-01 1.08435705e-01 1.02098560e+00 -2.00624578e-02
5.31644858e-02 -2.27201432e-01 -8.52138400e-01 7.40860760e-01
6.26428723e-01 2.34452575e-01 4.53601927e-01 7.60804832e-01
1.03743374e-01 -5.65182902e-02 -1.00092196e+00 1.37565064e+00
1.69735342e-01 -8.06339741e-01 7.32180551e-02 3.07792872e-01
5.93488276e-01 -5.23974717e-01 6.87533000e-04 5.89558594e-02
1.08619682e-01 -9.09507811e-01 1.35362344e-02 7.28599951e-02
8.39236617e-01 -1.20313215e+00 5.98780394e-01 2.73959070e-01
-1.71008825e+00 6.35911971e-02 -5.84362268e-01 2.21153572e-01
3.96712078e-03 9.33277607e-01 -4.27876949e-01 6.00101292e-01
6.02248251e-01 6.74414814e-01 -6.04862869e-01 8.96808803e-01
-1.76470727e-01 4.94804561e-01 -3.78611803e-01 -2.00729333e-02
-6.73073754e-02 -6.88803911e-01 -6.84200041e-03 8.91293705e-01
2.74037987e-01 3.68468732e-01 5.33205688e-01 3.86766046e-01
-4.56477433e-01 5.12190402e-01 -5.05744517e-01 1.78714115e-02
6.82120740e-01 1.70871258e+00 -1.37526691e+00 -1.13876380e-01
-4.98412609e-01 8.79134119e-01 6.14256263e-01 9.96899009e-02
-8.92495632e-01 -2.77995139e-01 5.74292839e-01 2.88591385e-01
4.83429432e-01 -2.19372183e-01 3.19762319e-01 -1.05406821e+00
2.06647441e-01 -7.44721234e-01 5.74645281e-01 -2.40447894e-01
-1.46419859e+00 5.89625001e-01 -2.33717993e-01 -1.13499677e+00
1.04184277e-01 -4.59535450e-01 -5.98531783e-01 2.67384171e-01
-1.17620766e+00 -1.10232055e+00 -5.58219194e-01 8.74855101e-01
2.13332400e-01 6.86574206e-02 5.17141998e-01 6.13960385e-01
-9.61020052e-01 7.35171080e-01 3.72156166e-02 4.43128586e-01
8.02856982e-01 -9.06140566e-01 -1.22158259e-01 6.41492128e-01
2.54746765e-01 6.46737516e-01 9.33638513e-02 -6.53568089e-01
-1.57263565e+00 -1.20697820e+00 4.21636760e-01 3.48487198e-02
4.53353673e-01 -5.52427113e-01 -8.86547208e-01 2.33401656e-01
-1.42241523e-01 4.08097923e-01 9.52950954e-01 3.96370620e-01
-5.40866554e-01 -4.85021859e-01 -1.19753385e+00 5.50566971e-01
1.50654221e+00 -4.10782576e-01 4.66764979e-02 5.21412134e-01
3.06221455e-01 1.43449321e-01 -9.32440221e-01 3.32303047e-01
3.39595407e-01 -1.05337346e+00 8.59696209e-01 -2.51716882e-01
2.13768154e-01 -5.62046468e-01 1.10894889e-01 -1.04433250e+00
-3.13867301e-01 -5.46516061e-01 -4.04556841e-02 1.68243551e+00
1.61713734e-01 -7.65729547e-01 1.04491174e+00 4.45419580e-01
4.02351201e-01 -7.60550201e-01 -8.67386281e-01 -7.15937853e-01
-3.89077723e-01 -7.34659955e-02 6.95826232e-01 1.07727957e+00
8.40374008e-02 6.43576980e-01 -2.52013177e-01 2.72762924e-01
9.82604563e-01 4.51297224e-01 7.51129210e-01 -1.58108914e+00
-1.79511949e-01 -4.35609251e-01 -7.01842666e-01 -9.84835684e-01
4.81195778e-01 -9.56302762e-01 6.60528056e-03 -1.27884126e+00
5.68343401e-01 -6.72562301e-01 -1.92135856e-01 1.94801152e-01
-4.95579958e-01 5.61040342e-01 -1.51814958e-02 3.33560348e-01
-1.17189634e+00 6.49567366e-01 1.07566130e+00 -1.94804445e-01
1.06196860e-02 -2.95355190e-02 -6.42454445e-01 7.04641223e-01
6.94152772e-01 -3.72778237e-01 -7.25250602e-01 -1.55227274e-01
-1.52071834e-01 -3.36366221e-02 -2.02589147e-02 -1.04246569e+00
4.21389282e-01 -1.11299224e-01 4.51394379e-01 -4.70865577e-01
3.55390489e-01 -1.06539130e+00 1.40329495e-01 3.14619869e-01
2.90294029e-02 -1.26142323e-01 -2.96205044e-01 9.24313486e-01
-2.97348589e-01 1.32908195e-01 8.53131890e-01 1.67296708e-01
-4.33576018e-01 8.61930609e-01 4.41877097e-02 1.26051724e-01
1.21333075e+00 7.33258426e-02 -1.12098694e-01 -3.18962812e-01
-5.81593513e-01 4.10886407e-01 5.31460583e-01 2.57646114e-01
6.34713948e-01 -1.40547729e+00 -5.17575979e-01 2.86166131e-01
1.28792927e-01 9.09931213e-02 3.19287568e-01 9.76034939e-01
-1.73424646e-01 1.47424534e-01 2.36077473e-01 -8.01550746e-01
-1.68083048e+00 8.95626664e-01 -3.34801152e-02 -6.52552098e-02
-4.63327259e-01 7.18784630e-01 3.53936404e-01 -1.95378795e-01
2.40474761e-01 2.95637459e-01 -2.99816906e-01 2.39544272e-01
6.20436609e-01 3.07667285e-01 4.90972772e-03 -7.36356139e-01
-6.11159384e-01 9.07221615e-01 -1.07865930e-01 3.29753667e-01
1.17282665e+00 -1.49817675e-01 -4.17459518e-01 2.13823412e-02
1.35546732e+00 1.12281136e-01 -1.08252668e+00 -2.76958555e-01
1.56899408e-01 -7.33616829e-01 -8.42127353e-02 -7.87332132e-02
-1.78116500e+00 8.99230063e-01 3.66046667e-01 3.03493172e-01
1.22821975e+00 1.52853459e-01 7.57863343e-01 3.36698145e-01
5.22882581e-01 -9.59951222e-01 3.48040983e-02 -1.00074276e-01
3.11542720e-01 -1.35540748e+00 2.78734207e-01 -1.17907643e+00
-3.48838836e-01 9.33522582e-01 8.12944174e-01 7.23880902e-02
1.04478014e+00 6.89768344e-02 -1.15337722e-01 -3.27433527e-01
-3.94753695e-01 -1.98682174e-01 2.15577260e-01 4.76429194e-01
3.27992380e-01 9.37881544e-02 -3.43717366e-01 3.24801475e-01
1.13329649e-01 -3.34034234e-01 -1.53985456e-01 5.36397159e-01
-3.84326845e-01 -1.23906040e+00 -2.50695854e-01 6.72783971e-01
-3.64866167e-01 1.48567617e-01 -3.76928657e-01 6.05905294e-01
6.65513799e-02 1.34187543e+00 3.84770297e-02 -4.41753685e-01
-1.35477945e-01 -1.17744647e-01 4.90689069e-01 -4.43752587e-01
-1.69517979e-01 2.21293584e-01 -1.48365974e-01 -6.30342185e-01
-5.67614317e-01 -6.41549647e-01 -1.30214047e+00 -5.74145019e-01
-4.78285313e-01 5.48846543e-01 4.19847757e-01 9.27162409e-01
4.97118950e-01 1.51917666e-01 9.81159687e-01 -7.42466986e-01
-2.60965765e-01 -7.62286901e-01 -7.46243417e-01 5.24918616e-01
-2.55752534e-01 -8.58181953e-01 -4.19816732e-01 -1.28935158e-01] | [13.467852592468262, 1.051177740097046] |
5553b5fa-0ef9-415d-a834-1b3c68d73e32 | trusted-multi-view-classification-1 | 2102.02051 | null | https://arxiv.org/abs/2102.02051v1 | https://arxiv.org/pdf/2102.02051v1.pdf | Trusted Multi-View Classification | Multi-view classification (MVC) generally focuses on improving classification accuracy by using information from different views, typically integrating them into a unified comprehensive representation for downstream tasks. However, it is also crucial to dynamically assess the quality of a view for different samples in order to provide reliable uncertainty estimations, which indicate whether predictions can be trusted. To this end, we propose a novel multi-view classification method, termed trusted multi-view classification, which provides a new paradigm for multi-view learning by dynamically integrating different views at an evidence level. The algorithm jointly utilizes multiple views to promote both classification reliability and robustness by integrating evidence from each view. To achieve this, the Dirichlet distribution is used to model the distribution of the class probabilities, parameterized with evidence from different views and integrated with the Dempster-Shafer theory. The unified learning framework induces accurate uncertainty and accordingly endows the model with both reliability and robustness for out-of-distribution samples. Extensive experimental results validate the effectiveness of the proposed model in accuracy, reliability and robustness. | ['Joey Tianyi Zhou', 'Huazhu Fu', 'Changqing Zhang', 'Zongbo Han'] | 2021-02-03 | trusted-multi-view-classification | https://openreview.net/forum?id=OOsR8BzCnl5 | https://openreview.net/pdf?id=OOsR8BzCnl5 | iclr-2021-1 | ['multi-view-learning'] | ['computer-vision'] | [-3.83373767e-01 -1.29720837e-01 -4.25942779e-01 -7.20126748e-01
-1.23778033e+00 -6.42236412e-01 6.69019282e-01 3.48311305e-01
2.95510083e-01 7.05277443e-01 4.20330688e-02 2.51558214e-01
-3.56348932e-01 -8.40233207e-01 -4.79696065e-01 -1.14743674e+00
4.57354009e-01 4.89604175e-01 1.01616502e-01 2.88013816e-01
1.21234030e-01 1.23361990e-01 -1.62449336e+00 5.55810034e-01
1.13750219e+00 1.35089731e+00 -1.79887921e-01 3.46684664e-01
3.43107820e-01 6.40898764e-01 -4.93395180e-01 -8.38137746e-01
-2.45250240e-02 -9.86974537e-02 -3.24763000e-01 2.73656189e-01
1.96379907e-02 -3.56003463e-01 5.84290802e-01 1.26608419e+00
2.18711719e-01 -1.03606217e-01 1.03091133e+00 -1.45187199e+00
-3.98241967e-01 6.32201254e-01 -6.42104447e-01 -1.39079124e-01
2.72878528e-01 -1.96452841e-01 1.17536235e+00 -1.08074498e+00
3.01452667e-01 1.32249248e+00 3.70709300e-01 1.28640145e-01
-9.79940474e-01 -5.51898420e-01 4.52465713e-01 4.90994602e-01
-1.20752800e+00 -3.03718209e-01 7.62072206e-01 -5.18104017e-01
-5.95049635e-02 2.02228501e-02 5.06129682e-01 1.21940160e+00
6.09527469e-01 8.60485196e-01 1.54420173e+00 -9.29952711e-02
5.45899808e-01 6.02210283e-01 2.75896132e-01 4.07010674e-01
4.93049920e-01 4.12992109e-03 -6.82774007e-01 -4.07490671e-01
6.19678497e-02 1.32836431e-01 -3.18667531e-01 -7.21044958e-01
-7.66829133e-01 7.21965075e-01 1.59917012e-01 5.35066320e-05
-2.86946744e-01 -3.00058722e-01 6.18878424e-01 2.89652422e-02
9.07110631e-01 -4.45305288e-01 -4.12799865e-01 2.53276199e-01
-7.08562851e-01 -1.86181873e-01 5.93967259e-01 8.84251595e-01
4.09866989e-01 -6.99822754e-02 -1.14820391e-01 8.21211874e-01
8.11307013e-01 6.28666282e-01 1.10869847e-01 -9.18051362e-01
4.50619251e-01 6.20152712e-01 1.64357483e-01 -8.15807462e-01
-2.49481723e-02 -6.83449686e-01 -1.20961511e+00 4.80990559e-01
1.34838030e-01 7.38310367e-02 -3.62056434e-01 1.56069803e+00
6.76673174e-01 -3.75062460e-03 4.26626712e-01 7.17597961e-01
6.10998929e-01 4.79834020e-01 -1.60240501e-01 -4.76229370e-01
1.43779325e+00 -6.65425420e-01 -6.73051119e-01 3.91031384e-01
3.01187962e-01 -4.19083089e-01 5.06364763e-01 1.00310910e+00
-7.88740873e-01 -6.52263522e-01 -1.35710990e+00 5.12784660e-01
-2.05997393e-01 2.71268010e-01 1.12650879e-01 8.13973665e-01
-5.40949762e-01 5.03077745e-01 -8.31296802e-01 2.21815258e-01
4.11957234e-01 -9.40646790e-03 -4.68860298e-01 -3.42134506e-01
-1.13764071e+00 8.82365882e-01 4.21338588e-01 1.02691464e-01
-1.07057965e+00 -5.62945306e-01 -6.69403672e-01 1.39036834e-01
3.58346194e-01 -7.07234800e-01 8.92594755e-01 -6.82398379e-01
-1.35555351e+00 3.98350179e-01 -9.33150873e-02 -1.86263382e-01
7.91845620e-01 -2.00488433e-01 -4.58793670e-01 2.91979939e-01
7.10776597e-02 -2.34271288e-02 1.15697527e+00 -1.99277329e+00
-7.23216295e-01 -8.76805007e-01 -7.12310821e-02 2.76923627e-01
-1.85338199e-01 -2.91311204e-01 -3.21876228e-01 -3.49502206e-01
3.74294251e-01 -7.12892532e-01 1.58606678e-01 -6.90914318e-02
-1.71704873e-01 -3.49686623e-01 6.63091958e-01 -6.58258438e-01
9.71852064e-01 -1.92389750e+00 2.72206187e-01 4.80923593e-01
2.98861951e-01 -6.38942346e-02 4.69917476e-01 2.62677848e-01
1.68651491e-01 -3.24581331e-03 -9.08685178e-02 -4.05482292e-01
-1.79329947e-01 3.20104539e-01 -1.57607347e-01 7.20694780e-01
-1.10686578e-01 3.73505861e-01 -8.69871438e-01 -5.48343599e-01
2.11627334e-01 4.34899837e-01 -9.12927315e-02 1.13362543e-01
8.88636932e-02 5.78399599e-01 -5.15108228e-01 7.95204997e-01
9.63061929e-01 -4.69203740e-01 3.58325005e-01 -3.60121518e-01
3.77752095e-01 -3.08351696e-01 -1.41098464e+00 1.11437678e+00
-6.14258885e-01 -2.38613024e-01 1.73571840e-01 -1.00984216e+00
9.25553858e-01 5.65800488e-01 3.61197293e-01 -5.06758951e-02
1.39717191e-01 1.88748091e-01 -3.47898215e-01 -1.66626781e-01
6.07174076e-02 -2.57455289e-01 -1.72774762e-01 5.40291667e-01
2.04146653e-01 -4.76011708e-02 -2.06944525e-01 1.78104982e-01
3.64776254e-01 1.49627373e-01 6.19847715e-01 -1.19759835e-01
8.47143471e-01 -5.81805706e-01 7.53562331e-01 4.66062546e-01
-1.86407015e-01 3.87933940e-01 4.84605670e-01 -1.36306018e-01
-9.08742309e-01 -1.14321899e+00 -3.61863345e-01 7.51801372e-01
2.31977001e-01 -2.63716578e-01 -5.69309652e-01 -1.15129304e+00
5.21855522e-03 7.76424646e-01 -5.61232150e-01 -2.52271086e-01
3.48369390e-01 -6.42361403e-01 1.20560169e-01 6.16635859e-01
5.50130129e-01 -7.41824508e-04 -2.23796263e-01 -9.89356637e-03
-2.96521962e-01 -1.08487153e+00 1.60287060e-02 1.69098675e-01
-9.69061792e-01 -1.30224276e+00 -6.35986209e-01 8.02784339e-02
5.94440341e-01 3.18014175e-01 9.62805033e-01 -2.08318874e-01
5.55570662e-01 7.31058538e-01 -5.81232846e-01 -2.20448941e-01
-7.32348561e-01 -3.68820637e-01 4.38502103e-01 4.78695840e-01
-7.33256415e-02 -6.23756349e-01 -3.63685757e-01 5.92529655e-01
-8.04062307e-01 -4.44692634e-02 4.68363822e-01 1.00967228e+00
7.19280660e-01 3.62466365e-01 8.75001192e-01 -9.27119136e-01
2.40817174e-01 -5.86622596e-01 -8.01003397e-01 9.22184587e-01
-1.09871149e+00 -5.43993786e-02 6.99848771e-01 6.96970522e-02
-1.56794560e+00 -3.83206576e-01 1.45840183e-01 -6.91370726e-01
1.91444084e-02 6.04828715e-01 -7.36639440e-01 1.36499733e-01
3.48964036e-01 2.55883157e-01 -6.08498231e-02 -2.23910362e-01
4.82741803e-01 8.66209149e-01 1.89602345e-01 -7.55416155e-01
5.20764649e-01 6.95752978e-01 2.07629204e-01 -1.21343836e-01
-1.24059224e+00 -4.53791142e-01 -9.34246182e-01 -7.08501816e-01
5.34155130e-01 -1.29359043e+00 -8.11945081e-01 5.59776962e-01
-9.52720582e-01 5.79116702e-01 2.33060867e-01 6.04361236e-01
-5.19952238e-01 5.62428653e-01 -3.51467133e-01 -1.27703595e+00
-3.72497380e-01 -1.26271439e+00 1.14373291e+00 7.32169300e-02
3.38125229e-01 -1.12415671e+00 -3.36059868e-01 7.10330606e-01
-2.56692916e-01 1.08262539e-01 8.19214284e-01 -8.31694067e-01
-5.25422394e-01 -2.82571048e-01 -1.81626260e-01 7.59320796e-01
7.23404735e-02 1.06566384e-01 -1.32745504e+00 -3.87633175e-01
4.43360776e-01 -4.25755352e-01 9.40190375e-01 2.24726975e-01
1.25584233e+00 -1.64821610e-01 -2.36542776e-01 1.54326662e-01
1.39957690e+00 6.60394356e-02 1.71905011e-01 -1.11481786e-01
3.66450250e-01 7.29464591e-01 1.06956363e+00 1.02145410e+00
5.67028642e-01 4.30870086e-01 6.40738130e-01 6.86645210e-01
4.48808610e-01 -5.64161688e-02 4.69045788e-01 1.07745492e+00
-1.51493013e-01 -4.05601710e-01 -6.69884980e-01 1.34902984e-01
-1.90820479e+00 -1.01300883e+00 -5.00521511e-02 2.48060060e+00
4.87852871e-01 2.42032379e-01 -1.15660779e-01 3.05658281e-01
8.97696853e-01 6.86837360e-02 -6.02439046e-01 9.78241861e-02
-6.58490509e-02 -5.20271897e-01 5.70240840e-02 2.96304166e-01
-9.31331754e-01 1.63024873e-01 5.37179279e+00 1.11557198e+00
-6.04202151e-01 3.20135951e-01 1.00721979e+00 1.98602527e-01
-7.06315100e-01 8.86074156e-02 -1.05637980e+00 5.74220359e-01
6.63038194e-01 -2.07538202e-01 3.62385884e-02 9.72576976e-01
1.82009459e-01 -5.64612269e-01 -1.21442580e+00 8.57278109e-01
3.33323121e-01 -9.80288088e-01 9.97275040e-02 1.07981563e-01
8.70781243e-01 -4.35589284e-01 1.50807828e-01 8.26265663e-02
3.39741349e-01 -4.28736329e-01 7.79195070e-01 9.31832075e-01
5.57417572e-01 -1.07363009e+00 9.44163084e-01 6.94735885e-01
-1.20649529e+00 -2.81669766e-01 -2.61634320e-01 4.71176594e-01
6.10452630e-02 1.27895856e+00 -4.22683775e-01 1.41725802e+00
5.50803602e-01 9.51121092e-01 -4.59586561e-01 6.67083859e-01
-2.96406239e-01 4.08057511e-01 -6.91602528e-02 3.08308572e-01
-3.11278671e-01 -4.27471459e-01 4.40872937e-01 4.85852242e-01
5.68149209e-01 7.06527904e-02 2.88990855e-01 5.92263460e-01
1.44361988e-01 -1.43221289e-01 -4.78250563e-01 4.85999852e-01
6.67228997e-01 1.33683717e+00 -5.85625529e-01 -7.41317332e-01
-3.61981690e-01 5.40226519e-01 3.26167911e-01 1.15403019e-01
-9.45623755e-01 3.76421422e-01 1.90434307e-01 -5.46351373e-01
2.41194248e-01 6.28188178e-02 -3.89000088e-01 -1.56384957e+00
1.60409063e-01 -8.62126231e-01 6.56920075e-01 -8.87926936e-01
-1.91135347e+00 2.44107977e-01 2.46166199e-01 -1.89219666e+00
-1.46431595e-01 -5.46869218e-01 -3.43177944e-01 6.87302649e-01
-1.41547346e+00 -1.44252014e+00 -2.65059739e-01 5.55019200e-01
4.20661956e-01 -1.87828943e-01 6.24746919e-01 7.74901873e-03
-4.68978822e-01 4.71904695e-01 4.42235321e-01 -3.75768483e-01
9.60678995e-01 -1.27328467e+00 -5.46743274e-01 6.08072519e-01
-5.77298517e-04 2.72152573e-01 4.57793236e-01 -6.60235941e-01
-1.08569956e+00 -1.03813541e+00 1.29163995e-01 -5.55520952e-01
5.44405997e-01 -2.79235374e-03 -9.37316239e-01 3.30292255e-01
-4.28687781e-02 2.15950519e-01 1.19545603e+00 2.93171674e-01
-6.43957138e-01 -5.18845737e-01 -1.26647353e+00 7.62790293e-02
4.82830733e-01 -5.85599005e-01 -3.58873725e-01 1.82660654e-01
4.16717887e-01 -3.14093232e-01 -1.33270097e+00 6.23970628e-01
7.19310641e-01 -1.35274529e+00 7.71586359e-01 -2.85336137e-01
5.48908114e-01 -3.04785669e-01 -5.68048835e-01 -1.60082543e+00
-2.23076746e-01 2.83638626e-01 -4.13667619e-01 1.71955740e+00
2.09417447e-01 -7.99324155e-01 3.76717269e-01 5.09566426e-01
-1.92580316e-02 -8.49893570e-01 -9.97282028e-01 -7.79623032e-01
-1.60660483e-02 -6.74641848e-01 5.46122253e-01 6.93445206e-01
-8.99954215e-02 9.57057476e-02 -5.35449088e-01 7.36628413e-01
1.10860825e+00 4.36542571e-01 5.16263068e-01 -1.58420813e+00
-2.90354699e-01 -8.42132717e-02 -2.80376613e-01 -6.03072286e-01
2.52139956e-01 -7.66476750e-01 -4.46977243e-02 -1.32776177e+00
6.59179270e-01 -3.40641767e-01 -4.58424211e-01 4.22969880e-03
-3.73791218e-01 -1.56014591e-01 1.69240251e-01 4.51646596e-01
-8.27481568e-01 9.24480021e-01 1.20086396e+00 -6.43330365e-02
3.43102127e-01 5.07241011e-01 -7.07879424e-01 9.31638718e-01
4.42351401e-01 -3.04303914e-01 -5.84528327e-01 1.01674400e-01
3.52158040e-01 6.14283919e-01 5.20349443e-01 -1.03176153e+00
1.56338513e-01 -1.05571859e-01 6.50547326e-01 -9.81981933e-01
4.65484560e-01 -1.14662373e+00 6.02662489e-02 2.22236812e-01
-2.25730121e-01 -3.41820538e-01 -2.49750197e-01 1.38646686e+00
-4.08222347e-01 -2.64172614e-01 9.56055522e-01 -7.33749196e-02
-2.75123894e-01 1.90569431e-01 -1.51298895e-01 -2.09199995e-01
1.25711179e+00 1.60808817e-01 -3.00270408e-01 -5.81508100e-01
-1.02652180e+00 5.62708080e-01 3.28870416e-01 9.21041146e-02
6.85424924e-01 -1.64598191e+00 -5.91071069e-01 4.04204354e-02
5.36206186e-01 -1.30590379e-01 6.51206672e-01 8.95398974e-01
3.42612624e-01 1.69011816e-01 7.88090378e-02 -1.05655408e+00
-1.42510343e+00 6.99431360e-01 1.22384422e-01 -5.00390649e-01
1.97356343e-02 5.25544703e-01 5.41679896e-02 -3.88419002e-01
-1.50071522e-02 1.23544736e-02 -5.78654706e-01 6.51804566e-01
4.56464171e-01 6.33108139e-01 1.65783495e-01 -5.29492795e-01
-2.16219395e-01 5.30395627e-01 1.90908071e-02 -2.09878996e-01
1.12607968e+00 -5.87647736e-01 -1.92169592e-01 1.01228392e+00
8.24853897e-01 2.30809357e-02 -1.34885812e+00 -3.80879134e-01
-2.19812259e-01 -5.38297415e-01 4.04924601e-02 -9.18322206e-01
-9.66640770e-01 1.04667008e+00 5.13866007e-01 2.24623531e-01
1.00204098e+00 -6.77106306e-02 2.22087637e-01 1.40855968e-01
8.69815528e-01 -9.80752051e-01 2.21982569e-01 2.32873067e-01
8.27062726e-01 -1.56878221e+00 3.40424031e-01 -4.79739338e-01
-9.51902628e-01 1.25204289e+00 6.46117449e-01 3.67026746e-01
9.30527031e-01 -5.70717547e-03 -2.49430344e-01 -7.64760226e-02
-1.10509181e+00 2.58359849e-01 5.45736969e-01 5.12860060e-01
4.44404408e-02 1.84922546e-01 1.07461901e-03 1.00220084e+00
4.14048135e-01 -1.19131103e-01 4.12798405e-01 5.46195924e-01
-4.26819772e-01 -1.07054257e+00 -6.03026927e-01 4.96663779e-01
-2.81893879e-01 3.24036777e-01 -5.04403375e-02 2.63132423e-01
1.67582482e-01 1.19525480e+00 -5.13367116e-01 -4.39498693e-01
-4.44711708e-02 2.52797008e-01 2.97537059e-01 -5.83493829e-01
-1.85840458e-01 2.47364148e-01 4.15686220e-02 -2.91944057e-01
-7.49113023e-01 -8.86251986e-01 -7.91877806e-01 -1.07859336e-01
-7.37500191e-01 2.67199397e-01 5.85883260e-01 1.10696483e+00
3.68360370e-01 4.90248322e-01 1.20082057e+00 -6.11976624e-01
-1.16057658e+00 -7.41548955e-01 -9.10720229e-01 5.41281030e-02
5.05387336e-02 -9.96060312e-01 -6.08904123e-01 -8.56512934e-02] | [8.51975154876709, 4.5271124839782715] |
feee37ea-12dd-43d2-9913-0e1e0c3b7826 | the-architecture-of-a-biologically-plausible | 2306.15364 | null | https://arxiv.org/abs/2306.15364v1 | https://arxiv.org/pdf/2306.15364v1.pdf | The Architecture of a Biologically Plausible Language Organ | We present a simulated biologically plausible language organ, made up of stylized but realistic neurons, synapses, brain areas, plasticity, and a simplified model of sensory perception. We show through experiments that this model succeeds in an important early step in language acquisition: the learning of nouns, verbs, and their meanings, from the grounded input of only a modest number of sentences. Learning in this system is achieved through Hebbian plasticity, and without backpropagation. Our model goes beyond a parser previously designed in a similar environment, with the critical addition of a biologically plausible account for how language can be acquired in the infant's brain, not just processed by a mature brain. | ['Christos H. Papadimitriou', 'Daniel Mitropolsky'] | 2023-06-27 | null | null | null | null | ['language-acquisition'] | ['natural-language-processing'] | [ 1.33469924e-01 7.02795029e-01 3.09974760e-01 2.10311660e-03
2.17887312e-01 -3.95033002e-01 8.45477939e-01 3.15732569e-01
-5.79151809e-01 7.68682539e-01 4.34737414e-01 -2.43549511e-01
1.97594240e-01 -9.22486305e-01 -9.77458954e-01 -2.72180885e-01
-3.55954506e-02 3.92070025e-01 5.93518436e-01 -5.32360017e-01
3.69893193e-01 5.87682784e-01 -1.39709783e+00 3.52123976e-01
6.26067102e-01 3.28850120e-01 7.09969103e-01 5.33515930e-01
-3.97614360e-01 9.89228427e-01 -2.50016898e-01 -5.87563992e-01
-1.54787779e-01 -5.71293414e-01 -9.89186704e-01 -5.06289005e-01
-2.26737395e-01 -3.00355971e-01 -2.17269421e-01 8.19905102e-01
1.11844860e-01 -2.55520195e-01 6.90889060e-01 -4.67677712e-01
-1.24995828e+00 1.25486708e+00 -6.93662763e-02 1.62200809e-01
2.66995966e-01 1.77809432e-01 7.71262765e-01 -1.08976495e+00
5.47377884e-01 1.67766666e+00 5.34954309e-01 1.25343728e+00
-1.27651978e+00 -3.14102829e-01 1.62789688e-01 -2.41404623e-01
-1.06846082e+00 -4.23647434e-01 4.09125388e-01 -4.41759109e-01
1.57888639e+00 -3.78455341e-01 1.49926531e+00 1.08885813e+00
6.76610112e-01 5.97033024e-01 1.19581044e+00 -8.22320402e-01
2.56674647e-01 -4.18709032e-03 1.77249894e-01 8.69042814e-01
4.56628233e-01 4.19446319e-01 -9.84999478e-01 3.62231314e-01
1.21411228e+00 -3.49197388e-01 3.53482626e-02 1.09748989e-01
-1.19788456e+00 2.90019691e-01 2.98501104e-01 6.65318131e-01
-4.43682998e-01 6.62967861e-01 8.24194700e-02 2.53251076e-01
-1.00809157e-01 5.14532149e-01 -6.06310010e-01 1.90499634e-01
-5.68911731e-01 -3.16858530e-01 6.66501224e-01 6.22268796e-01
3.79533529e-01 3.84853125e-01 3.48413289e-01 6.39693797e-01
7.29568601e-01 7.39163876e-01 6.68901920e-01 -8.32408667e-01
-1.00069799e-01 1.75962508e-01 -2.34914035e-01 -2.47263849e-01
-3.73007566e-01 -1.87174007e-01 -2.10333183e-01 4.41794187e-01
5.16113937e-01 -1.37768194e-01 -1.04686511e+00 2.01491690e+00
-1.78632066e-01 -2.25036256e-02 4.62956071e-01 3.32419693e-01
7.46936679e-01 7.17421055e-01 7.15254545e-01 -7.97190741e-02
1.14862311e+00 -6.15534604e-01 -2.19878241e-01 -5.17104626e-01
2.50381202e-01 -3.06433529e-01 1.18865192e+00 5.07303119e-01
-1.78832686e+00 -4.03898984e-01 -1.20355582e+00 -2.71252871e-01
-3.40263456e-01 -6.28313363e-01 1.21618986e+00 6.97565794e-01
-1.72020888e+00 9.20069933e-01 -8.97788882e-01 -8.35450888e-01
4.03679460e-01 5.55888891e-01 -4.49227333e-01 2.89621532e-01
-1.01609421e+00 1.55568027e+00 6.02170408e-01 -8.36834218e-03
-1.08158052e+00 -3.91527116e-01 -6.46081924e-01 2.31944039e-01
-2.81675756e-01 -1.15508759e+00 1.48002398e+00 -1.23842287e+00
-1.70084310e+00 1.23293531e+00 -1.57427311e-01 -4.44059938e-01
-3.79213244e-01 -1.69373259e-01 -6.20905161e-02 3.38209957e-01
-5.13783574e-01 1.09873974e+00 4.68656391e-01 -1.49653423e+00
-1.95874721e-01 -2.71466255e-01 8.45569000e-03 9.53890607e-02
2.35238463e-01 3.67143184e-01 4.62575965e-02 -5.69761872e-01
2.53442258e-01 -4.65856940e-01 -1.48737267e-01 3.14070173e-02
2.26980537e-01 -1.82752520e-01 -5.44637680e-01 -7.01092303e-01
3.74841124e-01 -2.03138542e+00 3.57625484e-02 1.10969625e-01
3.20724249e-02 -1.10428035e-01 -4.66470391e-01 7.06361949e-01
-1.09144039e-01 4.03784573e-01 -1.84552759e-01 -1.51383840e-02
-4.16435488e-02 4.38891739e-01 -4.41847444e-01 1.40505144e-02
5.59757710e-01 1.32963133e+00 -9.42384064e-01 -4.37669784e-01
-1.86919495e-01 3.78673881e-01 -6.23532593e-01 -8.35929066e-02
-1.83396623e-01 4.66208369e-01 -2.81054169e-01 3.81065190e-01
7.99213722e-03 1.05373830e-01 2.06385776e-01 5.77428818e-01
-9.95410793e-03 7.54752219e-01 -5.62122822e-01 1.67927778e+00
-6.35226429e-01 5.49796522e-01 5.01515009e-02 -9.29002762e-01
6.60277009e-01 5.41917443e-01 -3.11432213e-01 -8.94152343e-01
3.29619586e-01 4.23064053e-01 6.75335944e-01 -3.69500667e-01
-2.00529188e-01 -6.61590040e-01 1.00143842e-01 6.03489459e-01
3.92454982e-01 -8.46567988e-01 1.56482279e-01 1.35635793e-01
8.60747159e-01 3.43441010e-01 3.37527812e-01 -5.54284930e-01
2.38700196e-01 -2.33840942e-01 3.16885024e-01 7.31769025e-01
1.01024531e-01 2.72694707e-01 2.03133807e-01 -2.48463765e-01
-1.09427345e+00 -1.75984085e+00 -1.20750174e-01 1.14457381e+00
2.29945350e-02 2.50260115e-01 -1.13073194e+00 1.61601558e-01
-3.10424238e-01 1.35029268e+00 -6.64048672e-01 -4.33328450e-01
-7.78492093e-01 -4.83466327e-01 4.61260051e-01 6.58331454e-01
2.37622149e-02 -2.19392323e+00 -8.94218087e-01 5.43741465e-01
6.28803492e-01 -6.38936460e-01 1.70014009e-01 4.20096695e-01
-1.31642330e+00 -6.33584321e-01 -3.19135427e-01 -1.31169128e+00
7.45770156e-01 -2.23903075e-01 1.30008698e+00 5.05552471e-01
-8.13746899e-02 5.10740757e-01 4.95572016e-02 -8.28510344e-01
-8.97347927e-01 -5.67769527e-01 1.22464083e-01 -8.67232203e-01
2.89426923e-01 -1.09662032e+00 -3.26217294e-01 -4.20159936e-01
-9.03860569e-01 5.29948235e-01 6.46343052e-01 7.91571259e-01
2.59449571e-01 -5.32778442e-01 7.08810568e-01 -6.22423649e-01
6.90063000e-01 -2.14969829e-01 -2.02870667e-01 4.14129525e-01
-2.61213243e-01 3.78457397e-01 6.01216376e-01 -3.56919140e-01
-1.28953135e+00 -2.32081473e-01 -5.87414086e-01 8.10219467e-01
-3.57464463e-01 3.32810283e-01 -6.26197457e-02 2.76823211e-02
7.31676459e-01 6.59313500e-01 -3.10512632e-01 -3.69008303e-01
6.70149744e-01 2.00686932e-01 7.53376007e-01 -9.17049766e-01
6.95811749e-01 2.00359046e-01 -1.63520515e-01 -8.37836146e-01
-2.74506330e-01 4.97031599e-01 -1.02005076e+00 2.04781726e-01
7.14283884e-01 -8.98133218e-01 -6.13571644e-01 5.44274569e-01
-1.62315536e+00 -8.17770362e-01 -9.30800736e-01 9.64202881e-01
-7.60317981e-01 5.39797591e-03 -1.12260127e+00 -5.88228583e-01
-3.20084989e-01 -5.36878705e-01 5.08707166e-01 4.62495744e-01
-4.13870066e-01 -9.99774694e-01 7.89730772e-02 -2.45039433e-01
2.86491930e-01 -2.82206595e-01 1.69813681e+00 -6.33869350e-01
-5.90543270e-01 3.08728278e-01 4.21345457e-02 3.13359499e-01
-1.41902074e-01 3.51429105e-01 -9.41469550e-01 1.93479985e-01
2.60111362e-01 -5.74957848e-01 1.11350536e+00 2.50457823e-01
6.26350522e-01 1.03223845e-02 -2.66759008e-01 4.44344997e-01
1.43356049e+00 6.88386440e-01 5.28652966e-01 -1.12004153e-01
1.33449554e-01 9.93511856e-01 -3.88915211e-01 -2.45717973e-01
4.25262094e-01 -3.12390804e-01 1.82722226e-01 7.62717128e-02
-4.75670636e-01 -5.20253003e-01 4.78978217e-01 1.55304921e+00
-3.54330897e-01 -6.48205802e-02 -1.04637814e+00 9.19912994e-01
-1.24345052e+00 -1.04354274e+00 3.62496942e-01 1.89037800e+00
1.29926789e+00 6.09855533e-01 -1.79817110e-01 1.49407893e-01
4.56260622e-01 -4.01365131e-01 -4.97698903e-01 -1.24248707e+00
-3.40055823e-01 1.15999484e+00 2.23421976e-02 7.21904933e-01
-1.31496251e-01 1.45111775e+00 8.70861435e+00 3.17256361e-01
-1.02484238e+00 6.44527152e-02 3.37184966e-01 1.45922542e-01
-8.27671826e-01 -7.09387213e-02 -5.35622656e-01 1.27073050e-01
1.29886568e+00 -2.15208888e-01 8.79811585e-01 2.97667950e-01
1.44175008e-01 -3.80006939e-01 -1.49049675e+00 4.30031627e-01
-1.06359586e-01 -1.23788190e+00 3.54771286e-01 -4.82731998e-01
3.45242381e-01 7.44142607e-02 -9.96116921e-02 1.62417859e-01
6.96791351e-01 -1.37477911e+00 1.18904042e+00 6.64103210e-01
4.94949907e-01 -2.25898147e-01 5.77755384e-02 6.35167599e-01
-7.64540493e-01 -1.74649537e-01 -3.68893057e-01 -5.61677158e-01
1.12371342e-02 -1.78156253e-02 -4.93306607e-01 -6.16680324e-01
3.31475496e-01 1.51827753e-01 -5.04655421e-01 9.13216710e-01
-7.99867749e-01 8.17709208e-01 -3.70294064e-01 -6.89140499e-01
-3.70727368e-02 2.91201711e-01 9.65610147e-02 1.29293156e+00
3.02917719e-01 6.27250910e-01 -7.22235978e-01 9.78943110e-01
8.75889435e-02 4.31305915e-01 -6.25444829e-01 7.59524032e-02
6.18009448e-01 7.55232871e-01 -9.81262445e-01 -2.95019269e-01
-4.44848031e-01 7.78128207e-01 3.66415501e-01 2.76451826e-01
-2.69249976e-01 -3.37527506e-02 1.71067536e-01 -7.18838796e-02
3.28064471e-01 -4.63852376e-01 -5.51947534e-01 -8.76151681e-01
-2.69203573e-01 -4.55229998e-01 -5.44317484e-01 -1.12251079e+00
-1.12564492e+00 4.60769981e-01 -5.12763560e-02 -4.26356234e-02
-3.28042895e-01 -9.72437561e-01 -8.93364966e-01 9.85716224e-01
-9.87298608e-01 -1.22974300e+00 4.47439015e-01 4.82705683e-01
4.21174407e-01 -5.53798117e-02 1.20393443e+00 -2.91034102e-01
-2.16560632e-01 2.22828075e-01 -7.37016648e-02 9.69676822e-02
-1.92196682e-01 -1.35846221e+00 1.03760278e+00 6.54269695e-01
5.03075123e-01 1.10651958e+00 6.62598014e-01 -5.08225381e-01
-1.01062870e+00 -3.06299597e-01 1.10891676e+00 -2.13354334e-01
6.51526392e-01 -8.08644474e-01 -8.99063289e-01 7.02943027e-01
7.70287991e-01 -3.12624305e-01 6.20189488e-01 -1.61748692e-01
-2.87402838e-01 -5.25474213e-02 -1.14259279e+00 9.04034555e-01
1.42860138e+00 -6.37342155e-01 -1.63026118e+00 9.89582837e-02
9.72894907e-01 2.65148342e-01 -5.57872534e-01 4.60592620e-02
9.82312143e-01 -8.32243681e-01 7.98093438e-01 -6.80477262e-01
3.70862752e-01 -6.60874173e-02 1.25936314e-01 -1.44091439e+00
-4.46896106e-01 -5.07469654e-01 1.66038156e-01 9.25165415e-01
7.21213162e-01 -7.10882783e-01 4.16253209e-01 6.04490995e-01
-2.16761693e-01 -6.71152115e-01 -8.91848087e-01 -5.02393246e-01
7.59539366e-01 -3.14938009e-01 4.37242746e-01 3.11297327e-01
3.28234553e-01 5.68494141e-01 3.15578461e-01 -3.61875951e-01
3.87109965e-01 -1.60787806e-01 5.68523491e-03 -1.41796970e+00
-3.88472468e-01 -6.63215876e-01 6.34272816e-03 -9.52752054e-01
1.27601057e-01 -9.16978002e-01 4.35258687e-01 -1.82992375e+00
2.23571941e-01 -1.04695857e-01 -4.49368715e-01 3.90700281e-01
2.32359722e-01 -5.25304936e-02 3.57412249e-01 -3.04760486e-02
8.53292346e-02 2.18572333e-01 1.37665021e+00 1.58293605e-01
-1.08813839e-02 -4.05512869e-01 -9.45981622e-01 1.50590968e+00
7.82002628e-01 -7.05142081e-01 -3.83338332e-01 -7.96564043e-01
6.23481154e-01 1.23897806e-01 3.04452002e-01 -9.83070374e-01
1.16289087e-01 -2.32108250e-01 6.82397962e-01 -2.84770042e-01
1.52471051e-01 -5.95524311e-01 -1.06993623e-01 9.58710253e-01
-4.45604265e-01 2.35074446e-01 4.69824344e-01 1.03548020e-01
2.65599012e-01 -6.29780412e-01 9.33891177e-01 -6.57398045e-01
-9.95223999e-01 -3.41440320e-01 -9.91388679e-01 5.70701696e-02
7.65787721e-01 -6.33385599e-01 -4.15888846e-01 8.72584730e-02
-1.24945581e+00 -1.94316268e-01 4.69419986e-01 5.52557595e-02
8.74569654e-01 -8.18495452e-01 -5.96544206e-01 1.75385818e-01
-5.59489727e-01 -3.53250474e-01 -3.44559222e-01 2.03985870e-01
-1.04481685e+00 4.57039088e-01 -7.50635207e-01 -1.90279894e-02
-4.86477673e-01 7.05154240e-01 3.77753437e-01 1.53165147e-01
-2.99114168e-01 1.22962141e+00 6.60777569e-01 -5.73770516e-02
-1.22125402e-01 -7.70238221e-01 -4.65699285e-01 -3.90581787e-01
4.47085083e-01 -2.88056374e-01 -6.47354007e-01 -5.91477096e-01
-3.59909743e-01 8.98939371e-01 3.10534745e-01 -4.86400932e-01
1.33664060e+00 -2.01799572e-01 -3.91225815e-01 9.68081534e-01
3.28784227e-01 3.59080106e-01 -9.73962426e-01 7.77271092e-02
3.55651453e-02 3.61391783e-01 -6.75649405e-01 -1.03921986e+00
-5.17355740e-01 1.40500104e+00 1.79149956e-01 4.29910123e-02
1.04180622e+00 2.85001129e-01 7.88345993e-01 6.55981541e-01
6.30824924e-01 -9.22893047e-01 1.95592597e-01 6.99918389e-01
1.08423066e+00 -6.15723550e-01 -2.79498160e-01 -2.39042059e-01
-2.96912193e-01 1.14915979e+00 5.47753572e-01 -5.76062620e-01
6.97429657e-01 5.10273993e-01 -9.30176005e-02 -4.90981489e-02
-1.14812779e+00 -2.59758651e-01 -2.28011794e-02 7.54941702e-01
7.45474994e-01 -1.08321115e-01 -5.37770569e-01 6.83018088e-01
-3.93548191e-01 -2.78490018e-02 5.61660290e-01 7.60929108e-01
-1.07644618e+00 -9.92138267e-01 -1.88453615e-01 5.63147962e-02
-7.55364597e-01 -7.86391973e-01 -4.25567478e-01 8.42632771e-01
6.94146931e-01 5.43682516e-01 1.22152708e-01 6.95852563e-02
1.53700396e-01 3.97258639e-01 1.26539230e+00 -1.30335510e+00
-7.41763294e-01 -9.89523977e-02 -4.81934398e-02 -1.96359754e-01
-4.88080919e-01 -7.04425156e-01 -2.04252100e+00 1.77138653e-02
1.10373408e-01 5.28495349e-02 6.46506190e-01 1.11747873e+00
-2.27084473e-01 3.70909870e-01 1.08539350e-01 -6.99761569e-01
-1.87129974e-01 -8.13102722e-01 -8.11471462e-01 9.93490890e-02
8.87213871e-02 -2.97382236e-01 -8.25559944e-02 3.99734735e-01] | [10.217816352844238, 8.661703109741211] |
cdab9bb4-5c89-4ff4-ab2b-938919c61b45 | random-walk-model-from-the-point-of-view-of | 1908.04333 | null | https://arxiv.org/abs/1908.04333v1 | https://arxiv.org/pdf/1908.04333v1.pdf | Random walk model from the point of view of algorithmic trading | Despite the fact that an intraday market price distribution is not normal, the random walk model of price behaviour is as important for the understanding of basic principles of the market as the pendulum model is a starting point of many fundamental theories in physics. This model is a good zero order approximation for liquid fast moving markets where the queue position is less important than the price action. In this paper we present an exact solution for the cost of the static passive slice execution. It is shown, that if a price has a random walk behaviour, there is no optimal limit level for an order execution: all levels have the same execution cost as an immediate aggressive execution at the beginning of the slice. Additionally the estimations for the risk of a limit order as well as the probability of a limit order execution as functions of the slice time and standard deviation of the price are derived. | ['Alexandre Argenson', 'Bruce Bland', 'Oleh Danyliv'] | 2019-08-12 | null | null | null | null | ['algorithmic-trading'] | ['time-series'] | [-5.01666486e-01 -1.30478367e-01 -1.19683087e-01 -1.33519005e-02
-4.46298532e-02 -6.71666980e-01 6.03784204e-01 5.80598056e-01
-6.55376017e-01 6.73168182e-01 -4.45229769e-01 -8.27687502e-01
-4.82412517e-01 -1.09903002e+00 -6.35581911e-01 -7.28506386e-01
-2.75875211e-01 7.79661655e-01 8.56086671e-01 -3.28722954e-01
4.50329036e-01 5.27075350e-01 -1.20276439e+00 -8.54730681e-02
3.83232921e-01 1.16347694e+00 -2.71092385e-01 7.59014487e-01
-3.50489169e-01 4.38858330e-01 -7.05107033e-01 -3.19970399e-01
6.97557151e-01 -5.35282493e-01 -3.08543086e-01 -2.67802596e-01
-5.94243467e-01 -4.76308316e-01 7.94817284e-02 1.07438409e+00
3.01393904e-02 5.62406778e-02 6.03913784e-01 -1.26696396e+00
1.25446260e-01 8.03117990e-01 -5.91027856e-01 7.29665518e-01
5.02681956e-02 -3.21636349e-02 1.13293648e+00 3.85081582e-02
1.82698518e-01 5.31032026e-01 1.26252010e-01 -5.04290089e-02
-1.17212713e+00 -3.25066954e-01 7.26869702e-02 7.68725425e-02
-1.08269405e+00 3.52262229e-01 4.41146761e-01 -2.77346671e-01
7.30177581e-01 4.14228141e-01 8.99296761e-01 7.31101856e-02
1.17824066e+00 2.60955691e-01 1.10116374e+00 -4.49449658e-01
6.64852440e-01 1.81237459e-02 3.57655406e-01 -2.17402205e-01
6.87021852e-01 4.40069616e-01 -3.27135861e-01 -4.26911980e-01
5.54620206e-01 -5.99066690e-02 -1.31017327e-01 -5.23805022e-01
-5.37969053e-01 6.82213306e-01 4.85013500e-02 3.70859534e-01
-6.10852957e-01 8.20727497e-02 3.38220507e-01 4.76954639e-01
1.66713357e-01 1.77147150e-01 -5.74189007e-01 -3.61037403e-01
-1.01426542e+00 9.71696615e-01 1.35850549e+00 3.47588271e-01
1.51217133e-01 -1.09521642e-01 -4.26883362e-02 -3.10877860e-01
2.70489275e-01 5.74926138e-01 2.63482749e-01 -5.31011760e-01
2.54378468e-01 3.16301018e-01 4.91419405e-01 -2.68673748e-01
-4.24968392e-01 -4.47853655e-01 -1.85005501e-01 5.65235138e-01
9.18202460e-01 -1.35360584e-01 -2.06920937e-01 1.05413926e+00
4.16367017e-02 -1.38702884e-01 -1.62716478e-01 4.70967054e-01
-3.49457949e-01 6.54195905e-01 -2.83673227e-01 -7.53695428e-01
1.17794764e+00 -1.93971321e-01 -7.52725780e-01 5.30946493e-01
1.40255019e-01 -9.31279004e-01 5.66729188e-01 5.79641998e-01
-1.30215573e+00 -9.27103218e-04 -1.05661070e+00 7.43381202e-01
-2.02093780e-01 -9.38020587e-01 1.58663124e-01 7.85377085e-01
-6.76576912e-01 9.33320165e-01 -1.02914584e+00 3.18053097e-01
-5.79629660e-01 2.17918485e-01 3.95003438e-01 6.43475831e-01
-1.03323793e+00 8.87621582e-01 2.38190308e-01 1.47690773e-01
-1.12683333e-01 -7.80470550e-01 -2.62728006e-01 3.68565172e-01
4.58701968e-01 -1.85284212e-01 1.67348659e+00 -3.09847206e-01
-1.44783807e+00 2.81610578e-01 2.23680228e-01 -9.96356606e-01
9.41643417e-01 -1.29773259e-01 -2.89271325e-01 -1.09388761e-01
-5.52713545e-03 -5.89452147e-01 4.42857325e-01 -5.53329885e-01
-5.66092908e-01 -2.82936037e-01 1.12402335e-01 1.01475567e-01
6.66647434e-01 -3.93673293e-02 5.30180633e-01 -1.49795055e-01
7.55201951e-02 -8.55517626e-01 -3.67678553e-01 -5.97079694e-01
-1.94498032e-01 -1.17169835e-01 2.24979550e-01 -3.40370804e-01
1.35303831e+00 -1.79195762e+00 -4.15514767e-01 6.40127659e-01
-2.96775013e-01 -2.30714217e-01 8.79891992e-01 7.41485536e-01
-1.94896564e-01 2.50825077e-01 -2.18477231e-02 6.77277267e-01
4.35671121e-01 -4.34752218e-02 -5.74144840e-01 5.66436768e-01
-2.97625780e-01 4.83332247e-01 -4.11790490e-01 3.77473146e-01
7.34655708e-02 -1.40302449e-01 -4.43657488e-01 2.73905769e-02
-3.87736410e-01 3.22766006e-02 -3.91700923e-01 -1.02021292e-01
7.85120428e-01 5.92612103e-02 1.17346592e-01 2.19582796e-01
-8.00298870e-01 3.66340250e-01 -1.34952772e+00 3.57882828e-01
-8.47206563e-02 2.31105849e-01 -7.08649158e-02 -5.23397744e-01
8.06589484e-01 2.24156633e-01 5.90897501e-01 -1.02986360e+00
1.66747957e-01 5.08924961e-01 4.53081965e-01 1.71594232e-01
4.39567536e-01 -5.40441811e-01 -2.26639345e-01 8.47010911e-01
-7.57513344e-01 -2.47570217e-01 4.94207472e-01 -1.29922647e-02
8.28136802e-01 9.46143717e-02 3.80012333e-01 -6.10842824e-01
2.84784347e-01 -1.81300625e-01 3.53917390e-01 3.73108357e-01
-7.63523206e-02 6.15041777e-02 1.30036736e+00 -7.11080194e-01
-1.12496495e+00 -1.17357183e+00 -3.55316609e-01 4.57312077e-01
4.37289566e-01 -3.75898987e-01 -7.20156789e-01 -1.22458264e-01
2.28234619e-01 9.97738957e-01 -2.84009188e-01 4.81835529e-02
-6.27227843e-01 -1.07207561e+00 -2.36999646e-01 2.00705063e-02
3.95820171e-01 -8.14921319e-01 -1.22457707e+00 2.88901210e-01
4.94383007e-01 -9.80918050e-01 -3.90869319e-01 4.10226494e-01
-6.71912551e-01 -1.37068462e+00 -3.30907822e-01 3.58068764e-01
1.31705806e-01 -2.42274970e-01 1.10192144e+00 5.54223359e-02
1.30115412e-02 8.99404362e-02 -5.44531271e-02 -7.81272054e-01
-6.00736737e-01 -1.80155024e-01 -3.47759500e-02 -4.41957600e-02
3.86713237e-01 -2.63200045e-01 -7.56561816e-01 3.15271884e-01
-1.05515242e+00 -5.83608210e-01 -1.10784806e-01 2.03859895e-01
5.27245939e-01 5.67448795e-01 2.35420167e-01 -5.30682325e-01
8.86985540e-01 -2.40434498e-01 -1.38988912e+00 -5.75742386e-02
-1.00359666e+00 2.49957129e-01 6.43364310e-01 1.60611913e-01
-7.53081203e-01 -3.86760831e-01 2.49447748e-02 3.42056841e-01
2.25324422e-01 2.52758801e-01 1.15377083e-01 2.86953479e-01
1.52385056e-01 2.06547379e-01 1.56263798e-01 -4.99589503e-01
-1.01157799e-01 6.93495423e-02 -1.84478059e-01 -4.14469063e-01
8.10462713e-01 3.15788776e-01 5.37335634e-01 -7.57563770e-01
-2.28783295e-01 -2.52291590e-01 -3.73179078e-01 -1.75795645e-01
6.98978126e-01 -1.82764411e-01 -1.24835658e+00 4.60259855e-01
-7.97062576e-01 -3.17823917e-01 -6.44409418e-01 3.78990561e-01
-7.08965480e-01 2.36758813e-01 -5.45347154e-01 -1.21710992e+00
5.08151799e-02 -9.62958217e-01 2.72463948e-01 5.42211294e-01
-8.72562155e-02 -9.36380982e-01 1.40713722e-01 -3.16581219e-01
3.61339450e-01 4.19612855e-01 9.95071471e-01 -9.33486044e-01
-8.87167871e-01 -4.32444721e-01 2.88922966e-01 1.99996710e-01
-2.41640151e-01 3.30073595e-01 -1.10717177e-01 -3.82391997e-02
6.05889320e-01 6.52294815e-01 2.98667401e-01 7.31228411e-01
2.40813643e-01 -2.27686867e-01 1.49167534e-02 1.91032980e-02
1.70190144e+00 6.58745408e-01 3.70164245e-01 7.70868897e-01
-1.93900615e-01 3.32049012e-01 6.73567116e-01 6.77379906e-01
-5.06854653e-02 7.25183010e-01 4.29227412e-01 5.30926406e-01
7.91192055e-01 -4.31214981e-02 4.15329784e-01 3.42081755e-01
2.91776136e-02 1.17876835e-01 -8.15363288e-01 7.07695112e-02
-1.45561874e+00 -1.04029155e+00 -4.70400274e-01 2.90044379e+00
2.95805603e-01 1.02974808e+00 5.76974750e-01 3.72207433e-01
5.55540621e-01 -1.75941340e-03 -2.75074214e-01 -9.68306422e-01
1.37903273e-01 2.73489565e-01 1.18939686e+00 7.41220891e-01
-4.70978677e-01 -2.77274642e-02 7.18312263e+00 5.50729632e-01
-7.37798333e-01 -2.51128107e-01 6.16111994e-01 -1.85544983e-01
-3.46291512e-01 2.50237226e-01 -1.12954342e+00 1.01151145e+00
1.31097913e+00 -8.49590480e-01 1.12729169e-01 6.00099206e-01
5.62038660e-01 -7.96786606e-01 -8.29474986e-01 2.51893818e-01
-7.07586110e-01 -1.06198323e+00 -3.11880350e-01 6.65831804e-01
2.25578353e-01 -2.33481258e-01 -6.32606596e-02 4.63805012e-02
-7.66875297e-02 -3.57794583e-01 9.08304393e-01 3.92221779e-01
-1.74122959e-01 -1.20574701e+00 9.63125229e-01 9.05422330e-01
-1.08896637e+00 1.28569096e-01 -1.46313220e-01 -4.61930066e-01
7.12787330e-01 7.54451096e-01 -2.69470334e-01 6.76028669e-01
3.89011621e-01 -4.41452950e-01 -1.35472938e-01 1.24178743e+00
-6.85741454e-02 4.14333642e-01 -8.36168945e-01 -2.55036712e-01
2.62946010e-01 -7.69184530e-01 6.12426162e-01 4.33129370e-01
1.79901570e-01 -1.39268294e-01 -5.11078164e-02 6.93605363e-01
6.30391896e-01 8.13960209e-02 -2.98685580e-01 1.79427728e-01
2.08548196e-02 6.07123315e-01 -1.27407837e+00 6.08739369e-02
-3.56701970e-01 4.96878475e-01 -6.28854275e-01 2.80261010e-01
-8.88963819e-01 -2.09422603e-01 4.58837003e-01 8.07770491e-01
3.99623811e-01 -3.96017432e-01 -3.27015549e-01 -6.38147712e-01
4.18572009e-01 -2.31025547e-01 3.14195186e-01 -6.66462630e-02
-8.64619613e-01 1.95139408e-01 2.92068928e-01 -1.05971003e+00
-7.17552423e-01 -6.09232783e-01 -7.60230780e-01 9.92395461e-01
-1.41000676e+00 2.33330652e-02 6.90833926e-01 1.57468855e-01
2.71621615e-01 8.65452178e-03 4.13959831e-01 -4.18709368e-02
-1.98099747e-01 -6.84784278e-02 3.18569779e-01 -1.89887851e-01
-1.93645060e-01 -1.52210653e+00 4.92506087e-01 7.59823143e-01
6.31222278e-02 4.51184928e-01 1.34567475e+00 -8.60129833e-01
-1.09114909e+00 -1.16221964e-01 8.30079079e-01 -1.87246025e-01
1.16395473e+00 -7.76475854e-03 -8.26681435e-01 2.75394678e-01
3.19953054e-01 -1.88157409e-01 4.68239456e-01 -2.85497546e-01
3.75808358e-01 -1.07403830e-01 -9.35622036e-01 3.85401517e-01
8.72918367e-02 -1.13900498e-01 -5.40252447e-01 2.64617294e-01
5.23416221e-01 -3.19538742e-01 -6.52890682e-01 1.13423936e-01
6.73747241e-01 -1.59321880e+00 3.74452353e-01 2.02053003e-02
-3.05190355e-01 -1.04285531e-01 1.13814577e-01 -9.49480891e-01
4.36715744e-02 -1.18395841e+00 -8.96525104e-03 7.28352129e-01
5.00561237e-01 -8.13925087e-01 7.79350877e-01 1.12954283e+00
6.43291712e-01 -8.36287141e-01 -1.35010636e+00 -1.21031344e+00
1.28585339e-01 -3.04748625e-01 6.53637648e-01 1.03780441e-01
-2.22947411e-02 4.53774668e-02 1.83536991e-01 -6.95659667e-02
7.63135433e-01 3.25135380e-01 2.73709118e-01 -1.09563398e+00
-5.90639651e-01 -6.80680573e-01 -3.23804438e-01 -7.34396815e-01
-2.38959163e-01 -3.39614332e-01 -4.08071220e-01 -1.10461974e+00
-2.23655105e-01 -1.00012384e-02 -5.32017231e-01 -6.45675421e-01
2.62395650e-01 -4.89878446e-01 3.78092021e-01 2.28082255e-01
-1.51750579e-01 2.63095126e-02 9.14153814e-01 5.79390347e-01
-4.21814173e-01 9.90926921e-01 7.46690482e-03 5.69031715e-01
7.53864944e-01 -5.27270913e-01 -3.69834632e-01 5.96473753e-01
7.14685082e-01 3.48885387e-01 9.69931111e-02 -6.72122657e-01
1.84086159e-01 -2.03499034e-01 -1.38301328e-02 -9.54140902e-01
-1.18017584e-01 -1.03147233e+00 5.21312714e-01 9.90338445e-01
-1.10420048e-01 8.06043684e-01 6.69890419e-02 6.75743699e-01
-1.00066133e-01 -8.29446852e-01 5.76313257e-01 -3.20727110e-01
-1.65390715e-01 8.39549601e-02 -4.87652808e-01 -6.29394427e-02
1.42464256e+00 -3.38996887e-01 9.66603458e-02 -3.58427197e-01
-7.44396925e-01 3.43853712e-01 4.44268942e-01 9.19806957e-02
-3.00951768e-02 -8.19791257e-01 -2.76360571e-01 1.88059196e-01
-6.28394842e-01 -4.35380667e-01 1.36108443e-01 7.46322215e-01
-1.03763330e+00 5.17318130e-01 -1.91029549e-01 -3.54082018e-01
-8.01315069e-01 6.50804758e-01 6.46686077e-01 -7.14169800e-01
-6.25386715e-01 9.22039002e-02 -1.42419077e-02 3.55898172e-01
-1.51998580e-01 -4.61358666e-01 1.57912195e-01 3.00422221e-01
7.66620696e-01 5.88895857e-01 3.12639356e-01 -3.11722606e-01
-1.64945260e-01 5.42338252e-01 -2.41956266e-04 -5.60398400e-01
1.07273972e+00 -9.36223045e-02 -2.69856900e-01 1.05261385e+00
5.10719955e-01 8.80922601e-02 -1.18139637e+00 3.30533981e-01
6.40818298e-01 -3.56355429e-01 -2.11338222e-01 -3.93375427e-01
-5.71821213e-01 4.85047847e-01 2.56780893e-01 1.33271372e+00
6.47432387e-01 -4.22906995e-01 6.56603456e-01 -8.91549960e-02
5.76947927e-01 -1.16125500e+00 -5.98284304e-01 3.55124205e-01
5.81688643e-01 -3.42500001e-01 1.02351390e-01 -3.47452313e-01
-4.68913823e-01 1.06094646e+00 3.44108837e-03 -5.10139763e-01
1.14871514e+00 7.34006822e-01 -1.95465267e-01 7.31445327e-02
-6.41121745e-01 1.52645260e-02 -1.79140538e-01 3.43921073e-02
2.95127273e-01 3.29179376e-01 -1.08617675e+00 3.97301555e-01
-3.89619738e-01 1.04854032e-01 1.03489435e+00 9.57592130e-01
-5.51244676e-01 -1.33504522e+00 -3.63299251e-01 3.80526155e-01
-9.12356317e-01 1.16154402e-01 1.90565921e-02 1.20242023e+00
-3.18260461e-01 4.12744373e-01 4.43232536e-01 2.77292252e-01
7.33403862e-01 4.29796875e-01 3.73590291e-01 -4.95845288e-01
-7.52548397e-01 3.55039090e-01 -1.89730823e-01 -5.02272904e-01
4.87477407e-02 -6.76790357e-01 -1.45612812e+00 -6.48212433e-01
-1.72865897e-01 6.90979123e-01 7.33419538e-01 9.54560399e-01
-2.25591809e-01 2.53064364e-01 5.42001903e-01 -3.92252058e-01
-1.20656800e+00 -3.10205579e-01 -1.38044083e+00 -5.03301620e-04
2.99922705e-01 -4.77985799e-01 -6.80957258e-01 -5.54312408e-01] | [4.755852222442627, 4.060290813446045] |
290b56cd-5026-4b5d-8412-319c75e3ff6f | going-full-tilt-boogie-on-document | 2102.09550 | null | https://arxiv.org/abs/2102.09550v3 | https://arxiv.org/pdf/2102.09550v3.pdf | Going Full-TILT Boogie on Document Understanding with Text-Image-Layout Transformer | We address the challenging problem of Natural Language Comprehension beyond plain-text documents by introducing the TILT neural network architecture which simultaneously learns layout information, visual features, and textual semantics. Contrary to previous approaches, we rely on a decoder capable of unifying a variety of problems involving natural language. The layout is represented as an attention bias and complemented with contextualized visual information, while the core of our model is a pretrained encoder-decoder Transformer. Our novel approach achieves state-of-the-art results in extracting information from documents and answering questions which demand layout understanding (DocVQA, CORD, SROIE). At the same time, we simplify the process by employing an end-to-end model. | ['Gabriela Pałka', 'Michał Pietruszka', 'Tomasz Dwojak', 'Dawid Jurkiewicz', 'Łukasz Borchmann', 'Rafał Powalski'] | 2021-02-18 | null | null | null | null | ['document-image-classification'] | ['computer-vision'] | [ 3.25855762e-01 4.37869668e-01 1.62121788e-01 -4.99142259e-01
-7.68798590e-01 -8.73123884e-01 8.56811762e-01 4.26600009e-01
-3.87952477e-01 1.27422586e-01 6.33721769e-01 -6.99648440e-01
1.39105290e-01 -7.56291628e-01 -9.23234582e-01 -1.91585824e-01
2.65885085e-01 6.64683104e-01 7.81931952e-02 -2.78950542e-01
5.73287427e-01 6.08160757e-02 -1.05807602e+00 1.01657176e+00
8.95471632e-01 8.73078644e-01 3.95948917e-01 8.17953646e-01
-6.21399641e-01 1.16202486e+00 -6.35489166e-01 -4.86120492e-01
-3.40385318e-01 -3.83316308e-01 -1.22689128e+00 1.50358543e-01
8.47341835e-01 -5.47374845e-01 -4.54447895e-01 7.94179022e-01
1.96774647e-01 -1.04680039e-01 7.66422868e-01 -9.38750803e-01
-1.22773206e+00 6.89719915e-01 -6.17796123e-01 7.63687491e-02
7.09711194e-01 1.59130916e-01 1.48317480e+00 -9.42409217e-01
8.25150073e-01 1.57930124e+00 1.96532965e-01 4.46769863e-01
-1.32633305e+00 1.56920373e-01 6.69746339e-01 4.44384664e-01
-9.54898834e-01 -2.57869422e-01 9.75421965e-01 -4.71908927e-01
1.08335638e+00 2.14038584e-02 5.04040122e-01 1.37216985e+00
1.95425987e-01 1.28192508e+00 6.42863750e-01 -6.48439288e-01
6.32183999e-02 -1.11984357e-01 2.95111239e-01 9.37668383e-01
-2.03204062e-02 -4.87839788e-01 -3.81616652e-01 4.36380684e-01
4.41803873e-01 -3.53670008e-02 -2.85628021e-01 -6.95065856e-01
-1.16922688e+00 7.31236219e-01 7.69317210e-01 1.09468222e-01
-2.26647794e-01 2.30624363e-01 6.73165679e-01 3.55043292e-01
3.50290418e-01 6.30139709e-01 -3.08113754e-01 2.09298581e-01
-7.70479620e-01 1.78593189e-01 7.66063213e-01 1.15604603e+00
5.99013865e-01 -1.10515557e-01 -5.84693134e-01 6.13552690e-01
4.49199289e-01 6.13561571e-01 1.50562808e-01 -5.58627248e-01
1.12686610e+00 8.88060451e-01 -1.47977263e-01 -8.39770794e-01
-4.59272444e-01 -4.05857682e-01 -7.05332875e-01 -1.02985233e-01
1.76661119e-01 1.32874370e-01 -1.14836168e+00 1.56425214e+00
-1.04768299e-01 -6.07857049e-01 5.23754694e-02 8.54281068e-01
9.92159724e-01 8.61649811e-01 7.45911943e-03 4.06547219e-01
1.65456331e+00 -1.44812548e+00 -8.63970101e-01 -6.75276577e-01
6.27642274e-01 -4.09398586e-01 1.62243259e+00 3.51645887e-01
-1.06670105e+00 -5.06890893e-01 -1.10403979e+00 -9.99302328e-01
-6.78008497e-01 1.65723607e-01 1.19478449e-01 1.45771250e-01
-1.12562764e+00 -1.02201492e-01 -3.63854140e-01 -3.24351251e-01
6.61149442e-01 -5.69478050e-03 -2.75629699e-01 -3.95370126e-01
-8.82211506e-01 9.98911917e-01 3.05684566e-01 1.72522634e-01
-1.08624959e+00 -5.99774301e-01 -1.30908430e+00 4.70010966e-01
7.77547181e-01 -1.05256343e+00 1.53191102e+00 -9.10764515e-01
-1.22978258e+00 8.95216763e-01 -3.44140232e-01 -4.21758741e-01
4.30102795e-01 -6.45501614e-01 -4.22196165e-02 5.10715008e-01
1.43982172e-01 7.86414564e-01 6.67616546e-01 -1.43994224e+00
-9.85366926e-02 -4.48005617e-01 4.25350517e-01 1.76873431e-01
-2.00973958e-01 -2.62493759e-01 -6.71851456e-01 -5.41337729e-01
-2.59083509e-01 -5.04125237e-01 -1.66447461e-02 1.21413425e-01
-6.84242189e-01 -3.08907866e-01 1.13864803e+00 -9.86949027e-01
1.12112784e+00 -2.05976510e+00 6.27010405e-01 1.17950469e-01
5.47832429e-01 -1.81635860e-02 -6.22933984e-01 8.57070506e-01
5.19165024e-02 1.62206694e-01 -2.99344569e-01 -3.82841557e-01
4.92040396e-01 9.93836951e-03 -7.14406908e-01 1.12746902e-01
4.81901824e-01 1.62949848e+00 -8.25110257e-01 -4.27705228e-01
2.16164127e-01 3.88643712e-01 -6.99013710e-01 4.62195039e-01
-1.08282340e+00 1.95726514e-01 -3.63005042e-01 5.49198449e-01
4.58055794e-01 -6.66490018e-01 2.70660609e-01 -2.44487256e-01
1.02919534e-01 5.09348035e-01 -5.50820291e-01 2.21304297e+00
-5.76835811e-01 1.07806253e+00 4.02454622e-02 -1.09121430e+00
7.04686761e-01 1.26020640e-01 -2.32345313e-01 -1.27086353e+00
2.85685301e-01 -3.19976091e-01 -2.50362813e-01 -8.56839955e-01
5.73913932e-01 2.73024440e-01 -3.54323566e-01 5.83339274e-01
2.12741837e-01 -1.04157135e-01 2.71070749e-01 7.05326498e-01
8.78995538e-01 3.28618973e-01 7.32903853e-02 -1.83806211e-01
3.80913049e-01 6.12217654e-03 -4.26752388e-01 6.90503657e-01
2.26820424e-01 5.70887625e-01 1.17444980e+00 -3.25457335e-01
-1.28742993e+00 -1.19624877e+00 4.45089608e-01 1.49740350e+00
8.55048094e-03 -6.60677433e-01 -8.71516526e-01 -8.70517254e-01
-1.35464653e-01 9.63169098e-01 -9.60492373e-01 7.55500495e-02
-6.39532149e-01 2.48104438e-01 4.60743874e-01 8.29797864e-01
2.93110073e-01 -1.23790038e+00 -7.85944223e-01 -4.70061190e-02
-4.02771950e-01 -1.34409058e+00 -4.74628001e-01 1.25384763e-01
-5.11498868e-01 -1.07234800e+00 -6.10050440e-01 -9.38361108e-01
8.07262778e-01 2.86535889e-01 1.59156144e+00 2.07860827e-01
-3.18955243e-01 5.52426517e-01 -4.56379116e-01 -4.07550216e-01
-2.26955935e-02 3.68774474e-01 -1.00527346e+00 -2.34996796e-01
1.65156350e-01 -2.37144411e-01 -5.63447297e-01 -4.07825083e-01
-1.22831964e+00 4.30113733e-01 7.66548932e-01 6.79986358e-01
3.06361049e-01 -5.20765185e-01 3.04282278e-01 -1.15942168e+00
8.02849233e-01 -4.00354922e-01 -5.84602475e-01 6.89636886e-01
-2.39663333e-01 4.75459874e-01 8.73829424e-01 -6.19084574e-02
-1.10564327e+00 -2.03948468e-01 -1.63484693e-01 -3.10862362e-01
-2.86060512e-01 8.03531587e-01 -4.28571880e-01 6.65407181e-01
4.19968545e-01 1.49510622e-01 -2.83861607e-01 -6.91903472e-01
1.03275478e+00 4.14897472e-01 7.72841275e-01 -4.97889131e-01
7.11222172e-01 4.63488758e-01 -1.82720497e-01 -9.34568107e-01
-1.14149213e+00 -2.78581172e-01 -1.09620714e+00 1.64612215e-02
1.13532376e+00 -7.05967307e-01 -7.86813796e-01 2.79977005e-02
-1.49559224e+00 -4.66395736e-01 -1.41369075e-01 -3.35217208e-01
-5.25386631e-01 4.31112498e-01 -3.97074640e-01 -5.47246456e-01
-4.63286132e-01 -9.34858859e-01 1.48322368e+00 -1.66121677e-01
-1.41865194e-01 -1.05156946e+00 -7.99524337e-02 4.63945687e-01
3.00617605e-01 3.57468009e-01 1.52153015e+00 -4.52469409e-01
-1.01767075e+00 1.25213386e-02 -7.01704621e-01 -2.24556863e-01
-5.01054704e-01 -1.41763225e-01 -8.79350781e-01 -2.12802246e-01
-2.94274598e-01 -7.49206960e-01 1.25707328e+00 1.51065066e-01
1.53682578e+00 -5.21047056e-01 -2.08065093e-01 5.37789285e-01
1.54916000e+00 2.61850227e-02 7.46602237e-01 3.70393902e-01
1.07052636e+00 8.19392502e-01 8.16886649e-02 1.67313829e-01
9.14873838e-01 2.96073020e-01 7.61419237e-01 -3.89438212e-01
-1.73617601e-01 -6.65100396e-01 1.27607390e-01 5.47055244e-01
6.08118713e-01 -7.29091585e-01 -1.11485553e+00 7.33573437e-01
-1.85048854e+00 -9.60222363e-01 7.63708167e-03 1.33427656e+00
5.41241586e-01 1.30419552e-01 -1.38257414e-01 -1.96634885e-02
6.80005997e-02 6.18225396e-01 -5.34788609e-01 -6.85575366e-01
-9.28439945e-02 8.95957947e-02 2.53989063e-02 6.25407577e-01
-1.03788245e+00 1.12056148e+00 6.35002756e+00 1.91617876e-01
-1.03605723e+00 -2.00397894e-01 4.53089207e-01 -3.89218889e-02
-6.85845196e-01 -1.05886377e-01 -2.95874327e-01 -1.41295373e-01
6.97842836e-01 2.06135958e-01 4.34538782e-01 6.62307918e-01
-8.74971598e-02 -1.18679032e-01 -1.47252715e+00 6.00811243e-01
6.33328438e-01 -1.44836307e+00 4.57282126e-01 -2.25006580e-01
3.38469982e-01 -1.28505990e-01 2.76686639e-01 3.17786008e-01
2.50096858e-01 -1.37857163e+00 1.03997982e+00 6.15047991e-01
1.02763617e+00 -5.47054470e-01 4.68165636e-01 3.49042892e-01
-1.08891428e+00 -1.23004965e-01 -1.15381435e-01 -4.89044823e-02
2.39527762e-01 1.97918177e-01 -7.12565005e-01 6.01269722e-01
4.74613637e-01 7.77896941e-01 -1.00279117e+00 6.71580911e-01
-6.92143977e-01 2.39318267e-01 2.83143252e-01 -2.66280591e-01
6.18630052e-01 1.51733264e-01 1.24841101e-01 1.51739287e+00
-1.36552647e-01 -8.23808461e-02 7.95128271e-02 1.09029269e+00
-3.03323984e-01 6.56307712e-02 -6.98186100e-01 -3.35997045e-01
1.11292832e-01 9.64646518e-01 -5.04418552e-01 -5.10840774e-01
-6.78442836e-01 1.09017622e+00 8.26350033e-01 6.75903261e-01
-6.00616813e-01 -5.76727033e-01 -1.48968473e-01 1.67372618e-02
7.11570621e-01 -3.69553834e-01 -5.95665216e-01 -1.34718227e+00
2.81765878e-01 -1.06624317e+00 4.30431902e-01 -1.25576627e+00
-1.03191173e+00 5.87235570e-01 -6.56969026e-02 -7.70148873e-01
-2.07873419e-01 -1.10255861e+00 -5.97883463e-01 7.06650853e-01
-1.67538846e+00 -1.56189263e+00 -4.28601980e-01 3.57508838e-01
9.30482447e-01 3.35644707e-02 8.03475738e-01 -1.40113547e-01
-3.90896082e-01 2.51604855e-01 -5.64065352e-02 3.67557079e-01
4.43090767e-01 -1.60014009e+00 7.68367171e-01 6.79981411e-01
3.14574450e-01 6.75012469e-01 6.08902872e-01 -3.31190228e-01
-1.81383336e+00 -9.25493240e-01 1.03005075e+00 -6.65319502e-01
6.44603670e-01 -1.10566247e+00 -8.94767702e-01 9.77085829e-01
1.12058461e+00 -5.09237230e-01 5.54252028e-01 3.81011516e-01
-1.03627503e+00 2.51735061e-01 -5.02829015e-01 7.87684500e-01
9.76055086e-01 -1.03970039e+00 -1.10620916e+00 2.84266412e-01
9.62043405e-01 -3.89448702e-01 -2.84700125e-01 -3.29193287e-02
5.69157898e-01 -7.21740603e-01 9.26797271e-01 -8.60557616e-01
1.13846946e+00 -1.69500127e-01 -6.29290715e-02 -1.33076465e+00
-2.91342050e-01 -1.74017236e-01 -2.24739388e-01 8.49343896e-01
5.89343011e-01 -8.91587045e-03 5.29873669e-01 1.94744736e-01
-1.71116710e-01 -7.31972635e-01 -4.41518903e-01 -3.08629364e-01
1.53061852e-01 -3.10801685e-01 4.05866116e-01 6.47087395e-01
3.17229480e-01 1.21776724e+00 -3.56923819e-01 -3.41258347e-02
3.69286060e-01 4.02160317e-01 8.63570571e-01 -1.08203125e+00
-1.83110654e-01 -6.02000892e-01 2.42194757e-01 -1.61457348e+00
3.97646189e-01 -9.06292319e-01 1.30432740e-01 -2.39013839e+00
4.63789225e-01 6.24042809e-01 1.34221949e-02 4.50601190e-01
7.63006434e-02 -1.20262899e-01 4.05364722e-01 -1.98111400e-01
-1.16667223e+00 6.49593234e-01 1.48025084e+00 -6.88148916e-01
2.61617243e-01 -7.03008175e-01 -8.66963685e-01 4.83871996e-01
4.54250902e-01 2.24183723e-02 -6.94191217e-01 -1.29200065e+00
5.33161819e-01 9.50306356e-02 6.42800927e-01 -4.73636806e-01
3.37561995e-01 9.20501873e-02 6.06129229e-01 -1.14684808e+00
2.05976486e-01 -7.64462411e-01 -8.99588346e-01 1.93975404e-01
-8.50808740e-01 4.83785868e-01 3.08919966e-01 5.43158531e-01
-3.12311262e-01 -2.44445875e-02 2.70596355e-01 -1.99598029e-01
-6.92035913e-01 -2.13630162e-02 -5.01467168e-01 3.82108629e-01
5.54715097e-01 5.12199067e-02 -1.00929093e+00 -6.31821215e-01
-5.68135560e-01 5.43139517e-01 3.88299465e-01 6.53787255e-01
8.05630088e-01 -8.88879180e-01 -6.29029453e-01 2.28876725e-01
5.45698941e-01 1.28323585e-01 1.70748398e-01 4.60356504e-01
-6.51078522e-01 9.43756104e-01 -2.05859587e-01 -5.27993143e-01
-1.00942636e+00 1.00286651e+00 1.68404162e-01 -4.23298240e-01
-7.47145832e-01 7.42812991e-01 7.75505424e-01 -3.17263395e-01
5.13066232e-01 -5.32977343e-01 -3.74438673e-01 1.83657303e-01
4.80767339e-01 -2.05038562e-01 -6.06914647e-02 -3.60598654e-01
-2.36482233e-01 4.94266272e-01 -2.44184032e-01 -2.88996816e-01
1.34781933e+00 -2.80373573e-01 -2.07549572e-01 4.58598703e-01
1.40575600e+00 -1.61067069e-01 -1.12387109e+00 -2.89144427e-01
3.35190624e-01 -5.76375239e-02 -2.39577353e-01 -1.10019040e+00
-6.11429334e-01 1.56873131e+00 6.57010823e-02 2.48730212e-01
1.08808112e+00 2.31807783e-01 6.18653476e-01 7.95412004e-01
-2.78914779e-01 -9.08087075e-01 7.21642911e-01 9.59608078e-01
1.41182435e+00 -1.26233065e+00 -8.55489373e-02 -1.34406760e-01
-8.98060262e-01 1.18477631e+00 7.40188718e-01 -9.09201801e-02
2.65316486e-01 2.97658503e-01 1.39635548e-01 -6.60976887e-01
-1.25173533e+00 -1.84268549e-01 7.08501697e-01 4.55485165e-01
5.57915986e-01 -1.26102284e-01 1.66611791e-01 4.68260884e-01
-2.06863239e-01 -3.47661644e-01 2.58244634e-01 9.52901721e-01
-5.49210370e-01 -7.51833975e-01 -1.04310550e-01 2.13635355e-01
-1.28370136e-01 -3.59035730e-01 -8.19793820e-01 7.69639134e-01
-4.10054624e-01 7.34441519e-01 3.50250870e-01 -3.31311226e-02
5.33318698e-01 2.47074947e-01 5.22112429e-01 -5.62666178e-01
-4.86416787e-01 1.44381076e-01 1.52221814e-01 -4.63295192e-01
7.26603568e-02 -2.34876722e-01 -1.11949575e+00 9.66572296e-03
1.08239204e-01 -1.33436486e-01 5.15803814e-01 1.05119491e+00
3.41894865e-01 9.57860291e-01 1.58986196e-01 -4.73516703e-01
-2.92845607e-01 -9.93220568e-01 2.94446647e-02 5.63317955e-01
7.47958422e-01 -2.09592581e-01 2.59080708e-01 2.27112547e-01] | [11.122445106506348, 1.996045470237732] |
7a62dd80-f69a-40c9-80f2-114acdb2bdc2 | real-time-audio-video-enhancement-with-a | 2303.00949 | null | https://arxiv.org/abs/2303.00949v1 | https://arxiv.org/pdf/2303.00949v1.pdf | Real-time Audio Video Enhancement \\with a Microphone Array and Headphones | This paper presents a complete hardware and software pipeline for real-time speech enhancement in noisy and reverberant conditions. The device consists of a microphone array and a camera mounted on eyeglasses, connected to an embedded system that enhances speech and plays back the audio in headphones, with a latency of maximum 120 msec. The proposed approach relies on face detection, tracking and verification to enhance the speech of a target speaker using a beamformer and a postfiltering neural network. Results demonstrate the feasibility of the approach, and opens the door to the exploration and validation of a wide range of beamformer and speech enhancement methods for real-time speech enhancement. | ['François Grondin', 'Jérémy Bélec', 'Amélie Rioux-Joyal', 'Olivier Bergeron', 'Félix Ducharme-Turcotte', 'Francis Cardinal', 'Étienne Deshaies-Samson', 'Anthony Gosselin', 'Jacob Kealey'] | 2023-03-02 | null | null | null | null | ['face-detection', 'video-enhancement', 'speech-enhancement'] | ['computer-vision', 'computer-vision', 'speech'] | [ 3.61116320e-01 -1.52816884e-02 8.65486443e-01 -2.15325758e-01
-5.63165605e-01 -3.65605205e-01 3.04795325e-01 -1.91979215e-01
-5.45255125e-01 2.49258503e-01 2.84518421e-01 -5.42949021e-01
6.21000640e-02 -2.30363861e-01 -2.91787297e-01 -7.72844374e-01
-1.36170417e-01 -4.10590380e-01 3.07833225e-01 -5.22177406e-02
1.20044369e-02 6.93519950e-01 -1.99839783e+00 1.54687777e-01
9.04629007e-02 9.18242097e-01 4.61945266e-01 1.35989118e+00
5.42256355e-01 2.39578009e-01 -1.09512866e+00 -1.82868406e-01
-2.06443742e-02 1.65802706e-02 -1.23767726e-01 3.37934315e-01
5.39661050e-01 -6.50287211e-01 -3.21208298e-01 9.42958534e-01
1.22913098e+00 2.38581598e-01 -2.78378092e-02 -9.80336189e-01
-7.07637370e-02 2.47959837e-01 -2.30070740e-01 4.76077139e-01
8.69204521e-01 1.95223749e-01 1.89543471e-01 -1.21852231e+00
2.07742497e-01 1.19731236e+00 8.71562719e-01 5.47887146e-01
-8.90018463e-01 -7.86014438e-01 -3.32136720e-01 5.60871186e-03
-1.22845864e+00 -1.27522027e+00 6.41173422e-01 -1.39310688e-01
1.47798347e+00 5.28235495e-01 6.16621017e-01 1.01951408e+00
8.66308808e-02 3.17245811e-01 1.23782575e+00 -5.89246452e-01
4.18108672e-01 6.07750773e-01 1.59694552e-01 2.22794384e-01
-2.09594086e-01 7.38878906e-01 -6.55027866e-01 -2.42734507e-01
5.54333389e-01 -3.59486490e-01 -5.94475567e-01 3.75510067e-01
-6.27818942e-01 2.63428897e-01 -6.87174127e-02 4.88209993e-01
-6.72407746e-01 -2.73046523e-01 3.34696621e-01 2.46499211e-01
6.08812928e-01 1.02076136e-01 -2.46010035e-01 -1.20976217e-01
-1.23992312e+00 -4.28786650e-02 7.60878563e-01 7.18881071e-01
-2.34669857e-02 3.44161749e-01 3.66094299e-02 6.38199329e-01
8.57958138e-01 7.47185588e-01 2.56085038e-01 -5.88310540e-01
9.42676067e-02 -1.84166849e-01 4.42130454e-02 -6.94133401e-01
-4.54038382e-01 -4.97243822e-01 -3.78294289e-01 8.40183616e-01
-1.03103943e-01 -5.96291959e-01 -6.70401275e-01 1.21849346e+00
7.55521297e-01 7.28130579e-01 -8.86356533e-02 9.61381137e-01
7.73941278e-01 7.07736850e-01 -8.56080875e-02 -5.94781399e-01
1.51211190e+00 -6.30134225e-01 -1.13843775e+00 -9.42944176e-03
-1.82099953e-01 -1.37464094e+00 7.87279487e-01 6.77865446e-01
-1.43951106e+00 -8.72459650e-01 -1.14933658e+00 1.86869279e-01
-2.39904717e-01 6.18960261e-02 4.56461608e-02 1.54750383e+00
-1.68325746e+00 1.12910062e-01 -7.14754164e-01 -2.41459385e-01
-1.95682734e-01 8.09716403e-01 -4.59236711e-01 3.97075653e-01
-8.83158386e-01 7.87801206e-01 -8.50316286e-02 5.68190575e-01
-6.99210167e-01 -5.80044746e-01 -9.33564186e-01 1.64802760e-01
-3.92920971e-01 -6.19416296e-01 1.48621976e+00 -9.66063321e-01
-2.14686680e+00 6.25269055e-01 -3.20949674e-01 -2.57927656e-01
2.56409973e-01 -3.40749323e-01 -9.60935295e-01 1.78153411e-01
-5.61521709e-01 5.06724156e-02 1.47715831e+00 -8.40469003e-01
-7.39131808e-01 -4.51124609e-01 -3.37506384e-01 7.76673257e-02
-5.15964091e-01 9.34536159e-01 -1.45786434e-01 -6.21851444e-01
-9.20621157e-02 -3.62372249e-01 1.87500119e-01 -2.52245635e-01
7.04185143e-02 3.07839006e-01 1.23451602e+00 -1.22055554e+00
1.35766554e+00 -2.51069164e+00 -4.01253253e-01 2.00692698e-01
-7.87514448e-02 7.52668440e-01 -5.54783940e-02 1.40937835e-01
-3.96081418e-01 -4.74594086e-01 2.46991828e-01 -8.10515642e-01
-1.26804754e-01 -4.08594131e-01 -2.56674528e-01 7.42217720e-01
-2.29606569e-01 8.64090472e-02 -5.13632774e-01 -7.52547011e-03
7.74781704e-01 1.34663439e+00 -2.92674601e-01 6.05562568e-01
6.41008317e-01 3.47475678e-01 2.53048301e-01 4.17773575e-01
1.14492023e+00 6.53642118e-01 -1.76508069e-01 -8.68616477e-02
-5.62982142e-01 5.37752628e-01 -1.81691182e+00 1.06567156e+00
-6.73541188e-01 9.87393141e-01 1.17589355e+00 -2.43252203e-01
7.91257918e-01 1.06304622e+00 -1.40868574e-01 -4.23307180e-01
4.00619388e-01 2.50248730e-01 -3.45278203e-01 -7.89406061e-01
4.96392697e-01 -4.37404476e-02 7.76917040e-01 4.69450682e-01
9.32232365e-02 3.49427909e-02 -3.00328791e-01 -1.91861764e-01
1.11051691e+00 -3.26998204e-01 1.84950158e-01 -3.40107456e-02
1.00904751e+00 -9.68114078e-01 -1.35275364e-01 3.91503423e-01
-4.23316509e-01 4.15671825e-01 -4.62110221e-01 2.01428652e-01
-9.18278694e-01 -1.04119158e+00 -1.33544192e-01 1.31925416e+00
-3.03437084e-01 -2.12581873e-01 -1.24832046e+00 -8.43665376e-02
-5.09057939e-01 5.84083676e-01 -2.71904022e-01 -1.75828096e-02
-4.78619397e-01 -2.34766558e-01 4.86928821e-01 3.88444483e-01
2.42556602e-01 -1.02024651e+00 -6.69148862e-01 3.11811030e-01
1.40514061e-01 -1.01953745e+00 -6.44171715e-01 6.97497725e-02
-6.43550992e-01 -6.13861322e-01 -6.74794316e-01 -9.90512133e-01
5.89502454e-01 3.44984025e-01 5.81798315e-01 -9.03813839e-02
-3.60794634e-01 7.62600660e-01 -1.61998600e-01 -5.51251173e-01
-5.96812546e-01 -5.72914600e-01 2.00442910e-01 1.92955002e-01
2.30281785e-01 -6.27079070e-01 -8.47379804e-01 3.11025798e-01
-5.08612633e-01 -4.84388500e-01 1.31448850e-01 5.91980934e-01
-7.79413292e-03 3.19359958e-01 3.97588491e-01 -1.00325458e-01
9.35084283e-01 -4.34267223e-02 -1.04520822e+00 -9.34706777e-02
-3.05176258e-01 -7.37832129e-01 2.13981867e-01 -4.79654044e-01
-1.62006974e+00 1.34301245e-01 -9.37449396e-01 -1.24387279e-01
-4.55427945e-01 -7.34792650e-02 -5.50272763e-01 -2.59659857e-01
6.07937992e-01 1.03444576e-01 8.23284388e-02 -5.87633073e-01
1.75071150e-01 1.42912471e+00 8.55216563e-01 3.83315682e-01
5.23794830e-01 1.59261361e-01 -4.40323740e-01 -1.38522339e+00
2.19178811e-01 -8.30700040e-01 -2.77204514e-01 -6.56383276e-01
4.71287131e-01 -1.05671561e+00 -7.93646812e-01 9.74970400e-01
-1.49826515e+00 -2.75922924e-01 2.02573184e-02 7.27026880e-01
-1.68354779e-01 2.35087961e-01 -7.48570859e-01 -1.47058177e+00
-7.14151025e-01 -1.03923261e+00 1.26416755e+00 3.36048990e-01
-2.42702767e-01 -9.31917489e-01 7.22860396e-02 3.82204235e-01
7.43268490e-01 -5.30152857e-01 2.48505175e-02 -2.09822074e-01
-1.02148235e-01 -5.02953708e-01 2.01101065e-01 6.27939165e-01
1.64263979e-01 9.65741575e-02 -1.87330186e+00 -5.10983825e-01
7.24403620e-01 3.03793132e-01 2.44462818e-01 8.49399090e-01
5.80974698e-01 -1.76608548e-01 -2.04064980e-01 5.28071344e-01
9.71822798e-01 5.71929157e-01 7.37142086e-01 -7.96940848e-02
-6.38078749e-02 6.81018949e-01 3.46196055e-01 4.79722500e-01
-2.71917909e-01 7.06347585e-01 4.39403653e-01 -4.14522558e-01
-5.00934243e-01 1.81383014e-01 6.38467848e-01 5.85625172e-01
2.68089294e-01 -3.93071324e-01 -5.40441155e-01 3.65673631e-01
-8.70224476e-01 -9.70006406e-01 -1.29014716e-01 2.38577628e+00
6.18481398e-01 -1.68055087e-01 2.34670311e-01 7.74103940e-01
1.13258135e+00 -1.47394419e-01 1.02984443e-01 -5.70797443e-01
2.03429028e-01 7.29911327e-01 1.49693787e-01 1.26064730e+00
-1.16122115e+00 5.54001451e-01 7.31193113e+00 3.98908496e-01
-1.50523555e+00 4.25056696e-01 1.79759458e-01 -4.00267728e-02
2.61566281e-01 -7.12288618e-01 -8.09553564e-01 2.40144104e-01
1.31930006e+00 9.34870169e-02 5.59317529e-01 7.71771848e-01
8.20630491e-01 -9.50624794e-02 -7.33785689e-01 1.02810872e+00
2.48543680e-01 -7.73922861e-01 -8.39041412e-01 -1.97459105e-02
1.93376258e-01 -2.10605398e-01 3.70537192e-01 -9.99715254e-02
-4.31829572e-01 -7.10975289e-01 7.47842968e-01 -8.94488208e-03
5.84253490e-01 -6.54872537e-01 6.09223306e-01 1.49893090e-01
-1.16830277e+00 -1.10373117e-01 3.89002310e-03 -3.00978333e-01
3.76661301e-01 3.56790245e-01 -1.17316604e+00 -1.16672128e-01
7.43238509e-01 -8.41723084e-02 -2.19706789e-01 1.46538734e+00
-3.83626312e-01 7.39350557e-01 -4.57520157e-01 8.97936970e-02
-3.89719367e-01 2.77327210e-01 8.94313812e-01 1.72495842e+00
5.13687670e-01 3.07949781e-01 -5.01445651e-01 1.77460641e-01
3.28129917e-01 9.70923975e-02 -4.23397362e-01 3.25167179e-01
4.59244460e-01 1.56181931e+00 -4.53738362e-01 -2.15857133e-01
-3.47548693e-01 8.92375767e-01 -6.02575302e-01 4.46587652e-01
-7.69491732e-01 -7.52399683e-01 6.58084214e-01 3.28803480e-01
3.42846245e-01 -2.57282648e-02 -2.03130603e-01 -5.27786970e-01
1.80697665e-01 -1.06791162e+00 1.44513339e-01 -1.23205280e+00
-6.95503592e-01 1.17746663e+00 -4.21461076e-01 -7.27850258e-01
-1.99095964e-01 -9.68592644e-01 -7.77756989e-01 1.35498559e+00
-1.13725674e+00 -8.23698819e-01 -2.98442155e-01 8.98122728e-01
4.79133576e-01 -1.65162683e-01 9.66528416e-01 7.71280885e-01
-5.03827870e-01 7.10863888e-01 -1.27933726e-01 -2.50354558e-01
5.81754446e-01 -1.04637337e+00 3.99406105e-01 1.17055833e+00
-2.09161937e-01 6.85746968e-01 1.15882182e+00 -4.04844791e-01
-1.28380489e+00 -8.29273164e-01 1.16960597e+00 1.42241102e-02
4.21107799e-01 -6.92663252e-01 -8.79612207e-01 4.07841116e-01
8.23680520e-01 -1.70282394e-01 7.68028677e-01 -7.46161267e-02
4.47173975e-02 -2.49093831e-01 -1.39362657e+00 3.56664181e-01
5.13148665e-01 -8.13646376e-01 -7.10009634e-01 2.79042214e-01
4.53973830e-01 -5.37285864e-01 -4.92810220e-01 -1.00242149e-03
6.13366008e-01 -9.95958745e-01 8.50579321e-01 9.63417292e-02
-3.79282564e-01 -3.59867215e-01 4.80121970e-02 -1.32707143e+00
-2.96693780e-02 -1.29472494e+00 -1.33404344e-01 1.44146240e+00
2.75074750e-01 -8.15329909e-01 5.02644718e-01 5.03769636e-01
-3.16317976e-01 -1.25593483e-01 -1.14962029e+00 -4.03748691e-01
-7.72247136e-01 -8.55093122e-01 5.43058753e-01 1.97773710e-01
3.01195532e-01 1.68146506e-01 -2.18761548e-01 9.38707769e-01
4.59170133e-01 -7.92753637e-01 4.65800345e-01 -6.04701638e-01
-3.31434667e-01 -2.26223275e-01 -4.75122720e-01 -7.94850767e-01
3.00448039e-03 -1.43338487e-01 2.94190764e-01 -1.05850446e+00
-4.57913548e-01 3.65469486e-01 -1.27604753e-01 -2.07215045e-02
-1.97487280e-01 1.50065169e-01 -2.45406091e-01 -7.07845509e-01
-3.26630473e-03 2.35806808e-01 4.20682669e-01 1.11654408e-01
-3.69159549e-01 4.37921107e-01 -2.91439891e-01 8.51555109e-01
3.99634331e-01 -3.21225047e-01 -4.81308490e-01 -2.72346914e-01
-1.54310182e-01 1.98570564e-01 6.72499001e-01 -1.05238914e+00
8.03753436e-01 6.18808866e-01 3.17925483e-01 -5.15374064e-01
5.79770088e-01 -1.30086350e+00 3.07353944e-01 5.27368128e-01
-2.25661695e-02 3.16535771e-01 6.37625217e-01 2.46230498e-01
-3.75833690e-01 -1.34447470e-01 9.17243242e-01 5.18782318e-01
-3.88057649e-01 -1.69367328e-01 -8.65026414e-01 -7.43891954e-01
9.05912161e-01 -4.04598713e-01 -1.29314885e-01 -4.76598978e-01
-1.11755490e+00 -4.59190190e-01 -1.39153406e-01 3.17559063e-01
9.78448331e-01 -8.38596702e-01 -6.62618518e-01 1.05038869e+00
-5.97753823e-01 -7.69382834e-01 4.97973114e-01 8.41629863e-01
-4.34189320e-01 1.87767059e-01 2.98344810e-02 -6.05918109e-01
-2.23607564e+00 4.59443867e-01 9.57411051e-01 3.58254611e-01
-4.09246743e-01 1.21781921e+00 -1.20887004e-01 1.11194715e-01
6.54024959e-01 -2.42623836e-02 -3.17567348e-01 -1.58276200e-01
1.32722306e+00 6.54009700e-01 5.32353282e-01 -5.96336603e-01
-3.47738475e-01 2.11726069e-01 3.23146015e-01 -7.49606609e-01
1.13644862e+00 -4.58690882e-01 -1.46092430e-01 -7.12443963e-02
9.69945967e-01 4.64496195e-01 -1.01077783e+00 1.46472260e-01
-3.43311429e-01 -7.31810272e-01 6.40697002e-01 -9.25871551e-01
-9.16845977e-01 9.60610390e-01 1.49399424e+00 1.27437070e-01
1.76698494e+00 -4.14842457e-01 3.49470675e-01 -3.91895287e-02
6.04236796e-02 -9.45686698e-01 -7.62285590e-02 2.95639247e-01
1.05719161e+00 -7.79682040e-01 -3.37359756e-01 -4.90440726e-01
2.50323070e-03 1.16100156e+00 2.76360184e-01 1.52702123e-01
1.06054890e+00 1.15329659e+00 4.31620210e-01 5.02575710e-02
-5.61994433e-01 -2.12208461e-02 2.41781622e-01 1.04055774e+00
6.91542745e-01 -2.28175089e-01 2.52698641e-02 4.77310240e-01
-4.04288620e-01 -3.98039185e-02 1.97282255e-01 6.39770985e-01
-6.06466353e-01 -7.40256488e-01 -1.04408538e+00 -3.74200702e-01
-7.51218438e-01 -3.02619457e-01 -1.47227958e-01 3.30567569e-01
2.67763317e-01 1.78115880e+00 1.14771016e-01 -3.11002791e-01
7.77520657e-01 3.92345227e-02 4.04170036e-01 -3.44206214e-01
-1.37701094e+00 8.20436656e-01 8.83555114e-02 -3.28547329e-01
-1.07794620e-01 -8.14887762e-01 -7.70995378e-01 -1.67635694e-01
-7.47147202e-01 -2.31754519e-02 1.17875075e+00 6.40396476e-01
3.34031940e-01 7.17949212e-01 8.83283973e-01 -8.57074499e-01
-4.23421502e-01 -1.20277047e+00 -7.84519672e-01 -1.87185869e-01
7.20949531e-01 -1.55076385e-01 -6.18613839e-01 1.59996167e-01] | [14.954854965209961, 5.7980217933654785] |
b11f5dfb-d1bf-43c3-b8dc-f32782536eea | cholectriplet2021-a-benchmark-challenge-for | 2204.04746 | null | https://arxiv.org/abs/2204.04746v2 | https://arxiv.org/pdf/2204.04746v2.pdf | CholecTriplet2021: A benchmark challenge for surgical action triplet recognition | Context-aware decision support in the operating room can foster surgical safety and efficiency by leveraging real-time feedback from surgical workflow analysis. Most existing works recognize surgical activities at a coarse-grained level, such as phases, steps or events, leaving out fine-grained interaction details about the surgical activity; yet those are needed for more helpful AI assistance in the operating room. Recognizing surgical actions as triplets of <instrument, verb, target> combination delivers comprehensive details about the activities taking place in surgical videos. This paper presents CholecTriplet2021: an endoscopic vision challenge organized at MICCAI 2021 for the recognition of surgical action triplets in laparoscopic videos. The challenge granted private access to the large-scale CholecT50 dataset, which is annotated with action triplet information. In this paper, we present the challenge setup and assessment of the state-of-the-art deep learning methods proposed by the participants during the challenge. A total of 4 baseline methods from the challenge organizers and 19 new deep learning algorithms by competing teams are presented to recognize surgical action triplets directly from surgical videos, achieving mean average precision (mAP) ranging from 4.2% to 38.1%. This study also analyzes the significance of the results obtained by the presented approaches, performs a thorough methodological comparison between them, in-depth result analysis, and proposes a novel ensemble method for enhanced recognition. Our analysis shows that surgical workflow analysis is not yet solved, and also highlights interesting directions for future research on fine-grained surgical activity recognition which is of utmost importance for the development of AI in surgery. | ['Nicolas Padoy', 'Cristians Gonzalez', 'Barbara Seeliger', 'Pietro Mascagni', 'Didier Mutter', 'Danail Stoyanov', 'Alexander Jenke', 'Lalithkumar Seenivasan', 'Mobarakol Islam', 'Mengya Xu', 'Nicolas Elini van der Kar', 'Jakob-Anton Aschenbrenner', 'Shuai Ding', 'Yuanbo Zhu', 'Imanol Luengo', 'Debdoot Sheet', 'Velmurugan Balasubramanian', 'Zhen Li', 'GuiBin Bian', 'Chen Qian', 'Inge Nicole Wijma', 'Ruby Mae Egging', 'Jaime Fonseca', 'Pedro Morais', 'João L. Vilaça', 'Tobias Czempiel', 'Finn Gaida', 'Li Ling', 'Helena R. Torres', 'Bruno Oliveira', 'Nithya Bhasker', 'Sebastian Bodenstedt', 'Mohammad Hasan Sarhan', 'Julian Ronald Abbing', 'Hongliang Ren', 'Winnie Pang', 'Satoshi Kondo', 'Rong Tao', 'Rachana Sathish', 'Sista Raviteja', 'Beerend Gerats', 'Bokai Zhang', 'Liansheng Wang', 'Jiacheng Wang', 'Huabin Chen', 'Li Zhang', 'Maria Robu', 'Ricardo Sanchez-Matilla', 'Neil Getty', 'Xiaotian Duan', 'Guoyan Zheng', 'Derong Yu', 'Hao Wang', 'Yuxuan Yang', 'Fucang Jia', 'Tong Xia', 'Zixuan Zhao', 'Fangfang Xia', 'Armine Vardazaryan', 'Tong Yu', 'Deepak Alapatt', 'Chinedu Innocent Nwoye'] | 2022-04-10 | null | null | null | null | ['action-triplet-recognition'] | ['computer-vision'] | [ 4.04021263e-01 2.97491640e-01 -5.26082993e-01 -2.81183124e-01
-8.69706571e-01 -6.84033513e-01 4.49767828e-01 3.78017247e-01
-5.80543935e-01 2.94167161e-01 9.13187742e-01 -4.68878597e-01
-6.93794906e-01 -1.55769676e-01 -5.29976666e-01 -9.17676747e-01
-4.26983535e-01 2.69685805e-01 -2.96938956e-01 -1.56377759e-02
3.17078471e-01 6.53846502e-01 -1.48700154e+00 1.09859633e+00
3.02664220e-01 1.02033830e+00 9.46112201e-02 9.75616157e-01
2.54404575e-01 1.31502581e+00 -5.93083501e-01 1.77110761e-01
4.73401248e-01 -3.39824349e-01 -9.13590729e-01 -6.48553967e-02
5.64894080e-01 -3.16224247e-01 -4.34588253e-01 5.25840998e-01
5.51962316e-01 -1.52079716e-01 3.70845109e-01 -7.28614450e-01
3.21229756e-01 6.89376950e-01 2.10327655e-02 4.70388383e-01
3.96881938e-01 4.31985527e-01 5.27356267e-01 -4.04175192e-01
9.80193794e-01 6.23966694e-01 5.84072053e-01 5.18581092e-01
-6.93604708e-01 -5.10691464e-01 1.49100214e-01 2.16437101e-01
-8.38164926e-01 -3.26373726e-01 4.20575619e-01 -7.12760210e-01
1.07875955e+00 6.59272254e-01 1.10339761e+00 1.46757197e+00
7.18546987e-01 9.31157768e-01 1.03766394e+00 -4.25499141e-01
-3.06567028e-02 -2.33889088e-01 -4.65319268e-02 1.13512909e+00
2.76825756e-01 5.42605519e-01 -7.01281190e-01 4.52652872e-02
7.43479967e-01 3.37838531e-01 -4.05707270e-01 -6.84557557e-01
-2.12536144e+00 5.12402236e-01 5.07960618e-01 7.50949383e-01
-6.71733201e-01 1.71183065e-01 8.37380111e-01 1.59831747e-01
-1.63329095e-01 1.00659287e+00 -4.51081216e-01 -7.45552063e-01
-7.48856008e-01 -1.32572129e-01 9.91504669e-01 8.64909291e-01
7.93892965e-02 -3.95707250e-01 -7.83416867e-01 3.14564347e-01
-5.96237816e-02 -3.08343261e-01 7.04210818e-01 -1.14127707e+00
1.67277545e-01 7.95015574e-01 1.17732570e-01 -6.87976360e-01
-9.85070884e-01 -4.70254660e-01 -7.68936396e-01 2.13188872e-01
3.46062869e-01 -1.00988254e-01 -1.11972380e+00 1.17514813e+00
9.47516784e-02 3.67339104e-02 -9.40905139e-02 7.57240653e-01
1.13846755e+00 -5.37062176e-02 1.73518911e-01 -3.34766984e-01
1.47451675e+00 -1.30359709e+00 -9.68357325e-01 -1.86875284e-01
1.26547122e+00 -6.90109789e-01 6.50104344e-01 7.38127470e-01
-9.47698712e-01 -3.87546003e-01 -8.51182759e-01 2.56096236e-02
-2.34277323e-01 7.00318992e-01 1.04626143e+00 3.30005735e-01
-9.41951632e-01 5.76790929e-01 -1.44177008e+00 -2.81276554e-01
4.02658820e-01 7.58654118e-01 -6.71994507e-01 -4.27838564e-02
-6.08704329e-01 1.12656474e+00 2.63046861e-01 4.35021311e-01
-1.33694482e+00 -9.33003128e-01 -1.08627653e+00 -1.37986571e-01
7.35289812e-01 -7.02671349e-01 1.55022156e+00 -6.55655980e-01
-1.32060289e+00 1.31290531e+00 8.74572396e-02 -5.18681407e-01
7.26313233e-01 -3.21006507e-01 -2.33181298e-01 5.79192601e-02
-2.99029440e-01 3.11639696e-01 3.44821393e-01 -7.34419882e-01
-8.46431911e-01 -2.89741457e-01 4.43463564e-01 2.93819875e-01
-9.99503061e-02 -9.42562222e-02 -4.16513026e-01 -6.16565764e-01
-2.21608981e-01 -1.27698135e+00 -5.27981877e-01 6.26003072e-02
-3.54616374e-01 -3.46225314e-02 1.48800150e-01 -4.37857389e-01
1.63752306e+00 -2.17806792e+00 3.59259218e-01 -4.47608083e-02
4.68059182e-01 1.51877880e-01 1.51509851e-01 6.87394440e-01
-4.64453876e-01 -2.65093386e-01 1.93961233e-01 -2.40691885e-01
-1.40546814e-01 3.19582134e-01 2.04905629e-01 6.37186825e-01
-4.68987226e-01 7.90432394e-01 -1.04825830e+00 -4.36980158e-01
7.11489260e-01 4.90369871e-02 -5.90728879e-01 2.96589792e-01
3.38920087e-01 6.67814136e-01 -3.01796883e-01 9.76080954e-01
-4.24573049e-02 -1.36619315e-01 5.26553214e-01 -7.78587401e-01
-2.73823947e-01 2.31069803e-01 -8.40278506e-01 2.41465878e+00
-5.97427070e-01 4.98966485e-01 1.03411384e-01 -9.98055696e-01
2.81473756e-01 4.04400915e-01 1.13295186e+00 -4.59271669e-01
3.88475299e-01 1.26519427e-01 3.98061246e-01 -7.87002742e-01
1.94689244e-01 1.77736014e-01 -4.35386688e-01 -4.23279330e-02
3.84694129e-01 7.84663260e-02 4.26049054e-01 1.42239090e-02
1.62135613e+00 1.02399834e-01 1.19446826e+00 -4.23442513e-01
4.05509412e-01 4.60431993e-01 3.74729484e-01 1.19205928e+00
-6.12396121e-01 3.41801137e-01 5.13838530e-01 -1.12713981e+00
-3.16886783e-01 -8.17949235e-01 -2.44754180e-02 9.72721219e-01
1.42596364e-01 -7.18950689e-01 -5.18114686e-01 -1.17180586e+00
-1.84848234e-01 1.99144751e-01 -1.47201395e+00 -3.31156760e-01
-8.57812285e-01 -4.95357007e-01 1.64176434e-01 7.28237391e-01
-3.26673239e-02 -1.24904883e+00 -1.00706923e+00 3.09409142e-01
-1.94287613e-01 -1.11290240e+00 -4.68454599e-01 4.77044761e-01
-8.35078895e-01 -1.62609208e+00 -3.75847638e-01 -7.23187268e-01
6.97965860e-01 1.05513446e-01 1.09577572e+00 -5.99445999e-02
-9.93883550e-01 6.19873166e-01 -4.16460752e-01 -6.60703957e-01
-5.10558546e-01 -4.84398343e-02 -3.31345648e-02 -2.17721045e-01
2.74489850e-01 7.62618408e-02 -9.83854890e-01 3.47334594e-01
-5.58152676e-01 3.62180799e-01 1.19407094e+00 9.80645001e-01
5.34682214e-01 -5.83529890e-01 -4.46597517e-01 -1.03916299e+00
2.54658371e-01 -2.36741215e-01 -3.17591220e-01 2.77026802e-01
-4.34289545e-01 4.55608033e-02 1.65240362e-01 -3.95582229e-01
-8.30580890e-01 3.33766580e-01 1.93219602e-01 -5.08482277e-01
-3.96969646e-01 5.66376328e-01 5.25586545e-01 -2.89568752e-01
9.10701156e-01 2.83873398e-02 4.51021232e-02 -1.35732919e-01
-2.26665035e-01 2.73225665e-01 6.79767132e-01 -3.45157951e-01
1.09622195e-01 5.56173623e-01 1.31472796e-01 -3.09459686e-01
-1.06827033e+00 -9.30405438e-01 -4.94303912e-01 -3.69008958e-01
8.21591616e-01 -8.95236075e-01 -1.13546181e+00 3.19843501e-01
-7.33368099e-01 -4.55880404e-01 -3.57943535e-01 8.79046142e-01
-8.54952514e-01 3.02107390e-02 -8.49839032e-01 -3.27709943e-01
-3.74125212e-01 -1.54738081e+00 1.28174770e+00 1.16086239e-03
-5.77939808e-01 -9.17528570e-01 2.53721744e-01 6.57963753e-01
3.39439899e-01 7.85824299e-01 4.36406344e-01 -7.78585851e-01
-6.72542095e-01 -5.65852404e-01 2.90184736e-01 -5.19086840e-03
4.16715413e-01 -1.28751144e-01 -5.59309244e-01 -3.99819672e-01
-1.12727076e-01 1.49408787e-01 5.74983299e-01 5.68618715e-01
1.75155306e+00 -2.65132248e-01 -7.73336112e-01 8.58652055e-01
1.06769443e+00 3.82831812e-01 5.09890497e-01 6.01432800e-01
5.51959515e-01 5.29947340e-01 1.06556213e+00 4.90536213e-01
2.84183268e-02 7.03633785e-01 8.52136791e-01 -1.30725652e-01
-1.83933049e-01 3.52023244e-01 1.23700865e-01 4.50024694e-01
-5.97613215e-01 1.14111610e-01 -1.19751847e+00 5.93822777e-01
-1.77509987e+00 -9.31067467e-01 2.01991931e-01 2.11179304e+00
6.60162449e-01 -2.07062513e-02 -2.44204924e-01 -7.45350644e-02
5.37664928e-02 1.70664176e-01 -2.69122809e-01 -3.45347971e-01
5.67604125e-01 2.36511841e-01 7.12260187e-01 9.94581357e-02
-1.40815866e+00 4.56053942e-01 6.14379978e+00 5.70794463e-01
-1.20267582e+00 -8.13992396e-02 1.82815179e-01 -6.47727549e-01
5.59948802e-01 -4.28900719e-01 -6.10068142e-01 2.99737036e-01
8.34272265e-01 -8.22001472e-02 1.47195086e-01 9.13482130e-01
3.67302299e-01 -2.41418198e-01 -1.55639887e+00 1.16499698e+00
2.22982228e-01 -1.92923486e+00 -2.81096101e-01 3.11329901e-01
5.61410785e-01 2.45415211e-01 -2.20252916e-01 4.22837108e-01
1.86197206e-01 -1.01427686e+00 2.87462533e-01 7.37633705e-01
7.30254471e-01 -2.31729522e-01 1.03858006e+00 1.63100824e-01
-8.92753363e-01 -5.87260723e-01 5.78009009e-01 9.94825885e-02
-1.90951064e-01 7.06696287e-02 -1.28454161e+00 7.82796621e-01
8.86420965e-01 1.02193797e+00 -2.57476121e-01 1.12400460e+00
-5.53715751e-02 2.24387929e-01 -5.12129031e-02 9.92698893e-02
4.26514626e-01 3.93716037e-01 4.78388399e-01 1.53107202e+00
2.81647474e-01 1.20069496e-01 2.28909299e-01 -1.43880725e-01
5.94891831e-02 -1.50764525e-01 -5.93355358e-01 -2.78488964e-01
-1.53413415e-01 1.46227896e+00 -4.95558709e-01 -2.85718352e-01
-2.64236778e-01 7.08860576e-01 1.36754727e-02 -9.37267914e-02
-6.82298601e-01 -6.96156994e-02 8.36400986e-01 5.50984666e-02
-7.70259947e-02 -7.28863431e-03 -1.81140199e-01 -1.11086702e+00
-5.61444610e-02 -1.18210435e+00 8.57256591e-01 -5.52506506e-01
-5.00180125e-01 6.13497138e-01 -3.08105592e-02 -1.91345704e+00
-5.32689035e-01 -1.06959307e+00 -2.73317486e-01 3.43018711e-01
-1.12006295e+00 -1.14840329e+00 -1.12307870e+00 5.77668965e-01
7.30878651e-01 5.28561287e-02 1.24943411e+00 6.81371847e-03
-3.44080657e-01 4.51717347e-01 -2.18773991e-01 3.00675482e-01
9.57942724e-01 -1.29972172e+00 -3.23208958e-01 4.24089998e-01
9.31208581e-02 7.72650599e-01 6.36654258e-01 -2.83711046e-01
-1.70202434e+00 -6.75735891e-01 6.33676291e-01 -9.39101934e-01
5.35387814e-01 -1.29746661e-01 -1.76323488e-01 9.05322373e-01
2.61366367e-01 2.14077592e-01 1.10682690e+00 2.48931319e-01
1.34500176e-01 -1.95275143e-01 -8.98727000e-01 5.87777257e-01
1.26894081e+00 -1.83468878e-01 -8.32268000e-01 7.61735737e-01
3.00238729e-01 -1.37906373e+00 -1.20131719e+00 9.67830539e-01
7.64362514e-01 -1.00346231e+00 9.62329507e-01 -8.14893544e-01
5.82652926e-01 2.67957570e-03 2.77815461e-01 -1.24744260e+00
-2.42795572e-01 -6.83638871e-01 -2.93830663e-01 -7.85858706e-02
1.22230634e-01 -4.93600816e-02 8.23971033e-01 3.33670765e-01
-8.83230090e-01 -1.08886027e+00 -8.11575294e-01 -4.09387201e-01
-5.01068413e-01 -3.14874560e-01 -1.06868744e-01 9.35504079e-01
2.74641335e-01 -5.15732765e-01 -2.37413660e-01 -7.33145624e-02
3.08831096e-01 2.19008982e-01 7.67663956e-01 -9.81490493e-01
-2.76638597e-01 -6.90528393e-01 -6.07698977e-01 -3.19300175e-01
-4.19037431e-01 -7.62778580e-01 3.12641263e-02 -1.82393026e+00
1.99746042e-01 2.09656376e-02 -7.79905438e-01 7.61206746e-01
2.00769261e-01 -3.27863395e-02 1.43619120e-01 1.56554952e-01
-7.28977799e-01 -2.34077647e-01 1.47839141e+00 -1.89020321e-01
-1.63946062e-01 7.29008168e-02 -6.27529860e-01 7.78791130e-01
3.74766529e-01 -4.30793494e-01 -6.86154291e-02 -3.52323979e-01
-3.79941426e-02 2.86918044e-01 3.12382281e-01 -1.15784180e+00
6.43345475e-01 -1.59457192e-01 2.10484236e-01 -4.61200804e-01
3.46005112e-01 -1.02903521e+00 2.57858217e-01 9.73523140e-01
-4.41268086e-01 -2.13559434e-01 4.27548170e-01 4.98586327e-01
-4.69701201e-01 1.41096175e-01 5.10638416e-01 -5.38818896e-01
-9.27938521e-01 3.86193931e-01 -5.32021105e-01 -1.84660763e-01
1.41798306e+00 -3.17706019e-01 -1.10747673e-01 1.40807420e-01
-1.31079924e+00 8.42361003e-02 4.22823317e-02 7.36048937e-01
2.29132086e-01 -8.66245508e-01 -5.77421010e-01 1.30615756e-01
7.00231671e-01 -3.56433615e-02 9.91521180e-01 1.71071696e+00
-8.26983929e-01 8.13279212e-01 -3.24618816e-01 -7.48688817e-01
-1.53961790e+00 5.74583173e-01 6.23427987e-01 -8.47197175e-01
-7.99246728e-01 7.98719227e-01 3.95203352e-01 -2.36529216e-01
6.07056677e-01 -7.64582992e-01 -3.51625651e-01 2.16839314e-01
6.49934471e-01 -6.43384457e-02 5.29722631e-01 9.59849742e-04
-6.75581515e-01 3.76502693e-01 -4.55034554e-01 5.76856196e-01
1.21841037e+00 3.60062391e-01 1.90422639e-01 2.02793345e-01
1.04465377e+00 -1.57096937e-01 -9.58814025e-01 -2.14133020e-02
9.41409264e-03 -3.24367940e-01 1.95131406e-01 -1.57918143e+00
-8.97230804e-01 6.26768410e-01 6.81451201e-01 -2.57863432e-01
1.30081153e+00 -7.29157077e-03 2.54585057e-01 4.38202560e-01
5.58467329e-01 -8.84604514e-01 -2.52484460e-03 4.00356501e-01
1.13622451e+00 -1.43745911e+00 -1.16087519e-01 -3.71028990e-01
-7.36094475e-01 1.28373635e+00 6.34502113e-01 -5.65700606e-03
5.05178511e-01 6.23314559e-01 2.43724883e-01 -4.77597266e-01
-7.41252065e-01 4.83922521e-03 4.48564082e-01 1.38707072e-01
4.83037949e-01 2.16366157e-01 -4.56180453e-01 4.91387606e-01
6.44651055e-02 5.84386587e-01 5.86571515e-01 1.42971134e+00
5.53998873e-02 -7.57184744e-01 9.79534537e-02 6.93598807e-01
-6.84722602e-01 -4.26869392e-02 2.57969312e-02 1.02980077e+00
8.08422416e-02 4.96429682e-01 -1.72468349e-01 -2.02820674e-01
7.81772912e-01 -3.55784804e-01 6.69789553e-01 -7.46247172e-01
-1.27691305e+00 -2.28441730e-02 4.89336312e-01 -1.59879470e+00
-5.10827422e-01 -7.05993652e-01 -9.99237716e-01 2.27432057e-01
3.80797774e-01 4.07307893e-02 8.75396073e-01 7.40544438e-01
4.73488033e-01 1.29361522e+00 2.58576155e-01 -9.80046034e-01
-3.61456126e-01 -8.78279567e-01 -1.54137716e-01 2.56973416e-01
5.90600491e-01 -8.27679336e-01 -6.17475152e-01 -6.36373758e-02] | [14.057258605957031, -3.4049859046936035] |
b2ebb58c-0e4c-480a-bd55-783e3b0f8cd8 | brazilian-lyrics-based-music-genre | 2003.05377 | null | https://arxiv.org/abs/2003.05377v1 | https://arxiv.org/pdf/2003.05377v1.pdf | Brazilian Lyrics-Based Music Genre Classification Using a BLSTM Network | Organize songs, albums, and artists in groups with shared similarity could be done with the help of genre labels. In this paper, we present a novel approach for automatic classifying musical genre in Brazilian music using only the song lyrics. This kind of classification remains a challenge in the field of Natural Language Processing. We construct a dataset of 138,368 Brazilian song lyrics distributed in 14 genres. We apply SVM, Random Forest and a Bidirectional Long Short-Term Memory (BLSTM) network combined with different word embeddings techniques to address this classification task. Our experiments show that the BLSTM method outperforms the other models with an F1-score average of $0.48$. Some genres like "gospel", "funk-carioca" and "sertanejo", which obtained 0.89, 0.70 and 0.69 of F1-score, respectively, can be defined as the most distinct and easy to classify in the Brazilian musical genres context. | ['Hélio Cortês Vieira Lopes', 'Rômulo César Costa de Sousa', 'Raul de Araújo Lima', 'Simone Diniz Junqueira Barbosa'] | 2020-03-06 | null | null | null | null | ['genre-classification'] | ['computer-vision'] | [-1.96970895e-01 -4.12514865e-01 4.09460478e-02 -5.11967354e-02
-5.69622159e-01 -8.42913687e-01 4.25530553e-01 2.74327904e-01
-6.36066318e-01 7.55869150e-01 3.86231095e-01 1.17272735e-01
-3.34147871e-01 -7.91699767e-01 -3.14398140e-01 -7.45966673e-01
-5.51460199e-02 3.51144254e-01 1.03535596e-02 -3.39323968e-01
6.82290196e-01 8.94431546e-02 -1.69862890e+00 4.75302458e-01
3.23054671e-01 1.07157636e+00 1.89904511e-01 8.04279149e-01
-2.58697093e-01 7.21079826e-01 -1.01704693e+00 -6.01414263e-01
-4.42126021e-02 -3.24470103e-01 -9.05366898e-01 -5.35541177e-01
3.87454420e-01 1.41355231e-01 -9.64588374e-02 9.31058884e-01
6.33550465e-01 4.20503527e-01 8.02837551e-01 -8.82335305e-01
-5.46801209e-01 1.05424297e+00 -3.91954124e-01 2.52272218e-01
3.85627866e-01 -2.63775259e-01 1.50698149e+00 -7.01961935e-01
5.55605948e-01 9.31297183e-01 7.88720310e-01 4.94275808e-01
-1.08132398e+00 -1.08060586e+00 -1.55459031e-01 5.29293776e-01
-1.53470838e+00 -1.25325903e-01 9.12593305e-01 -7.62712121e-01
9.54005361e-01 4.20832753e-01 7.89180815e-01 1.13186049e+00
2.49220431e-01 2.53487229e-01 1.20640111e+00 -5.55466652e-01
-8.96295458e-02 9.45788324e-02 3.45020831e-01 4.79551166e-01
-1.09976061e-01 -3.08162659e-01 -9.76345062e-01 -2.14584440e-01
5.37571251e-01 -1.49913877e-01 2.55322810e-02 2.10405067e-01
-1.34274960e+00 8.58044565e-01 2.68356800e-01 9.77765381e-01
-5.11581413e-02 7.22682774e-02 7.11418390e-01 5.10016441e-01
5.03841817e-01 8.90396535e-01 -4.13326323e-01 -4.28333193e-01
-1.06769145e+00 2.22259462e-01 8.53827417e-01 4.32482362e-01
2.69325912e-01 1.90444186e-01 -2.90345158e-02 1.28633249e+00
2.15249300e-01 2.72034466e-01 9.98012364e-01 -7.69043028e-01
1.68774635e-01 3.38230819e-01 -1.53786018e-01 -1.17037833e+00
-2.37725154e-01 -6.55473471e-01 -7.32484519e-01 5.71372956e-02
5.19186914e-01 7.43714124e-02 -3.74755204e-01 1.67268586e+00
-1.44777253e-01 2.13524967e-01 -8.84856805e-02 7.01790869e-01
1.08486581e+00 7.12281168e-01 -8.33356306e-02 -7.62877986e-02
1.51693666e+00 -1.11200345e+00 -7.27961063e-01 2.40865350e-01
1.34043217e-01 -1.27988338e+00 1.42778552e+00 7.78515577e-01
-8.70378137e-01 -9.63311672e-01 -1.23892319e+00 1.71790961e-02
-5.21883130e-01 1.67654172e-01 3.25920850e-01 6.31533265e-01
-7.01258659e-01 9.38490987e-01 -4.61154759e-01 -2.50094324e-01
1.19196296e-01 3.20154786e-01 -1.89173266e-01 6.75477266e-01
-9.68778133e-01 5.29579878e-01 2.24392176e-01 -3.07786703e-01
-8.86147857e-01 -5.48563838e-01 -3.44423145e-01 3.07030156e-02
3.62707302e-02 -4.37527716e-01 1.04859495e+00 -9.70266581e-01
-1.49586666e+00 1.16849959e+00 1.01006076e-01 -4.57694560e-01
6.82465434e-02 -3.01744014e-01 -6.53144121e-01 -1.73518434e-01
2.54079908e-01 4.34635341e-01 6.09573722e-01 -8.15999448e-01
-8.56388986e-01 -4.40936089e-01 7.56414458e-02 5.65878442e-03
-8.33397329e-01 3.43863666e-01 1.39402673e-01 -1.10498905e+00
-2.04025403e-01 -8.84545088e-01 1.35387480e-01 -4.96983260e-01
-3.51473302e-01 -7.18178689e-01 1.68493107e-01 -7.34987557e-01
1.68500555e+00 -2.26661849e+00 2.50551283e-01 -2.62793130e-03
6.00338914e-02 -1.50973886e-01 7.69106150e-02 5.15485108e-01
8.67532641e-02 1.97501838e-01 4.20648046e-02 -4.51502278e-02
1.84561253e-01 4.99944836e-02 -4.23987538e-01 6.57718554e-02
-2.28815719e-01 4.61539865e-01 -7.99474895e-01 -5.02453029e-01
-2.08849072e-01 2.73710877e-01 -3.65109533e-01 2.76186824e-01
3.69016267e-02 3.43053371e-01 -6.89577088e-02 5.72444439e-01
9.85392332e-02 2.13834196e-01 1.37935251e-01 -1.04491457e-01
-2.60065466e-01 5.31254888e-01 -1.18253863e+00 2.17832255e+00
-4.87235844e-01 7.73158550e-01 -2.92196780e-01 -8.55684280e-01
1.31014705e+00 3.90561908e-01 4.23183829e-01 -3.43151361e-01
1.36477783e-01 4.51172084e-01 2.80956000e-01 -5.37748754e-01
7.01255620e-01 -6.58382535e-01 -3.87384176e-01 4.11712855e-01
3.66195261e-01 1.52019095e-02 4.55571294e-01 -3.38412523e-01
1.00749338e+00 6.57021701e-02 2.72276312e-01 -3.78459841e-01
5.79780757e-01 -1.81254610e-01 4.81235385e-01 5.39601922e-01
-1.17231645e-01 6.49875641e-01 3.74284983e-01 -6.12017691e-01
-9.34011996e-01 -1.01632202e+00 -7.57227391e-02 1.65759659e+00
-2.60582775e-01 -5.52270114e-01 -4.87049550e-01 -4.20784324e-01
-1.69314504e-01 4.52675134e-01 -5.06993592e-01 8.44041109e-02
-6.09984457e-01 -5.04975557e-01 1.01830328e+00 2.45871499e-01
2.10089028e-01 -1.51982522e+00 -3.51716459e-01 2.29120299e-01
-2.50758439e-01 -6.42816246e-01 -5.31877697e-01 2.62395412e-01
-6.71515286e-01 -9.22040999e-01 -7.96432436e-01 -1.20981336e+00
-2.54679799e-01 -3.30935687e-01 1.42811322e+00 -3.70587528e-01
-2.66251087e-01 -7.38677308e-02 -4.86325681e-01 -3.60311240e-01
-6.80239126e-02 5.63583910e-01 1.98133767e-01 1.60743162e-01
5.22199571e-01 -9.82056797e-01 -3.18515867e-01 6.46417141e-02
-5.03628373e-01 -2.59461224e-01 1.81516364e-01 8.32809329e-01
5.79883158e-01 -1.62722245e-01 7.91846395e-01 -6.78269684e-01
9.18297946e-01 -4.34320867e-01 2.53314283e-02 -3.17081138e-02
-4.30536509e-01 -2.16878921e-01 8.87067318e-01 -7.29916513e-01
-4.94384319e-01 -2.87443340e-01 -3.89509141e-01 -2.04272464e-01
-2.16474965e-01 3.89735579e-01 1.39760375e-01 1.77449003e-01
6.94282651e-01 1.85580745e-01 -4.36038315e-01 -1.10226059e+00
1.21631464e-02 1.06438005e+00 4.69723135e-01 -5.57474375e-01
4.92620349e-01 2.40820885e-01 -3.86760384e-01 -8.61237168e-01
-9.71716642e-01 -5.54018319e-01 -5.10948062e-01 -3.55818272e-01
1.02733457e+00 -7.04356611e-01 -8.56682599e-01 2.20657349e-01
-1.09045553e+00 1.30749345e-01 -5.61476767e-01 6.26313150e-01
-4.32276517e-01 1.08276591e-01 -7.44850218e-01 -7.50217676e-01
-5.47363520e-01 -7.38670826e-01 8.42830956e-01 1.53966174e-01
-7.19793916e-01 -7.85157621e-01 5.53273380e-01 4.29131836e-01
2.13746205e-01 2.58245945e-01 1.12414408e+00 -8.21384311e-01
1.62172303e-01 4.14216220e-02 1.06252469e-01 5.25800407e-01
1.47939846e-01 -7.23228455e-02 -1.25803399e+00 -1.77804381e-01
7.36933351e-02 -3.90469134e-01 8.87731850e-01 1.73823670e-01
1.17192566e+00 -1.23354770e-01 2.31782958e-01 5.73356152e-01
1.28686917e+00 4.67510879e-01 3.00575763e-01 6.41189218e-01
6.98550165e-01 4.41851079e-01 3.74891043e-01 4.59816903e-01
2.20866114e-01 7.99627304e-01 1.67299397e-02 3.83348376e-01
-3.31884772e-01 -4.00771618e-01 7.01528192e-01 1.58982193e+00
-3.84784967e-01 1.00133076e-01 -8.60751212e-01 7.08996534e-01
-1.59238815e+00 -1.06690216e+00 -2.71475967e-02 2.07452679e+00
1.05957425e+00 1.08362369e-01 3.83223146e-01 8.07445884e-01
6.16350114e-01 4.16322470e-01 2.63695456e-02 -9.02373075e-01
-1.11195296e-01 7.73626626e-01 8.81247371e-02 2.05162689e-01
-1.20588815e+00 9.72530425e-01 6.04750109e+00 1.36132991e+00
-1.11852157e+00 3.42374325e-01 3.06938589e-01 -3.27194154e-01
1.03973038e-02 -1.60196483e-01 -6.84073508e-01 6.01331830e-01
1.06805623e+00 6.52528182e-02 4.80302095e-01 5.94033480e-01
-1.31704817e-02 3.91768157e-01 -9.14632440e-01 1.14570761e+00
4.06315744e-01 -1.17744303e+00 1.39344946e-01 -1.44792944e-01
6.71597958e-01 -9.22720730e-02 -3.49682719e-02 5.05454898e-01
1.87327042e-02 -1.21862650e+00 7.19671130e-01 5.70958853e-01
7.49626994e-01 -9.18962419e-01 7.00567544e-01 1.65762559e-01
-1.20226586e+00 -5.57993054e-02 -4.40218210e-01 -5.11307776e-01
-1.21632859e-01 3.40746552e-01 -4.64295268e-01 5.13432622e-01
1.05896378e+00 9.98543441e-01 -4.83315170e-01 8.04391265e-01
2.88598854e-02 9.10369158e-01 -8.48333016e-02 -5.17879486e-01
1.71903238e-01 -2.32067242e-01 5.59633076e-01 1.26797688e+00
5.30281425e-01 -4.74608332e-01 1.38357639e-01 4.23363805e-01
1.82785967e-03 6.90916955e-01 -4.62610155e-01 -2.16235891e-01
1.97685957e-01 1.19368565e+00 -8.29951763e-01 -3.50786269e-01
-2.31518764e-02 7.39845693e-01 1.46991149e-01 -7.41950348e-02
-8.86495352e-01 -8.99112701e-01 5.30562222e-01 1.50570676e-01
2.08639964e-01 -1.49825186e-01 -3.85830432e-01 -1.12282407e+00
-3.51921432e-02 -8.64651442e-01 5.41619360e-01 -7.24433064e-01
-1.56230056e+00 9.24511313e-01 -4.20847505e-01 -1.39834774e+00
-1.54513240e-01 -6.28439844e-01 -3.73505086e-01 5.19699991e-01
-1.04378366e+00 -1.09930503e+00 -1.22608535e-01 5.51612735e-01
8.33862722e-01 -8.61790180e-01 1.28962624e+00 6.02738678e-01
-1.89128175e-01 4.63598073e-01 2.63722092e-01 2.84647584e-01
9.60795820e-01 -1.48190200e+00 -2.22685695e-01 1.36449561e-01
1.08022487e+00 3.99480194e-01 5.97872376e-01 -1.08116589e-01
-7.17372417e-01 -6.89947665e-01 1.39635754e+00 -2.75293529e-01
1.09821916e+00 -3.56679738e-01 -5.70335567e-01 2.83587754e-01
6.34152770e-01 -6.12262428e-01 1.26908612e+00 5.63573480e-01
-6.14118695e-01 -5.22194147e-01 -6.51593626e-01 4.36033309e-01
1.03976679e+00 -6.95185125e-01 -9.11163747e-01 3.38371396e-01
5.01408815e-01 1.80952653e-01 -1.22514200e+00 7.06242919e-02
1.02806222e+00 -9.03726935e-01 8.49147022e-01 -6.20201170e-01
6.57892048e-01 -2.86329091e-01 -3.36465597e-01 -1.05498517e+00
-2.20359787e-01 -4.69997674e-01 2.90844768e-01 1.36703181e+00
4.04483825e-01 -2.77733088e-01 5.09960711e-01 -6.80656195e-01
2.16745716e-02 -4.08773631e-01 -1.21540964e+00 -8.13788056e-01
2.83833951e-01 -5.91568232e-01 4.27597314e-01 1.11489284e+00
1.48956394e-02 7.00224638e-01 -4.96849477e-01 -5.86407006e-01
3.00439835e-01 4.74070311e-01 6.11250877e-01 -1.44392788e+00
-6.17966712e-01 -7.88893282e-01 -6.49386883e-01 -6.25530005e-01
2.06699073e-01 -1.21308875e+00 -2.67716259e-01 -1.16388333e+00
1.96341664e-01 -2.53480285e-01 -9.90593672e-01 3.85893345e-01
3.97606343e-01 9.28501785e-01 2.94933438e-01 2.03820512e-01
-6.05405271e-01 2.20761657e-01 8.41865420e-01 -3.63749117e-01
-2.21901070e-02 5.72511889e-02 -5.19062161e-01 9.56843555e-01
1.04965973e+00 -7.69334197e-01 5.02292849e-02 -4.69790846e-01
4.15908813e-01 -1.55581191e-01 -1.21046379e-02 -1.18832421e+00
3.35752591e-02 8.57598335e-02 7.17838481e-02 -2.66399205e-01
5.34863830e-01 -5.26911616e-01 1.19325466e-01 5.49555361e-01
-5.69241703e-01 2.41722897e-01 1.89205073e-02 2.41221920e-01
-6.46352410e-01 -5.32072961e-01 4.26351935e-01 -9.51395463e-03
-3.19226950e-01 -2.69603521e-01 -5.18371344e-01 1.10332832e-01
6.98745728e-01 -7.63676688e-03 2.00865790e-01 -1.48764968e-01
-1.11096823e+00 -4.93696570e-01 -8.64400715e-02 8.57141197e-01
3.14641386e-01 -1.58940029e+00 -8.35374296e-01 1.01078257e-01
2.02229038e-01 -5.62901139e-01 7.53281265e-02 5.64852297e-01
-6.84442639e-01 1.71777144e-01 -4.56546605e-01 -3.06395978e-01
-1.75713742e+00 1.52598843e-01 -2.86137331e-02 -5.19208431e-01
-2.30783731e-01 8.74750376e-01 -1.92927927e-01 -3.63432139e-01
4.45348680e-01 -3.88680577e-01 -8.15083027e-01 6.41863346e-01
1.35061488e-01 4.96285439e-01 -9.93587747e-02 -7.70743132e-01
-3.77016962e-01 9.14434612e-01 3.38923275e-01 -2.47979313e-01
1.45308387e+00 3.03075701e-01 -4.98579025e-01 1.41175306e+00
1.02640963e+00 5.25689960e-01 -1.85504705e-01 1.36571243e-01
3.49405527e-01 -3.47447038e-01 -2.80035555e-01 -8.70979786e-01
-9.87571359e-01 1.13960755e+00 7.89072096e-01 7.20716119e-01
9.65068579e-01 -1.81545660e-01 7.62744904e-01 2.51124889e-01
3.22258979e-01 -1.22009480e+00 4.20409031e-02 6.00189567e-01
8.48207176e-01 -7.47408748e-01 -3.26054633e-01 2.58490294e-01
-4.23205137e-01 1.25570989e+00 2.41324201e-01 -7.27659822e-01
8.11996102e-01 1.61818936e-01 4.95742485e-02 1.35579184e-01
-7.29781151e-01 -6.88741431e-02 5.00748634e-01 1.91629767e-01
9.27177966e-01 2.78033167e-01 -1.00865805e+00 1.33071828e+00
-9.38321173e-01 -2.29626417e-01 2.92430282e-01 3.38155270e-01
-4.90420848e-01 -1.19779420e+00 -3.62815678e-01 3.64692211e-01
-1.01983321e+00 -1.26916170e-01 -6.05370939e-01 5.51361799e-01
6.71857774e-01 1.05205667e+00 2.14500457e-01 -8.08405161e-01
2.23499715e-01 2.92047471e-01 4.92865622e-01 -6.93837106e-01
-1.28780496e+00 3.43750358e-01 1.40108824e-01 -2.08263788e-02
-5.65621734e-01 -2.95770019e-01 -8.86654496e-01 -3.42907608e-01
-1.41953632e-01 4.46600795e-01 6.10433817e-01 7.37830281e-01
-7.67601049e-03 8.21683705e-01 5.85342824e-01 -6.16644442e-01
-2.90624678e-01 -1.39611733e+00 -8.97047400e-01 5.75194716e-01
-4.93874215e-02 -4.82126653e-01 -4.20888931e-01 1.84337616e-01] | [15.87685489654541, 5.23203182220459] |
c0ed55cd-6707-4567-8e04-6e9ec776c2d8 | modeling-what-to-ask-and-how-to-ask-for | 2305.03088 | null | https://arxiv.org/abs/2305.03088v1 | https://arxiv.org/pdf/2305.03088v1.pdf | Modeling What-to-ask and How-to-ask for Answer-unaware Conversational Question Generation | Conversational Question Generation (CQG) is a critical task for machines to assist humans in fulfilling their information needs through conversations. The task is generally cast into two different settings: answer-aware and answer-unaware. While the former facilitates the models by exposing the expected answer, the latter is more realistic and receiving growing attentions recently. What-to-ask and how-to-ask are the two main challenges in the answer-unaware setting. To address the first challenge, existing methods mainly select sequential sentences in context as the rationales. We argue that the conversation generated using such naive heuristics may not be natural enough as in reality, the interlocutors often talk about the relevant contents that are not necessarily sequential in context. Additionally, previous methods decide the type of question to be generated (boolean/span-based) implicitly. Modeling the question type explicitly is crucial as the answer, which hints the models to generate a boolean or span-based question, is unavailable. To this end, we present SG-CQG, a two-stage CQG framework. For the what-to-ask stage, a sentence is selected as the rationale from a semantic graph that we construct, and extract the answer span from it. For the how-to-ask stage, a classifier determines the target answer type of the question via two explicit control signals before generating and filtering. In addition, we propose Conv-Distinct, a novel evaluation metric for CQG, to evaluate the diversity of the generated conversation from a context. Compared with the existing answer-unaware CQG models, the proposed SG-CQG achieves state-of-the-art performance. | ['Ai Ti Aw', 'Nancy F. Chen', 'Liangming Pan', 'Anh Tai Tran', 'Shafiq Joty', 'Bowei Zou', 'Xuan Long Do'] | 2023-05-04 | null | null | null | null | ['question-generation'] | ['natural-language-processing'] | [ 2.69580781e-01 4.56807852e-01 5.68538867e-02 -6.08126640e-01
-8.43475342e-01 -7.97257841e-01 7.37240911e-01 1.50416434e-01
3.70590799e-02 7.01180935e-01 5.59994161e-01 -5.60543239e-01
-1.25444055e-01 -9.02862668e-01 -1.91796809e-01 -2.93836594e-01
6.05806708e-01 6.22560620e-01 4.12552625e-01 -6.30188048e-01
4.77780282e-01 -3.89842056e-02 -1.64209294e+00 8.44885349e-01
1.24554765e+00 1.12255955e+00 4.00578439e-01 8.86811018e-01
-9.81255829e-01 1.04649103e+00 -8.57898653e-01 -6.59674644e-01
-6.10307828e-02 -9.73024726e-01 -1.38858986e+00 1.84903339e-01
3.38370174e-01 -8.84080902e-02 -2.37904731e-02 9.01110649e-01
3.16876024e-01 2.45278865e-01 4.11320567e-01 -1.57204235e+00
-3.71961296e-01 8.37915659e-01 2.24688962e-01 4.66680713e-02
1.01956141e+00 3.92841965e-01 1.34948015e+00 -7.56017804e-01
5.44134736e-01 1.56765378e+00 1.63758054e-01 9.76726055e-01
-8.10120881e-01 -3.54490042e-01 4.68242258e-01 3.61116111e-01
-8.72061193e-01 -5.20576000e-01 1.02137232e+00 -4.02227014e-01
5.55862308e-01 7.64947116e-01 4.12861705e-01 1.06660688e+00
-2.27040157e-01 9.58427548e-01 1.17175984e+00 -2.84046769e-01
4.11234558e-01 2.96995342e-01 3.85383189e-01 4.39663172e-01
-2.42459193e-01 -4.23145860e-01 -6.94419742e-01 -4.39590782e-01
1.62458196e-01 -3.18839669e-01 -6.80832207e-01 1.03136785e-01
-8.49779427e-01 1.03611875e+00 2.86349028e-01 1.01687104e-01
-3.37069243e-01 -2.71521002e-01 1.89947248e-01 3.25717390e-01
1.53350949e-01 7.16581523e-01 -4.46613520e-01 -2.79150218e-01
-7.37418950e-01 6.96808815e-01 1.38633978e+00 1.23694694e+00
7.22767472e-01 -5.49506485e-01 -8.97876143e-01 7.13410378e-01
1.96905494e-01 1.93916932e-02 1.49177611e-01 -1.08581340e+00
7.87373781e-01 1.12650037e+00 5.02588272e-01 -1.06897414e+00
-2.69421756e-01 -1.90315500e-01 -5.93609273e-01 -4.78450388e-01
7.37550020e-01 -1.29263774e-01 -2.64500082e-01 1.80497897e+00
5.78200042e-01 -1.19762570e-01 1.30150259e-01 1.06157959e+00
1.40286493e+00 5.97726405e-01 -3.42653394e-02 -3.21354061e-01
1.77991450e+00 -1.30468225e+00 -9.86515641e-01 -4.79808927e-01
4.60312754e-01 -8.02925706e-01 1.50615299e+00 1.49155036e-01
-7.91140079e-01 -4.75661963e-01 -7.58749843e-01 -9.43272784e-02
-1.13607988e-01 -5.17631918e-02 4.63005841e-01 5.30806959e-01
-8.85169685e-01 1.71971962e-01 1.88583180e-01 -4.00554627e-01
-8.52489620e-02 -1.63996324e-01 1.10508882e-01 -3.23778480e-01
-1.72198379e+00 5.70149422e-01 3.45106535e-02 1.27946004e-01
-4.27690983e-01 -4.37424719e-01 -7.24253058e-01 2.42509946e-01
8.36873293e-01 -9.85447526e-01 1.62518752e+00 -7.54354417e-01
-1.62131941e+00 7.76261210e-01 -6.62251174e-01 -2.52837896e-01
7.02723503e-01 4.06783298e-02 -4.13365841e-01 3.94892722e-01
3.76361877e-01 5.45049608e-01 9.12077010e-01 -1.35266387e+00
-6.41003013e-01 -2.07665682e-01 7.66558230e-01 3.34210604e-01
9.52219740e-02 -1.15647286e-01 -4.91335392e-01 -3.15619051e-01
2.76241839e-01 -6.73315883e-01 1.48272805e-03 -2.93790936e-01
-6.79064453e-01 -8.40196252e-01 6.22490704e-01 -4.81664151e-01
1.49390781e+00 -1.64221704e+00 -2.59423763e-01 -1.88756704e-01
3.06287527e-01 9.48232114e-02 -1.97990239e-01 8.15752983e-01
3.06420952e-01 2.81457454e-01 -1.90279722e-01 -2.86582708e-01
1.47491023e-01 1.16588160e-01 -4.77764577e-01 -2.92298794e-01
3.53624672e-01 7.81786323e-01 -1.16757178e+00 -5.25293350e-01
-2.64311910e-01 -1.82268694e-01 -6.52470410e-01 9.64105010e-01
-7.89900839e-01 6.03417993e-01 -7.42638946e-01 2.84742028e-01
7.33378947e-01 -4.15440619e-01 2.56991506e-01 -2.16265976e-01
2.17916250e-01 8.22188497e-01 -1.15045571e+00 1.32923830e+00
-5.02169669e-01 1.61762536e-01 3.36986691e-01 -6.52359247e-01
1.03547490e+00 3.36852580e-01 -1.39548361e-01 -4.04601306e-01
-8.93616527e-02 2.33647883e-01 -6.40768651e-03 -9.28442478e-01
7.57656991e-01 -7.37577751e-02 -3.33064735e-01 5.46079636e-01
-1.25456899e-01 -4.46959615e-01 3.33589107e-01 5.65863311e-01
1.21561122e+00 -5.24246180e-03 3.65060091e-01 -2.28860587e-01
1.05633140e+00 6.32665008e-02 6.12640500e-01 1.08993936e+00
-3.52533758e-01 7.33009636e-01 1.03554392e+00 -1.52872443e-01
-1.69609532e-01 -7.96971500e-01 4.27912652e-01 1.09173000e+00
3.77824187e-01 -6.04613721e-01 -1.01123559e+00 -1.20846629e+00
-2.75273532e-01 1.16449583e+00 -3.66347730e-01 2.76658982e-02
-7.08907783e-01 -5.60099967e-02 2.20293880e-01 -2.88455468e-02
6.42249405e-01 -1.34395146e+00 -4.41930175e-01 3.44102144e-01
-1.10500908e+00 -1.36249423e+00 -9.43651319e-01 -3.39871943e-01
-4.84141648e-01 -1.50026393e+00 -2.79154897e-01 -4.68099982e-01
5.72259426e-01 3.51028889e-01 1.43268394e+00 5.10239244e-01
3.43238086e-01 4.36321646e-01 -6.65950954e-01 -9.79536697e-02
-6.21189892e-01 1.06471829e-01 -5.48148632e-01 4.76678491e-01
4.99060810e-01 -2.96126455e-01 -1.00599456e+00 5.88007927e-01
-8.26299071e-01 1.43847525e-01 9.95037407e-02 7.09018111e-01
3.23731042e-02 -2.54732966e-01 1.03503942e+00 -1.18630791e+00
1.37246060e+00 -5.43294966e-01 -8.19855407e-02 4.56738055e-01
-3.52536857e-01 4.91232574e-02 9.82180655e-01 -1.23201013e-01
-1.21596110e+00 -3.11464399e-01 -3.36879969e-01 1.18708089e-01
-2.01294124e-01 3.30334544e-01 -6.30587578e-01 5.26401162e-01
5.98874152e-01 2.23874465e-01 -1.16459303e-01 -2.79226005e-01
5.05466759e-01 8.84759128e-01 2.30861321e-01 -8.99424016e-01
4.43845659e-01 1.26202717e-01 -4.52888072e-01 -6.87302709e-01
-1.23927021e+00 -8.07630718e-01 -1.31123364e-01 -4.35395122e-01
7.42627978e-01 -3.59449089e-01 -1.06800079e+00 1.65428609e-01
-1.68261862e+00 -1.73506439e-01 -9.75811929e-02 -1.57335088e-01
-4.96087253e-01 4.00197834e-01 -3.56153518e-01 -1.24460757e+00
-4.47311848e-01 -1.22094166e+00 9.55733538e-01 3.60192448e-01
-6.40291393e-01 -8.46613467e-01 -4.35874790e-01 1.00261414e+00
5.15060127e-01 -1.05041049e-01 1.21700323e+00 -1.08731186e+00
-6.84912324e-01 2.73007154e-02 -2.57846981e-01 -2.26539858e-02
2.76626736e-01 -4.03758377e-01 -1.08159423e+00 1.83965057e-01
4.56753016e-01 -3.50136220e-01 6.69973135e-01 -2.05929741e-01
1.08987570e+00 -9.29257512e-01 -3.47896665e-02 -7.41868764e-02
8.73017907e-01 1.40659079e-01 5.76645553e-01 -6.78768978e-02
2.87801206e-01 1.23624372e+00 7.17591226e-01 3.00796986e-01
9.98647332e-01 7.28190243e-01 4.38151747e-01 4.53819007e-01
-1.74593389e-01 -6.79131627e-01 1.91671863e-01 7.96085775e-01
6.22354865e-01 -6.08273804e-01 -5.60939789e-01 4.82350230e-01
-2.03647947e+00 -1.03947902e+00 -4.68514174e-01 1.93947327e+00
1.09217405e+00 1.44262150e-01 3.70241031e-02 1.96295932e-01
8.89052689e-01 4.21209276e-01 -4.91068125e-01 -4.06577438e-01
5.41766398e-02 -6.34948686e-02 -5.62612116e-01 8.03020716e-01
-5.92210591e-01 9.40061986e-01 4.79589128e+00 8.28541458e-01
-7.09248364e-01 -2.80754175e-02 6.46883368e-01 3.18516791e-01
-8.86923969e-01 3.87841582e-01 -9.31086481e-01 5.61498106e-01
5.16908944e-01 -7.79790506e-02 4.47693139e-01 6.95611715e-01
5.25116444e-01 -2.65705913e-01 -1.36007297e+00 8.79312277e-01
1.70222849e-01 -1.23944438e+00 2.55638689e-01 -5.21040559e-01
3.00256252e-01 -8.58189046e-01 -4.14263874e-01 5.38701952e-01
7.81092122e-02 -7.80803859e-01 8.37252200e-01 5.19215167e-01
3.71450812e-01 -3.64157349e-01 6.63564444e-01 8.08289468e-01
-1.24975407e+00 -9.47870985e-02 -5.90586737e-02 -2.47772723e-01
3.29819620e-01 7.42738008e-01 -1.01899850e+00 6.48143947e-01
3.82752508e-01 -6.36516197e-04 -5.93798399e-01 6.35887444e-01
-8.02473962e-01 7.60116458e-01 3.95580269e-02 -6.06622040e-01
2.89471954e-01 -2.02400252e-01 6.54862463e-01 1.05577731e+00
2.07152605e-01 3.92897546e-01 3.10276359e-01 1.16612923e+00
-1.85105965e-01 2.34397039e-01 -3.03594977e-01 1.05021402e-01
8.64187300e-01 1.38700497e+00 -4.99783188e-01 -3.03478181e-01
-9.00706947e-02 9.30497766e-01 2.97554940e-01 4.73047405e-01
-4.77321148e-01 -4.44068402e-01 3.89613628e-01 3.31869423e-01
-4.71649468e-02 2.52888262e-01 -2.06612065e-01 -1.06888604e+00
3.72764885e-01 -1.21586668e+00 5.98757923e-01 -9.53463078e-01
-1.49439275e+00 8.12670708e-01 -1.53483972e-01 -1.09345984e+00
-5.11473060e-01 -3.18488747e-01 -9.81993556e-01 1.05537784e+00
-1.62536800e+00 -9.71015632e-01 -6.85974777e-01 3.81813318e-01
8.95511091e-01 2.40181610e-01 6.35165691e-01 -1.33648187e-01
-2.67675191e-01 5.09331346e-01 -8.95231485e-01 1.90276653e-02
6.08958602e-01 -1.23577249e+00 3.06066096e-01 6.64850056e-01
-2.08622009e-01 8.57165337e-01 9.10522461e-01 -5.06506443e-01
-1.30379522e+00 -9.65139806e-01 1.41526198e+00 -4.52938229e-01
3.50262403e-01 -4.60792899e-01 -1.04359460e+00 1.27691686e-01
3.95089239e-01 -4.01359379e-01 4.84462529e-01 -3.58543210e-02
-3.20037395e-01 -3.38181183e-02 -1.29335630e+00 7.85970807e-01
1.09111428e+00 -5.26992977e-01 -9.38810587e-01 3.59725416e-01
1.11754990e+00 -2.08109230e-01 -1.59538791e-01 1.47769183e-01
3.48437689e-02 -1.41604066e+00 4.75863963e-01 -6.06469452e-01
5.42937160e-01 -4.35166240e-01 -1.50078386e-01 -1.30243099e+00
1.91027910e-01 -1.19580138e+00 -1.50147557e-01 1.53260040e+00
5.43313742e-01 -6.77264988e-01 7.98458636e-01 8.15073371e-01
-2.59785414e-01 -8.42805803e-01 -8.30653965e-01 -4.14856881e-01
-2.00189441e-01 -4.00217623e-01 1.13627529e+00 7.71636486e-01
2.79318899e-01 7.89481521e-01 -6.52957931e-02 -1.30978003e-01
3.37581396e-01 5.89715660e-01 9.11489546e-01 -1.16595566e+00
-2.32258409e-01 -4.84757006e-01 3.95358384e-01 -1.77519751e+00
1.07718311e-01 -7.80245781e-01 3.00502092e-01 -2.03675103e+00
-1.05636336e-01 -3.71618330e-01 3.39619339e-01 8.30022544e-02
-7.25812793e-01 -6.52773738e-01 2.48130485e-01 2.89389938e-02
-5.50302088e-01 6.46786034e-01 1.53356278e+00 -1.79836541e-01
-2.63614327e-01 3.66346389e-01 -1.07508087e+00 6.24499559e-01
8.23358297e-01 -2.76922822e-01 -7.28238881e-01 4.44678217e-02
4.93623823e-01 6.67038620e-01 1.95080906e-01 -5.63669086e-01
5.07224143e-01 -4.28628236e-01 -4.73223507e-01 -7.48056531e-01
3.87497216e-01 -4.15910453e-01 -3.43542010e-01 1.77457482e-01
-6.79022312e-01 -1.64648108e-02 -4.07483310e-01 6.13474667e-01
-3.82911086e-01 -5.61990201e-01 4.51264292e-01 -4.61373240e-01
-4.16780114e-01 2.17311293e-01 -3.58082980e-01 7.43828058e-01
5.07795036e-01 -1.52966052e-01 -6.59588635e-01 -1.02399921e+00
-4.28542465e-01 6.29608631e-01 -2.04850305e-02 6.44089758e-01
8.03079128e-01 -1.13411999e+00 -7.41475344e-01 -1.13469392e-01
4.74648356e-01 7.24208727e-02 3.96797240e-01 7.08532810e-01
-8.13815072e-02 3.90324384e-01 6.07832313e-01 -3.63557726e-01
-9.88635123e-01 4.48455185e-01 3.20340514e-01 -4.98937100e-01
-8.52597952e-02 7.47115970e-01 3.80101115e-01 -7.62401998e-01
1.78510904e-01 -2.48674512e-01 -6.68637872e-01 2.26880714e-01
5.91999114e-01 2.38460451e-01 1.58120207e-02 -3.13472390e-01
-2.36466780e-01 7.64550343e-02 1.63133591e-01 -1.66214183e-01
6.61874235e-01 -3.86416316e-01 -3.65939498e-01 2.90092409e-01
9.88703012e-01 1.50845692e-01 -8.61891866e-01 -2.39908233e-01
3.08280408e-01 -3.50652635e-01 -6.05990767e-01 -9.26830471e-01
-5.67556977e-01 9.56574202e-01 -2.90591657e-01 9.14599299e-01
8.93351555e-01 1.52449429e-01 1.05046046e+00 3.56846660e-01
2.91935295e-01 -1.05231297e+00 4.05446917e-01 7.65044153e-01
1.41773033e+00 -1.10447478e+00 -4.43574548e-01 -9.22312200e-01
-7.66576231e-01 1.15560400e+00 1.05080283e+00 4.35838163e-01
3.34417641e-01 -2.34619603e-01 2.75861859e-01 -3.77922118e-01
-1.07407033e+00 -2.78902799e-01 3.08218598e-01 3.79614949e-01
4.24278408e-01 9.31716263e-02 -5.57632148e-01 1.00695360e+00
-5.92826843e-01 -2.49437898e-01 6.64602935e-01 6.32052720e-01
-5.91132462e-01 -1.01514518e+00 -3.23596328e-01 5.18752396e-01
-1.44285575e-01 -1.18556961e-01 -9.66268659e-01 2.67253578e-01
-1.99281320e-01 1.99035120e+00 -2.57628024e-01 -4.08384591e-01
4.98094618e-01 3.68431419e-01 9.21506658e-02 -8.19610894e-01
-1.04712439e+00 -3.41654658e-01 7.59560049e-01 -4.70999360e-01
-2.80362457e-01 -2.85157949e-01 -1.24937868e+00 -9.90360379e-02
-4.54310209e-01 7.05141366e-01 2.94019610e-01 1.18791270e+00
5.56509972e-01 2.94034034e-01 9.87940788e-01 -1.26719192e-01
-7.93273747e-01 -9.83829260e-01 -9.13601816e-02 6.69571042e-01
2.16657951e-01 -3.12139690e-01 -7.83282459e-01 -2.09611908e-01] | [11.842032432556152, 8.070246696472168] |
1eeba2dc-2f81-47e5-a557-f916b2217930 | dependency-grammar-induction-with-neural | 1708.00801 | null | http://arxiv.org/abs/1708.00801v1 | http://arxiv.org/pdf/1708.00801v1.pdf | Dependency Grammar Induction with Neural Lexicalization and Big Training Data | We study the impact of big models (in terms of the degree of lexicalization)
and big data (in terms of the training corpus size) on dependency grammar
induction. We experimented with L-DMV, a lexicalized version of Dependency
Model with Valence and L-NDMV, our lexicalized extension of the Neural
Dependency Model with Valence. We find that L-DMV only benefits from very small
degrees of lexicalization and moderate sizes of training corpora. L-NDMV can
benefit from big training data and lexicalization of greater degrees,
especially when enhanced with good model initialization, and it achieves a
result that is competitive with the current state-of-the-art. | ['Kewei Tu', 'Yong Jiang', 'Wenjuan Han'] | 2017-08-02 | dependency-grammar-induction-with-neural-1 | https://aclanthology.org/D17-1176 | https://aclanthology.org/D17-1176.pdf | emnlp-2017-9 | ['dependency-grammar-induction'] | ['natural-language-processing'] | [-5.23891330e-01 6.02722168e-01 -5.56078792e-01 -3.31446737e-01
-6.04508460e-01 -7.97851562e-01 5.40784061e-01 1.05630226e-01
-6.06974006e-01 7.54023910e-01 4.86758381e-01 -7.38270164e-01
3.07643831e-01 -8.37793887e-01 -6.66764140e-01 -3.10618341e-01
-1.34813473e-01 1.02980816e+00 1.14636287e-01 -5.39306998e-01
-2.26914465e-01 1.04431540e-01 -9.51285958e-01 1.44955099e-01
6.74757719e-01 3.90151739e-01 3.09747010e-01 3.12371165e-01
-1.35296121e-01 6.58139348e-01 -4.77353215e-01 -4.37232256e-01
1.42405599e-01 -3.23937804e-01 -8.14649045e-01 -1.82581440e-01
2.12429658e-01 -6.42772987e-02 -9.99863967e-02 7.51260042e-01
6.50484383e-01 8.63328725e-02 5.72564960e-01 -7.92418778e-01
-7.93141305e-01 1.54291534e+00 -5.16300678e-01 6.37103558e-01
-1.49544090e-01 2.41031259e-01 1.12400961e+00 -9.31772590e-01
1.11601746e+00 1.56904352e+00 4.63001728e-01 6.64596200e-01
-1.54506278e+00 -5.91378093e-01 5.39493203e-01 -1.33571759e-01
-1.25258219e+00 -2.70630866e-01 4.11003292e-01 -3.80892634e-01
1.92074037e+00 -3.55390489e-01 4.12505627e-01 1.23103023e+00
4.75930050e-02 4.26715702e-01 1.03109407e+00 -6.86575115e-01
2.37858698e-01 -1.14386693e-01 7.38646209e-01 5.53633571e-01
3.74367893e-01 3.24366391e-01 -2.57415593e-01 1.09145269e-01
7.75606453e-01 -7.81861663e-01 1.17545120e-01 3.92874703e-02
-7.81669497e-01 1.33124483e+00 2.18358770e-01 4.28206772e-01
-2.84736693e-01 1.96122646e-01 5.76690555e-01 6.88212633e-01
4.83210623e-01 7.22054124e-01 -9.52491701e-01 -1.32591501e-01
-4.78763610e-01 3.29548344e-02 9.83698368e-01 1.25945342e+00
7.13249445e-01 4.47620302e-01 -1.37369037e-01 9.05264556e-01
2.22076520e-01 4.79037434e-01 4.60265726e-01 -8.39422464e-01
8.94323528e-01 5.08221030e-01 -3.88077825e-01 -5.04197404e-02
-9.25759375e-01 -5.46741009e-01 -4.96926337e-01 -1.53224140e-01
3.16495448e-01 -7.44591475e-01 -1.13339365e+00 2.50623012e+00
3.92909050e-02 -1.74314678e-01 1.19232126e-01 3.82684857e-01
9.77749407e-01 6.36187851e-01 4.75139141e-01 -5.02594054e-01
1.33588278e+00 -7.77026355e-01 -6.58007026e-01 -9.25094068e-01
1.38060379e+00 -3.38905871e-01 1.49893999e+00 2.66023219e-01
-1.34395921e+00 -6.67293668e-01 -1.00024784e+00 -1.82926297e-01
-1.76381618e-01 -4.11387056e-01 1.08061647e+00 8.19932938e-01
-1.29137003e+00 4.09055591e-01 -7.10929990e-01 -3.76006633e-01
7.53933787e-02 6.75150812e-01 -5.28825045e-01 -2.26316392e-01
-1.51268828e+00 1.20228028e+00 8.40589225e-01 -2.02280909e-01
-7.81574130e-01 -4.35705453e-01 -1.16246426e+00 -1.41451340e-02
4.53330487e-01 -7.10590184e-01 1.20746756e+00 -4.55779940e-01
-1.00121844e+00 8.14400017e-01 1.61926791e-01 -1.79637685e-01
-1.85621351e-01 1.24141254e-01 -1.12495795e-01 -3.34488362e-01
-8.13628361e-03 9.78036463e-01 5.49593866e-01 -1.08506811e+00
-4.84465629e-01 -6.38561964e-01 2.40632251e-01 2.02078179e-01
-2.68078715e-01 1.63442984e-01 -5.89296818e-01 -4.88507926e-01
-2.26861879e-01 -1.18176162e+00 -3.70704174e-01 -1.20434320e+00
-2.20618695e-01 -6.18823886e-01 4.89673108e-01 -4.91448939e-01
1.42961955e+00 -1.86683977e+00 5.18617332e-01 1.26458138e-01
2.08767161e-01 1.58289924e-01 -4.77607459e-01 2.96516091e-01
-5.45558445e-02 6.04984403e-01 1.45279258e-01 -4.68921572e-01
-3.13899964e-02 7.41185904e-01 2.30068237e-01 -3.31936926e-02
1.07189015e-01 1.13586450e+00 -5.78940630e-01 -6.30447805e-01
-4.61210683e-02 3.05113524e-01 -8.16884756e-01 7.43989870e-02
-5.65933526e-01 1.78472430e-01 -2.45769039e-01 3.62371862e-01
3.41055185e-01 -2.85700887e-01 6.86638594e-01 1.29169360e-01
5.34471869e-02 5.98251939e-01 -9.95446265e-01 1.61204600e+00
-8.01985621e-01 3.12718987e-01 1.55877635e-01 -6.17821395e-01
6.55604541e-01 3.62991452e-01 -2.42764428e-01 -7.97807515e-01
4.50824261e-01 1.49086744e-01 6.26733065e-01 -1.77268296e-01
2.00114921e-01 -3.97586107e-01 -4.29349452e-01 6.95267260e-01
5.22072196e-01 -5.31620160e-03 7.82429278e-01 5.99056959e-01
9.79616880e-01 -1.93890586e-01 2.67774224e-01 -5.08198023e-01
-1.34516433e-01 -1.68640226e-01 6.27855122e-01 6.34529650e-01
2.19318807e-01 -4.16715480e-02 9.83378768e-01 -2.57233530e-01
-1.08460617e+00 -7.14770555e-01 -3.35209072e-01 1.70575511e+00
-2.53506929e-01 -7.55596399e-01 -8.76603186e-01 -9.68677104e-01
-2.02797025e-01 1.14691663e+00 -7.44470716e-01 -8.91252831e-02
-1.12890828e+00 -1.24174714e+00 5.45496166e-01 1.08728266e+00
-2.91071348e-02 -1.62321675e+00 -1.84620008e-01 2.49613479e-01
-3.32957394e-02 -1.40384793e+00 -2.67577946e-01 1.01322412e+00
-1.07702708e+00 -4.85237956e-01 -3.92924808e-03 -1.23000944e+00
4.70814764e-01 -4.97516870e-01 1.92811513e+00 1.46664277e-01
1.80221334e-01 -4.24437433e-01 -5.50649405e-01 -4.16041851e-01
-7.83491731e-01 5.98138928e-01 -7.31644556e-02 -9.84880269e-01
4.56976593e-01 -7.38875389e-01 1.70573257e-02 1.46958083e-01
-7.74423003e-01 -1.60908297e-01 7.92665839e-01 8.68911088e-01
4.58230227e-01 -2.76527107e-01 7.07843721e-01 -1.47399211e+00
8.50893855e-01 -3.81005138e-01 -4.66537863e-01 -4.95177433e-02
-8.48823071e-01 6.73169136e-01 5.27765036e-01 -4.67171758e-01
-9.11757946e-01 -4.14588243e-01 -6.01668358e-01 -9.87474900e-03
-3.06274043e-03 8.11726630e-01 -4.74154234e-01 1.53840840e-01
7.71461248e-01 -5.93542278e-01 -3.98517102e-01 -5.36264300e-01
9.16511297e-01 4.15494144e-01 1.35241926e-01 -7.45051742e-01
1.60178974e-01 -3.09661746e-01 -1.96833476e-01 -5.34664929e-01
-6.74738824e-01 9.34052542e-02 -9.14948583e-01 5.18686116e-01
9.05120194e-01 -9.86270189e-01 -1.99512571e-01 1.75271466e-01
-1.13861442e+00 -9.73798454e-01 -3.10242474e-01 5.39960563e-01
-2.14437231e-01 1.94072351e-01 -1.26681411e+00 -4.90590110e-02
-3.65065336e-01 -1.13109338e+00 7.62834311e-01 -2.04685465e-01
-3.49545300e-01 -1.41822863e+00 2.16784611e-01 1.04812672e-03
2.10857689e-01 -7.05357082e-03 1.68815303e+00 -8.82927954e-01
-9.74369347e-02 1.15457386e-01 -1.81775503e-02 3.35871309e-01
-3.52592081e-01 -2.99810052e-01 -7.03552365e-01 -3.56081516e-01
-9.31793451e-02 -7.08688140e-01 1.05845606e+00 8.88667345e-01
3.27245593e-01 -1.21498562e-01 -3.37035716e-01 7.32281268e-01
1.35831940e+00 6.97642118e-02 4.92897213e-01 2.53888637e-01
1.04546285e+00 3.84073079e-01 4.03830945e-01 2.89181501e-01
6.15873814e-01 6.43880308e-01 4.12548602e-01 -2.72005081e-01
-3.25943261e-01 -1.81603089e-01 3.73390436e-01 1.29226482e+00
-1.17837109e-01 -4.59181368e-01 -1.03400469e+00 4.42426115e-01
-1.62514400e+00 -5.14859974e-01 -5.20907752e-02 1.90165651e+00
1.21739757e+00 6.62697613e-01 2.25102514e-01 -1.14242196e-01
5.69093764e-01 3.71811181e-01 -3.35776478e-01 -1.14549005e+00
-4.14020449e-01 5.67341089e-01 4.41841662e-01 9.49340582e-01
-8.86090636e-01 1.80559289e+00 7.51537371e+00 8.86078298e-01
-8.67540300e-01 5.57364404e-01 4.78373587e-01 -2.14532524e-01
-5.02183318e-01 1.54424593e-01 -1.55485356e+00 1.47260934e-01
1.37355530e+00 1.43246651e-01 3.64418685e-01 8.91994774e-01
-2.82999665e-01 1.80754215e-01 -1.43114245e+00 5.29619694e-01
4.75933077e-03 -1.11794364e+00 7.74892373e-03 3.63876998e-01
8.18076968e-01 6.44462764e-01 -3.16963941e-01 9.56319988e-01
8.63356113e-01 -1.06920302e+00 4.04309183e-01 -5.59109986e-01
8.32832932e-01 -9.65609193e-01 9.62069809e-01 6.07238531e-01
-1.05475557e+00 -2.26948395e-01 -6.62483990e-01 -3.32528979e-01
3.72257859e-01 4.76710379e-01 -6.70912087e-01 2.22509243e-02
2.00392857e-01 2.55147576e-01 -8.08716357e-01 -1.09674901e-01
-8.64137292e-01 8.66990089e-01 -5.24052918e-01 3.06275394e-02
2.69892126e-01 9.11261290e-02 4.32170890e-02 1.35731518e+00
-1.45924300e-01 4.41467986e-02 4.63910371e-01 5.04965603e-01
-4.01013821e-01 2.65633166e-01 -6.95891261e-01 8.35597739e-02
8.58848631e-01 1.14940965e+00 -6.46075547e-01 -4.08186704e-01
-5.14286160e-01 1.80970967e-01 8.76364768e-01 1.58372328e-01
-7.29825616e-01 -1.11371838e-02 3.36123824e-01 1.21469572e-02
6.17911041e-01 -3.24596137e-01 -4.93139684e-01 -1.02281070e+00
-3.05688173e-01 -1.08717024e+00 6.18134439e-01 -3.81087035e-01
-9.19939518e-01 1.13749766e+00 1.80996686e-01 -3.41213733e-01
-8.74049246e-01 -5.80883503e-01 -3.48531812e-01 8.91063631e-01
-1.06495202e+00 -1.33782697e+00 3.12389225e-01 7.24321365e-01
5.78815460e-01 4.29968052e-02 1.07890308e+00 9.44841281e-02
-8.22754264e-01 9.86092985e-01 -2.39086837e-01 2.32860625e-01
5.27540207e-01 -1.26310766e+00 1.04767680e+00 8.11706185e-01
3.19622129e-01 6.40070856e-01 6.78723097e-01 -7.20806301e-01
-1.09012628e+00 -1.05339277e+00 1.13877106e+00 -8.41210127e-01
5.64750373e-01 -7.78268397e-01 -7.93900311e-01 1.23585618e+00
3.63116056e-01 -5.90922832e-02 6.40906155e-01 1.05889392e+00
-5.03629148e-01 2.62626827e-01 -7.93583930e-01 4.23664659e-01
1.24342442e+00 -2.78699815e-01 -7.91142166e-01 3.34260672e-01
1.09023106e+00 -1.57868415e-01 -1.11609983e+00 4.62172687e-01
1.37217522e-01 -7.15214849e-01 6.92343295e-01 -7.68489778e-01
1.86002165e-01 5.02687156e-01 -2.51017064e-01 -1.69361794e+00
-7.56621301e-01 -3.47138464e-01 -1.09024107e-01 1.14273965e+00
1.04770386e+00 -4.92398977e-01 6.18411899e-01 5.15005291e-01
-3.10845643e-01 -6.98696733e-01 -7.63952672e-01 -1.04866755e+00
7.59158254e-01 -5.78149498e-01 4.10338551e-01 9.98104155e-01
2.70152032e-01 1.40270495e+00 -1.24795079e-01 -1.14594758e-01
3.78561497e-01 7.72094205e-02 6.14655852e-01 -1.36791909e+00
-6.64507210e-01 -1.94879532e-01 -6.55275211e-02 -7.48709738e-01
6.12727463e-01 -1.08390045e+00 -1.43253163e-01 -1.63684344e+00
1.52337730e-01 -3.38632613e-01 -2.48452023e-01 8.55457425e-01
-4.27288234e-01 1.04058929e-01 2.80813813e-01 1.11221084e-02
-3.68996322e-01 1.73492298e-01 1.02237427e+00 2.25657076e-01
-6.19671524e-01 -4.72005337e-01 -1.10243738e+00 8.01029980e-01
7.68360794e-01 -7.24749625e-01 -6.14633679e-01 -7.21465051e-01
5.38541198e-01 1.88260242e-01 -5.72212160e-01 -5.38722396e-01
-3.28402400e-01 2.26212572e-03 1.24272406e-01 -2.90609717e-01
1.07257485e-01 6.96811751e-02 -2.60483235e-01 2.65696913e-01
-3.54815871e-01 4.91967291e-01 4.87807781e-01 1.57807246e-02
1.58468261e-02 -1.55653313e-01 8.35314512e-01 -3.55591089e-01
-7.75076509e-01 2.30998963e-01 -3.67388427e-01 7.62993693e-01
6.90544903e-01 9.81191024e-02 -6.19701326e-01 -1.35343730e-01
-9.65792239e-01 2.36929804e-01 4.14545685e-01 5.49499512e-01
1.11056291e-01 -1.08827841e+00 -7.59893477e-01 5.61737001e-01
4.17764299e-02 -3.42788883e-02 -2.13645846e-01 7.72844195e-01
-1.78183690e-01 5.03218472e-01 -1.28385618e-01 -2.95546979e-01
-1.20978880e+00 1.02090216e+00 -2.77869433e-01 -8.50778639e-01
-5.00169516e-01 1.17750788e+00 5.58144212e-01 -6.36654317e-01
2.16179863e-01 -5.63276589e-01 -3.70228767e-01 2.03278363e-01
1.50188476e-01 -9.75449532e-02 2.62615651e-01 -5.12623310e-01
-3.79454166e-01 5.75001061e-01 -4.12726104e-01 -3.67229193e-01
1.28745031e+00 9.03848037e-02 -5.96562177e-02 2.01621875e-01
8.81068945e-01 -6.89735590e-03 -8.84247184e-01 -8.17841887e-02
9.75148454e-02 1.42942667e-01 3.05480212e-01 -8.98313701e-01
-1.21243930e+00 9.96429801e-01 1.23924546e-01 -9.27374586e-02
8.80252361e-01 5.79754233e-01 6.19944811e-01 6.40961230e-01
6.15221918e-01 -1.25368881e+00 -8.51974413e-02 1.05318820e+00
5.64225852e-01 -1.05714488e+00 1.49341524e-02 -4.21621919e-01
-9.73235190e-01 4.11783189e-01 8.80762577e-01 -1.30292132e-01
8.09845626e-01 9.18684304e-01 4.09699142e-01 -3.54196191e-01
-1.26435637e+00 -4.48732018e-01 -2.86241382e-01 6.47192538e-01
6.37868106e-01 1.69986457e-01 -6.01671934e-01 9.71619308e-01
-3.04849625e-01 -4.09515351e-01 2.77791321e-01 7.79064178e-01
-6.36408091e-01 -1.59754455e+00 -9.62865818e-03 3.54685098e-01
-5.15968442e-01 -8.29649866e-01 -5.41772664e-01 1.25234556e+00
2.95184672e-01 9.80133712e-01 6.22263439e-02 -4.44630235e-01
3.47005844e-01 2.33715251e-01 7.16075838e-01 -1.19903076e+00
-9.46231544e-01 5.67876518e-01 7.56441414e-01 -4.39390182e-01
1.51180342e-01 -3.81307483e-01 -1.51718903e+00 -3.49808812e-01
-7.89731860e-01 2.37496495e-01 5.23778260e-01 1.01585019e+00
2.84836143e-01 3.45452577e-01 2.99887747e-01 -7.94316709e-01
-6.83786392e-01 -1.45964301e+00 -9.25403595e-01 4.92617607e-01
-2.33446404e-01 -5.90180159e-01 -4.25145745e-01 -2.32473299e-01] | [10.609827041625977, 9.514890670776367] |
a01c00d5-49e4-4d22-9244-ba3b29340685 | a-study-on-the-integration-of-pipeline-and | 2305.01620 | null | https://arxiv.org/abs/2305.01620v2 | https://arxiv.org/pdf/2305.01620v2.pdf | A Study on the Integration of Pipeline and E2E SLU systems for Spoken Semantic Parsing toward STOP Quality Challenge | Recently there have been efforts to introduce new benchmark tasks for spoken language understanding (SLU), like semantic parsing. In this paper, we describe our proposed spoken semantic parsing system for the quality track (Track 1) in Spoken Language Understanding Grand Challenge which is part of ICASSP Signal Processing Grand Challenge 2023. We experiment with both end-to-end and pipeline systems for this task. Strong automatic speech recognition (ASR) models like Whisper and pretrained Language models (LM) like BART are utilized inside our SLU framework to boost performance. We also investigate the output level combination of various models to get an exact match accuracy of 80.8, which won the 1st place at the challenge. | ['Shinji Watanabe', 'Brian Yan', 'Emiru Tsunoo', 'Yosuke Kashiwagi', 'Yifan Peng', 'Jessica Huynh', 'Shih-Lun Wu', 'Hayato Futami', 'Siddhant Arora'] | 2023-05-02 | null | null | null | null | ['spoken-language-understanding', 'semantic-parsing', 'spoken-language-understanding'] | ['natural-language-processing', 'natural-language-processing', 'speech'] | [ 1.10712051e-01 3.99234384e-01 3.42573285e-01 -1.10164690e+00
-1.53479481e+00 -5.68541348e-01 6.47836804e-01 -2.68861592e-01
-5.57779908e-01 3.44400436e-01 7.49029219e-01 -4.09022242e-01
5.06977260e-01 -1.89136505e-01 -6.20774448e-01 2.92384950e-03
1.30930077e-02 5.73444724e-01 1.95196152e-01 -3.63071084e-01
1.00707479e-01 -1.00269079e-01 -1.05895782e+00 7.90683329e-01
7.54275024e-01 1.00964499e+00 3.19668740e-01 1.32748032e+00
-5.90276361e-01 7.92870820e-01 -7.61526227e-01 7.44783208e-02
-9.16435048e-02 -6.08078182e-01 -1.30821884e+00 -1.28529534e-01
4.34288681e-01 -1.34828508e-01 -1.19150355e-01 9.42612350e-01
7.74864912e-01 2.05189139e-01 2.67959744e-01 -1.16965950e+00
-4.09312040e-01 1.14065421e+00 -2.67704993e-01 3.51108074e-01
7.24576950e-01 -2.86614113e-02 1.08044183e+00 -1.17958629e+00
3.46994221e-01 1.88939655e+00 2.63436973e-01 9.38593328e-01
-5.05200922e-01 -6.85368240e-01 3.66417438e-01 3.05724978e-01
-1.17624176e+00 -1.10781145e+00 4.16332752e-01 2.53817886e-01
1.61327744e+00 9.20836031e-02 -1.97375327e-01 1.23003423e+00
-4.07722563e-01 1.51774764e+00 1.04997778e+00 -5.23890793e-01
3.98140907e-01 5.62685803e-02 8.83804321e-01 6.47257805e-01
-7.43988693e-01 -3.06916058e-01 -9.61577117e-01 1.25764966e-01
1.77317664e-01 -8.39825749e-01 -3.64790231e-01 5.52350163e-01
-1.04112720e+00 6.11366630e-01 2.36282066e-01 3.09902787e-01
-1.46946058e-01 2.56167114e-01 6.28768206e-01 5.28427422e-01
6.59994364e-01 9.67561230e-02 -8.52515459e-01 -3.80774617e-01
-8.48891497e-01 -5.44650033e-02 1.17805421e+00 9.36397731e-01
2.88262039e-01 2.46840566e-01 -3.04992259e-01 1.42241156e+00
6.49708629e-01 5.02091110e-01 4.77817416e-01 -8.85424256e-01
7.37304270e-01 3.37759666e-02 -1.96841493e-01 -5.39424047e-02
-5.70062220e-01 -2.34959930e-01 -5.36744475e-01 -4.15818214e-01
2.27569267e-01 -2.96172202e-01 -1.41719830e+00 1.74777663e+00
-8.96307230e-02 7.08485723e-01 6.88977420e-01 1.00136077e+00
1.28983223e+00 1.16562700e+00 4.77667391e-01 1.38034552e-01
1.71503663e+00 -1.53873456e+00 -8.97521973e-01 -8.09275806e-01
8.76303077e-01 -8.12227428e-01 1.31395006e+00 4.61311698e-01
-1.12026381e+00 -5.68574131e-01 -9.21694219e-01 -2.05883726e-01
-2.71246791e-01 7.46330917e-02 9.96739864e-02 6.51803017e-01
-1.57131600e+00 1.67243525e-01 -7.55938411e-01 -6.51592493e-01
2.75256597e-02 -5.52582666e-02 -1.81450874e-01 -1.37337729e-01
-1.52160072e+00 8.33690941e-01 2.49416664e-01 -3.58069874e-02
-9.83987451e-01 -6.68021798e-01 -1.04866087e+00 7.32494518e-02
2.50486940e-01 -3.49923760e-01 1.87469399e+00 -8.42922568e-01
-1.94676352e+00 1.08240342e+00 -6.19362533e-01 -7.89411128e-01
-1.85928419e-02 -6.74350560e-01 -6.03830338e-01 1.73643440e-01
-1.55849293e-01 1.01327991e+00 4.71358925e-01 -1.01121390e+00
-5.35135984e-01 -4.32677627e-01 -3.10381263e-01 4.25845444e-01
1.47470444e-01 6.19921148e-01 -4.37368274e-01 -3.46674651e-01
2.62885422e-01 -5.82613409e-01 -7.92436525e-02 -6.92052245e-01
-4.55208033e-01 -6.51692927e-01 7.72980392e-01 -1.17750907e+00
1.04315376e+00 -2.05913043e+00 1.28885005e-02 -9.96975452e-02
-5.28601229e-01 2.66283184e-01 -5.76370776e-01 3.57725859e-01
4.15615514e-02 2.50428617e-01 -3.28740001e-01 -1.15797782e+00
2.44780868e-01 3.74556243e-01 -5.92864752e-01 -1.97964057e-01
4.22805637e-01 1.06836867e+00 -7.05243587e-01 4.21361299e-03
2.59898156e-01 3.00077766e-01 -2.31084898e-01 5.48769951e-01
-3.55700195e-01 4.77542311e-01 -3.73793840e-01 5.53023875e-01
5.42483747e-01 1.01344727e-01 -1.77047938e-01 2.04446971e-01
3.58991548e-02 1.10820282e+00 -1.07897484e+00 2.34818888e+00
-7.35420465e-01 3.66704941e-01 6.12795591e-01 -9.74519491e-01
9.14666593e-01 7.08172023e-01 -2.89049447e-01 -8.53117824e-01
2.17703074e-01 2.65838355e-01 -2.49563038e-01 -2.98003048e-01
2.90198505e-01 -1.10614501e-01 -3.15349728e-01 4.19545740e-01
6.29741549e-01 -4.88115132e-01 -4.30504501e-01 3.99422169e-01
1.17178369e+00 -7.01537281e-02 2.56338090e-01 -6.13962173e-01
7.39627898e-01 -1.35405898e-01 4.70559150e-02 9.10799026e-01
-4.69620943e-01 9.65662301e-01 1.93424806e-01 -8.59045237e-02
-7.83255994e-01 -1.17091906e+00 1.48126706e-01 1.38885701e+00
-2.31675342e-01 -4.06711042e-01 -1.13183606e+00 -5.62724411e-01
-6.78436697e-01 1.08124530e+00 2.46102400e-02 8.23086351e-02
-5.07148445e-01 -5.45790136e-01 1.09253824e+00 6.98667765e-01
8.18620980e-01 -1.27287567e+00 1.13212153e-01 4.35148925e-01
-3.56346190e-01 -1.95586526e+00 -5.00999510e-01 4.55710255e-02
-5.15079141e-01 -7.74528623e-01 -7.82843292e-01 -8.55428815e-01
-1.40405074e-01 1.38637289e-01 1.24385464e+00 -2.20050290e-01
-8.11044406e-03 6.38689041e-01 -7.35428810e-01 -6.52356386e-01
-7.25216985e-01 1.46442488e-01 -7.36575499e-02 -6.40277863e-02
4.37362105e-01 -8.55665877e-02 -3.54949147e-01 1.56432152e-01
-6.00948751e-01 1.21287726e-01 9.52284038e-02 6.54717326e-01
2.11863577e-01 -6.89884126e-01 1.13434482e+00 -6.20955706e-01
8.13153923e-01 -2.29066461e-01 -2.38337874e-01 5.04175663e-01
4.80574630e-02 5.80813251e-02 4.19010639e-01 2.23450467e-01
-1.24509394e+00 1.04890741e-01 -1.00157452e+00 1.88605879e-02
-4.83810067e-01 5.15071630e-01 -5.37119746e-01 3.90186548e-01
4.16528285e-01 4.23827648e-01 -6.79683536e-02 -6.48619890e-01
5.85408330e-01 1.24959767e+00 7.81928003e-01 -5.22268176e-01
7.83060938e-02 2.33303029e-02 -8.61790597e-01 -1.40072536e+00
-1.18665099e+00 -8.45614910e-01 -2.80721337e-01 1.19554915e-01
1.13782573e+00 -1.40295613e+00 -3.78177285e-01 8.29947293e-01
-1.57073200e+00 -5.37730753e-01 2.26538181e-01 4.23866838e-01
-4.86787081e-01 4.47462618e-01 -7.78341413e-01 -1.08188009e+00
-9.33671117e-01 -1.12281799e+00 1.53833508e+00 2.42305547e-01
-3.43226939e-01 -8.71231914e-01 -1.39219120e-01 8.42902899e-01
7.31270075e-01 -6.08373463e-01 4.36953604e-01 -1.23059773e+00
-3.19775760e-01 1.32094711e-01 -2.66239345e-01 8.49794865e-01
-3.61038804e-01 -3.81177843e-01 -1.58059883e+00 -4.22393121e-02
1.23497704e-02 -6.26500189e-01 1.02788115e+00 3.34733635e-01
9.82315302e-01 9.22934636e-02 5.92191704e-02 2.35638559e-01
7.61387110e-01 1.53488323e-01 7.18594253e-01 -1.60542473e-01
5.18903315e-01 8.25336218e-01 3.09743345e-01 9.56418514e-02
7.94742346e-01 5.53948998e-01 -8.01241770e-02 1.49537213e-02
-3.02245349e-01 -1.05940796e-01 8.20195198e-01 1.21479964e+00
6.63442135e-01 -5.13179302e-01 -1.20111382e+00 5.29343367e-01
-1.76261902e+00 -3.02026838e-01 -2.59525180e-01 1.80820203e+00
7.15168118e-01 1.95101127e-01 -2.63298601e-01 -1.91787139e-01
3.64186257e-01 3.96328956e-01 -6.50498495e-02 -8.26716542e-01
-1.64858356e-01 6.29183769e-01 1.18506297e-01 1.22546554e+00
-1.09154356e+00 1.68491745e+00 6.19265842e+00 8.67015302e-01
-8.38788927e-01 4.09222394e-01 8.46709251e-01 2.77112544e-01
-1.94302395e-01 -1.61471695e-01 -1.07807732e+00 7.22038373e-02
1.74064136e+00 -4.05798890e-02 2.76443899e-01 6.82806671e-01
5.55748880e-01 -3.03564429e-01 -9.17086959e-01 8.76004219e-01
3.07869703e-01 -9.55762863e-01 1.14215612e-01 -9.14661169e-01
2.74242491e-01 7.73381233e-01 -4.77604955e-01 6.81265414e-01
3.43536258e-01 -1.27095664e+00 4.68463004e-01 5.69696277e-02
4.24404472e-01 -6.78365648e-01 9.07550573e-01 4.47312653e-01
-1.14614892e+00 4.34485078e-01 -7.69867152e-02 5.12580434e-03
8.22482169e-01 3.01428437e-01 -1.08513737e+00 4.97498602e-01
6.93051636e-01 5.11761487e-01 -9.10048634e-02 6.62827611e-01
-5.85611820e-01 1.16206861e+00 -6.07002378e-01 1.34161348e-02
6.16448820e-01 9.33369175e-02 5.63677967e-01 1.58841169e+00
1.43304512e-01 5.21355927e-01 2.68362850e-01 5.00577509e-01
-2.79528737e-01 1.03040613e-01 -2.78826982e-01 -1.79006055e-01
5.05036473e-01 9.10679460e-01 -5.50080299e-01 -6.76038504e-01
-5.75980604e-01 1.36482966e+00 -5.71303023e-03 3.45004052e-01
-4.72804129e-01 5.25901280e-02 9.10573661e-01 -4.92076039e-01
-2.79425830e-01 -4.78434145e-01 -2.37714484e-01 -1.19436371e+00
-2.34023884e-01 -8.01401615e-01 3.83051932e-01 -9.48184311e-01
-1.20371783e+00 9.68452334e-01 -2.42156535e-01 -2.86937684e-01
-3.73531222e-01 -7.06053197e-01 -8.89834464e-01 1.17937279e+00
-1.83109272e+00 -1.25057447e+00 -2.06621990e-01 4.35716301e-01
1.47540617e+00 -1.86752573e-01 1.12764585e+00 2.47503355e-01
-4.34100747e-01 5.56376696e-01 -3.91502380e-01 1.52777895e-01
7.68462241e-01 -1.28311586e+00 1.17614770e+00 8.25052202e-01
3.36286545e-01 1.69631362e-01 6.89795375e-01 -2.92390049e-01
-1.20956719e+00 -9.12434816e-01 1.16185451e+00 -6.67579353e-01
8.28554034e-01 -7.18138933e-01 -1.23988569e+00 6.30881131e-01
6.08254850e-01 -4.06139165e-01 5.62551558e-01 1.04711175e-01
-3.82789612e-01 3.82428885e-01 -7.44498134e-01 6.40314594e-02
8.76618564e-01 -8.58950973e-01 -8.29325080e-01 2.00251713e-01
1.33941495e+00 -4.60538775e-01 -4.51018959e-01 4.91202831e-01
1.09615587e-01 -5.47198653e-01 8.71525109e-01 -6.96606874e-01
2.44166002e-01 -6.21116720e-03 -4.84926611e-01 -1.61582148e+00
5.06897628e-01 -4.43202406e-01 4.39115435e-01 1.25417352e+00
6.57593906e-01 -4.54632282e-01 5.31217515e-01 5.17421007e-01
-7.30862737e-01 -3.17054272e-01 -1.20537543e+00 -7.05153704e-01
6.85277432e-02 -9.06114340e-01 5.80383316e-02 6.44070268e-01
4.72966991e-02 1.05520737e+00 -2.25913033e-01 4.30520922e-01
5.02790749e-01 -3.64417195e-01 5.79872787e-01 -6.35837674e-01
-2.30719730e-01 -4.67027068e-01 -1.59709156e-02 -1.48461664e+00
4.17187065e-01 -9.00646448e-01 3.87634069e-01 -1.75345349e+00
-1.92878693e-01 1.47002593e-01 -2.84464866e-01 5.56341887e-01
-3.02882612e-01 -2.14889437e-01 3.06426495e-01 -3.30017656e-01
-8.08048904e-01 8.84159863e-01 6.97597325e-01 -1.06095828e-01
-2.06341445e-01 -9.37086437e-03 -5.94894409e-01 6.54968798e-01
8.44861150e-01 -1.02613769e-01 -2.99480170e-01 -8.10852766e-01
-5.46961963e-01 2.08834320e-01 1.98407054e-01 -7.89672315e-01
2.27128953e-01 4.22030360e-01 -3.28851581e-01 -7.11276829e-01
4.52293396e-01 -3.25672209e-01 -5.77750444e-01 2.27028891e-01
-4.64814693e-01 -2.84545332e-01 5.01364768e-01 9.06149074e-02
-7.72193134e-01 -2.15268835e-01 8.55985105e-01 -1.67857319e-01
-1.28833127e+00 -1.10859469e-01 -3.85022521e-01 2.52945900e-01
4.41403180e-01 2.95812726e-01 -5.19154072e-01 -8.17918122e-01
-9.22161043e-01 9.54803169e-01 -4.64896619e-01 9.44387138e-01
7.12745547e-01 -6.04260802e-01 -1.18554378e+00 1.25072792e-01
2.16954127e-01 -2.88184248e-02 4.42561626e-01 5.86934984e-01
-3.40391874e-01 7.02146232e-01 2.49009326e-01 -4.94350016e-01
-1.24339879e+00 -2.36939311e-01 4.71526265e-01 -9.68565270e-02
-3.67296666e-01 1.42437637e+00 2.64528036e-01 -1.03519356e+00
6.18003607e-01 -4.80949730e-01 -1.09883845e-01 -8.14349279e-02
1.06438732e+00 9.11022723e-02 2.66916901e-01 -3.25195462e-01
-7.90525019e-01 2.35545218e-01 4.77990657e-02 -5.74513793e-01
1.14458919e+00 -4.08377856e-01 1.41567960e-01 4.60828185e-01
1.30329955e+00 -4.98719037e-01 -8.85350049e-01 -4.49063450e-01
5.57713687e-01 3.87533695e-01 3.13593090e-01 -1.37547219e+00
-5.97754121e-01 1.32899785e+00 5.01423240e-01 -1.08347107e-02
9.61820602e-01 3.71600926e-01 1.22609365e+00 5.30577183e-01
3.62204194e-01 -1.13715768e+00 -1.25084430e-01 1.30657005e+00
1.22735655e+00 -1.54183352e+00 -8.72668207e-01 -5.26830852e-01
-9.17528987e-01 9.06040370e-01 5.34060419e-01 -1.05253518e-01
5.88494062e-01 4.81688619e-01 5.87050676e-01 -2.48122737e-02
-9.37620401e-01 -4.97742265e-01 3.65481913e-01 1.77983016e-01
5.38761437e-01 2.04735816e-01 -1.93125278e-01 1.12951446e+00
-2.24955723e-01 -1.24448515e-01 3.68452579e-01 5.40676475e-01
-9.44865584e-01 -1.07765317e+00 -2.00909480e-01 -8.42302106e-03
-4.46966678e-01 -6.78844690e-01 -7.77418554e-01 4.06838000e-01
-7.97873914e-01 1.31443310e+00 2.05242019e-02 -1.89517468e-01
6.78156734e-01 7.32038438e-01 2.13484824e-01 -9.76350963e-01
-5.20114422e-01 3.15766841e-01 7.19956577e-01 -8.90287399e-01
-1.23493105e-01 -1.54567584e-01 -1.97707903e+00 2.59459406e-01
1.46885619e-01 3.11985731e-01 9.83734250e-01 1.22644639e+00
4.20343071e-01 5.06676435e-01 4.21435177e-01 -4.76329386e-01
-6.36048615e-01 -1.45240080e+00 -3.07912320e-01 3.71051803e-02
1.22111909e-01 -8.37523490e-02 -3.65249872e-01 2.07760744e-02] | [14.031609535217285, 7.009063243865967] |
9bdb2fe2-2ddc-4776-90ce-5fbf511b1159 | grenzlinie-at-semeval-2021-task-7-detecting | null | null | https://aclanthology.org/2021.semeval-1.34 | https://aclanthology.org/2021.semeval-1.34.pdf | Grenzlinie at SemEval-2021 Task 7: Detecting and Rating Humor and Offense | This paper introduces the result of Team Grenzlinie{'}s experiment in SemEval-2021 task 7: HaHackathon: Detecting and Rating Humor and Offense. This task has two subtasks. Subtask1 includes the humor detection task, the humor rating prediction task, and the humor controversy detection task. Subtask2 is an offensive rating prediction task. Detection task is a binary classification task, and the rating prediction task is a regression task between 0 to 5. 0 means the task is not humorous or not offensive, 5 means the task is very humorous or very offensive. For all the tasks, this paper chooses RoBERTa as the pre-trained model. In classification tasks, Bi-LSTM and adversarial training are adopted. In the regression task, the Bi-LSTM is also adopted. And then we propose a new approach named compare method. Finally, our system achieves an F1-score of 95.05{\%} in the humor detection task, F1-score of 61.74{\%} in the humor controversy detection task, 0.6143 RMSE in humor rating task, 0.4761 RMSE in the offensive rating task on the test datasets. | ['Xiaobing Zhou', 'Renyuan Liu'] | 2021-08-01 | null | null | null | semeval-2021 | ['humor-detection'] | ['natural-language-processing'] | [-3.55941772e-01 1.03452660e-01 2.66173501e-02 4.59388457e-02
-3.10842633e-01 -1.91636056e-01 6.27564430e-01 -2.69278109e-01
-2.72003472e-01 9.21331048e-01 6.64551854e-01 -2.27124602e-01
3.42594683e-01 -6.32040858e-01 -2.20267117e-01 -4.85031098e-01
3.87642950e-01 1.94172725e-01 -3.53749208e-02 -7.02928901e-01
6.88024640e-01 -6.63737431e-02 -8.98310900e-01 8.35464060e-01
7.34239995e-01 1.21013093e+00 -2.35155210e-01 1.01361644e+00
4.75563556e-01 2.31840920e+00 -1.18325007e+00 -5.53808689e-01
1.45457923e-01 -7.23684013e-01 -8.70623648e-01 -3.79039168e-01
1.29418865e-01 -3.84490728e-01 -4.73392218e-01 1.24515736e+00
7.51393020e-01 2.78714478e-01 8.84954751e-01 -1.25458372e+00
-8.95413339e-01 8.28812182e-01 -5.81451535e-01 4.90474068e-02
3.91766697e-01 2.52327383e-01 9.62687731e-01 -1.05350065e+00
2.12536916e-01 1.15942955e+00 7.48236299e-01 6.46823883e-01
-6.94964945e-01 -8.94879997e-01 -5.59804678e-01 6.62103117e-01
-1.03481209e+00 -1.24186397e-01 9.13820684e-01 -1.05585766e+00
8.40472877e-01 1.62888065e-01 3.57726127e-01 1.24690402e+00
4.46812719e-01 7.32943475e-01 1.38809216e+00 9.81021076e-02
1.43722743e-01 4.37083960e-01 2.44486704e-01 3.84442508e-01
-2.64114380e-01 2.22444311e-01 -4.31388348e-01 -8.70000292e-03
6.00302100e-01 -1.13740660e-01 -2.70145297e-01 8.82607818e-01
-9.24808264e-01 1.14094985e+00 7.70708323e-01 3.67454708e-01
-3.56327891e-01 7.93249309e-02 8.85176063e-01 8.12482953e-01
3.99706870e-01 5.79576492e-01 9.27792415e-02 -4.98962472e-04
-9.74585950e-01 4.92344499e-01 9.53889012e-01 6.07282162e-01
1.10494368e-01 5.71500123e-01 -5.31411588e-01 1.13064134e+00
-1.05344921e-01 4.44464862e-01 8.56586516e-01 -1.07265043e+00
4.27776664e-01 6.95314586e-01 3.24803412e-01 -1.14178193e+00
-6.15187168e-01 -6.18406177e-01 -1.19359469e+00 7.93898284e-01
1.81350514e-01 -3.35638762e-01 -6.55862451e-01 1.53333533e+00
-4.57888454e-01 -3.16836745e-01 1.98922083e-01 1.29420853e+00
1.63371360e+00 8.28890979e-01 8.70678667e-03 -5.45331120e-01
1.15705860e+00 -1.26778531e+00 -8.22810948e-01 -1.05552211e-01
3.72487336e-01 -9.90949631e-01 1.24438250e+00 7.06170619e-01
-1.30256426e+00 -7.53552973e-01 -1.26015842e+00 -1.38761848e-01
-1.99128509e-01 6.28102198e-02 9.28612251e-04 1.46623299e-01
-6.50893629e-01 5.91482759e-01 2.25765184e-01 1.11175850e-02
2.22514063e-01 1.45159051e-01 -2.03130513e-01 3.73690635e-01
-1.80606580e+00 1.53485751e+00 4.02029753e-01 -1.72715157e-01
-1.13350713e+00 -2.90971011e-01 -5.20517647e-01 1.26525968e-01
-5.74528053e-02 -4.08385694e-01 1.46567452e+00 -1.24314392e+00
-1.24901092e+00 1.11629164e+00 3.70319366e-01 -5.50191641e-01
8.08258176e-01 -2.35653371e-01 -5.08748889e-01 -2.25125894e-01
2.25846142e-01 2.29273364e-02 6.75879955e-01 -1.15527320e+00
-4.05835241e-01 -6.70945868e-02 9.76118743e-02 4.06428754e-01
-2.47058913e-01 4.16364729e-01 4.99954969e-01 -6.57057464e-01
-4.68941718e-01 -6.39703393e-01 1.74771965e-01 -7.04165399e-01
-5.30975282e-01 -2.96028703e-01 6.32864952e-01 -1.11716926e+00
1.79536545e+00 -1.94858563e+00 1.03310414e-01 -2.26218417e-01
4.60814595e-01 4.07978892e-01 1.83881193e-01 3.25553209e-01
-2.71715105e-01 4.85886373e-02 -1.39993548e-01 1.07214905e-01
1.66957840e-01 -4.81148720e-01 -5.27444899e-01 2.56395400e-01
-3.67563426e-01 9.63182449e-01 -7.18178689e-01 -4.11419779e-01
1.58154611e-02 1.28053287e-02 -2.30020657e-01 6.42026007e-01
1.89919472e-01 1.43043837e-02 -6.14161044e-02 5.77000976e-01
3.08517188e-01 -1.35152131e-01 -3.49064112e-01 1.97149385e-02
-1.58883214e-01 2.72045642e-01 -6.28775954e-01 7.20644176e-01
-5.08989692e-01 7.84940422e-01 -1.48217648e-01 -5.63693345e-01
1.51176012e+00 4.43235785e-01 5.78534305e-02 -7.09090531e-01
6.97761893e-01 2.75795311e-01 2.81135231e-01 -7.02387631e-01
6.78745031e-01 -1.00866354e+00 -4.42374498e-01 5.29858470e-01
-3.47616404e-01 -3.93305749e-01 2.67894119e-02 2.09258690e-01
1.13182485e+00 -3.33918005e-01 5.39915621e-01 -2.67271072e-01
7.81327963e-01 8.13142397e-03 5.64589024e-01 5.58584690e-01
-4.97866541e-01 6.42648995e-01 8.53044569e-01 -6.38215363e-01
-1.20065737e+00 -6.30903840e-01 1.40925914e-01 1.33597720e+00
-1.11089863e-01 -4.53656353e-02 -7.22103477e-01 -6.23427868e-01
-1.12649418e-01 1.31459165e+00 -9.27294850e-01 -5.37792385e-01
-4.66218203e-01 -6.91749454e-01 5.79182982e-01 6.08172894e-01
9.71829951e-01 -1.73525155e+00 -5.21879017e-01 3.79395224e-02
-4.68344927e-01 -8.05676103e-01 -3.92989606e-01 1.80973515e-01
-3.99302602e-01 -1.08564925e+00 -6.12985909e-01 -8.26222003e-01
-2.71465015e-02 6.90028295e-02 1.23736119e+00 2.10533753e-01
2.38221243e-01 -4.23416138e-01 -4.32030618e-01 -2.54353970e-01
-5.77988565e-01 -3.84474307e-01 8.80912840e-02 -4.32181120e-01
5.20825624e-01 -5.51833689e-01 -4.39063311e-01 5.64438522e-01
-2.53066778e-01 6.99486434e-02 1.76333427e-01 1.19640839e+00
-5.59035651e-02 -1.58472449e-01 9.27777648e-01 -8.45350504e-01
1.20007920e+00 -6.01773202e-01 1.00426763e-01 -2.58903623e-01
-5.24426281e-01 -6.87296152e-01 1.36999977e+00 -5.56678593e-01
-7.94265509e-01 -5.71444869e-01 -1.19761102e-01 -3.90395164e-01
3.03513259e-01 4.49580491e-01 7.01128468e-02 3.57436001e-01
1.38559437e+00 -1.12210698e-02 -3.55451733e-01 -1.42421126e-01
-5.02848588e-02 1.16489077e+00 9.44871008e-01 -1.38804182e-01
7.79690146e-01 -1.04677685e-01 -2.57425994e-01 -3.77436846e-01
-1.31244338e+00 -2.62995064e-01 -2.87844360e-01 -5.43219686e-01
9.26371336e-01 -1.04522872e+00 -9.28933442e-01 6.29265964e-01
-1.43574560e+00 -4.54854459e-01 -1.52742445e-01 3.05536091e-01
-7.26322532e-01 -1.21190969e-03 -9.90996182e-01 -9.64153469e-01
-1.06523800e+00 -7.94373453e-01 1.94529548e-01 4.84031141e-01
-6.04307115e-01 -7.43591905e-01 1.65954083e-01 8.91509116e-01
3.05551171e-01 4.85147804e-01 8.71554494e-01 -1.06892896e+00
4.58611161e-01 -4.63185370e-01 -4.18188542e-01 8.65036070e-01
-3.23014021e-01 -3.47647369e-01 -1.07378495e+00 -1.18880816e-01
5.39947033e-01 -1.08805096e+00 9.74223673e-01 9.59144682e-02
9.33216035e-01 -6.48000956e-01 5.68172991e-01 3.21149826e-01
9.00473833e-01 4.46673214e-01 1.01909816e+00 5.59164703e-01
7.08989382e-01 4.41771984e-01 7.63527036e-01 5.91649234e-01
3.51295054e-01 4.94325697e-01 6.25099123e-01 -1.14990771e-01
-1.34940207e-01 -4.15599704e-01 9.03343022e-01 8.96694183e-01
-3.68939161e-01 -3.16768251e-02 -7.96484232e-01 2.02938333e-01
-1.93313110e+00 -1.60483730e+00 -7.65734017e-01 2.12785053e+00
9.04911399e-01 2.73640215e-01 4.53583837e-01 4.17032659e-01
9.78352010e-01 3.80957723e-01 -4.70411360e-01 -9.85679150e-01
-3.10758978e-01 -2.19432548e-01 -6.25282824e-02 8.75662565e-01
-9.76952851e-01 1.25523818e+00 5.36774683e+00 7.43053436e-01
-8.82157981e-01 3.94740909e-01 7.21873343e-01 -3.00264537e-01
-4.19231020e-02 -3.70311916e-01 -4.11145002e-01 7.06034720e-01
8.02360415e-01 -3.17478955e-01 4.79993105e-01 1.14297032e+00
4.89812106e-01 -2.75293458e-02 -8.34694386e-01 1.03088200e+00
6.75557792e-01 -6.65828288e-01 -1.89746588e-01 -3.46662700e-01
8.19893479e-01 -3.24818283e-01 3.03840600e-02 1.10669768e+00
5.17127991e-01 -1.63357806e+00 7.56006718e-01 5.79653621e-01
7.73478448e-01 -8.81089330e-01 1.16213584e+00 5.77614903e-01
-6.45475030e-01 -4.89613742e-01 -7.61359513e-01 -6.80187225e-01
8.65220800e-02 5.68719804e-01 -7.11231053e-01 -5.84321022e-02
5.83801150e-01 6.30073547e-01 -3.79054844e-01 8.46327841e-01
-9.38340783e-01 6.32059693e-01 3.75203311e-01 -2.07500413e-01
9.88729671e-02 1.20964356e-01 5.71921706e-01 1.47613716e+00
1.15308292e-01 2.59879798e-01 1.48070663e-01 9.13868010e-01
-4.31194752e-01 1.44910574e-01 -3.88964862e-01 3.81173879e-01
3.65544379e-01 1.43365967e+00 -4.66517871e-03 -5.64000487e-01
1.84850588e-01 8.90445411e-01 2.38250464e-01 7.32935071e-02
-1.18851924e+00 -6.40672266e-01 -1.03436619e-01 9.49708894e-02
-2.94549972e-01 2.32044846e-01 -1.01190555e+00 -9.01118934e-01
-4.11522895e-01 -1.27786732e+00 5.13909280e-01 -1.26208723e+00
-1.33749163e+00 7.70327508e-01 -3.72235149e-01 -1.08461976e+00
-1.02590367e-01 -6.36179924e-01 -1.42454350e+00 9.30471897e-01
-1.04735863e+00 -1.19650841e+00 -7.08866954e-01 6.31151795e-01
7.69560039e-01 -6.49369061e-01 1.91312239e-01 1.24734990e-01
-4.35029060e-01 5.12705624e-01 -2.08554372e-01 2.98884809e-01
8.63030970e-01 -1.33829045e+00 -2.06143081e-01 3.64013135e-01
-7.65707731e-01 2.47971684e-01 9.43227530e-01 -6.66776359e-01
-4.19878811e-01 -9.40614879e-01 1.22513056e+00 -4.40636665e-01
8.72441173e-01 2.96637937e-02 -1.00947392e+00 3.40727121e-01
2.84248739e-01 -5.14089763e-01 5.24317920e-01 -3.19888823e-05
-5.53157687e-01 1.88506752e-01 -1.19718289e+00 4.13563907e-01
4.01507348e-01 -5.88798940e-01 -9.55717385e-01 4.35879916e-01
5.05106390e-01 -2.26582900e-01 -9.94503498e-01 5.08558273e-01
4.84349638e-01 -1.30475390e+00 5.71355700e-01 -7.75270462e-01
1.69324028e+00 -3.17478091e-01 -3.62548739e-01 -1.11143732e+00
-7.63978302e-01 -3.57763618e-01 -1.66637927e-01 8.87160480e-01
3.29505533e-01 -6.26141280e-02 5.43284476e-01 1.78333059e-01
-3.44445527e-01 -6.96771204e-01 -6.99536920e-01 -4.98168409e-01
5.69229066e-01 -7.43606836e-02 8.07284191e-02 1.36316454e+00
5.20414472e-01 1.03747535e+00 -1.07301569e+00 -5.72133005e-01
3.63664687e-01 1.14394657e-01 7.74498045e-01 -1.01082969e+00
-4.31000292e-01 -8.56967330e-01 -8.11258852e-02 -6.61084652e-01
2.22992405e-01 -9.53418314e-01 5.02520502e-02 -1.58873785e+00
7.91030228e-01 4.79711980e-01 -5.43847978e-01 6.99484110e-01
-2.74998993e-01 4.53453153e-01 5.35556376e-01 4.26715374e-01
-4.72941816e-01 7.20665038e-01 1.23305237e+00 -2.59150475e-01
-2.34269962e-01 8.15501884e-02 -8.26727509e-01 9.51240242e-01
1.04893672e+00 -2.06655964e-01 -4.23369296e-02 1.97439417e-01
3.33698153e-01 4.58478332e-01 5.24366975e-01 -9.33465421e-01
9.12961811e-02 -3.85350496e-01 5.39459288e-01 -5.49260139e-01
2.61392236e-01 -2.30596766e-01 -2.41701290e-01 8.40461075e-01
-5.03430843e-01 -5.44693507e-02 -1.78970933e-01 -1.65602341e-01
-2.46623188e-01 -6.13626003e-01 1.40354598e+00 -2.91539818e-01
-3.05302083e-01 -3.04863513e-01 -3.36199790e-01 2.76234478e-01
7.85047889e-01 -1.43428668e-01 -1.03520060e+00 -9.09192443e-01
-7.55641222e-01 3.00026953e-01 3.19120020e-01 3.43163133e-01
7.73022771e-01 -1.45621407e+00 -1.27171874e+00 -5.53025544e-01
-1.88923046e-01 -4.76144850e-01 1.75391585e-01 9.64720845e-01
-3.34635586e-01 -1.35369867e-01 -5.19275963e-01 1.26170188e-01
-1.30639637e+00 4.94945467e-01 5.45253098e-01 -5.03852069e-01
-2.25775868e-01 7.03453004e-01 2.46597052e-01 -3.03388923e-01
8.85841921e-02 6.32799864e-01 -8.48829508e-01 3.10038716e-01
6.57214522e-01 1.02220547e+00 -3.39189082e-01 -8.68804753e-01
-1.47616416e-02 2.13465318e-01 -8.05873573e-02 -3.76167893e-02
1.02500749e+00 4.04236138e-01 -4.12961572e-01 7.23838329e-01
7.57084966e-01 -9.77564752e-02 -7.01972306e-01 1.78440288e-01
-2.25986123e-01 -6.22744039e-02 1.55138820e-01 -1.70532167e+00
-7.66077161e-01 1.15259194e+00 2.31416613e-01 3.69362801e-01
9.85205352e-01 -4.57119733e-01 7.64879882e-01 3.68814588e-01
3.80957709e-03 -1.48002410e+00 6.14381731e-01 1.29257858e+00
1.66646802e+00 -1.10969520e+00 2.96798170e-01 3.16595495e-01
-1.35665786e+00 1.07254827e+00 1.07795906e+00 -4.10160840e-01
3.70498262e-02 -7.47344643e-02 1.35086209e-01 -2.14416951e-01
-9.28870320e-01 1.44576594e-01 3.00153494e-01 1.45786941e-01
7.51109779e-01 1.25789568e-01 -8.70395124e-01 1.37204373e+00
-8.79519641e-01 -3.79286222e-02 9.16142404e-01 8.30344260e-02
-1.06388330e+00 -1.33450910e-01 -5.06322980e-01 3.87280196e-01
-5.52668214e-01 -2.63506889e-01 -1.04298663e+00 6.65113866e-01
1.98510990e-01 1.23821139e+00 -4.92326736e-01 -1.34837401e+00
5.26417673e-01 1.86264440e-02 -1.88220978e-01 -3.57701927e-01
-1.48356879e+00 -1.16555884e-01 3.93219799e-01 -1.87816396e-01
2.23858073e-01 -4.30429913e-02 -1.26271713e+00 -8.73755395e-01
-2.23605484e-01 3.17614377e-01 1.96060970e-01 7.48351216e-01
-4.71686006e-01 4.84566182e-01 1.10851693e+00 -5.74334800e-01
-9.25626278e-01 -1.47628427e+00 -5.47715425e-01 6.72942042e-01
-9.68188699e-03 -5.10056078e-01 -7.42284656e-01 -1.73288733e-01] | [8.867189407348633, 11.078749656677246] |
b0af9c95-bca5-47b9-9a87-c9253d45535b | semantic-and-effective-communication-for | 2301.05901 | null | https://arxiv.org/abs/2301.05901v1 | https://arxiv.org/pdf/2301.05901v1.pdf | Semantic and Effective Communication for Remote Control Tasks with Dynamic Feature Compression | The coordination of robotic swarms and the remote wireless control of industrial systems are among the major use cases for 5G and beyond systems: in these cases, the massive amounts of sensory information that needs to be shared over the wireless medium can overload even high-capacity connections. Consequently, solving the effective communication problem by optimizing the transmission strategy to discard irrelevant information can provide a significant advantage, but is often a very complex task. In this work, we consider a prototypal system in which an observer must communicate its sensory data to an actor controlling a task (e.g., a mobile robot in a factory). We then model it as a remote Partially Observable Markov Decision Process (POMDP), considering the effect of adopting semantic and effective communication-oriented solutions on the overall system performance. We split the communication problem by considering an ensemble Vector Quantized Variational Autoencoder (VQ-VAE) encoding, and train a Deep Reinforcement Learning (DRL) agent to dynamically adapt the quantization level, considering both the current state of the environment and the memory of past messages. We tested the proposed approach on the well-known CartPole reference control problem, obtaining a significant performance increase over traditional approaches | ['Michele Zorzi', 'Andrea Zanella', 'Federico Chiariotti', 'Francesco Pase', 'Pietro Talli'] | 2023-01-14 | null | null | null | null | ['feature-compression'] | ['computer-vision'] | [ 1.18906744e-01 4.85346764e-01 -2.59496160e-02 -2.63589267e-02
-3.39558989e-01 -1.71512663e-01 4.13566947e-01 2.66568124e-01
-4.16932911e-01 9.58303034e-01 -2.81108558e-01 -2.95826942e-01
-5.09094536e-01 -9.51939762e-01 -6.39745176e-01 -1.19778919e+00
-2.79864609e-01 6.55541718e-01 8.13898444e-02 -3.67618769e-01
-4.04263958e-02 3.45810771e-01 -1.33777058e+00 -2.59805560e-01
5.05338550e-01 1.55607200e+00 8.12071204e-01 7.72787929e-01
9.89385396e-02 1.01340854e+00 -1.05393183e+00 -1.11455075e-03
3.39891575e-02 -3.62958938e-01 -8.45403790e-01 4.43089247e-01
-7.91486740e-01 -2.56621450e-01 -4.83459495e-02 1.03068531e+00
6.24806702e-01 3.65342379e-01 4.57117826e-01 -1.31208646e+00
-1.38038814e-01 7.66116619e-01 -2.46252585e-02 6.84611276e-02
-2.32851994e-03 8.44728127e-02 7.39629805e-01 -6.09325767e-02
6.33588493e-01 1.26080751e+00 -1.05697989e-01 3.74101728e-01
-9.99383450e-01 -2.88452119e-01 3.32365602e-01 6.22845232e-01
-1.10893488e+00 -3.84220183e-01 6.54765248e-01 -2.59205312e-01
6.74910665e-01 2.58479547e-02 8.49698544e-01 1.13504648e+00
8.14487398e-01 3.30982983e-01 6.88709199e-01 -2.42037281e-01
9.19503808e-01 1.60128936e-01 -3.33384693e-01 4.63867337e-01
-1.13093808e-01 3.67787927e-02 -2.29033872e-01 -1.54245198e-01
5.49959421e-01 -8.77112523e-02 -2.04242870e-01 -4.70685840e-01
-1.25108123e+00 7.82626629e-01 5.06251752e-01 3.30170155e-01
-1.04176319e+00 3.04802984e-01 7.68606663e-02 7.62914896e-01
1.75288573e-01 4.49634373e-01 -3.82619739e-01 -2.13220641e-01
-4.10538852e-01 -1.38047531e-01 1.08811665e+00 8.45812380e-01
5.65607488e-01 2.85816103e-01 -1.58907279e-01 2.68804640e-01
4.35522676e-01 5.08944213e-01 1.65543705e-01 -1.48787022e+00
2.54298210e-01 -2.48938985e-02 4.92101312e-01 -8.47679019e-01
-4.97020066e-01 -6.61025107e-01 -1.13441980e+00 3.80353630e-01
5.73848821e-02 -7.83945739e-01 -3.13424349e-01 1.61736429e+00
5.72962105e-01 1.40254498e-01 4.28052783e-01 1.00723803e+00
8.72737840e-02 9.70056176e-01 -7.69945607e-02 -7.44754195e-01
1.02685511e+00 -6.89343870e-01 -1.10122669e+00 1.64330006e-01
2.73760259e-01 -5.27668953e-01 2.07380608e-01 6.80388749e-01
-9.39154148e-01 -4.75913584e-01 -1.29680657e+00 4.04674888e-01
-2.00442895e-01 -1.02342786e-02 2.31166989e-01 2.67166704e-01
-9.86959040e-01 1.00078201e+00 -9.31115091e-01 -4.31950271e-01
2.17452529e-03 6.85109556e-01 1.53287157e-01 1.20040372e-01
-1.16930318e+00 8.72804523e-01 3.82056445e-01 3.18386704e-01
-1.28607786e+00 4.23438884e-02 -2.88633436e-01 4.65377361e-01
9.84440684e-01 -8.37525070e-01 1.19023621e+00 -9.09154117e-01
-2.12853599e+00 -1.47884011e-01 2.42418841e-01 -6.05097353e-01
3.12158138e-01 3.18219632e-01 -3.24447542e-01 2.09210604e-01
-2.18468472e-01 2.62183696e-01 1.27270234e+00 -1.05476475e+00
-7.98327386e-01 -3.45705509e-01 4.59219337e-01 3.78471762e-01
-1.96741626e-01 -5.35140216e-01 -6.06864430e-02 -2.44452089e-01
4.41776123e-03 -1.19518483e+00 -5.18824339e-01 -2.60518730e-01
-3.35471183e-01 -4.08304065e-01 9.22884703e-01 -5.14653981e-01
7.89568841e-01 -2.00744104e+00 8.03075552e-01 1.98886812e-01
-1.82625633e-02 -1.41217113e-01 5.14175147e-02 5.01826823e-01
5.09937763e-01 -8.91933441e-02 8.59695971e-02 -3.71894509e-01
4.50446531e-02 4.81960893e-01 1.36918863e-02 4.05894637e-01
-1.08333647e-01 2.28561372e-01 -9.42231715e-01 -4.17594016e-01
4.15924728e-01 3.71955991e-01 -3.44030321e-01 1.27129436e-01
-5.29289544e-01 1.02768052e+00 -1.02971888e+00 1.91793397e-01
2.01285809e-01 -4.06374723e-01 4.85553205e-01 -1.67032413e-03
-4.17311117e-02 -2.48615444e-01 -1.33056509e+00 1.70133364e+00
-8.50771189e-01 4.03107107e-01 7.77716100e-01 -1.33331895e+00
5.96043885e-01 6.05675459e-01 9.89944518e-01 -4.53545243e-01
5.19695282e-01 -4.63491119e-02 -6.04673661e-03 -4.39181149e-01
4.17096913e-01 1.74201056e-01 -1.85089603e-01 2.04033092e-01
2.72993207e-01 -1.80655435e-01 -1.00204416e-01 8.75660926e-02
1.38669872e+00 -2.08179310e-01 2.76558936e-01 -1.42617330e-01
5.96512496e-01 -9.97243226e-02 5.93254924e-01 9.32030141e-01
-3.16910982e-01 -1.84146971e-01 4.17011976e-01 -1.99203506e-01
-7.33285189e-01 -8.08831215e-01 2.89306432e-01 7.98497617e-01
6.83013320e-01 -1.36442646e-01 -6.09527588e-01 -3.89928013e-01
-3.00143272e-01 7.92684376e-01 -1.79086789e-01 -3.06401849e-01
-1.69810280e-01 -6.49054408e-01 -3.14479321e-01 -2.55653143e-01
5.21120727e-01 -1.03074062e+00 -9.36103880e-01 7.69306719e-01
-2.62934387e-01 -1.21852100e+00 2.37094939e-01 4.26577389e-01
-4.41610605e-01 -7.87085533e-01 -4.74853396e-01 -3.97611141e-01
1.81519374e-01 1.22185564e-02 7.24309981e-01 -3.93351242e-02
-2.20351480e-03 6.48223102e-01 -4.17905509e-01 -4.74664986e-01
-6.62751436e-01 2.18821853e-01 1.44505039e-01 3.66043180e-01
-4.60222542e-01 -4.87134963e-01 -4.40374821e-01 1.67095736e-01
-6.03454113e-01 -2.73635387e-01 5.56989312e-01 7.98671305e-01
5.56271851e-01 8.04314911e-01 5.90577304e-01 -3.00106198e-01
7.48352706e-01 -6.43921494e-01 -7.72088230e-01 1.47619009e-01
-3.27756912e-01 1.89806148e-01 9.54653502e-01 -3.57162863e-01
-1.12965596e+00 5.70821762e-02 -4.46489938e-02 -3.88142318e-01
1.89735796e-02 2.17671230e-01 -2.57149339e-01 -6.06967211e-02
-2.73406319e-02 8.92525464e-02 -1.36175796e-01 -4.49106432e-02
2.52954572e-01 8.38397801e-01 1.52780175e-01 -4.64338452e-01
4.29281384e-01 4.91912633e-01 4.77493286e-01 -1.20901060e+00
-5.21809340e-01 -9.48221982e-02 -3.82214606e-01 -5.45386314e-01
1.13883197e+00 -6.70882404e-01 -1.37434995e+00 1.41269743e-01
-1.39323962e+00 -3.01167697e-01 -4.69413310e-01 7.33489573e-01
-8.13101768e-01 8.09128508e-02 -4.76597428e-01 -8.58402073e-01
1.23321220e-01 -1.49228120e+00 9.21962261e-01 2.11870130e-02
2.13462800e-01 -7.89876759e-01 -1.97495982e-01 1.04367957e-01
5.54431856e-01 2.44495139e-01 8.56930673e-01 -2.37431839e-01
-8.90808761e-01 1.14285886e-01 4.20803994e-01 2.97996163e-01
-1.80862606e-01 -3.61723542e-01 -6.57730043e-01 -4.93679643e-01
3.22730660e-01 -1.05722949e-01 2.36974418e-01 4.30323720e-01
1.07876647e+00 -3.48554552e-01 -4.06652719e-01 1.02679893e-01
1.23498285e+00 4.71153975e-01 4.28246334e-02 -7.56328553e-02
2.58438200e-01 6.12892330e-01 7.67028451e-01 1.00806057e+00
3.57173741e-01 9.28948104e-01 1.09986711e+00 3.31370145e-01
3.09410810e-01 2.60950446e-01 3.70857716e-01 9.11211908e-01
-1.71760023e-01 -7.27394938e-01 -2.93928623e-01 8.61922503e-02
-2.09084249e+00 -8.24888229e-01 3.42876673e-01 2.06528258e+00
8.33864957e-02 -2.00452916e-02 -3.39754283e-01 8.44868571e-02
8.05200219e-01 6.18655942e-02 -7.17680871e-01 -4.70373929e-01
2.78210551e-01 -2.92663574e-01 5.48543155e-01 4.54667151e-01
-1.06340992e+00 5.34717858e-01 4.59198046e+00 7.65069008e-01
-1.21323061e+00 4.38011736e-01 3.22253585e-01 -1.61334231e-01
3.44626248e-01 -1.97856337e-01 -3.85124445e-01 5.92911780e-01
1.36031115e+00 8.53840262e-03 1.01201153e+00 7.25023270e-01
4.47374314e-01 -3.37760597e-01 -1.11660159e+00 1.06796741e+00
-4.37734187e-01 -1.08402479e+00 -3.07783693e-01 3.61443460e-01
5.43749869e-01 9.78789851e-02 2.89969612e-03 3.26384604e-01
5.44778049e-01 -4.66712534e-01 9.50958908e-01 6.98091507e-01
3.82894985e-02 -8.10779810e-01 9.74102080e-01 8.71126890e-01
-9.85181332e-01 -7.33063102e-01 -4.00467455e-01 -1.99567407e-01
4.60961223e-01 5.90781152e-01 -5.22202909e-01 9.29528236e-01
7.81791866e-01 5.00005305e-01 3.89034525e-02 7.09935188e-01
-4.87616472e-03 2.66537637e-01 -4.65906709e-01 -4.88859206e-01
2.54954427e-01 -3.25007826e-01 9.80611861e-01 2.99543202e-01
6.41285598e-01 2.27948919e-01 4.13383424e-01 6.19098485e-01
-1.17461327e-02 -1.30625069e-01 -5.77909350e-01 1.62082866e-01
5.15309930e-01 1.09932530e+00 -8.42565000e-01 -4.58896101e-01
-2.10105002e-01 1.13930893e+00 9.02789906e-02 4.37759876e-01
-7.51136541e-01 -2.86558598e-01 7.36685872e-01 -4.54074711e-01
5.53289354e-01 -2.14122295e-01 3.98154259e-01 -9.56231296e-01
-4.86076735e-02 -3.75856519e-01 9.17740315e-02 -6.84768021e-01
-9.10812974e-01 5.83122313e-01 -3.29550117e-01 -1.23123205e+00
-5.17113984e-01 -1.02057673e-01 -8.88929293e-02 3.63559753e-01
-1.29103005e+00 -4.69629586e-01 -1.54934779e-01 6.41796768e-01
5.29957116e-01 -1.48294747e-01 1.01443601e+00 2.48825669e-01
-5.77419937e-01 -3.29368353e-01 4.99485761e-01 -4.71702993e-01
1.82300791e-01 -1.06876493e+00 -4.61833209e-01 4.85470593e-01
-1.00498714e-01 -4.62348014e-02 8.64825726e-01 -2.81110704e-01
-1.79032302e+00 -1.01147747e+00 2.53681183e-01 5.30149676e-02
6.59606040e-01 -2.70290345e-01 -2.76039243e-01 5.14017463e-01
4.38043624e-01 -2.15403605e-02 -5.24762981e-02 -3.15576315e-01
8.05944920e-01 -3.31784695e-01 -1.20893896e+00 4.68214154e-01
6.05487227e-01 -8.15212429e-02 -1.05117328e-01 4.05405253e-01
8.42645288e-01 -2.61462122e-01 -1.13719821e+00 7.56775364e-02
1.92354947e-01 -6.65985048e-01 6.69104218e-01 -2.59165645e-01
-4.64810692e-02 -2.77202576e-01 -2.61181593e-01 -1.77468252e+00
-2.21114844e-01 -9.45284843e-01 -2.29954958e-01 9.10215378e-01
1.27313524e-01 -5.96710682e-01 5.94933152e-01 1.42797858e-01
-2.15312913e-02 -5.27339339e-01 -1.53083158e+00 -4.67882633e-01
-3.94243628e-01 -3.18125963e-01 6.96151853e-01 5.89614093e-01
-1.95589028e-02 4.79304224e-01 -4.05851603e-01 6.69979572e-01
8.10301602e-01 -4.67901118e-02 5.38155913e-01 -1.26579046e+00
-5.59044302e-01 2.72419304e-02 -4.26546782e-01 -1.14364135e+00
2.27913171e-01 -6.39712572e-01 2.42527202e-01 -1.39262080e+00
-5.35483897e-01 -4.63442236e-01 -3.25281262e-01 -1.27518490e-01
5.89922667e-01 -4.00473982e-01 5.28575301e-01 5.64231202e-02
-1.02672470e+00 8.69988799e-01 1.35881329e+00 -4.93471056e-01
-1.80304945e-01 4.84661341e-01 -1.76990405e-01 5.82113624e-01
7.49967337e-01 -4.89686906e-01 -5.39670229e-01 -4.44509655e-01
1.73186615e-01 1.13701963e+00 3.33369583e-01 -1.24886358e+00
6.70147359e-01 -1.31127313e-01 3.73970866e-02 -2.75063723e-01
8.20289433e-01 -1.56604600e+00 1.50226206e-01 6.37372911e-01
-2.85167545e-01 -6.92821667e-02 -3.93053859e-01 1.11522353e+00
-1.65170103e-01 -1.44464925e-01 5.63833952e-01 -1.29921883e-01
-7.89886594e-01 1.56686932e-01 -1.03511024e+00 -5.05135179e-01
1.25770414e+00 4.15418208e-01 3.23832110e-02 -6.15871966e-01
-1.33582485e+00 4.80865359e-01 7.29319975e-02 4.38184083e-01
3.18868846e-01 -8.70179057e-01 -2.28803650e-01 1.04127703e-02
-3.14029098e-01 -7.13340789e-02 3.52248490e-01 8.20326030e-01
-1.06582381e-01 2.77133793e-01 -1.99605078e-01 -7.75463402e-01
-8.33883762e-01 6.68595433e-01 4.47872937e-01 -2.68558204e-01
-3.64050984e-01 2.55892277e-01 -3.11684221e-01 -1.91390797e-01
4.69637603e-01 -4.57751006e-01 -4.33083206e-01 5.02668060e-02
6.29235730e-02 7.39201963e-01 8.39832276e-02 -3.65422338e-01
-1.39236711e-02 2.31931925e-01 5.66859066e-01 -1.66382417e-01
1.24416995e+00 -7.32059717e-01 -7.84118101e-02 4.84802008e-01
9.24407542e-01 -5.33454955e-01 -1.06152546e+00 -1.13989338e-01
-1.37386963e-01 2.36146078e-01 5.70481122e-01 -4.20524657e-01
-1.19124413e+00 8.05680573e-01 7.99031258e-01 7.72893965e-01
9.85355854e-01 1.40242368e-01 6.22215688e-01 9.40331578e-01
1.00643468e+00 -1.38190961e+00 -6.48929849e-02 3.36012602e-01
5.05576849e-01 -1.10210872e+00 -1.66964173e-01 -3.41802925e-01
-4.69945222e-01 1.12772977e+00 1.87544227e-02 1.09371513e-01
1.00042844e+00 1.87821507e-01 -1.19915910e-01 -5.79457246e-02
-1.08606386e+00 -1.83121204e-01 -5.76635718e-01 4.72974986e-01
-4.33603942e-01 3.88255656e-01 1.02279387e-01 4.28896695e-01
-6.10140860e-02 -5.24282493e-02 7.85513520e-01 9.38078821e-01
-4.91616815e-01 -7.72933543e-01 -4.33766752e-01 2.70652890e-01
-4.51032013e-01 6.38653338e-01 2.17976943e-01 4.46524441e-01
3.04949135e-01 1.50177455e+00 1.02400042e-01 -2.61136264e-01
2.09877267e-01 -4.83249962e-01 2.82224923e-01 -4.35051620e-01
-2.77845323e-01 6.98377639e-02 1.59582822e-04 -8.36605251e-01
-6.02440298e-01 -6.93142474e-01 -1.22935271e+00 -1.73244283e-01
-2.90553868e-01 2.53854364e-01 1.10743499e+00 1.08512187e+00
4.17472601e-01 1.20624566e+00 8.77596378e-01 -1.00835609e+00
-9.63967323e-01 -7.06258714e-01 -8.18078935e-01 -8.19981620e-02
6.05010390e-01 -8.33427608e-01 -3.37128937e-01 -1.81324124e-01] | [4.502495288848877, 1.9630489349365234] |
c0494c9d-55e7-40f2-9808-412139f50699 | on-the-structural-generalization-in-text-to | 2301.04790 | null | https://arxiv.org/abs/2301.04790v2 | https://arxiv.org/pdf/2301.04790v2.pdf | On the Structural Generalization in Text-to-SQL | Exploring the generalization of a text-to-SQL parser is essential for a system to automatically adapt the real-world databases. Previous works provided investigations focusing on lexical diversity, including the influence of the synonym and perturbations in both natural language questions and databases. However, research on the structure variety of database schema~(DS) is deficient. Specifically, confronted with the same input question, the target SQL is probably represented in different ways when the DS comes to a different structure. In this work, we provide in-deep discussions about the structural generalization of text-to-SQL tasks. We observe that current datasets are too templated to study structural generalization. To collect eligible test data, we propose a framework to generate novel text-to-SQL data via automatic and synchronous (DS, SQL) pair altering. In the experiments, significant performance reduction when evaluating well-trained text-to-SQL models on the synthetic samples demonstrates the limitation of current research regarding structural generalization. According to comprehensive analysis, we suggest the practical reason is the overfitting of (NL, SQL) patterns. | ['Kai Yu', 'Hanchong Zhang', 'Zhi Chen', 'Hongshen Xu', 'Su Zhu', 'Ruisheng Cao', 'Lu Chen', 'Jieyu Li'] | 2023-01-12 | null | null | null | null | ['text-to-sql'] | ['computer-code'] | [ 2.74038196e-01 1.16659418e-01 -7.61574507e-02 -8.38394046e-01
-6.15832865e-01 -7.66153276e-01 3.27195048e-01 3.50537747e-01
-6.06606230e-02 5.68663836e-01 1.22532077e-01 -6.38405740e-01
-1.71585754e-02 -1.07751083e+00 -1.05880439e+00 -5.35643138e-02
3.60078782e-01 6.23565972e-01 4.58641022e-01 -6.19547427e-01
3.93234104e-01 5.55496275e-01 -1.92209077e+00 8.82726371e-01
7.92988122e-01 6.43101990e-01 1.91189960e-01 3.32346261e-01
-7.83058643e-01 5.18886805e-01 -8.63975465e-01 -8.15526426e-01
2.27526769e-01 -3.91000748e-01 -7.77192712e-01 -7.87888989e-02
7.10930526e-01 1.46862671e-01 -8.30863044e-02 1.18214679e+00
4.44432855e-01 -3.23719591e-01 3.83102655e-01 -9.43280220e-01
-5.87953746e-01 1.08583772e+00 -3.16061080e-02 2.19776168e-01
9.27907228e-01 1.06237009e-01 9.48994875e-01 -9.62531090e-01
9.35664117e-01 1.47435760e+00 5.48587382e-01 5.19113779e-01
-1.25070035e+00 -3.97812933e-01 6.12742640e-02 1.27368495e-01
-1.42630589e+00 -4.20765787e-01 8.14501941e-01 -1.58447191e-01
1.03177679e+00 6.68727279e-01 1.74289197e-01 1.49074578e+00
1.73751503e-01 5.68661571e-01 9.64539528e-01 -7.73187518e-01
1.10584788e-01 8.89537096e-01 1.55231461e-01 4.03734714e-01
5.93079031e-01 -1.75186068e-01 -7.21199572e-01 -2.02143967e-01
1.66312829e-01 -4.56473619e-01 -2.79789325e-02 -3.29365849e-01
-8.67476523e-01 6.52083158e-01 -1.52681112e-01 4.34442043e-01
1.27643696e-03 -5.75012743e-01 6.51806116e-01 7.13327944e-01
-8.32250789e-02 7.42762625e-01 -7.69634902e-01 -2.86176443e-01
-4.15524274e-01 5.66299200e-01 1.18551147e+00 1.64914870e+00
7.86961377e-01 -1.04625352e-01 1.24131419e-01 7.03661382e-01
6.44587651e-02 4.98589337e-01 6.76888406e-01 -6.05304480e-01
9.53407466e-01 1.18011296e+00 -1.62185565e-01 -1.04626691e+00
-4.20391232e-01 -3.41632664e-01 -5.71154594e-01 -3.70248497e-01
3.36152226e-01 -4.29710187e-03 -3.39357704e-01 1.77063429e+00
2.48572886e-01 -6.27632022e-01 4.27554131e-01 2.46994197e-01
9.02622044e-01 4.39629972e-01 -5.31377234e-02 -4.23821449e-01
1.18647659e+00 -2.62967795e-01 -7.91204035e-01 -2.07132414e-01
1.03541255e+00 -9.14341033e-01 1.81398189e+00 3.43298465e-01
-1.07277811e+00 -8.33276987e-01 -9.88649547e-01 2.13029400e-01
-8.75169814e-01 -1.11902542e-01 2.78111786e-01 1.02508485e+00
-5.22758782e-01 3.13222378e-01 -4.46334690e-01 -1.01116431e+00
-2.46090159e-01 -7.04909042e-02 -3.26430291e-01 1.50537631e-02
-1.27458334e+00 9.03913081e-01 7.45476842e-01 -2.54173219e-01
-7.18788803e-02 -7.90503323e-01 -7.46295273e-01 -8.32949951e-02
7.16871440e-01 -7.67857432e-01 1.20937622e+00 -4.74864662e-01
-1.36464798e+00 8.90392900e-01 -2.48764485e-01 -4.17649478e-01
5.32948792e-01 3.38087752e-02 -8.95657599e-01 -3.39188486e-01
9.24424231e-02 1.10547908e-01 2.76752710e-01 -1.21682107e+00
-5.51223457e-01 -5.31185627e-01 -1.34298667e-01 -7.12791756e-02
-2.01632202e-01 1.93093374e-01 -3.28078866e-01 -8.10643137e-01
3.85390460e-01 -7.58302033e-01 1.33751497e-01 -5.95861435e-01
-6.84712470e-01 -1.84170306e-01 5.35664082e-01 -2.26814300e-01
1.91986191e+00 -2.09727573e+00 -1.75837830e-01 3.04730922e-01
-3.80019784e-01 1.31388396e-01 2.13558562e-02 1.06307983e+00
-4.13814604e-01 6.10632479e-01 -1.13148302e-01 1.80498138e-01
1.23105988e-01 2.96446711e-01 -8.06041658e-01 -2.14208156e-01
1.70381829e-01 6.75171435e-01 -3.53529364e-01 -5.68714142e-01
-2.64139861e-01 -4.94084209e-01 -7.46158004e-01 4.70353961e-01
-7.25053787e-01 -9.88384485e-02 -4.29944873e-01 7.35584617e-01
6.43812597e-01 1.54492289e-01 3.81448358e-01 -5.50207853e-01
-1.01258568e-02 5.09225011e-01 -1.45078111e+00 1.43762565e+00
-2.20494241e-01 1.94331855e-01 -6.58481658e-01 -9.50685918e-01
1.27484381e+00 -6.60253093e-02 1.02185063e-01 -9.78443861e-01
-2.88119078e-01 4.50266629e-01 8.40375423e-02 -1.17525029e+00
6.69212043e-01 -8.35750848e-02 -5.37187219e-01 1.91410288e-01
6.86813379e-03 -3.01779121e-01 5.07814288e-01 1.23729497e-01
9.75923419e-01 4.93733473e-02 3.90128672e-01 -2.18200684e-01
6.46045268e-01 9.85515118e-02 4.65233207e-01 1.08613360e+00
3.40468377e-01 3.36002618e-01 7.73377299e-01 -4.11965281e-01
-1.01392400e+00 -1.25003862e+00 -2.89432913e-01 1.09613681e+00
-1.30778864e-01 -7.09776223e-01 -6.49631560e-01 -7.22307444e-01
1.49722397e-01 1.50754952e+00 -3.31053585e-01 -3.12986284e-01
-9.08766806e-01 -6.73305988e-01 9.80669260e-01 2.20066860e-01
1.82038486e-01 -1.08137941e+00 -4.96093810e-01 3.05175245e-01
-6.15110323e-02 -1.15904880e+00 -1.93649992e-01 2.92541653e-01
-1.00125360e+00 -1.13843167e+00 3.14253867e-01 -7.37180650e-01
3.32301587e-01 -2.40736097e-01 1.29780221e+00 -1.91733137e-01
-1.33799553e-01 3.02319109e-01 -3.89543831e-01 -6.06576979e-01
-1.20879769e+00 4.52037245e-01 -1.63029045e-01 -1.76690653e-01
7.51141667e-01 -3.66930127e-01 9.55564603e-02 4.80074257e-01
-1.46900356e+00 -1.70454860e-01 4.85878617e-01 5.61272740e-01
4.50984359e-01 2.23999441e-01 6.53434694e-01 -1.49820793e+00
9.39450622e-01 -4.19022024e-01 -4.51537609e-01 7.99962461e-01
-8.47083688e-01 4.91329998e-01 7.81841218e-01 -5.55805087e-01
-1.28134894e+00 -2.53374785e-01 -1.72730431e-01 1.22462094e-01
-4.39094305e-01 9.28279161e-01 -7.20364630e-01 3.70245248e-01
1.16477036e+00 4.59024251e-01 -9.42762420e-02 -5.58919787e-01
3.94830048e-01 6.75539196e-01 4.21271473e-01 -1.10772896e+00
7.76421070e-01 -1.13754049e-01 -1.86881632e-01 -8.40635598e-01
-4.08264548e-01 -4.13033292e-02 -5.09108543e-01 2.88770527e-01
1.98334962e-01 -4.38268661e-01 -3.88481885e-01 3.32881629e-01
-1.25000525e+00 4.93522659e-02 -4.10983920e-01 2.02230606e-02
-4.92940485e-01 5.96999705e-01 -1.83744386e-01 -4.33420420e-01
7.06133246e-02 -1.08141959e+00 6.91514671e-01 -1.67997237e-02
-4.62495416e-01 -8.66090119e-01 1.19529739e-01 1.14316113e-01
4.45221841e-01 3.01879048e-02 1.82997561e+00 -1.30125916e+00
-5.21935940e-01 -2.11458772e-01 2.39154220e-01 1.96753383e-01
3.09774250e-01 4.37085748e-01 -5.03721535e-01 -1.78307500e-02
3.45890611e-01 -2.37080246e-01 5.74475713e-02 -2.71909416e-01
1.30836582e+00 -4.14778054e-01 -9.83901024e-02 5.12352288e-01
1.46001232e+00 4.65013534e-01 6.82420313e-01 5.54468811e-01
1.40336752e-01 7.95394242e-01 8.20500433e-01 4.70834106e-01
4.83011186e-01 8.31856430e-01 1.16199717e-01 3.56751591e-01
4.24412340e-02 -5.24902046e-01 2.96335667e-01 8.76414359e-01
6.73783243e-01 -5.24479508e-01 -1.02147770e+00 2.98910946e-01
-1.44787538e+00 -9.06074822e-01 -1.25085935e-01 2.26618576e+00
1.15678227e+00 5.24563789e-01 -1.57937109e-01 -6.20890483e-02
6.20427787e-01 -5.78704290e-02 -4.90822375e-01 -5.96836746e-01
-6.16454244e-01 1.53734684e-01 1.09416790e-01 2.25880489e-01
-5.54808676e-01 1.06955361e+00 6.01186514e+00 6.57877445e-01
-1.12319469e+00 -5.08710444e-01 2.79088050e-01 1.84639573e-01
-8.32728267e-01 2.74812765e-02 -1.15383840e+00 3.88855547e-01
1.11481869e+00 -6.23116016e-01 2.30077386e-01 7.65887439e-01
-1.77897550e-02 1.11318707e-01 -1.67673695e+00 7.80333221e-01
1.97963983e-01 -1.15927732e+00 8.22288454e-01 -3.54443729e-01
2.10705712e-01 -3.76035362e-01 -4.00189683e-02 5.53148746e-01
-2.13221908e-01 -6.51828349e-01 7.18818903e-01 6.96247935e-01
5.40794551e-01 -4.84960467e-01 6.55922651e-01 5.69805384e-01
-7.87711620e-01 2.97710765e-03 -5.56281030e-01 3.59703511e-01
-2.09850222e-01 3.37084413e-01 -1.14002621e+00 7.91572869e-01
7.91552007e-01 2.25678742e-01 -1.17682362e+00 4.63197738e-01
2.11420387e-01 3.89835864e-01 -2.91949660e-01 -3.16693664e-01
-2.81683385e-01 -2.21780330e-01 5.08702517e-01 1.39957583e+00
5.64116299e-01 -1.80800378e-01 -1.84371948e-01 1.04644179e+00
5.83135858e-02 4.80191976e-01 -1.08989012e+00 -2.17763513e-01
7.30524004e-01 5.49261034e-01 -6.82049841e-02 -2.50012428e-01
-5.54544330e-01 5.22448897e-01 2.49352157e-01 4.36124474e-01
-7.01832116e-01 -4.71847117e-01 3.80212367e-01 3.60083252e-01
1.02805182e-01 1.59367561e-01 -4.94679302e-01 -1.20902479e+00
6.54603243e-01 -1.62582099e+00 5.68943202e-01 -7.79224694e-01
-1.54735827e+00 7.73739636e-01 3.14009815e-01 -1.05252874e+00
-4.43846256e-01 -5.28683305e-01 -4.28735644e-01 8.63429010e-01
-8.29598963e-01 -8.51842880e-01 -1.25838846e-01 5.38385749e-01
6.18949711e-01 -6.52360082e-01 9.31394279e-01 2.99064964e-01
-6.00285053e-01 1.09522355e+00 -2.19949752e-01 -1.00673333e-01
1.06559908e+00 -1.13484240e+00 6.02384090e-01 8.30360353e-01
-4.18348759e-02 1.28314352e+00 9.41035986e-01 -7.88524330e-01
-1.69594646e+00 -9.89835024e-01 1.19450986e+00 -6.15649521e-01
8.07391763e-01 -4.27519023e-01 -1.43088448e+00 7.96028078e-01
2.21543267e-01 -4.79710639e-01 8.39490354e-01 -2.52652671e-02
-4.44243848e-01 -4.64428544e-01 -1.08254707e+00 9.29830551e-01
1.09599543e+00 -6.75406098e-01 -1.04063904e+00 1.68483257e-01
1.05778778e+00 -4.69396472e-01 -1.05397356e+00 6.93323851e-01
2.36365974e-01 -1.20930958e+00 7.70622194e-01 -1.23053336e+00
3.00494790e-01 -2.22120062e-01 -7.06119895e-01 -9.09841061e-01
2.33328804e-01 -5.55811584e-01 5.92777058e-02 1.46502244e+00
7.56663382e-01 -7.76518643e-01 8.68715465e-01 6.90822363e-01
-2.16088161e-01 -6.36414945e-01 -8.65818083e-01 -7.99434423e-01
1.73523262e-01 -5.98673582e-01 1.00442779e+00 9.13285494e-01
1.88846551e-02 4.08058733e-01 2.59281188e-01 2.69353867e-01
1.14348680e-01 1.19223900e-01 1.18489718e+00 -9.41473365e-01
-3.05713892e-01 -3.51581097e-01 -1.95252895e-01 -1.13068402e+00
9.19932313e-03 -8.70927393e-01 -3.23152125e-01 -7.57318616e-01
-2.66721789e-02 -4.42083538e-01 1.68886155e-01 -1.07027777e-02
-1.44636407e-01 -7.83268273e-01 1.38543293e-01 -5.34019321e-02
-2.10737169e-01 4.30987120e-01 7.61943877e-01 -1.24974184e-01
-2.54468679e-01 3.15923065e-01 -7.71748006e-01 4.91334110e-01
8.01195383e-01 -7.49033093e-01 -6.84654057e-01 -3.89779508e-01
7.01876819e-01 2.16983467e-01 -9.09744650e-02 -8.03164124e-01
2.88363069e-01 -3.86958003e-01 -1.32768944e-01 -8.20829391e-01
-1.24697559e-01 -8.76743197e-01 4.43532348e-01 4.28739637e-01
-6.15125835e-01 6.83340549e-01 3.72950107e-01 2.95326233e-01
-4.75628585e-01 -6.66281641e-01 4.21645850e-01 -2.43800536e-01
-4.99897033e-01 -1.63936928e-01 -2.66229510e-01 5.00549674e-01
7.22487748e-01 -3.69187266e-01 -5.43121159e-01 -1.59954131e-01
-4.56247658e-01 -1.89055905e-01 5.50324380e-01 6.98595583e-01
5.84429860e-01 -1.10403097e+00 -5.20572543e-01 5.97849786e-01
6.70449913e-01 -1.07155226e-01 -5.35475574e-02 2.18369335e-01
-6.89588010e-01 5.91172457e-01 -8.75117332e-02 -5.94169676e-01
-9.89793897e-01 7.82325566e-01 3.79465044e-01 -1.56793445e-01
-1.18762217e-01 4.33668792e-01 -1.55991046e-02 -1.02246606e+00
3.73011678e-01 -6.90630913e-01 -7.64077017e-03 -1.27750427e-01
3.17841023e-02 2.84298033e-01 5.12998819e-01 -1.92364193e-02
-1.87502250e-01 3.99071693e-01 -3.18781465e-01 1.80919115e-02
1.01114368e+00 -5.59355319e-02 -3.28091949e-01 7.88971722e-01
1.14833629e+00 2.81848967e-01 -2.36449510e-01 -4.96833354e-01
7.42752552e-01 -4.51121092e-01 -9.94861126e-01 -7.34343529e-01
-4.30038035e-01 6.48967862e-01 4.67019707e-01 5.73470831e-01
1.04337430e+00 1.16046972e-03 3.43507260e-01 1.11633778e+00
3.87680680e-01 -1.06042123e+00 1.21918023e-01 5.69011867e-01
1.25559247e+00 -1.05349958e+00 -1.77236125e-01 -5.35596132e-01
-7.06259549e-01 1.49694419e+00 9.99854028e-01 4.90820646e-01
5.33115387e-01 2.89000452e-01 2.18764082e-01 -2.62119859e-01
-9.36870277e-01 2.66175121e-01 -5.72557747e-03 4.64566946e-01
5.52148759e-01 -2.83775002e-01 -2.56979525e-01 6.35630429e-01
-6.40820622e-01 -2.67519206e-01 8.58793616e-01 1.07416368e+00
-2.74147272e-01 -1.67287385e+00 -3.54723364e-01 4.88096207e-01
-2.53938138e-01 -9.23533067e-02 -5.68679988e-01 1.18743420e+00
4.50886600e-02 8.79410923e-01 -2.02519745e-02 -4.37405974e-01
1.08933544e+00 4.44902629e-01 3.54387313e-01 -9.01041269e-01
-7.22084224e-01 -3.26886177e-01 3.24929267e-01 -4.38509762e-01
-4.76320311e-02 -7.78355718e-01 -9.80814278e-01 -3.67786646e-01
-9.86718237e-02 1.33304298e-01 5.80467224e-01 6.61192596e-01
4.86875921e-01 3.19144666e-01 4.91810918e-01 4.02581841e-01
-1.12168396e+00 -9.56468821e-01 -3.10624152e-01 8.77511024e-01
-3.32811087e-01 -2.15965763e-01 -1.99959591e-01 9.09840018e-02] | [9.827136039733887, 7.834438323974609] |
485792ef-6f0b-47da-9328-957b21955a91 | cream-weakly-supervised-object-localization | 2205.13922 | null | https://arxiv.org/abs/2205.13922v1 | https://arxiv.org/pdf/2205.13922v1.pdf | CREAM: Weakly Supervised Object Localization via Class RE-Activation Mapping | Weakly Supervised Object Localization (WSOL) aims to localize objects with image-level supervision. Existing works mainly rely on Class Activation Mapping (CAM) derived from a classification model. However, CAM-based methods usually focus on the most discriminative parts of an object (i.e., incomplete localization problem). In this paper, we empirically prove that this problem is associated with the mixup of the activation values between less discriminative foreground regions and the background. To address it, we propose Class RE-Activation Mapping (CREAM), a novel clustering-based approach to boost the activation values of the integral object regions. To this end, we introduce class-specific foreground and background context embeddings as cluster centroids. A CAM-guided momentum preservation strategy is developed to learn the context embeddings during training. At the inference stage, the re-activation mapping is formulated as a parameter estimation problem under Gaussian Mixture Model, which can be solved by deriving an unsupervised Expectation-Maximization based soft-clustering algorithm. By simply integrating CREAM into various WSOL approaches, our method significantly improves their performance. CREAM achieves the state-of-the-art performance on CUB, ILSVRC and OpenImages benchmark datasets. Code will be available at https://github.com/Jazzcharles/CREAM. | ['Shang Gao', 'Xuequan Lu', 'Tao Zhang', 'Rui-Wei Zhao', 'Rui Feng', 'Yuejie Zhang', 'Junlin Hou', 'Jilan Xu'] | 2022-05-27 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Xu_CREAM_Weakly_Supervised_Object_Localization_via_Class_RE-Activation_Mapping_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Xu_CREAM_Weakly_Supervised_Object_Localization_via_Class_RE-Activation_Mapping_CVPR_2022_paper.pdf | cvpr-2022-1 | ['weakly-supervised-object-localization'] | ['computer-vision'] | [-6.55398443e-02 -1.58793375e-01 -4.67498034e-01 -3.56904387e-01
-8.80636871e-01 -3.60898077e-01 5.90779066e-01 1.18895367e-01
-4.86877978e-01 2.86677241e-01 -7.64866620e-02 1.31914228e-01
5.44409417e-02 -3.97318661e-01 -7.82373071e-01 -1.16343689e+00
3.06809634e-01 2.55959123e-01 4.31549191e-01 2.38282472e-01
1.88151762e-01 3.68419439e-01 -1.28394103e+00 4.49688226e-01
8.31142128e-01 1.05170608e+00 4.52010304e-01 3.42992604e-01
-2.29911923e-01 8.78169835e-01 -2.01287270e-01 -1.64014027e-01
1.27903044e-01 -4.56170708e-01 -6.90812528e-01 2.44901583e-01
4.06134456e-01 -1.84032880e-02 1.25842693e-03 1.36088717e+00
2.60010391e-01 1.16217703e-01 7.29800642e-01 -1.28696585e+00
-6.32597506e-01 3.40609610e-01 -8.84387195e-01 4.92640018e-01
-2.69138783e-01 -1.07779585e-01 9.78823602e-01 -1.37626684e+00
3.88756514e-01 9.72236037e-01 4.21670198e-01 4.97320384e-01
-1.33585835e+00 -4.63837028e-01 5.62446773e-01 4.31774467e-01
-1.81340921e+00 -2.09574699e-01 1.06093276e+00 -4.53735530e-01
3.14560950e-01 1.48202240e-01 4.32092339e-01 8.22494864e-01
-1.31906554e-01 1.11346531e+00 1.07766116e+00 -4.05558676e-01
4.58371729e-01 5.59033811e-01 1.31464779e-01 6.68391943e-01
1.65583462e-01 -2.82878548e-01 -4.32529420e-01 -2.38250159e-02
6.52529418e-01 1.31875038e-01 -3.51224482e-01 -7.23696589e-01
-1.04208803e+00 7.91251242e-01 6.80468619e-01 3.39963794e-01
-3.24557155e-01 9.81631428e-02 6.54096678e-02 -3.72209460e-01
6.72210693e-01 -1.45593463e-02 -4.34470892e-01 2.46668518e-01
-1.00738847e+00 2.02981103e-02 4.08704907e-01 8.66262257e-01
9.27535653e-01 -3.09843332e-01 -2.25752562e-01 9.50475812e-01
6.13962889e-01 1.86559141e-01 4.45919245e-01 -7.34318852e-01
2.31956333e-01 7.24955916e-01 5.15764989e-02 -9.72477317e-01
-1.31980896e-01 -6.83960915e-01 -6.11055255e-01 1.46528378e-01
3.98383021e-01 1.65096521e-01 -7.76862919e-01 1.69205177e+00
8.33220959e-01 7.60868967e-01 -2.45396756e-02 1.10811365e+00
5.91928124e-01 6.63993895e-01 1.24014020e-01 -9.46541876e-02
1.34945655e+00 -1.40232790e+00 -6.41531467e-01 -2.68222541e-01
5.34962177e-01 -5.89195073e-01 1.17195821e+00 1.53370917e-01
-8.54585648e-01 -5.87296486e-01 -9.28136706e-01 -7.63358474e-02
-3.59029800e-01 6.13342762e-01 4.20343667e-01 4.11734849e-01
-8.20793569e-01 3.76046509e-01 -1.09938788e+00 -1.72166497e-01
7.92471528e-01 1.94600001e-01 -2.14018613e-01 -9.67008844e-02
-5.54272175e-01 6.27165318e-01 4.55009788e-01 2.58026749e-01
-9.40640032e-01 -6.58556521e-01 -7.22762227e-01 -1.05544917e-01
5.81530511e-01 -2.89856821e-01 8.59526515e-01 -1.18553710e+00
-1.27364194e+00 9.71099854e-01 -2.50180066e-01 -3.26808035e-01
3.22267532e-01 -2.64814466e-01 -6.33797646e-02 3.44992042e-01
2.30985254e-01 7.07355082e-01 1.01257420e+00 -1.58439600e+00
-6.99328840e-01 -4.66771185e-01 -1.25498205e-01 2.15978861e-01
-3.65300298e-01 6.92111477e-02 -9.25755441e-01 -7.73182392e-01
2.32836455e-01 -8.76153886e-01 -1.51081994e-01 1.00476958e-01
-2.35003233e-01 -4.48040456e-01 8.19940448e-01 -4.95394051e-01
1.13812220e+00 -2.34406996e+00 2.72200197e-01 3.30894813e-02
1.96726307e-01 3.26192945e-01 -5.29722497e-02 -1.44379884e-01
-8.50442946e-02 -3.98502275e-02 -3.54707479e-01 -6.88466668e-01
-8.09233114e-02 1.90684885e-01 -1.76652223e-01 8.11341226e-01
5.39647102e-01 1.04226673e+00 -9.00195301e-01 -8.12897205e-01
3.22073072e-01 6.25889182e-01 -6.35759652e-01 2.02781826e-01
-1.47699684e-01 5.35484254e-01 -4.10442591e-01 8.75681400e-01
8.48443687e-01 -4.47586864e-01 5.29786162e-02 -4.49380636e-01
-1.00589365e-01 -1.07656635e-01 -1.16740441e+00 1.82466185e+00
-3.20507079e-01 3.64455819e-01 2.69263804e-01 -1.30576563e+00
5.31583548e-01 -7.32530653e-02 3.74164790e-01 -2.93563783e-01
2.72694200e-01 1.11345001e-01 -1.91075653e-01 -5.00990093e-01
9.86910984e-02 -4.09626812e-02 2.29925886e-01 1.62256092e-01
2.37105414e-01 2.55277663e-01 1.72921419e-02 1.51706353e-01
6.16975248e-01 4.07981634e-01 1.78557575e-01 -4.37329233e-01
6.82501853e-01 -9.20423791e-02 6.70573533e-01 5.76477468e-01
-3.37222993e-01 6.71759307e-01 3.79371345e-01 5.79058193e-03
-5.79959631e-01 -1.07003927e+00 -3.69991869e-01 1.17883945e+00
4.61220145e-01 -3.66073459e-01 -1.03621840e+00 -8.64073336e-01
-1.58026293e-01 4.39989507e-01 -8.19038451e-01 -2.77065963e-01
-5.53040624e-01 -1.02174771e+00 2.02348381e-01 6.74330235e-01
4.31052536e-01 -7.18603790e-01 -3.93841296e-01 -4.04359773e-02
-1.40810266e-01 -1.17421246e+00 -5.17073452e-01 4.46019888e-01
-8.21514964e-01 -9.96363640e-01 -8.13739777e-01 -1.01231480e+00
1.03582215e+00 3.32336009e-01 7.60037243e-01 6.13184497e-02
-3.13736528e-01 3.63093466e-01 -4.14363623e-01 -1.68157160e-01
1.32602125e-01 -6.09285682e-02 1.70993861e-02 5.58155298e-01
4.28150505e-01 -3.69951606e-01 -6.75625145e-01 4.67859983e-01
-9.12386596e-01 2.07298137e-02 6.58596694e-01 8.97492409e-01
1.07601655e+00 -1.50121767e-02 4.25921947e-01 -7.86243796e-01
3.49204913e-02 -6.14113867e-01 -6.52852595e-01 3.35311711e-01
-5.00492454e-01 -4.69287597e-02 3.77914488e-01 -6.64491355e-01
-9.75803435e-01 3.45647544e-01 9.17582121e-03 -7.92951405e-01
-2.25533694e-01 2.95263290e-01 -5.05611300e-01 -1.90052390e-01
3.55935305e-01 4.02893990e-01 -2.81255692e-01 -5.75832248e-01
5.66331685e-01 6.38332009e-01 6.28711760e-01 -7.13010848e-01
7.85292923e-01 8.24895322e-01 -3.80240351e-01 -6.76025987e-01
-1.20580983e+00 -8.73192012e-01 -7.98273802e-01 -2.34232962e-01
1.00360024e+00 -1.02524483e+00 -2.99982309e-01 2.57818609e-01
-8.88780594e-01 -4.88535136e-01 -2.90980965e-01 5.44139266e-01
-3.70350838e-01 3.15347284e-01 -3.52750212e-01 -7.49188960e-01
8.99873078e-02 -1.15719914e+00 1.20515037e+00 3.46970081e-01
2.37714455e-01 -9.77458239e-01 5.38009172e-03 4.33929592e-01
1.01875104e-01 4.11489941e-02 5.75736344e-01 -5.81535041e-01
-7.02286541e-01 -1.90359443e-01 -5.05988419e-01 6.33336365e-01
1.08519822e-01 -2.54726738e-01 -1.15705073e+00 -2.24707916e-01
1.28219455e-01 -1.21622100e-01 1.11947703e+00 4.62473720e-01
1.46672416e+00 -2.64291048e-01 -5.07190526e-01 8.00054669e-01
1.40000355e+00 -4.68170792e-02 3.47500235e-01 1.36870712e-01
9.57786977e-01 4.87542659e-01 7.93549955e-01 3.22345018e-01
3.30462754e-01 7.29604661e-01 4.54305977e-01 -1.38660669e-01
-2.20077008e-01 -1.84478357e-01 3.11007261e-01 6.99378550e-01
8.40283185e-02 -1.82511061e-02 -8.02506089e-01 6.23671234e-01
-1.99926484e+00 -5.93181491e-01 -1.40240699e-01 2.05680013e+00
9.46992576e-01 7.90270790e-02 5.08804508e-02 -1.54202119e-01
8.21163774e-01 -9.52638744e-04 -5.49640238e-01 3.72182995e-01
-4.81475741e-02 7.45415017e-02 3.60454112e-01 5.30245125e-01
-1.43150878e+00 1.07874811e+00 4.65610456e+00 1.16734636e+00
-1.04628396e+00 6.32925689e-01 7.08136499e-01 -2.81246305e-01
9.46243107e-02 8.37623775e-02 -1.03837311e+00 7.89178371e-01
4.64197993e-01 2.43541449e-01 2.94174373e-01 1.06698239e+00
7.20782131e-02 -2.31314853e-01 -1.06443512e+00 1.08090806e+00
3.41827691e-01 -1.22676980e+00 -1.95337027e-01 7.23615382e-03
8.14345837e-01 3.42749022e-02 1.14798084e-01 3.39159399e-01
-2.39086792e-01 -7.38122344e-01 9.11561728e-01 3.65402102e-01
4.75503922e-01 -6.17232144e-01 7.53324807e-01 4.25247610e-01
-1.29279983e+00 -1.28982410e-01 -5.04802108e-01 2.66394556e-01
-3.49207744e-02 7.08162010e-01 -3.68926615e-01 3.65196884e-01
8.54363739e-01 7.45183587e-01 -7.63409257e-01 9.55879867e-01
-4.32160705e-01 7.56036520e-01 -3.36271465e-01 1.68223575e-01
2.09851265e-01 -2.78981894e-01 3.97478729e-01 1.30612242e+00
-8.85846987e-02 8.60105548e-03 2.75860548e-01 1.23249710e+00
-1.76223982e-02 8.55448321e-02 -9.14636403e-02 1.68489397e-01
2.51257509e-01 1.66225803e+00 -1.10833788e+00 -3.53864878e-01
-3.17182481e-01 1.18212914e+00 5.77433228e-01 3.95994693e-01
-1.13659167e+00 -9.97443795e-02 3.99392039e-01 1.17826268e-01
5.08116424e-01 -1.52559444e-01 -1.86402574e-01 -1.28916132e+00
1.26012564e-01 -5.01843512e-01 2.64732987e-01 -5.84780991e-01
-1.18400788e+00 3.80601197e-01 1.57511786e-01 -1.11981285e+00
4.25125778e-01 -6.40231848e-01 -7.82272100e-01 5.41871190e-01
-1.72746980e+00 -1.18526053e+00 -5.36672950e-01 5.04487276e-01
4.94563848e-01 8.63086432e-02 3.47625494e-01 6.42188191e-01
-1.03531289e+00 5.53148448e-01 4.39066328e-02 1.66625559e-01
6.38427079e-01 -1.26622808e+00 -3.88560921e-01 9.40769792e-01
2.98947334e-01 7.14068770e-01 3.11281115e-01 -4.89506274e-01
-1.00619292e+00 -1.45676672e+00 5.63249409e-01 -5.16930938e-01
6.70567930e-01 -5.46320081e-01 -1.03056955e+00 5.12365878e-01
2.59900670e-02 6.23617768e-01 7.05124021e-01 -1.73557833e-01
-1.62227720e-01 -6.87626973e-02 -8.90865803e-01 4.38710630e-01
8.69574249e-01 -5.01513958e-01 -4.73128557e-01 3.44985664e-01
5.82745969e-01 -2.41208509e-01 -5.95698953e-01 3.87036800e-01
1.06267564e-01 -6.99717581e-01 1.04592264e+00 -4.53047484e-01
2.31566533e-01 -8.55751038e-01 -2.87959665e-01 -9.47866619e-01
-2.95593172e-01 -8.76279026e-02 -4.74660099e-01 1.43432748e+00
1.11268103e-01 -3.56627613e-01 7.85613954e-01 1.24887235e-01
-1.08121827e-01 -1.09618831e+00 -9.03245211e-01 -7.31660724e-01
-3.69939953e-02 -3.92769516e-01 1.69493496e-01 1.00752175e+00
-2.15178624e-01 2.58251339e-01 -3.48094888e-02 5.23410499e-01
8.21204901e-01 1.36514410e-01 5.91235816e-01 -8.94651234e-01
-2.30439574e-01 -4.85982597e-01 -2.95030415e-01 -1.04341269e+00
3.33233297e-01 -1.21845007e+00 2.81205148e-01 -1.37658679e+00
3.88307422e-01 -5.90128660e-01 -6.41089737e-01 5.82745254e-01
-4.68862802e-01 4.25076604e-01 1.76805124e-01 3.01201940e-01
-1.09738708e+00 8.05140853e-01 9.58067775e-01 -1.60471529e-01
-7.60095641e-02 -3.69878951e-03 -5.96595347e-01 8.80762815e-01
9.34022903e-01 -6.45924211e-01 -2.25152895e-01 -2.43842259e-01
1.48025127e-02 -5.80630004e-01 8.91116023e-01 -9.60851610e-01
3.49079907e-01 -4.77655195e-02 4.18447137e-01 -7.59397566e-01
2.73813576e-01 -7.37995088e-01 -2.91595817e-01 2.54201889e-01
-3.18111092e-01 -4.44723129e-01 1.27527103e-01 7.59417057e-01
-2.23203152e-01 -3.77054065e-01 1.04322004e+00 2.78829690e-02
-7.20957458e-01 3.27874750e-01 -1.99810952e-01 1.31961703e-01
1.19744921e+00 2.33719833e-02 -1.30490260e-02 2.03988999e-01
-7.68513322e-01 2.99643248e-01 3.57125103e-01 3.32863986e-01
5.54427564e-01 -1.52261198e+00 -5.03011703e-01 1.45514250e-01
2.54087240e-01 4.12053883e-01 2.00126156e-01 1.41577232e+00
-2.36058846e-01 1.06677346e-01 2.19230890e-01 -9.24829245e-01
-1.19073343e+00 5.98947048e-01 3.94494742e-01 -1.29887253e-01
-4.98901963e-01 1.25054967e+00 5.94983876e-01 -2.85110831e-01
4.50130582e-01 -4.13108289e-01 -1.02489799e-01 1.12956040e-01
5.93171299e-01 2.13417351e-01 -1.08988330e-01 -6.43828392e-01
-7.02116191e-01 6.02043450e-01 -8.74589980e-02 1.27701849e-01
1.17118168e+00 -1.31238014e-01 -1.95238307e-01 5.00722647e-01
1.38876951e+00 -1.76006416e-03 -1.55894101e+00 -4.11681443e-01
8.94298106e-02 -4.97904539e-01 4.31729615e-01 -4.30869520e-01
-1.26738501e+00 9.24677849e-01 9.40296113e-01 -2.22620934e-01
1.07485771e+00 5.24620712e-01 3.80408645e-01 1.27701029e-01
1.18876562e-01 -1.22912753e+00 3.23058724e-01 1.31710947e-01
7.19351590e-01 -1.41991115e+00 -4.79942784e-02 -3.84064257e-01
-6.42100215e-01 7.40862072e-01 7.87213206e-01 -2.44274512e-01
8.59858811e-01 3.06541800e-01 -7.57634118e-02 -8.89973789e-02
-4.67765242e-01 -4.60329413e-01 4.09422636e-01 4.13573086e-01
2.24649370e-01 4.71217034e-04 -9.11080465e-02 9.43243921e-01
4.62278813e-01 -1.36825562e-01 -2.33631432e-02 9.27194118e-01
-4.24473077e-01 -9.38289046e-01 -4.44085360e-01 1.54684529e-01
-4.74278957e-01 -1.79822713e-01 -2.71574706e-01 5.06907642e-01
6.10191941e-01 7.78793454e-01 9.78316143e-02 -1.72055110e-01
-8.91334713e-02 -3.80272605e-02 4.59108233e-01 -7.43545532e-01
-1.52783111e-01 5.36699712e-01 -6.13999188e-01 -5.82121968e-01
-6.58540010e-01 -7.46140301e-01 -1.42199636e+00 3.91375661e-01
-6.49711430e-01 1.23485535e-01 5.64444363e-01 7.50130951e-01
2.86713392e-01 6.19708121e-01 5.12688577e-01 -1.07748580e+00
-3.40583622e-01 -8.18961978e-01 -5.70960939e-01 3.16495657e-01
2.30875760e-01 -9.22826886e-01 -4.07430857e-01 1.62902445e-01] | [9.62022876739502, 0.774587869644165] |
f6af8459-4d90-4f14-a2c8-61b849af8e39 | robust-point-light-source-estimation-using | 1812.04857 | null | http://arxiv.org/abs/1812.04857v1 | http://arxiv.org/pdf/1812.04857v1.pdf | Robust Point Light Source Estimation Using Differentiable Rendering | Illumination estimation is often used in mixed reality to re-render a scene
from another point of view, to change the color/texture of an object, or to
insert a virtual object consistently lit into a real video or photograph.
Specifically, the estimation of a point light source is required for the
shadows cast by the inserted object to be consistent with the real scene. We
tackle the problem of illumination retrieval given an RGBD image of the scene
as an inverse problem: we aim to find the illumination that minimizes the
photometric error between the rendered image and the observation. In particular
we propose a novel differentiable renderer based on the Blinn-Phong model with
cast shadows. We compare our differentiable renderer to state-of-the-art
methods and demonstrate its robustness to an incorrect reflectance estimation. | ['Grégoire Nieto', 'Philippe Robert', 'Salma Jiddi'] | 2018-12-12 | null | null | null | null | ['light-source-estimation'] | ['computer-vision'] | [ 7.78831363e-01 -1.17651515e-01 6.21288836e-01 -3.50835532e-01
-2.96679944e-01 -6.01842523e-01 4.52515960e-01 -4.02449042e-01
-2.02847496e-01 5.60310006e-01 -3.92035931e-01 -6.17531203e-02
1.31592840e-01 -7.66611159e-01 -8.22107971e-01 -9.19146299e-01
7.73470938e-01 3.19348454e-01 1.68919042e-02 -6.30146638e-02
2.96171427e-01 7.23929346e-01 -1.67456436e+00 -8.17410722e-02
5.90403497e-01 9.57179666e-01 5.66530049e-01 8.43510032e-01
-3.85403819e-02 5.24201572e-01 -5.05334973e-01 -2.19777077e-01
4.87744123e-01 -4.65629846e-01 -5.74612260e-01 7.25114405e-01
6.64124787e-01 -5.46443164e-01 -5.27734347e-02 1.05120850e+00
2.01746926e-01 3.56873691e-01 4.41740900e-01 -9.56057787e-01
-4.68251079e-01 -5.54822803e-01 -6.91847146e-01 -3.40862423e-01
7.41461933e-01 -5.93756288e-02 5.35586059e-01 -8.29338551e-01
7.01945007e-01 9.22838748e-01 5.05088866e-02 4.81237531e-01
-1.43414164e+00 -2.34884188e-01 2.52898544e-01 -3.60387051e-03
-1.22022319e+00 -5.88962018e-01 1.09089243e+00 -2.04653770e-01
2.39136115e-01 5.71402669e-01 8.71338665e-01 8.09594810e-01
2.00272709e-01 9.74373668e-02 1.52540481e+00 -6.12404168e-01
4.87793535e-01 2.07872316e-01 -2.43831992e-01 7.66743302e-01
7.29013160e-02 1.36340126e-01 -5.79924285e-01 -2.38463134e-01
8.92777801e-01 1.83966100e-01 -6.56049728e-01 -5.56393683e-01
-9.67705309e-01 1.85796857e-01 3.25189263e-01 -1.84248805e-01
-5.14573157e-01 2.02762470e-01 -3.39075863e-01 4.90940250e-02
6.41488314e-01 2.19598174e-01 -2.35035717e-01 4.13914919e-02
-6.69055998e-01 -4.25803810e-02 7.23259151e-01 5.59964776e-01
1.01869810e+00 -4.71587144e-02 2.82604039e-01 5.22375703e-01
5.28821468e-01 8.25092077e-01 -1.35651574e-01 -1.23500693e+00
1.37389779e-01 5.33676684e-01 4.34928060e-01 -8.79514933e-01
1.15749478e-01 -1.03232346e-01 -4.44599807e-01 8.58932137e-01
3.31242055e-01 1.80418149e-01 -7.64006972e-01 1.36280310e+00
8.96250308e-01 5.13982475e-01 3.82975745e-03 1.21780729e+00
6.06270671e-01 5.84958553e-01 -7.11216331e-01 -3.41812223e-01
1.11303508e+00 -5.87244272e-01 -6.96193814e-01 -4.47102815e-01
-3.25602852e-02 -1.06189215e+00 1.05356610e+00 7.78187990e-01
-1.11530638e+00 -3.99914861e-01 -1.06274211e+00 -2.33316049e-01
1.33533776e-01 2.32396260e-01 1.89509526e-01 6.89733684e-01
-8.00878525e-01 4.52742398e-01 -6.79929674e-01 -1.85448617e-01
-2.36555025e-01 8.87606964e-02 -2.53191918e-01 -2.99360156e-01
-4.84502584e-01 7.80178487e-01 6.20634342e-03 6.05902195e-01
-8.98475528e-01 -4.72623467e-01 -5.53386629e-01 -1.40996397e-01
5.71259022e-01 -7.75536835e-01 1.02119803e+00 -1.46436548e+00
-2.24964666e+00 1.08961749e+00 -5.58277249e-01 2.73845613e-01
6.31011009e-01 -2.65440226e-01 -6.89342991e-02 2.02907979e-01
-3.86131555e-01 -1.34645104e-01 1.28518486e+00 -1.92791367e+00
-2.45592609e-01 -5.42467058e-01 1.80393144e-01 5.14778614e-01
3.05935621e-01 -2.37578258e-01 -6.19717419e-01 -1.03114493e-01
6.38314724e-01 -9.71364737e-01 -6.59021316e-05 2.56973535e-01
-5.76247752e-01 5.67360938e-01 9.30174768e-01 -6.57631457e-01
3.69783282e-01 -2.19325089e+00 2.80378878e-01 1.99042380e-01
1.40432149e-01 -6.01346008e-02 -7.08104968e-02 2.21078500e-01
-1.14274085e-01 -5.29292285e-01 -1.62517741e-01 -7.37783074e-01
-2.75503635e-01 3.87385398e-01 -4.34286237e-01 9.49050426e-01
-4.27526206e-01 4.74409610e-01 -9.41057444e-01 -2.77548909e-01
6.59741044e-01 8.74994278e-01 -1.45557493e-01 7.18437076e-01
-2.72496670e-01 9.81212854e-01 -1.52580917e-01 3.07183921e-01
1.01776409e+00 1.31948978e-01 2.16899380e-01 -3.49086612e-01
-5.32412469e-01 2.69762218e-01 -1.40736711e+00 1.74671352e+00
-6.84959948e-01 5.53273916e-01 3.70965302e-01 -3.20503592e-01
1.00779498e+00 2.16751620e-01 4.05173600e-01 -7.26177096e-01
9.97888893e-02 1.17406383e-01 -5.87874234e-01 -4.25333977e-01
7.41157532e-01 -1.93000033e-01 5.88229120e-01 4.62018698e-01
-6.35449648e-01 -6.60395622e-01 -3.60454261e-01 -8.78013223e-02
7.41496801e-01 5.71295500e-01 3.49270739e-02 1.74155589e-02
4.26497459e-01 -4.27718550e-01 4.78173256e-01 4.16919470e-01
4.19087201e-01 8.62059891e-01 1.25408426e-01 -2.27501050e-01
-8.63481283e-01 -1.07631886e+00 -4.39196527e-02 6.59009635e-01
4.96598661e-01 1.54015020e-01 -7.67225027e-01 -8.25148821e-02
-2.64360130e-01 7.94734597e-01 -4.49246436e-01 1.83898896e-01
-4.47144121e-01 -3.27286601e-01 -1.84976384e-01 -1.70233607e-01
5.97395718e-01 -6.26260936e-01 -1.12272584e+00 -1.08878121e-01
-2.96049327e-01 -1.27715933e+00 -2.08175406e-01 -3.18046659e-02
-7.80763805e-01 -1.16857374e+00 -5.88492215e-01 -2.58307427e-01
1.03048658e+00 6.00422323e-01 1.09128952e+00 1.16728351e-01
-2.84432262e-01 7.63997197e-01 -2.83169687e-01 -8.89325291e-02
-4.39864844e-01 -6.49231732e-01 -1.15569539e-01 7.44436443e-01
-5.13036430e-01 -3.67679894e-01 -8.17036450e-01 5.77610135e-01
-1.24245465e+00 5.17338157e-01 -2.85115708e-02 3.46881092e-01
7.36857653e-01 -1.04464233e-01 -4.82704818e-01 -7.72564709e-01
3.05255130e-02 1.91129684e-01 -1.08842683e+00 4.27459389e-01
-3.44476640e-01 -2.04410911e-01 3.87964070e-01 -1.44089654e-01
-1.60938418e+00 3.95105481e-01 1.77771032e-01 -5.74441969e-01
-6.94844127e-02 -6.93613887e-02 -4.26022321e-01 -3.51542622e-01
3.85540634e-01 2.15066627e-01 -3.19742054e-01 -5.67417860e-01
4.17202175e-01 4.84309912e-01 6.25795782e-01 -6.24520123e-01
8.42469215e-01 1.16565573e+00 3.86294693e-01 -1.19657695e+00
-8.19490492e-01 -4.72113431e-01 -9.31297958e-01 -5.07457376e-01
8.38596880e-01 -5.23523033e-01 -9.67299640e-01 5.49415886e-01
-1.46736300e+00 -6.18635595e-01 -3.88826877e-01 3.55233073e-01
-5.96665442e-01 5.87032914e-01 -1.63656801e-01 -1.00291574e+00
8.43682140e-03 -1.26330519e+00 1.47336757e+00 1.85458824e-01
3.30892354e-01 -9.33585823e-01 3.17218393e-01 4.20102239e-01
1.12660965e-02 4.39311564e-01 7.07902014e-01 6.56893790e-01
-9.42942381e-01 6.18801191e-02 -2.83205301e-01 4.17260617e-01
5.15819430e-01 4.21334445e-01 -1.37573481e+00 -2.13030264e-01
3.63311499e-01 2.45487973e-01 4.92889315e-01 2.83156365e-01
9.79251921e-01 -7.01847747e-02 -1.02036133e-01 9.80596960e-01
1.80855870e+00 1.78386271e-01 8.71123314e-01 1.59141317e-01
7.55491912e-01 5.53368151e-01 6.28790140e-01 3.99836600e-01
1.17875159e-01 9.98449445e-01 9.21884775e-01 -3.54591250e-01
-9.46141109e-02 -1.16958814e-02 2.02826411e-01 6.07520223e-01
-3.42362136e-01 -3.97137076e-01 -4.57047790e-01 -8.71319249e-02
-1.60650027e+00 -5.57525337e-01 -5.04189432e-01 2.92187452e+00
3.82162809e-01 -3.76021862e-01 -5.18394768e-01 8.13625678e-02
5.44597983e-01 1.33337557e-01 -4.94701713e-01 -3.60173345e-01
-2.05788508e-01 -9.89820659e-02 5.45469999e-01 8.42738211e-01
-4.76467967e-01 6.49644852e-01 6.08631897e+00 7.18475506e-02
-1.37698174e+00 4.28929739e-02 2.44813994e-01 2.99034640e-02
-5.22848427e-01 3.26419890e-01 -3.68559808e-01 2.26804763e-01
3.07770878e-01 1.27290085e-01 9.48207915e-01 3.02479059e-01
3.60888869e-01 -7.96449900e-01 -1.26327944e+00 1.21890223e+00
3.56246769e-01 -6.50220513e-01 -2.39968657e-01 1.93218604e-01
7.65899956e-01 -2.05380306e-01 1.81924328e-01 -5.70628345e-01
2.55528931e-03 -5.31700432e-01 7.84104645e-01 9.95987296e-01
6.97605848e-01 -1.24078892e-01 5.69902472e-02 1.60374135e-01
-1.00034714e+00 4.05133754e-01 -3.49363923e-01 -6.12855181e-02
4.36792970e-01 7.18914390e-01 -7.98818886e-01 5.53476334e-01
3.51408929e-01 3.15038145e-01 -3.18229347e-01 9.11145449e-01
-4.35669214e-01 1.83129892e-01 -2.34618530e-01 3.57477188e-01
-8.66969451e-02 -9.34607923e-01 7.62462974e-01 5.86347044e-01
2.26579919e-01 2.12401196e-01 1.84982330e-01 1.08830059e+00
3.59885059e-02 -4.48483117e-02 -4.60061997e-01 5.71864724e-01
-2.04774171e-01 1.26701534e+00 -8.03741455e-01 -2.56579667e-01
-2.38827407e-01 1.71513557e+00 6.42480925e-02 1.03339636e+00
-7.91106224e-01 1.54079422e-01 5.44292927e-01 1.54564276e-01
-3.05140633e-02 -3.37665975e-01 1.70877799e-02 -1.21975279e+00
2.36519665e-01 -4.43333983e-01 -2.68481702e-01 -1.68062162e+00
-7.91495979e-01 5.51295638e-01 -2.11057097e-01 -1.22297692e+00
1.88985586e-01 -5.06638288e-01 -4.07019198e-01 1.02624559e+00
-1.67246711e+00 -8.90036702e-01 -7.90499449e-01 6.55169427e-01
3.77077401e-01 5.80046833e-01 8.46921742e-01 1.25380412e-01
-2.56553322e-01 -3.67413163e-01 5.06371081e-01 -2.83302903e-01
6.79044902e-01 -1.10873210e+00 1.21769927e-01 8.98053229e-01
9.07429680e-02 4.45494205e-01 9.99237597e-01 -5.30150235e-01
-1.93289292e+00 -6.97083354e-01 1.67150736e-01 -2.99396306e-01
1.57553107e-01 -3.83446842e-01 -7.86159158e-01 8.04044008e-01
1.90413892e-01 1.78407043e-01 4.47255485e-02 -2.58715659e-01
-2.50993788e-01 -3.90494674e-01 -1.14044976e+00 5.36640704e-01
7.86075890e-01 -7.58116841e-01 -1.77594692e-01 2.86248237e-01
4.07260269e-01 -9.28589642e-01 -3.60077083e-01 6.32569790e-02
7.37612784e-01 -1.28376305e+00 1.07018292e+00 -1.32134110e-01
1.65043950e-01 -5.43400228e-01 -3.24227899e-01 -1.55193698e+00
6.28074631e-02 -6.86786413e-01 1.93390831e-01 1.00394595e+00
-1.56585172e-01 -9.57694709e-01 6.32979333e-01 9.11101758e-01
5.16151078e-02 -2.34166726e-01 -9.90264535e-01 -4.47316498e-01
-6.83597803e-01 -3.04207981e-01 4.70609307e-01 6.98077083e-01
-6.19720459e-01 5.97514622e-02 -4.37776923e-01 6.53869450e-01
9.80525255e-01 4.88454551e-01 1.09402859e+00 -1.22252262e+00
-4.69829053e-01 1.65723488e-01 -2.23191977e-01 -1.10487366e+00
1.42620534e-01 -3.98505777e-01 3.31684947e-01 -1.59355938e+00
1.55790105e-01 -6.48537934e-01 1.80888370e-01 -1.06126562e-01
-8.65599141e-02 3.05474877e-01 2.99604475e-01 1.42692700e-01
-3.70574504e-01 5.46205699e-01 1.42612565e+00 -1.00018559e-02
-4.26523268e-01 1.55053109e-01 -1.07202999e-01 7.91261017e-01
4.28292781e-01 -4.95687693e-01 -2.25620583e-01 -8.35903406e-01
7.77241349e-01 2.57764012e-01 8.12307000e-01 -6.60739779e-01
2.68386677e-02 -3.19258362e-01 2.87684768e-01 -5.41126370e-01
9.04769063e-01 -1.31629181e+00 6.83772027e-01 1.37863919e-01
-6.50956333e-02 -1.55353010e-01 -1.07865863e-01 5.27662516e-01
2.65616298e-01 -5.28347850e-01 6.64318085e-01 -2.04272941e-01
-3.53909016e-01 1.56420022e-01 -2.63825376e-02 -2.62152821e-01
7.87601769e-01 -2.63810724e-01 -2.86512285e-01 -3.21955949e-01
-5.02217412e-01 -4.76024538e-01 1.10344517e+00 9.08917096e-03
9.36851144e-01 -9.93197799e-01 -3.76294404e-01 2.65953928e-01
6.77926615e-02 1.43877149e-01 1.97339803e-01 7.38546669e-01
-9.63161826e-01 -7.85235539e-02 1.63943142e-01 -8.08727622e-01
-1.70743966e+00 3.50980818e-01 7.31710374e-01 2.81692445e-01
-7.09854960e-01 4.03178900e-01 4.66150165e-01 -4.19677466e-01
-3.19818668e-02 -5.34016192e-01 3.99891168e-01 -5.12338996e-01
4.61481929e-01 5.64647436e-01 2.15536088e-01 -6.69028938e-01
-8.88081864e-02 9.42632973e-01 6.81593597e-01 -4.33294117e-01
1.23638237e+00 -5.55705905e-01 -4.63097572e-01 8.26883197e-01
1.20254385e+00 3.62692654e-01 -1.50521612e+00 -2.52426863e-01
-7.56637514e-01 -1.14293385e+00 4.41437185e-01 -6.45973802e-01
-1.09520030e+00 7.10997045e-01 6.94595158e-01 1.53272748e-01
1.21405315e+00 -1.70052633e-01 4.28714037e-01 3.01149309e-01
5.61422944e-01 -9.55232859e-01 -4.91824420e-03 1.75557137e-01
8.58926296e-01 -9.86226499e-01 4.16657478e-01 -5.15267551e-01
-3.85424852e-01 1.25187564e+00 1.37085110e-01 8.10934976e-02
5.23925900e-01 1.16960995e-01 2.96319574e-01 -3.09015751e-01
-1.98071018e-01 -1.30465254e-01 2.97165394e-01 2.41017431e-01
1.46278888e-01 5.14767729e-02 2.81621307e-01 -6.42916143e-01
1.45329654e-01 -2.88530663e-02 6.90279186e-01 7.28205144e-01
-1.95805162e-01 -8.85650992e-01 -9.40646708e-01 -1.78368017e-01
-2.09546480e-02 1.47805095e-01 -3.95977348e-01 2.80211478e-01
2.16929466e-02 8.72986197e-01 1.97433144e-01 2.30008468e-01
3.87744397e-01 -6.43891171e-02 9.26174939e-01 -5.22197783e-01
-2.06230164e-01 3.70836049e-01 -3.30668688e-01 -7.23822474e-01
-8.44911754e-01 -7.31616557e-01 -1.10207212e+00 -8.89505297e-02
-4.99490380e-01 -2.02807084e-01 1.38179731e+00 8.46466422e-01
-1.62466709e-02 5.11467457e-01 9.96571839e-01 -1.12959290e+00
1.61192164e-01 -4.93307292e-01 -1.00372779e+00 3.10299903e-01
5.84449768e-01 -4.66572225e-01 -6.26298428e-01 2.39144221e-01] | [9.78016185760498, -3.0352940559387207] |
b1c694b5-409c-4019-bb58-2052d8b5384f | peer-learning-for-unbiased-scene-graph | 2301.00146 | null | https://arxiv.org/abs/2301.00146v2 | https://arxiv.org/pdf/2301.00146v2.pdf | Peer Learning for Unbiased Scene Graph Generation | Unbiased scene graph generation (USGG) is a challenging task that requires predicting diverse and heavily imbalanced predicates between objects in an image. To address this, we propose a novel framework peer learning that uses predicate sampling and consensus voting (PSCV) to encourage multiple peers to learn from each other. Predicate sampling divides the predicate classes into sub-distributions based on frequency, and assigns different peers to handle each sub-distribution or combinations of them. Consensus voting ensembles the peers' complementary predicate knowledge by emphasizing majority opinion and diminishing minority opinion. Experiments on Visual Genome show that PSCV outperforms previous methods and achieves a new state-of-the-art on SGCls task with 31.6 mean. | ['Yangsheng Xu', 'Tin Lun Lam', 'Yuhongze Zhou', 'Junjie Hu', 'Liguang Zhou'] | 2022-12-31 | null | null | null | null | ['scene-graph-generation', 'unbiased-scene-graph-generation'] | ['computer-vision', 'computer-vision'] | [ 3.03889364e-01 4.61364955e-01 -5.20470798e-01 -4.43263203e-01
-8.03448677e-01 -1.55282065e-01 5.57303488e-01 3.66696239e-01
2.22452044e-01 1.10752869e+00 3.24152499e-01 1.15026914e-01
7.10140541e-02 -9.55184042e-01 -1.00473166e+00 -7.12437928e-01
2.29471028e-02 8.93444479e-01 8.60782921e-01 1.23180404e-01
1.88106552e-01 3.35591622e-02 -1.86090350e+00 8.84828389e-01
1.25262105e+00 8.69075298e-01 1.66059062e-02 5.44133127e-01
-3.39791715e-01 1.41985631e+00 -9.08550799e-01 -6.51195765e-01
2.10131526e-01 -5.07009566e-01 -6.57503664e-01 4.23406065e-02
9.48294044e-01 -3.94688286e-02 -5.59599884e-02 1.22309291e+00
6.80500329e-01 -4.81129847e-02 8.73205006e-01 -1.82376182e+00
-9.21227813e-01 9.47069764e-01 -1.06786287e+00 2.21017078e-01
3.37284565e-01 1.15255900e-01 1.31468821e+00 -6.53308094e-01
8.09373736e-01 1.43646252e+00 2.99315691e-01 3.76890838e-01
-1.17358685e+00 -8.87376666e-01 6.47188246e-01 7.48508632e-01
-1.22334993e+00 2.72743292e-02 8.74787271e-01 -3.90202165e-01
6.62537396e-01 3.37862998e-01 8.41813028e-01 8.62434387e-01
2.38431785e-02 1.35299528e+00 1.29471385e+00 -8.38960558e-02
3.50791484e-01 8.14061239e-02 5.57429940e-02 5.16369820e-01
7.04282761e-01 -3.53502035e-01 -1.16237414e+00 -4.10623699e-01
8.58246982e-02 -3.46613437e-01 -4.43701565e-01 -6.24572873e-01
-1.13480222e+00 7.41598189e-01 6.78944945e-01 -4.00224805e-01
-4.98636961e-01 6.60389066e-02 1.99937895e-01 -2.16934606e-02
6.79564476e-01 2.71510810e-01 -3.63397449e-01 2.79897690e-01
-6.65042996e-01 5.29151380e-01 8.87238860e-01 9.61129963e-01
9.51398790e-01 -1.06681086e-01 -8.91624153e-01 6.62645280e-01
4.62106854e-01 6.48492098e-01 8.22793022e-02 -8.40989053e-01
3.18511575e-01 8.36754143e-01 -2.45394841e-01 -1.11607134e+00
3.71516675e-01 -4.45394009e-01 -7.17378557e-01 2.82459468e-01
5.57294227e-02 4.72655892e-02 -1.06008601e+00 1.59214175e+00
9.41268027e-01 3.72156531e-01 -1.05748802e-01 6.97935283e-01
1.35043597e+00 7.12106407e-01 3.39179397e-01 -2.68858105e-01
9.37706530e-01 -1.28279936e+00 -3.27944726e-01 -7.78980628e-02
-2.30815232e-01 -5.77019691e-01 8.44239056e-01 3.96787375e-01
-7.99888313e-01 -3.12732369e-01 -7.51053333e-01 2.36187607e-01
-4.98912148e-02 -5.82464784e-03 6.07792675e-01 3.39047849e-01
-1.17450714e+00 5.59694588e-01 -2.30287120e-01 2.25703120e-01
1.22395706e+00 1.41479924e-01 -1.26685590e-01 -1.08289801e-01
-9.07004178e-01 1.98442712e-01 2.87312686e-01 -5.60492277e-01
-1.08548939e+00 -1.05580139e+00 -7.00624645e-01 9.08897892e-02
3.63961369e-01 -1.00463343e+00 1.07569015e+00 -1.22104406e+00
-1.33736980e+00 1.07126808e+00 -3.55539232e-01 -4.74365562e-01
7.24297881e-01 1.39226569e-02 -9.17150378e-02 1.67328969e-01
4.80266571e-01 9.19919014e-01 1.09473288e+00 -1.67312431e+00
-8.64154935e-01 -3.18402052e-01 -1.04169495e-01 7.19970465e-01
-4.69067087e-03 -1.73824966e-01 -4.38817173e-01 -4.66121942e-01
5.46625555e-02 -6.21671796e-01 -1.45348623e-01 -7.10345358e-02
-8.06322813e-01 -9.00748253e-01 8.51786375e-01 -1.74048066e-01
8.49104106e-01 -1.68269467e+00 -1.19805910e-01 1.91448182e-01
6.92712963e-01 9.65155214e-02 -8.36339369e-02 1.08720027e-01
1.09887170e-02 -1.57543346e-01 5.44070713e-02 -1.21999167e-01
-3.76435257e-02 9.07449722e-02 -4.68160361e-01 3.51865292e-01
1.30873606e-01 7.46745229e-01 -1.30753458e+00 -7.78741479e-01
-6.84690401e-02 2.74695098e-01 -6.48919702e-01 1.19104028e-01
-7.03874290e-01 3.91077220e-01 -4.60744739e-01 9.57393348e-01
9.21843529e-01 -6.92443430e-01 3.48430306e-01 -1.73290104e-01
4.53061044e-01 2.47251943e-01 -1.20213008e+00 8.64584923e-01
5.70710659e-01 4.87940758e-01 -3.86755735e-01 -9.72247958e-01
1.00416911e+00 -1.40754491e-01 4.30757374e-01 -3.98715764e-01
-4.11681794e-02 1.83363426e-02 -1.49529561e-01 -1.73990026e-01
3.86702418e-01 3.25428218e-01 2.15968072e-01 3.46527755e-01
-1.83377545e-02 -2.62728512e-01 5.20165503e-01 5.52918673e-01
1.12470436e+00 9.34148282e-02 2.71654785e-01 -2.94109821e-01
4.93711680e-01 1.69728294e-01 1.03755116e+00 9.59574342e-01
-5.33978760e-01 6.81634486e-01 9.31169808e-01 -3.12940806e-01
-6.23052299e-01 -1.37753832e+00 3.20835292e-01 1.29640532e+00
5.85728049e-01 -4.27044451e-01 -3.98138493e-01 -1.31704497e+00
4.10217226e-01 5.35546899e-01 -6.40153170e-01 1.05032519e-01
-2.89396971e-01 -4.82540578e-01 -2.13463083e-02 3.81724596e-01
6.25630200e-01 -1.31080413e+00 -1.07991152e-01 -1.23611316e-01
-2.72132486e-01 -9.29983318e-01 -3.32468629e-01 -3.02972525e-01
-3.67389679e-01 -1.45642626e+00 -6.29128456e-01 -8.83841872e-01
1.15485561e+00 4.45082396e-01 1.77221465e+00 -7.17176497e-02
-4.27181646e-02 2.42843390e-01 -4.25114572e-01 -8.30556035e-01
-1.52102306e-01 -8.72963518e-02 -1.02481939e-01 1.96133271e-01
3.49903494e-01 -4.48434681e-01 -7.76226580e-01 1.30695373e-01
-4.23675328e-01 1.59085199e-01 4.74864304e-01 6.63673818e-01
1.24958098e+00 9.86721888e-02 5.11280358e-01 -1.37380028e+00
5.35827339e-01 -5.58425665e-01 -5.00460267e-01 5.15568733e-01
-5.09521246e-01 -2.68328190e-01 3.17676008e-01 -4.21388894e-01
-1.17899680e+00 -8.57886747e-02 4.53530401e-01 -6.05056643e-01
-4.70970273e-02 6.48991242e-02 -2.27945700e-01 -1.22690937e-02
6.20070398e-01 1.22731902e-01 -3.71589512e-01 1.23822622e-01
4.31635976e-01 2.60633796e-01 5.55580378e-01 -6.96335137e-01
8.01909029e-01 5.76843381e-01 -1.52953804e-01 -5.41038871e-01
-1.08457029e+00 -4.41493809e-01 -7.30842203e-02 -6.73894167e-01
7.58170962e-01 -1.24952149e+00 -7.11076915e-01 7.40029991e-01
-9.83367503e-01 -4.14277613e-01 -8.44088316e-01 8.91599208e-02
-3.08532059e-01 1.56576350e-01 -2.81748414e-01 -5.96701741e-01
-4.41930324e-01 -8.91303957e-01 1.00809026e+00 7.56917477e-01
-1.04024179e-01 -4.52497095e-01 2.96849549e-01 7.17566431e-01
2.17768446e-01 2.65171707e-01 6.62300110e-01 -7.27986693e-01
-1.14414954e+00 4.99241017e-02 -4.73110735e-01 3.28497022e-01
-2.61060912e-02 3.60656679e-01 -8.86644006e-01 -2.82050282e-01
-5.97353160e-01 -5.88217735e-01 1.04560959e+00 4.21546489e-01
1.54861605e+00 -2.78823733e-01 -4.66688037e-01 2.89613754e-01
1.25526869e+00 -2.76236773e-01 5.48405766e-01 -2.86324527e-02
8.75114143e-01 5.61504364e-01 7.55676925e-01 5.63413978e-01
1.06104624e+00 8.68897066e-02 5.98614752e-01 2.44522896e-02
-6.85075164e-01 -7.28029847e-01 1.28767341e-01 3.33723128e-01
1.43792778e-01 -3.84524494e-01 -8.54069054e-01 8.19260955e-01
-1.98347294e+00 -1.00193536e+00 -3.04306090e-01 1.88127220e+00
1.14508498e+00 2.11230919e-01 -8.71819779e-02 -1.37227580e-01
1.06291115e+00 5.02481282e-01 -8.60477090e-01 1.49706125e-01
-5.02752244e-01 1.04073085e-01 4.63126659e-01 2.60883719e-01
-1.12560725e+00 1.00321138e+00 6.25787687e+00 1.15866172e+00
-9.53440726e-01 -5.51372953e-02 1.21477771e+00 9.31011736e-02
-6.34502113e-01 1.46562606e-01 -9.61809635e-01 6.34742320e-01
-3.22722015e-03 -4.77512360e-01 4.57237773e-02 1.12468433e+00
-3.31057936e-01 -5.29327273e-01 -8.00701916e-01 9.68987823e-01
3.87121528e-01 -1.48259294e+00 4.37945664e-01 -2.09029213e-01
1.64616406e+00 3.25592041e-01 -3.13612819e-02 3.65850359e-01
1.14623964e+00 -8.62957478e-01 6.02805018e-01 3.56704503e-01
4.58614081e-01 -4.44002807e-01 4.51367348e-01 3.07322472e-01
-1.12823367e+00 1.57295942e-01 -3.77397209e-01 -6.66443333e-02
-1.54124007e-01 1.13884950e+00 -9.90624905e-01 4.12079722e-01
1.09741104e+00 8.54479015e-01 -5.89436531e-01 1.33346462e+00
-8.41866016e-01 7.37020552e-01 -3.18702757e-01 -2.17205897e-01
-3.33552867e-01 1.43901519e-02 6.18237019e-01 7.28725016e-01
-6.02351800e-02 -1.06589861e-01 5.86445928e-01 5.69727302e-01
-6.04273856e-01 2.03995943e-01 -3.46205115e-01 3.66883188e-01
8.32655311e-01 1.30188119e+00 -8.45415950e-01 -6.94249213e-01
1.09431800e-02 7.26027489e-01 6.36144400e-01 2.76661366e-01
-7.16378152e-01 1.19851111e-02 7.57663190e-01 -1.33099314e-02
7.72516727e-01 5.25671601e-01 -3.52458537e-01 -1.14701927e+00
1.08634211e-01 -9.86379981e-01 7.24504650e-01 -9.43692148e-01
-1.68236852e+00 2.60980964e-01 7.83532485e-02 -1.21701896e+00
2.34381735e-01 -1.74260244e-01 -9.67547297e-01 5.08366883e-01
-2.03450203e+00 -1.16546595e+00 -5.30940354e-01 6.05093062e-01
4.59775150e-01 -8.66888836e-02 3.23791921e-01 -1.17702201e-01
-3.71216446e-01 4.05723691e-01 -4.13362503e-01 -1.98730573e-01
9.01211023e-01 -1.54835653e+00 1.66428432e-01 5.89051425e-01
2.03443334e-01 -1.36135340e-01 6.87462628e-01 -1.06930172e+00
-7.89852500e-01 -1.28892231e+00 1.03702736e+00 -3.93081218e-01
3.41210991e-01 6.05060719e-02 -8.38984191e-01 3.70326847e-01
3.92519891e-01 3.86446804e-01 8.32320094e-01 -4.63416241e-03
-6.27854407e-01 -4.49609309e-01 -1.31872797e+00 5.56283772e-01
1.07613051e+00 -2.27764428e-01 -3.82663250e-01 7.29365706e-01
6.97973728e-01 -5.50183892e-01 -4.23265427e-01 7.24653482e-01
6.70748055e-02 -1.34631062e+00 7.80093133e-01 -5.14644921e-01
7.04038560e-01 -8.24287355e-01 -5.42552210e-02 -1.44915009e+00
-3.61185521e-01 -4.36698020e-01 -3.64450306e-01 1.24551368e+00
3.87834996e-01 -5.90032101e-01 1.43491542e+00 2.63725072e-01
1.05972052e-01 -9.91022170e-01 -6.98597729e-01 -3.16224933e-01
-3.13812464e-01 -2.39748821e-01 8.68824601e-01 9.69124258e-01
-2.97356278e-01 3.45285058e-01 -2.03656808e-01 2.83271164e-01
1.08801711e+00 6.20702446e-01 1.12163305e+00 -1.27396739e+00
-3.96856874e-01 -4.08140957e-01 -6.12028420e-01 -8.99656057e-01
2.52528042e-01 -8.60438228e-01 6.06869273e-02 -1.81067443e+00
6.28351748e-01 -4.72420007e-01 -2.58522302e-01 5.35303414e-01
-8.52093518e-01 4.95921880e-01 1.55656531e-01 3.31438601e-01
-1.39467049e+00 8.37756693e-01 1.41302657e+00 -5.12754083e-01
3.31048965e-02 -2.60417555e-02 -1.16050375e+00 8.81507397e-01
7.49334574e-01 -5.08292913e-01 -5.90234339e-01 -2.25179434e-01
4.40927386e-01 -4.34358090e-01 2.40241602e-01 -9.33738947e-01
4.21897024e-01 -2.08392948e-01 6.42978668e-01 -1.12755001e+00
-9.60699469e-02 -2.63321936e-01 -7.08348975e-02 3.44434887e-01
-1.99665159e-01 -3.51999760e-01 -3.01967651e-01 9.08770621e-01
-3.15411478e-01 4.01079625e-01 7.24775612e-01 -9.83566120e-02
-8.12564015e-01 6.30945921e-01 1.78004861e-01 6.18516862e-01
1.39225256e+00 -1.08197905e-01 -8.43731642e-01 -4.78064239e-01
-5.18393517e-01 8.10070693e-01 4.25474823e-01 2.33075544e-01
8.43579113e-01 -1.45075130e+00 -1.24630857e+00 1.55638978e-01
2.80950695e-01 4.70654190e-01 4.08465236e-01 4.87779349e-01
-4.60900486e-01 -3.05472493e-01 -6.24182560e-02 -9.13734853e-01
-1.58871412e+00 -2.46445574e-02 2.44039312e-01 -6.16423130e-01
-3.03592235e-01 1.63438940e+00 3.73514324e-01 -5.62977374e-01
2.28665709e-01 2.63240039e-01 -5.15960574e-01 1.24631666e-01
3.18796366e-01 4.03359652e-01 -2.01867655e-01 -5.89923501e-01
-3.80214751e-01 2.87637144e-01 -2.86519796e-01 5.44585228e-01
1.24674618e+00 2.51645714e-01 -4.54119533e-01 1.72315642e-01
5.96033931e-01 1.36329249e-01 -1.55684638e+00 -4.43272680e-01
-3.71162623e-01 -8.45270574e-01 -3.12743813e-01 -7.93087959e-01
-1.20985937e+00 1.02497853e-01 1.08414739e-01 2.40630552e-01
1.08978212e+00 5.35648286e-01 4.58749920e-01 6.39913306e-02
3.57446820e-01 -1.09139693e+00 3.99147809e-01 3.99639070e-01
8.14605176e-01 -1.36739457e+00 4.28936988e-01 -9.45109904e-01
-1.05428886e+00 4.32965577e-01 1.12827587e+00 -3.49250972e-01
6.95551753e-01 -4.17009778e-02 6.34355843e-02 -3.71083945e-01
-1.14374864e+00 -1.74143150e-01 4.10510570e-01 9.78572845e-01
1.22004360e-01 4.76182461e-01 -3.51122767e-01 2.69551098e-01
-3.10064465e-01 -1.53974354e-01 3.05764347e-01 6.95553064e-01
-4.03650939e-01 -1.03180301e+00 -2.60241538e-01 9.74049389e-01
-3.93786252e-01 -7.89198354e-02 -5.67175627e-01 1.53588966e-01
4.42163885e-01 9.09003854e-01 1.77422032e-01 -2.52327591e-01
1.70625597e-01 -4.33072597e-01 4.70173150e-01 -6.76791906e-01
-4.81569350e-01 -2.89764643e-01 1.41735375e-01 -5.77129602e-01
-6.79728150e-01 -7.39126682e-01 -1.08476079e+00 -2.33984485e-01
-3.19424480e-01 1.23300388e-01 2.12546840e-01 5.75485647e-01
4.40355480e-01 7.91868687e-01 8.63510251e-01 -7.00563788e-01
-3.96954417e-01 -6.02312148e-01 -5.56629479e-01 6.29837930e-01
-4.69634347e-02 -4.73577619e-01 -4.09738362e-01 -2.84038261e-02] | [10.235395431518555, 1.8018825054168701] |
2d70c0d5-c26e-4d6b-9d58-0a89a958a6f0 | knowledge-graph-papers-iclr-2021 | null | null | https://openreview.net/forum?id=2P7cGsM14Mj | https://openreview.net/pdf?id=2P7cGsM14Mj | Knowledge Graph Papers @ ICLR 2021 | This post aims at providing an overview of ICLR 2021 papers focusing on knowledge graphs (KGs). In particular, we highlight the research in four wide areas: complex query answering and reasoning in KGs, temporal logics and KGs, NLP point of view and entity linking, multimodal question answering with KGs. We hope this post could be of use for researchers and practitioners from adjacent areas to demonstrate the benefits of employing knowledge graphs in those various domains solving important problems. | ['Anonymous'] | 2022-01-17 | null | null | null | iclr-track-blog-2022-5 | ['complex-query-answering'] | ['knowledge-base'] | [-4.90500689e-01 8.73377264e-01 -5.67186832e-01 -2.19881326e-01
-6.11005962e-01 -8.89725208e-01 4.96317297e-01 7.27847576e-01
-1.18313596e-01 1.12568009e+00 3.14463139e-01 -4.05651331e-01
-9.01975811e-01 -1.13928819e+00 -6.18248582e-01 2.14229152e-01
-3.71900976e-01 8.62285376e-01 7.45337188e-01 -5.00930965e-01
7.48922229e-02 5.84495246e-01 -1.58366394e+00 3.09386045e-01
6.26911938e-01 9.77238536e-01 -4.72183943e-01 3.64707738e-01
-3.80326867e-01 1.44763148e+00 -3.61054569e-01 -9.07776117e-01
-2.96027839e-01 3.23996335e-01 -1.71096194e+00 -6.76216066e-01
4.00092959e-01 3.32253039e-01 -5.59580684e-01 1.07157254e+00
4.37227130e-01 5.72852612e-01 2.20904917e-01 -1.37772775e+00
-9.01811242e-01 8.46778452e-01 8.23716894e-02 2.28961051e-01
1.49185526e+00 -3.82383466e-01 1.27915156e+00 -3.58157754e-01
1.24324095e+00 1.60011709e+00 5.49112737e-01 2.04375505e-01
-5.95349133e-01 -1.70196071e-01 9.37872529e-02 1.00289166e+00
-1.56846464e+00 -3.70952606e-01 3.21730286e-01 -1.44029945e-01
1.72929394e+00 4.97409195e-01 6.07085228e-01 4.98275518e-01
-1.69328496e-01 9.11333978e-01 7.14128792e-01 -7.66186178e-01
5.54099083e-02 3.11477423e-01 5.93086779e-01 8.45036387e-01
5.20648777e-01 -1.28672998e-02 -1.04571581e+00 -3.75527978e-01
3.62675250e-01 -7.97156751e-01 -2.01048970e-01 -3.64389807e-01
-1.08007646e+00 6.43016517e-01 1.68109462e-01 4.51879084e-01
-4.07370210e-01 2.67385334e-01 6.03675365e-01 5.08935750e-01
1.67078990e-02 5.69608986e-01 -8.85079384e-01 -1.23690642e-01
-3.85026336e-01 6.63577199e-01 1.44679940e+00 1.50940704e+00
6.03382647e-01 -3.35201532e-01 -1.52855903e-01 3.96897674e-01
7.30979323e-01 5.42994738e-01 -1.38805315e-01 -1.25427997e+00
5.37613213e-01 9.70310271e-01 3.46186191e-01 -1.03361142e+00
-4.57741171e-01 2.71896511e-01 -7.23984540e-02 -6.58168077e-01
5.58249578e-02 4.91615236e-02 -5.33994377e-01 1.44410801e+00
6.22092426e-01 -2.14013293e-01 8.35520089e-01 3.65776986e-01
1.52409422e+00 4.07073766e-01 7.19359100e-01 -3.40423226e-01
1.77506900e+00 -5.89718163e-01 -1.23991799e+00 -3.76450084e-02
9.03698206e-01 -5.22389948e-01 5.59881270e-01 3.20384614e-02
-1.21967196e+00 2.28263792e-02 -6.49083138e-01 -4.30102825e-01
-1.43474531e+00 -1.88015059e-01 1.14421606e+00 2.92704731e-01
-1.31626034e+00 1.35014609e-01 -5.11449754e-01 -9.97158766e-01
1.24215096e-01 5.99683702e-01 -2.92942971e-01 -3.02814841e-01
-2.10233855e+00 1.44591510e+00 1.29022300e+00 6.81346878e-02
-2.04084009e-01 -7.12891638e-01 -1.13990664e+00 -1.47094771e-01
9.14660692e-01 -1.34427524e+00 9.96166408e-01 1.60054624e-01
-9.59693611e-01 1.04294860e+00 -2.66879797e-01 -6.74495518e-01
-1.94973797e-02 -4.93079036e-01 -1.20139909e+00 1.36911437e-01
1.85336754e-01 3.76561105e-01 6.44506216e-02 -6.72117114e-01
-5.06029010e-01 -6.97111309e-01 4.59208935e-01 3.59560519e-01
1.65557414e-01 4.05356348e-01 -7.16396630e-01 -4.92057428e-02
-1.02720045e-01 -4.25790191e-01 1.97677881e-01 -5.03191292e-01
-4.80210662e-01 -8.02448511e-01 1.02345419e+00 -7.95489788e-01
1.55000389e+00 -1.66251838e+00 2.51100272e-01 3.46902728e-01
-6.62697032e-02 2.00083226e-01 3.49621564e-01 1.28133345e+00
2.48786107e-01 2.42841274e-01 4.00978804e-01 7.11562812e-01
4.61121798e-01 4.86092776e-01 -4.67722595e-01 -2.88391978e-01
-9.32609215e-02 1.75042868e+00 -1.00220191e+00 -1.08210409e+00
1.49726093e-01 1.38634250e-01 3.09207141e-02 -1.21161051e-01
-8.38764906e-01 -5.13151214e-02 -8.30572248e-01 1.23757112e+00
6.18669540e-02 -3.14096868e-01 6.21668220e-01 -5.51873863e-01
1.56712845e-01 -2.83789970e-02 -1.25766945e+00 1.83425438e+00
4.53774668e-02 4.51439619e-01 -3.49826336e-01 -6.57859504e-01
6.27227306e-01 7.81760931e-01 8.04469585e-02 -7.60933876e-01
-3.57072979e-01 1.67836681e-01 -5.64504027e-01 -1.24378395e+00
6.96843624e-01 -1.98032670e-02 -2.93423593e-01 5.89058362e-03
8.76534209e-02 -2.68759638e-01 4.57107127e-01 6.66187763e-01
9.96906102e-01 3.66308570e-01 6.20273650e-01 -1.77176774e-01
8.53428602e-01 4.50527430e-01 1.02524936e-01 6.58609331e-01
-1.72263697e-01 -5.68227112e-01 3.12986910e-01 -5.73742986e-01
-4.82228041e-01 -9.31447327e-01 -7.90964812e-02 1.08794796e+00
2.75219887e-01 -7.14226484e-01 -2.39738107e-01 -6.85086310e-01
3.07700098e-01 9.31933522e-01 -4.08151209e-01 -1.12746976e-01
-4.45021659e-01 -3.79464686e-01 1.04261267e+00 7.17431962e-01
5.28116584e-01 -1.50162899e+00 -4.01836932e-01 2.05190793e-01
-4.76311833e-01 -1.41422725e+00 4.45201010e-01 -1.78748786e-01
-8.88384342e-01 -1.58517361e+00 -2.87416042e-04 -7.99275279e-01
6.22354038e-02 -3.63382936e-01 1.64034235e+00 -3.16174366e-02
-3.57403159e-01 1.24494588e+00 -4.67999160e-01 -6.76235437e-01
1.89383641e-01 4.50976677e-02 -1.60919711e-01 -7.74257123e-01
8.98910224e-01 -3.85366857e-01 -2.69587606e-01 5.02607301e-02
-9.59580898e-01 -4.87848163e-01 8.21490213e-02 8.39982703e-02
4.88177717e-01 4.02080059e-01 6.77425444e-01 -1.19145489e+00
8.40442419e-01 -6.29778147e-01 -5.76163054e-01 1.33088529e+00
-7.03471422e-01 2.26263747e-01 -2.84698784e-01 2.96647221e-01
-1.19458425e+00 -3.52111667e-01 1.98386252e-01 1.20041519e-01
-2.09365919e-01 9.93000329e-01 -2.97693014e-01 -2.04457909e-01
5.64796567e-01 -7.08762780e-02 -5.78758717e-01 -2.13565901e-01
9.62473810e-01 1.60610855e-01 5.46431243e-01 -1.09159124e+00
4.29053664e-01 2.96042979e-01 2.07455248e-01 -4.95225310e-01
-7.55386353e-01 -7.90135443e-01 -5.27459681e-01 -2.86189944e-01
8.19691479e-01 -7.50786006e-01 -1.21925056e+00 3.33136432e-02
-1.17512584e+00 2.61463616e-02 -7.18621314e-01 1.81543857e-01
-5.95860422e-01 3.53803992e-01 -5.23706555e-01 -1.03021431e+00
-4.90279227e-01 -2.37744465e-01 1.12785828e+00 5.15331507e-01
-3.07080179e-01 -1.45814431e+00 1.19230248e-01 8.95140052e-01
2.25732684e-01 3.77329767e-01 1.38278949e+00 -7.81600654e-01
-5.64896226e-01 -3.29687923e-01 -3.28970701e-01 -3.67973179e-01
-1.77331567e-01 1.23342127e-01 -6.29355788e-01 2.14357629e-01
-7.66742289e-01 -6.84717000e-01 1.98931009e-01 6.94991872e-02
5.59905171e-01 -2.42427692e-01 -7.05952704e-01 -2.49985129e-01
1.84120238e+00 6.39564216e-01 9.73871827e-01 6.11247838e-01
3.96247268e-01 9.45590436e-01 9.54249799e-01 5.06911390e-02
9.73350525e-01 6.80897951e-01 2.41965279e-01 6.56297743e-01
-8.06566626e-02 -3.20942789e-01 -2.37994254e-01 7.01018333e-01
-5.45117021e-01 -3.30907702e-01 -1.41549802e+00 7.78181374e-01
-2.46592331e+00 -1.04682732e+00 -2.50217348e-01 1.52754879e+00
8.60806227e-01 -4.89212275e-01 -2.00491384e-01 -8.74916613e-02
6.76514268e-01 -3.64954509e-02 -4.52590317e-01 -3.36891145e-01
-5.78692913e-01 1.53557077e-01 2.57653534e-01 5.61807811e-01
-9.40054595e-01 1.53581607e+00 7.29964209e+00 7.94408619e-01
-7.61229172e-02 -1.17058203e-01 -2.26678655e-01 5.25602818e-01
-4.14056182e-01 4.12218809e-01 -9.95287001e-01 -2.37605542e-01
1.20155370e+00 -5.25661409e-01 5.54184914e-01 5.01500368e-01
-2.13433400e-01 -2.93414980e-01 -9.90410149e-01 8.40720534e-01
-1.35697499e-01 -1.64565480e+00 2.72868872e-01 -2.79154539e-01
4.92904186e-01 -2.42148221e-01 -4.35835689e-01 6.28102899e-01
7.15814412e-01 -1.02243519e+00 4.81265634e-01 1.17516625e+00
2.84418374e-01 -5.09954274e-01 8.80652905e-01 4.40955199e-02
-1.38588059e+00 2.39145048e-02 -1.86622947e-01 3.15937638e-01
4.60842758e-01 2.81603158e-01 -6.99687183e-01 1.78796756e+00
9.05279517e-01 5.77944756e-01 -5.68087518e-01 9.17485774e-01
-2.53824472e-01 1.63145643e-02 -3.14040095e-01 -6.78727962e-03
-8.36623535e-02 -4.14602496e-02 3.50033671e-01 1.30067909e+00
-3.93753015e-02 5.20182014e-01 -2.42227152e-01 5.02154350e-01
5.61921783e-02 3.05638015e-01 -9.85168219e-01 -3.77868563e-01
7.09086597e-01 6.80587292e-01 -3.29616576e-01 -6.19395971e-01
-3.85252893e-01 4.68197495e-01 3.76805991e-01 6.98731065e-01
-5.66542625e-01 -6.24845982e-01 8.13410506e-02 -1.20189451e-01
2.04578042e-01 -8.96585211e-02 2.19667614e-01 -1.10085082e+00
3.78890336e-02 -8.07907879e-01 1.45973706e+00 -1.24844646e+00
-1.15139854e+00 3.67913157e-01 6.66393876e-01 -3.07815939e-01
-3.86181235e-01 -3.98160189e-01 1.61689878e-01 4.90605026e-01
-1.62700045e+00 -1.57568550e+00 -1.43160194e-01 1.09862053e+00
-2.91162521e-01 4.87503000e-02 1.10458612e+00 6.62578046e-01
-2.29365021e-01 1.85530484e-02 -5.52248597e-01 -1.08790495e-01
5.53196371e-01 -1.28030205e+00 -5.83853107e-03 2.23396882e-01
3.33139449e-01 8.42268646e-01 5.20741820e-01 -1.04748118e+00
-1.86152387e+00 -8.98155212e-01 1.38231361e+00 -8.48008633e-01
9.04493392e-01 3.58963609e-01 -7.21770167e-01 1.38162637e+00
2.93844819e-01 -1.10431621e-02 6.75532758e-01 4.37335491e-01
-5.18332124e-01 -1.08451452e-02 -1.27914929e+00 2.31080741e-01
1.28598726e+00 -9.80172753e-01 -1.00563121e+00 7.61996806e-01
8.98715377e-01 -4.98773992e-01 -1.63292646e+00 8.79024744e-01
4.99550641e-01 -4.59260046e-01 1.27795172e+00 -1.35292923e+00
-5.35776198e-01 -5.83645463e-01 -5.15418828e-01 -4.49150711e-01
3.38824317e-02 -6.64066374e-01 -9.01730418e-01 1.29038393e+00
5.64578533e-01 -7.76939571e-01 6.37409210e-01 1.11708844e+00
1.26104251e-01 -5.52701652e-01 -1.04143429e+00 -7.52324164e-01
-4.45296615e-01 -7.36969292e-01 7.14347661e-01 1.01734221e+00
4.66639102e-01 3.99145037e-01 3.67902741e-02 4.35401917e-01
3.43073040e-01 2.23208874e-01 2.72642761e-01 -1.34514380e+00
1.43838674e-01 7.73310587e-02 -7.85088420e-01 -5.03292680e-01
2.47403204e-01 -6.28179193e-01 -6.43603027e-01 -2.23700690e+00
-2.11714432e-01 -3.07339698e-01 -1.58172339e-01 8.04765403e-01
2.82653362e-01 -1.49865672e-01 -1.06874797e-02 1.08958028e-01
-1.40548229e+00 1.34809449e-01 1.24877262e+00 -2.00104728e-01
-1.28203668e-02 -3.50480020e-01 -6.80858910e-01 5.88438928e-01
3.83526862e-01 -1.63532883e-01 -7.50685871e-01 -2.57703692e-01
1.11981237e+00 3.04568321e-01 4.70330805e-01 -5.70772648e-01
9.41700041e-01 -3.66571158e-01 -2.66838789e-01 -6.86991334e-01
4.45245713e-01 -1.00327194e+00 7.77006030e-01 2.36706913e-01
-2.02887077e-02 9.49989259e-02 5.81503808e-01 6.85210407e-01
-7.32836187e-01 -2.99949616e-01 -3.70463841e-02 -4.17658657e-01
-1.53980494e+00 3.56870368e-02 6.00702390e-02 5.62208354e-01
1.33249307e+00 -6.41868263e-02 -7.16249347e-01 -3.29424143e-01
-1.18420982e+00 7.62316704e-01 -7.23593608e-02 2.79652357e-01
5.11282861e-01 -1.28530240e+00 -1.73655182e-01 -5.45697331e-01
4.01552498e-01 -2.15786681e-01 2.28745759e-01 7.42183268e-01
-6.60705805e-01 1.39232838e+00 1.91253334e-01 6.89457208e-02
-1.16823661e+00 7.08302617e-01 1.27510756e-01 -4.07254308e-01
-4.26621079e-01 6.80314541e-01 -6.45040512e-01 -5.90025783e-01
5.51493883e-01 -6.72498271e-02 -7.18423724e-01 2.59773284e-01
2.21004561e-01 6.39580190e-01 1.16292223e-01 -3.27187985e-01
-9.41271245e-01 6.50922358e-01 1.48338363e-01 -9.53048989e-02
1.15120816e+00 -2.81539202e-01 -7.29992688e-01 6.00232899e-01
6.45976603e-01 -8.09363276e-02 1.82580054e-01 -3.44314814e-01
8.65698993e-01 9.56579298e-02 -4.07343149e-01 -1.35410666e+00
-5.74004352e-01 1.00609913e-01 2.81351894e-01 5.89995146e-01
8.65317822e-01 7.81178832e-01 7.34122157e-01 9.74165320e-01
6.97841465e-01 -1.27565467e+00 -5.52977204e-01 4.50163424e-01
1.02975130e+00 -9.15068746e-01 1.89703956e-01 -6.78491652e-01
-6.90285563e-01 1.10851705e+00 4.19357687e-01 4.80194211e-01
7.24777758e-01 1.91680416e-02 -2.19827555e-02 -1.27048779e+00
-8.78420591e-01 -4.89628613e-01 4.36757803e-01 8.63093555e-01
2.04678908e-01 8.61552507e-02 -1.68180570e-01 2.09269583e-01
-6.41157925e-02 3.95919889e-01 -1.08227752e-01 1.30447507e+00
-2.01276079e-01 -1.47232735e+00 -1.02314845e-01 2.40928411e-01
-4.46700335e-01 -2.41033196e-01 -8.21155131e-01 1.33627462e+00
1.15191534e-01 9.49164867e-01 -4.66617227e-01 -2.19825096e-02
9.35555220e-01 5.61225832e-01 6.65395379e-01 -6.09046876e-01
-5.42132258e-01 -8.58782709e-01 1.01069176e+00 -6.99308872e-01
-1.05139470e+00 -5.40001214e-01 -1.38184154e+00 -3.51371557e-01
-4.20109272e-01 6.89481676e-01 3.91754955e-01 8.17813814e-01
5.27143359e-01 1.81000054e-01 -6.24089777e-01 5.22174418e-01
9.86490622e-02 -5.22225618e-01 -6.46459699e-01 4.67017367e-02
-3.59195292e-01 -6.44138157e-01 2.22864464e-01 1.81805015e-01] | [9.551115989685059, 8.019603729248047] |
94d7c04b-ed36-4b60-96cc-9d48e570ce8c | contrastive-entity-linkage-mining-variational | null | null | https://openreview.net/forum?id=fR44nF03Rb | https://openreview.net/pdf?id=fR44nF03Rb | Contrastive Entity Linkage: Mining Variational Attributes from Large Catalogs for Entity Linkage | Presence of near identical, but distinct, entities called entity variations makes the task of data integration challenging. For example, in the domain of grocery products, variations share the same value for attributes such as brand, manufacturer and product line, but differ in other attributes, called variational attributes, such as package size and color. Identifying variations across data sources is an important task in itself and is crucial for identifying duplicates. However, this task is challenging as the variational attributes are often present as a part of unstructured text and are domain dependent. In this work, we propose our approach, Contrastive entity linkage, to identify both entity pairs that are the same and pairs that are variations of each other. We propose a novel unsupervised approach, VarSpot, to mine domain-dependent variational attributes present in unstructured text. The proposed approach reasons about both similarities and differences between entities and can easily scale to large sources containing millions of entities. We show the generality of our approach by performing experimental evaluation on three different domains. Our approach significantly outperforms state-of-the-art learning-based and rule-based entity linkage systems by up to 4% F1 score when identifying duplicates, and up to 41% when identifying entity variations. | ['Lise Getoor', 'Christos Faloutsos', 'Xin Luna Dong', 'Hao Wei', 'Bunyamin Sisman', 'Varun Embar'] | 2020-02-14 | null | null | null | akbc-2020-6 | ['data-integration'] | ['knowledge-base'] | [-2.97329664e-01 -7.03278705e-02 -2.84368515e-01 -4.95305389e-01
-7.24541128e-01 -9.79527950e-01 4.11994636e-01 1.04583097e+00
-1.96134105e-01 8.86431873e-01 -9.88644809e-02 2.03902796e-01
-3.32175672e-01 -9.35496926e-01 -1.00196624e+00 -2.65074193e-01
1.71499443e-03 1.03533137e+00 4.51064169e-01 -2.46550456e-01
1.14282243e-01 3.11982781e-01 -1.66775775e+00 3.02821934e-01
1.14823222e+00 7.84988821e-01 -2.50396207e-02 -2.05879062e-02
-8.90419364e-01 5.50162733e-01 -7.75153935e-01 -9.96498346e-01
2.14532375e-01 -8.54289308e-02 -6.38157845e-01 -4.16450590e-01
5.60522974e-01 2.78577805e-01 2.17804894e-01 1.22082245e+00
3.34099770e-01 -1.57309875e-01 7.93526649e-01 -1.67592454e+00
-7.81085014e-01 9.88438964e-01 -7.36897886e-01 -2.52409279e-03
4.90065038e-01 -5.97771525e-01 1.28411984e+00 -8.76407683e-01
1.04603779e+00 1.17208123e+00 8.39445889e-01 2.50281710e-02
-1.18031132e+00 -8.40891659e-01 1.94827363e-01 1.40346736e-01
-1.65972650e+00 -4.33151931e-01 5.54923177e-01 -6.42241418e-01
8.41856897e-01 1.75976783e-01 1.05291888e-01 6.56050622e-01
-5.62719628e-03 5.92867732e-01 5.01566768e-01 -3.63518685e-01
3.48240852e-01 4.09403861e-01 2.03369930e-01 2.45169565e-01
9.02397692e-01 -7.17653930e-01 -5.02573967e-01 -6.16734266e-01
3.15032393e-01 2.76327841e-02 1.55383497e-02 -6.39566064e-01
-1.06163609e+00 7.39415467e-01 1.01111807e-01 3.38312447e-01
-3.48068774e-01 -1.94342971e-01 3.65637511e-01 2.03053087e-01
4.29729432e-01 6.28505826e-01 -8.28234971e-01 -1.31292626e-01
-6.31698906e-01 5.45385242e-01 1.25683522e+00 1.65781593e+00
1.06419861e+00 -6.13673627e-01 -3.19376364e-02 1.14123511e+00
1.62347421e-01 4.90142316e-01 2.89132297e-01 -8.63967299e-01
8.13652873e-01 1.10666680e+00 4.48683441e-01 -1.20737696e+00
-4.21352267e-01 -3.45111266e-02 -6.06085002e-01 -2.77773440e-01
3.20242643e-01 -2.75102317e-01 -5.74138045e-01 1.84640932e+00
6.64540827e-01 -3.41091938e-02 2.52735138e-01 4.54998583e-01
1.22295821e+00 3.28711182e-01 2.36484364e-01 -1.64431646e-01
1.47377431e+00 -3.73347759e-01 -8.44333827e-01 -1.10598937e-01
6.60585701e-01 -9.67952251e-01 2.41482899e-01 -1.30438954e-01
-9.06270385e-01 -2.83217460e-01 -7.77580440e-01 -6.49907589e-02
-8.89312983e-01 8.32567811e-02 4.79873180e-01 4.48829114e-01
-3.25759262e-01 7.17786789e-01 -3.85136753e-01 -6.08575940e-01
7.76065439e-02 4.73929763e-01 -5.97334027e-01 5.83871230e-02
-1.44599712e+00 6.97248578e-01 5.77274144e-01 -4.51855183e-01
3.22552800e-01 -1.15695643e+00 -1.04169142e+00 7.19910040e-02
6.53410375e-01 -7.57594824e-01 9.98585343e-01 -6.70186281e-01
-6.61502123e-01 8.73781025e-01 -3.50538880e-01 -2.34833315e-01
3.74144644e-01 -3.42113644e-01 -1.03108358e+00 -2.99721271e-01
6.25645876e-01 2.84297854e-01 3.44267309e-01 -1.29677558e+00
-1.04413712e+00 -4.98700887e-01 -6.89263130e-03 -6.69672564e-02
-1.47175923e-01 1.78709432e-01 -7.36192107e-01 -5.83723605e-01
1.05473608e-01 -8.91744673e-01 9.47337896e-02 -5.23013175e-02
-4.26715463e-01 -3.98522437e-01 6.88507855e-01 -6.53989553e-01
1.42625260e+00 -1.99987221e+00 -1.07834809e-01 4.15603518e-01
2.72959113e-01 1.29624143e-01 2.57822007e-01 6.36885226e-01
-2.86279316e-03 2.54762024e-01 -5.07211685e-01 -5.21670766e-02
2.60218292e-01 2.01744318e-01 -3.19661722e-02 1.17979169e-01
5.13259649e-01 7.48939395e-01 -9.33550537e-01 -8.27958584e-01
-1.91678286e-01 2.65420228e-01 -8.99427459e-02 -3.00650895e-02
-2.48826817e-01 -4.71295714e-02 -4.66171294e-01 8.10255647e-01
9.66662765e-01 -1.74821138e-01 5.04813552e-01 -4.12229687e-01
-1.09333247e-01 9.53742936e-02 -1.83943760e+00 1.47546029e+00
-1.94027886e-01 4.06363338e-01 -1.07941441e-02 -8.08534503e-01
1.19638932e+00 8.66563246e-02 7.02043056e-01 -5.59541762e-01
-1.00949578e-01 4.90655094e-01 -1.92814663e-01 -4.70290899e-01
9.89922106e-01 2.30884284e-01 -4.70279723e-01 1.84848651e-01
-3.67372185e-02 2.43259385e-01 7.89888740e-01 3.64488989e-01
1.13449085e+00 -2.78450307e-02 6.24463379e-01 -1.73964038e-01
4.55999136e-01 1.29012287e-01 1.17769313e+00 5.42695582e-01
1.83062240e-01 4.70750719e-01 6.62434280e-01 -9.14971754e-02
-1.12353086e+00 -1.11476886e+00 -3.30988556e-01 8.14487278e-01
4.53545839e-01 -7.69288301e-01 -3.22299540e-01 -7.67454088e-01
8.09762537e-01 6.17762744e-01 -5.40545523e-01 2.63895273e-01
-5.75400114e-01 -7.15888262e-01 4.83668894e-01 6.32277906e-01
2.49414638e-01 -7.05861807e-01 -1.72498468e-02 5.55605412e-01
-2.93498576e-01 -1.27066350e+00 -1.71124101e-01 9.70343053e-02
-5.70834935e-01 -1.25149667e+00 -5.97583354e-01 -6.97144926e-01
4.89877999e-01 -1.31677330e-01 1.57927704e+00 -6.78015575e-02
-3.02466452e-01 7.66664296e-02 -3.03530455e-01 -6.29550636e-01
-4.27751452e-01 2.34894425e-01 1.15097672e-01 -5.58772050e-02
8.28915179e-01 -4.27328378e-01 -1.72352105e-01 4.83096093e-01
-7.79032886e-01 -6.66507006e-01 5.64892054e-01 4.28033113e-01
8.06002259e-01 1.64299786e-01 6.71612442e-01 -1.57299042e+00
4.47519094e-01 -1.02541900e+00 -5.78741193e-01 7.64416277e-01
-6.34148180e-01 3.06451112e-01 5.62964380e-01 -3.23541969e-01
-1.11055291e+00 2.49294788e-01 2.39617273e-01 -1.50803998e-01
-1.19097106e-01 5.23196578e-01 -5.19371331e-01 2.97837943e-01
4.82886940e-01 -3.06082547e-01 -3.32774460e-01 -7.06030846e-01
3.92583221e-01 7.57448137e-01 6.72669947e-01 -7.26414979e-01
7.82342851e-01 3.33615810e-01 -6.07150085e-02 -6.09873772e-01
-4.14033175e-01 -8.13249469e-01 -8.54785264e-01 3.20454240e-01
5.16600251e-01 -1.20047212e+00 -4.54934806e-01 3.66608053e-01
-1.17145920e+00 5.40407538e-01 -2.46473595e-01 1.08998314e-01
-1.05219088e-01 3.58531207e-01 -3.62865716e-01 -4.73513812e-01
-2.38503203e-01 -7.18700469e-01 9.83993709e-01 3.44879836e-01
-5.32479823e-01 -9.25618172e-01 2.26557836e-01 6.25537336e-02
1.72058210e-01 5.12261331e-01 1.10310817e+00 -1.25940073e+00
-3.57817441e-01 -2.24610925e-01 -1.77384302e-01 -2.91982859e-01
5.73646665e-01 4.18335527e-01 -4.33230340e-01 6.43813759e-02
-7.17012227e-01 1.92892238e-01 4.95517135e-01 5.71619645e-02
5.71300685e-01 -3.78386974e-02 -9.19922292e-01 2.88159668e-01
1.45631266e+00 1.64017946e-01 4.62465942e-01 3.78067076e-01
8.17072928e-01 9.71900523e-01 9.56168532e-01 6.06720090e-01
6.57393336e-01 9.69277620e-01 -2.87750829e-02 3.29872444e-02
1.60665914e-01 -7.31929243e-02 -1.13229260e-01 5.30617177e-01
1.95746124e-01 -3.44023913e-01 -9.21692967e-01 8.91723275e-01
-2.10340142e+00 -1.01737940e+00 -7.30071485e-01 2.23201013e+00
1.03570199e+00 -9.18594822e-02 2.14383304e-01 -1.76226556e-01
1.24182999e+00 -3.50633651e-01 -7.19267666e-01 -2.93924540e-01
-3.84375572e-01 -3.49726975e-02 7.07150459e-01 4.32980955e-02
-1.21328771e+00 7.68895328e-01 5.59528828e+00 6.36844397e-01
-6.06337965e-01 -1.80387348e-01 1.51092187e-01 1.01865023e-01
-4.93378192e-01 -1.60289332e-02 -1.21645463e+00 7.52167344e-01
7.17773080e-01 -5.00925779e-01 -1.72707170e-01 8.59378517e-01
-3.67827892e-01 -2.50444822e-02 -1.28718257e+00 9.69823718e-01
4.67077047e-02 -1.20778716e+00 -3.58350947e-02 -1.38044253e-01
8.21498871e-01 -2.45099068e-01 -3.16747785e-01 1.32022053e-01
5.89187086e-01 -6.19897366e-01 6.16851211e-01 3.84848833e-01
5.11927366e-01 -9.70295668e-01 8.96488249e-01 -3.90052609e-02
-1.60760796e+00 3.13215889e-02 -5.24814427e-01 5.02940357e-01
9.02238041e-02 1.00669265e+00 -5.80372810e-01 1.04592013e+00
1.04290736e+00 9.06319618e-01 -5.39716601e-01 1.14956379e+00
-1.93189997e-02 -5.54584116e-02 -4.15170997e-01 6.51489273e-02
-3.62700045e-01 -1.29600853e-01 3.93599659e-01 1.34945321e+00
4.72874671e-01 1.26732767e-01 -1.11144349e-01 9.93292034e-01
-6.47729516e-01 4.12150294e-01 -5.53066075e-01 2.27008737e-03
1.21586347e+00 1.21627402e+00 -4.95175034e-01 -4.19614404e-01
-4.56606865e-01 8.90606046e-01 3.56272846e-01 2.06777483e-01
-7.83946753e-01 -8.55760992e-01 1.02734661e+00 1.95934288e-02
7.10784435e-01 1.69651732e-01 -1.96905792e-01 -1.00623369e+00
5.43240309e-01 -7.06956506e-01 6.65095508e-01 -3.58832151e-01
-1.72890949e+00 3.73768449e-01 2.04172125e-03 -1.39592648e+00
-3.93398523e-01 -2.56113768e-01 -4.11219656e-01 9.26285863e-01
-1.36013484e+00 -8.91513884e-01 -4.23516124e-01 4.88228798e-01
9.41887498e-02 -2.30759978e-01 6.91522539e-01 6.54449761e-01
-5.66508710e-01 8.20508897e-01 7.69189835e-01 3.29976767e-01
1.36016464e+00 -1.39669144e+00 7.73083746e-01 5.44852018e-01
2.05490366e-01 9.29155350e-01 7.36049831e-01 -1.05956244e+00
-1.26703107e+00 -1.23093033e+00 1.40933204e+00 -6.74844027e-01
6.42415822e-01 -3.36249202e-01 -1.15562201e+00 6.17531359e-01
-5.55713102e-02 -6.72570691e-02 9.36955333e-01 3.97748053e-01
-6.02086902e-01 -1.48866802e-01 -1.48720753e+00 1.67149052e-01
1.09686577e+00 -2.89467037e-01 -8.33558083e-01 2.07683474e-01
6.40435815e-01 -4.01700079e-01 -1.40350974e+00 3.30861181e-01
4.74659890e-01 -6.78818583e-01 1.03010941e+00 -5.12312889e-01
3.20145130e-01 -5.51435411e-01 -2.42811307e-01 -1.29895508e+00
-2.44211048e-01 -3.67614537e-01 -2.03349650e-01 2.24084425e+00
8.04568827e-01 -6.18594170e-01 5.95560730e-01 1.12822914e+00
1.96100324e-01 -1.78195372e-01 -7.40476251e-01 -1.15870500e+00
5.52906655e-02 -3.10124387e-03 1.16548419e+00 1.41254127e+00
-1.83846839e-02 2.25264698e-01 2.09071543e-02 3.20859134e-01
6.37438774e-01 4.36635226e-01 7.75474429e-01 -1.94974351e+00
6.71787560e-03 -2.24326521e-01 -6.29262984e-01 -5.43498456e-01
2.69900262e-01 -7.23695576e-01 -1.43838510e-01 -1.61139309e+00
1.08732700e-01 -9.92721260e-01 -1.23570316e-01 4.04687434e-01
-3.29361379e-01 -1.64693281e-01 5.27514778e-02 3.61805022e-01
-6.44850433e-01 1.46859214e-01 2.87352383e-01 -3.01831990e-01
-3.82896096e-01 -1.48550704e-01 -7.72062600e-01 5.88781476e-01
4.77660626e-01 -8.00271213e-01 -3.36431228e-02 -3.81831020e-01
5.46067119e-01 -3.26626867e-01 -1.32240459e-01 -7.72356570e-01
4.63192195e-01 -3.21108550e-02 4.36634809e-01 -6.42964542e-01
9.59729552e-02 -9.77980614e-01 6.22097313e-01 -1.21561326e-02
-1.34043142e-01 3.44500542e-01 2.32549161e-01 6.03504896e-01
-3.70290339e-01 -3.67767781e-01 2.16258273e-01 -5.03305793e-02
-1.02680540e+00 8.28236043e-02 1.99491993e-01 5.10406017e-01
1.13590932e+00 -4.67839539e-02 -5.97950220e-01 5.15839085e-02
-4.23049450e-01 4.33986932e-01 8.40582907e-01 7.00887740e-01
2.28552029e-01 -1.38983643e+00 -8.94522905e-01 -2.43044913e-01
6.40015483e-01 -6.37439042e-02 2.14920472e-02 3.87057304e-01
-2.53053635e-01 1.03298940e-01 -2.07677349e-01 -4.74095881e-01
-1.47975266e+00 6.43877685e-01 -1.31634846e-02 -1.18890680e-01
-3.35401088e-01 6.10553682e-01 -3.12396996e-02 -6.51703537e-01
1.10497899e-01 -1.68301195e-01 -3.49245250e-01 6.27694488e-01
3.98788631e-01 4.66357470e-01 2.65177041e-01 -7.05421150e-01
-7.72274256e-01 8.44889283e-01 -2.33799860e-01 3.42965722e-01
1.30504346e+00 -1.61738336e-01 -3.25869739e-01 5.34575582e-01
1.12164080e+00 3.16830724e-01 -5.95305026e-01 -5.10971427e-01
5.21155298e-01 -5.14552593e-01 -6.17318749e-01 -8.39672923e-01
-1.07583344e+00 1.89824134e-01 3.60482663e-01 3.42518389e-01
7.94389725e-01 3.70065570e-01 7.84927845e-01 5.13383210e-01
4.44772273e-01 -1.22506475e+00 -6.46003544e-01 3.73900741e-01
4.68434602e-01 -1.52100194e+00 1.32993490e-01 -9.43933129e-01
-6.92311168e-01 8.71867478e-01 6.41728938e-01 2.03019217e-01
4.87890214e-01 3.41952115e-01 -4.95039709e-02 -3.36237311e-01
-5.03010511e-01 -2.83950806e-01 3.91914397e-01 6.46367669e-01
6.64023936e-01 1.06134497e-01 -4.72606510e-01 7.40093172e-01
-1.12649100e-02 -2.33947262e-01 3.80015910e-01 1.19195342e+00
-1.83665499e-01 -1.55029917e+00 -3.78621697e-01 6.43691957e-01
-5.36149800e-01 -1.33579254e-01 -8.53220642e-01 1.01493800e+00
4.20364499e-01 8.62545609e-01 3.31322253e-01 -1.24891959e-01
6.82527900e-01 1.94571912e-01 1.82768747e-01 -4.37613368e-01
-8.66455197e-01 -4.60958719e-01 4.26501662e-01 -3.06992084e-01
-5.31911314e-01 -9.75505471e-01 -1.58520997e+00 -4.37928259e-01
-2.97189415e-01 3.60721052e-01 6.50495231e-01 8.02710474e-01
7.31557965e-01 2.31133357e-01 4.33701485e-01 1.52749307e-02
-1.41574070e-01 -6.41266346e-01 -7.22547233e-01 9.43513930e-01
-1.09675974e-02 -9.99068379e-01 -1.01442620e-01 1.62685558e-01] | [9.249128341674805, 8.13854694366455] |
b3cf0579-f9b9-422f-9e6b-4630dccdee51 | optimizing-a-digital-twin-for-fault-diagnosis | 2212.03564 | null | https://arxiv.org/abs/2212.03564v1 | https://arxiv.org/pdf/2212.03564v1.pdf | Optimizing a Digital Twin for Fault Diagnosis in Grid Connected Inverters -- A Bayesian Approach | In this paper, a hyperparameter tuning based Bayesian optimization of digital twins is carried out to diagnose various faults in grid connected inverters. As fault detection and diagnosis require very high precision, we channelize our efforts towards an online optimization of the digital twins, which, in turn, allows a flexible implementation with limited amount of data. As a result, the proposed framework not only becomes a practical solution for model versioning and deployment of digital twins design with limited data, but also allows integration of deep learning tools to improve the hyperparameter tuning capabilities. For classification performance assessment, we consider different fault cases in virtual synchronous generator (VSG) controlled grid-forming converters and demonstrate the efficacy of our approach. Our research outcomes reveal the increased accuracy and fidelity levels achieved by our digital twin design, overcoming the shortcomings of traditional hyperparameter tuning methods. | ['Pedro H. J. Nardelli', 'Charalampos Kalalas', 'Subham Sahoo', 'Pavol Mulinka'] | 2022-12-07 | null | null | null | null | ['fault-detection'] | ['miscellaneous'] | [-3.76671314e-01 -2.26950362e-01 -1.05913512e-01 -5.04029691e-02
-6.06484771e-01 -5.46932578e-01 2.57394552e-01 -2.83591390e-01
4.96310085e-01 8.45921338e-01 -4.35724646e-01 -4.13665861e-01
-7.11444318e-01 -7.93291628e-01 -2.79524088e-01 -1.00734985e+00
-1.78088620e-01 8.21531057e-01 -1.13295078e-01 -1.71361879e-01
1.33209303e-01 5.10183930e-01 -1.40865171e+00 -4.33720291e-01
1.11993313e+00 1.09743726e+00 4.21980083e-01 3.16281199e-01
5.01838088e-01 6.78219125e-02 -1.04121244e+00 2.64546033e-02
1.72590852e-01 -6.32257313e-02 -5.41110039e-01 2.85644174e-01
-1.21717475e-01 -4.40992624e-01 -3.79356921e-01 1.01165569e+00
8.42997074e-01 -2.96711829e-02 5.80966473e-01 -1.41688335e+00
-2.50529826e-01 8.81903589e-01 -3.57842982e-01 1.78247467e-01
2.25844592e-01 4.81897920e-01 7.41301477e-01 -3.11769336e-01
-7.59736821e-02 8.72353971e-01 6.59336090e-01 6.06621662e-03
-1.29663157e+00 -8.51986647e-01 -6.12015799e-02 5.52040696e-01
-1.61628211e+00 -1.08691953e-01 7.12549090e-01 -5.13839066e-01
1.26970005e+00 1.31581560e-01 1.02265346e+00 1.00715399e+00
2.54057348e-01 4.56388891e-01 9.59037125e-01 -3.30640614e-01
5.48184156e-01 -1.40553012e-01 -3.34764570e-02 3.62673551e-01
4.81791645e-01 1.22913755e-01 -3.15853029e-01 -9.98880267e-02
7.45236218e-01 -4.77413565e-01 -6.39835954e-01 -1.60927415e-01
-8.79039407e-01 7.82196581e-01 2.10230867e-03 1.47493660e-01
-2.94475466e-01 2.17763230e-01 3.44192058e-01 3.13266188e-01
3.83256048e-01 7.24369824e-01 -8.80066276e-01 -2.88036346e-01
-7.52942502e-01 5.32180183e-02 7.74840653e-01 8.35161984e-01
3.31100494e-01 8.19748223e-01 -2.15985313e-01 9.00652468e-01
3.01403582e-01 4.10813361e-01 6.72335923e-01 -6.91169381e-01
1.73270017e-01 4.59238350e-01 2.30336338e-01 -5.24402976e-01
-7.16481626e-01 -8.88725877e-01 -9.20757234e-01 1.27440393e-01
2.41276268e-02 -4.06532466e-01 -6.58072531e-01 1.43404889e+00
2.38078713e-01 3.24698359e-01 -5.08007742e-02 7.73713946e-01
1.63927212e-01 4.21465099e-01 -2.90047288e-01 -3.26588571e-01
1.39263642e+00 -3.70668977e-01 -1.05995238e+00 1.56061843e-01
4.73559231e-01 -5.55271864e-01 1.07522047e+00 9.52753186e-01
-6.78219199e-01 -3.25794578e-01 -1.48284554e+00 7.30246723e-01
-1.22104712e-01 3.24900061e-01 6.06408536e-01 1.01754630e+00
-8.15109372e-01 8.33455980e-01 -9.02157187e-01 -1.51068896e-01
3.46091837e-01 5.32284498e-01 7.65725449e-02 1.93083897e-01
-1.39783633e+00 1.15699363e+00 6.21105552e-01 2.23812670e-01
-1.05715585e+00 -7.78072715e-01 -4.36332464e-01 2.97462881e-01
5.60722530e-01 -7.17194676e-01 1.21264076e+00 -2.53531039e-01
-2.16875529e+00 -1.81159768e-02 4.62270170e-01 -6.05176270e-01
3.08185607e-01 -2.05682829e-01 -5.12965918e-01 3.23290192e-02
-2.80607790e-01 -1.65161207e-01 1.00009489e+00 -8.23856413e-01
-4.71367598e-01 -1.76723629e-01 -1.91891596e-01 8.59135315e-02
-5.27487099e-01 -2.89656907e-01 -1.19959094e-01 -6.94006443e-01
4.96009067e-02 -6.11858130e-01 -2.59141102e-02 -6.38994217e-01
-4.49822068e-01 -3.56613219e-01 8.45553398e-01 -6.61369383e-01
1.27100348e+00 -1.71448112e+00 1.78272337e-01 3.83774430e-01
-1.19402342e-01 2.32358560e-01 3.57116997e-01 5.32606721e-01
-3.77648808e-02 -8.88040662e-02 -1.19940490e-01 9.18405280e-02
3.48764688e-01 2.32034221e-01 -1.66728064e-01 6.36541665e-01
3.46738636e-01 4.24709171e-01 -5.16715944e-01 -5.61936852e-03
6.26056492e-01 2.90929049e-01 -3.63920689e-01 2.12597206e-01
-4.07077730e-01 2.95402586e-01 -6.31927609e-01 6.61922812e-01
5.85962713e-01 -3.32790524e-01 3.88396949e-01 -5.34912765e-01
1.82040870e-01 7.27306306e-02 -1.47113872e+00 1.28350067e+00
-7.00605392e-01 3.28090817e-01 7.84640610e-02 -1.24393463e+00
1.06736481e+00 6.55337751e-01 5.88251948e-01 -6.05176687e-01
2.89593369e-01 7.27485567e-02 -1.14086069e-01 -6.26538873e-01
1.56875312e-01 2.03783020e-01 9.25372615e-02 3.65622342e-01
4.98324543e-01 -6.74351275e-01 1.82611793e-01 -3.21028262e-01
9.17110264e-01 9.91344899e-02 2.20817924e-01 -5.06589174e-01
2.91642398e-01 -5.58045283e-02 8.65450621e-01 4.58793044e-01
1.71323732e-01 3.48234653e-01 5.63373744e-01 -1.53725803e-01
-9.28680599e-01 -8.68047297e-01 -6.87271833e-01 2.10983858e-01
-1.44742504e-01 -3.47601265e-01 -8.08331490e-01 -4.30395186e-01
1.49731308e-01 8.33940029e-01 -9.32801813e-02 -3.61827433e-01
-2.92333871e-01 -1.48396802e+00 4.89212632e-01 4.33942258e-01
4.73056942e-01 -3.60290796e-01 -5.44482291e-01 3.54972273e-01
7.47948587e-02 -9.65512812e-01 8.72685909e-02 5.58661401e-01
-8.67689669e-01 -1.17953014e+00 -3.10931265e-01 -5.32581747e-01
1.65136755e-01 -4.11657214e-01 1.12948513e+00 1.12853780e-01
-4.90377903e-01 1.02867536e-01 -3.81003201e-01 -2.41358206e-01
-3.29362065e-01 8.76219347e-02 4.50604558e-01 -2.97559917e-01
-1.34973750e-01 -8.47432792e-01 -4.04665798e-01 5.59657812e-01
-5.03056884e-01 -2.04765931e-01 4.12027538e-01 1.23439050e+00
3.41732085e-01 8.35945964e-01 1.16526401e+00 -5.15945911e-01
7.10188925e-01 -5.17372191e-01 -1.54795444e+00 2.59118557e-01
-1.20757556e+00 -8.65700170e-02 7.99931645e-01 -5.37451744e-01
-8.50059450e-01 -8.91237855e-02 -1.21213317e-01 -6.19521797e-01
2.93922462e-02 6.06454968e-01 -4.90110099e-01 -2.97190934e-01
4.37561989e-01 -1.05026379e-01 -7.09714219e-02 -7.05252111e-01
1.52247220e-01 8.68112445e-01 3.69040728e-01 -6.11466110e-01
7.20130444e-01 -1.75811529e-01 1.64773971e-01 -6.82062149e-01
-1.32014319e-01 1.55206218e-01 -3.44613433e-01 -2.37385973e-01
2.97044158e-01 -1.07148147e+00 -9.22375202e-01 8.51615667e-01
-6.17242873e-01 -4.48985636e-01 -7.31634423e-02 5.68843842e-01
-5.75531185e-01 2.05566719e-01 -6.48872554e-01 -6.40514851e-01
-3.13027829e-01 -1.57994878e+00 8.88448298e-01 1.67716548e-01
-1.03113197e-01 -1.05748045e+00 -3.25632632e-01 3.94134462e-01
4.30235088e-01 1.98031947e-01 1.21182787e+00 -5.96646965e-01
-6.49076164e-01 -3.06674063e-01 3.29253703e-01 3.47577065e-01
1.70372635e-01 6.73636273e-02 -1.05299497e+00 -7.81535864e-01
1.96840957e-01 -2.62142777e-01 -5.53178973e-02 4.57824439e-01
1.37827432e+00 -5.89105152e-02 -2.04880327e-01 8.54110003e-01
1.53413403e+00 2.93174833e-01 3.58492792e-01 4.47790802e-01
3.98427695e-01 8.38047341e-02 6.41431570e-01 8.91331613e-01
2.42406771e-01 9.02394116e-01 6.49232805e-01 1.99218869e-01
4.79591079e-02 1.63904682e-01 1.80750955e-02 9.76366401e-01
3.52946669e-01 -5.38635969e-01 -9.38050330e-01 4.23007667e-01
-1.48489809e+00 -3.24095041e-01 4.85327914e-02 2.09032512e+00
6.71232283e-01 -5.06015308e-03 -2.72627741e-01 5.58803499e-01
6.76643610e-01 -3.65278214e-01 -6.47965848e-01 -6.78932574e-03
-1.06727861e-01 3.37071776e-01 3.98069888e-01 1.64336488e-01
-8.93630803e-01 5.75001717e-01 6.84708357e+00 1.10546577e+00
-1.07878995e+00 5.94677217e-02 4.83529270e-01 6.11127019e-02
1.33623198e-01 -1.07001506e-01 -4.80001599e-01 7.09273517e-01
1.04896605e+00 -3.39879721e-01 6.83736205e-01 7.22700059e-01
4.84854341e-01 -2.64155060e-01 -1.06082177e+00 9.81266618e-01
-2.85017580e-01 -1.24000394e+00 -2.12502748e-01 2.69726384e-02
8.67966056e-01 -1.97740048e-01 -8.85815918e-02 8.44678581e-02
2.93407053e-01 -1.04807568e+00 5.62188864e-01 2.44819820e-01
5.41206479e-01 -1.09412086e+00 1.05369067e+00 1.21479116e-01
-8.08965981e-01 -6.65972114e-01 -2.14726016e-01 4.86322679e-02
1.24975309e-01 1.00560272e+00 -1.22377551e+00 1.05522740e+00
9.77076590e-01 3.53401572e-01 -4.49609876e-01 1.33144736e+00
-2.92706519e-01 9.95098948e-01 -3.53599966e-01 1.31692775e-02
-3.70898664e-01 -4.05786276e-01 5.22851408e-01 5.92385173e-01
6.22017920e-01 -1.77118197e-01 2.49649271e-01 8.57485235e-01
1.23275034e-01 -3.16831023e-01 -2.93065280e-01 1.14966668e-01
1.07905912e+00 1.25702667e+00 -6.06720150e-01 -1.36874720e-01
7.15470165e-02 3.67507786e-01 -1.36426717e-01 3.56975079e-01
-8.22166204e-01 -4.30923551e-01 7.76297331e-01 -3.55415165e-01
3.02738428e-01 -1.19011976e-01 -6.15939915e-01 -8.59018564e-01
-1.15763910e-01 -1.20129323e+00 2.03609392e-01 -8.94622266e-01
-1.20849252e+00 3.62402320e-01 3.14971626e-01 -1.21682596e+00
-5.52612662e-01 -5.85929334e-01 -6.92237616e-01 9.87363279e-01
-1.11258996e+00 -7.72254527e-01 -1.81895465e-01 3.06755126e-01
5.42234600e-01 -5.20796716e-01 7.94898570e-01 3.71076941e-01
-1.12867618e+00 5.65407455e-01 5.66808999e-01 -4.03981984e-01
3.15958202e-01 -1.26136971e+00 2.53386289e-01 8.91951263e-01
-2.92422235e-01 1.36058077e-01 1.00047338e+00 -5.15205681e-01
-1.90031528e+00 -9.80611682e-01 -9.83506367e-02 9.11256373e-02
8.86721849e-01 -2.37463400e-01 -8.51263344e-01 2.47989267e-01
3.97565544e-01 -3.75857323e-01 3.58282924e-01 2.15007979e-02
3.90962690e-01 -3.35469306e-01 -1.34686863e+00 3.99177402e-01
3.24840248e-01 -4.91892248e-01 -3.79608333e-01 6.05751872e-01
4.83331501e-01 -3.31791133e-01 -1.49249327e+00 6.75735176e-01
3.36859711e-02 -4.14631218e-01 8.85752499e-01 -6.32759631e-02
-4.75031018e-01 -4.96494085e-01 -1.59568682e-01 -2.02694821e+00
-1.75138921e-01 -1.01229882e+00 -3.29322249e-01 1.50003481e+00
1.98779836e-01 -9.03983414e-01 4.92049724e-01 1.32844076e-01
-4.35068399e-01 -7.74761379e-01 -1.17139232e+00 -6.66771233e-01
4.62744124e-02 -3.60268533e-01 1.19404030e+00 1.02880394e+00
6.72624931e-02 1.61009371e-01 -9.25516188e-02 7.84671664e-01
7.41543174e-01 1.31858056e-02 3.01995754e-01 -1.33341801e+00
-4.83662724e-01 -3.13352227e-01 -3.73404294e-01 -3.75126958e-01
1.74512938e-01 -5.64905941e-01 -3.32395099e-02 -1.06046999e+00
-3.36309493e-01 -5.82820773e-01 -1.52156383e-01 4.97280508e-01
-8.77939463e-02 1.37511790e-01 -2.93169409e-01 -1.35492831e-01
8.93989354e-02 7.87353039e-01 9.17824209e-01 -1.70877859e-01
7.91760162e-02 1.15071043e-01 -1.33156911e-01 4.97647166e-01
1.26755178e+00 -2.06208542e-01 -6.40940547e-01 -3.43448490e-01
1.05570234e-01 3.05802733e-01 1.43207759e-01 -1.27034020e+00
2.94542462e-02 8.01897328e-03 4.75741237e-01 -4.32297438e-01
2.06284344e-01 -7.57479906e-01 4.26851630e-01 4.52423155e-01
4.33819473e-01 2.11121961e-01 2.81054616e-01 3.62573564e-01
3.20326234e-03 -2.94989377e-01 7.23220587e-01 4.15576190e-01
-4.64571267e-01 2.29387712e-02 -8.15076232e-01 -3.42399240e-01
8.51965547e-01 -5.41201495e-02 -3.58317316e-01 -1.15972787e-01
-5.56895256e-01 7.14229763e-01 4.09301043e-01 3.36230934e-01
3.17341954e-01 -1.09808207e+00 -4.91329432e-01 5.07457554e-01
-1.66338727e-01 -2.81330291e-02 1.91433474e-01 8.03115964e-01
-4.57430363e-01 3.23134005e-01 -2.13697985e-01 -8.44391882e-01
-8.26962233e-01 1.62671700e-01 6.49854422e-01 -9.97791812e-03
-7.24613726e-01 4.16595459e-01 -3.71005237e-01 -3.77347380e-01
2.74665713e-01 -3.33662719e-01 -2.40631685e-01 1.18809119e-01
-2.57700738e-02 6.31239474e-01 8.18021894e-01 7.31102675e-02
-5.67294806e-02 3.37004364e-01 3.87735724e-01 1.83347195e-01
1.44418156e+00 -1.05386980e-01 3.62946442e-03 2.75696635e-01
5.28716683e-01 -6.49703145e-01 -1.40418184e+00 4.33953553e-01
2.82834638e-02 -3.61111701e-01 6.49452627e-01 -1.13590109e+00
-1.37725616e+00 4.79027748e-01 1.05080163e+00 2.82161951e-01
1.35481465e+00 -4.28686202e-01 2.52063781e-01 5.47532737e-01
7.91489542e-01 -1.29741621e+00 -2.44159624e-01 4.10220027e-01
5.60576677e-01 -8.47766697e-01 4.22806069e-02 -1.17691502e-01
-3.10518324e-01 1.18436980e+00 5.38307667e-01 -4.68863882e-02
5.14561653e-01 7.46411204e-01 -8.94414932e-02 -1.31312504e-01
-1.06449354e+00 2.44203418e-01 -4.67308201e-02 9.10266936e-01
1.12506822e-01 1.81629524e-01 -1.33573040e-01 6.97000504e-01
-2.84058601e-01 -4.29800525e-02 6.69332981e-01 7.09853590e-01
-2.80587852e-01 -1.05587590e+00 -7.14400232e-01 7.28768349e-01
-4.38157111e-01 1.78159833e-01 3.32745582e-01 8.53619933e-01
2.92895511e-02 1.16419864e+00 -1.06888376e-02 -4.56591755e-01
3.63646150e-01 7.45426118e-02 6.17885232e-01 -2.23929077e-01
-5.93856692e-01 7.88037330e-02 1.92644402e-01 -4.43448305e-01
3.75585169e-01 -5.39208353e-01 -9.14070606e-01 -4.94550541e-02
-9.47863817e-01 4.22117114e-01 8.08331490e-01 9.52001691e-01
2.71862954e-01 1.13319123e+00 1.00983393e+00 -9.48984742e-01
-1.19759333e+00 -1.22235250e+00 -9.05341864e-01 -2.48239815e-01
1.03425279e-01 -1.17071915e+00 -4.62472320e-01 -3.10931742e-01] | [6.009566783905029, 2.671999216079712] |
ae8e65a1-ce5f-47e2-8729-029d9d2ce7fb | cgnn-traffic-classification-with-graph-neural | 2110.09726 | null | https://arxiv.org/abs/2110.09726v1 | https://arxiv.org/pdf/2110.09726v1.pdf | CGNN: Traffic Classification with Graph Neural Network | Traffic classification associates packet streams with known application labels, which is vital for network security and network management. With the rise of NAT, port dynamics, and encrypted traffic, it is increasingly challenging to obtain unified traffic features for accurate classification. Many state-of-the-art traffic classifiers automatically extract features from the packet stream based on deep learning models such as convolution networks. Unfortunately, the compositional and causal relationships between packets are not well extracted in these deep learning models, which affects both prediction accuracy and generalization on different traffic types. In this paper, we present a chained graph model on the packet stream to keep the chained compositional sequence. Next, we propose CGNN, a graph neural network based traffic classification method, which builds a graph classifier over automatically extracted features over the chained graph. Extensive evaluation over real-world traffic data sets, including normal, encrypted and malicious labels, show that, CGNN improves the prediction accuracy by 23\% to 29\% for application classification, by 2\% to 37\% for malicious traffic classification, and reaches the same accuracy level for encrypted traffic classification. CGNN is quite robust in terms of the recall and precision metrics. We have extensively evaluated the parameter sensitivity of CGNN, which yields optimized parameters that are quite effective for traffic classification. | ['Yan Jia', 'Qing Liao', 'Ye Wang', 'Siyuan Ren', 'Yongquan Fu', 'Bo Pang'] | 2021-10-19 | null | null | null | null | ['traffic-classification'] | ['miscellaneous'] | [ 1.57511830e-02 -4.72211331e-01 -4.71010536e-01 -4.12204295e-01
1.95753574e-01 -5.00646830e-01 2.85825670e-01 1.00165576e-01
-7.81686902e-02 5.13957739e-01 -4.69036192e-01 -9.87676322e-01
-1.89051345e-01 -1.15321267e+00 -3.17151874e-01 -3.58594239e-01
-4.66425329e-01 3.72223318e-01 5.79455197e-01 -1.33155897e-01
1.57372057e-01 7.91340292e-01 -1.39620686e+00 4.70809907e-01
4.24418896e-01 1.66376722e+00 -3.45662981e-01 7.20993400e-01
-7.83763230e-01 1.07700825e+00 -6.49078608e-01 -8.39823067e-01
3.00375998e-01 -2.23041639e-01 -5.65087199e-01 -1.80194408e-01
3.51777464e-01 -1.15624644e-01 -7.75782108e-01 1.03703964e+00
1.03572212e-01 -2.86699742e-01 5.32215893e-01 -1.80185580e+00
-3.53692710e-01 7.54594803e-01 -2.70598412e-01 5.09181559e-01
-1.17609635e-01 2.89102107e-01 1.15407860e+00 -9.95516777e-02
3.65426689e-01 1.11236548e+00 7.60907710e-01 2.76409507e-01
-9.61731672e-01 -1.32538402e+00 1.69395998e-01 4.66673136e-01
-1.01953995e+00 -2.50880301e-01 9.03288960e-01 -4.88673627e-01
8.65688980e-01 2.51863807e-01 5.40566444e-01 9.31216121e-01
3.30134541e-01 3.53190333e-01 6.85851216e-01 6.07507750e-02
-1.47802085e-01 1.96190789e-01 6.26095116e-01 8.53539944e-01
3.76118869e-01 2.10195616e-01 -1.33692175e-01 -2.11318552e-01
4.15653288e-01 3.76235485e-01 -9.34161097e-02 -1.29377469e-01
-6.98129654e-01 7.06445456e-01 5.10600626e-01 2.22056270e-01
-3.31417203e-01 3.63912970e-01 7.52853513e-01 7.03296721e-01
5.30199111e-01 1.50611609e-01 -4.90138769e-01 -3.08373451e-01
-5.66634655e-01 -2.65870363e-01 1.03184438e+00 1.04890692e+00
9.58593965e-01 4.62683707e-01 -2.05880310e-02 6.53706074e-01
1.75885513e-01 6.61117494e-01 2.11731106e-01 -4.04315650e-01
5.24622023e-01 9.23092723e-01 -7.31922746e-01 -1.29005003e+00
-5.01227319e-01 -6.79810345e-01 -1.01285505e+00 -1.28341228e-01
4.44278419e-01 -1.89308614e-01 -6.73369586e-01 1.59590197e+00
-7.78357908e-02 6.73126578e-01 -1.45732984e-01 3.84141475e-01
8.07110965e-01 5.54549575e-01 3.87468487e-01 -8.90065134e-02
1.17890930e+00 -6.06700718e-01 -4.79756027e-01 2.50226021e-01
8.23424935e-01 -5.82420707e-01 7.28418767e-01 1.25267237e-01
-4.14858878e-01 -6.71288252e-01 -8.24366033e-01 5.93304455e-01
-6.81234121e-01 -1.75838232e-01 9.02160168e-01 1.05489731e+00
-8.03586185e-01 6.92707360e-01 -3.05525333e-01 -3.71816248e-01
9.40303624e-01 6.56816423e-01 -2.54594952e-01 -1.30389333e-01
-1.19199085e+00 2.56100744e-01 5.70057213e-01 -2.61863321e-01
-6.29272819e-01 -7.58756757e-01 -5.54987371e-01 3.99919361e-01
2.76138365e-01 -2.99643636e-01 8.23323667e-01 -9.92616653e-01
-1.25099552e+00 4.87183750e-01 3.56148481e-02 -5.41788816e-01
1.86106250e-01 3.37178022e-01 -1.15009320e+00 -1.45894527e-01
-2.01887518e-01 8.19631442e-02 8.80614400e-01 -9.50645089e-01
-9.46810603e-01 -1.60692453e-01 1.14609964e-01 -7.04964161e-01
-5.92599988e-01 2.59177797e-02 -1.15808599e-01 -4.31653410e-01
-2.12681815e-01 -8.21855307e-01 1.02530144e-01 -2.83734024e-01
-3.63281190e-01 -4.01866972e-01 1.39587331e+00 -1.55101568e-01
1.41499329e+00 -2.07726383e+00 -6.09422982e-01 6.33628249e-01
7.39596546e-01 6.49563849e-01 -2.31492788e-01 2.44117275e-01
-1.68525964e-01 3.74891728e-01 1.14012077e-01 -5.20346453e-03
-6.49458962e-03 1.11660063e-01 -5.21326721e-01 1.06668681e-01
2.34757826e-01 1.05300581e+00 -7.68666029e-01 -3.48125845e-01
2.28061125e-01 2.25806877e-01 -4.67775524e-01 2.57407963e-01
-2.07520261e-01 1.44648075e-01 -4.48659390e-01 5.98041594e-01
8.44976068e-01 -5.06474853e-01 3.08738351e-01 -3.76616597e-01
3.18322450e-01 2.53826559e-01 -6.88606739e-01 7.43133307e-01
-6.15749776e-01 9.55066681e-01 -3.02404702e-01 -1.29056394e+00
1.26304579e+00 2.09920332e-01 6.72466218e-01 -7.20082879e-01
3.68587136e-01 1.68024838e-01 3.25997323e-01 -3.49513113e-01
3.46000195e-02 3.09592206e-02 1.25506386e-01 5.98812342e-01
7.41729140e-02 5.35306692e-01 2.68730611e-01 9.42267179e-02
1.31758630e+00 -6.59792840e-01 -3.97274382e-02 -7.65670836e-02
9.47886348e-01 -2.44364455e-01 4.65870202e-01 6.88188732e-01
-2.95740366e-01 -4.82792221e-02 1.02971995e+00 -8.32966864e-01
-6.99172616e-01 -9.51206267e-01 4.65140566e-02 1.28180206e+00
5.63627668e-02 -4.35770661e-01 -5.54443002e-01 -1.19396329e+00
1.54391915e-01 4.12088543e-01 -2.31338948e-01 -3.83887947e-01
-8.36208105e-01 -8.70680988e-01 8.72504473e-01 2.90925801e-01
7.64791489e-01 -1.08334017e+00 1.80694729e-01 2.84376830e-01
1.93070248e-01 -1.78304470e+00 -1.94932640e-01 8.96948576e-02
-6.36034012e-01 -1.48666751e+00 1.12470165e-01 -8.24429929e-01
3.86334747e-01 2.12947935e-01 1.22952211e+00 5.66436887e-01
-5.04512228e-02 -7.39220679e-02 -4.42103297e-01 -2.13146865e-01
-6.46016777e-01 4.97530788e-01 1.75333302e-02 5.63017488e-01
5.93096077e-01 -9.54604268e-01 -1.99547604e-01 4.53748077e-01
-6.65281534e-01 -3.48283559e-01 6.09541655e-01 4.41176832e-01
-1.63285490e-02 3.76658082e-01 7.89209187e-01 -1.28758597e+00
7.18161166e-01 -7.49302030e-01 -6.61084473e-01 1.09242141e-01
-7.16136873e-01 -6.98463768e-02 1.28231108e+00 -5.59291482e-01
-3.94804567e-01 -4.51030552e-01 -2.35937208e-01 -5.98890424e-01
-1.83170751e-01 1.34934977e-01 -2.35226557e-01 -3.43035877e-01
4.42103088e-01 2.61974752e-01 3.38514298e-02 -1.31659344e-01
1.21173829e-01 8.72803748e-01 1.03508532e-01 -4.78553146e-01
9.35454249e-01 3.03900570e-01 4.59870815e-01 -8.42775166e-01
-4.60142583e-01 -1.38863161e-01 -3.92037898e-01 -3.05981666e-01
5.80149233e-01 -2.71972358e-01 -1.28671777e+00 4.87048596e-01
-1.09449220e+00 -1.25949398e-01 -7.05681145e-02 4.06816274e-01
-6.86665848e-02 5.47285140e-01 -7.38803804e-01 -4.72356617e-01
-3.83668751e-01 -1.32068741e+00 4.88578111e-01 -7.27444738e-02
2.71077268e-02 -1.25611269e+00 -4.32637215e-01 4.80477177e-02
8.16090167e-01 4.83690985e-02 1.60795045e+00 -1.20339584e+00
-8.19842160e-01 -3.39306593e-01 -8.88266087e-01 5.70233524e-01
3.60364676e-01 1.00513667e-01 -8.09212327e-01 -1.31220683e-01
-4.32019770e-01 2.13466689e-01 7.95310438e-01 -4.62240446e-03
1.51572907e+00 -3.38595212e-01 -3.63024145e-01 1.02385020e+00
1.15793097e+00 5.66136658e-01 5.60752869e-01 -1.77386347e-02
9.79014814e-01 4.75242376e-01 -8.98369849e-02 1.68640614e-01
9.54366848e-02 3.66097718e-01 6.10448599e-01 2.08286390e-01
-3.14684868e-01 -1.35310084e-01 1.50879472e-01 9.15662766e-01
1.37828931e-01 -7.13943422e-01 -9.37188685e-01 -1.04911976e-01
-1.54645896e+00 -9.73322511e-01 -2.85589427e-01 1.78676987e+00
1.17394902e-01 7.04802632e-01 2.46374622e-01 3.40598971e-01
9.30049539e-01 2.69171178e-01 -3.27214062e-01 -5.03055096e-01
8.98703784e-02 5.38003981e-01 7.48178482e-01 2.91455775e-01
-1.01560473e+00 1.03604078e+00 6.09276962e+00 1.07624173e+00
-1.60268199e+00 -7.31169805e-02 6.88651204e-01 5.30764997e-01
-1.12218343e-01 -3.38932574e-02 -8.08478534e-01 8.76964808e-01
1.29633713e+00 -9.51177627e-02 5.49111605e-01 7.52841771e-01
4.38116565e-02 7.13486493e-01 -1.02810907e+00 1.19590998e+00
-2.41118535e-01 -1.52907431e+00 5.95230579e-01 2.10628703e-01
3.04641366e-01 1.14405654e-01 8.37989151e-02 6.53309047e-01
3.51475209e-01 -9.50024724e-01 1.04326673e-01 2.21561402e-01
8.50880563e-01 -9.47896421e-01 1.03878796e+00 8.62761438e-02
-1.51030183e+00 -3.38485897e-01 -2.22437635e-01 -5.18907420e-02
-5.71424849e-02 5.73722005e-01 -8.49238753e-01 6.15857065e-01
3.34471226e-01 1.02757728e+00 -6.24426842e-01 7.04243362e-01
1.19536392e-01 1.00249076e+00 -8.33970904e-02 -2.80742586e-01
2.13289231e-01 -2.49168038e-01 3.01634282e-01 1.34615934e+00
1.78183272e-01 -1.51167214e-01 3.29138756e-01 7.24654436e-01
-5.82754493e-01 1.07328765e-01 -6.70829892e-01 -2.67336071e-01
5.14126778e-01 1.25745904e+00 -8.33428204e-01 -4.07699615e-01
-4.25709426e-01 3.73436540e-01 1.95873573e-01 4.66266483e-01
-1.05181599e+00 -7.76331246e-01 1.05075848e+00 3.75500649e-01
1.64341450e-01 -1.28492832e-01 -1.84953108e-01 -8.90621841e-01
-1.85883164e-01 -6.90247655e-01 3.40385914e-01 -1.46802545e-01
-1.52076125e+00 8.95004809e-01 -2.31873676e-01 -1.38356471e+00
3.04359775e-02 -1.00159860e+00 -8.91854644e-01 5.14569461e-01
-1.52200711e+00 -1.00419986e+00 -5.81942916e-01 6.19294226e-01
1.39192089e-01 -8.11177552e-01 4.53128159e-01 7.96888828e-01
-8.04304063e-01 9.92626250e-01 -1.66188985e-01 6.69373810e-01
2.04250634e-01 -6.60780251e-01 6.37382865e-01 5.45012772e-01
1.48437738e-01 3.53377014e-01 1.01116747e-01 -5.34049630e-01
-1.33491731e+00 -1.59007013e+00 6.00560188e-01 -1.70901924e-01
8.83526444e-01 -2.93539256e-01 -8.48936260e-01 6.37391329e-01
-2.15077922e-01 4.30855334e-01 6.61280990e-01 -2.55172793e-02
-8.45041096e-01 -8.24501753e-01 -1.15380919e+00 4.05908018e-01
1.22961378e+00 -6.46670997e-01 2.34969646e-01 2.49546602e-01
9.51170266e-01 1.93766400e-01 -9.03398573e-01 4.88295257e-01
5.39608419e-01 -9.84621584e-01 7.47633696e-01 -9.97289479e-01
-1.17575087e-01 -1.34228364e-01 -2.81253040e-01 -9.78501558e-01
-4.57411408e-01 -5.28332889e-01 -3.65784347e-01 1.22058558e+00
4.32532459e-01 -1.14934671e+00 1.11607635e+00 -1.54136801e-02
2.15183124e-02 -7.13981092e-01 -7.37984419e-01 -8.88098419e-01
-1.85342312e-01 -9.53171372e-01 1.14387047e+00 1.00399446e+00
-3.59179676e-01 5.38135827e-01 -1.32421851e-01 1.00348637e-01
6.88515902e-01 6.87620267e-02 9.11687315e-01 -1.64636779e+00
6.00204431e-02 -9.15856481e-01 -1.17308652e+00 -9.74879205e-01
6.31875634e-01 -1.17869949e+00 -7.87469506e-01 -1.01619053e+00
-1.81480810e-01 -7.70825088e-01 -5.57301044e-01 1.90375492e-01
2.26326406e-01 2.23666683e-01 1.49842426e-01 1.59472317e-01
-6.05344594e-01 2.09126428e-01 9.95243609e-01 -3.84228975e-01
4.59427051e-02 3.64000589e-01 -4.64338839e-01 4.91880506e-01
1.07087588e+00 -4.17611837e-01 -6.29066110e-01 -1.60989925e-01
-4.68732379e-02 -1.78102583e-01 1.19789086e-01 -1.17287731e+00
1.40166819e-01 -1.80718318e-01 1.92017928e-01 -3.62902224e-01
1.01836160e-01 -1.07797825e+00 1.13315023e-01 4.90199804e-01
-1.59656137e-01 2.04632893e-01 5.61102778e-02 5.59142232e-01
-1.13208771e-01 1.86788380e-01 7.40818322e-01 1.61115497e-01
-8.15724790e-01 1.01159990e+00 -3.68539780e-01 3.19201589e-01
9.54285979e-01 -2.62416184e-01 -5.26735723e-01 -5.81454217e-01
-4.69975829e-01 -3.06598321e-02 7.43332058e-02 5.79832315e-01
5.83051920e-01 -1.43056035e+00 -5.48841476e-01 5.76639175e-01
1.00851029e-01 -6.15910828e-01 -7.27775097e-02 6.58263922e-01
-7.27591157e-01 6.21015131e-01 -2.91951895e-01 -6.98678255e-01
-1.26361084e+00 6.99162006e-01 4.11053717e-01 -2.91639477e-01
-2.40706459e-01 3.71127576e-01 1.13441031e-02 -4.02419627e-01
2.71096110e-01 -4.21822369e-01 -1.97083279e-01 -1.50737107e-01
3.17074507e-01 5.63601196e-01 2.16923252e-01 -6.17331326e-01
-3.72548521e-01 6.86005890e-01 -1.18308887e-01 6.52385414e-01
8.66561174e-01 2.06089646e-01 -2.53392249e-01 -5.66706294e-04
1.72886324e+00 -1.61102355e-01 -5.37974775e-01 -2.44488180e-01
5.77175803e-02 -4.11295146e-01 -1.28761649e-01 -2.63833612e-01
-1.77681470e+00 1.08947802e+00 5.10734856e-01 8.48088741e-01
1.01035142e+00 -3.52750093e-01 1.13457692e+00 5.84223688e-01
3.36745769e-01 -2.93750018e-01 -9.22348648e-02 7.79718101e-01
-8.97785053e-02 -1.03273880e+00 -3.38359952e-01 -6.96304917e-01
-2.35248014e-01 1.44238496e+00 7.53433704e-01 -9.33275744e-02
1.29533780e+00 1.69913575e-01 -6.54832274e-02 -2.89030910e-01
-7.99072981e-01 -3.07987988e-01 5.13051637e-02 6.62555039e-01
2.56245524e-01 1.85613975e-01 1.83848843e-01 2.74457693e-01
-2.59153187e-01 -2.49838576e-01 2.74627030e-01 4.46327269e-01
-5.17648160e-01 -1.15173268e+00 1.34143978e-01 9.08808053e-01
-5.87189436e-01 -1.00793548e-01 -2.10581467e-01 7.97410965e-01
2.09446579e-01 1.05291963e+00 3.36718023e-01 -1.22505105e+00
3.06215733e-01 1.00099079e-01 -1.95651762e-02 -2.63828427e-01
-6.31545544e-01 -6.40136003e-01 1.86897859e-01 -5.20142555e-01
4.72087488e-02 1.00075856e-01 -1.14914882e+00 -9.90280867e-01
-2.59403825e-01 3.57058823e-01 5.41917324e-01 8.26560795e-01
4.28656369e-01 7.13451028e-01 1.09426498e+00 -2.11784646e-01
-2.35772014e-01 -7.33907819e-01 -4.85842019e-01 5.36055565e-01
3.11896592e-01 -6.41179383e-01 -5.26661277e-01 -3.24317098e-01] | [5.063797473907471, 7.238203525543213] |
a2088f43-758f-4610-a3a1-607fb0ae6f29 | exploring-timbre-disentanglement-in-non | 2110.07192 | null | https://arxiv.org/abs/2110.07192v3 | https://arxiv.org/pdf/2110.07192v3.pdf | Exploring Timbre Disentanglement in Non-Autoregressive Cross-Lingual Text-to-Speech | In this paper, we study the disentanglement of speaker and language representations in non-autoregressive cross-lingual TTS models from various aspects. We propose a phoneme length regulator that solves the length mismatch problem between IPA input sequence and monolingual alignment results. Using the phoneme length regulator, we present a FastPitch-based cross-lingual model with IPA symbols as input representations. Our experiments show that language-independent input representations (e.g. IPA symbols), an increasing number of training speakers, and explicit modeling of speech variance information all encourage non-autoregressive cross-lingual TTS model to disentangle speaker and language representations. The subjective evaluation shows that our proposed model can achieve decent naturalness and speaker similarity in cross-language voice cloning. | ['Yue Lin', 'Yang Zhang', 'Haitong Zhang', 'Xinyuan Yu', 'Haoyue Zhan'] | 2021-10-14 | null | null | null | null | ['voice-cloning'] | ['speech'] | [-1.22162983e-01 -5.40437996e-02 -4.30389017e-01 -4.81640100e-01
-1.16940379e+00 -8.60420883e-01 4.01531368e-01 -7.83388674e-01
-7.70107582e-02 3.70350897e-01 4.58641440e-01 -8.21703970e-01
2.16633052e-01 5.90340756e-02 -5.43725073e-01 -5.16034722e-01
2.49708742e-01 4.11701709e-01 -3.40818346e-01 -3.06415945e-01
-9.19292942e-02 4.90472287e-01 -1.11281967e+00 9.12374929e-02
1.15256321e+00 3.44825923e-01 2.40192458e-01 9.02894557e-01
-1.96554586e-01 3.39851648e-01 -7.01035380e-01 -4.13082063e-01
2.93060541e-01 -4.83271718e-01 -5.69104195e-01 -2.44007111e-01
6.07586324e-01 9.12399441e-02 -3.65172207e-01 9.16619956e-01
7.92581439e-01 2.15389729e-02 8.81465793e-01 -1.07151377e+00
-9.07571852e-01 9.86162841e-01 -7.21310437e-01 3.20259959e-01
2.51648128e-01 4.36696894e-02 1.10235977e+00 -7.92141736e-01
-7.56145492e-02 1.86753845e+00 5.16225457e-01 6.70655072e-01
-1.35121536e+00 -1.16926205e+00 2.84677893e-01 1.88509300e-01
-1.72437632e+00 -8.81973267e-01 9.96469975e-01 -3.43890607e-01
8.52105618e-01 7.24954605e-01 1.60771444e-01 1.41858459e+00
3.19859356e-01 8.57284367e-01 1.25221097e+00 -7.87680805e-01
-4.46938515e-01 4.83071446e-01 1.52638614e-01 3.69322747e-01
-2.55451709e-01 4.94619578e-01 -6.78165197e-01 -2.29112297e-01
6.99028134e-01 -7.73228228e-01 -2.61250556e-01 -1.13309190e-01
-7.53712595e-01 7.35399485e-01 -1.53528467e-01 4.07490611e-01
-9.03203785e-02 -1.04696453e-01 5.68009436e-01 5.08138895e-01
2.68466890e-01 3.80767137e-01 -6.83163464e-01 -2.49023184e-01
-6.56886280e-01 -3.47977787e-01 5.48816383e-01 1.14793289e+00
1.57014921e-01 8.85580361e-01 -2.35545542e-02 1.44531095e+00
5.90051830e-01 1.20652902e+00 9.55113173e-01 -7.84910738e-01
6.81633770e-01 -3.34301442e-01 -3.22989583e-01 -8.17951918e-01
8.45318139e-02 -5.25660157e-01 -6.47356272e-01 6.62333006e-03
1.16921179e-01 -5.94781488e-02 -6.03843510e-01 2.22667170e+00
-1.08835571e-01 8.89919028e-02 3.77565742e-01 5.99549592e-01
4.59474504e-01 9.33979571e-01 1.77715749e-01 -6.01317585e-01
1.60181248e+00 -1.07993650e+00 -1.11013567e+00 -2.28338122e-01
5.19552231e-01 -1.32934535e+00 1.46699011e+00 5.07891178e-01
-1.08916283e+00 -1.04037321e+00 -9.47431505e-01 -8.94660205e-02
-1.04353704e-01 4.76963788e-01 5.36537766e-02 1.27799022e+00
-7.04793811e-01 7.55698830e-02 -4.57438350e-01 2.19336748e-02
-4.18854386e-01 4.88419622e-01 -2.98825532e-01 2.51839995e-01
-1.40505898e+00 9.14953709e-01 1.10720620e-01 -9.29028355e-03
-4.63278323e-01 -4.76923168e-01 -9.04420614e-01 -6.59746025e-03
-2.92342424e-01 -4.38615978e-01 1.26238728e+00 -1.14818919e+00
-2.10367060e+00 7.63732910e-01 -6.28225982e-01 -2.16931254e-01
1.96869582e-01 -7.40731955e-01 -9.63221133e-01 -4.09806430e-01
-1.83605880e-01 6.13473237e-01 1.07361877e+00 -1.35839391e+00
-2.12454095e-01 -3.24981183e-01 -7.93199897e-01 5.63815653e-01
-1.44446909e-01 4.04896200e-01 -3.71811479e-01 -9.38512921e-01
8.99584126e-03 -1.26898825e+00 2.65665770e-01 -7.52463758e-01
-6.40573502e-01 -1.15683049e-01 7.13254869e-01 -1.13855255e+00
1.39960110e+00 -2.25807047e+00 2.72484392e-01 1.02816038e-01
-3.87018830e-01 4.09635484e-01 -3.80340874e-01 2.56782472e-01
-4.37832028e-01 3.09050292e-01 2.01203808e-01 -6.15447521e-01
-9.17798653e-02 3.78983557e-01 -6.98511600e-01 3.52233410e-01
-4.86411937e-02 6.92119658e-01 -4.83089179e-01 -5.58730006e-01
1.23550177e-01 6.25096679e-01 -4.39708233e-01 3.77851456e-01
4.19526279e-01 6.88676178e-01 -7.66447708e-02 2.62516111e-01
7.84365714e-01 7.55319178e-01 2.88453460e-01 -2.02311620e-01
-1.12970874e-01 8.63045692e-01 -7.78481603e-01 1.39913940e+00
-9.48095083e-01 7.24600911e-01 9.35281739e-02 -3.65151018e-01
1.27066922e+00 7.60403872e-01 -1.96709409e-01 -6.21487200e-01
3.06804091e-01 3.61831158e-01 5.30879736e-01 -2.48973474e-01
6.09427989e-01 -4.12689388e-01 -2.39459679e-01 2.36843184e-01
-2.98672486e-02 -2.90469617e-01 -6.12985849e-01 -2.87533075e-01
1.00075200e-01 4.20029089e-02 4.86239433e-01 -3.92249376e-01
8.10079694e-01 -7.13867307e-01 6.62238300e-01 5.29069424e-01
-2.79073536e-01 5.64200342e-01 3.03083230e-02 2.15345308e-01
-8.85949433e-01 -1.20922780e+00 -1.29121616e-01 1.39914143e+00
-2.59817392e-01 -2.93857455e-01 -7.55562603e-01 -4.25371945e-01
-2.79875010e-01 1.29782236e+00 -1.38095208e-02 -2.53569841e-01
-9.97195005e-01 -1.46541879e-01 1.11320281e+00 3.28792423e-01
-1.06337205e-01 -7.96752095e-01 5.45497179e-01 8.11083913e-02
-3.25617462e-01 -1.12270594e+00 -1.19048965e+00 1.28261849e-01
-7.66905904e-01 -3.99167985e-01 -5.53790450e-01 -1.04802012e+00
3.62407297e-01 2.17548683e-01 7.53902733e-01 -6.16846621e-01
2.72496790e-01 2.14754507e-01 -9.88292843e-02 -3.62594187e-01
-1.01146090e+00 2.56656259e-01 8.19196939e-01 -2.62607299e-02
4.79326963e-01 -7.20884919e-01 -3.84029038e-02 5.31439483e-01
-2.07835451e-01 -3.54140885e-02 4.73847419e-01 8.27959955e-01
4.14519459e-01 -4.12259340e-01 7.19282925e-01 -6.71886027e-01
8.61858249e-01 -2.29052886e-01 -4.95983928e-01 2.51551747e-01
-4.06690806e-01 3.08970422e-01 9.39226329e-01 -9.15135682e-01
-1.20502651e+00 -6.40770718e-02 -4.04841542e-01 -7.18503892e-01
-1.43495932e-01 6.25659600e-02 -6.69624150e-01 3.07111800e-01
5.51035464e-01 4.99138623e-01 3.40587012e-02 -7.86302269e-01
8.02772224e-01 1.15443873e+00 5.78342676e-01 -8.48995984e-01
8.39643955e-01 -3.79249424e-01 -6.18734360e-01 -1.02265179e+00
-2.25189343e-01 -4.49302167e-01 -7.08781421e-01 1.43390402e-01
5.86625159e-01 -1.08773243e+00 -7.91256189e-01 3.89660209e-01
-1.54286706e+00 1.17165945e-01 -1.14928775e-01 9.74640846e-01
-7.51757026e-01 5.25710106e-01 -6.22445405e-01 -1.20405447e+00
-4.19468343e-01 -1.27848721e+00 8.83590877e-01 -1.16993420e-01
-5.93653023e-01 -8.36861789e-01 2.81633288e-01 5.44158638e-01
2.35221386e-01 -7.31553435e-01 1.11846983e+00 -9.34478223e-01
-3.37220460e-01 7.88195617e-03 2.57164180e-01 7.27545142e-01
5.42506397e-01 -7.37319374e-03 -1.33093977e+00 -1.60046965e-01
2.75885493e-01 4.96308431e-02 7.34156072e-02 3.36242706e-01
7.92011380e-01 -7.17256784e-01 -4.89461906e-02 6.26587629e-01
6.88207448e-01 4.96149093e-01 5.89393616e-01 -2.29794025e-01
1.03734386e+00 8.13094556e-01 4.68606740e-01 -6.23120070e-02
2.37321690e-01 9.19507444e-01 -2.28910327e-01 -3.81355669e-04
-2.97750562e-01 -5.25715232e-01 9.27391231e-01 2.30479884e+00
1.96521729e-01 -3.49577338e-01 -5.44678032e-01 4.10373181e-01
-1.31143582e+00 -8.19131613e-01 -2.73018539e-01 2.54179072e+00
1.09931874e+00 -1.05828933e-01 1.06818095e-01 6.34209439e-03
9.23077524e-01 1.01379856e-01 -3.83042485e-01 -1.23852336e+00
-2.52187967e-01 -5.85949328e-03 4.42962229e-01 1.08308542e+00
-6.75298989e-01 1.52672136e+00 6.85952425e+00 1.28722489e+00
-1.17027640e+00 1.04512647e-01 4.61416066e-01 4.67307791e-02
-5.38839400e-01 -2.74418265e-01 -1.04724348e+00 3.58929485e-01
1.30379796e+00 -2.91372031e-01 5.71098387e-01 7.68287301e-01
3.99827659e-01 5.66283762e-01 -1.19773209e+00 1.20287800e+00
2.90982932e-01 -4.37507242e-01 3.31306964e-01 1.30088195e-01
4.01220620e-01 8.11346695e-02 4.25665587e-01 4.10091639e-01
3.61044943e-01 -1.18264484e+00 7.40871310e-01 4.79167290e-02
8.70633543e-01 -8.83606911e-01 3.17836821e-01 3.92643929e-01
-1.18953657e+00 1.25703007e-01 -2.09566638e-01 3.55796635e-01
2.65381157e-01 -1.12673938e-01 -9.26543534e-01 6.61829054e-01
9.53821614e-02 1.75771281e-01 -2.11860746e-01 5.47989070e-01
-2.67383575e-01 9.34105456e-01 -1.88476726e-01 2.33239591e-01
-1.82479218e-01 -3.10008645e-01 8.78064454e-01 1.49063313e+00
5.47945857e-01 -2.25855127e-01 -1.33923590e-02 6.45883083e-01
2.17410311e-01 6.35890424e-01 -8.43200386e-01 -5.03690206e-02
8.60848308e-01 6.31759286e-01 2.69174844e-01 -8.74739289e-02
-2.02397168e-01 1.04135287e+00 2.34150752e-01 4.62579906e-01
-7.16617644e-01 -1.01676196e-01 1.13191843e+00 -2.98656642e-01
6.17450885e-02 -5.79966366e-01 -3.35067898e-01 -8.47609699e-01
-3.23302716e-01 -1.38812125e+00 1.86245088e-02 -8.04004312e-01
-1.39130783e+00 1.01329923e+00 -1.49531633e-01 -1.40971923e+00
-6.94206774e-01 -5.40391326e-01 -5.22514462e-01 1.43929720e+00
-1.42698395e+00 -1.25354528e+00 7.98352659e-01 5.49110949e-01
9.56007779e-01 -5.51177025e-01 1.16287625e+00 2.87780583e-01
-5.90986192e-01 1.41202819e+00 9.04310346e-02 -2.35554613e-02
8.30905318e-01 -9.40760672e-01 6.03653967e-01 6.14994764e-01
5.30757844e-01 1.14812577e+00 9.24455643e-01 -3.68657857e-01
-1.23135352e+00 -7.89609313e-01 1.22498226e+00 -5.14259696e-01
4.30666804e-01 -4.36097413e-01 -1.09802055e+00 7.74557412e-01
6.34968698e-01 -4.52752948e-01 1.15059566e+00 5.07450521e-01
-6.69965386e-01 -1.64750054e-01 -6.31229520e-01 1.04096913e+00
6.81561351e-01 -1.14762938e+00 -9.15224433e-01 -7.52181411e-02
1.16614532e+00 -1.33967131e-01 -5.29607594e-01 1.75971463e-01
8.57079208e-01 -4.54922616e-01 1.02413058e+00 -6.24282300e-01
-2.91810036e-01 -2.62479007e-01 -4.34832335e-01 -1.52775514e+00
-2.98303723e-01 -1.10780525e+00 2.96229780e-01 1.61509991e+00
6.15541637e-01 -5.92442930e-01 1.66954041e-01 1.80669099e-01
-2.06415087e-01 -1.48272589e-01 -1.16258848e+00 -1.38264263e+00
5.03824353e-01 -5.97682416e-01 3.93424571e-01 9.42129612e-01
1.37190655e-01 9.75075245e-01 -1.06409431e+00 5.22383213e-01
2.30986252e-01 -1.89853191e-01 8.02357972e-01 -8.62976253e-01
-6.13812685e-01 -2.96890825e-01 -5.28781749e-02 -1.56010878e+00
5.90064108e-01 -9.15953338e-01 2.41335005e-01 -5.84740102e-01
-2.67240375e-01 -4.63146359e-01 -3.88981193e-01 6.65072948e-02
-1.75900772e-01 -3.20807844e-01 3.01079690e-01 2.43157923e-01
2.09047765e-01 8.75931084e-01 1.08886039e+00 2.03413349e-02
-5.76261938e-01 2.56837636e-01 -6.18332267e-01 7.21469939e-01
1.01959765e+00 -7.96323597e-01 -6.16713405e-01 -4.30854231e-01
-4.43691880e-01 1.78643078e-01 -2.97672421e-01 -7.05130041e-01
-7.56960958e-02 -1.08113840e-01 -1.55692160e-01 -4.52899694e-01
8.94272029e-01 -9.09880579e-01 2.10884735e-02 3.03326160e-01
-5.00760794e-01 2.60811418e-01 5.04301190e-01 3.64065140e-01
-1.54891565e-01 -2.39106685e-01 6.18927836e-01 2.21725836e-01
-3.64967510e-02 -1.07737236e-01 -7.29815662e-01 2.55890060e-02
4.33936268e-01 -1.57533079e-01 -2.07902789e-01 -5.24133861e-01
-6.50620759e-01 -2.24869907e-01 -2.14216355e-02 1.03165233e+00
1.95526376e-01 -1.56223738e+00 -8.36546063e-01 6.01504087e-01
1.54120801e-02 -8.69361758e-01 1.93093747e-01 5.85053027e-01
1.50779471e-01 8.18539858e-01 -9.88341644e-02 -5.65216005e-01
-1.89576232e+00 5.49776137e-01 2.32262507e-01 -1.59788370e-01
-2.47687884e-02 9.73766744e-01 8.53689849e-01 -1.03081703e+00
2.85694093e-01 -1.54937983e-01 -1.34688124e-01 -1.76773980e-01
2.17385709e-01 1.30584955e-01 -3.67656767e-01 -1.30697048e+00
-4.07718241e-01 8.51115763e-01 -3.84059966e-01 -5.54308236e-01
4.63863105e-01 -7.29167581e-01 3.82357836e-01 1.03490078e+00
1.33629847e+00 1.02206397e+00 -7.32799232e-01 -3.10425073e-01
-2.49262109e-01 -4.99266267e-01 -2.83229411e-01 -6.56175554e-01
-6.35884523e-01 1.31176960e+00 7.76550889e-01 1.43375655e-04
7.81199634e-01 -1.04828022e-01 8.56290162e-01 1.31151333e-01
1.39028579e-01 -1.04612041e+00 -1.41800940e-01 7.68458903e-01
1.31692398e+00 -8.84306550e-01 -4.13634777e-01 -7.46136546e-01
-1.15021539e+00 8.29769731e-01 6.87189341e-01 1.58366799e-01
4.39467162e-01 1.76431134e-01 6.53438330e-01 4.05555010e-01
-7.58867979e-01 4.02367078e-02 5.27120769e-01 8.39663684e-01
7.93755054e-01 5.22720516e-01 -1.83432892e-01 6.58595979e-01
-9.89370346e-01 -7.06326544e-01 1.69550091e-01 1.13953933e-01
-3.01768810e-01 -1.58646536e+00 -7.24960327e-01 -3.66189688e-01
-5.45629799e-01 -5.83988547e-01 -3.61928880e-01 5.49870312e-01
-1.75882116e-01 1.08931398e+00 4.78202924e-02 -6.11671329e-01
4.13003415e-01 5.19226134e-01 3.53198558e-01 -4.45705920e-01
-6.70132697e-01 5.18992603e-01 3.96385908e-01 -4.11642343e-02
1.91413790e-01 -5.27097702e-01 -9.84003544e-01 -3.14014107e-01
-6.08727038e-01 2.90767401e-01 7.88168192e-01 8.19792390e-01
4.97650206e-01 5.77079415e-01 1.04996598e+00 -4.82916653e-01
-7.49878109e-01 -1.13429868e+00 -6.08530819e-01 -2.19451319e-02
3.80251467e-01 -1.76532939e-01 -5.05916715e-01 7.85920620e-02] | [14.790995597839355, 6.687067985534668] |
f4284085-52c0-4227-8282-b3d4d95cb947 | towards-bio-inspired-unsupervised | 2106.09326 | null | https://arxiv.org/abs/2106.09326v1 | https://arxiv.org/pdf/2106.09326v1.pdf | Towards bio-inspired unsupervised representation learning for indoor aerial navigation | Aerial navigation in GPS-denied, indoor environments, is still an open challenge. Drones can perceive the environment from a richer set of viewpoints, while having more stringent compute and energy constraints than other autonomous platforms. To tackle that problem, this research displays a biologically inspired deep-learning algorithm for simultaneous localization and mapping (SLAM) and its application in a drone navigation system. We propose an unsupervised representation learning method that yields low-dimensional latent state descriptors, that mitigates the sensitivity to perceptual aliasing, and works on power-efficient, embedded hardware. The designed algorithm is evaluated on a dataset collected in an indoor warehouse environment, and initial results show the feasibility for robust indoor aerial navigation. | ['Bart Dhoedt', 'Matthias Hartmann', 'Tim Verbelen', 'Ozan Catal', 'Ni Wang'] | 2021-06-17 | null | null | null | null | ['drone-navigation'] | ['computer-vision'] | [ 2.14659378e-01 -3.58414352e-01 1.35957837e-01 -1.28528103e-01
3.94211262e-02 -7.85333395e-01 3.87857705e-01 -1.30785340e-02
-4.52587575e-01 8.26829135e-01 -2.51444638e-01 7.71830380e-02
-4.89008576e-01 -9.17476058e-01 -4.81515288e-01 -8.05710018e-01
-4.34096217e-01 2.54881918e-01 1.68708235e-01 -4.50803548e-01
2.36950219e-01 8.23252201e-01 -1.82038975e+00 -4.34446752e-01
7.42463470e-01 8.55344951e-01 6.11168087e-01 6.30412519e-01
5.80048203e-01 6.78258061e-01 -5.15703857e-01 4.04102445e-01
5.54731488e-01 -1.36836961e-01 -2.00366303e-01 -2.02499926e-01
4.38862979e-01 -2.79484034e-01 -6.02276564e-01 9.22444642e-01
6.97418809e-01 5.08763552e-01 5.22786915e-01 -1.20068693e+00
-4.35242772e-01 -1.60291940e-01 2.21604422e-01 1.36424616e-01
4.14349914e-01 6.86624721e-02 5.17573357e-01 -3.92416000e-01
4.32572991e-01 7.67233670e-01 7.42871523e-01 4.82796431e-01
-1.04008627e+00 -4.95261669e-01 2.06197402e-03 -2.26124480e-01
-1.65620601e+00 -3.75370830e-01 5.13971210e-01 -3.21957916e-01
1.06994307e+00 -3.78937460e-02 9.02717113e-01 1.08025992e+00
1.29888046e+00 2.50713557e-01 8.39940369e-01 7.49889016e-02
8.92469168e-01 -8.98639485e-02 -4.94787276e-01 1.09113860e+00
6.85051978e-01 4.55153078e-01 -6.78048611e-01 -6.49861917e-02
8.11625481e-01 3.46327484e-01 -4.87900496e-01 -1.08095598e+00
-1.22584224e+00 5.88093877e-01 9.28943157e-01 1.69857576e-01
-4.97832716e-01 4.86667335e-01 -4.65976559e-02 1.63550347e-01
-6.58927811e-03 7.45855689e-01 -7.29060054e-01 2.83333263e-03
-9.53928828e-01 1.92489609e-01 8.29803705e-01 1.23392284e+00
7.23391294e-01 6.42986178e-01 5.72128057e-01 1.74982194e-02
4.99807149e-01 9.33161736e-01 9.03270185e-01 -7.80417919e-01
-6.67246655e-02 6.29121900e-01 3.54272336e-01 -1.49556696e+00
-8.28779638e-01 -5.81315577e-01 -7.20175922e-01 4.11142021e-01
-4.86741513e-01 -3.25298309e-01 -8.26136827e-01 1.74382544e+00
2.18828544e-01 1.72480553e-01 5.84002554e-01 1.10962987e+00
7.70887792e-01 7.52671242e-01 -2.02332422e-01 -9.51373833e-04
1.05161035e+00 -6.58093512e-01 -6.08536363e-01 -5.63473701e-01
5.95649660e-01 -2.40051597e-01 5.70373654e-01 4.15921748e-01
-4.07865316e-01 -7.24228799e-01 -1.70716059e+00 -1.82659961e-02
-7.93001831e-01 3.64454657e-01 9.74364698e-01 7.96579897e-01
-1.36607611e+00 4.96587485e-01 -1.11304581e+00 -9.41155374e-01
-1.42515406e-01 8.65668058e-01 -6.03828013e-01 9.05308127e-02
-9.93413210e-01 9.75325763e-01 4.96819019e-01 1.26492545e-01
-1.16543639e+00 -5.34127466e-03 -1.12404382e+00 -3.53240967e-02
1.13779902e-02 -9.19869602e-01 5.85472822e-01 -3.20825130e-01
-1.75335848e+00 4.84703064e-01 1.12642072e-01 -6.78751767e-01
1.44252200e-02 -3.90016228e-01 -3.92838001e-01 -1.28406808e-01
-2.62838905e-03 8.26537549e-01 8.34662557e-01 -1.10222340e+00
-5.25690794e-01 -5.94516098e-01 1.75811216e-01 5.83740294e-01
-2.20913608e-02 -8.68698239e-01 2.85273999e-01 -1.67912066e-01
8.16420317e-01 -1.25612640e+00 -6.03101492e-01 6.29995242e-02
3.38805407e-01 3.38818431e-01 8.52380097e-01 -1.20883361e-01
4.37736362e-01 -2.03341722e+00 3.74389797e-01 -2.22807080e-01
-3.48666199e-02 2.51829270e-02 1.56219319e-01 6.05673969e-01
4.94471550e-01 -4.99814898e-01 -9.65528339e-02 -1.29400283e-01
-1.51190564e-01 5.95569134e-01 -5.19885004e-01 8.81555438e-01
-2.40202501e-01 5.45919418e-01 -1.04173923e+00 -1.46317288e-01
4.27634358e-01 3.56627554e-01 -4.39060569e-01 -4.05688249e-02
1.02102675e-01 6.10640287e-01 -3.69841933e-01 8.76406789e-01
5.91220260e-01 2.41920069e-01 1.04273871e-01 -1.99588224e-01
-5.98311543e-01 6.57456985e-04 -1.15403104e+00 2.39578009e+00
-6.55037522e-01 9.74726379e-01 9.94203463e-02 -5.20228148e-01
1.35000646e+00 -1.14474371e-01 4.38663781e-01 -6.02001309e-01
5.28838873e-01 9.32542756e-02 -2.56351024e-01 -2.33018294e-01
9.55375731e-01 4.87393476e-02 -4.97645020e-01 2.76699681e-02
6.02654397e-01 -5.66006720e-01 -2.90645987e-01 -6.11650310e-02
1.20926309e+00 6.28377616e-01 4.55575049e-01 -6.57286942e-01
4.08974707e-01 1.65483356e-01 5.63927591e-01 5.63450694e-01
-3.55223089e-01 7.88666606e-02 -3.53777736e-01 -9.91175354e-01
-4.24296439e-01 -1.30398035e+00 4.49862257e-02 6.17059588e-01
1.04057670e+00 -4.60603744e-01 -4.81000096e-01 -1.92180529e-01
9.84997749e-02 5.57772577e-01 -1.81313962e-01 -5.83457410e-01
-1.86624870e-01 -5.87134600e-01 4.54897612e-01 1.59642026e-01
5.20972848e-01 -7.20406651e-01 -1.80076706e+00 1.85651094e-01
9.78973955e-02 -1.09020710e+00 4.86395538e-01 7.30484903e-01
-9.46295321e-01 -7.50819862e-01 -1.42998189e-01 -8.44594002e-01
6.53877079e-01 5.05730927e-01 7.72891104e-01 -3.10387403e-01
-5.54355800e-01 6.79031610e-01 -2.49025136e-01 -3.12196761e-01
4.24794734e-01 -9.37680621e-03 1.01752663e+00 -3.60890567e-01
1.87964439e-01 -8.23449075e-01 -3.58451098e-01 3.68865401e-01
-4.43332821e-01 -2.03441188e-01 6.64885402e-01 6.10399783e-01
7.63195693e-01 5.07645071e-01 2.05568224e-01 1.59938969e-02
3.20507675e-01 -3.22890252e-01 -1.23583186e+00 -2.14246735e-01
-4.18246806e-01 5.68876117e-02 7.36987472e-01 -9.86424387e-02
-4.94829804e-01 6.38635814e-01 3.25901210e-01 -2.49367952e-02
-3.30357671e-01 3.04768920e-01 -2.66748756e-01 -7.85724163e-01
6.82425976e-01 7.34358251e-01 -3.59218687e-01 -3.39030325e-02
3.73443246e-01 3.60884279e-01 6.91464126e-01 -2.82076359e-01
9.60512638e-01 5.70967019e-01 4.62102860e-01 -1.22538292e+00
-2.47486681e-01 -3.10937583e-01 -1.00943089e+00 -2.59559423e-01
6.98274672e-01 -1.38961709e+00 -5.99101961e-01 -2.37424951e-02
-8.91053259e-01 -2.23319441e-01 -1.76528603e-01 8.50277007e-01
-8.82581651e-01 5.16652614e-02 1.04788586e-01 -9.43054855e-01
-1.49658233e-01 -9.33359802e-01 8.95596564e-01 5.73906302e-01
-8.93878788e-02 -6.89178288e-01 5.54229021e-01 -1.42800301e-01
3.04299027e-01 4.73061532e-01 3.83229613e-01 -4.25317794e-01
-1.16680384e+00 -3.46370786e-01 2.54693806e-01 -1.57277673e-01
2.04662398e-01 -2.06084967e-01 -9.20791507e-01 -6.33686304e-01
2.15304002e-01 -3.30511421e-01 6.10190511e-01 2.82726645e-01
3.67796659e-01 2.79681459e-02 -6.81241214e-01 9.94567454e-01
1.84740496e+00 4.39723194e-01 3.04994673e-01 2.50533581e-01
3.59712720e-01 2.37702340e-01 7.92156696e-01 4.77935255e-01
2.68420726e-01 3.41771275e-01 8.83512974e-01 1.21119261e-01
4.40115541e-01 -4.22604054e-01 4.43647087e-01 9.38102782e-01
1.39628083e-01 -7.12578118e-01 -6.73235655e-01 4.19761360e-01
-1.87373602e+00 -4.58135337e-01 3.15988570e-01 2.02377439e+00
-2.54173219e-01 4.73274216e-02 -7.34046519e-01 1.46006420e-02
2.04128698e-01 2.35344306e-01 -7.51650989e-01 -3.45294982e-01
-1.93222180e-01 -1.24001103e-02 9.47726429e-01 4.06869262e-01
-1.25183272e+00 1.26968813e+00 6.71231747e+00 1.39839351e-01
-1.22573698e+00 -8.58362317e-02 -4.08248514e-01 -5.84176518e-02
1.95605367e-01 6.70192689e-02 -6.98246717e-01 2.12568447e-01
1.08188832e+00 1.05311736e-01 4.15360093e-01 1.34196508e+00
-7.25631043e-02 -5.23024201e-01 -1.03236616e+00 1.29726934e+00
3.55897129e-01 -1.08391511e+00 -1.08433895e-01 2.52298892e-01
6.87868237e-01 3.32323164e-01 1.28054217e-01 2.69193411e-01
1.92769155e-01 -7.66701102e-01 6.35018110e-01 6.11285031e-01
3.88952583e-01 -9.41714644e-01 9.61607158e-01 6.48495197e-01
-1.42367244e+00 -4.60203022e-01 -1.02240050e+00 -5.90155303e-01
1.20699987e-01 2.32718989e-01 -7.92539239e-01 5.56616366e-01
7.65385509e-01 7.05076814e-01 -5.89245439e-01 1.13759530e+00
-2.69937485e-01 -1.94193363e-01 -4.43843007e-01 -4.52539355e-01
2.35541105e-01 -2.93710142e-01 5.04190326e-01 7.18584001e-01
7.85124123e-01 1.88166246e-01 4.32816863e-01 7.14009225e-01
2.06176385e-01 -2.16278657e-01 -1.41084409e+00 -1.86126456e-02
3.32498938e-01 1.23825502e+00 -1.02024627e+00 1.49676464e-02
1.57937244e-01 1.18657923e+00 -3.96659002e-02 1.36873931e-01
-9.46387768e-01 -4.66319948e-01 8.02325726e-01 -3.72904032e-01
7.61534795e-02 -9.93847132e-01 1.83699846e-01 -1.11806953e+00
-3.60494077e-01 -2.42377803e-01 -2.37986296e-01 -1.18216348e+00
-5.22874475e-01 7.88202882e-01 -4.12153691e-01 -1.52444637e+00
-1.48350820e-01 -9.00588512e-01 4.09102626e-02 5.55002391e-01
-1.31492472e+00 -1.09326732e+00 -7.76551902e-01 5.48935950e-01
5.72767615e-01 -5.24508297e-01 1.40511441e+00 -4.65889182e-03
-2.88264781e-01 -2.70433282e-03 2.02133104e-01 -4.42495286e-01
4.21334684e-01 -1.05825520e+00 3.86084467e-01 1.03719449e+00
4.11259860e-01 6.82078719e-01 8.34418535e-01 -7.12426126e-01
-2.02075028e+00 -9.96046782e-01 4.67719644e-01 -4.65923935e-01
2.88640678e-01 -5.96070230e-01 -1.90267578e-01 6.17460966e-01
1.57694449e-05 4.05350253e-02 6.90343082e-01 -3.78467798e-01
3.12885433e-01 -1.83661222e-01 -1.46286273e+00 6.51131570e-01
1.23861933e+00 -4.75510448e-01 -6.40765965e-01 3.68443906e-01
7.74802864e-01 -4.68510419e-01 -5.36972106e-01 4.86383110e-01
6.13650739e-01 -1.03696823e+00 8.76620293e-01 -1.66407913e-01
-4.18193936e-01 -1.01020849e+00 -6.93089664e-01 -1.26077211e+00
-7.05049753e-01 -7.38904655e-01 -2.23475099e-02 7.38312781e-01
-7.69106001e-02 -6.32538915e-01 8.52787733e-01 1.42239422e-01
-3.03048998e-01 -3.60632628e-01 -1.21916592e+00 -8.06973219e-01
-8.05803597e-01 1.22901335e-01 3.56411129e-01 5.98378479e-01
-2.71094590e-01 9.59054902e-02 -4.63382810e-01 1.05624211e+00
4.33404446e-01 1.05084844e-01 9.49933410e-01 -1.62423849e+00
3.00139248e-01 4.08388883e-01 -1.29645765e+00 -1.16214788e+00
2.81507522e-01 -5.81675172e-01 4.71769840e-01 -1.69634724e+00
-7.87562966e-01 -1.71252444e-01 -2.56741375e-01 2.61923343e-01
7.39999473e-01 1.51663870e-01 -2.90962279e-01 8.69705155e-02
-6.95383072e-01 9.64492023e-01 6.00183666e-01 -2.28746943e-02
-2.09443495e-01 -1.43773690e-01 -1.40562654e-01 8.25085342e-01
7.77677953e-01 -6.36025846e-01 -8.55538249e-01 -4.27682042e-01
1.24100178e-01 -1.07031681e-01 2.55874842e-01 -2.11955285e+00
7.05939889e-01 -3.78331989e-02 8.57723057e-01 -1.01480949e+00
8.82489145e-01 -1.29569423e+00 1.15595780e-01 1.11986434e+00
2.25199714e-01 4.94506627e-01 3.76317084e-01 8.77232552e-01
-8.06114152e-02 1.15714304e-03 6.34667158e-01 -2.83408582e-01
-1.28962994e+00 1.81816757e-01 -1.02224207e+00 -6.46192491e-01
1.11484993e+00 -4.39739823e-01 -5.55864312e-02 -8.81963521e-02
-2.93379724e-01 -2.88346130e-03 8.83297205e-01 4.19045806e-01
9.13031638e-01 -1.28086674e+00 1.04459472e-01 8.46539617e-01
8.29650387e-02 -1.42116189e-01 2.42835462e-01 3.67887229e-01
-1.17806935e+00 7.55965412e-01 -8.81160259e-01 -7.40715325e-01
-7.43862033e-01 7.36390173e-01 3.36976528e-01 3.70288491e-01
-4.44485217e-01 9.58760917e-01 -9.08563733e-02 -7.34514594e-01
-3.66036943e-03 -6.91425920e-01 -1.47364601e-01 -1.63850039e-01
2.28908464e-01 1.92586064e-01 -7.40013272e-02 -7.91200221e-01
-7.35683441e-01 8.82569849e-01 7.15913892e-01 -9.59449634e-02
1.11127925e+00 -4.60084409e-01 7.22229704e-02 4.81879264e-01
6.66051984e-01 4.39999029e-02 -1.12754118e+00 2.12404072e-01
-7.30709285e-02 -4.34910923e-01 3.35237801e-01 -5.13392687e-01
-6.07954621e-01 6.57718658e-01 1.28655910e+00 -3.09976116e-02
9.64319110e-01 -7.43210614e-01 4.88313973e-01 1.39736485e+00
1.29448032e+00 -8.45385551e-01 -6.24798238e-02 7.19059646e-01
5.15007854e-01 -1.18457031e+00 2.33933255e-01 3.82611714e-02
-2.48379514e-01 1.23033309e+00 6.98373020e-01 -6.49285257e-01
5.18860161e-01 2.46872872e-01 3.62922847e-02 -1.41258135e-01
-6.56798542e-01 -1.95141539e-01 -1.55886844e-01 1.04284620e+00
-3.05752177e-02 -7.24561652e-03 2.16667145e-01 4.92097139e-01
-5.68044007e-01 -1.04339518e-01 6.10381901e-01 1.46695399e+00
-9.53961074e-01 -4.79804248e-01 -1.18745238e-01 -4.45161819e-01
2.94072956e-01 6.88700899e-02 -3.47800136e-01 8.43002200e-01
7.07241118e-01 8.87697458e-01 -2.30731256e-02 -6.78630114e-01
3.19484562e-01 -3.02860171e-01 5.99909604e-01 -4.92642462e-01
-1.56432912e-01 -1.86705291e-01 -3.00779462e-01 -7.83866227e-01
-3.86233538e-01 -1.25782102e-01 -1.27034616e+00 1.16792016e-01
-3.63657802e-01 1.79602355e-01 1.20424700e+00 7.12041438e-01
5.71073949e-01 6.33098185e-01 4.58769888e-01 -1.23738647e+00
-5.61183803e-02 -4.01269466e-01 -7.75597215e-01 -5.84527433e-01
5.43959200e-01 -1.15469921e+00 -3.67428333e-01 -1.64190575e-01] | [7.330127239227295, -1.9147083759307861] |
db61bd60-b148-4809-aa46-37452a318009 | improving-the-intent-classification-accuracy | 2303.06585 | null | https://arxiv.org/abs/2303.06585v1 | https://arxiv.org/pdf/2303.06585v1.pdf | Improving the Intent Classification accuracy in Noisy Environment | Intent classification is a fundamental task in the spoken language understanding field that has recently gained the attention of the scientific community, mainly because of the feasibility of approaching it with end-to-end neural models. In this way, avoiding using intermediate steps, i.e. automatic speech recognition, is possible, thus the propagation of errors due to background noise, spontaneous speech, speaking styles of users, etc. Towards the development of solutions applicable in real scenarios, it is interesting to investigate how environmental noise and related noise reduction techniques to address the intent classification task with end-to-end neural models. In this paper, we experiment with a noisy version of the fluent speech command data set, combining the intent classifier with a time-domain speech enhancement solution based on Wave-U-Net and considering different training strategies. Experimental results reveal that, for this task, the use of speech enhancement greatly improves the classification accuracy in noisy conditions, in particular when the classification model is trained on enhanced signals. | ['Daniele Falavigna', 'Alessio Brutti', 'Mohamed Nabih Ali'] | 2023-03-12 | null | null | null | null | ['spoken-language-understanding', 'intent-classification', 'spoken-language-understanding', 'speech-enhancement'] | ['natural-language-processing', 'natural-language-processing', 'speech', 'speech'] | [ 3.14250588e-01 1.54361427e-01 6.82258666e-01 -5.17118156e-01
-4.58109379e-01 -1.41230956e-01 5.54925382e-01 1.91281170e-01
-8.89551759e-01 4.16247278e-01 4.58617985e-01 -1.13913499e-01
-2.32424557e-01 -5.96495807e-01 -1.83517471e-01 -6.69711113e-01
-9.81662497e-02 3.12179267e-01 -1.29543602e-01 -4.75741118e-01
1.57772720e-01 4.38019693e-01 -2.04504585e+00 2.69410044e-01
7.25963473e-01 9.78451967e-01 4.56180781e-01 7.61313260e-01
-2.53743201e-01 7.63840556e-01 -8.19628537e-01 4.89183217e-02
-6.16974868e-02 -3.17657977e-01 -6.08553171e-01 9.23490245e-03
-5.31060100e-02 -1.21423058e-01 2.75538359e-02 9.15547431e-01
7.11634338e-01 5.29813051e-01 4.99204636e-01 -6.73947573e-01
9.37018916e-02 5.82928479e-01 3.71846348e-01 1.10759757e-01
6.02104247e-01 -1.83802322e-02 7.00844646e-01 -6.69779420e-01
3.43070298e-01 9.76266682e-01 4.96826470e-01 3.45769733e-01
-1.28905571e+00 -2.40033209e-01 3.88853811e-02 4.04331535e-01
-1.24178958e+00 -7.41221011e-01 9.61164474e-01 -3.89887869e-01
1.04114449e+00 3.59023362e-01 3.05567294e-01 1.32123399e+00
-9.45830345e-02 6.60148144e-01 1.06377459e+00 -7.55767226e-01
6.54871523e-01 4.84007329e-01 3.95719439e-01 1.09455355e-01
-3.29658806e-01 1.92148566e-01 -6.04877353e-01 9.21151713e-02
-2.36071661e-01 -4.43226337e-01 -6.07204437e-01 3.65959629e-02
-5.30694842e-01 6.01185143e-01 1.32872805e-01 1.00383055e+00
-7.54726231e-01 -2.46857345e-01 5.61408103e-01 3.80906373e-01
9.38398719e-01 2.32858673e-01 -3.64224076e-01 -5.00138104e-01
-1.14320576e+00 -1.61149111e-02 9.86593246e-01 2.69415915e-01
4.52671826e-01 3.99136037e-01 4.55684960e-02 1.10317683e+00
2.54222006e-01 1.88602433e-01 6.91003025e-01 -3.60821217e-01
2.37172201e-01 4.55232710e-01 -2.08094031e-01 -8.14616621e-01
-6.48696065e-01 -8.53231609e-01 -8.15729320e-01 3.94010752e-01
1.56126902e-01 -3.97693574e-01 -7.19645143e-01 1.70263410e+00
1.41674399e-01 3.17548104e-02 5.38322389e-01 8.50695908e-01
6.06368124e-01 8.67126703e-01 9.71698314e-02 -4.04576838e-01
1.22356629e+00 -5.07888973e-01 -1.04914522e+00 -3.15859705e-01
6.12442791e-01 -7.24380732e-01 8.88259947e-01 7.84187138e-01
-7.27855444e-01 -7.78766870e-01 -8.62595677e-01 3.16664338e-01
-5.68276286e-01 2.50752270e-01 1.83068812e-02 9.62711990e-01
-1.05355394e+00 6.47425354e-01 -5.25167406e-01 -3.89555901e-01
-3.10461730e-01 3.46778184e-01 -3.64893079e-01 1.88446045e-01
-1.33845460e+00 1.01324928e+00 5.40016472e-01 3.71490121e-01
-3.83731335e-01 -4.24775511e-01 -7.43049681e-01 3.90206188e-01
1.51425451e-01 -1.48959458e-01 1.15403247e+00 -1.00697672e+00
-1.80897880e+00 4.25920278e-01 -1.06500424e-01 -5.53710341e-01
4.99645561e-01 -3.53277266e-01 -5.67279041e-01 -8.74643400e-02
-3.34612757e-01 2.23723426e-01 8.67478848e-01 -1.15947676e+00
-5.19095540e-01 -4.31448251e-01 -4.20918494e-01 2.60101169e-01
-6.64863884e-01 -5.20876900e-04 1.56815201e-01 -5.02089500e-01
2.79590152e-02 -6.08311117e-01 -5.69241308e-02 -4.30578440e-01
-9.90810692e-02 -1.88896760e-01 8.00189376e-01 -9.90610480e-01
1.06525588e+00 -2.43193150e+00 1.43924400e-01 3.59257311e-01
-3.74659121e-01 5.36466718e-01 3.64991426e-02 6.02796555e-01
-2.77880281e-01 -2.00173035e-01 -2.82186180e-01 -6.79151595e-01
-5.36783449e-02 1.02883078e-01 -9.51014385e-02 2.90397286e-01
-1.96851008e-02 4.93812934e-02 -4.31498766e-01 -1.25827774e-01
5.53708494e-01 7.12797105e-01 -3.60034436e-01 4.85485256e-01
-1.34917498e-01 4.06297952e-01 -1.89717159e-01 -7.41394758e-02
5.40134370e-01 5.33381820e-01 9.48005468e-02 -1.66047752e-01
-5.47090709e-01 4.49920207e-01 -1.34766710e+00 1.58345699e+00
-9.83211815e-01 8.55455399e-01 5.32880306e-01 -1.24548435e+00
1.24495375e+00 7.70253837e-01 2.58209467e-01 -8.71542096e-01
4.95350748e-01 3.67720634e-01 9.18321013e-02 -6.72802925e-01
3.81711572e-01 -1.50180832e-01 3.18032742e-01 2.16111839e-01
2.18628779e-01 -5.43681718e-02 -1.10504963e-01 -4.90287364e-01
7.95495272e-01 -1.68284088e-01 1.84670895e-01 -2.01787800e-01
7.93431699e-01 -3.39946896e-01 -5.64996265e-02 5.49829781e-01
-1.50445141e-02 4.59921747e-01 3.87905687e-02 -1.68886691e-01
-6.33752882e-01 -5.35482883e-01 -1.06405385e-01 9.45750117e-01
-4.38029915e-01 -3.77659984e-02 -9.55508232e-01 -1.45201519e-01
-5.49931109e-01 1.18164396e+00 -2.77608812e-01 -3.25185716e-01
-3.71351391e-01 -6.31752968e-01 4.28247184e-01 -2.51611564e-02
5.76414585e-01 -1.15447259e+00 -5.84267616e-01 6.68390334e-01
-3.73375744e-01 -1.03327966e+00 1.11721680e-01 7.66864002e-01
-8.13202977e-01 -4.18984234e-01 -7.34747946e-01 -5.55063426e-01
1.21025719e-01 -9.18991044e-02 6.83685839e-01 2.61960309e-02
-3.03607155e-02 4.39134687e-01 -7.79780686e-01 -4.23912704e-01
-9.89105523e-01 1.56430721e-01 1.19218431e-01 4.19690281e-01
4.20457304e-01 -6.50569201e-01 -9.20365229e-02 1.51470497e-01
-9.94422793e-01 -1.40459880e-01 3.99763167e-01 7.05845952e-01
-1.87907025e-01 2.55741745e-01 6.50672853e-01 -3.34452569e-01
9.39838886e-01 -1.21089093e-01 -4.58348423e-01 6.01446740e-02
-4.16221231e-01 3.75249386e-02 7.51721382e-01 -2.25426346e-01
-1.32396984e+00 4.37475704e-02 -1.08937824e+00 2.66611464e-02
-8.85523379e-01 6.01717055e-01 -3.32150668e-01 -2.29144413e-02
8.54993343e-01 5.36260188e-01 1.93255827e-01 -6.18191302e-01
-4.89763869e-03 1.34035039e+00 2.34811500e-01 -1.09362090e-02
4.06180233e-01 -5.19041903e-02 -3.05537015e-01 -1.77326882e+00
-4.33434904e-01 -7.89674997e-01 -3.22990596e-01 -5.53514123e-01
7.47136235e-01 -6.89534247e-01 -5.16934574e-01 7.31258690e-01
-1.37309444e+00 -2.40983203e-01 -1.77960962e-01 8.20427656e-01
-5.27130723e-01 3.57256174e-01 -3.50596666e-01 -1.36342025e+00
-5.14361918e-01 -1.28076184e+00 6.79294825e-01 1.66398630e-01
-3.76414418e-01 -9.36492622e-01 -2.42931470e-02 4.01484817e-01
6.76887095e-01 -3.44116896e-01 6.96791530e-01 -9.90779638e-01
-8.87030959e-02 -2.57170230e-01 4.64422256e-01 9.18779850e-01
1.50690839e-01 -2.78661579e-01 -1.58168232e+00 2.44506896e-02
6.48047507e-01 -1.28210694e-01 6.04637980e-01 3.36158067e-01
6.70327961e-01 -1.00959070e-01 1.56101003e-01 1.02742508e-01
1.13610113e+00 4.99524146e-01 7.86848724e-01 5.31812310e-02
3.42875905e-02 9.91862178e-01 4.79363322e-01 3.95976394e-01
-1.70320332e-01 8.75127614e-01 3.55732262e-01 -1.91617206e-01
-1.28466338e-01 2.01885313e-01 3.22819382e-01 7.99637556e-01
6.56473860e-02 -4.91641134e-01 -7.85749674e-01 4.09109592e-01
-1.42749977e+00 -9.43811417e-01 -1.19636461e-01 2.15307641e+00
3.73419076e-01 2.51457274e-01 -1.37312010e-01 7.55640626e-01
5.43454587e-01 1.93265423e-01 6.14686124e-02 -6.44663393e-01
7.92064369e-02 1.56959057e-01 -5.85999414e-02 6.92608118e-01
-8.79283130e-01 3.56566310e-01 4.89403296e+00 7.81513453e-01
-1.34788096e+00 6.76424131e-02 4.93748575e-01 2.37661853e-01
-8.76835212e-02 -4.43351090e-01 -5.37635207e-01 4.43824857e-01
1.26296616e+00 3.80972505e-01 4.34520543e-01 8.49630058e-01
7.57582784e-01 -3.19610357e-01 -8.57714295e-01 7.60864735e-01
3.18888277e-01 -7.79202282e-01 -4.07203794e-01 -2.74049938e-01
2.35693660e-02 -7.14902952e-02 -1.62457407e-01 2.73056537e-01
-4.76791203e-01 -7.26316690e-01 7.50548542e-01 4.14888084e-01
9.77135152e-02 -7.54285932e-01 1.06367099e+00 8.36732626e-01
-8.81118953e-01 -1.23247892e-01 1.27379745e-01 -1.56000063e-01
4.55703586e-01 8.55283022e-01 -1.07888138e+00 4.77236807e-01
5.31905353e-01 8.13391209e-02 -7.84865543e-02 9.58374262e-01
-7.65544251e-02 8.57516408e-01 -7.04365253e-01 -4.99525666e-01
1.37415051e-01 -2.36235842e-01 8.20959032e-01 1.36399877e+00
4.16302174e-01 3.43450978e-02 -2.55003899e-01 7.36494362e-01
3.32589924e-01 3.69356692e-01 -6.97931826e-01 6.28952980e-02
1.33294398e-02 9.82331872e-01 -6.41270339e-01 8.35797489e-02
-1.03644311e-01 9.79947209e-01 -6.61937743e-02 3.82834136e-01
-3.38015139e-01 -4.20146078e-01 5.62343359e-01 -1.12968718e-03
1.03121944e-01 -2.62226135e-01 -1.60555512e-01 -6.60322666e-01
1.71410605e-01 -8.03023934e-01 -1.60057142e-01 -7.48231351e-01
-7.47913241e-01 8.95135522e-01 -1.16005003e-01 -9.80330884e-01
-5.40439069e-01 -6.56419456e-01 -7.09713399e-01 1.09069335e+00
-1.28443420e+00 -6.97857738e-01 -1.13241829e-01 2.39176914e-01
8.17726433e-01 -1.83856964e-01 9.90639567e-01 6.03204012e-01
-2.53793478e-01 2.78599739e-01 9.91963893e-02 -1.80507913e-01
3.30902368e-01 -8.40722322e-01 1.41811013e-01 8.79431427e-01
1.92928225e-01 3.06663334e-01 1.24695110e+00 -3.17692965e-01
-8.53999674e-01 -6.52362049e-01 1.15579128e+00 3.30779791e-01
3.63192379e-01 -6.56388462e-01 -1.13284540e+00 4.49090973e-02
3.30190301e-01 -4.74091560e-01 5.81188321e-01 2.46514663e-01
3.15154910e-01 -1.88221991e-01 -1.03527963e+00 5.90403974e-01
5.99119127e-01 -6.68404460e-01 -7.79533088e-01 1.38183177e-01
4.88675803e-01 2.78721447e-03 -5.68597913e-01 1.85457975e-01
2.11804852e-01 -1.24388862e+00 7.54173875e-01 -3.64234187e-02
-1.21107295e-01 -5.98153360e-02 -1.88550651e-01 -1.71129453e+00
1.93249375e-01 -3.24712545e-01 4.30563897e-01 1.36232710e+00
5.28333604e-01 -5.90135992e-01 6.59895480e-01 3.94507706e-01
-2.95160919e-01 -2.54054815e-01 -1.29897952e+00 -4.81099218e-01
-3.20163399e-01 -8.09565246e-01 9.80371907e-02 3.42719674e-01
6.66309819e-02 3.19524974e-01 -3.98171455e-01 3.54492813e-01
2.59024352e-01 -3.01271170e-01 4.61277187e-01 -1.32538760e+00
-3.30522507e-01 -3.42323124e-01 -4.43747967e-01 -8.52301359e-01
2.39171743e-01 -3.89342159e-01 5.64625263e-01 -1.34909201e+00
-7.08353579e-01 -2.23307669e-01 1.77171022e-01 1.94321036e-01
2.81206165e-02 -2.02142626e-01 2.23872244e-01 -3.80722851e-01
-3.40073593e-02 9.25713956e-01 5.39006650e-01 -8.48953053e-02
-3.22316170e-01 4.98880506e-01 -6.83673797e-03 7.15026379e-01
8.05487156e-01 -4.13713962e-01 -5.35629332e-01 1.58908051e-02
-4.54664379e-02 2.51109749e-01 1.75015256e-01 -1.49042237e+00
3.37825209e-01 3.35364580e-01 -1.04952253e-01 -5.44283509e-01
8.00806463e-01 -1.22988832e+00 2.33515799e-01 5.63942313e-01
-3.98577005e-01 -4.62245524e-01 3.11019331e-01 2.75030643e-01
-5.53156853e-01 -8.86505902e-01 6.80272460e-01 9.07982215e-02
-8.12792242e-01 -4.66819495e-01 -1.13646424e+00 -3.55320543e-01
6.01440728e-01 -3.06202412e-01 5.99006228e-02 -7.40365505e-01
-9.70004618e-01 -2.05502078e-01 -2.12894708e-01 3.56414676e-01
4.22704995e-01 -5.41070938e-01 -6.78813398e-01 3.73018116e-01
5.90573587e-02 -2.29971007e-01 6.36830330e-01 6.79785192e-01
-2.36356899e-01 3.74937057e-01 6.63301647e-02 -4.98617202e-01
-1.50191784e+00 2.26040781e-01 6.26505673e-01 -6.12592585e-02
-4.09039766e-01 5.33214271e-01 -3.21219712e-01 -4.88342136e-01
6.44044399e-01 -4.03786421e-01 -6.45774424e-01 2.79115468e-01
5.23678064e-01 4.06164110e-01 8.11518669e-01 -5.17854989e-01
-2.86608666e-01 2.86781400e-01 2.09465414e-01 -4.15749311e-01
1.33616567e+00 -2.02880442e-01 2.63463467e-01 5.56622386e-01
1.19117367e+00 1.64473951e-02 -7.41971552e-01 1.08731315e-01
2.19638079e-01 1.04654960e-01 5.06557167e-01 -8.92385483e-01
-7.09851623e-01 9.95441854e-01 1.02279043e+00 5.85882723e-01
1.37021303e+00 -4.79087919e-01 3.07194084e-01 6.17596865e-01
4.31362867e-01 -1.20905101e+00 -1.95035025e-01 6.93025827e-01
9.51629043e-01 -1.09401214e+00 -4.50330794e-01 -2.99785525e-01
-3.98643911e-01 1.26246846e+00 9.43036079e-02 2.56525904e-01
7.77956545e-01 4.96010870e-01 4.15627629e-01 1.24397501e-01
-4.14240301e-01 -4.90162849e-01 2.60609668e-03 6.18724048e-01
4.20613855e-01 -8.57675672e-02 -3.66366506e-01 5.31227708e-01
-1.70123160e-01 1.52331233e-01 4.61803794e-01 6.82231069e-01
-6.20236397e-01 -1.16401446e+00 -6.17048860e-01 2.47855484e-01
-2.44528025e-01 -9.97009501e-02 -2.46084988e-01 4.94640261e-01
1.59057692e-01 1.47055328e+00 -8.69917050e-02 -4.50836867e-01
7.08541393e-01 5.27066469e-01 -1.23028614e-01 -4.54957366e-01
-9.81364369e-01 9.16053653e-02 3.94150019e-01 -1.21468633e-01
-4.86030579e-01 -6.02727234e-01 -1.01305056e+00 8.95979330e-02
-4.89065379e-01 4.24683541e-01 1.29065621e+00 1.19661152e+00
1.42907843e-01 8.29243720e-01 6.35205388e-01 -9.48805690e-01
-6.30541146e-01 -1.39911926e+00 -6.05243802e-01 1.84143826e-01
4.70378399e-01 -3.68569732e-01 -6.41451836e-01 -1.20526962e-01] | [14.846537590026855, 5.899835109710693] |
604ef74a-4043-42d0-9c74-a9d52f852afd | better-early-than-late-fusing-topics-with | 2007.11314 | null | https://arxiv.org/abs/2007.11314v1 | https://arxiv.org/pdf/2007.11314v1.pdf | Better Early than Late: Fusing Topics with Word Embeddings for Neural Question Paraphrase Identification | Question paraphrase identification is a key task in Community Question Answering (CQA) to determine if an incoming question has been previously asked. Many current models use word embeddings to identify duplicate questions, but the use of topic models in feature-engineered systems suggests that they can be helpful for this task, too. We therefore propose two ways of merging topics with word embeddings (early vs. late fusion) in a new neural architecture for question paraphrase identification. Our results show that our system outperforms neural baselines on multiple CQA datasets, while an ablation study highlights the importance of topics and especially early topic-embedding fusion in our architecture. | ['Nicole Peinelt', 'Dong Nguyen', 'Maria Liakata'] | 2020-07-22 | null | null | null | null | ['paraphrase-identification'] | ['natural-language-processing'] | [-1.59537923e-02 1.64618358e-01 2.17378005e-01 -3.58660638e-01
-1.33302057e+00 -6.41133010e-01 8.06469500e-01 6.62767470e-01
-4.82687593e-01 2.62390733e-01 9.68619525e-01 -4.91362691e-01
-1.83987632e-01 -7.95309007e-01 -3.69414866e-01 -2.28320435e-02
5.81252158e-01 5.32305121e-01 5.19344270e-01 -3.34143907e-01
4.76139367e-01 -7.79355988e-02 -1.18570077e+00 5.58405519e-01
9.45589662e-01 6.47696733e-01 2.28269887e-03 1.03323650e+00
-7.55568624e-01 9.47727680e-01 -1.01053178e+00 -7.06956387e-01
-1.49165317e-01 -4.31926757e-01 -1.51700115e+00 -4.34022337e-01
1.18929958e+00 -4.71781790e-01 -5.42960584e-01 5.32582700e-01
5.56222439e-01 3.49469304e-01 6.35723472e-01 -1.29164362e+00
-1.38424182e+00 1.97316289e-01 1.32554993e-01 7.74807692e-01
7.39813745e-01 1.92152299e-02 1.62627530e+00 -1.03202605e+00
4.33339655e-01 1.25517333e+00 8.17450583e-01 4.91936833e-01
-1.28376722e+00 -3.44232708e-01 -6.85333535e-02 6.69605196e-01
-8.64959836e-01 -4.07223910e-01 7.07750499e-01 -4.80749577e-01
1.56320095e+00 3.81103158e-01 4.86809015e-01 1.21915019e+00
1.83733523e-01 7.98632085e-01 8.31124365e-01 -3.83131713e-01
2.38223895e-01 2.72182584e-01 1.02226067e+00 3.17110628e-01
3.07143807e-01 -6.11807406e-01 -7.31723845e-01 -9.59829330e-01
1.51274446e-02 2.60958709e-02 -3.91550124e-01 -1.41976491e-01
-9.66496706e-01 1.41629696e+00 3.37834924e-01 2.84180969e-01
-2.48133346e-01 3.16352069e-01 3.91985446e-01 7.67271042e-01
4.48274702e-01 1.34132755e+00 -2.99107194e-01 -2.08222285e-01
-1.00545275e+00 7.49382019e-01 1.04970002e+00 7.13877022e-01
7.65518069e-01 -6.63278341e-01 -8.68580043e-01 1.26687765e+00
2.32866466e-01 -1.64327547e-02 7.76953459e-01 -1.16972804e+00
4.09205168e-01 8.41456950e-01 1.04190767e-01 -1.02357221e+00
-1.86710596e-01 -1.60865173e-01 -1.83741957e-01 -6.18674159e-01
4.44256186e-01 -5.27782142e-02 -5.84032476e-01 1.53411484e+00
1.83579456e-02 -1.95660237e-02 1.34336064e-02 4.99752343e-01
1.26329708e+00 7.46606350e-01 1.60449848e-01 3.54384154e-01
1.95116484e+00 -1.25205922e+00 -8.41768384e-01 -5.10589361e-01
5.73447883e-01 -8.19612563e-01 1.27923393e+00 -1.70949250e-01
-1.00044191e+00 -5.61083913e-01 -7.26884186e-01 -8.26221764e-01
-6.12230718e-01 -3.11958224e-01 4.16230232e-01 8.44895840e-01
-1.38143778e+00 3.12558889e-01 -2.39576817e-01 -9.70811486e-01
4.39187974e-01 -7.62673318e-02 -6.32886440e-02 -3.37399065e-01
-1.47855949e+00 9.33824539e-01 -1.62593678e-01 -4.85673338e-01
-4.72570330e-01 -1.00084865e+00 -7.32809424e-01 4.41257417e-01
-2.85477191e-03 -1.15270281e+00 1.65692306e+00 -2.76702523e-01
-1.01525831e+00 9.65641320e-01 -6.66115046e-01 -4.36644405e-01
-2.72775322e-01 -4.72861528e-01 -1.40378475e-01 5.59207380e-01
4.30440217e-01 8.44130397e-01 9.24527109e-01 -5.80017030e-01
-4.25107598e-01 -1.99204758e-01 4.33264107e-01 8.23468566e-02
-8.00920367e-01 3.31003159e-01 -1.38710499e-01 -5.70974469e-01
4.89627616e-03 -5.33799648e-01 1.23451911e-01 2.82578290e-01
6.94345161e-02 -1.24789190e+00 8.42118382e-01 -1.14793634e+00
1.25147164e+00 -1.78156960e+00 -3.04330945e-01 -3.34313214e-01
4.94449556e-01 8.44679326e-02 -4.12100583e-01 9.99717236e-01
-6.57364307e-03 3.28381866e-01 -1.10442422e-01 -3.42862278e-01
1.05932392e-01 -4.79794741e-02 -6.89931452e-01 -1.13887019e-01
4.78941053e-01 1.14008069e+00 -1.04121459e+00 -5.10078788e-01
-3.75826418e-01 2.34874740e-01 -8.17214131e-01 6.56456172e-01
-3.06130618e-01 -4.65377629e-01 -2.28783727e-01 4.04409140e-01
4.04676288e-01 -5.50370932e-01 -3.08948606e-01 1.14103183e-01
3.43568355e-01 1.01192367e+00 -5.57023406e-01 1.77248132e+00
-4.75358635e-01 1.14365494e+00 -9.75711942e-02 -5.98058760e-01
8.02988946e-01 5.98540246e-01 1.40007764e-01 -5.94671249e-01
-2.12175593e-01 -5.80651015e-02 -7.49007314e-02 -8.08854342e-01
9.86522257e-01 1.50687471e-01 7.08747134e-02 1.04451966e+00
3.41214806e-01 -4.55657840e-01 1.80053234e-01 7.40515590e-01
1.56684244e+00 -4.82293457e-01 -4.03657323e-03 -5.07136643e-01
4.58721846e-01 2.87774026e-01 8.62117298e-03 1.20772707e+00
-6.29912019e-01 1.02628410e+00 6.10284209e-01 -2.13497803e-01
-1.00724459e+00 -1.17869031e+00 -1.16412282e-01 1.47437847e+00
-1.72283322e-01 -5.57084799e-01 -5.52448869e-01 -7.85418391e-01
2.59080321e-01 9.63366330e-01 -7.12208927e-01 -4.12675619e-01
-5.99904478e-01 -3.27093393e-01 5.42066336e-01 6.19365513e-01
1.41543463e-01 -8.67332458e-01 -3.91213059e-01 4.67352957e-01
-4.75246251e-01 -7.48716712e-01 -7.71550477e-01 -8.83303657e-02
-7.61697829e-01 -1.21050751e+00 -1.01195204e+00 -1.01707149e+00
3.56669016e-02 9.13919985e-01 1.56476367e+00 1.61447972e-01
-1.20173790e-01 1.15316129e+00 -4.42439109e-01 -1.51974201e-01
-3.36553156e-01 4.11361039e-01 -3.55942190e-01 -4.60325986e-01
1.01772976e+00 -5.61312616e-01 -1.01011729e+00 4.18700278e-02
-8.44592273e-01 -5.06843090e-01 7.88249075e-02 9.23892558e-01
-3.01863104e-01 -9.12582517e-01 1.10976648e+00 -6.56665683e-01
1.81511259e+00 -8.86775553e-01 8.81845877e-02 6.14578962e-01
-4.95773077e-01 4.21560444e-02 3.52923274e-01 -2.73970276e-01
-7.84609199e-01 -7.41190553e-01 -5.49478590e-01 -1.35493234e-01
-1.73812628e-01 2.08889842e-01 3.03794742e-01 9.66915116e-02
9.81760204e-01 2.05361009e-01 1.41232997e-01 -6.27418816e-01
6.46378517e-01 9.27840352e-01 2.00386271e-01 -4.54707980e-01
4.03065711e-01 1.39214694e-01 -8.21832836e-01 -7.99444199e-01
-8.13929439e-01 -1.13599098e+00 -1.82418182e-01 -1.44496178e-02
9.21097994e-01 -1.02121079e+00 -4.87622291e-01 1.18261099e-01
-1.68488467e+00 3.28991860e-01 -5.14297843e-01 2.99996957e-02
-1.60387322e-01 6.15395486e-01 -7.05018163e-01 -3.62536371e-01
-6.88376665e-01 -8.70269060e-01 9.64197695e-01 4.10495669e-01
-8.39475870e-01 -1.08759594e+00 7.19388843e-01 8.85858655e-01
9.83326972e-01 -7.03078985e-01 1.32384861e+00 -1.37919152e+00
-3.75213683e-01 -1.81527674e-01 -5.40939569e-01 2.44420588e-01
1.80528313e-01 -3.62595320e-01 -1.26993072e+00 -2.45278850e-01
2.52993941e-01 -7.09501505e-01 1.22898221e+00 -4.81799990e-02
9.05573308e-01 -2.48412415e-01 -8.60196352e-02 -6.73488602e-02
9.61280465e-01 -3.15122873e-01 3.74218881e-01 3.59495878e-01
3.23833972e-01 7.63712525e-01 -2.33456016e-01 -3.66953239e-02
8.57362211e-01 4.13073659e-01 8.65816921e-02 5.25080502e-01
-4.60793704e-01 -2.59494901e-01 3.41550499e-01 1.10578597e+00
1.08707678e+00 -3.29439670e-01 -9.61353600e-01 1.22951543e+00
-1.51814020e+00 -8.75256360e-01 -3.80200654e-01 1.73101449e+00
9.71137524e-01 -4.27790642e-01 7.04718381e-03 -2.20015630e-01
4.19769019e-01 5.30584395e-01 -4.76702839e-01 -8.76124442e-01
9.99994799e-02 5.34613907e-01 -2.08648294e-01 5.36424279e-01
-8.61022592e-01 4.55165088e-01 7.42284012e+00 5.43865681e-01
-3.52751017e-01 6.02106631e-01 2.83964604e-01 9.61581767e-02
-8.53775859e-01 1.31804436e-01 -5.74931324e-01 3.13654631e-01
1.03600347e+00 -3.82139981e-01 -1.23887196e-01 8.91497552e-01
-2.82541901e-01 -1.18400365e-01 -1.30644643e+00 7.13304281e-01
6.57538474e-01 -1.43423319e+00 2.74806678e-01 -5.46553731e-01
6.30095899e-01 1.01581357e-01 -4.49365526e-02 7.42293835e-01
4.56911415e-01 -8.12816978e-01 -3.89635563e-03 3.62089634e-01
6.49565365e-03 -1.21680103e-01 7.00637102e-01 6.99011534e-02
-6.73083484e-01 -1.30938739e-01 -6.31896853e-01 5.67249907e-03
1.21595107e-01 3.95408750e-01 -1.08080888e+00 1.27318427e-01
6.91196918e-01 2.59817034e-01 -1.27403963e+00 1.42989755e+00
-2.52202481e-01 9.20926869e-01 -3.07937145e-01 -4.41544294e-01
3.53854418e-01 2.02860788e-01 4.46905226e-01 1.17342722e+00
5.86124063e-02 -3.15181553e-01 -4.12980586e-01 1.10740244e+00
-3.85387510e-01 3.12323004e-01 -3.71680111e-01 -3.51761468e-02
5.59720159e-01 9.96843100e-01 -1.91284731e-01 -4.53685313e-01
-7.22965896e-01 1.10804558e+00 5.79409540e-01 3.39127421e-01
-3.05172503e-01 -6.35943294e-01 8.26711059e-01 -2.99896090e-03
2.53348470e-01 -1.09910034e-01 -5.17993309e-02 -1.29120004e+00
1.52827948e-01 -8.55123222e-01 6.94213331e-01 -9.57059085e-01
-1.86039317e+00 2.93136835e-01 -7.87814036e-02 -4.50451434e-01
-4.80738074e-01 -3.46098363e-01 -1.26327181e+00 1.12908602e+00
-1.61944985e+00 -9.11976397e-01 -3.54293197e-01 3.88354003e-01
1.05530274e+00 2.15776507e-02 9.36181426e-01 1.37827531e-01
-2.48377815e-01 7.47593939e-01 3.52632344e-01 9.52261239e-02
1.08229208e+00 -1.46270800e+00 8.72515559e-01 6.60251558e-01
3.81576329e-01 1.07174683e+00 5.15188813e-01 -3.97134215e-01
-1.12253511e+00 -8.64268482e-01 1.69584870e+00 -1.04548573e+00
7.35934556e-01 -5.42082787e-01 -1.50678825e+00 3.65695119e-01
1.08349025e+00 -7.07078099e-01 1.04867792e+00 5.80964506e-01
-8.27703416e-01 8.77027512e-02 -9.42200780e-01 5.25110185e-01
4.57460076e-01 -1.25065756e+00 -1.59550118e+00 7.28120029e-01
1.42901850e+00 5.32787859e-01 -8.08745384e-01 -1.19086199e-01
3.28016460e-01 -9.30626333e-01 9.95232463e-01 -9.62024331e-01
4.01152730e-01 1.09750032e-01 4.54375632e-02 -1.17100990e+00
-5.12776911e-01 -4.48066413e-01 -3.23642790e-01 1.30643082e+00
3.34866405e-01 -6.85388446e-01 9.43846226e-01 8.95077288e-01
4.80889603e-02 -5.63381255e-01 -1.26202273e+00 -5.17107546e-01
6.96574032e-01 -3.10411528e-02 5.33656359e-01 1.03211951e+00
2.25275904e-01 7.72914588e-01 3.02227974e-01 -1.70620546e-01
2.99887508e-01 4.88117598e-02 5.94978392e-01 -1.15144110e+00
-5.82952723e-02 -6.35401726e-01 -2.14788299e-02 -1.37570632e+00
1.39626071e-01 -9.15687919e-01 -2.52563089e-01 -2.09198666e+00
4.61260796e-01 1.42916486e-01 -8.77534747e-02 -1.97007111e-03
-7.88328111e-01 -7.48953596e-02 1.29205987e-01 1.73420206e-01
-6.90005720e-01 8.16371083e-01 5.84697008e-01 -3.88380259e-01
1.67715594e-01 -3.89584824e-02 -1.05109453e+00 3.00787717e-01
8.22752237e-01 -6.51124597e-01 -3.64795238e-01 -8.51725459e-01
4.30324167e-01 -2.77697235e-01 5.40442705e-01 -9.77495372e-01
5.85012078e-01 4.17389810e-01 -8.41255337e-02 -7.28117824e-01
4.09870327e-01 -4.77644563e-01 -8.15771520e-01 3.56926888e-01
-9.26986158e-01 5.41886151e-01 9.44600329e-02 6.99817479e-01
-3.86397958e-01 -8.51405203e-01 3.35444421e-01 -2.15279073e-01
-3.93346399e-01 -1.29707173e-01 -8.31217706e-01 8.09664309e-01
2.86996096e-01 -9.88910496e-02 -5.57217419e-01 -8.34276974e-01
-2.92742789e-01 6.29300296e-01 1.68617651e-01 8.65138590e-01
7.35274255e-01 -1.28345144e+00 -8.11509907e-01 -2.91812010e-02
5.19185424e-01 -3.86791617e-01 2.70755649e-01 3.13809246e-01
-2.92662621e-01 8.30793858e-01 2.33218238e-01 -4.55111921e-01
-1.08003378e+00 4.27998632e-01 1.99286655e-01 -4.49881971e-01
-1.45905018e-01 1.30174589e+00 -7.51526188e-03 -6.23809218e-01
2.65451223e-01 -4.75187123e-01 -4.79435384e-01 3.85423630e-01
6.52339995e-01 4.42306817e-01 2.13490769e-01 -2.00272631e-02
-2.34532818e-01 2.63202429e-01 -6.05309248e-01 -1.03347354e-01
1.02659702e+00 -2.13354930e-01 -2.51151145e-01 3.18098187e-01
1.65294683e+00 -1.89564839e-01 -3.76769722e-01 -5.45151949e-01
2.03915820e-01 -3.87421846e-01 -1.23812191e-01 -6.33586466e-01
-1.91805884e-01 1.27371848e+00 5.06792843e-01 5.00085652e-01
6.73888683e-01 2.60081232e-01 1.10840881e+00 9.53802943e-01
-4.36033458e-01 -1.03689504e+00 5.31634867e-01 6.98760033e-01
1.14256680e+00 -1.25866938e+00 -1.77767705e-02 4.43803743e-02
-2.83746511e-01 1.14866781e+00 7.18470514e-01 -3.38167995e-01
6.11859620e-01 -6.29486680e-01 -3.80691476e-02 -5.04355192e-01
-1.11214375e+00 -1.46832928e-01 1.98963642e-01 5.86519003e-01
4.37853694e-01 -5.92184484e-01 -3.60011637e-01 6.33159399e-01
-3.85684103e-01 -3.30780745e-01 6.49147272e-01 8.82837355e-01
-6.50382698e-01 -9.60001469e-01 -3.38014334e-01 8.32921445e-01
-5.55746913e-01 -5.65806746e-01 -6.93512440e-01 3.92202646e-01
-4.07615334e-01 1.30386090e+00 3.63191664e-01 -1.40885428e-01
3.55413258e-01 8.23507011e-01 8.92258883e-02 -8.91665995e-01
-1.38996148e+00 -6.31649554e-01 3.16978484e-01 -4.59909886e-01
-1.22721121e-01 -6.89080000e-01 -3.04864526e-01 2.53174640e-02
-4.65890437e-01 6.21510863e-01 2.86570877e-01 7.28175044e-01
8.66479158e-01 2.41507173e-01 5.30503273e-01 1.81638628e-01
-8.38280916e-01 -1.37095058e+00 6.96893735e-03 5.89199483e-01
3.93698305e-01 -1.75827384e-01 -5.76157331e-01 -2.28517115e-01] | [11.3596773147583, 8.080748558044434] |
6048ba1b-ede1-46dc-88f5-2a6663f20a18 | tokenwise-contrastive-pretraining-for-finer | 2204.05188 | null | https://arxiv.org/abs/2204.05188v2 | https://arxiv.org/pdf/2204.05188v2.pdf | Tokenwise Contrastive Pretraining for Finer Speech-to-BERT Alignment in End-to-End Speech-to-Intent Systems | Recent advances in End-to-End (E2E) Spoken Language Understanding (SLU) have been primarily due to effective pretraining of speech representations. One such pretraining paradigm is the distillation of semantic knowledge from state-of-the-art text-based models like BERT to speech encoder neural networks. This work is a step towards doing the same in a much more efficient and fine-grained manner where we align speech embeddings and BERT embeddings on a token-by-token basis. We introduce a simple yet novel technique that uses a cross-modal attention mechanism to extract token-level contextual embeddings from a speech encoder such that these can be directly compared and aligned with BERT based contextual embeddings. This alignment is performed using a novel tokenwise contrastive loss. Fine-tuning such a pretrained model to perform intent recognition using speech directly yields state-of-the-art performance on two widely used SLU datasets. Our model improves further when fine-tuned with additional regularization using SpecAugment especially when speech is noisy, giving an absolute improvement as high as 8% over previous results. | ['Brian Kingsbury', 'Hong-Kwang J. Kuo', 'Samuel Thomas', 'Eric Fosler-Lussier', 'Vishal Sunder'] | 2022-04-11 | null | null | null | null | ['intent-recognition'] | ['natural-language-processing'] | [ 3.24752480e-01 3.35842341e-01 -1.96637902e-02 -9.30125356e-01
-1.21267509e+00 -4.15112674e-01 7.76614606e-01 1.94530457e-01
-8.25795829e-01 4.81094778e-01 9.27141488e-01 -4.89872366e-01
4.03022289e-01 -5.24147809e-01 -8.05676997e-01 -3.47326756e-01
4.45895791e-02 6.02288783e-01 -1.15234852e-02 -3.32281590e-01
-3.40657793e-02 -1.91155763e-03 -1.41355622e+00 5.88921845e-01
6.99130356e-01 9.06682074e-01 2.55891025e-01 1.01954091e+00
-4.85512406e-01 7.37914383e-01 -5.24433672e-01 -2.08262950e-02
-1.49745941e-01 -4.46810007e-01 -1.20309842e+00 8.25663749e-03
3.96877915e-01 -3.76633704e-01 -1.55008346e-01 5.62350750e-01
6.01624846e-01 4.58420217e-01 6.44673765e-01 -6.02071464e-01
-8.25188637e-01 7.66869843e-01 8.09001178e-02 9.48296785e-02
3.87509614e-01 -1.04326233e-01 1.50549543e+00 -1.10458255e+00
3.43838602e-01 1.63958061e+00 4.06850189e-01 7.68636346e-01
-1.41219079e+00 -4.12877411e-01 3.03457886e-01 2.63277382e-01
-1.14636815e+00 -8.19382012e-01 4.66074795e-01 -2.88250390e-02
1.91520667e+00 7.72482157e-02 1.69256955e-01 1.08394814e+00
-1.82948977e-01 1.09459376e+00 9.72225666e-01 -9.59404647e-01
2.74517775e-01 1.24316804e-01 9.87782329e-02 5.85933149e-01
-4.62778777e-01 1.01634808e-01 -6.00973427e-01 1.93825275e-01
2.58469760e-01 -2.62650579e-01 -1.82761192e-01 6.75045028e-02
-1.10376513e+00 1.03768837e+00 6.48134649e-01 5.03103077e-01
-1.89269558e-01 3.17275196e-01 8.26436102e-01 4.06517714e-01
7.87504077e-01 3.69960397e-01 -7.17466354e-01 -4.29108948e-01
-9.88013387e-01 -2.74449080e-01 8.85072351e-01 5.36782324e-01
7.85605848e-01 4.10710901e-01 -1.33614853e-01 1.25017285e+00
4.76905972e-01 2.54095793e-01 9.50074196e-01 -4.44717109e-01
4.68123704e-01 3.41781601e-02 -3.50564718e-01 -9.05134622e-03
-1.87540315e-02 -2.85756260e-01 -3.04885924e-01 8.30999240e-02
1.54605194e-03 -2.49924138e-01 -1.37898445e+00 1.87465441e+00
2.77107172e-02 4.14026886e-01 5.16654730e-01 6.40882134e-01
8.22717011e-01 1.00960815e+00 4.25952107e-01 2.96295226e-01
1.40401113e+00 -1.31015480e+00 -7.40755081e-01 -5.79872012e-01
1.03045118e+00 -6.84871972e-01 1.41581595e+00 8.11380148e-02
-8.83401215e-01 -5.61277747e-01 -1.08711255e+00 -4.42739695e-01
-6.53249323e-01 7.78722316e-02 2.15909749e-01 7.20559180e-01
-1.29883647e+00 3.93224776e-01 -8.54381442e-01 -5.01667380e-01
2.59774715e-01 2.72173703e-01 -4.98378932e-01 -3.87645103e-02
-1.41733360e+00 1.03839242e+00 6.12890124e-01 -1.97685152e-01
-8.99991989e-01 -7.34564900e-01 -1.38593566e+00 2.77417004e-01
3.20186675e-01 -4.37754720e-01 1.65955389e+00 -8.63146484e-01
-1.96806490e+00 7.82953918e-01 -5.02998650e-01 -7.64363647e-01
-4.34485935e-02 -5.11596084e-01 -2.90009469e-01 2.11755298e-02
-7.41082896e-03 8.74620855e-01 7.81279981e-01 -1.03983414e+00
-5.20234227e-01 -8.50822926e-02 -1.55276775e-01 3.08953047e-01
-4.12342817e-01 1.53000221e-01 -1.01407990e-01 -6.26195490e-01
-1.80252597e-01 -5.39917946e-01 6.39399439e-02 -3.51566792e-01
-1.44448400e-01 -7.02555656e-01 9.57674921e-01 -6.99335515e-01
1.04775560e+00 -2.02235699e+00 1.69166252e-01 -2.29277000e-01
-2.99212843e-01 5.91454804e-01 -4.54014242e-01 8.37192655e-01
-1.80456638e-01 2.48445958e-01 -2.17913553e-01 -1.14477706e+00
2.63894528e-01 6.16907775e-01 -5.54321289e-01 1.62072387e-02
6.62423015e-01 9.87063944e-01 -9.38676000e-01 -2.34024227e-02
5.46901166e-01 6.45576477e-01 -7.96898246e-01 5.88525414e-01
-3.11985344e-01 2.99374521e-01 -9.83818397e-02 2.93545872e-01
2.23375171e-01 1.83581010e-01 5.98867014e-02 5.58411293e-02
-2.53046155e-02 1.25256598e+00 -7.29862630e-01 1.97461224e+00
-1.21787548e+00 7.46120632e-01 1.59989387e-01 -1.26366234e+00
8.96595120e-01 7.04393506e-01 -1.02520451e-01 -4.23280388e-01
1.37837008e-01 4.10023183e-01 3.22406995e-03 -2.33299628e-01
4.71518278e-01 -7.55965173e-01 -4.69007529e-02 4.77278203e-01
5.95594287e-01 -4.03808385e-01 -2.29243904e-01 9.88046378e-02
9.55654562e-01 2.66089253e-02 4.06860620e-01 -1.91785440e-01
6.15562141e-01 -3.25962126e-01 1.82672106e-02 7.27389753e-01
-1.10650413e-01 5.79065025e-01 7.55694360e-02 -1.61348179e-01
-1.00634980e+00 -9.93027389e-01 -1.26263082e-01 1.58317304e+00
-6.74234092e-01 -5.40284753e-01 -8.49270642e-01 -8.62958908e-01
1.58305466e-03 1.18668067e+00 -5.88599622e-01 -2.08060429e-01
-5.23473680e-01 -2.91442841e-01 7.33162463e-01 9.16988671e-01
3.99277002e-01 -1.28231752e+00 -2.98243556e-02 5.25684536e-01
-1.08337775e-01 -1.37417519e+00 -5.42548001e-01 7.09012151e-01
-6.12585127e-01 -2.65376121e-01 -6.65478885e-01 -9.36769783e-01
4.65177953e-01 -6.33907737e-03 1.14120805e+00 -6.66532964e-02
-1.21080481e-01 3.25255930e-01 -6.92412376e-01 -4.62742627e-01
-6.65929317e-01 2.01079682e-01 1.84444055e-01 -4.41465341e-02
6.02313697e-01 -2.82871276e-01 -2.24214926e-01 -7.30795711e-02
-9.24688458e-01 -2.49870181e-01 5.93394160e-01 1.25411963e+00
2.95209706e-01 -5.37559509e-01 9.18619156e-01 -7.73282468e-01
6.63302124e-01 -3.21308285e-01 -9.78242308e-02 5.59083223e-02
-3.17026943e-01 4.48369831e-01 8.29811931e-01 -6.58219531e-02
-1.12115681e+00 -1.20715536e-01 -9.24364746e-01 -3.66965741e-01
-5.78329563e-01 6.67713106e-01 -7.20983893e-02 2.16302872e-01
3.73456001e-01 2.85604894e-01 -3.79785858e-02 -6.19348705e-01
8.48722458e-01 1.19997144e+00 4.40679580e-01 -6.29074097e-01
3.88108313e-01 1.37992874e-01 -6.16002619e-01 -1.31941414e+00
-1.04682922e+00 -8.01695526e-01 -7.73891807e-01 3.44023883e-01
9.94910240e-01 -8.37411046e-01 -3.63605529e-01 2.45312318e-01
-1.18348968e+00 -7.01131284e-01 -3.62623632e-01 5.97468972e-01
-5.81291854e-01 3.48232508e-01 -6.43805206e-01 -9.81935501e-01
-4.45471913e-01 -1.04791796e+00 1.35933554e+00 -1.63395137e-01
-3.56236041e-01 -1.24642766e+00 1.33024663e-01 5.79631865e-01
6.58266187e-01 -5.74861467e-01 8.92247021e-01 -1.15015900e+00
-1.93003371e-01 7.97576737e-03 -2.27547094e-01 9.72609460e-01
1.80938885e-01 -4.32475060e-01 -1.51764953e+00 -3.61156523e-01
-1.22763984e-01 -7.68010557e-01 1.17860115e+00 1.94461420e-01
8.78182530e-01 -3.19171458e-01 -3.07644103e-02 5.10934651e-01
1.24304664e+00 4.57061157e-02 4.69549716e-01 1.56813115e-01
6.25591815e-01 6.15853846e-01 3.47468227e-01 2.67629139e-02
3.19883436e-01 6.35531545e-01 1.93018436e-01 1.99417081e-02
-3.94820929e-01 -3.47875327e-01 5.79489231e-01 1.14144170e+00
4.20790285e-01 -3.07951152e-01 -8.69558692e-01 1.06327689e+00
-1.71518028e+00 -7.56485403e-01 5.53648949e-01 2.02768326e+00
1.08594894e+00 1.40321180e-01 -1.76615477e-01 -9.45874024e-03
1.70826510e-01 3.98444891e-01 -9.26419646e-02 -1.06326449e+00
2.84612447e-01 6.71891272e-01 1.86075881e-01 1.03055656e+00
-1.06169474e+00 1.51167309e+00 5.96449614e+00 9.58145797e-01
-1.32722557e+00 3.02007973e-01 5.05106330e-01 1.09002367e-01
-2.90364683e-01 -1.28667980e-01 -8.56419146e-01 5.24551757e-02
1.63512862e+00 1.23272978e-01 2.78891504e-01 8.63790393e-01
2.72706926e-01 1.67542752e-02 -1.28999245e+00 7.28739619e-01
3.33242804e-01 -1.38201225e+00 6.44262582e-02 -2.19127104e-01
5.08998334e-01 4.08003181e-01 -6.17725626e-02 6.83946908e-01
3.84254515e-01 -1.27028680e+00 5.92205822e-01 -1.09606080e-01
8.07270527e-01 -7.70552635e-01 8.80031466e-01 2.50102907e-01
-1.29790378e+00 1.94314837e-01 -2.83912003e-01 -1.97700605e-01
4.89784867e-01 1.21692047e-01 -1.65595925e+00 4.80268717e-01
4.36000973e-01 6.26716614e-01 1.13411978e-01 4.93417591e-01
-5.12520254e-01 1.09854543e+00 -3.68279159e-01 -2.35802814e-01
9.17522252e-01 1.27037451e-01 4.75013912e-01 1.80533385e+00
3.64237688e-02 -1.98035184e-02 1.81885153e-01 6.98428988e-01
-3.34186405e-01 2.74730593e-01 -6.84688270e-01 -2.84210414e-01
3.91620785e-01 8.58576238e-01 -2.33967617e-01 -6.11803889e-01
-6.63932025e-01 1.28351462e+00 6.27525985e-01 2.70314276e-01
-2.99016863e-01 -5.37363291e-01 1.01179206e+00 -2.00100526e-01
7.87728667e-01 -5.42099059e-01 -7.95892328e-02 -1.04708159e+00
-1.60412386e-01 -6.07986808e-01 6.71549290e-02 -5.84017217e-01
-1.34991419e+00 9.52052653e-01 -7.17750937e-02 -7.20090568e-01
-7.96129644e-01 -8.41980159e-01 -7.60768354e-01 1.19825613e+00
-1.90781307e+00 -1.42317939e+00 2.31318980e-01 2.93372780e-01
1.18457878e+00 -2.58722633e-01 1.34971428e+00 1.51697949e-01
-3.35625708e-01 6.90838933e-01 1.73485383e-01 3.44628721e-01
7.41734624e-01 -1.50191903e+00 7.55605817e-01 8.36317360e-01
5.16100705e-01 6.64499819e-01 6.03206515e-01 -3.55061352e-01
-9.38880622e-01 -1.01722300e+00 1.23765099e+00 -5.29427350e-01
8.98462832e-01 -8.23218226e-01 -1.09512722e+00 9.08134162e-01
6.16981626e-01 1.02992587e-01 7.67454028e-01 5.28257191e-01
-4.83911783e-01 7.34126940e-02 -7.64267504e-01 3.93653393e-01
7.20370114e-01 -1.12780797e+00 -1.13383329e+00 1.41399875e-01
1.10238171e+00 -1.83701769e-01 -6.19680345e-01 1.08423404e-01
2.66115457e-01 -5.95353544e-01 8.29770446e-01 -8.92536581e-01
1.98556513e-01 1.07219685e-02 -4.23578262e-01 -1.78980625e+00
1.91547796e-01 -5.48883677e-01 2.28483722e-01 1.16936326e+00
5.17096937e-01 -7.58999944e-01 5.67335248e-01 1.31177753e-01
-7.33866394e-01 -8.20494294e-01 -1.15024829e+00 -8.52503121e-01
2.60698050e-01 -7.77401984e-01 2.15088159e-01 6.27137303e-01
2.42399693e-01 7.96088755e-01 -2.66687155e-01 1.79168191e-02
2.48327419e-01 -4.03529376e-01 5.26187658e-01 -7.14366496e-01
-2.43018791e-01 -3.34360629e-01 -5.27030885e-01 -1.54743004e+00
6.85445130e-01 -1.04872394e+00 5.40141881e-01 -1.68935525e+00
-2.94681579e-01 -3.18191916e-01 -3.34009707e-01 6.78527415e-01
-1.90472588e-01 8.67216811e-02 1.75377950e-01 -1.81279153e-01
-3.69057506e-01 1.09479547e+00 7.02982485e-01 -5.29657938e-02
-6.96896985e-02 -2.87430823e-01 -3.41081947e-01 4.96427745e-01
6.18300319e-01 -2.51033843e-01 -3.63002717e-01 -6.44264698e-01
-6.12319291e-01 -2.88859218e-01 1.21200398e-01 -9.11284745e-01
1.17890455e-01 3.69220555e-01 -1.61604181e-01 -4.92563188e-01
7.59855330e-01 -5.84668279e-01 -8.88290524e-01 3.61798823e-01
-5.90975583e-01 -3.38708609e-01 5.62192202e-01 5.16559958e-01
-6.68418348e-01 -4.20685351e-01 7.33040929e-01 -9.88077745e-02
-1.15414917e+00 -1.60510093e-01 -5.61024129e-01 1.79455623e-01
5.22086620e-01 -1.09108344e-01 -1.39657363e-01 -7.37031341e-01
-8.29199255e-01 2.70835496e-02 -1.51282266e-01 6.35787010e-01
5.15728593e-01 -1.02044082e+00 -8.23683083e-01 5.80709755e-01
1.97287038e-01 -9.83804464e-02 1.12059116e-01 5.69821775e-01
-1.64682522e-01 9.99117732e-01 1.03760995e-01 -6.23685241e-01
-1.26181984e+00 1.40836269e-01 4.07228559e-01 -2.12793559e-01
-5.57294071e-01 1.33456230e+00 3.90899450e-01 -8.95771563e-01
3.81546021e-01 -7.56599426e-01 9.00089368e-02 -5.49672171e-02
6.58147395e-01 -1.33344367e-01 4.26959544e-01 -6.96561038e-01
-5.91722906e-01 3.38358581e-01 -3.85049194e-01 -5.41158378e-01
1.42509520e+00 -2.10615993e-01 3.48337322e-01 6.13509357e-01
1.58311009e+00 2.98995897e-03 -1.31341612e+00 -4.73716319e-01
2.23508496e-02 -1.81824341e-01 3.15519422e-01 -9.21743453e-01
-5.39139330e-01 1.32883203e+00 2.54174948e-01 1.56000063e-01
6.96306109e-01 3.48947555e-01 9.72435057e-01 5.51843226e-01
2.83745956e-02 -1.02503800e+00 2.71645039e-01 1.14204717e+00
9.68471169e-01 -1.47722614e+00 -5.85933030e-01 -1.66862458e-01
-7.58077621e-01 1.02279639e+00 2.99864441e-01 -2.41071835e-01
6.75584793e-01 2.40677759e-01 2.98596412e-01 6.69801682e-02
-8.62765193e-01 -5.09443462e-01 4.29718643e-01 5.31143367e-01
8.59282315e-01 1.74296513e-01 8.09537619e-02 3.76894027e-01
-2.35910863e-01 -3.17449301e-01 2.75637180e-01 9.12647128e-01
-7.12848783e-01 -1.40486813e+00 -7.69593939e-02 2.25221500e-01
-4.30314600e-01 -6.05438292e-01 -1.56344369e-01 6.18360639e-01
-3.68504792e-01 1.05433023e+00 2.73308486e-01 -1.40385717e-01
2.92050779e-01 6.66967094e-01 1.24379866e-01 -1.34374619e+00
-4.71682757e-01 9.10648555e-02 6.04826629e-01 -4.31814820e-01
-2.11901978e-01 -4.46733177e-01 -1.51974487e+00 1.68065041e-01
-3.15429837e-01 2.29504794e-01 8.82718086e-01 1.41466248e+00
2.99976975e-01 8.10210526e-01 3.90405148e-01 -9.97344911e-01
-7.49420047e-01 -1.29906631e+00 -2.69361645e-01 2.57998347e-01
7.06806004e-01 -5.02004802e-01 -4.86256391e-01 -7.22007230e-02] | [14.02847671508789, 6.9848737716674805] |
49d8efc8-b2e5-4f85-ac44-4387dfff3425 | policy-diagnosis-via-measuring-role-diversity | 2207.05683 | null | https://arxiv.org/abs/2207.05683v1 | https://arxiv.org/pdf/2207.05683v1.pdf | Policy Diagnosis via Measuring Role Diversity in Cooperative Multi-agent RL | Cooperative multi-agent reinforcement learning (MARL) is making rapid progress for solving tasks in a grid world and real-world scenarios, in which agents are given different attributes and goals, resulting in different behavior through the whole multi-agent task. In this study, we quantify the agent's behavior difference and build its relationship with the policy performance via {\bf Role Diversity}, a metric to measure the characteristics of MARL tasks. We define role diversity from three perspectives: action-based, trajectory-based, and contribution-based to fully measure a multi-agent task. Through theoretical analysis, we find that the error bound in MARL can be decomposed into three parts that have a strong relation to the role diversity. The decomposed factors can significantly impact policy optimization on three popular directions including parameter sharing, communication mechanism, and credit assignment. The main experimental platforms are based on {\bf Multiagent Particle Environment (MPE)} and {\bf The StarCraft Multi-Agent Challenge (SMAC). Extensive experiments} clearly show that role diversity can serve as a robust measurement for the characteristics of a multi-agent cooperation task and help diagnose whether the policy fits the current multi-agent system for a better policy performance. | ['Xiaojun Chang', 'Xiaodan Liang', 'Chuanlong Xie', 'Siyi Hu'] | 2022-06-01 | null | null | null | null | ['smac-1', 'smac'] | ['playing-games', 'playing-games'] | [-7.40678430e-01 -5.47917664e-01 -1.80840269e-01 1.19326442e-01
-4.35554236e-01 -4.96914297e-01 7.21039057e-01 3.79884958e-01
-7.91381836e-01 1.15086722e+00 -5.80812953e-02 2.77675893e-02
-6.26156628e-01 -5.67810059e-01 -3.19517553e-01 -1.11838305e+00
-4.66508359e-01 7.93029547e-01 3.24349344e-01 -8.12140524e-01
3.05597752e-01 3.05652380e-01 -1.30569422e+00 -2.17631176e-01
1.12098980e+00 8.83783817e-01 5.96788287e-01 5.59490502e-01
3.08419019e-01 1.00898480e+00 -1.19946384e+00 1.79941818e-01
4.04143840e-01 -3.17086339e-01 -5.48186779e-01 -5.55002177e-03
-8.16893518e-01 -8.59785303e-02 8.85050967e-02 1.06198096e+00
7.24951923e-01 3.35393280e-01 6.73568487e-01 -1.97052419e+00
-2.03073174e-01 7.57195413e-01 -7.02310681e-01 4.28378314e-01
1.79330140e-01 5.38626969e-01 6.47775531e-01 -8.90678242e-02
2.43695095e-01 1.33259237e+00 2.54036099e-01 2.85408884e-01
-6.12039983e-01 -4.47593868e-01 4.11193401e-01 6.47533894e-01
-1.01905596e+00 5.35018146e-02 3.97694349e-01 -4.04512554e-01
5.94439507e-01 1.32853895e-01 7.68376350e-01 1.04712546e+00
6.17923975e-01 5.56641459e-01 1.38579583e+00 -2.41291329e-01
6.83846593e-01 3.72581333e-02 -1.34058371e-01 4.44007933e-01
3.29019994e-01 2.65459031e-01 -4.14374411e-01 -3.97499740e-01
7.58856237e-01 -3.44013333e-01 -2.04828799e-01 -1.98495373e-01
-1.45145214e+00 9.27029014e-01 1.12527870e-01 1.83789849e-01
-7.55233109e-01 1.60513118e-01 5.07068396e-01 7.20406175e-01
1.95878282e-01 7.59166121e-01 -4.84092087e-01 -4.31463331e-01
1.02761045e-01 6.30230904e-01 8.12209666e-01 6.54662073e-01
3.75413746e-01 2.83666253e-01 -3.03253412e-01 6.95473373e-01
2.92596221e-01 5.26958287e-01 7.78226435e-01 -1.39535606e+00
4.29389983e-01 2.84534216e-01 6.11633122e-01 -6.82042480e-01
-9.18198228e-01 -3.21145982e-01 -6.76106870e-01 7.03008592e-01
6.26630902e-01 -7.46401608e-01 -3.16057913e-03 1.66095400e+00
4.61930603e-01 -2.15693284e-02 3.27234745e-01 1.13858366e+00
3.43568891e-01 5.50580859e-01 1.39541626e-02 -5.34106255e-01
1.34709918e+00 -1.09099293e+00 -5.08533597e-01 1.22531084e-02
5.65055966e-01 -5.86077631e-01 8.56923521e-01 4.20676708e-01
-1.04685152e+00 -3.78089637e-01 -1.07943106e+00 1.08232880e+00
-2.08904594e-01 -2.04653025e-01 3.99179548e-01 5.51083446e-01
-8.47447634e-01 4.32975829e-01 -5.75863183e-01 -1.32704481e-01
-2.35680327e-01 1.79504767e-01 1.88175201e-01 3.66600186e-01
-1.25735009e+00 1.16720021e+00 5.25307894e-01 -3.27818245e-01
-1.15685856e+00 -5.17735243e-01 -1.52590960e-01 1.28598347e-01
7.99214900e-01 -5.82843959e-01 1.22760940e+00 -6.65507078e-01
-1.77075315e+00 2.98408661e-02 8.46415937e-01 -4.14248437e-01
6.15540206e-01 2.91785508e-01 -3.75395328e-01 -3.65103744e-02
8.79111066e-02 1.03278227e-01 7.33522356e-01 -1.43300295e+00
-1.24988902e+00 -2.65484661e-01 5.01299798e-01 8.80459428e-01
-1.81664348e-01 9.16705430e-02 2.34195173e-01 -6.03902221e-01
-6.93223417e-01 -1.11209142e+00 -4.03368592e-01 -7.34470487e-01
1.74836993e-01 -6.61784708e-01 4.32904184e-01 -5.26292101e-02
8.99271071e-01 -1.81759131e+00 4.16024029e-01 4.45726328e-02
2.57463127e-01 1.92806050e-01 -2.86659956e-01 5.96874118e-01
5.67047477e-01 -1.01027541e-01 2.13282481e-01 5.56949414e-02
3.39381039e-01 3.54238093e-01 1.57349855e-01 4.65356857e-01
-3.92423272e-01 4.73613083e-01 -9.82169628e-01 -2.24861473e-01
6.32631453e-03 -1.04967095e-01 -7.39084259e-02 3.76259744e-01
-5.00085831e-01 5.77130914e-01 -9.40139115e-01 4.24540788e-01
5.24359226e-01 -1.15459137e-01 3.26409161e-01 2.55269140e-01
-3.08604658e-01 -3.16043913e-01 -1.32569027e+00 1.35054469e+00
-3.09097379e-01 1.32303506e-01 4.70244557e-01 -1.13987517e+00
8.62828434e-01 2.00575858e-01 9.23227549e-01 -9.39830363e-01
3.12043607e-01 1.95020139e-01 5.66915810e-01 -5.38214028e-01
6.65130138e-01 3.36576045e-01 -1.15603052e-01 8.04013312e-01
-2.73883551e-01 7.76233375e-02 5.67539454e-01 -4.17646319e-02
1.29590428e+00 -1.11939013e-01 3.98524463e-01 -6.36101663e-01
4.96882677e-01 1.25191316e-01 7.63316214e-01 9.40443695e-01
-7.52061069e-01 -2.18977407e-01 7.60935605e-01 -2.07651109e-01
-7.75891781e-01 -6.46781743e-01 3.52113694e-01 1.40679097e+00
7.15432703e-01 -1.12116979e-02 -5.04012764e-01 -6.15724266e-01
3.14581126e-01 5.13585865e-01 -5.91063619e-01 -1.64560124e-01
-5.95497906e-01 -1.24747002e+00 3.73021841e-01 8.70614722e-02
5.77392459e-01 -1.30171883e+00 -1.12197542e+00 3.22822630e-01
-6.10846980e-03 -1.04611397e+00 -4.46977526e-01 9.78780612e-02
-2.14899153e-01 -1.27037358e+00 -7.42731690e-01 -3.04035157e-01
1.32656083e-01 4.47466940e-01 1.13012242e+00 8.24223608e-02
7.85436705e-02 8.35973799e-01 -8.51651847e-01 -6.78357482e-01
-5.99550307e-01 7.82424435e-02 5.23889184e-01 -1.04902379e-01
-1.19434036e-01 -4.03691292e-01 -7.02761948e-01 8.25356305e-01
-6.32501960e-01 -3.16758037e-01 6.47225320e-01 6.83828831e-01
2.39146709e-01 4.36260372e-01 9.75314736e-01 -3.03571165e-01
1.44730258e+00 -7.86323726e-01 -7.24033773e-01 4.81854528e-01
-7.95602262e-01 -9.65555608e-02 8.65251541e-01 -5.79066455e-01
-9.57812011e-01 -7.19608247e-01 3.67098242e-01 -1.17841549e-01
3.21063660e-02 5.07779300e-01 6.92569688e-02 -4.65527214e-02
4.59322870e-01 2.28310555e-01 4.61366594e-01 -8.03479329e-02
-3.24633643e-02 4.49385017e-01 -5.08942902e-02 -1.03251529e+00
4.31814343e-01 9.15025473e-02 1.49614871e-01 -4.26554948e-01
-2.50163913e-01 -8.09613243e-02 1.04547299e-01 -5.50884306e-01
6.01581395e-01 -7.80058503e-01 -1.60627842e+00 7.21104383e-01
-1.06191313e+00 -6.85342014e-01 -1.44157410e-01 7.68564880e-01
-7.15385735e-01 2.63951629e-01 -4.46198314e-01 -8.07358742e-01
-1.15069918e-01 -1.44188643e+00 3.61892223e-01 5.45182168e-01
5.83014429e-01 -1.10359406e+00 4.34695065e-01 9.38569084e-02
5.99940836e-01 1.03853218e-01 8.06124032e-01 -6.92612231e-01
-2.88342774e-01 4.53850240e-01 7.36780092e-02 1.17996819e-01
-1.46495685e-01 -2.08794191e-01 -2.76455075e-01 -8.53358388e-01
1.93802521e-01 -4.59682673e-01 1.47331730e-01 4.48148072e-01
7.34412968e-01 -2.42298722e-01 -1.33844361e-01 9.48445126e-02
1.36404455e+00 6.96853101e-01 2.03071058e-01 8.92974377e-01
2.45458946e-01 4.13872987e-01 1.19964612e+00 1.14639854e+00
6.45225704e-01 8.42226088e-01 9.07887995e-01 2.79501647e-01
3.34087819e-01 4.94159132e-01 7.39087760e-01 8.43324602e-01
-5.29792249e-01 -5.61348319e-01 -7.63024867e-01 -2.06633750e-03
-2.43741870e+00 -1.08808661e+00 1.48387253e-01 2.10646510e+00
5.14888167e-01 5.68143912e-02 7.12722719e-01 -2.12484479e-01
6.73074424e-01 1.49023488e-01 -7.81702876e-01 -2.52205640e-01
-2.39409402e-01 -5.62748194e-01 6.46175146e-01 2.75326520e-01
-6.40484273e-01 5.27684093e-01 6.05605268e+00 1.23943591e+00
-9.76044953e-01 3.19488466e-01 4.51451361e-01 -1.07246086e-01
1.02051094e-01 -3.15786272e-01 -7.07575798e-01 8.78028452e-01
8.07202816e-01 -8.17766905e-01 8.63580167e-01 5.61199129e-01
6.30081773e-01 -4.65980023e-01 -7.71350503e-01 9.72455680e-01
-1.26711011e-01 -1.10599339e+00 -3.48587990e-01 3.09163660e-01
5.91306388e-01 2.85150290e-01 4.38802391e-02 6.05991662e-01
9.44938958e-01 -7.65478075e-01 7.50830770e-01 5.11752903e-01
1.82249963e-01 -7.79954016e-01 7.35923648e-01 6.93765998e-01
-1.21367455e+00 -5.69702387e-01 -5.36684692e-01 -2.88469791e-01
-1.92069691e-02 7.97614604e-02 -6.47694647e-01 8.52584004e-01
6.91967309e-01 2.77203023e-01 -3.10095966e-01 1.00162840e+00
2.29288161e-01 3.00741583e-01 -4.73898277e-02 -5.39316297e-01
4.28228527e-01 -5.58002651e-01 8.41435492e-01 4.72863257e-01
3.17826122e-01 1.80201620e-01 8.21835220e-01 5.64873159e-01
3.28116059e-01 1.44677786e-02 -1.27338395e-01 1.71477571e-01
8.58583570e-01 1.41383791e+00 -7.00529814e-01 -8.38072971e-02
-2.83541858e-01 5.54094553e-01 3.41329604e-01 3.72981846e-01
-1.25764406e+00 -4.26049642e-02 9.86628354e-01 -3.73976916e-01
-6.53194189e-02 -4.39052135e-01 9.22494456e-02 -8.23861659e-01
-1.00813992e-01 -1.17290568e+00 3.30882967e-01 -4.99257147e-01
-1.40693951e+00 6.42999351e-01 7.03452667e-03 -1.44999385e+00
-3.15063655e-01 -5.11338294e-01 -6.51834488e-01 6.07443750e-01
-1.64440334e+00 -6.01075053e-01 -3.96435112e-01 6.95032477e-01
5.85839748e-01 -1.01895916e+00 6.47438884e-01 2.42741182e-01
-9.38111961e-01 2.89249957e-01 6.03252113e-01 -3.66323471e-01
6.20962203e-01 -1.27336395e+00 -2.48409927e-01 3.46315056e-01
-4.55967188e-01 -1.65708363e-01 7.62392342e-01 -2.64224380e-01
-1.50060940e+00 -7.54422069e-01 -3.41817170e-01 -1.36722028e-01
8.26668262e-01 2.10776404e-01 -4.99240190e-01 1.00752510e-01
5.36700904e-01 -2.26619765e-01 4.25561756e-01 -1.75466239e-02
3.09664071e-01 -1.24851689e-01 -1.04077852e+00 4.16823506e-01
7.82994211e-01 2.11257219e-01 -1.60712406e-01 4.42750245e-01
7.37595260e-01 -3.32276016e-01 -1.05565035e+00 2.02414706e-01
2.06980631e-01 -1.17710066e+00 7.23276854e-01 -5.25149822e-01
1.50156366e-02 -2.60874897e-01 -5.84178381e-02 -2.25081658e+00
-6.07495070e-01 -7.19610810e-01 3.74408509e-03 8.46385658e-01
1.45802125e-01 -1.03845167e+00 5.77848792e-01 2.90627867e-01
-8.17976370e-02 -6.77302718e-01 -1.08710515e+00 -1.18939459e+00
2.88600802e-01 1.91807345e-01 9.34252501e-01 9.58074987e-01
9.34249237e-02 2.88894832e-01 -4.02651429e-01 1.59719482e-01
7.40504622e-01 -1.42874286e-01 7.71558464e-01 -1.15904951e+00
-5.35426259e-01 -8.69916081e-01 7.00814463e-03 -8.06590736e-01
2.60327816e-01 -4.45275545e-01 -1.56699643e-01 -1.27692318e+00
4.18834165e-02 -1.14057302e+00 -4.99343157e-01 1.62440836e-01
-1.91176713e-01 -4.80647147e-01 5.61722696e-01 6.38046384e-01
-1.05804789e+00 8.07793438e-01 1.38367581e+00 -2.05779657e-01
-2.73563445e-01 1.28491566e-01 -3.47182751e-01 6.14931524e-01
1.09507585e+00 -3.94645303e-01 -4.65194017e-01 -4.68555599e-01
1.54800758e-01 6.79195166e-01 7.94389844e-02 -1.05845129e+00
2.95166075e-01 -1.05785346e+00 -3.37031037e-01 1.32481428e-02
1.58838287e-01 -8.35618615e-01 9.23063979e-03 7.65760183e-01
-1.29818872e-01 5.92446506e-01 -6.96929544e-02 6.99822128e-01
-2.60340758e-02 -3.53410780e-01 6.08389735e-01 -2.70957321e-01
-8.62804353e-01 2.47873887e-01 -6.16770029e-01 2.69537270e-01
1.52424777e+00 2.08989576e-01 -8.54304552e-01 -4.24045086e-01
-4.24948275e-01 1.01171267e+00 1.44012481e-01 4.27608222e-01
1.48031577e-01 -1.19469285e+00 -1.12433255e+00 -3.52831393e-01
-2.12207675e-01 -5.70418596e-01 2.16412961e-01 9.02763247e-01
-4.50689286e-01 1.81385964e-01 -7.83564150e-01 -3.50306839e-01
-1.14464307e+00 2.68732905e-01 6.41475379e-01 -7.21924424e-01
-1.81572363e-01 2.46997729e-01 -1.47057801e-01 -2.23908082e-01
3.55216175e-01 4.57568280e-02 -4.73162204e-01 2.24434569e-01
3.32154483e-01 9.14566517e-01 -2.30308071e-01 -3.19405556e-01
-1.74428761e-01 2.08345309e-01 1.03879154e-01 -1.90286696e-01
1.17823720e+00 -2.74457395e-01 4.24080752e-02 2.16181338e-01
3.55455190e-01 -3.86526823e-01 -1.33298922e+00 -2.27591276e-01
-2.65684575e-01 -2.69233763e-01 -7.42402375e-02 -9.83721316e-01
-1.00193155e+00 9.80756059e-02 5.57794333e-01 8.82761002e-01
7.90342271e-01 -2.95410097e-01 4.95234042e-01 4.77267474e-01
9.37673926e-01 -1.63832974e+00 4.75088716e-01 6.96266294e-01
1.10499668e+00 -1.31626201e+00 1.19358838e-01 -8.47891346e-02
-1.18851805e+00 9.42817152e-01 9.59695399e-01 -1.06422044e-01
5.75743020e-01 3.07688177e-01 9.35064070e-03 -1.77935928e-01
-1.00697207e+00 -3.86441469e-01 -3.51216614e-01 8.06669414e-01
-1.04343124e-01 4.38104719e-01 -6.03116214e-01 7.07271874e-01
-2.11434644e-02 -3.53462487e-01 8.47369552e-01 8.31892133e-01
-7.42850423e-01 -1.17026973e+00 -6.25449717e-01 3.14798474e-01
8.36677700e-02 5.92825770e-01 -6.20462280e-03 8.60414565e-01
1.19957015e-01 1.19374394e+00 3.59969400e-02 -2.83357233e-01
3.98345679e-01 -4.92455125e-01 4.61568356e-01 -1.99250311e-01
-8.50991726e-01 -1.40231237e-01 2.72825807e-02 -4.62434173e-01
-6.00912154e-01 -5.80226302e-01 -1.37427831e+00 -6.93570018e-01
3.79939377e-02 5.37733555e-01 4.35101181e-01 8.58151197e-01
4.49645430e-01 8.53531241e-01 1.06806695e+00 -8.57917309e-01
-1.21352112e+00 -9.96714056e-01 -8.88215601e-01 2.50244260e-01
-1.31850541e-02 -1.24361837e+00 -3.96219134e-01 -7.36715794e-01] | [3.7713568210601807, 1.974765419960022] |
6f999e85-6fc9-4fa7-9dc1-beabb3aec8f1 | evaluating-foveated-video-quality-using | 2106.06817 | null | https://arxiv.org/abs/2106.06817v1 | https://arxiv.org/pdf/2106.06817v1.pdf | Evaluating Foveated Video Quality Using Entropic Differencing | Virtual Reality is regaining attention due to recent advancements in hardware technology. Immersive images / videos are becoming widely adopted to carry omnidirectional visual information. However, due to the requirements for higher spatial and temporal resolution of real video data, immersive videos require significantly larger bandwidth consumption. To reduce stresses on bandwidth, foveated video compression is regaining popularity, whereby the space-variant spatial resolution of the retina is exploited. Towards advancing the progress of foveated video compression, we propose a full reference (FR) foveated image quality assessment algorithm, which we call foveated entropic differencing (FED), which employs the natural scene statistics of bandpass responses by applying differences of local entropies weighted by a foveation-based error sensitivity function. We evaluate the proposed algorithm by measuring the correlations of the predictions that FED makes against human judgements on the newly created 2D and 3D LIVE-FBT-FCVR databases for Virtual Reality (VR). The performance of the proposed algorithm yields state-of-the-art as compared with other existing full reference algorithms. Software for FED has been made available at: http://live.ece.utexas.edu/research/Quality/FED.zip | ['Alan Bovik', 'Anjul Patney', 'Yize Jin'] | 2021-06-12 | null | null | null | null | ['foveation'] | ['computer-vision'] | [ 9.75046754e-02 -4.21476066e-01 1.82483748e-01 -9.90154594e-02
-8.32563698e-01 -2.44860068e-01 2.55503803e-01 -2.91120976e-01
-6.55110002e-01 6.51831210e-01 2.04081237e-01 -2.47750804e-01
-3.36796135e-01 -4.01965559e-01 -6.42133057e-01 -5.27928352e-01
-2.86760390e-01 -7.63317406e-01 3.40535104e-01 -5.06802425e-02
5.99818587e-01 3.48275214e-01 -2.12042141e+00 2.85054505e-01
9.85693336e-01 1.35392654e+00 6.54775023e-01 1.00636601e+00
5.78163564e-01 7.41470456e-01 -2.34502777e-01 -3.06235105e-01
3.67708772e-01 -2.63480216e-01 -3.51494610e-01 -9.38938931e-02
4.04259890e-01 -5.57861447e-01 -7.32567668e-01 1.21071064e+00
7.63674736e-01 3.99537861e-01 2.83198327e-01 -7.76436388e-01
-7.82741785e-01 -1.37015060e-01 -5.71729481e-01 9.69625056e-01
8.23469639e-01 4.77959588e-02 7.51617312e-01 -1.03211915e+00
6.50083244e-01 7.45626450e-01 2.55961239e-01 2.43045434e-01
-8.86405289e-01 -2.18967497e-01 -2.43249223e-01 7.97117233e-01
-1.82085764e+00 -8.22528899e-01 5.36763191e-01 -3.46948475e-01
1.05017483e+00 4.40069914e-01 6.13750637e-01 6.88057244e-01
3.79353344e-01 5.64539790e-01 1.10785151e+00 -2.68246561e-01
3.32462221e-01 2.78279632e-01 -3.61663014e-01 5.84405303e-01
1.69923916e-01 5.20835400e-01 -4.91917372e-01 1.49526820e-01
8.56473565e-01 -3.97215664e-01 -8.76300275e-01 -4.27414596e-01
-7.72613108e-01 4.30034101e-01 3.48112851e-01 1.24190733e-01
-6.31514192e-01 -9.88954455e-02 3.22175831e-01 3.06267530e-01
3.97666633e-01 2.16504246e-01 -1.28047556e-01 -6.07680202e-01
-7.67848194e-01 1.27490416e-01 2.30389521e-01 8.44501078e-01
2.53677428e-01 1.40495032e-01 1.02868848e-01 8.96426380e-01
3.09736341e-01 4.68999624e-01 7.00017810e-01 -1.33052981e+00
2.28669435e-01 7.23796412e-02 3.33618462e-01 -1.17096019e+00
-1.82097718e-01 -5.15427232e-01 -4.50119466e-01 4.18467760e-01
-3.07267578e-03 1.12591423e-01 -4.02979195e-01 1.30375636e+00
2.18107909e-01 2.09026456e-01 1.70716494e-01 1.31028736e+00
6.44672990e-01 5.40903330e-01 -3.79788816e-01 -5.35731554e-01
1.03580415e+00 -3.76084507e-01 -7.52140820e-01 3.23145092e-01
2.72933662e-01 -6.89564586e-01 9.96658087e-01 7.38073826e-01
-1.26189864e+00 -8.02357852e-01 -1.12494385e+00 1.10683322e-01
3.32019366e-02 3.02398670e-03 2.77542084e-01 1.08033955e+00
-1.17635250e+00 3.04225266e-01 -5.80260754e-01 -2.80497134e-01
1.81339383e-01 1.26679480e-01 -3.39953870e-01 -1.33847356e-01
-1.05350780e+00 7.03887224e-01 1.74974561e-01 -6.05487078e-02
-3.62179250e-01 -6.17178082e-01 -8.25996161e-01 -4.35503535e-02
2.99689889e-01 -3.86299878e-01 1.16145802e+00 -7.74945259e-01
-1.65237713e+00 7.04377532e-01 -2.81496216e-02 -6.06497765e-01
5.33933580e-01 -1.47012919e-01 -7.82469749e-01 5.24872303e-01
-2.19475016e-01 4.68021512e-01 8.34249258e-01 -1.10794210e+00
-8.44366431e-01 -3.88976783e-01 1.66564837e-01 5.47245443e-01
1.25501789e-02 9.09550786e-02 -4.07550365e-01 -2.52410740e-01
1.29597053e-01 -6.37162447e-01 3.20936441e-01 7.59105291e-03
1.92278132e-01 2.60417283e-01 6.68004870e-01 -8.73003662e-01
1.44232523e+00 -2.36066985e+00 -2.71496087e-01 -1.79799765e-01
1.26682937e-01 4.26791668e-01 3.26625146e-02 -3.57440412e-02
-2.07743654e-03 -2.58508861e-01 1.50742441e-01 9.25737396e-02
-3.57140571e-01 -1.95088536e-01 3.10408939e-02 6.98728025e-01
-2.31979772e-01 4.73684698e-01 -9.35100198e-01 -3.74788880e-01
6.55857384e-01 7.40691185e-01 -7.81290293e-01 4.43892665e-02
4.50764656e-01 2.81605482e-01 -1.33964345e-01 5.74059963e-01
9.52964604e-01 -4.66488861e-03 -1.95571527e-01 -1.70151219e-01
-5.62495589e-01 5.21804206e-02 -1.23171508e+00 1.63445020e+00
-4.92946982e-01 1.00384486e+00 -7.24034533e-02 -7.10923076e-01
8.00206602e-01 2.37698823e-01 4.12746996e-01 -1.31456661e+00
1.83009923e-01 3.25759143e-01 -2.78210491e-02 -7.79326677e-01
9.97980237e-01 2.68737763e-01 5.34789026e-01 -2.44279400e-01
-1.10013969e-01 -5.62003739e-02 1.48976505e-01 1.13328502e-01
1.00695145e+00 3.44000518e-01 5.03635287e-01 1.50702475e-02
5.67261398e-01 -4.40072924e-01 2.83434629e-01 3.73680085e-01
-6.51526749e-01 7.47155547e-01 -3.42072584e-02 -1.87465281e-03
-1.13518262e+00 -1.22767901e+00 -5.14456689e-01 5.21457911e-01
4.84253407e-01 -3.34034026e-01 -6.62043333e-01 7.82471150e-02
-3.08391720e-01 7.98122108e-01 -3.02608818e-01 -9.44618955e-02
-6.67122453e-02 -3.91277313e-01 2.81513363e-01 2.30067804e-01
7.41241693e-01 -6.69096172e-01 -1.28724897e+00 -8.71019289e-02
-3.86874706e-01 -1.23096955e+00 -2.86534965e-01 -4.34976757e-01
-8.05982292e-01 -9.56881285e-01 -8.85292947e-01 -8.24773684e-02
7.70567805e-02 8.85375857e-01 6.29070103e-01 -3.71063232e-01
-3.28271478e-01 7.78136849e-01 -5.21132052e-01 2.03125607e-02
-1.94741841e-02 -5.48905015e-01 2.17516735e-01 1.19939178e-01
2.26350531e-01 -4.78879690e-01 -1.00253522e+00 4.27157640e-01
-8.80472481e-01 -9.29551497e-02 3.61548841e-01 6.26586199e-01
6.53972030e-01 2.71845162e-01 3.24391216e-01 -2.09479406e-01
6.02128685e-01 -4.79244232e-01 -8.39771867e-01 -1.00123271e-01
-4.32561815e-01 -4.01431352e-01 3.32451850e-01 -2.17861086e-01
-1.28216553e+00 -5.12864113e-01 9.60810408e-02 -6.28540516e-01
-5.99226840e-02 3.84107977e-01 1.25650823e-01 -2.17665106e-01
8.27204943e-01 2.50316381e-01 -1.63548768e-01 -3.12550142e-02
6.66582957e-02 1.17395198e+00 8.04560423e-01 7.58780614e-02
2.39722803e-01 4.28569227e-01 -2.17645407e-01 -1.31495142e+00
-8.11158493e-02 -6.39444113e-01 2.67404946e-03 -7.21112192e-01
7.47836351e-01 -9.68264937e-01 -8.78178120e-01 2.77597219e-01
-7.92574227e-01 5.83741628e-02 -1.86020166e-01 1.18029809e+00
-9.00763631e-01 6.03534818e-01 -3.06661993e-01 -1.26107192e+00
1.14168651e-01 -1.11155963e+00 7.75483251e-01 4.74902242e-01
1.15356825e-01 -5.35331070e-01 2.67273560e-02 5.07510543e-01
4.43513900e-01 1.24483276e-02 3.09136122e-01 1.19367704e-01
-7.42377579e-01 -2.72723556e-01 -4.36209351e-01 6.58945322e-01
-8.52480084e-02 1.46151915e-01 -1.15149844e+00 -1.83599144e-01
2.61305898e-01 -2.17881668e-02 4.81164575e-01 8.79809558e-01
9.95740652e-01 1.19944714e-01 1.98349148e-01 7.74304926e-01
1.64906216e+00 6.74176991e-01 1.18667972e+00 4.38513488e-01
8.40018168e-02 4.02137637e-01 9.45782781e-01 8.62366915e-01
1.30705908e-01 8.91821623e-01 5.22537768e-01 3.20755064e-01
-3.96050662e-02 -8.32107216e-02 3.03512394e-01 7.62697697e-01
-6.62669182e-01 -3.13409418e-01 -5.15368402e-01 3.55186284e-01
-1.44451571e+00 -1.20771182e+00 4.41098697e-02 2.73269701e+00
1.97636768e-01 6.12529702e-02 -9.10810083e-02 3.31309497e-01
5.98344922e-01 4.12959717e-02 -5.24838030e-01 -5.39597750e-01
-1.94953620e-01 -7.13897645e-02 5.80991507e-01 4.51819688e-01
-8.05209696e-01 3.22255313e-01 5.44476080e+00 8.61646056e-01
-1.24206269e+00 -1.31805003e-01 4.19678092e-01 -2.98100710e-01
-1.02964707e-01 -3.45152289e-01 -2.24580571e-01 5.35308063e-01
1.22305906e+00 -4.29482996e-01 6.64863110e-01 8.43177915e-01
7.63689101e-01 -6.56130612e-01 -5.57380855e-01 1.59559107e+00
2.20423669e-01 -1.01235998e+00 -3.99759144e-01 3.54857862e-01
4.95281339e-01 6.15373626e-02 5.37857592e-01 -1.75869428e-02
-4.84850854e-01 -6.68867528e-01 7.47040033e-01 8.54488254e-01
1.14935493e+00 -7.49516368e-01 7.22248614e-01 1.23666056e-01
-1.06527722e+00 -2.94501781e-01 -6.47221863e-01 9.50754527e-03
-3.24921049e-02 2.44415224e-01 -4.84653234e-01 7.81215012e-01
9.77146626e-01 4.54039991e-01 -3.08564812e-01 1.44357157e+00
3.32827032e-01 1.63782746e-01 -6.35163188e-02 1.56862766e-01
-1.99836493e-01 -3.44548374e-01 8.13711703e-01 7.96904147e-01
7.90439308e-01 4.05903935e-01 -5.84559798e-01 5.28411269e-01
2.37952337e-01 2.41452843e-01 -7.55713582e-01 6.80338293e-02
4.35999990e-01 9.12787855e-01 -2.61608541e-01 -1.80101432e-02
-7.25420773e-01 9.64327216e-01 -2.23241478e-01 4.81197447e-01
-9.28591669e-01 -5.40085912e-01 6.77340806e-01 2.15639710e-01
4.20542896e-01 -2.36871958e-01 1.59917533e-01 -1.13312256e+00
1.13147341e-01 -7.68548787e-01 9.27291363e-02 -1.58480370e+00
-5.31032264e-01 7.73117185e-01 1.22298049e-02 -1.71390808e+00
-3.13778907e-01 -5.92570543e-01 3.10457200e-02 8.03283215e-01
-1.51748252e+00 -3.01844895e-01 -5.04296958e-01 7.06866562e-01
6.21831179e-01 -1.78021386e-01 5.61619639e-01 5.91013789e-01
-2.12084100e-01 6.32048786e-01 2.74404258e-01 -4.38256353e-01
6.00192666e-01 -7.99935520e-01 -2.80070081e-02 9.64496255e-01
-4.69830111e-02 3.11646491e-01 9.25393224e-01 -3.42991889e-01
-1.45827746e+00 -5.48537850e-01 4.47065741e-01 2.62295417e-02
3.48833561e-01 8.48502070e-02 -7.19635963e-01 1.55062228e-01
2.48253364e-02 1.50731996e-01 6.01675570e-01 -4.70194668e-01
-1.44321352e-01 -1.22765988e-01 -1.43252265e+00 5.96319973e-01
1.05330813e+00 -6.44805610e-01 -3.28656524e-01 -2.92957425e-01
5.56917608e-01 -4.34827030e-01 -8.08897495e-01 3.97899479e-01
8.57519150e-01 -1.59156764e+00 7.72378325e-01 1.39507502e-01
3.14636588e-01 -4.23605591e-01 -8.18215132e-01 -1.19575870e+00
-2.75198430e-01 -6.58574879e-01 -1.12437427e-01 5.40232420e-01
1.05077930e-01 -7.48549938e-01 4.86867934e-01 4.25621957e-01
-1.24966674e-01 -5.36266565e-01 -1.13217187e+00 -7.58700311e-01
-7.16242909e-01 -7.13305056e-01 7.80588910e-02 5.48003316e-01
1.81077689e-01 -5.83276674e-02 -3.51594836e-01 2.81945556e-01
6.57512307e-01 -4.10066724e-01 5.19581676e-01 -9.15336251e-01
-4.75008130e-01 -1.33084387e-01 -1.03012574e+00 -1.20542312e+00
-6.23078585e-01 -3.97361100e-01 -2.51451731e-01 -1.12359035e+00
-1.27728179e-01 9.35136154e-02 -3.53745252e-01 -7.22312450e-01
1.59797087e-01 3.98863524e-01 1.10698625e-01 8.59799534e-02
-4.80366886e-01 6.57950640e-01 1.21430635e+00 3.30481827e-01
-3.74025881e-01 1.39884681e-01 -4.58112448e-01 5.04129171e-01
6.87760890e-01 1.07794218e-01 -7.14978337e-01 -1.99670359e-01
2.05995679e-01 6.33956790e-01 3.59032482e-01 -1.44051754e+00
1.74598724e-01 1.46351010e-01 2.38124728e-01 -5.09802938e-01
6.90429688e-01 -6.78614974e-01 1.96592674e-01 2.65221983e-01
-1.86715081e-01 1.16831876e-01 1.51286870e-01 6.67024136e-01
-3.53745013e-01 -5.97927123e-02 1.01430476e+00 4.98078465e-02
-9.51225340e-01 7.85693377e-02 -4.94957954e-01 -2.26274848e-01
9.08607423e-01 -8.72922957e-01 -2.80815005e-01 -6.31146491e-01
-4.93235171e-01 -3.22867155e-01 5.37161946e-01 2.46176690e-01
1.15808105e+00 -9.92341638e-01 -5.83995759e-01 3.07675540e-01
1.61862060e-01 -7.18758464e-01 9.64225709e-01 1.01586366e+00
-6.91729605e-01 6.42146409e-01 -4.37635183e-01 -6.30791128e-01
-1.47365701e+00 7.68528342e-01 4.51303750e-01 3.33525717e-01
-5.72179914e-01 7.63696313e-01 2.65861690e-01 3.60389829e-01
8.71911496e-02 -2.35034332e-01 -3.24681908e-01 -3.72181833e-01
8.63175333e-01 8.04816425e-01 1.56566188e-01 -6.82478845e-01
-1.37260497e-01 4.12198603e-01 6.21726327e-02 -3.92439455e-01
1.04158533e+00 -8.18482935e-01 4.94671315e-01 3.86839956e-01
1.12590504e+00 1.48942426e-01 -1.27560377e+00 -1.19047295e-02
-3.79969060e-01 -1.20217693e+00 5.51569581e-01 -5.97241640e-01
-8.13497722e-01 6.95525169e-01 1.30626225e+00 2.28376314e-01
1.56773806e+00 -3.94264966e-01 4.61780876e-01 1.14507385e-01
5.98145247e-01 -1.07974839e+00 -1.04688451e-01 9.15350020e-02
7.53813326e-01 -1.09384334e+00 3.61890602e-03 -4.74951774e-01
-7.32406497e-01 7.89636493e-01 2.76105493e-01 -8.44512582e-02
5.60481727e-01 6.60644248e-02 -3.45762260e-02 1.70424134e-01
-7.12714136e-01 -2.99122214e-01 1.71282157e-01 9.11579847e-01
3.83734256e-01 1.26710972e-02 -4.55792189e-01 1.32248968e-01
-1.02618560e-01 2.99826026e-01 8.08454871e-01 6.99274838e-01
-5.18224180e-01 -2.72369653e-01 -3.91564280e-01 3.08574468e-01
-5.87556303e-01 -1.74812213e-01 3.21721852e-01 5.59150338e-01
-7.65198246e-02 1.18302321e+00 -6.62884163e-03 -5.61579645e-01
4.27926391e-01 -3.94377619e-01 6.69181406e-01 -1.77922437e-03
1.05739214e-01 1.45228177e-01 3.07112075e-02 -8.03266466e-01
-4.93715584e-01 -6.63107276e-01 -9.58560765e-01 -5.21836042e-01
-2.98312008e-01 1.22339159e-01 1.04361880e+00 3.62554312e-01
5.56300640e-01 5.81933796e-01 7.09376693e-01 -8.61330688e-01
-3.16040516e-01 -7.00407267e-01 -9.21058655e-01 4.38946150e-02
3.02511603e-01 -8.00340652e-01 -4.37068492e-01 -5.02773188e-02] | [11.565446853637695, -1.8800292015075684] |
907e9d41-d2c4-4671-bb57-e3503c869aa6 | deepfake-captcha-a-method-for-preventing-fake | 2301.03064 | null | https://arxiv.org/abs/2301.03064v1 | https://arxiv.org/pdf/2301.03064v1.pdf | Deepfake CAPTCHA: A Method for Preventing Fake Calls | Deep learning technology has made it possible to generate realistic content of specific individuals. These `deepfakes' can now be generated in real-time which enables attackers to impersonate people over audio and video calls. Moreover, some methods only need a few images or seconds of audio to steal an identity. Existing defenses perform passive analysis to detect fake content. However, with the rapid progress of deepfake quality, this may be a losing game. In this paper, we propose D-CAPTCHA: an active defense against real-time deepfakes. The approach is to force the adversary into the spotlight by challenging the deepfake model to generate content which exceeds its capabilities. By doing so, passive detection becomes easier since the content will be distorted. In contrast to existing CAPTCHAs, we challenge the AI's ability to create content as opposed to its ability to classify content. In this work we focus on real-time audio deepfakes and present preliminary results on video. In our evaluation we found that D-CAPTCHA outperforms state-of-the-art audio deepfake detectors with an accuracy of 91-100% depending on the challenge (compared to 71% without challenges). We also performed a study on 41 volunteers to understand how threatening current real-time deepfake attacks are. We found that the majority of the volunteers could not tell the difference between real and fake audio. | ['Yisroel Mirsky', 'Fred M. Grabovski', 'Guy Frankovits', 'Lior Yasur'] | 2023-01-08 | null | null | null | null | ['face-swapping'] | ['computer-vision'] | [ 1.05070181e-01 7.10219219e-02 2.09554762e-01 1.64122239e-01
-1.02490103e+00 -1.13662076e+00 6.44956470e-01 -2.35190257e-01
-4.16558832e-01 6.44317508e-01 1.85662225e-01 -4.09532227e-02
5.58587790e-01 -6.42521024e-01 -5.11138380e-01 -6.12451017e-01
-5.50647154e-02 5.65353990e-01 4.59243357e-01 -1.97597057e-01
5.93184568e-02 2.50859469e-01 -1.31160975e+00 7.69703269e-01
2.57741451e-01 5.75188279e-01 -3.74255240e-01 1.32520008e+00
2.00324357e-01 8.20955873e-01 -1.68550706e+00 -7.54897654e-01
3.19285601e-01 -5.72089136e-01 -5.48521399e-01 -3.46207589e-01
7.21977711e-01 -8.78730416e-01 -6.43285215e-01 9.65112805e-01
1.09660327e+00 -5.48570275e-01 1.86402246e-01 -1.53240657e+00
-3.17992181e-01 5.86820662e-01 -3.61560553e-01 3.45797002e-01
7.85884082e-01 5.47909260e-01 4.04791176e-01 -2.48202816e-01
3.72725666e-01 1.37592793e+00 1.08056414e+00 1.00300264e+00
-1.07380795e+00 -1.15559185e+00 -5.36107302e-01 -3.28426249e-02
-1.45296800e+00 -6.27806485e-01 8.14026058e-01 -4.47538525e-01
6.10191524e-01 4.15650129e-01 7.94407189e-01 1.51378953e+00
1.59045327e-02 7.29035258e-01 1.00886738e+00 -4.24740165e-01
2.24446354e-04 2.52633810e-01 -3.56162250e-01 3.47506583e-01
2.75992453e-01 5.13182759e-01 -5.95231771e-01 -6.28610671e-01
5.41429818e-01 -7.79924989e-01 -3.21930677e-01 2.25978643e-01
-1.04205847e+00 1.00449586e+00 8.80137011e-02 3.99422050e-01
-1.42313898e-01 6.26983762e-01 5.96421659e-01 4.65839624e-01
1.50591299e-01 6.87747538e-01 3.53122413e-01 -6.75972044e-01
-1.17132974e+00 4.65247750e-01 8.24155092e-01 2.15639263e-01
6.28748164e-02 5.21580100e-01 8.00628588e-02 4.06589895e-01
-4.23804112e-02 7.57942557e-01 4.42036211e-01 -1.01914299e+00
5.01704216e-01 -2.02990398e-01 2.96059340e-01 -1.52696490e+00
7.49848261e-02 -5.11985302e-01 -3.55825037e-01 2.36935779e-01
8.60812664e-01 -6.91011071e-01 -7.19481111e-01 1.58584487e+00
3.77431437e-02 4.51336145e-01 -5.13506532e-02 9.63213444e-01
5.24189174e-01 8.03445339e-01 -3.17782581e-01 2.21444786e-01
1.28921187e+00 -3.59386444e-01 -8.11774909e-01 -1.87708095e-01
6.96285665e-01 -9.78173375e-01 7.52001345e-01 8.69672716e-01
-1.16022182e+00 -4.36927080e-01 -1.17360914e+00 5.24271250e-01
-2.20502898e-01 -2.28398845e-01 3.55879277e-01 1.71224165e+00
-1.13518739e+00 2.03799963e-01 -4.26141322e-01 -4.68023717e-02
3.46710622e-01 5.07153749e-01 -3.08164299e-01 2.07818374e-01
-1.73375332e+00 7.19588876e-01 2.30443209e-01 -5.82088046e-02
-1.30774128e+00 -4.21279132e-01 -4.51049119e-01 -1.47686481e-01
8.18562657e-02 -3.47876906e-01 1.51674139e+00 -1.29082108e+00
-1.30436909e+00 9.58949924e-01 3.64694715e-01 -8.33999336e-01
1.09328699e+00 -2.52860397e-01 -7.03506291e-01 7.35385120e-01
-1.32330712e-02 8.19563329e-01 1.17730916e+00 -1.34102380e+00
-4.79337007e-01 4.03521881e-02 1.97577387e-01 -5.42080477e-02
-3.52666050e-01 2.38999322e-01 -3.56614776e-02 -7.92134285e-01
-6.47947550e-01 -1.05321026e+00 3.73254597e-01 -7.69825131e-02
-4.33063835e-01 2.77746797e-01 1.40838301e+00 -6.60612762e-01
1.03773487e+00 -2.18520880e+00 -5.75788021e-01 1.73596472e-01
2.76996166e-01 9.17605937e-01 -6.64936677e-02 5.71148276e-01
5.37236109e-02 3.77022684e-01 2.93902367e-01 -3.75719368e-01
-7.77876303e-02 -4.86761592e-02 -7.30993688e-01 6.33385658e-01
-6.28919825e-02 6.09513164e-01 -9.74458635e-01 -3.19311380e-01
-2.18023732e-02 7.09217370e-01 -6.48737788e-01 1.14910200e-01
1.14204019e-01 2.14632466e-01 -9.85299647e-02 3.89862776e-01
5.76637983e-01 4.37581390e-01 -7.27153048e-02 1.71994895e-01
-1.19127315e-02 2.58233249e-01 -9.32768524e-01 9.90884602e-01
-1.75280169e-01 1.23406541e+00 3.86783421e-01 -7.50823379e-01
7.39865720e-01 8.02549839e-01 2.10691616e-01 -5.23153484e-01
4.38603312e-01 3.23634207e-01 1.95825949e-01 -3.22946429e-01
5.24683297e-01 -1.40596002e-01 -4.05190617e-01 6.88520133e-01
-3.43973219e-01 -2.50046134e-01 -2.29226619e-01 3.13846499e-01
1.16237092e+00 -5.09857178e-01 -4.33784693e-01 9.70938280e-02
3.25411230e-01 -2.50815451e-02 5.16812652e-02 1.23963046e+00
-4.99676049e-01 7.05542088e-01 4.31067944e-01 -3.82708371e-01
-1.01454747e+00 -1.02074361e+00 4.48257178e-01 7.84649909e-01
-5.67951389e-02 -4.26615149e-01 -1.37776875e+00 -6.10417187e-01
-2.44371533e-01 5.37214875e-01 -4.09623891e-01 -3.92516822e-01
-7.64854252e-01 -4.16210234e-01 1.90577376e+00 8.61773416e-02
8.44459474e-01 -9.52901125e-01 -7.54941344e-01 3.84503245e-01
-5.35277665e-01 -1.20003080e+00 -4.68059927e-01 -5.19896865e-01
-1.78422183e-01 -8.21100295e-01 -8.77542317e-01 -8.05836380e-01
1.94199026e-01 3.56408775e-01 7.75962174e-01 2.44112700e-01
-3.12283188e-01 4.65985566e-01 -3.28386903e-01 -6.01746738e-01
-9.58990514e-01 -2.23177806e-01 1.25880465e-01 -9.75751728e-02
3.62349689e-01 -4.22249556e-01 -5.07637799e-01 3.78667653e-01
-9.71270084e-01 -3.89204800e-01 1.93290010e-01 6.97011709e-01
-3.17427576e-01 4.14039254e-01 5.64132571e-01 -4.98143882e-01
9.77474689e-01 -1.74025789e-01 -3.03666204e-01 -1.76478624e-01
1.90501392e-01 -4.23592299e-01 4.42643970e-01 -1.02417600e+00
-5.91912329e-01 -1.45980284e-01 -3.00082147e-01 -3.00110728e-01
-6.79738298e-02 -4.52442430e-02 6.06796816e-02 -3.45052391e-01
9.06103551e-01 3.19074005e-01 8.58618170e-02 -4.99716960e-02
1.22876264e-01 1.03791237e+00 8.84905756e-01 -3.53776157e-01
1.08062625e+00 5.71133614e-01 -4.99416023e-01 -1.07590902e+00
-2.69328445e-01 -7.06047937e-02 1.22489199e-01 -7.10309207e-01
7.15739489e-01 -9.53484237e-01 -1.04429543e+00 1.07205868e+00
-1.45891964e+00 -2.66393036e-01 1.02451108e-01 4.36437190e-01
-2.99054682e-01 7.48266697e-01 -9.49422181e-01 -9.51900363e-01
-1.98070690e-01 -1.01188910e+00 8.22533250e-01 -1.80207446e-01
-5.56643665e-01 -8.05261910e-01 2.47861281e-01 7.22095847e-01
5.89090586e-01 4.25089359e-01 2.10637614e-01 -6.65742517e-01
-2.89082170e-01 -7.73924351e-01 3.89582328e-02 2.85004348e-01
6.92556892e-03 -9.22233537e-02 -1.32324672e+00 -5.48152268e-01
2.56313421e-02 -5.10198176e-01 5.29971480e-01 1.01778604e-01
7.30744779e-01 -6.00910544e-01 -1.15086004e-01 1.36912540e-01
8.34759951e-01 3.32278311e-01 1.04860890e+00 2.88034350e-01
4.40755934e-01 7.27076411e-01 3.64266843e-01 3.02752256e-01
-1.54170441e-02 8.78208697e-01 4.77154881e-01 -1.97652820e-03
-3.05460632e-01 -4.99233425e-01 7.89696753e-01 3.60131651e-01
2.76944548e-01 -7.99456239e-01 -8.07047844e-01 4.68795270e-01
-1.17620742e+00 -1.42406750e+00 -2.98390955e-01 2.15305448e+00
6.72649622e-01 6.91294253e-01 6.42283916e-01 6.06461525e-01
1.17145073e+00 1.06772907e-01 9.88268852e-02 -7.37304509e-01
-9.73854586e-02 1.66427702e-01 6.19275451e-01 7.48528898e-01
-1.05188215e+00 1.03360105e+00 6.41092014e+00 7.88108706e-01
-1.45990705e+00 2.09618434e-01 2.78516263e-01 -2.21265629e-01
-5.07587567e-02 -3.16873133e-01 -6.13499761e-01 8.73869598e-01
1.17413449e+00 6.29307851e-02 2.41612598e-01 5.37496030e-01
3.23188871e-01 9.14112628e-02 -6.92680597e-01 1.17091751e+00
3.37497890e-01 -1.28285098e+00 -8.89323652e-02 3.11969012e-01
2.74481386e-01 -3.40470284e-01 3.73218477e-01 1.43756479e-01
4.87739474e-01 -1.00857747e+00 9.31940317e-01 -2.12896213e-01
8.16653192e-01 -7.96516538e-01 7.85137951e-01 3.56179804e-01
-7.54817188e-01 -3.76334488e-02 -1.33722927e-02 -1.71912298e-01
3.41506004e-01 3.03374827e-01 -1.27064073e+00 -1.89942166e-01
5.20852506e-01 -1.92037418e-01 -2.68723994e-01 1.07596016e+00
-2.03759491e-01 9.92100954e-01 -5.16811371e-01 -6.75389320e-02
4.39993471e-01 7.84654140e-01 8.48924994e-01 1.43754947e+00
4.23044831e-01 -1.05240293e-01 -6.93179369e-02 4.60736752e-01
9.62523222e-02 -4.55833942e-01 -7.71871269e-01 -3.32330883e-01
6.20951414e-01 6.42265558e-01 -5.35545826e-01 -2.70289063e-01
1.83279335e-01 1.17754257e+00 -5.73912442e-01 -1.61889102e-02
-1.05040646e+00 -6.65858805e-01 5.73388100e-01 4.27406937e-01
1.23831481e-01 -2.26500899e-01 1.69819877e-01 -9.04227734e-01
-3.03230226e-01 -1.31580627e+00 3.07100594e-01 -6.94078386e-01
-9.42946017e-01 6.69761181e-01 -1.10118479e-01 -1.11072826e+00
-6.38536930e-01 -2.80081153e-01 -6.20018244e-01 4.73976135e-01
-8.53605151e-01 -1.08642459e+00 -1.66275412e-01 6.20174050e-01
2.41920292e-01 -1.57262266e-01 6.67352080e-01 5.14577985e-01
3.96077670e-02 1.04746759e+00 -2.42966145e-01 6.97538376e-01
7.57053375e-01 -7.86280394e-01 7.17955232e-01 9.16793287e-01
2.13924110e-01 3.22242469e-01 1.21775806e+00 -8.17243874e-01
-9.67232406e-01 -5.20658612e-01 7.11302876e-01 -4.79668796e-01
6.34652257e-01 -6.90847278e-01 -7.78083265e-01 2.34835699e-01
1.91049978e-01 -3.67084056e-01 6.79367840e-01 -6.39521301e-01
-5.55725694e-01 4.08939598e-03 -1.46477592e+00 4.33087915e-01
5.03509343e-01 -8.04444373e-01 -5.55910408e-01 3.07245832e-02
5.84978402e-01 -1.92316040e-01 -9.75785181e-02 -3.82299237e-02
8.91201735e-01 -1.20278800e+00 9.95222211e-01 -3.44983429e-01
-5.96376285e-02 -2.28542611e-01 2.15869069e-01 -1.10392225e+00
9.57905650e-02 -1.10037899e+00 2.98792068e-02 1.33035159e+00
-4.05664183e-03 -6.01977527e-01 1.26305950e+00 1.66090637e-01
1.23799667e-01 2.88273036e-01 -1.01435101e+00 -9.32477176e-01
-2.60403380e-04 -6.59786165e-01 3.44961226e-01 1.09115601e+00
-6.95613399e-02 -1.21254899e-01 -1.00508046e+00 3.54598880e-01
6.63594127e-01 -6.96952581e-01 1.01097667e+00 -9.64695334e-01
-5.49780011e-01 -2.08669960e-01 -8.91860008e-01 -8.39066863e-01
-1.57030866e-01 -2.53433913e-01 -7.51111582e-02 -7.87325621e-01
-3.03606302e-01 -1.33508414e-01 2.61830956e-01 2.71799892e-01
4.37102556e-01 9.92222607e-01 4.39795256e-01 1.50955305e-01
-1.76347643e-01 -1.54395006e-03 7.58623898e-01 -2.10229427e-01
-2.17449173e-01 1.58132300e-01 -4.67481226e-01 6.57864332e-01
1.00401151e+00 -7.80453742e-01 -3.59198719e-01 -3.92482668e-01
1.71859875e-01 8.28188434e-02 7.31185794e-01 -1.39732718e+00
1.05803065e-01 3.58732581e-01 2.85359234e-01 -3.07624251e-01
7.40804374e-01 -6.12838626e-01 2.27533147e-01 1.12827563e+00
-3.51920575e-01 -3.70257050e-02 3.24977487e-01 3.94254804e-01
-1.35798484e-01 -1.50690481e-01 7.84651697e-01 -1.82499811e-01
-2.26544589e-01 -3.26003104e-01 -1.08924866e+00 1.15109399e-01
8.26587737e-01 -5.03061175e-01 -7.36677289e-01 -1.31584263e+00
-5.19853711e-01 -1.89458653e-01 4.83204067e-01 3.47672224e-01
6.74778044e-01 -1.08702934e+00 -6.84720397e-01 9.24337134e-02
-3.40903878e-01 -6.82892382e-01 2.74730235e-01 3.18906039e-01
-1.13493788e+00 3.59000266e-01 -2.58341432e-01 -4.26082700e-01
-1.79757607e+00 4.15584087e-01 4.74490821e-01 9.73048508e-02
-2.10614458e-01 9.59947228e-01 9.45591275e-03 1.68277636e-01
4.99902010e-01 2.97871023e-01 2.72461902e-02 9.53107104e-02
9.66659427e-01 2.46571288e-01 -2.41646051e-01 -8.15439880e-01
-4.36382473e-01 1.56047031e-01 -1.67012870e-01 -5.80168664e-01
8.19362640e-01 1.29815727e-01 3.08375001e-01 -5.08571081e-02
1.21569848e+00 5.93351424e-01 -1.14804685e+00 2.30139777e-01
-3.71006906e-01 -7.74160087e-01 -5.27942181e-02 -7.33001828e-01
-8.63328218e-01 1.13752651e+00 8.04469466e-01 7.29059756e-01
8.01502109e-01 -1.99311465e-01 1.29329681e+00 1.33431345e-01
3.05350363e-01 -8.69233251e-01 7.22233772e-01 1.08329333e-01
5.58885753e-01 -5.75140297e-01 -3.93596560e-01 -3.35412413e-01
-5.91165245e-01 1.02742493e+00 4.49886858e-01 -1.31461322e-01
1.08369954e-01 4.68215853e-01 3.50459099e-01 1.58210117e-02
-4.17070597e-01 2.60902762e-01 -2.53532052e-01 1.12887609e+00
-9.33939964e-02 -5.74224815e-02 1.12295277e-01 1.71239600e-01
-7.33536303e-01 -1.52853966e-01 1.03290725e+00 8.12397659e-01
-6.24397218e-01 -1.18504918e+00 -9.78833258e-01 -1.98201030e-01
-1.09308922e+00 2.09548742e-01 -1.11697471e+00 7.51781881e-01
2.18712687e-01 1.43477082e+00 -5.82596995e-02 -7.62489796e-01
-5.02473935e-02 -2.81642944e-01 3.19210738e-01 -2.19029248e-01
-8.22526395e-01 9.04764831e-02 5.49003661e-01 -1.76324472e-01
-1.91956356e-01 -5.33611357e-01 -7.75556624e-01 -8.82386923e-01
-3.14401120e-01 4.64925796e-01 6.61768794e-01 5.89693367e-01
-2.52062827e-02 -1.95558190e-01 6.40941679e-01 -7.91646957e-01
-3.89907688e-01 -5.00878394e-01 -1.84405476e-01 3.03494930e-01
6.24620318e-01 -2.58721501e-01 -7.07215786e-01 -1.25481635e-01] | [12.577014923095703, 1.3027251958847046] |
b4570c5c-945d-47e0-a781-eafb4d55e82f | arm-order-recognition-in-multi-armed-bandit | 2005.13085 | null | https://arxiv.org/abs/2005.13085v1 | https://arxiv.org/pdf/2005.13085v1.pdf | Arm order recognition in multi-armed bandit problem with laser chaos time series | By exploiting ultrafast and irregular time series generated by lasers with delayed feedback, we have previously demonstrated a scalable algorithm to solve multi-armed bandit (MAB) problems utilizing the time-division multiplexing of laser chaos time series. Although the algorithm detects the arm with the highest reward expectation, the correct recognition of the order of arms in terms of reward expectations is not achievable. Here, we present an algorithm where the degree of exploration is adaptively controlled based on confidence intervals that represent the estimation accuracy of reward expectations. We have demonstrated numerically that our approach did improve arm order recognition accuracy significantly, along with reduced dependence on reward environments, and the total reward is almost maintained compared with conventional MAB methods. This study applies to sectors where the order information is critical, such as efficient allocation of resources in information and communications technology. | ['Nicolas Chauvet', 'Naoki Narisawa', 'Mikio Hasegawa', 'Makoto Naruse'] | 2020-05-26 | null | null | null | null | ['irregular-time-series'] | ['time-series'] | [ 3.87700784e-05 -2.52047777e-01 -4.02437180e-01 1.86525747e-01
-7.80460358e-01 -9.97041464e-01 5.17817378e-01 -2.08173454e-01
-6.22223556e-01 1.08172548e+00 -4.55084592e-02 -5.50903499e-01
-8.12472880e-01 -6.41685426e-01 -3.48936051e-01 -9.07861829e-01
-2.11139560e-01 6.28957808e-01 -2.46678740e-01 -5.20773567e-02
9.00032818e-01 8.18114221e-01 -1.33429241e+00 1.49092600e-01
7.06286490e-01 1.56644678e+00 -8.04099292e-02 8.38852286e-01
-6.30283505e-02 6.34821236e-01 -5.23667455e-01 -1.41854025e-02
7.20552146e-01 -6.75039589e-01 -3.76318842e-01 -4.17184174e-01
-3.98567766e-01 -3.72730851e-01 -3.96429241e-01 8.67715120e-01
4.85476106e-01 1.18728615e-01 7.40906000e-01 -8.90230358e-01
-1.98860422e-01 6.68824136e-01 -7.55446851e-01 6.94020510e-01
1.63783580e-02 2.69122154e-01 1.09661138e+00 -3.72216016e-01
1.12833746e-01 8.32640827e-01 2.19590917e-01 2.02219814e-01
-1.29495549e+00 -9.10160184e-01 -2.27287620e-01 2.12481350e-01
-1.18672955e+00 -4.51764166e-01 7.32902110e-01 -2.17616379e-01
5.35918832e-01 5.27741253e-01 7.93451965e-01 3.41759920e-01
3.06083471e-01 4.10894185e-01 1.51767659e+00 -6.91892862e-01
5.02484441e-01 -1.56700891e-02 6.93089887e-02 5.61436832e-01
3.85736436e-01 8.00658405e-01 -7.26122439e-01 -4.05442506e-01
8.09709251e-01 -2.53227830e-01 1.05983220e-01 -1.93981349e-01
-1.09083402e+00 1.04552209e+00 2.40052581e-01 4.52926368e-01
-8.57301116e-01 6.22890651e-01 -1.93028286e-01 4.25959349e-01
4.43392515e-01 8.19757402e-01 -1.46077164e-02 -2.54189372e-01
-1.24798393e+00 -2.37942040e-02 7.16670454e-01 4.64965105e-01
6.11349106e-01 1.03849478e-01 -7.81535923e-01 -3.42416428e-02
4.73830625e-02 8.11774313e-01 4.04660612e-01 -1.14810860e+00
3.17871988e-01 3.96142751e-02 8.67282450e-01 -6.98334336e-01
-5.47469735e-01 -7.75725901e-01 -6.41343951e-01 1.23988822e-01
6.68501019e-01 -5.90075314e-01 -7.12006271e-01 1.40098226e+00
1.42855808e-01 2.17389762e-01 7.35442191e-02 7.82110095e-01
-1.20375559e-01 5.45910120e-01 -5.20512879e-01 -8.29031944e-01
1.15955877e+00 -3.47529233e-01 -8.20461154e-01 4.17985208e-03
3.50488245e-01 -7.71243453e-01 4.47956473e-01 4.68461126e-01
-1.19344008e+00 1.06561810e-01 -8.45304310e-01 9.19680893e-01
2.37139966e-02 1.38712555e-01 9.05130982e-01 1.17540383e+00
-9.54453707e-01 5.16598821e-01 -5.01366794e-01 1.99788168e-01
4.51452047e-01 6.09653711e-01 5.08597314e-01 4.15642679e-01
-8.39648843e-01 8.52412283e-01 2.46232256e-01 2.53728777e-01
-4.59277481e-01 -5.64657450e-01 -6.26062155e-02 1.63108498e-01
2.36009136e-01 -4.08606231e-01 1.37422156e+00 -5.59198737e-01
-1.91747677e+00 2.13370860e-01 -2.54331052e-01 -9.50657010e-01
4.74633992e-01 1.84811756e-01 -1.66979749e-02 4.43789631e-01
-1.06632821e-01 1.09853923e-01 1.18382180e+00 -7.84935415e-01
-7.06934214e-01 -2.67337680e-01 -2.13461429e-01 -5.74826449e-02
-2.08595902e-01 -5.76797202e-02 6.69875443e-01 -2.29396939e-01
2.19568714e-01 -1.14531362e+00 -4.26377743e-01 -4.89992678e-01
-2.74140984e-01 -1.32501304e-01 4.16883498e-01 -5.42581417e-02
1.10514379e+00 -1.74227905e+00 -3.05126131e-01 7.04975426e-01
-1.02157362e-01 -3.88613865e-02 -1.55090943e-01 5.86917996e-01
2.80661345e-01 1.19198434e-01 3.44364613e-01 3.52813780e-01
1.04540552e-03 -2.09969088e-01 -9.09381986e-01 5.76451004e-01
3.36929224e-02 9.33555484e-01 -6.68111205e-01 -7.14399666e-02
-1.94000285e-02 -2.81533748e-01 -4.34005260e-01 7.86741972e-02
-2.97286063e-01 6.79515064e-01 -8.94572079e-01 5.70907772e-01
3.88330519e-01 -5.31114101e-01 -4.27349322e-02 1.58410057e-01
-5.54261506e-01 7.69097060e-02 -1.09253514e+00 1.08783591e+00
-4.41307455e-01 4.50595319e-01 -2.94964500e-02 -1.00318360e+00
8.16565752e-01 2.05739081e-01 8.76565933e-01 -8.64320159e-01
2.03997657e-01 3.82960498e-01 2.42695019e-01 -1.36858404e-01
5.63311160e-01 -1.66446447e-01 -1.78689584e-02 1.19159555e+00
-6.49250031e-01 -3.80132049e-01 5.38903028e-02 -1.07614532e-01
1.10839736e+00 -5.59435546e-01 1.32668331e-01 -2.54766732e-01
1.46332800e-01 1.76890537e-01 -1.29309162e-01 1.26465464e+00
-1.45968765e-01 -2.11525466e-02 3.91929328e-01 -4.44236249e-01
-9.08874333e-01 -8.85386348e-01 -7.08150938e-02 9.33255613e-01
1.50621146e-01 2.08691016e-01 -8.85581598e-02 -1.08234614e-01
2.65851974e-01 8.21172416e-01 -4.83669132e-01 3.69119160e-02
-4.61885512e-01 -9.82406795e-01 3.71637464e-01 -6.92962483e-02
5.07751644e-01 -6.85043454e-01 -1.10663223e+00 4.06800151e-01
9.65456665e-02 -8.86611819e-01 -4.34622556e-01 2.86308199e-01
-7.57582963e-01 -8.99260700e-01 -7.11334586e-01 4.08849977e-02
3.09219927e-01 4.81286556e-01 6.79337502e-01 -3.60347897e-01
-2.84119219e-01 6.62693739e-01 -1.41750380e-01 -6.93363011e-01
-1.52046129e-01 -1.76789969e-01 2.04143405e-01 1.86747998e-01
1.93658963e-01 -5.13176084e-01 -9.39206362e-01 2.37351537e-01
-3.81117314e-01 -3.63540471e-01 9.29045856e-01 8.81483972e-01
3.26992422e-01 -2.81997845e-02 8.13060343e-01 -1.59845278e-01
9.20331538e-01 -2.08829165e-01 -1.28979361e+00 2.33021632e-01
-6.91509604e-01 3.85961771e-01 1.33966610e-01 -5.37769973e-01
-6.99127972e-01 2.18677521e-02 7.06913292e-01 -2.81454176e-01
3.91535163e-01 5.79141259e-01 1.16030669e+00 -6.22890890e-01
5.98375857e-01 2.94591218e-01 3.79614294e-01 1.13390319e-01
4.71686810e-01 6.09310389e-01 2.13815495e-01 -5.58482766e-01
6.40996039e-01 4.84563053e-01 6.69730604e-01 -5.94543099e-01
-7.18846023e-01 -3.57288659e-01 -1.08587861e-01 -2.74507701e-01
2.46635720e-01 -6.51748955e-01 -1.52976358e+00 4.07582894e-02
-1.16657519e+00 -1.38259262e-01 -3.95997167e-01 1.01611435e+00
-9.58674729e-01 1.26912892e-01 -4.35277551e-01 -1.74278843e+00
-3.84926856e-01 -6.70000970e-01 4.81897742e-01 2.67160088e-01
-2.67618503e-02 -3.35003436e-01 -2.52571404e-02 1.12817891e-01
8.26111019e-01 4.52342406e-02 8.57186139e-01 -5.65478742e-01
-1.03754735e+00 -3.69735003e-01 -1.04360290e-01 -3.79560560e-01
1.44704878e-01 -2.49033123e-01 -7.96052694e-01 -3.72289270e-01
2.32319370e-01 -3.46382082e-01 7.58512914e-01 1.07803309e+00
1.09823918e+00 -5.07708848e-01 -2.64894545e-01 -4.21872176e-03
1.13949561e+00 4.79985893e-01 4.01156604e-01 3.83750856e-01
-4.12552685e-01 4.74130839e-01 7.60037780e-01 1.02300870e+00
-1.94712698e-01 4.92388487e-01 5.12046695e-01 4.78555888e-01
5.87565780e-01 1.56978473e-01 1.83474749e-01 8.63354951e-02
-2.28972822e-01 -8.26730058e-02 -6.01553500e-01 3.37510467e-01
-1.82928169e+00 -1.19212747e+00 2.91468531e-01 2.53941870e+00
6.31802678e-01 8.23047757e-02 4.93528396e-01 -1.07891880e-01
8.43660891e-01 -1.12085000e-01 -9.48655486e-01 -5.55496871e-01
1.68812662e-01 4.11514997e-01 1.01049209e+00 2.26654902e-01
-6.50356591e-01 4.98722792e-01 7.94930124e+00 9.49365914e-01
-1.02642775e+00 -9.05125123e-03 7.84733355e-01 -6.75359845e-01
-1.54974446e-01 -1.98480919e-01 -5.94779968e-01 5.43737590e-01
1.20000052e+00 -6.50291502e-01 9.01045084e-01 3.16534311e-01
5.53954124e-01 -4.14674640e-01 -7.19540358e-01 9.76008058e-01
-6.01757705e-01 -1.72934449e+00 -4.45941716e-01 3.91162306e-01
7.10196495e-01 -9.21311677e-02 7.27550268e-01 -6.70267791e-02
5.05173028e-01 -1.11933374e+00 5.29038727e-01 5.92323244e-01
7.71411359e-01 -9.90018904e-01 5.54303110e-01 6.15978539e-01
-8.16586375e-01 -8.10069084e-01 -2.03005835e-01 -4.34140623e-01
-1.10796012e-01 8.12332988e-01 -8.77560318e-01 3.76502872e-01
3.16001266e-01 -6.84234202e-02 1.66180015e-01 1.18413496e+00
9.56102684e-02 5.87864280e-01 -7.22974420e-01 -8.28951597e-01
4.87773299e-01 -6.12196028e-01 5.99305630e-01 5.53985417e-01
1.00172162e+00 5.71554482e-01 -1.55455828e-01 8.18648160e-01
3.07734966e-01 -2.05640271e-01 -5.59518337e-01 -5.03385365e-01
7.89684296e-01 9.57745016e-01 -6.96430445e-01 2.11107247e-02
-1.60256661e-02 3.17784488e-01 -1.09900221e-01 3.27318996e-01
-5.26146472e-01 -1.48578733e-01 2.69482762e-01 -1.62419632e-01
6.77803755e-01 -4.44399327e-01 -4.71316189e-01 -5.46535552e-01
-1.79335073e-01 -3.32618892e-01 4.49622035e-01 -3.62937540e-01
-1.13046396e+00 3.85353982e-01 -1.64727509e-01 -1.29478908e+00
-7.07369685e-01 -3.88938069e-01 -5.69579780e-01 9.34109926e-01
-1.42486823e+00 -5.10860503e-01 4.44590002e-01 4.07882452e-01
4.60468084e-02 -4.79861677e-01 5.55797338e-01 -4.44482446e-01
-2.77358770e-01 3.80782038e-01 6.79100811e-01 -4.24058318e-01
1.94042131e-01 -8.19514930e-01 -2.97650844e-01 4.73799020e-01
1.81661084e-01 4.15810943e-01 1.19781661e+00 -4.06477243e-01
-1.66101146e+00 -5.58158636e-01 1.71769246e-01 -1.57265231e-01
1.05290914e+00 8.83387029e-02 -3.14852260e-02 1.93328366e-01
2.36644536e-01 -5.87043762e-02 6.99051261e-01 -4.94139604e-02
4.87639988e-03 -2.25658908e-01 -1.34379590e+00 3.02409172e-01
6.01348758e-01 -4.41266388e-01 -1.51562020e-01 5.89394569e-01
2.70102978e-01 -3.28429341e-01 -4.68065798e-01 1.96069643e-01
8.51150393e-01 -9.22214806e-01 7.61131823e-01 -5.20938158e-01
1.62033252e-02 2.89060678e-02 -1.45735130e-01 -1.38459361e+00
-4.29223359e-01 -1.24942851e+00 -4.56705391e-01 4.30897713e-01
4.45209384e-01 -1.11978304e+00 1.12179005e+00 5.92569947e-01
4.49413508e-01 -6.47518635e-01 -1.70714545e+00 -1.08795094e+00
-3.66926007e-02 -1.51744649e-01 6.63412690e-01 2.31924698e-01
1.83853284e-01 -1.60123676e-01 -5.44477940e-01 2.11867586e-01
7.37537205e-01 8.27121079e-01 3.49579781e-01 -1.08896434e+00
-3.61571223e-01 -6.09059930e-01 -1.30432814e-01 -9.62486386e-01
-1.10195279e-01 -5.66803455e-01 -8.27682298e-03 -9.59061682e-01
1.17340744e-01 -5.24290383e-01 -8.25831950e-01 -5.10372259e-02
3.35378200e-01 2.09376872e-01 1.94234252e-01 3.80216449e-01
-4.47718322e-01 4.66323256e-01 1.17855835e+00 -1.45658195e-01
-5.08323073e-01 7.23365188e-01 -6.63537145e-01 5.38780689e-02
9.29819882e-01 -5.14863253e-01 -1.77450001e-01 1.37216002e-01
3.10793102e-01 5.58746815e-01 3.82440656e-01 -7.77563572e-01
6.47236526e-01 -4.92316484e-01 3.08848649e-01 -9.18304741e-01
4.63883758e-01 -7.79501617e-01 4.57557254e-02 1.09636807e+00
-6.36950552e-01 1.55141264e-01 1.56435519e-01 9.64649439e-01
3.23225766e-01 -2.73065269e-01 5.73289514e-01 -5.57295047e-02
-9.21648666e-02 1.58028051e-01 -7.77852178e-01 -4.78518993e-01
1.16291559e+00 1.21075027e-02 -5.02123415e-01 -8.14637423e-01
-4.00571346e-01 2.15807304e-01 -2.95727342e-01 -2.04344764e-01
5.10771871e-01 -1.33432019e+00 -6.22093976e-01 -4.55115214e-02
-4.66757357e-01 -7.75823116e-01 -8.38449895e-02 9.91105735e-01
-1.73357930e-02 1.06062651e+00 -8.71522278e-02 -5.20990729e-01
-9.20710266e-01 5.95060170e-01 2.71530300e-01 -5.41391730e-01
-1.47230811e-02 5.65701246e-01 -4.86438215e-01 4.39930022e-01
-5.80116874e-03 -1.24467298e-01 2.27367394e-02 2.23369390e-01
6.38691545e-01 4.04982865e-01 -6.37811497e-02 -2.99789831e-02
-1.25135154e-01 5.91472745e-01 -4.65592481e-02 -5.42260766e-01
1.24691355e+00 -1.59915566e-01 -1.76769450e-01 3.06516171e-01
5.18710911e-01 -7.04935715e-02 -1.09265900e+00 -3.01534444e-01
1.03833184e-01 -7.08217144e-01 2.98074573e-01 -7.44760275e-01
-6.38328254e-01 5.78024387e-01 8.65434885e-01 8.11733127e-01
9.78781879e-01 -2.63929576e-01 5.01287282e-01 7.66800940e-01
8.87151778e-01 -1.03288507e+00 4.45118964e-01 3.79517704e-01
6.40162647e-01 -9.71703470e-01 1.68179899e-01 -1.28381103e-01
-3.25249374e-01 1.22707772e+00 -1.24190666e-01 -1.07895061e-01
5.93767822e-01 1.05856642e-01 -3.65940213e-01 -1.51723459e-01
-7.45788455e-01 -4.09838200e-01 1.16688363e-01 1.62022650e-01
-1.85313597e-01 1.81642026e-01 -5.09374976e-01 3.60590309e-01
-2.27553427e-01 5.80933578e-02 6.19923055e-01 7.84387052e-01
-8.75090539e-01 -7.78813481e-01 -6.19391441e-01 9.38255548e-01
-5.30077159e-01 -4.70780134e-02 -5.58569990e-02 4.01245594e-01
-7.72290766e-01 1.34538841e+00 2.08046526e-01 -2.35409632e-01
-6.61586598e-02 -1.55131087e-01 6.84991002e-01 -2.46451378e-01
-2.14075282e-01 2.34269947e-02 8.73785466e-02 -4.39555168e-01
-3.72512668e-01 -5.76243937e-01 -9.27886486e-01 -4.37841505e-01
-5.69380879e-01 6.48004830e-01 7.55162776e-01 1.06032991e+00
6.11252666e-01 5.66829853e-02 1.39820516e+00 -8.65183890e-01
-1.24256909e+00 -6.21873081e-01 -8.28015745e-01 -6.11390591e-01
6.01577461e-01 -5.86748540e-01 -5.71527004e-01 -8.84193301e-01] | [4.564005374908447, 3.315481424331665] |
e58e6048-ea76-4358-8b68-8a0ce49b37d8 | two-parents-one-child-dual-transfer-for-low | null | null | https://aclanthology.org/2021.findings-acl.241 | https://aclanthology.org/2021.findings-acl.241.pdf | Two Parents, One Child: Dual Transfer for Low-Resource Neural Machine Translation | null | ['Qun Liu', 'Liangyou Li', 'Meng Zhang'] | null | null | null | null | findings-acl-2021-8 | ['low-resource-neural-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] | [-7.375491619110107, 3.628530263900757] |
976ab4ae-b7c5-4001-9ea8-6a7d242e97a3 | multi-timescale-event-detection-in | 2211.10721 | null | https://arxiv.org/abs/2211.10721v1 | https://arxiv.org/pdf/2211.10721v1.pdf | Multi-timescale Event Detection in Nonintrusive Load Monitoring based on MDL Principle | Load event detection is the fundamental step for the event-based non-intrusive load monitoring (NILM). However, existing event detection methods with fixed parameters may fail in coping with the inherent multi-timescale characteristics of events and their event detection accuracy is easily affected by the load fluctuation. In this regard, this paper extends our previously designed two-stage event detection framework, and proposes a novel multi-timescale event detection method based on the principle of minimum description length (MDL). Following the completion of step-like event detection in the first stage, a long-transient event detection scheme with variable-length sliding window is designed for the second stage, which is intended to provide the observation and characterization of the same event at different time scales. In that, the context information in the aggregated load data is mined by motif discovery, and then based on the MDL principle, the proper observation scales are selected for different events and the corresponding detection results are determined. In the post-processing step, a load fluctuation location method based on voice activity detection (VAD) is proposed to identify and remove the unreasonable events caused by fluctuations. Based on newly proposed evaluation metrics, the comparison tests on public and private datasets demonstrate that our method achieves higher detection accuracy and integrity for events of various appliances across different scenarios. | ['Yixin Yu', 'Zishuai Liu', 'Wenpeng Luan', 'Jianfeng Zhang', 'Bo Liu'] | 2022-11-19 | null | null | null | null | ['activity-detection', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['computer-vision', 'knowledge-base', 'miscellaneous', 'time-series'] | [ 2.60568172e-01 -6.32363141e-01 -2.56000787e-01 -1.46068379e-01
-5.59160233e-01 -3.43328238e-01 2.40456462e-01 5.77066720e-01
-1.59941971e-01 6.17214382e-01 2.35537246e-01 -7.82361776e-02
-4.10152525e-01 -8.54032695e-01 1.90547686e-02 -7.20675349e-01
-1.05495185e-01 2.09031124e-02 5.05264938e-01 1.87702850e-01
3.22106361e-01 4.97548431e-01 -1.59273756e+00 8.14792439e-02
6.98582590e-01 1.13352585e+00 2.73200154e-01 3.42423022e-01
-2.71047592e-01 5.02757311e-01 -9.08318102e-01 4.86223489e-01
-1.51166946e-01 -6.65904403e-01 -2.46285886e-01 1.50280759e-01
-6.82303607e-01 -2.64954120e-01 6.53982535e-02 7.15705633e-01
7.17550576e-01 2.62058884e-01 3.57088536e-01 -1.45221686e+00
1.38255209e-01 5.12817144e-01 -6.93169177e-01 7.39701867e-01
6.21898890e-01 3.05117480e-03 7.61352956e-01 -5.66081643e-01
2.74940372e-01 9.00038064e-01 5.54922163e-01 1.58168763e-01
-1.11012638e+00 -7.47373939e-01 2.16988429e-01 3.99222165e-01
-1.70274234e+00 -1.97492138e-01 1.05238962e+00 -1.65467709e-01
9.12783146e-01 4.74893391e-01 3.83344024e-01 9.86334383e-01
3.90562475e-01 4.23029870e-01 6.43459916e-01 -3.58800650e-01
4.90802258e-01 -1.84181973e-01 1.31606370e-01 4.43438534e-03
2.90827483e-01 -1.38030946e-01 -5.29532313e-01 -6.13634229e-01
3.50209326e-01 3.91838580e-01 -8.12654495e-02 3.50713879e-01
-9.71870780e-01 6.16738081e-01 -3.09565693e-01 5.36476135e-01
-7.85349250e-01 -5.25604129e-01 9.64765787e-01 1.20388448e-01
4.32084799e-01 -1.78788036e-01 -3.96837890e-01 -3.74740630e-01
-1.06416428e+00 -1.01154614e-02 8.18217218e-01 5.76384366e-01
4.52828825e-01 3.49348672e-02 -6.06767654e-01 3.68633211e-01
2.88993746e-01 2.08199754e-01 7.49977648e-01 -2.31891781e-01
2.36174226e-01 9.29139316e-01 1.93269417e-01 -1.03794861e+00
-6.73319936e-01 -3.21688771e-01 -8.87523293e-01 -8.20810124e-02
-5.97615028e-03 -1.96133465e-01 -3.06899458e-01 1.69242871e+00
6.57614470e-01 6.02667511e-01 -1.49799492e-02 4.67813700e-01
4.08318102e-01 7.98420489e-01 4.16614294e-01 -1.25122309e+00
1.62275839e+00 -1.89670105e-03 -1.14676929e+00 2.83364296e-01
1.25027865e-01 -8.42151403e-01 8.47017527e-01 3.67300719e-01
-6.53565228e-01 -5.31107247e-01 -1.05983019e+00 6.20392561e-01
-2.26689577e-01 -4.16775197e-02 3.40639323e-01 5.04577458e-01
-1.51424110e-01 2.90264755e-01 -7.31761634e-01 -7.27072358e-01
2.67591514e-02 2.13418454e-02 2.44928196e-01 5.18914282e-01
-1.20835173e+00 5.40194035e-01 6.72718525e-01 -1.79749753e-04
-7.35220373e-01 -4.42604482e-01 -4.22225773e-01 1.15324043e-01
3.98561716e-01 -1.35426074e-01 1.05668128e+00 -4.34074253e-01
-1.31910002e+00 2.49950796e-01 -4.27728474e-01 -5.18175900e-01
3.53003055e-01 2.36025695e-02 -1.05312991e+00 1.72208622e-01
1.12476923e-01 -5.34871817e-01 7.93539464e-01 -6.10632241e-01
-9.09647167e-01 -2.32118323e-01 -4.80040461e-01 2.27259099e-01
-3.64701688e-01 5.00239789e-01 -1.72724426e-01 -7.53022850e-01
9.07307491e-02 -3.06327552e-01 5.01699895e-02 -7.75714636e-01
-4.06852573e-01 -5.96717179e-01 1.11779761e+00 -5.27613997e-01
2.18506026e+00 -2.39527750e+00 -4.21777219e-01 4.52892631e-01
-1.21207193e-01 1.86433315e-01 4.24927175e-01 8.57825220e-01
-7.89537206e-02 -2.58295864e-01 -3.33855785e-02 -1.07403532e-01
1.97263788e-02 4.95419577e-02 -5.00199318e-01 4.26277786e-01
-2.37465333e-02 3.35088730e-01 -8.13154817e-01 -6.62416399e-01
6.57572567e-01 1.33971691e-01 6.87293336e-02 4.66654867e-01
-8.24338123e-02 5.69296479e-01 -5.52009046e-01 6.17625535e-01
5.22911370e-01 -1.91635489e-01 2.63544202e-01 -3.27417463e-01
-4.13233042e-01 2.45369330e-01 -1.76253557e+00 1.07071054e+00
-4.17448819e-01 8.19270685e-02 -2.35902309e-01 -9.56988454e-01
1.04539895e+00 7.79739678e-01 8.74607503e-01 -6.46738827e-01
1.47908628e-01 2.02555418e-01 -3.62979144e-01 -6.51557386e-01
1.99460864e-01 1.86476961e-01 -1.77387536e-01 5.35020590e-01
-5.02378285e-01 6.92798376e-01 3.84965718e-01 -1.00603737e-01
1.27601516e+00 -1.88458130e-01 8.06466758e-01 -3.28241736e-02
6.71128869e-01 -2.81010360e-01 1.00763738e+00 5.50347149e-01
-4.61982459e-01 1.18950605e-01 2.03434527e-01 -3.56347144e-01
-5.68170071e-01 -1.09715080e+00 -3.14322501e-01 1.03407168e+00
2.49254778e-01 -5.05161643e-01 -4.69573200e-01 -4.52839255e-01
-1.27648771e-01 8.88411760e-01 -3.58009428e-01 -2.50690490e-01
-5.06975830e-01 -1.33211052e+00 4.69630718e-01 3.03391963e-01
5.12718141e-01 -1.38005769e+00 -9.17704344e-01 9.39999461e-01
-2.36833990e-01 -8.35481286e-01 -4.16872352e-01 3.37181568e-01
-6.37527645e-01 -1.09668815e+00 4.56035370e-04 -5.50596774e-01
2.78610736e-01 1.95218265e-01 7.99405038e-01 -1.91180855e-01
-4.30072069e-01 3.86113450e-02 -4.63378727e-01 -5.16427815e-01
-3.86872172e-01 -3.70422117e-02 1.36212721e-01 4.19138461e-01
9.40344632e-01 -9.78591740e-01 -8.43598127e-01 4.58710521e-01
-1.05552971e+00 -4.74941552e-01 4.48934048e-01 3.89060110e-01
8.48208189e-01 6.67782903e-01 1.30580652e+00 -3.96642864e-01
7.29314446e-01 -9.55128849e-01 -6.00578487e-01 1.89514697e-01
-8.37594390e-01 -6.03550136e-01 8.34639907e-01 -8.69069338e-01
-1.02616143e+00 4.00667358e-03 3.64989741e-03 -3.58963013e-02
-3.45359147e-01 2.74898738e-01 -4.51998413e-01 5.65066516e-01
4.73760128e-01 4.87270981e-01 -4.10207748e-01 -5.71250916e-01
-9.87184048e-02 8.59517515e-01 3.20247710e-01 -1.59203574e-01
4.41611469e-01 5.19345582e-01 -1.18728377e-01 -5.94633758e-01
-4.66944963e-01 -6.80436969e-01 -1.22042507e-01 -2.10033655e-01
8.05602193e-01 -8.11152518e-01 -9.94782507e-01 4.55790192e-01
-1.02303374e+00 3.55681300e-01 -2.87798405e-01 6.21713161e-01
-1.96204737e-01 4.17587698e-01 -6.75916255e-01 -1.22648501e+00
-6.30011618e-01 -7.39269912e-01 8.34594429e-01 5.34319460e-01
-5.73315322e-01 -6.28534973e-01 9.80601758e-02 -2.63543576e-01
3.56603742e-01 4.90095228e-01 7.84783900e-01 -9.40225184e-01
-2.32573807e-01 -4.92275923e-01 3.54818776e-02 -4.08415496e-03
5.89174449e-01 2.59327260e-03 -7.86818743e-01 -2.82274693e-01
3.93654764e-01 4.09597307e-01 2.59071201e-01 4.51229662e-01
1.11728668e+00 -2.11848721e-01 -4.65518773e-01 1.90808922e-01
1.52280855e+00 6.29054010e-01 6.37240887e-01 3.83568436e-01
3.14072549e-01 1.55465841e-01 7.94894993e-01 1.08089924e+00
2.66147435e-01 5.48841476e-01 4.08898205e-01 -2.92753726e-02
3.10994864e-01 -3.24604511e-01 5.01817465e-01 7.39353001e-01
3.62881422e-01 -4.34979498e-01 -3.46195459e-01 5.71983457e-01
-1.99644256e+00 -1.26561785e+00 -3.97607014e-02 2.42328072e+00
8.10205758e-01 5.12775004e-01 4.73438054e-01 6.58535004e-01
1.13839567e+00 1.66451842e-01 -7.68770814e-01 -3.59918207e-01
-1.79361269e-01 -5.90779819e-03 3.07785034e-01 1.53438179e-02
-9.48892474e-01 2.09707201e-01 5.81210756e+00 1.14116156e+00
-9.78092492e-01 1.63627282e-01 2.78879285e-01 -2.36895576e-01
2.00898439e-01 -1.37958631e-01 -9.37800348e-01 1.10141194e+00
1.16904223e+00 -4.33114886e-01 2.12525725e-01 5.83292544e-01
1.10302794e+00 -3.86894733e-01 -7.32732892e-01 1.01184869e+00
-3.50384146e-01 -8.98848176e-01 -9.90618765e-02 -3.49033140e-02
2.55261630e-01 -2.63566434e-01 -5.31299591e-01 8.07622448e-02
-2.23396495e-01 -4.03600872e-01 3.41694921e-01 6.37980461e-01
3.44604820e-01 -9.57194865e-01 5.56271195e-01 3.98956120e-01
-1.76053929e+00 -2.57184148e-01 -9.09760743e-02 -1.28210858e-01
4.28354830e-01 1.28455877e+00 -6.08308375e-01 7.90297627e-01
5.88774621e-01 3.85292798e-01 -9.15927812e-02 9.95920718e-01
3.04077026e-02 1.09024191e+00 -7.09658921e-01 5.90647310e-02
-4.43171173e-01 -1.83520056e-02 8.05228114e-01 1.20876694e+00
3.88542831e-01 -3.80052216e-02 4.50860083e-01 5.66792905e-01
2.64555752e-01 4.12958443e-01 -1.82632685e-01 3.72221917e-01
1.08734488e+00 1.27332211e+00 -1.04237175e+00 -3.00198585e-01
-4.42413568e-01 9.04113829e-01 -2.92193174e-01 2.03449354e-01
-1.07372081e+00 -7.34229267e-01 4.00379598e-01 -1.40150651e-01
3.05185646e-01 -3.02069681e-03 -2.08775640e-01 -8.96459341e-01
3.76956195e-01 -6.18217945e-01 8.98805022e-01 -2.10165098e-01
-1.38766551e+00 2.60534048e-01 -1.12202950e-03 -1.60614836e+00
-3.11766982e-01 2.46896267e-01 -1.37711906e+00 8.52098584e-01
-1.25408888e+00 -6.43307447e-01 -1.56806573e-01 7.28892148e-01
5.47973990e-01 2.85212129e-01 9.03335989e-01 5.90442359e-01
-1.00869310e+00 4.88725692e-01 2.08554208e-01 -1.48637131e-01
5.41476607e-01 -9.29605186e-01 2.74312310e-02 9.85011160e-01
-2.93253034e-01 3.09172839e-01 8.39610100e-01 -9.50580060e-01
-1.15048146e+00 -1.06954527e+00 9.84085739e-01 6.42520711e-02
6.95450127e-01 -2.42075309e-01 -1.00850832e+00 3.56986672e-01
-1.28182039e-01 -2.88231939e-01 7.82493472e-01 -1.35639921e-01
4.69080023e-02 -4.45652395e-01 -1.33413899e+00 2.76444614e-01
7.37896562e-01 -4.72072423e-01 -7.16742516e-01 1.95994571e-01
6.74800694e-01 -8.99560228e-02 -1.03810966e+00 4.66441840e-01
4.09298837e-01 -6.66952431e-01 6.96038544e-01 -1.47089496e-01
-3.92200142e-01 -8.46296728e-01 -1.81338757e-01 -7.50431299e-01
-2.90063769e-01 -9.93386924e-01 -4.88704890e-01 1.88727939e+00
2.03680322e-02 -6.06050491e-01 2.11852282e-01 1.02374658e-01
1.66534737e-01 -5.03083110e-01 -1.14634681e+00 -8.03415895e-01
-9.43834424e-01 -4.69055116e-01 9.63991344e-01 7.05964983e-01
6.44535795e-02 2.22406894e-01 -2.81404555e-01 5.69243789e-01
4.75989699e-01 4.25008416e-01 3.23555380e-01 -1.19715512e+00
-1.13407128e-01 -2.15999901e-01 -2.90280461e-01 -4.98793691e-01
-3.51029634e-01 -4.03949171e-01 -7.50396475e-02 -1.12282419e+00
1.80784255e-01 3.47550847e-02 -8.77567112e-01 2.69472212e-01
-3.63927186e-01 4.15445231e-02 -2.27337912e-01 4.65563148e-01
-8.07697177e-01 3.83750945e-01 3.56903613e-01 4.02351230e-01
-7.38904297e-01 3.12046051e-01 -2.44776383e-01 6.88016295e-01
1.13854825e+00 -4.36002463e-01 -4.34983552e-01 4.15546149e-01
8.80851503e-03 6.33531958e-02 2.79352874e-01 -1.03140211e+00
4.72035222e-02 -4.85897899e-01 3.65298390e-01 -1.07852519e+00
-2.11811244e-01 -8.49972069e-01 4.28233355e-01 5.11578739e-01
-4.64244373e-03 2.28873581e-01 2.51502454e-01 7.68698812e-01
-7.12382421e-02 2.31685415e-02 4.40571547e-01 1.96913511e-01
-8.08834314e-01 1.88807055e-01 -5.71532309e-01 -1.04020871e-01
1.18967330e+00 -9.90720466e-02 -8.55063125e-02 4.38057818e-02
-7.65894175e-01 2.20286131e-01 4.74235527e-02 3.44572097e-01
2.13925809e-01 -1.42698371e+00 -6.29134476e-01 1.51434287e-01
3.21217418e-01 -3.13547432e-01 3.88729721e-01 8.52589250e-01
1.21056072e-01 2.82142609e-02 1.56388357e-01 -5.01504421e-01
-1.11504900e+00 6.70217454e-01 1.02175884e-01 -4.97901559e-01
-7.23445117e-01 8.43657553e-02 -2.12689683e-01 5.02013147e-01
3.67007345e-01 -1.42616466e-01 -4.27767247e-01 4.48973536e-01
9.55739200e-01 7.42221057e-01 2.66102970e-01 -4.29366320e-01
-5.77199996e-01 2.28256375e-01 1.26367748e-01 1.14035793e-01
1.06271172e+00 -7.45888233e-01 -1.35776952e-01 7.86718905e-01
9.90942359e-01 1.58462897e-01 -9.68312442e-01 -2.65958518e-01
2.95020252e-01 -2.60960180e-02 -6.06023669e-02 -6.31281555e-01
-7.07032859e-01 2.38170698e-01 9.36396241e-01 7.90638804e-01
1.72029948e+00 -2.24200353e-01 1.15407717e+00 -2.76111275e-01
2.58622468e-01 -1.19396818e+00 -4.16771412e-01 6.86660558e-02
3.80433261e-01 -9.02038217e-01 -1.46115154e-01 -1.73421249e-01
-1.27110660e-01 9.14114177e-01 4.52786654e-01 1.18282922e-01
7.24269331e-01 2.84123152e-01 -1.93876266e-01 -7.92483464e-02
-6.23623908e-01 -2.94582456e-01 -1.19581580e-01 3.25676292e-01
8.32094774e-02 3.23724598e-01 -1.08516848e+00 8.54826331e-01
2.60372430e-01 2.36498520e-01 1.87533811e-01 1.05658376e+00
-5.80074549e-01 -9.57314491e-01 -5.56977212e-01 4.93092179e-01
-6.07932568e-01 1.69705376e-01 1.09385163e-01 3.45841557e-01
4.29905772e-01 1.45693231e+00 2.01372564e-01 -4.30045038e-01
4.76896465e-01 3.38676780e-01 -1.51056468e-01 -3.51222068e-01
-5.48780739e-01 2.23011389e-01 -1.34898171e-01 -6.05792105e-01
-4.12034005e-01 -8.04470897e-01 -1.57180226e+00 -2.50908524e-01
-3.83472800e-01 2.59376675e-01 4.65038121e-01 1.00207686e+00
6.86331511e-01 7.15082645e-01 1.22715843e+00 -4.47978050e-01
-4.94131088e-01 -1.03880966e+00 -7.00252593e-01 6.34610474e-01
2.11695820e-01 -5.74022830e-01 -6.84002578e-01 -1.00965679e-01] | [6.168606758117676, 2.646420478820801] |
4ddb5b4d-fcad-4971-9e80-ec9041474e5c | the-computational-limits-of-deep-learning | 2007.05558 | null | https://arxiv.org/abs/2007.05558v2 | https://arxiv.org/pdf/2007.05558v2.pdf | The Computational Limits of Deep Learning | Deep learning's recent history has been one of achievement: from triumphing over humans in the game of Go to world-leading performance in image classification, voice recognition, translation, and other tasks. But this progress has come with a voracious appetite for computing power. This article catalogs the extent of this dependency, showing that progress across a wide variety of applications is strongly reliant on increases in computing power. Extrapolating forward this reliance reveals that progress along current lines is rapidly becoming economically, technically, and environmentally unsustainable. Thus, continued progress in these applications will require dramatically more computationally-efficient methods, which will either have to come from changes to deep learning or from moving to other machine learning methods. | ['Kristjan Greenewald', 'Keeheon Lee', 'Neil C. Thompson', 'Gabriel F. Manso'] | 2020-07-10 | null | null | null | null | ['game-of-go'] | ['playing-games'] | [ 6.36763424e-02 -2.91839302e-01 -7.01589212e-02 -1.56571984e-01
-7.41622746e-01 -4.92972702e-01 7.30637014e-01 -1.86885640e-01
-5.79011381e-01 4.40194070e-01 1.57970294e-01 -6.64461434e-01
-8.78389552e-03 -6.07990801e-01 -2.23716781e-01 -5.61852574e-01
-1.94781899e-01 3.82030338e-01 1.33781908e-02 -4.00484264e-01
4.71926570e-01 6.53092384e-01 -1.37883520e+00 -2.77447049e-02
4.25949693e-01 8.93971443e-01 -1.87803701e-01 7.71850228e-01
-1.66188821e-01 1.01511180e+00 -4.10684377e-01 -6.03512228e-01
1.54736012e-01 -5.50189555e-01 -9.70853090e-01 1.97486021e-02
5.19910492e-02 -1.84765533e-01 -6.62043225e-03 8.64436984e-01
6.72225773e-01 1.10181302e-01 2.92793453e-01 -9.59220648e-01
-6.77866399e-01 6.98085800e-02 -6.02464080e-01 2.39080369e-01
1.19227208e-01 3.66631240e-01 1.08662009e+00 -7.88852811e-01
4.75952566e-01 1.01180613e+00 8.30514669e-01 3.96132529e-01
-7.12251008e-01 -4.67164963e-01 -1.67439029e-01 2.01323330e-01
-1.03160787e+00 -5.11270881e-01 6.02387071e-01 -6.69824183e-01
1.21222186e+00 3.18729103e-01 7.63228238e-01 7.53288031e-01
2.25496590e-01 6.72591329e-01 1.05819023e+00 -5.37241280e-01
2.15756491e-01 -4.69298847e-02 -5.05655944e-01 8.66871655e-01
2.25019574e-01 1.90390602e-01 -5.38535714e-01 1.81469634e-01
5.53550363e-01 -2.99265742e-01 -1.97214857e-01 -7.45888352e-02
-1.09867883e+00 7.96875179e-01 1.62830755e-01 6.22496605e-01
-1.10623091e-01 8.87566656e-02 6.22908413e-01 3.77216011e-01
4.52126533e-01 8.20851624e-01 -4.57996964e-01 -9.80816483e-01
-9.39461529e-01 2.79676676e-01 9.01048124e-01 3.99905533e-01
4.04372722e-01 4.13907796e-01 5.79988956e-01 7.35319078e-01
-8.59205648e-02 2.76943535e-01 7.58267879e-01 -1.08888626e+00
5.46771474e-02 4.94361579e-01 -3.17924172e-02 -1.19480407e+00
-5.03672361e-01 -5.02191067e-01 -7.65241325e-01 6.58489704e-01
5.57549477e-01 -2.11347148e-01 -9.53823984e-01 1.37679100e+00
8.54320526e-02 -2.49451518e-01 -3.73872787e-01 9.92513835e-01
4.18962777e-01 4.55076784e-01 -3.54120992e-02 1.67080835e-01
1.07295024e+00 -7.54091620e-01 -3.89094174e-01 -2.71027178e-01
5.58956563e-01 -8.19899499e-01 1.18406022e+00 8.69962335e-01
-1.08556366e+00 -4.54230428e-01 -1.15317929e+00 -2.56289423e-01
-3.47484291e-01 -2.85867304e-01 9.23876286e-01 9.45278168e-01
-9.28244889e-01 7.56229043e-01 -8.50488722e-01 -4.19010490e-01
5.82759142e-01 5.63066006e-01 -4.27431822e-01 6.19070344e-02
-9.35112238e-01 1.15374410e+00 8.77125636e-02 2.45872363e-01
-4.72141713e-01 -2.96613097e-01 -6.49278700e-01 -9.11666974e-02
4.05498743e-01 -6.70990407e-01 1.28582489e+00 -1.06620181e+00
-1.47351754e+00 1.14996302e+00 1.37504160e-01 -4.90280062e-01
5.65714002e-01 -2.15977073e-01 -3.17629665e-01 -2.66746610e-01
-4.34258431e-01 4.55431581e-01 8.35487545e-01 -7.38119602e-01
-9.14891243e-01 -4.24561679e-01 -7.49527737e-02 2.60345161e-01
-5.14426112e-01 4.37482983e-01 -7.83724934e-02 -2.79436052e-01
1.57028958e-01 -1.00244665e+00 -2.56918192e-01 6.89485669e-02
9.14230496e-02 -3.21821421e-01 7.33976185e-01 -4.91663814e-01
1.02003801e+00 -2.19767833e+00 5.20904250e-02 -3.14298481e-01
3.79626274e-01 3.19869846e-01 1.50907069e-01 4.85297382e-01
1.05394185e-01 1.19476736e-01 1.46079063e-02 -5.34603372e-02
5.83539680e-02 2.95757819e-02 -1.38186783e-01 5.97797096e-01
1.23659924e-01 9.15951908e-01 -1.02670431e+00 -1.77285254e-01
2.09027186e-01 5.16616285e-01 -3.97086740e-01 -4.28052276e-01
1.95227966e-01 3.83038849e-01 -2.45030701e-01 6.44256592e-01
1.40745431e-01 -3.07332098e-01 1.97252244e-01 2.93629140e-01
-4.58131641e-01 3.55876029e-01 -5.32697618e-01 1.69736278e+00
-3.18616062e-01 9.42510724e-01 2.76548207e-01 -1.18902194e+00
6.54214084e-01 2.16074899e-01 7.10269451e-01 -9.09005284e-01
4.93913621e-01 4.18347061e-01 4.84040558e-01 -7.24622250e-01
6.27700984e-01 -5.01551628e-01 1.22685581e-01 3.09339136e-01
-1.07534073e-01 -4.30766374e-01 -1.75669968e-01 -3.01485270e-01
9.03699815e-01 2.31936291e-01 1.06903948e-01 -2.55906343e-01
2.87395030e-01 3.32271546e-01 5.84994555e-01 4.13598001e-01
-6.11169815e-01 2.32978046e-01 1.63837180e-01 -6.85650527e-01
-1.19059122e+00 -8.41189563e-01 -7.04807509e-03 1.24985147e+00
-3.91049474e-01 -1.21864825e-01 -7.74543524e-01 -3.61987144e-01
-2.41959170e-01 1.56465754e-01 -1.69218317e-01 -6.33071363e-02
-7.34255493e-01 -8.20616245e-01 4.73073214e-01 4.33672220e-01
4.94591594e-01 -9.01177883e-01 -9.45774615e-01 1.38313189e-01
2.44786531e-01 -1.08988595e+00 2.95160357e-02 2.05686361e-01
-9.80660379e-01 -8.36010158e-01 -7.36479044e-01 -7.70685434e-01
8.01588595e-02 2.23036423e-01 1.02797687e+00 1.12744309e-01
-5.14557183e-01 3.02449435e-01 -2.28699788e-01 -7.54983723e-01
-2.87091792e-01 2.05394194e-01 1.84880525e-01 -4.65258747e-01
5.52210808e-01 -7.40799785e-01 -5.67738116e-01 -1.33997008e-01
-6.42862737e-01 -1.03736408e-02 5.68559587e-01 8.44668925e-01
6.39006421e-02 2.30835736e-01 7.52246678e-01 -7.32294559e-01
5.85123777e-01 -3.88290137e-01 -3.62733096e-01 -2.01800987e-01
-8.09510648e-01 -2.02727571e-01 4.72622514e-01 -7.36919865e-02
-6.92605495e-01 8.84804782e-03 -4.87268776e-01 -1.32920668e-01
-1.28369913e-01 6.19534612e-01 1.81633309e-02 -1.95952635e-02
6.78759038e-01 6.57959133e-02 9.26147401e-02 -3.50203574e-01
3.99972558e-01 8.15367341e-01 7.65464604e-01 -6.51985526e-01
7.78259814e-01 4.46824670e-01 4.09983471e-02 -1.27649760e+00
-5.28106272e-01 -9.83874723e-02 -4.20756400e-01 -3.65005195e-01
8.50504160e-01 -6.52656734e-01 -6.73826694e-01 5.88098943e-01
-8.71759653e-01 -3.22133422e-01 -2.51312017e-01 5.96651375e-01
-2.90610105e-01 2.98352450e-01 -5.72613657e-01 -6.73194289e-01
-1.02782734e-01 -1.16175401e+00 5.90452850e-01 2.49084264e-01
-4.48117316e-01 -1.09402108e+00 1.50075689e-01 5.61722815e-01
5.25723636e-01 1.92799300e-01 7.50516951e-01 -6.20384991e-01
-5.50044358e-01 -5.09699941e-01 -1.29633069e-01 6.17570221e-01
2.86754429e-01 -1.30065037e-02 -1.19541872e+00 -1.96864024e-01
2.57591367e-01 -4.90931720e-01 5.29525936e-01 1.97271243e-01
1.08771634e+00 -8.57983828e-02 -7.51895607e-02 5.56188285e-01
1.30971181e+00 6.12723410e-01 5.57022750e-01 4.65301216e-01
5.74367702e-01 6.64803982e-01 2.14161173e-01 7.22815171e-02
2.68352240e-01 4.81616378e-01 2.45929047e-01 -2.44504765e-01
-8.64141136e-02 -1.38979256e-01 1.05551913e-01 1.19603884e+00
-5.18094420e-01 5.36546446e-02 -1.22214866e+00 4.70045298e-01
-1.59743738e+00 -1.17113495e+00 7.67224841e-03 2.10112619e+00
5.43601453e-01 4.59880918e-01 2.64524281e-01 3.64836931e-01
2.42402077e-01 9.26079378e-02 -7.03485668e-01 -8.29674959e-01
1.60452604e-01 2.66611725e-01 1.84110060e-01 1.64370149e-01
-8.28517497e-01 8.94522548e-01 7.68298578e+00 7.85773158e-01
-1.74365878e+00 5.97105920e-02 8.37354064e-01 3.49656455e-02
-1.81705028e-01 -3.66570115e-01 -3.24901462e-01 3.75910401e-01
7.94840574e-01 -2.34657288e-01 6.41482711e-01 9.35020268e-01
2.11173251e-01 -3.16198379e-01 -7.73274004e-01 1.18219531e+00
9.61527303e-02 -1.34616590e+00 -3.04500699e-01 3.30217063e-01
4.98162836e-01 3.11835617e-01 4.20578688e-01 2.54452288e-01
2.83864379e-01 -1.30797970e+00 5.56636810e-01 2.93699764e-02
6.82639122e-01 -6.32306218e-01 5.25510728e-01 3.33098114e-01
-7.73153841e-01 -2.86863148e-01 -1.17899448e-01 -6.93143606e-01
-1.02651782e-01 6.14143729e-01 -6.70905948e-01 2.60914892e-01
6.95547104e-01 3.43595237e-01 -1.41981319e-01 9.64083493e-01
2.76259407e-02 6.74974263e-01 1.99041981e-02 -1.58813581e-01
3.43920082e-01 -2.50352383e-01 4.58523661e-01 1.28331542e+00
7.35004619e-02 6.67555183e-02 -7.90515728e-03 2.52641290e-01
-1.05771795e-01 2.08406262e-02 -8.07598174e-01 -5.73489428e-01
4.70020950e-01 1.30561388e+00 -1.02805090e+00 -2.12082654e-01
-8.17439616e-01 1.01057947e+00 9.10975784e-02 -3.78202796e-02
-8.80025744e-01 -3.80270243e-01 8.93403947e-01 1.79678313e-02
-8.35290272e-03 -7.24975169e-01 -6.81477308e-01 -8.92432988e-01
-5.29383644e-02 -9.53422487e-01 1.74369253e-02 -4.81932551e-01
-8.67333710e-01 6.15372002e-01 -5.61717451e-01 -8.71035755e-01
-2.71512806e-01 -7.29460895e-01 -3.95331711e-01 7.41026580e-01
-1.41514111e+00 -8.78475249e-01 -4.90256539e-03 1.34392440e-01
5.09507477e-01 -2.72454500e-01 7.86728203e-01 5.81475496e-01
-3.47034842e-01 4.07357305e-01 1.05412357e-01 2.43023306e-01
4.25361127e-01 -8.52062225e-01 4.92756754e-01 6.13433003e-01
4.46290195e-01 3.91167670e-01 7.02246308e-01 6.39281003e-03
-1.48154593e+00 -3.71287197e-01 9.22772586e-01 -4.93682444e-01
8.55152547e-01 -4.13833946e-01 -4.94214833e-01 5.72198391e-01
1.97505847e-01 -3.45351666e-01 6.39919698e-01 5.03910601e-01
-3.52175564e-01 -1.37040198e-01 -7.58902907e-01 5.64690948e-01
7.87818313e-01 -9.05636311e-01 -3.61382842e-01 1.40602887e-01
2.92020500e-01 -4.30340916e-01 -5.83477437e-01 4.51152146e-01
9.83955204e-01 -9.85191703e-01 8.37125182e-01 -8.80172193e-01
6.28269017e-01 -1.42138015e-04 -1.03715666e-01 -1.06579900e+00
-3.99760127e-01 -8.08193743e-01 3.19111973e-01 8.39063585e-01
2.65289068e-01 -3.60089302e-01 1.04501140e+00 7.91761041e-01
-5.41479409e-01 -1.25647080e+00 -8.66225898e-01 -8.20025861e-01
5.13069928e-01 -7.36081362e-01 3.60888541e-01 1.16933596e+00
1.72639757e-01 5.42962432e-01 -2.56901532e-01 -3.46757293e-01
2.16283351e-01 5.36683910e-02 7.26152837e-01 -1.14695299e+00
-2.79455423e-01 -8.62198949e-01 -8.72647464e-01 -8.44716728e-01
-2.87953645e-01 -6.63672328e-01 1.78554989e-02 -1.45178771e+00
1.76699281e-01 -1.80732980e-01 -2.45262980e-01 5.38569570e-01
6.78148791e-02 5.62170506e-01 2.99517989e-01 2.00667530e-01
-3.68324786e-01 8.04720521e-02 1.37869847e+00 -6.30721375e-02
-9.96331424e-02 -1.49868056e-01 -9.62628186e-01 1.06595922e+00
8.44470203e-01 5.62659465e-02 -2.96687573e-01 -5.94317913e-01
6.35708511e-01 -1.84512243e-01 1.73516482e-01 -1.32952261e+00
1.94356054e-01 -2.08928227e-01 2.50405967e-01 3.41672525e-02
4.32775170e-01 -6.22779071e-01 5.80415800e-02 5.42723060e-01
-2.30038483e-02 4.78050075e-02 2.76880503e-01 7.16008842e-02
-2.22387567e-01 9.12684724e-02 9.61627364e-01 -2.43669525e-01
-9.65820312e-01 1.26449034e-01 -4.93761390e-01 2.51576573e-01
8.16048145e-01 -5.81530154e-01 -1.63544372e-01 -4.09668446e-01
-8.09449911e-01 -2.61966467e-01 4.42375124e-01 4.23941255e-01
3.20547253e-01 -8.53570879e-01 -3.94812912e-01 -7.43911490e-02
-3.41756076e-01 -2.70987213e-01 9.91718173e-02 9.19856250e-01
-5.73907971e-01 5.68292916e-01 -3.04918051e-01 -3.82596761e-01
-1.05402565e+00 3.81659448e-01 3.93940210e-01 -6.24691844e-02
-8.08398426e-01 1.06973600e+00 -1.59922853e-01 -1.71355247e-01
3.05039346e-01 -1.38502106e-01 1.70466408e-01 -1.15252800e-01
5.18089473e-01 5.88559389e-01 1.69820771e-01 -2.79797196e-01
-3.07419151e-01 3.31088483e-01 -6.98976442e-02 -2.10000664e-01
1.47225773e+00 2.08165079e-01 -7.32652843e-02 5.15725195e-01
9.73066449e-01 2.69242764e-01 -1.01997960e+00 4.01028454e-01
1.17881395e-01 -1.90535247e-01 2.41809145e-01 -8.74041319e-01
-1.11121023e+00 1.05411136e+00 4.47305590e-01 2.67485112e-01
1.15429246e+00 -3.03528547e-01 1.02569973e+00 3.89212936e-01
6.25547051e-01 -1.26038134e+00 -8.60156566e-02 6.33395374e-01
4.36852843e-01 -1.10464430e+00 2.11308375e-01 1.56499207e-01
-3.55607957e-01 9.81472433e-01 2.78294802e-01 7.74325132e-02
8.99684787e-01 4.52136457e-01 1.29786059e-01 -3.79963934e-01
-4.23138171e-01 -2.45760500e-01 -1.33329809e-01 5.63230455e-01
8.83207560e-01 4.10578400e-02 -4.11206394e-01 2.23866105e-01
-4.74514902e-01 2.69201756e-01 2.04462945e-01 9.35146391e-01
-8.12343061e-01 -1.18010294e+00 -3.01859379e-01 4.36392486e-01
-6.39277637e-01 -8.51968005e-02 -4.33011651e-01 8.52668226e-01
3.83366823e-01 7.99912930e-01 2.67186463e-02 -4.64591891e-01
9.34572145e-02 8.42851251e-02 7.01418519e-01 -4.40991759e-01
-4.86390918e-01 -2.33477447e-02 1.34002835e-01 -4.26168591e-01
-3.23514372e-01 -6.65977359e-01 -1.23007226e+00 -8.84697437e-01
-1.66798085e-01 1.09430090e-01 1.15056551e+00 9.96192992e-01
3.77193987e-01 2.42171019e-01 3.02519709e-01 -9.23003197e-01
-5.44600070e-01 -5.99567413e-01 -6.10865891e-01 -2.42684528e-01
2.31265143e-01 -3.47220689e-01 -5.03607243e-02 -1.51498660e-01] | [8.990997314453125, 6.385815143585205] |
939a4d2e-33cf-4826-a2f9-84812d1834a4 | chid-a-large-scale-chinese-idiom-dataset-for | 1906.01265 | null | https://arxiv.org/abs/1906.01265v3 | https://arxiv.org/pdf/1906.01265v3.pdf | ChID: A Large-scale Chinese IDiom Dataset for Cloze Test | Cloze-style reading comprehension in Chinese is still limited due to the lack of various corpora. In this paper we propose a large-scale Chinese cloze test dataset ChID, which studies the comprehension of idiom, a unique language phenomenon in Chinese. In this corpus, the idioms in a passage are replaced by blank symbols and the correct answer needs to be chosen from well-designed candidate idioms. We carefully study how the design of candidate idioms and the representation of idioms affect the performance of state-of-the-art models. Results show that the machine accuracy is substantially worse than that of human, indicating a large space for further research. | ['Aixin Sun', 'Minlie Huang', 'Chujie Zheng'] | 2019-06-04 | chid-a-large-scale-chinese-idiom-dataset-for-1 | https://aclanthology.org/P19-1075 | https://aclanthology.org/P19-1075.pdf | acl-2019-7 | ['cloze-test'] | ['natural-language-processing'] | [-1.22327931e-01 -2.23077133e-01 -2.56020784e-01 -3.59328479e-01
-7.29597151e-01 -7.57898331e-01 4.80352730e-01 1.87891256e-02
-2.48620585e-01 6.17235065e-01 8.04131091e-01 -6.08592510e-01
1.27223909e-01 -6.10026240e-01 -2.35327870e-01 -1.17289253e-01
3.56316924e-01 6.51235282e-01 4.36304659e-01 -6.51595175e-01
6.14674807e-01 -2.76854008e-01 -1.02354324e+00 9.04289544e-01
1.46112287e+00 3.42906117e-01 3.72376472e-01 3.86856258e-01
-2.07610011e-01 1.17915452e+00 -1.02308834e+00 -4.71750706e-01
-8.75883847e-02 -1.36276627e+00 -1.20145357e+00 -2.97255665e-01
3.14348519e-01 -3.46977919e-01 1.49855629e-01 1.34377193e+00
4.87758994e-01 -2.45358050e-01 6.06129646e-01 -5.32979548e-01
-1.19046748e+00 9.34533775e-01 -5.99884391e-02 3.29759508e-01
1.03232360e+00 -2.91452199e-01 1.10713685e+00 -8.55549693e-01
9.00685132e-01 1.08197701e+00 2.03792378e-01 1.06930113e+00
-5.99101841e-01 -6.58877254e-01 2.75708616e-01 7.44066954e-01
-1.09747970e+00 -1.18204884e-01 9.53986704e-01 -1.73588917e-01
1.04583418e+00 4.25810575e-01 5.04380226e-01 9.74145591e-01
1.54818922e-01 7.46776342e-01 1.25967789e+00 -9.45566118e-01
1.81935966e-01 -3.13627385e-02 5.00134826e-01 2.77846336e-01
-3.09516545e-02 -3.30273062e-01 -4.41826284e-01 1.16000831e-01
3.39030951e-01 -3.95747751e-01 -6.47086143e-01 2.48908311e-01
-1.07896566e+00 1.02767754e+00 3.06970209e-01 5.62935174e-01
1.77993611e-01 -6.67758405e-01 4.69916403e-01 8.68192911e-01
5.59721515e-02 9.38226521e-01 -5.43130100e-01 -4.01230186e-01
-5.25496542e-01 5.34291387e-01 9.92172778e-01 1.14276397e+00
-1.13022894e-01 -3.83204788e-01 1.01366594e-01 8.64222050e-01
-2.34793425e-01 3.69694114e-01 8.77747059e-01 -7.07954824e-01
7.95655429e-01 8.30979288e-01 1.03110038e-01 -1.02300775e+00
-2.26617515e-01 -1.18521512e-01 -4.42884654e-01 -2.05465466e-01
6.95004344e-01 -6.56341016e-02 -3.22774261e-01 1.62451363e+00
-1.52368903e-01 -6.35125101e-01 2.80787647e-01 1.21053517e+00
9.14187849e-01 6.72729611e-01 -9.05422568e-02 -3.55582595e-01
1.51119661e+00 -1.24714875e+00 -9.74649847e-01 -5.34814119e-01
6.43420637e-01 -7.51356542e-01 1.84711397e+00 7.63752699e-01
-1.33212757e+00 -3.27985287e-01 -1.22029436e+00 -3.44845682e-01
-2.41804019e-01 3.39607745e-02 3.96009713e-01 5.28844476e-01
-5.63637793e-01 -1.89403608e-03 -3.86518836e-01 -1.91723719e-01
-1.16454557e-01 -1.99775696e-01 -1.10172808e-01 -4.76002216e-01
-1.46166229e+00 1.30405009e+00 6.64819539e-01 -4.48476858e-02
-5.88068843e-01 -1.85138345e-01 -7.12247849e-01 -2.83026583e-02
4.54635888e-01 -1.58041462e-01 1.35215247e+00 -1.10164177e+00
-1.41538489e+00 1.11863291e+00 -2.61299521e-01 -7.56559297e-02
5.53541481e-01 -5.49133778e-01 -6.87706649e-01 1.10746101e-01
2.43291110e-01 8.06130692e-02 3.36075932e-01 -7.92307496e-01
-4.87626582e-01 -1.08643860e-01 2.54625410e-01 3.58873755e-01
1.23194627e-01 7.39185870e-01 -3.28149438e-01 -7.29288697e-01
5.75757921e-01 -6.21115685e-01 2.57181495e-01 -7.14362144e-01
-4.38283645e-02 -4.09298033e-01 5.79072654e-01 -8.37857604e-01
1.85689104e+00 -2.01203203e+00 3.79156530e-01 -2.22149491e-01
1.14354879e-01 1.00779489e-01 -8.47664550e-02 7.64989793e-01
8.74691904e-02 4.05806124e-01 2.65796725e-02 3.71245593e-01
-1.15439124e-01 3.27028260e-02 -3.40848505e-01 9.72807780e-02
7.78212845e-02 1.03473914e+00 -1.03418267e+00 -1.96593300e-01
-3.56673419e-01 -3.92882407e-01 -6.95854545e-01 4.50549901e-01
-5.42288959e-01 4.89888787e-01 -4.12413031e-01 9.09756005e-01
7.09794581e-01 -1.75855368e-01 4.75809067e-01 5.46215415e-01
-1.21568806e-01 1.37179971e+00 -6.09149635e-01 1.53415847e+00
-3.23420286e-01 4.73610759e-01 -4.59760487e-01 -4.64408427e-01
8.02033842e-01 2.26155311e-01 -6.10528469e-01 -1.00950813e+00
2.16002375e-01 5.35239637e-01 6.87044024e-01 -8.17492545e-01
2.66572803e-01 -2.49263197e-01 -4.56645697e-01 5.69516063e-01
-5.33130884e-01 -2.16968939e-01 4.69924897e-01 1.66484803e-01
1.00435543e+00 -3.65673631e-01 8.04983974e-01 -6.36930227e-01
8.37330461e-01 4.32551116e-01 5.77808201e-01 7.46490896e-01
-3.48069444e-02 6.56274796e-01 7.62024820e-01 -2.03853160e-01
-7.13990331e-01 -1.14006209e+00 -2.29961962e-01 1.14348185e+00
2.71515280e-01 -7.69196451e-01 -1.15920150e+00 -8.85326445e-01
-6.72882915e-01 1.00872731e+00 -4.43521321e-01 4.41323221e-02
-1.27248538e+00 -4.84883636e-01 6.13760054e-01 7.26183772e-01
7.30338395e-01 -1.51932466e+00 -9.62622762e-01 2.92472333e-01
-5.55096030e-01 -9.96870697e-01 -5.64349592e-01 -4.69972845e-03
-6.97017133e-01 -1.33117843e+00 -1.76468402e-01 -1.54394174e+00
7.09782183e-01 1.50126651e-01 1.33011115e+00 6.85409307e-01
5.81333280e-01 -5.87997973e-01 -1.21637487e+00 -3.45197082e-01
-6.48015022e-01 9.87978056e-02 -4.14014935e-01 -9.54218745e-01
1.03165734e+00 -7.64760748e-02 -1.31135538e-01 2.85324991e-01
-6.25100911e-01 3.43634963e-01 1.64266467e-01 1.25387037e+00
-8.99409782e-03 -6.75804019e-02 6.45732284e-01 -1.21578503e+00
1.24458814e+00 -3.86246771e-01 -4.33783799e-01 4.21304435e-01
-3.10679555e-01 -2.24290103e-01 9.85005379e-01 -6.79121017e-01
-1.07915950e+00 -7.93524086e-01 -1.69930935e-01 5.98594606e-01
-1.41389802e-01 9.70512211e-01 -6.60884142e-01 4.81204957e-01
7.19801843e-01 1.46738201e-01 -2.67320484e-01 -6.19221687e-01
2.81638920e-01 8.35256279e-01 6.97740436e-01 -8.48814368e-01
3.58581871e-01 -2.08058909e-01 -7.99829543e-01 -5.29801011e-01
-1.03824127e+00 -1.65010661e-01 -6.30489349e-01 2.42352352e-01
7.63853371e-01 -8.71373415e-01 -2.81424373e-01 5.35178244e-01
-1.56297815e+00 -3.90466928e-01 3.77497852e-01 3.79652321e-01
-2.27551207e-01 2.11576432e-01 -1.22853565e+00 -3.34129870e-01
-2.07237378e-01 -1.13259995e+00 4.79425550e-01 6.55488074e-02
-7.77891278e-01 -8.22542608e-01 5.53821120e-03 3.87002945e-01
3.04815322e-01 -1.26228750e-01 1.69620824e+00 -9.55759585e-01
-4.94603246e-01 -1.41712110e-02 -1.83530107e-01 2.63919622e-01
8.29299912e-02 -4.70316529e-01 -5.81516922e-01 -5.36107607e-02
7.61722386e-01 -4.95571792e-01 3.41505915e-01 -9.61191803e-02
7.70452201e-01 -6.36395872e-01 1.94147587e-01 3.93199384e-01
1.20960057e+00 6.69119418e-01 7.99054742e-01 5.35662115e-01
3.82724196e-01 7.03971922e-01 7.00293183e-01 -1.98952723e-02
5.18205345e-01 3.95885319e-01 -1.08651169e-01 5.38091242e-01
1.36405565e-02 -5.39493263e-01 4.33900297e-01 1.62211776e+00
2.42709354e-01 -4.38591808e-01 -1.10949457e+00 6.20622873e-01
-1.64954543e+00 -8.25561643e-01 -5.07489145e-01 1.70905268e+00
1.30136669e+00 2.65714258e-01 -1.93207696e-01 3.13178301e-01
3.46588463e-01 4.06144671e-02 9.40048099e-02 -6.22638583e-01
-4.28804249e-01 3.10819656e-01 -2.86280125e-01 9.18913364e-01
-6.87174320e-01 1.40034366e+00 6.59935141e+00 5.25754213e-01
-1.08427441e+00 2.49488264e-01 1.85770616e-01 3.97453278e-01
-2.83483386e-01 1.29133329e-01 -6.22131586e-01 5.74943125e-01
5.21256089e-01 -2.18973204e-01 5.46455085e-01 7.54781067e-01
8.99539068e-02 -3.51952225e-01 -1.16221726e+00 5.73048472e-01
7.06764400e-01 -1.00750649e+00 6.79685734e-03 -4.89698559e-01
8.77076209e-01 -3.20679456e-01 -3.11742544e-01 2.20415354e-01
-1.32403094e-02 -1.20016611e+00 7.53933668e-01 -7.64324665e-02
6.24267161e-01 -4.90435839e-01 8.76539111e-01 7.48310089e-01
-6.26767039e-01 -3.52157615e-02 -5.08586049e-01 -9.50057745e-01
1.69056997e-01 -8.37633088e-02 -4.71489668e-01 2.37973407e-02
4.67261255e-01 4.48151469e-01 -9.90051806e-01 7.27067053e-01
-1.18220603e+00 9.03509855e-01 7.88906440e-02 -6.74060047e-01
2.54747540e-01 -1.71589583e-01 2.97990233e-01 1.19029844e+00
1.29389748e-01 4.84459192e-01 1.39581114e-01 8.75880718e-01
5.33894747e-02 4.64639336e-01 -3.25719416e-01 2.62061626e-01
7.02295482e-01 4.33040589e-01 -5.15733480e-01 -3.46490026e-01
-4.76868629e-01 1.09785450e+00 4.98653829e-01 1.99664116e-01
-5.61889112e-01 -2.76486725e-01 2.59964108e-01 1.34019420e-01
8.55199471e-02 -1.99394673e-01 -7.91238129e-01 -1.47653687e+00
3.45044523e-01 -1.61262298e+00 4.92032081e-01 -7.65971541e-01
-1.57256806e+00 7.44290411e-01 1.03227317e-01 -1.23539019e+00
-3.80602926e-01 -8.04695308e-01 -1.09902549e+00 8.85105073e-01
-1.16775262e+00 -8.53762269e-01 -3.13364089e-01 2.69634932e-01
1.01540506e+00 -1.04773246e-01 9.42098320e-01 1.62777320e-01
-2.84715503e-01 6.25854731e-01 -1.90396652e-01 5.00792265e-01
6.80407166e-01 -1.38616586e+00 4.97543335e-01 1.13099182e+00
-1.41391102e-02 1.10036147e+00 7.25157022e-01 -8.94497395e-01
-8.35262179e-01 -3.24206769e-01 1.72368777e+00 -8.93933713e-01
6.20244145e-01 -5.11849821e-01 -1.29181051e+00 5.71924984e-01
5.96113861e-01 -6.90942585e-01 5.81670642e-01 1.26552165e-01
-4.21722114e-01 3.16031456e-01 -9.00387645e-01 8.37931633e-01
1.09816301e+00 -6.37466848e-01 -1.45055950e+00 1.05498806e-01
6.76492751e-01 -7.22164452e-01 -2.47611150e-01 1.85427472e-01
3.60469967e-01 -1.01093507e+00 3.80039543e-01 -9.92753685e-01
9.38419282e-01 -4.18505073e-01 -1.38052791e-01 -1.47094727e+00
-2.42600054e-01 -2.79927820e-01 8.59203786e-02 1.04301345e+00
5.34958124e-01 -2.90165603e-01 4.62719768e-01 5.94329059e-01
-1.17188141e-01 -4.14743096e-01 -7.12838590e-01 -6.25483990e-01
6.84072793e-01 -2.99725294e-01 7.38046646e-01 1.01278508e+00
1.09911013e+00 8.33650708e-01 -1.91830754e-01 -1.91975877e-01
-1.53318360e-01 1.51058063e-01 5.58440566e-01 -8.47717345e-01
-3.40618283e-01 -4.83535022e-01 5.19400313e-02 -1.53112841e+00
3.09583664e-01 -6.39664888e-01 5.00959083e-02 -1.10522306e+00
1.72618419e-01 -1.80939317e-01 2.16773346e-01 1.74445197e-01
-6.03847027e-01 -1.00929074e-01 4.09023643e-01 1.98473796e-01
-5.64344943e-01 2.90417939e-01 1.40035880e+00 1.78525858e-02
-4.68580753e-01 -1.40183732e-01 -8.68268609e-01 8.22425663e-01
9.68771458e-01 -4.33998734e-01 -4.65700865e-01 -7.90440261e-01
3.03292543e-01 2.31501833e-01 -1.16832875e-01 -6.63134873e-01
3.58070910e-01 -3.89607280e-01 1.36858940e-01 -5.04298568e-01
-3.23476285e-01 -4.58689779e-01 -3.07757676e-01 4.74125266e-01
-5.81538439e-01 1.00926971e+00 -1.56787843e-01 -1.03436239e-01
-5.38010001e-01 -8.13400507e-01 4.91504759e-01 -5.06310761e-01
-7.14832008e-01 -6.93150342e-01 -4.76215094e-01 8.09338272e-01
6.72076106e-01 -9.30881873e-03 -6.39571071e-01 -5.17868578e-01
-1.03258215e-01 1.59043580e-01 7.20689893e-01 6.33713245e-01
7.46430397e-01 -1.12911987e+00 -8.46952677e-01 3.57084602e-01
3.52177143e-01 -1.69232069e-03 -2.10732013e-01 5.61511993e-01
-1.08890545e+00 4.36038107e-01 -2.85071433e-01 2.22548451e-02
-1.11283362e+00 5.37869811e-01 2.23309651e-01 -1.87075391e-01
-6.29862666e-01 5.66142738e-01 1.15637437e-01 -3.55893284e-01
1.42124742e-01 -4.48272377e-01 -3.86730462e-01 -1.75852567e-01
1.05521643e+00 2.36968502e-01 -1.42035345e-02 -4.72635567e-01
-2.95987785e-01 3.63397419e-01 -4.17393088e-01 -5.29388078e-02
7.44656444e-01 -2.23373979e-01 -6.04584634e-01 4.59688246e-01
7.95072794e-01 6.80511594e-01 -4.83007580e-01 6.20709881e-02
4.90493536e-01 -5.79187572e-01 -6.22004867e-01 -1.39852500e+00
-3.48560303e-01 7.88767278e-01 -1.95320725e-01 6.49905205e-02
1.16525996e+00 1.04971752e-01 8.42315197e-01 3.97504568e-01
4.49419349e-01 -1.07857335e+00 7.03513920e-02 1.19055426e+00
1.01218855e+00 -1.16494799e+00 -2.63923645e-01 -6.22153103e-01
-1.00662482e+00 1.14633131e+00 1.22515702e+00 -3.03708851e-01
2.87801355e-01 2.29079574e-01 8.37987006e-01 -2.44707853e-01
-8.93247306e-01 1.21275567e-01 3.31072718e-01 5.09621322e-01
8.60719979e-01 1.00671202e-01 -1.36259997e+00 1.24178696e+00
-8.72989655e-01 -3.85114402e-01 7.98415661e-01 8.19368839e-01
-3.82343262e-01 -1.30704796e+00 -2.58980125e-01 1.71114519e-01
-5.12539923e-01 -4.57160771e-01 -7.94972360e-01 8.93260241e-01
1.39115632e-01 1.18688262e+00 -5.22250235e-02 -2.41484687e-01
3.26223522e-01 1.10089593e-01 4.88302499e-01 -9.38022852e-01
-8.99410546e-01 3.75388339e-02 3.19862425e-01 -1.10270962e-01
-1.03863157e-01 -5.46216846e-01 -1.28679705e+00 -2.97334373e-01
-4.15480912e-01 3.92479002e-01 -7.10215569e-02 1.26062262e+00
-2.97972500e-01 -3.31284781e-03 2.12742001e-01 7.11346790e-02
-6.47064090e-01 -1.35176802e+00 -3.38127583e-01 6.04088664e-01
9.08433497e-02 -2.63893932e-01 -2.82688558e-01 -1.76899970e-01] | [10.915220260620117, 9.016057014465332] |
08a852a9-5bad-4456-b99b-8a6d35cf48df | collection-space-navigator-an-interactive | 2305.06809 | null | https://arxiv.org/abs/2305.06809v1 | https://arxiv.org/pdf/2305.06809v1.pdf | Collection Space Navigator: An Interactive Visualization Interface for Multidimensional Datasets | We introduce the Collection Space Navigator (CSN), a browser-based visualization tool to explore, research, and curate large collections of visual digital artifacts that are associated with multidimensional data, such as vector embeddings or tables of metadata. Media objects such as images are often encoded as numerical vectors, for e.g. based on metadata or using machine learning to embed image information. Yet, while such procedures are widespread for a range of applications, it remains a challenge to explore, analyze, and understand the resulting multidimensional spaces in a more comprehensive manner. Dimensionality reduction techniques such as t-SNE or UMAP often serve to project high-dimensional data into low dimensional visualizations, yet require interpretation themselves as the remaining dimensions are typically abstract. Here, the Collection Space Navigator provides a customizable interface that combines two-dimensional projections with a set of configurable multidimensional filters. As a result, the user is able to view and investigate collections, by zooming and scaling, by transforming between projections, by filtering dimensions via range sliders, and advanced text filters. Insights that are gained during the interaction can be fed back into the original data via ad hoc exports of filtered metadata and projections. This paper comes with a functional showcase demo using a large digitized collection of classical Western art. The Collection Space Navigator is open source. Users can reconfigure the interface to fit their own data and research needs, including projections and filter controls. The CSN is ready to serve a broad community. | ['Maximilian Schich', 'Andres Karjus', 'Mar Canet Solà', 'Tillmann Ohm'] | 2023-05-11 | null | null | null | null | ['dimensionality-reduction', 'data-visualization', 'data-visualization', 'embeddings-evaluation'] | ['methodology', 'methodology', 'miscellaneous', 'natural-language-processing'] | [-1.22345299e-01 -4.35404390e-01 1.07664630e-01 2.99703982e-02
-1.97873071e-01 -1.23256123e+00 7.70542324e-01 5.06606877e-01
-2.77230740e-01 1.60739556e-01 4.80297357e-01 -7.73048580e-01
-4.02601808e-01 -8.60161841e-01 -6.36009127e-02 -3.86878908e-01
-3.46260190e-01 2.55147010e-01 2.43296787e-01 -1.48517594e-01
5.63483894e-01 8.69524121e-01 -1.95209658e+00 1.10470921e-01
7.10297644e-01 7.06577957e-01 4.31907773e-01 7.80741990e-01
-6.76710308e-01 -4.43959326e-01 -5.37525594e-01 -3.50372940e-02
2.86805004e-01 -1.19777163e-02 -2.24926308e-01 -1.97947413e-01
5.63250244e-01 -2.20531702e-01 3.30819666e-01 9.40978408e-01
3.86672258e-01 -1.04493998e-01 3.99146378e-01 -1.22637963e+00
-6.73651695e-01 1.64363876e-01 -4.77036685e-01 2.28012383e-01
7.56532371e-01 4.55646843e-01 5.66123366e-01 -1.13585043e+00
1.10586703e+00 1.51579654e+00 4.28333372e-01 -2.15626098e-02
-1.66560638e+00 -4.48214769e-01 1.80179831e-02 -1.53932467e-01
-1.03344989e+00 -4.83570397e-02 7.31421292e-01 -9.30423915e-01
6.69936895e-01 1.04341531e+00 1.25043333e+00 7.06708968e-01
-1.18748583e-02 2.24831060e-01 1.06815171e+00 -4.58771914e-01
4.21979547e-01 3.90809059e-01 2.06190541e-01 3.70659500e-01
3.21365774e-01 -4.37985688e-01 -4.96896923e-01 -4.31529850e-01
9.32791412e-01 3.09430152e-01 -3.26981097e-01 -9.49225545e-01
-1.47107375e+00 4.80863988e-01 3.56144845e-01 2.29013264e-01
-2.14321792e-01 -2.72971451e-01 3.48135144e-01 1.95157200e-01
5.14373779e-01 6.98828399e-01 -1.52408510e-01 -4.64827031e-01
-8.08569968e-01 4.32465613e-01 5.58482707e-01 7.63537586e-01
8.67252111e-01 -4.46280062e-01 2.10373476e-01 6.28615975e-01
4.58863705e-01 3.53866786e-01 3.27329695e-01 -9.07028854e-01
2.67853677e-01 1.17615032e+00 1.68991894e-01 -1.08369255e+00
-6.04641855e-01 7.07134232e-02 -6.58698916e-01 1.07061195e+00
4.77941900e-01 2.58183450e-01 -7.97084689e-01 1.17877150e+00
8.00759256e-01 -6.19433999e-01 -4.88527685e-01 1.00232494e+00
6.18837237e-01 3.87557864e-01 2.65835017e-01 7.72433504e-02
1.63306689e+00 -8.57363939e-02 -8.83158922e-01 9.94439051e-02
8.00788343e-01 -6.72585785e-01 1.90022337e+00 3.90687734e-01
-1.12779427e+00 -1.25405103e-01 -1.23393929e+00 -5.09105921e-01
-1.28581023e+00 -1.84891403e-01 4.10110056e-01 5.12313247e-01
-1.08829904e+00 6.81041062e-01 -1.07904506e+00 -9.00365531e-01
4.68978018e-01 -1.08306237e-01 -4.58149195e-01 1.74932092e-01
-7.50864565e-01 8.37263465e-01 1.05126500e-01 -2.34938025e-01
6.81864843e-02 -1.34464657e+00 -8.69138598e-01 7.03722313e-02
-1.84687674e-01 -6.53079331e-01 6.34009004e-01 -2.13961545e-02
-9.14446533e-01 8.34059596e-01 1.54273286e-01 2.43722290e-01
6.97240889e-01 -3.23759347e-01 -3.79241794e-01 5.62064797e-02
1.51880130e-01 4.06608492e-01 3.02828521e-01 -1.25524032e+00
-5.02057791e-01 -7.44489729e-01 -6.74580634e-02 2.10003138e-01
-8.45485091e-01 3.49513143e-02 -4.46333051e-01 -6.35009468e-01
4.00406331e-01 -4.94364858e-01 -9.21919271e-02 9.48781073e-01
-4.33695972e-01 2.54235029e-01 1.62032068e+00 -7.72816718e-01
1.73389137e+00 -2.30359650e+00 2.59193629e-01 4.54570532e-01
4.50045586e-01 -5.42878024e-02 3.40016961e-01 7.19594777e-01
-1.03766017e-01 6.55165553e-01 -1.91168845e-01 -3.85071188e-01
2.43326813e-01 -2.43186504e-01 -2.32117977e-02 4.59165275e-01
-1.68303952e-01 3.25107694e-01 -8.71416152e-01 -3.82806957e-01
5.88909030e-01 7.16479182e-01 -3.85050327e-01 -1.29304141e-01
-1.49526462e-01 3.29609603e-01 -2.16917396e-01 4.73119825e-01
9.79501784e-01 -2.70962119e-01 3.00144941e-01 -1.90999746e-01
-8.67067456e-01 3.28821808e-01 -1.58123541e+00 1.89191329e+00
-3.08645397e-01 9.97617245e-01 3.44600886e-01 -2.04090342e-01
1.03105617e+00 -2.03944027e-01 2.27129772e-01 -5.19829392e-01
-4.21903461e-01 5.94409630e-02 -5.66798389e-01 -7.15273023e-01
8.60634744e-01 2.39848778e-01 6.71747141e-03 9.17114437e-01
-4.80979830e-01 -6.76597506e-02 4.29042965e-01 4.00908500e-01
8.62040758e-01 3.94925416e-01 1.83090433e-01 -4.78950590e-01
5.96970357e-02 1.17292926e-01 -1.11203730e-01 1.78812146e-01
4.31857944e-01 5.74406266e-01 7.41228580e-01 -6.15636230e-01
-1.42937136e+00 -1.25444937e+00 -6.30568922e-01 1.01289964e+00
-3.44654135e-02 -8.56538951e-01 -3.17722172e-01 -1.30749881e-01
3.56629312e-01 7.48487055e-01 -6.64971828e-01 2.62136102e-01
-3.19837004e-01 -2.66348302e-01 -5.12003042e-02 1.75408885e-01
1.26702385e-02 -7.30029285e-01 -1.22867095e+00 -2.33914211e-01
4.73149240e-01 -1.65224150e-01 -6.78038523e-02 2.67649796e-02
-1.10957372e+00 -8.38782847e-01 -6.61613464e-01 -2.88497746e-01
7.21415043e-01 2.90390313e-01 8.91014397e-01 7.05446973e-02
-7.08238482e-01 4.50080901e-01 -1.81402057e-01 -2.08869502e-01
-1.96920738e-01 2.31470447e-02 5.69235384e-02 -5.18999457e-01
2.23383516e-01 -1.09646475e+00 -8.07772398e-01 2.81015038e-01
-1.19156170e+00 3.96297544e-01 -5.19126803e-02 1.89169824e-01
6.23768568e-01 -3.74731451e-01 1.18590705e-01 -6.87366962e-01
9.59580600e-01 -6.54211760e-01 -6.86456800e-01 1.55530691e-01
-6.57636583e-01 8.97859037e-02 4.78701681e-01 -4.05784041e-01
-6.50969148e-01 -2.31311187e-01 5.61592162e-01 -4.80578125e-01
-2.88975120e-01 5.55247247e-01 -4.19718653e-01 2.51893699e-01
7.78839111e-01 -2.18730345e-01 2.14664117e-01 -1.13645470e+00
8.21402669e-01 8.81023884e-01 3.43336523e-01 -2.00931981e-01
8.57551754e-01 5.26692748e-01 -1.06837891e-01 -7.31104791e-01
2.31484428e-01 -2.36223489e-01 -9.82384026e-01 -4.23021525e-01
8.12923670e-01 -2.92987019e-01 -7.45086193e-01 -6.42163828e-02
-5.84110558e-01 -3.37221503e-01 -4.01646137e-01 3.94824266e-01
-2.34853432e-01 1.86982706e-01 -2.22669076e-02 -5.42753637e-01
-1.37081929e-02 -9.81275916e-01 8.77172470e-01 3.87398183e-01
-7.31582403e-01 -1.07352817e+00 2.01721147e-01 -3.16160530e-01
5.60527742e-01 7.55599022e-01 1.33499670e+00 -3.71839911e-01
-6.25283182e-01 -2.45668337e-01 -1.32046327e-01 -5.03165603e-01
3.82603079e-01 4.58186209e-01 -7.73289263e-01 -1.35083303e-01
-5.52951872e-01 2.13831693e-01 1.84425727e-01 -3.63372602e-02
1.23821044e+00 -4.03653204e-01 -6.84663892e-01 8.27305138e-01
1.32040060e+00 1.09288372e-01 6.28666759e-01 9.86810029e-01
5.92257917e-01 6.92218900e-01 3.24227124e-01 6.84875488e-01
2.33837336e-01 5.78632832e-01 5.82679510e-01 -2.09105179e-01
1.99685320e-01 -2.70667702e-01 -4.13946837e-01 2.91561276e-01
2.54348200e-02 5.59794195e-02 -1.13377130e+00 2.20983356e-01
-1.56196892e+00 -8.54150295e-01 -2.99467802e-01 2.53974295e+00
4.63496864e-01 2.25474797e-02 3.01338285e-01 1.53203577e-01
6.04065418e-01 3.05962622e-01 -6.33745611e-01 -5.14128327e-01
1.62893701e-02 -2.40536794e-01 3.15267056e-01 2.97558963e-01
-7.99743950e-01 3.47728312e-01 6.12181902e+00 2.80578047e-01
-1.28668797e+00 -3.16865116e-01 1.00642264e-01 -6.75359130e-01
-9.00365233e-01 1.22034803e-01 -3.44376147e-01 4.13264781e-01
6.14760160e-01 -5.18270433e-01 4.87816274e-01 8.84092152e-01
6.90226078e-01 -2.95588881e-01 -1.18261313e+00 1.14180362e+00
-4.22483653e-01 -1.69181085e+00 9.95252654e-02 4.31064188e-01
-1.08343409e-02 -5.11811793e-01 2.04575464e-01 -2.03754634e-01
1.68832719e-01 -9.33428168e-01 7.71443129e-01 7.05304861e-01
1.19963658e+00 -4.88509387e-01 -3.64533633e-01 -4.85119335e-02
-1.11750591e+00 -3.33975740e-02 -9.38568786e-02 -1.45157278e-01
4.39162910e-01 4.68167901e-01 -5.69046080e-01 2.36805752e-01
1.12619019e+00 6.72619879e-01 -9.74614978e-01 1.39458525e+00
3.90160918e-01 -3.92700881e-02 -5.74003518e-01 -1.42627329e-01
-2.85018504e-01 -6.67822480e-01 8.31057489e-01 1.48888862e+00
6.90849543e-01 -9.65864435e-02 -3.15972447e-01 1.02371669e+00
3.17726254e-01 1.79428205e-01 -7.70197392e-01 -1.12770341e-01
8.12732339e-01 1.63459778e+00 -8.90882552e-01 -1.76004991e-01
-1.41401172e-01 5.50255060e-01 1.18806548e-01 5.14391065e-01
-2.41144106e-01 -6.71508491e-01 1.12523210e+00 7.95159459e-01
8.70698467e-02 -7.17863381e-01 -7.11715162e-01 -7.60456264e-01
1.94543257e-01 -6.78660631e-01 2.78355092e-01 -1.25257123e+00
-8.04078460e-01 2.20304951e-01 4.14110720e-01 -1.48323333e+00
2.24268623e-02 -5.40217578e-01 -6.17154479e-01 1.10785758e+00
-6.03492618e-01 -9.64542449e-01 -6.12562656e-01 3.10663819e-01
6.94413483e-03 2.32708693e-01 9.66529131e-01 4.14074808e-01
-3.76258135e-01 -8.87151510e-02 5.31986892e-01 -2.22367987e-01
7.38845766e-01 -1.52275455e+00 5.50372362e-01 3.59837532e-01
-1.11241512e-01 9.91790593e-01 9.98111606e-01 -8.46356094e-01
-1.63016689e+00 -4.23410624e-01 2.77323753e-01 -5.21023929e-01
8.31690133e-01 -8.66646886e-01 -1.28114998e+00 5.82915246e-01
1.30116522e-01 -2.60985382e-02 9.59037781e-01 2.19072938e-01
-3.04044664e-01 1.27516285e-01 -1.12423706e+00 1.05190325e+00
1.02418625e+00 -4.09567565e-01 -2.56062418e-01 3.11465830e-01
4.90935564e-01 -3.47529918e-01 -1.17643297e+00 -2.31473431e-01
8.94108295e-01 -1.02457213e+00 9.94077563e-01 -6.01920128e-01
1.49029240e-01 -5.72085619e-01 -3.89035977e-02 -1.33133709e+00
-2.70699859e-01 -6.89363003e-01 3.00647803e-02 1.37313318e+00
4.48578387e-01 -4.81136173e-01 5.29301405e-01 1.04626787e+00
5.61636388e-02 -4.62700516e-01 -7.65180051e-01 -3.83647144e-01
-1.31184667e-01 -4.41189528e-01 9.58794475e-01 1.08683860e+00
6.30629957e-01 -2.12695286e-01 2.01438278e-01 -4.53350395e-02
5.95366836e-01 -3.06887887e-02 9.81900871e-01 -1.65959048e+00
2.99380749e-01 -7.36398518e-01 -4.68038380e-01 -6.02440834e-01
-8.01620364e-01 -7.37346113e-01 -7.32483208e-01 -1.90725243e+00
-2.46408731e-01 -7.63568521e-01 2.93787330e-01 2.45493934e-01
1.27520204e-01 1.69814408e-01 4.70157415e-01 5.07955074e-01
-5.62881641e-02 1.42444029e-01 1.12046206e+00 1.78737506e-01
-8.03198457e-01 -5.95001221e-01 -5.64432919e-01 6.56921089e-01
6.65601373e-01 -2.22590536e-01 -2.54845649e-01 -3.72142822e-01
5.70736527e-01 -2.06732824e-01 4.88592774e-01 -9.64568198e-01
1.16097257e-01 -1.57644898e-01 8.07917535e-01 -9.02113616e-01
3.44927222e-01 -7.73701489e-01 7.27525890e-01 1.54232144e-01
-2.81337053e-01 4.86162126e-01 4.60381478e-01 4.51489300e-01
4.52656835e-01 2.94301003e-01 3.44402581e-01 -6.49691820e-02
-4.25456524e-01 -1.69432640e-01 -4.14339066e-01 -3.82806510e-01
9.12459373e-01 -7.13514984e-01 -7.70420790e-01 -2.54557133e-01
-1.10100615e+00 2.72279024e-01 1.20986176e+00 5.43073475e-01
4.41093177e-01 -1.54941750e+00 -4.08026539e-02 3.93011361e-01
1.48900449e-01 2.21886933e-01 2.13539302e-01 5.08576870e-01
-7.27231085e-01 -3.60026024e-02 -6.76497221e-01 -7.00227737e-01
-1.28903830e+00 7.42168069e-01 -3.21674198e-02 1.26514047e-01
-1.17972410e+00 1.88603416e-01 1.85326003e-02 -3.51592720e-01
1.05899774e-01 -4.09301519e-01 -3.79117399e-01 6.51639521e-01
9.77317154e-01 6.41359508e-01 -4.57024202e-02 -1.91808820e-01
-4.65681434e-01 6.17811859e-01 3.17016810e-01 -4.15592968e-01
1.63054740e+00 -4.62112188e-01 -9.17543937e-03 8.99780095e-01
1.15940547e+00 9.03617218e-02 -1.40985811e+00 1.68269962e-01
-1.49916746e-02 -9.70094204e-01 -3.87867719e-01 -7.95334935e-01
-5.90620995e-01 9.51019585e-01 9.78761256e-01 8.38878155e-01
9.89884377e-01 1.13706373e-01 6.80832705e-03 -5.41997626e-02
-5.81787294e-03 -7.75078773e-01 -1.09725155e-01 -1.61761731e-01
1.18833816e+00 -7.11314559e-01 4.66762066e-01 -8.20141509e-02
-2.74650365e-01 1.55481756e+00 3.45397949e-01 2.33463868e-01
7.49017060e-01 6.52558386e-01 2.15586782e-01 -5.23971319e-01
-4.33254749e-01 1.47776425e-01 -5.39664216e-02 7.67902792e-01
4.38864172e-01 -3.68743539e-02 -3.73341143e-01 7.55522773e-02
-6.18241668e-01 -1.61847115e-01 4.80192482e-01 1.20376909e+00
-5.52036464e-01 -1.06774306e+00 -9.79528725e-01 6.27695739e-01
-2.31259353e-02 1.36924490e-01 -3.10038418e-01 9.05554116e-01
-1.37359351e-01 3.06756854e-01 7.51910090e-01 -1.75834656e-01
3.05371702e-01 1.49921700e-01 -1.00867376e-01 -3.59016329e-01
-4.52525765e-01 1.18209176e-01 -4.93176952e-02 -5.79192698e-01
4.73577231e-02 -8.35247099e-01 -1.08698761e+00 -4.12772804e-01
1.91919595e-01 -1.73275486e-01 1.29061341e+00 2.69591272e-01
7.44806111e-01 3.01994324e-01 -4.38114405e-02 -1.29679751e+00
1.14520028e-01 -9.18526590e-01 -6.55565858e-01 5.98019958e-01
4.26595420e-01 -7.10214376e-01 -4.40567374e-01 2.22947061e-01] | [7.997768878936768, 4.593198299407959] |
57ca94b0-ae31-440d-a6f5-36d39b0dd408 | riddle-reversible-and-diversified-de | 2303.05171 | null | https://arxiv.org/abs/2303.05171v3 | https://arxiv.org/pdf/2303.05171v3.pdf | RiDDLE: Reversible and Diversified De-identification with Latent Encryptor | This work presents RiDDLE, short for Reversible and Diversified De-identification with Latent Encryptor, to protect the identity information of people from being misused. Built upon a pre-learned StyleGAN2 generator, RiDDLE manages to encrypt and decrypt the facial identity within the latent space. The design of RiDDLE has three appealing properties. First, the encryption process is cipher-guided and hence allows diverse anonymization using different passwords. Second, the true identity can only be decrypted with the correct password, otherwise the system will produce another de-identified face to maintain the privacy. Third, both encryption and decryption share an efficient implementation, benefiting from a carefully tailored lightweight encryptor. Comparisons with existing alternatives confirm that our approach accomplishes the de-identification task with better quality, higher diversity, and stronger reversibility. We further demonstrate the effectiveness of RiDDLE in anonymizing videos. Code and models will be made publicly available. | ['Tieniu Tan', 'Jing Dong', 'Kang Zhao', 'Wei Wang', 'Dongze Li'] | 2023-03-09 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Li_RiDDLE_Reversible_and_Diversified_De-Identification_With_Latent_Encryptor_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Li_RiDDLE_Reversible_and_Diversified_De-Identification_With_Latent_Encryptor_CVPR_2023_paper.pdf | cvpr-2023-1 | ['de-identification'] | ['natural-language-processing'] | [ 1.99484900e-01 1.00849129e-01 -1.88030049e-01 2.61495709e-02
-2.68033653e-01 -1.17971587e+00 5.68202019e-01 -5.47563493e-01
-2.59218037e-01 7.43030488e-01 3.59841466e-01 -1.73326299e-01
2.50054210e-01 -8.19417179e-01 -3.17699552e-01 -9.07808363e-01
-1.53754532e-01 -1.11640878e-02 -4.77862686e-01 2.23957729e-02
-1.09656654e-01 6.72913492e-01 -1.35712528e+00 1.72025442e-01
6.23282254e-01 7.91878462e-01 -2.88036317e-01 6.95693314e-01
3.88737679e-01 5.15856266e-01 -5.84384859e-01 -1.11581314e+00
8.56597126e-01 -5.12201071e-01 -7.44085252e-01 -4.31982040e-01
4.36431080e-01 -7.81544030e-01 -5.09252071e-01 8.50237072e-01
7.33014286e-01 -5.86689472e-01 4.17455345e-01 -1.65279257e+00
-6.66230023e-01 4.15254682e-01 -3.11768621e-01 -4.06961828e-01
5.84407866e-01 4.85007912e-01 6.62468970e-01 -5.96233189e-01
6.50831044e-01 1.15510905e+00 9.10476506e-01 1.01536822e+00
-1.35419881e+00 -1.26383483e+00 -5.99527538e-01 -8.90484452e-02
-1.85298288e+00 -7.96468079e-01 3.96349907e-01 -1.92549393e-01
1.69874832e-01 6.31719172e-01 7.59456813e-01 1.38629377e+00
1.07960403e-01 3.89624983e-01 1.25181186e+00 -1.00020088e-01
-7.84261152e-02 5.31712353e-01 -3.82087976e-01 5.66920996e-01
5.16029894e-01 2.56614983e-01 -4.71643955e-01 -4.97277409e-01
5.50932169e-01 -3.47650684e-02 -6.04450762e-01 -3.60310644e-01
-1.04698801e+00 6.94375038e-01 -1.61971226e-01 1.29029220e-02
-3.49788904e-01 2.81476527e-01 4.63159084e-01 4.69027191e-01
2.49476314e-01 3.93976927e-01 1.05933726e-01 -3.42423528e-01
-9.19223726e-01 1.37623549e-01 1.04438639e+00 9.84869301e-01
8.00779641e-01 -4.67170291e-02 -1.23199493e-01 3.08705717e-01
7.31773898e-02 9.20069158e-01 4.79567081e-01 -9.32762444e-01
3.09414178e-01 2.19554305e-01 -1.30824009e-02 -1.23621976e+00
2.78142035e-01 5.36121726e-02 -1.15303767e+00 2.93074757e-01
3.05313259e-01 -4.53425139e-01 -5.11281073e-01 1.95778406e+00
1.42714322e-01 1.46853879e-01 2.86005735e-01 3.45962256e-01
5.63339174e-01 4.65278864e-01 1.85812321e-02 1.01326756e-01
1.52088833e+00 -4.44658756e-01 -7.44079411e-01 3.00180048e-01
3.58157635e-01 -6.47269070e-01 8.34035158e-01 4.33068424e-02
-9.86521721e-01 -3.51980954e-01 -1.15177834e+00 2.72646844e-02
-2.45432779e-01 1.65210500e-01 5.36758244e-01 1.75867808e+00
-1.35863221e+00 4.96360779e-01 -5.02437651e-01 -8.66566896e-02
6.68919802e-01 8.69074166e-01 -1.02186775e+00 8.70146528e-02
-1.35012972e+00 4.63636398e-01 4.37863201e-01 -1.73490837e-01
-7.67360449e-01 -8.54640484e-01 -9.01848137e-01 8.30507800e-02
-1.28066942e-01 -8.40597332e-01 8.47716272e-01 -1.02070391e+00
-1.82394552e+00 1.17017305e+00 -1.13762818e-01 -6.81410134e-01
1.03537166e+00 -1.34062683e-02 -4.56949800e-01 2.76566684e-01
1.36400843e-02 7.71497965e-01 1.26805556e+00 -1.05557287e+00
-3.35489482e-01 -2.49947146e-01 -1.76273286e-04 1.79708861e-02
-7.81718314e-01 -8.33620057e-02 -2.92236209e-01 -7.32604802e-01
-5.52432060e-01 -1.17452419e+00 2.13419601e-01 -5.19057401e-02
-5.78771651e-01 4.25803185e-01 1.31457055e+00 -7.40728855e-01
1.30675519e+00 -2.47277713e+00 -1.56557336e-01 4.05927032e-01
5.88416040e-01 6.72922790e-01 4.33802307e-02 6.16361797e-01
-2.82489657e-01 5.99251807e-01 -1.15709044e-01 -5.14145374e-01
-3.30472998e-02 -9.06075165e-02 -6.04991198e-01 7.75758445e-01
-1.68919683e-01 1.10313380e+00 -7.46533096e-01 -2.62849331e-01
7.99607560e-02 7.78933465e-01 -5.83809197e-01 4.31539007e-02
4.42408830e-01 4.74821478e-01 -1.80426966e-02 4.24368471e-01
1.22059667e+00 1.66265801e-01 4.62952018e-01 -9.46788788e-02
2.12591770e-03 1.73802137e-01 -1.34048927e+00 9.64917958e-01
-5.01540184e-01 7.75881469e-01 1.17900938e-01 -3.68700624e-02
7.23671138e-01 5.34840822e-01 3.65715504e-01 -3.29572707e-01
1.94183052e-01 1.82006180e-01 -3.74765635e-01 -2.15613008e-01
6.19266927e-01 2.02915922e-01 -1.26032516e-01 1.07973933e+00
-8.18801522e-02 3.19555372e-01 -1.53023884e-01 2.34830484e-01
6.74221218e-01 -5.00899144e-02 3.21708113e-01 -2.36952126e-01
6.69130564e-01 -6.52462959e-01 4.73103136e-01 6.20137036e-01
3.03299651e-02 4.27153975e-01 5.22006333e-01 -3.02164227e-01
-1.20849645e+00 -8.97203743e-01 7.92139992e-02 4.26002741e-01
2.62771457e-01 -8.22028577e-01 -1.18105185e+00 -7.52595425e-01
1.99829444e-01 3.05088103e-01 -5.66316307e-01 -2.30414793e-01
-6.67791843e-01 -3.34721208e-01 1.40480995e+00 5.71846403e-02
1.05508721e+00 -7.92991817e-01 -4.40082461e-01 -2.04531148e-01
-3.23276609e-01 -1.02895308e+00 -9.35967743e-01 -7.50010192e-01
-4.60511088e-01 -1.16736710e+00 -7.74073660e-01 -6.64665639e-01
6.90418780e-01 4.42378908e-01 6.30868018e-01 1.85682103e-01
-1.43126473e-01 5.53693712e-01 -2.50916299e-03 -1.70353368e-01
-8.77026558e-01 2.57456750e-01 4.19733018e-01 6.49280369e-01
2.42106557e-01 -6.64422810e-01 -7.53761470e-01 4.31117058e-01
-9.19327438e-01 1.03761494e-01 1.55729085e-01 6.95495367e-01
1.14103138e-01 2.16550708e-01 1.79741457e-01 -9.31899250e-01
6.26824617e-01 -2.91651696e-01 -4.33283776e-01 1.87571213e-01
-5.05748570e-01 5.00903726e-02 8.48335445e-01 -5.94715655e-01
-8.01073015e-01 8.42492953e-02 -7.57154152e-02 -1.31725669e-01
8.97021890e-02 -3.66557360e-01 -6.25435770e-01 -6.26472056e-01
3.27772886e-01 6.14652693e-01 4.10641581e-01 -2.56141603e-01
5.94043195e-01 9.81989324e-01 6.65870011e-01 -2.08106235e-01
1.26209545e+00 8.06851387e-01 -1.10775016e-01 -7.34460056e-01
2.00026914e-01 2.16439664e-02 -4.75564688e-01 -1.36175618e-01
5.62169611e-01 -1.08648384e+00 -1.44849622e+00 1.17208135e+00
-9.35920060e-01 -5.56024015e-02 -3.56333017e-01 2.69510418e-01
-4.70329612e-01 7.45435894e-01 -4.06521320e-01 -6.55085325e-01
-6.91997230e-01 -1.04988801e+00 9.08527613e-01 2.31372580e-01
-4.01111096e-01 -9.22937095e-01 8.14813673e-02 1.21341616e-01
5.47527313e-01 3.84955168e-01 5.20117939e-01 -3.82062435e-01
-6.18377328e-01 -4.45177674e-01 1.45715447e-02 4.76541042e-01
6.32856786e-01 -2.21143216e-02 -1.06722796e+00 -7.63226688e-01
1.04649864e-01 -8.55564773e-02 5.41130960e-01 -1.41248778e-01
1.00042808e+00 -1.04684019e+00 -3.27467352e-01 1.33258605e+00
1.21464801e+00 -9.32748839e-02 1.05737293e+00 2.43546605e-01
6.78176582e-01 4.72003222e-01 -2.65884046e-02 4.89184380e-01
4.65118885e-01 5.80416143e-01 2.12769911e-01 -1.19787060e-01
-1.80002376e-01 -7.25055039e-01 5.90553403e-01 3.94558340e-01
-9.14202705e-02 -2.33225182e-01 -3.98464590e-01 3.59501392e-01
-1.25623751e+00 -1.44425547e+00 1.50466427e-01 2.42455173e+00
1.02386987e+00 -5.13517737e-01 2.90871352e-01 7.60959089e-02
7.91138351e-01 2.45834708e-01 -2.31795952e-01 -4.00042415e-01
-3.82972330e-01 3.56778383e-01 1.08255148e+00 4.72161144e-01
-9.84410942e-01 1.05381262e+00 6.70407820e+00 9.03412402e-01
-1.24557602e+00 5.79994917e-02 6.17678940e-01 -3.56469937e-02
-4.25658554e-01 1.54161900e-01 -9.91039991e-01 9.24630821e-01
8.46648216e-01 -5.11381209e-01 6.55499518e-01 5.90563893e-01
-3.86168845e-02 2.90552735e-01 -9.83837187e-01 1.29418004e+00
1.44208446e-01 -1.49449718e+00 2.00827777e-01 5.57185709e-01
5.82860887e-01 -7.57108152e-01 5.13535440e-01 -1.62582383e-01
4.02944475e-01 -1.06525660e+00 6.68437421e-01 3.64951700e-01
1.52512419e+00 -1.06777656e+00 4.82665181e-01 6.13041706e-02
-1.16419399e+00 -1.07170902e-01 -2.22654983e-01 4.10805076e-01
1.81034088e-01 8.98991227e-02 -1.04805028e+00 7.12613165e-01
5.21890938e-01 5.84687293e-01 -4.89056110e-01 5.73364019e-01
-5.05672932e-01 4.27019387e-01 -4.09445524e-01 1.89493492e-01
-4.02929895e-02 -3.48642498e-01 7.07219601e-01 1.16637814e+00
3.93967479e-01 2.71834061e-02 -3.81311089e-01 6.95635498e-01
-4.52570528e-01 -5.60281873e-02 -9.64840591e-01 -2.28456795e-01
7.46488333e-01 1.13811624e+00 -1.68942377e-01 -1.38774320e-01
1.83690444e-01 1.46794081e+00 -2.29467466e-01 1.54505834e-01
-7.42741644e-01 -5.94852269e-01 1.13167238e+00 1.89792499e-01
2.39425123e-01 -2.91377217e-01 -1.23798169e-01 -1.22356808e+00
-2.55852044e-02 -1.28906381e+00 1.96161807e-01 -3.70484501e-01
-8.68140996e-01 7.00136662e-01 -1.79384694e-01 -1.28471231e+00
-3.12790960e-01 -1.78288653e-01 -3.76256078e-01 1.05163813e+00
-1.32795048e+00 -1.48116553e+00 -1.35417357e-01 8.96054745e-01
-3.62196207e-01 -4.16783899e-01 1.08451962e+00 3.64265054e-01
-3.37624162e-01 1.29746878e+00 1.17415033e-01 4.67901856e-01
7.10090458e-01 -8.74033988e-01 8.39884758e-01 9.41014707e-01
-1.57995895e-01 1.04276347e+00 3.73797446e-01 -7.66784668e-01
-1.49113119e+00 -9.80753124e-01 1.04179144e+00 -3.71805817e-01
1.88413367e-01 -8.76143277e-01 -4.71568495e-01 8.15369129e-01
3.54209036e-01 -5.26513875e-01 1.04925990e+00 -5.78663468e-01
-6.60624683e-01 -2.12762788e-01 -1.49724865e+00 8.28981400e-01
9.21299040e-01 -1.08664572e+00 6.75245821e-02 8.41760635e-03
4.77464646e-01 -2.27688476e-01 -5.60565472e-01 -9.04887095e-02
1.08323371e+00 -1.09446478e+00 1.13050163e+00 -3.14657182e-01
-9.56291258e-02 -3.03433031e-01 6.04695827e-02 -7.91983187e-01
-1.34648532e-02 -1.70626223e+00 -2.19721302e-01 1.60058081e+00
-7.80448765e-02 -1.10447037e+00 8.36321831e-01 5.23575544e-01
9.44568217e-01 -1.67104006e-01 -9.83897984e-01 -9.27553833e-01
-2.26363242e-01 -8.72125383e-03 1.33281672e+00 9.95163918e-01
-3.24685097e-01 -2.48310566e-01 -1.15568352e+00 3.29985946e-01
8.18495750e-01 -2.32437596e-01 1.29428875e+00 -1.00597382e+00
-3.15501690e-02 -9.29382071e-02 -6.10175848e-01 -9.56628740e-01
2.96706170e-01 -8.96045148e-01 -6.92413390e-01 -5.03735960e-01
1.28263116e-01 -5.66772401e-01 2.95555145e-01 6.62020147e-01
5.13280965e-02 7.56890237e-01 3.89699548e-01 3.36361170e-01
1.87302046e-02 4.91050720e-01 8.85611713e-01 -8.09474513e-02
-3.25912386e-01 1.58974037e-01 -1.03774536e+00 2.97503918e-01
9.26387548e-01 -6.93085134e-01 -4.94815975e-01 -4.80111428e-02
2.33137846e-01 -2.84664243e-01 5.23086190e-01 -8.97105217e-01
6.63357377e-02 1.93159223e-01 1.24283157e-01 -4.62019555e-02
3.58816296e-01 -9.10219669e-01 7.83390462e-01 6.89304531e-01
-7.34719262e-02 3.17119986e-01 1.04624189e-01 2.79584467e-01
7.12620094e-02 -2.30250340e-02 9.27020311e-01 6.00846298e-02
-3.65135640e-01 6.99585617e-01 -3.88043642e-01 -7.55058751e-02
1.21731007e+00 -7.66053379e-01 -1.15006350e-01 -7.50862539e-01
-3.43531042e-01 -4.41295356e-02 1.16306734e+00 3.39278013e-01
5.08296490e-01 -1.44804943e+00 -8.23513985e-01 8.19463432e-01
-1.41798824e-01 -6.90068722e-01 2.48295546e-01 4.48165268e-01
-7.56066084e-01 1.72584131e-01 -5.16779125e-01 -2.22594693e-01
-1.78772509e+00 4.68602508e-01 3.75991672e-01 -1.00660779e-01
-7.37529278e-01 5.37631094e-01 -4.14244607e-02 -2.57942766e-01
-8.08704346e-02 4.90274221e-01 3.07378056e-03 1.56781808e-01
1.01359093e+00 4.84538704e-01 -2.70307094e-01 -9.89652693e-01
-3.61058116e-01 6.76130712e-01 2.17284542e-02 -2.07579181e-01
1.02389276e+00 -4.61433291e-01 -3.93122435e-01 -5.56674659e-01
1.53722823e+00 7.46126950e-01 -1.19707906e+00 -3.45556182e-03
-4.92430657e-01 -8.02823842e-01 -3.92210364e-01 -2.73778349e-01
-1.13477290e+00 5.43332756e-01 5.83233058e-01 1.30252659e-01
1.08831596e+00 -6.71838284e-01 1.06851733e+00 -1.15641773e-01
3.95405918e-01 -5.69024146e-01 -3.84301752e-01 1.37889311e-01
3.60448271e-01 -5.85243523e-01 4.41242158e-02 -3.65499169e-01
-8.00821722e-01 9.95898306e-01 5.76250404e-02 1.39231488e-01
5.16162097e-01 2.36552358e-01 -3.01480293e-02 2.01146170e-01
-2.84252524e-01 3.41622502e-01 -6.59140572e-02 9.91560459e-01
-2.82203615e-01 1.06175303e-01 -1.24289356e-01 2.75559813e-01
-7.04953909e-01 -8.36857185e-02 7.10386753e-01 6.55491829e-01
3.83432120e-01 -1.73612082e+00 -4.33235765e-01 -3.17922354e-01
-7.52552748e-01 -2.23359361e-01 -4.49871063e-01 7.88552761e-01
1.84555709e-01 7.62730658e-01 -6.64139315e-02 -5.74300349e-01
-1.74305916e-01 -8.47212076e-02 2.50526339e-01 -1.98313221e-01
-7.92329848e-01 -6.24319792e-01 -4.88686822e-02 -6.09890342e-01
-3.04559797e-01 -4.08789247e-01 -7.03336716e-01 -1.26049817e+00
-8.11415166e-02 2.26298854e-01 5.19454002e-01 4.33034748e-01
8.41885149e-01 -2.62040466e-01 1.21411502e+00 -5.49872696e-01
-3.85918587e-01 -2.14794874e-01 -5.94125450e-01 3.49974990e-01
6.78273022e-01 -1.31560475e-01 -2.79754668e-01 2.19575822e-01] | [12.731184959411621, 0.7557722330093384] |
6d66af5a-8e06-4792-8866-5b3070cfca84 | semantic-query-by-example-speech-search-using | 1904.07078 | null | http://arxiv.org/abs/1904.07078v1 | http://arxiv.org/pdf/1904.07078v1.pdf | Semantic query-by-example speech search using visual grounding | A number of recent studies have started to investigate how speech systems can
be trained on untranscribed speech by leveraging accompanying images at
training time. Examples of tasks include keyword prediction and within- and
across-mode retrieval. Here we consider how such models can be used for
query-by-example (QbE) search, the task of retrieving utterances relevant to a
given spoken query. We are particularly interested in semantic QbE, where the
task is not only to retrieve utterances containing exact instances of the
query, but also utterances whose meaning is relevant to the query. We follow a
segmental QbE approach where variable-duration speech segments (queries, search
utterances) are mapped to fixed-dimensional embedding vectors. We show that a
QbE system using an embedding function trained on visually grounded speech data
outperforms a purely acoustic QbE system in terms of both exact and semantic
retrieval performance. | ['Karen Livescu', 'Aristotelis Anastassiou', 'Herman Kamper'] | 2019-04-15 | null | null | null | null | ['semantic-retrieval'] | ['natural-language-processing'] | [ 1.56948522e-01 1.76338479e-01 -3.86026800e-02 -4.81536478e-01
-1.61866808e+00 -5.36779523e-01 8.24291706e-01 1.30330101e-01
-4.76181418e-01 1.68742701e-01 5.50716162e-01 -8.74257162e-02
-1.81392003e-02 -2.57721543e-01 -7.37821400e-01 -5.27683556e-01
-1.34090498e-01 6.29787982e-01 2.44477868e-01 -2.36657694e-01
3.34929615e-01 3.65523160e-01 -1.68927240e+00 2.17509732e-01
1.80015266e-01 1.06675732e+00 7.14555919e-01 1.10836220e+00
-2.40704879e-01 4.74708349e-01 -9.58569527e-01 -6.77334964e-02
-2.87107766e-01 -3.00060868e-01 -1.07127023e+00 -1.75814843e-03
3.84142429e-01 -3.65299433e-01 -5.59051573e-01 8.27133954e-01
8.47695887e-01 5.75047195e-01 8.47010136e-01 -1.16057193e+00
-9.51206446e-01 1.29098132e-01 3.74167353e-01 3.26983571e-01
7.74266839e-01 -1.18311241e-01 1.36413670e+00 -1.25386000e+00
6.11000299e-01 1.55207264e+00 1.94105450e-02 8.24337363e-01
-9.95198727e-01 -3.41588646e-01 -4.01611552e-02 5.71869195e-01
-1.49994755e+00 -9.66444969e-01 6.27453625e-01 -2.84947962e-01
1.52394509e+00 4.22755092e-01 4.51014489e-01 1.05217350e+00
-1.49993852e-01 1.32494307e+00 4.12397861e-01 -6.16648436e-01
3.03025752e-01 3.50370944e-01 -4.08318862e-02 4.33644712e-01
-8.49487305e-01 1.70220196e-01 -8.98042917e-01 -8.08388069e-02
1.96312860e-01 -4.82008159e-01 -3.27907234e-01 -2.75218219e-01
-1.18953037e+00 1.05666673e+00 4.60651457e-01 3.50342363e-01
-3.41696233e-01 4.61055517e-01 5.22756517e-01 4.09661412e-01
5.65503657e-01 6.31999791e-01 -3.87857974e-01 -2.53915668e-01
-9.31572378e-01 2.77148604e-01 8.99493694e-01 1.09195960e+00
7.12500274e-01 8.54643732e-02 -3.34130287e-01 1.25718737e+00
5.01054466e-01 8.95567656e-01 8.64172637e-01 -9.20082390e-01
1.68142974e-01 -2.13879302e-01 1.18411459e-01 -6.07258856e-01
3.03472974e-03 6.39362782e-02 3.08579803e-02 -4.64476496e-01
-1.91954747e-01 4.23405141e-01 -1.01425374e+00 1.53297782e+00
1.00148134e-01 3.17821838e-02 6.46511197e-01 9.50869262e-01
1.23044086e+00 1.30370593e+00 1.74863994e-01 -1.30440816e-01
1.48281693e+00 -1.04573882e+00 -9.16639268e-01 -1.94873422e-01
5.60792625e-01 -8.29347432e-01 1.13552189e+00 8.73553827e-02
-9.81836617e-01 -5.34102380e-01 -7.25464821e-01 -2.70908564e-01
-6.40595794e-01 -4.61500064e-02 -1.25568151e-01 3.69667798e-01
-1.57215798e+00 -4.00174130e-03 -4.11371827e-01 -6.54708803e-01
-3.53120267e-02 2.85330743e-01 -3.03495497e-01 6.47544721e-03
-1.44716680e+00 9.27796423e-01 3.96077424e-01 -1.01136222e-01
-1.51806808e+00 -2.27428064e-01 -9.93469298e-01 3.10375571e-01
4.58409071e-01 -4.46647465e-01 1.82813883e+00 -9.20101464e-01
-1.43610799e+00 9.24830198e-01 -5.07684708e-01 -3.51461321e-01
-4.23555613e-01 -5.26339822e-02 -7.39504635e-01 7.73442924e-01
9.76126269e-02 1.08464158e+00 1.13655841e+00 -1.10345805e+00
-2.98089832e-01 -3.34459424e-01 -2.32924819e-02 5.08283854e-01
-2.05255434e-01 4.36044782e-01 -7.54868925e-01 -5.87242842e-01
-5.95276244e-02 -9.28762853e-01 2.36559525e-01 1.33025438e-01
-3.94499004e-01 -6.84620976e-01 9.91518676e-01 -8.59750867e-01
9.75557506e-01 -2.29881191e+00 1.43850759e-01 3.34299393e-02
-2.63421297e-01 2.47419536e-01 -4.13139045e-01 7.96325862e-01
2.17710599e-01 2.00991631e-01 5.22516966e-02 -4.16583747e-01
4.14765000e-01 2.92952508e-01 -7.01565504e-01 2.86932886e-02
2.41572276e-01 1.20092165e+00 -9.49885130e-01 -6.88130200e-01
4.44286950e-02 5.64658642e-01 -1.51984498e-01 5.74458897e-01
-3.06923807e-01 7.61181936e-02 -6.86516106e-01 6.11127019e-01
-1.84633896e-01 -1.32472232e-01 -3.78547788e-01 -7.18368143e-02
2.95904040e-01 4.13038820e-01 -5.57849765e-01 1.78126585e+00
-6.59623802e-01 1.05238402e+00 1.29462704e-01 -1.00397074e+00
9.32687163e-01 7.76902378e-01 2.04363942e-01 -1.04413021e+00
-4.02408272e-01 3.09319824e-01 -4.87929493e-01 -7.76835918e-01
8.07072341e-01 -2.88535595e-01 -6.47722855e-02 5.01057506e-01
3.28487217e-01 -6.35627627e-01 -3.40737253e-01 3.51913482e-01
8.83017302e-01 -1.72163665e-01 9.12796035e-02 6.49537221e-02
4.35327947e-01 5.44596575e-02 -3.71992290e-01 7.59335637e-01
-2.80320674e-01 5.44947803e-01 -1.55412816e-02 1.66361034e-01
-1.09276390e+00 -1.06508815e+00 -8.62137750e-02 1.49669445e+00
2.68055379e-01 -2.66067237e-01 -5.22695839e-01 -4.04281735e-01
-5.25718145e-02 8.59093904e-01 -2.98128009e-01 -2.89106995e-01
-1.09315172e-01 1.07669316e-01 5.23646235e-01 4.22985494e-01
-6.48434684e-02 -1.60790396e+00 -6.35573685e-01 3.27021658e-01
-2.07263932e-01 -1.29515707e+00 -8.97881448e-01 1.26731828e-01
-5.26523411e-01 -5.44208467e-01 -1.27781081e+00 -1.12383306e+00
3.16167444e-01 3.23137015e-01 1.19627285e+00 -1.55675352e-01
-1.99188337e-01 1.34580588e+00 -8.49876940e-01 -4.80227590e-01
-6.77898943e-01 -2.39159226e-01 -3.17622125e-02 -6.29368126e-02
4.48023468e-01 1.09814577e-01 -6.68966591e-01 3.62233728e-01
-1.09260249e+00 -3.78389269e-01 4.80043143e-01 9.85435069e-01
8.40632021e-01 -4.91313964e-01 8.21360111e-01 -5.12316003e-02
1.00169456e+00 -3.59433144e-01 -1.90265760e-01 6.60320342e-01
-4.40750033e-01 2.27507398e-01 9.50520188e-02 -5.29145598e-01
-6.17223322e-01 -8.22279006e-02 -3.44753534e-01 -6.97531104e-01
-1.03182994e-01 5.84540486e-01 5.16431928e-02 1.11295842e-01
4.01072472e-01 9.10438299e-01 1.78052504e-02 -3.73893946e-01
5.42954981e-01 1.19964075e+00 4.42527890e-01 -4.09864128e-01
1.22822925e-01 1.55965507e-01 -5.17514288e-01 -1.42691851e+00
-3.21202129e-01 -1.05187917e+00 -2.85679072e-01 -3.28147501e-01
1.10027313e+00 -8.40797365e-01 -4.92806643e-01 -1.08264580e-01
-1.20683205e+00 -2.60889083e-01 -3.41622531e-01 4.96225566e-01
-9.58445251e-01 2.80608773e-01 -3.56424809e-01 -1.21315205e+00
-4.06144083e-01 -1.27547729e+00 1.95749903e+00 -7.53133520e-02
-1.70777828e-01 -7.45949984e-01 9.40114483e-02 3.32532525e-01
3.69771063e-01 -6.85652137e-01 9.44689989e-01 -1.25835490e+00
-5.67111492e-01 -2.14255467e-01 -7.62816966e-02 2.72533506e-01
-1.18247427e-01 -3.20338398e-01 -1.38269305e+00 -3.22395772e-01
-1.80826753e-01 -7.17584372e-01 8.07652771e-01 5.39564788e-01
1.09695196e+00 -4.38582420e-01 -3.46615106e-01 1.03964485e-01
1.00441742e+00 6.10969007e-01 3.47366422e-01 -2.68404204e-02
2.79965430e-01 1.04570282e+00 6.66282237e-01 5.64416721e-02
3.66606385e-01 1.25105155e+00 2.57264823e-01 8.25450346e-02
-1.13846533e-01 -4.53110635e-01 4.28041935e-01 8.79590929e-01
7.53981233e-01 -6.39704883e-01 -8.92905712e-01 1.03147316e+00
-1.58305728e+00 -9.84789491e-01 7.25678980e-01 2.01371121e+00
6.28858387e-01 -2.61108935e-01 1.51411042e-01 -2.19969839e-01
4.06820446e-01 2.64157921e-01 -7.00130582e-01 -5.54626703e-01
-3.59288859e-03 1.20552741e-01 -7.28765056e-02 6.51475728e-01
-1.06087780e+00 9.61725533e-01 6.97848034e+00 9.60071445e-01
-1.08478391e+00 2.96279311e-01 4.26898926e-01 8.97986516e-02
-4.18652385e-01 -2.80067086e-01 -6.14101410e-01 7.52611235e-02
1.53728139e+00 -2.62398571e-01 3.09329182e-01 7.48834550e-01
2.83557475e-01 -5.54106757e-02 -1.28759909e+00 1.07002008e+00
4.33408320e-01 -1.09846020e+00 1.89591855e-01 -9.45320353e-02
1.66706473e-01 1.61446244e-01 2.45796293e-01 5.20372987e-01
7.15269614e-03 -1.23895860e+00 8.68146300e-01 4.93232518e-01
9.21890497e-01 -3.45727414e-01 3.76423627e-01 3.64987850e-01
-1.19944191e+00 1.43824503e-01 -2.77627975e-01 7.06299901e-01
3.14952523e-01 -3.34899873e-01 -1.29692197e+00 1.78136960e-01
8.24878573e-01 4.94524539e-01 -2.38994196e-01 9.15620506e-01
3.68496999e-02 4.79505897e-01 -2.30622604e-01 -6.29876673e-01
4.60206062e-01 1.97240666e-01 6.16769969e-01 1.38119733e+00
4.25099343e-01 1.74289241e-01 9.70977992e-02 8.05033863e-01
2.02198513e-02 4.90706712e-01 -9.17762220e-01 -4.86594081e-01
5.17720401e-01 5.60502231e-01 -4.65031892e-01 -6.12661004e-01
-4.19801056e-01 1.27169824e+00 -9.80414152e-02 7.89733350e-01
-2.24436313e-01 -4.59561020e-01 5.40867329e-01 -2.42023066e-01
7.16917992e-01 -1.61053777e-01 8.77786398e-01 -7.77882814e-01
1.12881586e-02 -7.63974726e-01 1.30782694e-01 -1.46475291e+00
-1.10415924e+00 6.83384418e-01 2.49501377e-01 -1.16639447e+00
-8.52934957e-01 -4.02285397e-01 -5.82013391e-02 1.03205466e+00
-1.57363498e+00 -1.07432318e+00 2.69084334e-01 5.42619646e-01
1.18675625e+00 -3.05427670e-01 1.22958958e+00 -1.05924457e-01
8.82792249e-02 4.15545225e-01 3.01827699e-01 4.67014201e-02
5.47411382e-01 -1.07788849e+00 2.96006322e-01 2.12281302e-01
9.42710161e-01 4.43052828e-01 8.84110808e-01 -3.21593016e-01
-1.46077013e+00 -6.98364019e-01 1.25782692e+00 -4.53803658e-01
5.99306107e-01 -4.45247680e-01 -1.02737117e+00 3.78475159e-01
1.39949501e-01 4.41706590e-02 5.97952902e-01 -1.94357917e-01
-3.79458874e-01 -4.89900187e-02 -8.33071768e-01 3.42045277e-01
5.61462402e-01 -1.55531454e+00 -7.42364407e-01 5.82165897e-01
1.18621838e+00 -1.37066856e-01 -7.87731349e-01 -5.75508028e-02
6.90477371e-01 -3.60340387e-01 1.31202638e+00 -6.72647595e-01
3.38488184e-02 -5.88078909e-02 -6.35925055e-01 -1.39914680e+00
4.36367571e-01 -4.54848737e-01 -1.91314787e-01 7.27708042e-01
4.64416504e-01 -1.02811009e-01 4.97863591e-01 3.93111140e-01
-2.58179694e-01 -7.20870972e-01 -1.34103858e+00 -9.37565863e-01
-7.30807036e-02 -5.32385588e-01 5.40350378e-01 3.05741161e-01
6.74110604e-03 5.68027675e-01 -3.18806559e-01 2.23624453e-01
2.32495554e-02 1.11175492e-01 4.92434651e-01 -8.26932847e-01
-3.08046043e-01 -4.22682166e-01 -5.38343668e-01 -1.72370160e+00
5.16325355e-01 -9.06318963e-01 6.54051960e-01 -1.53747821e+00
5.42908572e-02 9.52165388e-03 -2.52836317e-01 1.62869886e-01
8.66946205e-02 -3.25868465e-02 1.73894748e-01 5.26853204e-01
-7.48841107e-01 1.01702929e+00 1.14679599e+00 -5.30320168e-01
-1.09882891e-01 -1.60428323e-02 -1.54051006e-01 -4.20382693e-02
1.99034870e-01 -3.58292699e-01 -7.15942860e-01 -4.62416559e-01
3.87769635e-03 7.65677989e-01 5.54662645e-01 -5.24191976e-01
4.15728688e-01 1.07945047e-01 -2.06300899e-01 -5.93645155e-01
1.06979537e+00 -7.88435161e-01 -1.56491026e-01 6.20834976e-02
-9.27476585e-01 1.67766526e-01 2.56963521e-01 8.93877804e-01
-9.49694633e-01 -5.72557032e-01 2.30846986e-01 -1.03236869e-01
-1.32498896e+00 1.86129749e-01 -5.12202740e-01 -1.17645450e-01
7.06088662e-01 -3.03176969e-01 -4.16297764e-02 -1.17951858e+00
-1.15485263e+00 1.66439533e-01 -4.32680175e-02 5.77211738e-01
1.25900578e+00 -1.32242715e+00 -5.79996824e-01 -6.57183118e-03
8.64957213e-01 -4.91896808e-01 6.65399060e-02 4.30870265e-01
2.42162067e-02 1.05674267e+00 4.51124161e-01 -8.28727245e-01
-1.40466321e+00 5.80244362e-01 4.88351941e-01 3.11405301e-01
-4.80229527e-01 1.22355902e+00 3.64465415e-01 -4.08398300e-01
5.66805959e-01 -2.17060328e-01 -2.48321012e-01 7.61083961e-02
4.84003693e-01 -2.20973447e-01 1.10459805e-01 -1.19036388e+00
-4.78308856e-01 5.44613361e-01 1.02587104e-01 -7.74854064e-01
1.00327432e+00 -3.84532422e-01 4.46331292e-01 7.75286376e-01
1.69732666e+00 -4.74396348e-01 -9.22316909e-01 -4.46831524e-01
1.87389359e-01 -1.72882050e-01 2.10050404e-01 -8.01925838e-01
-5.88016570e-01 1.19162285e+00 8.52080464e-01 4.37260658e-01
1.01619780e+00 7.75962114e-01 7.96103001e-01 8.06028366e-01
3.69375139e-01 -1.16226768e+00 5.46369433e-01 4.54658538e-01
1.27401459e+00 -1.39809477e+00 -5.01207054e-01 9.14922357e-02
-8.42978239e-01 9.96201932e-01 -6.58400357e-02 2.34628320e-01
7.28571773e-01 -4.56310123e-01 2.34305769e-01 -5.52344024e-01
-9.02098894e-01 -5.60484946e-01 8.41756284e-01 6.44950032e-01
2.00351089e-01 6.75228909e-02 3.37297559e-01 5.09206846e-04
-1.86661154e-01 -4.83785778e-01 6.36151666e-03 8.09040725e-01
-6.71433628e-01 -8.53908896e-01 -2.10930556e-01 2.87861675e-01
-1.91329226e-01 -3.51194531e-01 -5.84871471e-01 5.94078302e-01
-7.46334076e-01 1.16001010e+00 1.13474466e-01 -3.53138000e-01
4.16218460e-01 5.07316053e-01 1.97954237e-01 -9.12529111e-01
-2.75098622e-01 3.82840306e-01 2.28157565e-01 -6.90714717e-01
-5.03466964e-01 -5.19957304e-01 -1.10156322e+00 5.89325368e-01
-5.06677687e-01 4.56399947e-01 1.04333436e+00 8.46851945e-01
2.75492042e-01 2.13426664e-01 4.99290645e-01 -6.23862207e-01
-7.78451264e-01 -1.01791775e+00 -6.65976644e-01 3.53430152e-01
8.47403049e-01 -4.75647926e-01 -3.67210269e-01 2.09935844e-01] | [10.724120140075684, 1.3386119604110718] |
b85265ca-2333-448d-a2e8-ece2590c7b76 | learning-to-predict-navigational-patterns | 2304.13242 | null | https://arxiv.org/abs/2304.13242v2 | https://arxiv.org/pdf/2304.13242v2.pdf | Learning to Predict Navigational Patterns from Partial Observations | Human beings cooperatively navigate rule-constrained environments by adhering to mutually known navigational patterns, which may be represented as directional pathways or road lanes. Inferring these navigational patterns from incompletely observed environments is required for intelligent mobile robots operating in unmapped locations. However, algorithmically defining these navigational patterns is nontrivial. This paper presents the first self-supervised learning (SSL) method for learning to infer navigational patterns in real-world environments from partial observations only. We explain how geometric data augmentation, predictive world modeling, and an information-theoretic regularizer enables our model to predict an unbiased local directional soft lane probability (DSLP) field in the limit of infinite data. We demonstrate how to infer global navigational patterns by fitting a maximum likelihood graph to the DSLP field. Experiments show that our SSL model outperforms two SOTA supervised lane graph prediction models on the nuScenes dataset. We propose our SSL method as a scalable and interpretable continual learning paradigm for navigation by perception. Code is available at https://github.com/robin-karlsson0/dslp. | ['Kazuya Takeda', 'Kento Ohtani', 'Keisuke Fujii', 'Francisco Lepe-Salazar', 'Alexander Carballo', 'Robin Karlsson'] | 2023-04-26 | null | null | null | null | ['lane-detection'] | ['computer-vision'] | [ 3.86703223e-01 6.84862792e-01 -4.51413929e-01 -7.92377293e-01
-3.38487178e-01 -5.23369193e-01 6.81783974e-01 -1.01472050e-01
-2.93790191e-01 9.84273255e-01 2.67632872e-01 -7.28132546e-01
-4.75947261e-01 -9.87402439e-01 -1.11480761e+00 -5.03399670e-01
-5.15620589e-01 7.67520547e-01 4.94204730e-01 -3.00119400e-01
4.11917597e-01 3.34402382e-01 -1.68335438e+00 -1.28861606e-01
1.10691893e+00 4.61327940e-01 5.39049327e-01 1.00962675e+00
2.21861787e-02 8.61533642e-01 3.34032357e-01 6.68592844e-03
4.03336376e-01 -1.06178276e-01 -6.65493071e-01 -2.30663661e-02
5.25282681e-01 -8.39424953e-02 -5.61589062e-01 7.23777235e-01
-6.19681180e-02 2.75151074e-01 1.14487064e+00 -1.46808934e+00
-4.32156682e-01 5.61892569e-01 -1.79483786e-01 -1.92727208e-01
3.90055746e-01 2.05312788e-01 1.14565074e+00 -4.76745129e-01
7.29086339e-01 1.26723909e+00 7.80337691e-01 1.27296597e-01
-1.27801430e+00 -1.16355039e-01 2.32865721e-01 4.89453286e-01
-1.29559648e+00 -3.93484890e-01 8.90755713e-01 -5.82918942e-01
7.32525349e-01 -8.75030458e-03 4.81882095e-01 1.09530234e+00
5.35828292e-01 9.75765884e-01 1.13211894e+00 -4.31142181e-01
6.32519782e-01 -2.11139128e-01 1.71621054e-01 1.18285990e+00
4.71166164e-01 1.18640259e-01 -7.32645273e-01 1.37189075e-01
7.10753739e-01 -2.80477613e-01 1.26003921e-01 -1.11652780e+00
-1.14788103e+00 6.36938274e-01 7.19800055e-01 -3.72823149e-01
-2.57064998e-01 3.44237268e-01 -9.63273570e-02 9.99377370e-02
4.82857004e-02 3.59930873e-01 -5.61470449e-01 -2.38829866e-01
-2.28029251e-01 2.57145435e-01 1.11102545e+00 1.24666619e+00
1.42366445e+00 -1.64748594e-01 8.61361563e-01 6.31728232e-01
4.96677279e-01 8.90147984e-01 6.79086745e-02 -1.49383545e+00
2.96532601e-01 4.68295038e-01 3.89743268e-01 -1.23623478e+00
-8.60795140e-01 -2.07568660e-01 -6.60178959e-01 2.53068745e-01
7.40339398e-01 -2.51612604e-01 -8.55234087e-01 1.63047361e+00
6.72281086e-02 -9.14953649e-02 2.71607041e-01 7.15465367e-01
2.60632575e-01 5.35971165e-01 -2.76931405e-01 2.84636974e-01
7.44357586e-01 -1.04044771e+00 -3.51721913e-01 -9.21995640e-01
1.00796902e+00 -6.77335858e-02 1.24902165e+00 4.03743148e-01
-2.77694702e-01 -3.32124203e-01 -1.15564656e+00 -2.32935548e-01
-6.94940925e-01 9.07931179e-02 9.94251192e-01 6.31698430e-01
-1.27128255e+00 6.93533346e-02 -1.02957249e+00 -6.89827025e-01
3.04138184e-01 1.57382265e-01 -5.13567686e-01 -5.95178045e-02
-8.86999488e-01 8.70784998e-01 5.00557661e-01 -3.19434181e-02
-6.58292532e-01 -1.56793296e-01 -1.31670773e+00 -5.18393874e-01
6.18088901e-01 -4.42479521e-01 1.00054705e+00 -3.66523474e-01
-1.31333566e+00 7.39524662e-01 -5.09497166e-01 -5.77082872e-01
4.87446070e-01 -2.25310773e-01 -1.63062245e-01 -3.02444100e-01
4.43851352e-01 1.00136673e+00 2.11541012e-01 -1.75897694e+00
-8.91576290e-01 -3.15884501e-01 -2.94351615e-02 5.35724521e-01
2.81691968e-01 -1.29305983e+00 -2.02703670e-01 5.40211424e-02
5.89657843e-01 -1.23719001e+00 -6.45209134e-01 1.00349821e-01
-7.20820785e-01 -1.65251374e-01 5.61709881e-01 -4.36938047e-01
7.09402144e-01 -1.82461870e+00 -2.49498356e-02 6.37903512e-01
1.06571749e-01 -6.24554276e-01 -4.60874327e-02 3.99005026e-01
6.53817356e-01 -1.25061929e-01 -5.63313127e-01 -2.73312032e-01
3.62045199e-01 7.89665937e-01 -4.11086321e-01 5.49681008e-01
-4.24320012e-01 8.25466931e-01 -1.23386097e+00 -3.03536206e-01
3.17655444e-01 -7.61800334e-02 -6.12959862e-01 2.41526254e-02
-3.31894010e-01 5.25492609e-01 -3.64791125e-01 5.90954542e-01
5.24817169e-01 -2.07326457e-01 4.01759267e-01 3.10790151e-01
-2.95807332e-01 3.15716445e-01 -1.06960464e+00 1.62410963e+00
-4.46735442e-01 9.67952132e-01 -2.90018976e-01 -1.06253636e+00
1.21697628e+00 -7.98433959e-01 1.64617635e-02 -7.47705042e-01
-2.57468998e-01 3.29650551e-01 -2.71772951e-01 -4.55775321e-01
6.47728860e-01 3.07002813e-01 -5.42348802e-01 2.20500439e-01
-1.14347190e-01 -2.70897090e-01 -2.95420196e-02 2.17130825e-01
1.14390779e+00 3.78322273e-01 4.72639561e-01 -3.59344035e-01
2.43470833e-01 3.34510744e-01 4.39469308e-01 1.32184374e+00
-2.39453703e-01 2.14070439e-01 4.72350061e-01 -7.02768445e-01
-1.07596028e+00 -1.61478388e+00 -7.29176402e-02 1.17921686e+00
7.01045930e-01 -2.32009947e-01 -3.51628363e-01 -5.07255375e-01
1.14684105e-01 1.03844893e+00 -5.98445654e-01 3.88585255e-02
-4.79266137e-01 -5.02920866e-01 4.46404159e-01 4.69424188e-01
5.95505416e-01 -8.59107196e-01 -5.51844001e-01 -1.76651493e-01
-2.36887574e-01 -1.09889829e+00 -2.59199720e-02 4.86221403e-01
-3.50167453e-01 -1.03717303e+00 6.20061122e-02 -8.41625869e-01
7.45499253e-01 3.86745483e-01 6.63359523e-01 -3.10227841e-01
-9.33583546e-03 6.75200880e-01 -2.42793843e-01 -2.80612648e-01
-2.62399971e-01 3.18032980e-01 3.93447816e-01 -1.65380314e-01
4.59143758e-01 -7.49705672e-01 -4.11590904e-01 6.09002888e-01
-1.09214172e-01 3.98042887e-01 6.74518347e-01 6.20695829e-01
7.82495856e-01 -2.36913655e-02 3.83680463e-01 -7.83599436e-01
3.47474188e-01 -8.34406495e-01 -8.14114153e-01 1.02240279e-01
-6.70460641e-01 4.41215038e-01 5.10904193e-01 1.71941686e-02
-1.18343472e+00 5.95147789e-01 1.29029289e-01 4.49529737e-01
-7.33969033e-01 5.81466138e-01 -2.75385737e-01 -1.18523538e-01
9.45145190e-01 4.85197395e-01 9.86299813e-02 -2.79114470e-02
8.35695148e-01 6.75379038e-01 7.53043354e-01 -5.67284822e-01
8.49571109e-01 7.59824812e-01 3.62314016e-01 -1.09279323e+00
-6.78089917e-01 -5.49205661e-01 -1.25654685e+00 -5.04762173e-01
7.08693027e-01 -7.82088041e-01 -6.48594618e-01 3.41352493e-01
-6.07765079e-01 -1.06471074e+00 4.49644076e-03 3.94905418e-01
-1.27732754e+00 3.88062537e-01 4.09469046e-02 -1.08132684e+00
5.85947931e-01 -6.65518761e-01 9.08943951e-01 2.09465057e-01
-4.26298201e-01 -1.33842254e+00 3.29193622e-01 4.53779936e-01
1.23723499e-01 2.06929564e-01 5.20668626e-01 -5.08371413e-01
-1.03510344e+00 -1.89696729e-01 -9.64921638e-02 -3.18402082e-01
2.16649711e-01 -4.58539605e-01 -7.81132460e-01 1.62279576e-01
-7.19242454e-01 -3.65262419e-01 9.52492952e-01 4.44171131e-01
8.49062562e-01 -4.41144973e-01 -6.57203019e-01 7.08364725e-01
1.10867226e+00 2.55336314e-01 4.32709783e-01 6.87703609e-01
7.99291492e-01 9.56130207e-01 7.23114073e-01 2.83199668e-01
1.25246716e+00 2.84276247e-01 6.30501747e-01 4.14585501e-01
1.20320119e-01 -9.75629985e-01 3.44074130e-01 4.61806774e-01
3.05222720e-02 -4.09569532e-01 -1.44508135e+00 7.09921241e-01
-2.30832505e+00 -8.87405932e-01 -3.49073112e-01 2.01958370e+00
2.21506447e-01 4.48247463e-01 -2.20704991e-02 -1.75678819e-01
2.85768747e-01 1.55956790e-01 -8.73270273e-01 -8.81224796e-02
-4.35123563e-01 -8.52117419e-01 1.09907758e+00 1.27352428e+00
-1.25817716e+00 1.18090904e+00 6.27001667e+00 5.12950957e-01
-5.93301177e-01 -3.20425868e-01 4.44128871e-01 6.49383307e-01
-5.65348029e-01 2.15311319e-01 -8.08314502e-01 1.62774980e-01
7.50866652e-01 2.11984381e-01 6.63015962e-01 1.13247764e+00
3.69478941e-01 -5.49673676e-01 -8.37368548e-01 7.98693955e-01
7.93760195e-02 -1.38525045e+00 -1.69061631e-01 3.50419968e-01
6.61534309e-01 4.38867927e-01 -1.66146588e-02 2.03795120e-01
9.93927658e-01 -9.34517145e-01 7.23501503e-01 8.77015114e-01
3.62596005e-01 -4.27022487e-01 4.58299756e-01 8.54649127e-01
-1.15243018e+00 -3.39460194e-01 -3.98582727e-01 -4.31516171e-01
3.65624756e-01 3.05071056e-01 -1.49376929e+00 3.58205408e-01
6.05949223e-01 1.01377678e+00 -6.43750250e-01 9.39995825e-01
-4.55473989e-01 7.06659198e-01 -7.28332520e-01 -4.22582358e-01
2.54156172e-01 -4.61386353e-01 7.14008510e-01 1.10220528e+00
2.09600270e-01 -4.65855636e-02 2.30048761e-01 7.19345868e-01
4.19238389e-01 -2.96274066e-01 -1.16186714e+00 4.17594254e-01
6.62559867e-01 8.10319602e-01 -9.28056240e-01 7.00133294e-02
-7.96308964e-02 7.49634027e-01 4.67294246e-01 4.88686949e-01
-3.69484127e-01 -3.17458987e-01 4.71797019e-01 2.90385813e-01
3.32331508e-02 -9.40263271e-01 -8.19936812e-01 -8.85715544e-01
-4.65192311e-02 -4.24292460e-02 1.23157233e-01 -9.74198699e-01
-1.22429585e+00 1.48299873e-01 1.37641549e-01 -1.21539831e+00
-4.73664045e-01 -8.72464776e-01 -3.73015165e-01 2.63500005e-01
-1.50249839e+00 -1.34261334e+00 -5.47562361e-01 2.57033587e-01
5.04184961e-01 -2.76843578e-01 5.56230128e-01 -5.30858099e-01
3.42329661e-03 1.42580435e-01 3.80209774e-01 -5.31623475e-02
4.55498487e-01 -1.65062571e+00 5.05094767e-01 8.19873750e-01
4.94906306e-02 3.29916239e-01 9.29848433e-01 -9.58237588e-01
-1.42655802e+00 -1.10299766e+00 6.84816062e-01 -7.59905636e-01
8.48192573e-01 -6.50267363e-01 -6.96089923e-01 1.10050118e+00
-2.29367658e-01 -2.81815171e-01 6.49195492e-01 2.88869858e-01
-3.53814632e-01 -1.33092955e-01 -8.61026824e-01 9.81387734e-01
1.63351893e+00 -3.79023969e-01 -4.99830335e-01 2.75711477e-01
5.53201973e-01 -1.84019595e-01 -3.28568846e-01 3.33451390e-01
6.66465223e-01 -9.19336021e-01 8.99942875e-01 -1.86113983e-01
7.61524662e-02 -5.94370008e-01 -5.39626062e-01 -1.42328417e+00
-3.31121355e-01 -5.62126577e-01 3.91944461e-02 4.77783650e-01
7.63717413e-01 -9.51149940e-01 1.13836670e+00 3.33277792e-01
-3.07710409e-01 -3.22111011e-01 -9.19802904e-01 -8.62332702e-01
-1.27423316e-01 -8.28945875e-01 1.61608145e-01 7.17405438e-01
2.00829789e-01 2.36117855e-01 -4.94994253e-01 7.14089751e-01
1.25782311e+00 -1.83887511e-01 1.29350734e+00 -1.19889402e+00
5.77761531e-02 -2.93492556e-01 -7.04638958e-01 -1.69142234e+00
4.43114758e-01 -8.55051696e-01 6.51474595e-01 -1.82949162e+00
-2.50360608e-01 -7.65652895e-01 2.24180385e-01 4.71080184e-01
2.89122880e-01 -2.21995845e-01 -3.57302606e-01 1.90884277e-01
-9.86657858e-01 5.47128558e-01 8.40643048e-01 -1.07503437e-01
-4.06331569e-01 8.47643986e-02 -5.04748762e-01 1.10491157e+00
1.18881142e+00 9.29523781e-02 -7.85264015e-01 -2.64660329e-01
6.21820152e-01 -2.26489097e-01 5.38464129e-01 -1.04528368e+00
8.92401636e-01 -6.04561508e-01 1.45833567e-01 -8.92260790e-01
4.20189917e-01 -5.31131804e-01 -9.45561305e-02 3.74050736e-01
-3.80425841e-01 -2.87286520e-01 -9.89527851e-02 1.19705403e+00
3.38556588e-01 1.29361838e-01 3.28473926e-01 6.06059423e-03
-1.57587099e+00 -1.77994639e-01 -9.52915370e-01 -1.62842259e-01
9.20521140e-01 -6.15881264e-01 -4.98322606e-01 -9.05297279e-01
-7.63406992e-01 6.92963541e-01 6.48940802e-01 3.91889602e-01
8.65854919e-01 -9.88275886e-01 -4.54337865e-01 3.01456511e-01
4.17273194e-01 1.68260172e-01 1.22913748e-01 5.45497894e-01
-8.89490485e-01 6.46001577e-01 -3.14090252e-01 -7.68097460e-01
-5.44972777e-01 1.96895167e-01 3.78038615e-01 2.42840484e-01
-5.34536481e-01 6.28482640e-01 1.63377985e-01 -1.33768618e+00
7.66159669e-02 -2.16479838e-01 -1.23956911e-01 -3.30357939e-01
1.80774033e-02 6.22334421e-01 -5.24780154e-01 -7.56123066e-01
-2.80370563e-01 5.77272773e-01 2.93987066e-01 -2.80817002e-01
1.19698215e+00 -8.36284280e-01 1.31601349e-01 8.76061082e-01
7.16231823e-01 -5.02223196e-03 -1.69523668e+00 -1.53050154e-01
4.29868877e-01 -2.48058572e-01 -3.41580570e-01 -6.64626837e-01
-1.49240410e-02 4.40925807e-01 3.39528978e-01 2.02580333e-01
3.97759467e-01 4.42098230e-01 3.66305828e-01 1.22101116e+00
1.10268569e+00 -1.09200537e+00 3.03111617e-02 8.55466485e-01
6.42673969e-01 -1.54009056e+00 -2.25725874e-01 -5.29749632e-01
-6.74431503e-01 9.74863648e-01 8.11744571e-01 -1.99859992e-01
8.09516847e-01 2.50215739e-01 -1.44249266e-02 -1.49188435e-03
-6.30560219e-01 -3.12598675e-01 1.73703581e-01 1.16453075e+00
-2.51998961e-01 5.33065915e-01 3.55025500e-01 3.23658705e-01
-9.26543415e-01 -4.39394861e-01 8.54988515e-01 8.97969544e-01
-1.10731602e+00 -6.17669225e-01 -2.98178732e-01 5.40642202e-01
7.41255224e-01 2.59829998e-01 -4.29995567e-01 8.62929046e-01
5.54770753e-02 9.56582844e-01 2.56671131e-01 -5.84315896e-01
8.43147933e-02 -1.59529567e-01 2.19701961e-01 -2.40057558e-01
1.04391384e+00 -4.54154134e-01 3.70215625e-01 -6.28224015e-01
-3.31576675e-01 -8.61828506e-01 -1.49071312e+00 -1.10871837e-01
2.70800382e-01 -9.38474387e-02 5.36332190e-01 1.08974373e+00
3.43051344e-01 3.63745429e-02 4.59931463e-01 -9.65864539e-01
6.73080757e-02 -4.98904198e-01 -4.49603468e-01 -6.72726482e-02
5.88496983e-01 -9.41773534e-01 -6.26877248e-01 2.43314028e-01] | [5.0732741355896, 0.6083695888519287] |
8a08c706-fc4f-4860-acc3-639e558555f7 | text-attentional-convolutional-neural-network | null | null | https://arxiv.org/abs/1510.03283 | https://arxiv.org/pdf/1510.03283.pdf | Text-attentional convolutional neural network for scene text detection | Recent deep learning models have demonstrated strong capabilities for classifying text and non-text components
in natural images. They extract a high-level feature computed globally from a whole image component (patch), where the cluttered background information may dominate true text features in the deep representation. This leads to less discriminative power and poorer robustness. In this work, we present a new system for scene text detection by proposing a novel Text-Attentional Convolutional Neural Network (Text-CNN) that particularly focuses on extracting text-related regions and features from the image components. We develop a new learning mechanism to train the Text-CNN with multi-level and rich supervised information, including text region mask, character label, and binary text/nontext information. The rich supervision information enables the Text CNN with a strong capability for discriminating ambiguous texts, and also increases its robustness against complicated background components. The training process is formulated as a multi-task learning problem, where low-level supervised information greatly facilitates main task of text/non-text classification. In addition, a powerful low-level detector called ContrastEnhancement Maximally Stable Extremal Regions (CE-MSERs) is developed, which extends the widely-used MSERs by enhancing intensity contrast between text patterns and background. This allows it to detect highly challenging text patterns, resulting in a higher recall. Our approach achieved promising results on the ICDAR 2013 dataset, with a F-measure of 0.82, improving the state-of-the-art results substantially. | ['Weilin Huang', 'Yu Qiao', 'Jian Yao', 'Tong He'] | 2016-03-24 | null | null | null | ieee-trans-on-image-processing-2016-2016-3 | ['scene-text-detection'] | ['computer-vision'] | [ 5.66950321e-01 -4.45494294e-01 -3.29610892e-02 -1.35411382e-01
-7.45809793e-01 -2.66151190e-01 7.80436277e-01 1.65801272e-01
-4.79822904e-01 4.47196186e-01 -8.22741389e-02 9.07215998e-02
3.38119090e-01 -7.26835966e-01 -5.75320780e-01 -1.18248689e+00
4.45006758e-01 2.41270810e-01 4.82323885e-01 -2.00725291e-02
2.45722368e-01 5.08301079e-01 -1.64376640e+00 7.95461237e-01
9.20672774e-01 1.10638201e+00 5.20993650e-01 5.48510909e-01
-4.95785534e-01 8.86619389e-01 -8.15828621e-01 -3.55980873e-01
9.29021239e-02 -4.83365387e-01 -4.31743890e-01 3.96569729e-01
6.46347642e-01 -3.12875986e-01 -2.43490130e-01 1.12994194e+00
5.11792183e-01 2.92876288e-02 8.72701883e-01 -8.06556106e-01
-5.28351247e-01 2.64450461e-01 -9.51480925e-01 3.72067630e-01
-5.94813824e-02 9.26883221e-02 9.27994132e-01 -1.16156411e+00
3.81343603e-01 1.12103796e+00 6.25375628e-01 3.04541916e-01
-1.07842565e+00 -5.12731135e-01 1.52284905e-01 1.76998436e-01
-1.43649161e+00 -2.59022862e-01 8.06586087e-01 -4.71673042e-01
8.28010619e-01 4.32250887e-01 4.70375031e-01 1.30072284e+00
5.83104312e-01 1.35497749e+00 1.09839082e+00 -5.48530579e-01
-3.87217700e-02 1.55705616e-01 -9.95662287e-02 7.74784923e-01
2.99392551e-01 -3.23307067e-01 -4.77827400e-01 2.44094640e-01
7.45101750e-01 3.01201969e-01 -3.20405155e-01 -2.17842497e-02
-1.22107565e+00 5.86654484e-01 3.20070684e-01 8.69265318e-01
-3.28597307e-01 -3.14282298e-01 3.40544552e-01 -6.07136451e-02
4.97495443e-01 -1.08598933e-01 -2.61925429e-01 3.35063607e-01
-1.16972423e+00 -2.25083739e-01 4.80519533e-01 5.35715759e-01
6.41371965e-01 1.79500848e-01 -4.98930365e-01 1.04952383e+00
1.45157412e-01 9.11689043e-01 5.94102502e-01 -3.61611657e-02
6.91309988e-01 8.47755075e-01 -2.53801882e-01 -1.21425200e+00
-5.01721382e-01 -6.61325634e-01 -1.40184438e+00 2.17031717e-01
3.65791291e-01 7.12229908e-02 -1.13226342e+00 1.12488711e+00
1.92093402e-01 -3.32915097e-01 -8.66884217e-02 8.59883368e-01
8.75633955e-01 8.52727950e-01 -1.84595153e-01 -2.56557405e-01
1.42156160e+00 -9.60220516e-01 -9.15590823e-01 -5.57156861e-01
4.36228395e-01 -9.28858340e-01 1.18081594e+00 5.00249565e-01
-7.48069882e-01 -7.04425693e-01 -9.50823188e-01 -4.84176874e-02
-6.35098100e-01 8.07466865e-01 3.19028199e-02 4.95723486e-01
-8.46259952e-01 1.45119265e-01 -6.08648181e-01 -4.23245490e-01
5.16466260e-01 2.71507919e-01 -2.71587729e-01 -8.53761286e-02
-9.23474789e-01 7.16408789e-01 4.07140583e-01 2.57374495e-01
-6.23421192e-01 1.19798884e-01 -7.99141645e-01 3.18040311e-01
4.69265819e-01 -3.07243168e-01 5.98927855e-01 -1.48323679e+00
-1.33934700e+00 9.88572240e-01 -1.92419410e-01 -2.13771626e-01
7.13515341e-01 -6.21686727e-02 -4.85997617e-01 4.72655147e-01
2.96315432e-01 5.13621271e-01 1.48240423e+00 -1.21439683e+00
-5.79949558e-01 -4.44462359e-01 -7.13623762e-01 3.49927366e-01
-6.03417039e-01 1.44004166e-01 -7.00747490e-01 -1.14491105e+00
1.22827396e-01 -4.65069443e-01 1.35981873e-01 1.60053268e-01
-5.19657493e-01 -1.99336231e-01 1.17146647e+00 -6.62539661e-01
1.07600677e+00 -2.19140601e+00 -1.28904074e-01 -1.80952236e-01
2.87030786e-01 3.45126778e-01 -1.42265037e-01 2.63246059e-01
7.36349672e-02 7.36545492e-03 -2.30922207e-01 -2.98148572e-01
-1.90112948e-01 -1.26842692e-01 -1.60147175e-01 5.19839764e-01
6.18422627e-01 9.02554035e-01 -4.27688241e-01 -9.21786606e-01
6.74465060e-01 5.24409115e-01 6.54680133e-02 -1.29015595e-01
-2.45230183e-01 1.57283157e-01 -5.16598821e-01 8.73561084e-01
8.77162576e-01 -4.89442736e-01 -1.61148831e-01 -2.36880660e-01
-6.35356456e-02 -2.87614137e-01 -1.16919482e+00 1.14162290e+00
-1.90361544e-01 1.05748618e+00 3.03498566e-01 -1.09664726e+00
1.06309736e+00 -2.62868758e-02 4.12901402e-01 -9.69141960e-01
3.89324009e-01 -4.28816825e-02 -2.88853288e-01 -7.15964913e-01
4.93934512e-01 4.55804057e-02 1.65449142e-01 1.20643616e-01
-1.56845331e-01 7.62662515e-02 7.34693483e-02 8.17385223e-03
7.60449708e-01 -7.32144713e-02 1.36989132e-01 -3.18986893e-01
8.44848275e-01 -8.76547247e-02 2.28181735e-01 8.43702912e-01
-1.88555107e-01 6.96873605e-01 3.16973954e-01 -5.02606690e-01
-8.15255344e-01 -8.65922213e-01 -3.93782377e-01 1.15579474e+00
3.15497577e-01 -2.37106532e-01 -5.30879557e-01 -8.24059248e-01
-2.24169865e-01 3.83228213e-01 -8.13807368e-01 1.66631758e-01
-5.67396998e-01 -1.00779426e+00 4.89344299e-01 5.43177307e-01
1.00550067e+00 -1.14329660e+00 -3.68172944e-01 4.13724370e-02
-5.01760483e-01 -1.30798805e+00 -5.38776100e-01 6.29506946e-01
-7.61975050e-01 -8.13175976e-01 -1.00568235e+00 -1.05957377e+00
6.84202671e-01 3.50512177e-01 8.30529213e-01 5.26974872e-02
-5.81254363e-01 2.15262137e-02 -3.47202986e-01 -2.41757572e-01
-3.42608482e-01 -2.16061205e-01 -4.33351547e-01 5.10379553e-01
3.65183115e-01 2.39098817e-01 -4.91693139e-01 4.64229196e-01
-1.15717220e+00 2.21395582e-01 1.10145545e+00 1.18789494e+00
6.63144112e-01 5.21845222e-01 3.77036750e-01 -3.49920183e-01
2.80211598e-01 9.66043845e-02 -4.28647459e-01 2.31980205e-01
-3.01055372e-01 -1.89883575e-01 1.02636409e+00 -5.08219600e-01
-1.19190264e+00 2.64838099e-01 -6.05537705e-02 -2.94052780e-01
-4.44689214e-01 2.39746988e-01 -2.54487216e-01 2.32735146e-02
5.93300104e-01 8.97724569e-01 -2.78935879e-01 -2.63428718e-01
-1.15763642e-01 9.20105994e-01 3.91865969e-01 -3.52378458e-01
5.97704113e-01 7.42628217e-01 1.01032052e-02 -1.39064145e+00
-8.81764054e-01 -6.94517136e-01 -8.14930379e-01 -2.65633970e-01
1.09914565e+00 -9.11076009e-01 -3.32627118e-01 9.48153973e-01
-9.09344673e-01 -4.30465251e-01 2.22621616e-02 3.11758310e-01
-2.24723130e-01 8.66231441e-01 -7.11962521e-01 -8.56195807e-01
-4.82029796e-01 -1.07547975e+00 1.62280834e+00 1.81669414e-01
4.64560777e-01 -9.90882874e-01 -4.28618282e-01 5.39093733e-01
3.58125538e-01 2.52325267e-01 7.07329452e-01 -6.68534696e-01
-6.02187872e-01 -2.37009957e-01 -5.75748980e-01 5.07655859e-01
4.59237508e-02 2.07801275e-02 -1.13137376e+00 -3.51562083e-01
7.40627125e-02 -2.50979662e-01 1.20677257e+00 5.48320711e-01
1.16037405e+00 -1.58870425e-02 -4.32069659e-01 4.80975747e-01
1.42865360e+00 1.04523385e-02 6.04037285e-01 4.93067950e-01
7.31408119e-01 3.85047138e-01 6.83707595e-01 3.37829739e-01
-8.63022432e-02 5.49580812e-01 4.16772634e-01 -6.32458329e-01
-1.57737195e-01 1.98379591e-01 4.23387915e-01 4.00185615e-01
3.65034193e-01 -6.85232580e-01 -8.84475648e-01 4.52756315e-01
-1.83228028e+00 -9.14038181e-01 -4.65339839e-01 1.90604496e+00
6.27612233e-01 2.54075378e-01 -8.75183195e-02 2.21874982e-01
1.18216693e+00 2.98604846e-01 -5.71852982e-01 1.80534683e-02
-6.57322586e-01 1.23303577e-01 3.54957670e-01 -1.58083402e-02
-1.57427502e+00 1.07563341e+00 5.75491810e+00 1.25012016e+00
-1.31765509e+00 -1.80937678e-01 7.75630713e-01 1.37923807e-01
4.45947677e-01 -6.83486223e-01 -8.51377249e-01 5.17989039e-01
4.32821929e-01 2.17771739e-01 1.67530533e-02 7.69835889e-01
2.42333770e-01 -3.94439638e-01 -6.34956539e-01 1.20373881e+00
5.28123736e-01 -1.16528285e+00 2.72583604e-01 6.87608942e-02
7.85006106e-01 7.59878159e-02 1.31047949e-01 1.98262006e-01
-2.73100019e-01 -1.03843915e+00 5.95279455e-01 2.56427854e-01
9.54121351e-01 -6.33280218e-01 9.56624866e-01 4.82716411e-01
-1.39396358e+00 -1.74185470e-01 -3.72219592e-01 3.51773679e-01
-3.26306731e-01 7.58123755e-01 -6.13687098e-01 6.08397007e-01
8.58317912e-01 9.78098333e-01 -8.13938677e-01 8.63407671e-01
4.05025929e-02 4.31600928e-01 -2.86939889e-01 -3.14094335e-01
3.54185998e-01 -9.86337885e-02 4.67329979e-01 1.74002182e+00
7.19381794e-02 -3.10358644e-01 3.04683656e-01 9.65490103e-01
-1.18870974e-01 5.76897562e-01 -4.52312887e-01 -3.74637283e-02
-8.56760889e-02 1.55141473e+00 -1.41948819e+00 -5.90756953e-01
-5.29518664e-01 1.29970086e+00 -1.05996370e-01 2.88948774e-01
-7.71692991e-01 -6.21580899e-01 -4.83962707e-02 -1.07238859e-01
6.19976938e-01 -4.29609744e-03 -4.90685791e-01 -1.51818311e+00
3.26736182e-01 -1.00684917e+00 4.44755882e-01 -9.17489290e-01
-1.18972445e+00 5.75786114e-01 -4.18003678e-01 -1.30318880e+00
3.41647923e-01 -9.36926842e-01 -6.66798770e-01 9.89606500e-01
-1.52549648e+00 -1.29659832e+00 -6.79043293e-01 9.23950016e-01
1.05942249e+00 -1.85226083e-01 3.96720409e-01 1.31797999e-01
-8.99477541e-01 5.85880220e-01 4.48715866e-01 5.13721943e-01
9.12287414e-01 -1.19983172e+00 7.38521367e-02 1.08917439e+00
-3.11774630e-02 2.28467390e-01 3.54501724e-01 -7.45710373e-01
-1.28039289e+00 -1.27566743e+00 4.23703820e-01 -3.39242965e-01
5.40736377e-01 -4.85272974e-01 -9.81527209e-01 2.66548038e-01
1.34992853e-01 -3.96447666e-02 2.57471293e-01 -3.65350097e-01
-1.12527788e-01 -6.77078739e-02 -1.02239263e+00 5.58374643e-01
4.38907802e-01 -4.43920135e-01 -6.07320487e-01 3.95258665e-01
1.82188362e-01 -2.07486659e-01 -4.36233461e-01 3.96072388e-01
3.90305072e-01 -1.09906638e+00 7.61306226e-01 1.56801477e-01
4.42797184e-01 -3.56973678e-01 -6.76491633e-02 -7.02046633e-01
-3.69477957e-01 -2.65168339e-01 2.15819422e-02 1.09726453e+00
4.55014482e-02 -4.93583918e-01 8.59968722e-01 -3.64364117e-01
-8.97238478e-02 -5.11474967e-01 -8.79755676e-01 -6.19410217e-01
3.73329744e-02 -1.77162394e-01 3.67902853e-02 1.02968252e+00
-3.66048545e-01 5.41351020e-01 -3.54579091e-01 2.01939464e-01
5.15055120e-01 2.38854796e-01 5.64818561e-01 -1.34822083e+00
-2.58782282e-02 -7.61132538e-01 -3.85170579e-01 -1.12527227e+00
1.45596918e-04 -8.00350130e-01 2.29243338e-01 -1.58122194e+00
4.21554655e-01 1.59411043e-01 -2.20224604e-01 4.35710251e-01
-4.62723196e-01 3.71987909e-01 3.19949500e-02 2.00042352e-01
-8.23344111e-01 6.38838768e-01 1.44212568e+00 -5.53294659e-01
-6.63183853e-02 -3.25931869e-02 -3.23123246e-01 6.73441887e-01
7.47998834e-01 -2.88392127e-01 1.10514350e-01 -2.40118593e-01
-9.88202021e-02 -2.02951849e-01 4.24229115e-01 -1.06291652e+00
2.35995278e-01 1.73439626e-02 1.21143830e+00 -1.25285184e+00
2.21769005e-01 -9.13610697e-01 -4.08551276e-01 5.05275965e-01
-1.72692642e-01 -2.52420962e-01 3.93730134e-01 5.88525593e-01
-3.25253755e-01 -1.96141005e-01 1.10171473e+00 -2.92613413e-02
-6.26142502e-01 -1.80514604e-02 -6.93514109e-01 -1.76582083e-01
8.56367052e-01 -3.95185441e-01 -5.19111931e-01 -1.43909991e-01
-5.34514844e-01 4.78188731e-02 4.85458732e-01 4.17206734e-01
7.16094136e-01 -8.90893459e-01 -6.93783164e-01 5.53386390e-01
2.68555224e-01 -8.61563459e-02 3.18031043e-01 1.09796000e+00
-4.53647345e-01 3.66459340e-01 -1.66576847e-01 -1.20145762e+00
-1.45735824e+00 5.76011837e-01 5.22465587e-01 -3.46940190e-01
-9.72774088e-01 4.67533439e-01 5.82606494e-01 -2.26483177e-02
3.80593181e-01 -4.31209028e-01 -2.69105494e-01 1.52376428e-01
7.08060205e-01 1.79662764e-01 2.67352521e-01 -7.91425645e-01
-3.08706164e-01 9.30628896e-01 -2.14923546e-01 2.73370326e-01
1.06664526e+00 -2.33846411e-01 -9.80309099e-02 3.40598762e-01
1.01750231e+00 1.59138873e-01 -1.22104037e+00 -4.00191277e-01
-2.32200488e-01 -4.02541012e-01 4.08155680e-01 -1.01308048e+00
-9.91446972e-01 1.14353192e+00 7.88138628e-01 3.76758665e-01
1.36781108e+00 -4.47055288e-02 5.82005322e-01 5.33430755e-01
-1.19116381e-01 -1.30241179e+00 6.56728446e-01 5.44516325e-01
7.64125764e-01 -1.55651367e+00 -2.35742354e-03 -1.83518440e-01
-6.73317254e-01 1.58576524e+00 6.84571147e-01 1.42244548e-01
2.65359908e-01 4.51164335e-01 1.45594776e-01 -2.58671314e-01
-5.29121041e-01 -5.73859513e-01 4.54801023e-01 5.16802788e-01
3.59382391e-01 -2.16528475e-01 6.82868063e-02 4.66976732e-01
3.75071794e-01 -4.40045625e-01 3.46830547e-01 7.92207420e-01
-8.33561778e-01 -5.57765126e-01 -7.75204301e-01 6.65623665e-01
-7.68823743e-01 -2.50109345e-01 -6.95434749e-01 9.29646730e-01
1.34203732e-01 1.02860522e+00 1.25900134e-01 -2.28108257e-01
6.01801202e-02 5.98144606e-02 2.26044968e-01 -3.78992975e-01
-5.63168108e-01 7.67751396e-01 -2.40960360e-01 -2.54919976e-01
-2.70934641e-01 -5.79405367e-01 -1.06593573e+00 1.94433611e-02
-7.93617964e-01 -1.79548964e-01 6.20189607e-01 8.52625728e-01
4.58070599e-02 6.94076002e-01 7.21472740e-01 -1.00234318e+00
-1.65516704e-01 -1.00621557e+00 -6.48014188e-01 3.06173384e-01
5.87125123e-01 -4.70251083e-01 -5.69250762e-01 3.34511578e-01] | [12.072467803955078, 2.2720956802368164] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.