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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
c3a5b1ab-499f-4ebf-8a47-89fbb08d35a9 | fine-grained-action-analysis-a-multi-modality | 2307.02730 | null | https://arxiv.org/abs/2307.02730v1 | https://arxiv.org/pdf/2307.02730v1.pdf | Fine-grained Action Analysis: A Multi-modality and Multi-task Dataset of Figure Skating | The fine-grained action analysis of the existing action datasets is challenged by insufficient action categories, low fine granularities, limited modalities, and tasks. In this paper, we propose a Multi-modality and Multi-task dataset of Figure Skating (MMFS) which was collected from the World Figure Skating Championships. MMFS, which possesses action recognition and action quality assessment, captures RGB, skeleton, and is collected the score of actions from 11671 clips with 256 categories including spatial and temporal labels. The key contributions of our dataset fall into three aspects as follows. (1) Independently spatial and temporal categories are first proposed to further explore fine-grained action recognition and quality assessment. (2) MMFS first introduces the skeleton modality for complex fine-grained action quality assessment. (3) Our multi-modality and multi-task dataset encourage more action analysis models. To benchmark our dataset, we adopt RGB-based and skeleton-based baseline methods for action recognition and action quality assessment. | ['Gui-Hong Lao', 'Hao liu', 'Ning Zhou', 'Wen-Yue Chen', 'Si-Fan Zhang', 'Yu-Ning Ding', 'Sheng-Lan Liu'] | 2023-07-06 | null | null | null | null | ['action-quality-assessment', 'action-analysis', 'action-recognition-in-videos', 'fine-grained-action-recognition'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 3.58222961e-01 -4.05199081e-01 -5.30981898e-01 -1.77923083e-01
-9.75451171e-01 -3.78289312e-01 5.12768865e-01 -3.23180735e-01
-3.76249433e-01 5.04737914e-01 1.01077521e+00 4.74801093e-01
-3.13586533e-01 -6.42701149e-01 -4.54653561e-01 -5.95708072e-01
-2.73142140e-02 -2.43664607e-01 7.01741219e-01 -2.60623962e-01
3.19109946e-01 4.16595072e-01 -1.87782049e+00 1.00662458e+00
5.38014352e-01 1.49924958e+00 -2.66545415e-01 1.03932071e+00
2.00268254e-01 1.35210109e+00 -4.58904445e-01 -2.24049717e-01
4.24402058e-01 -7.42358863e-01 -1.12205374e+00 2.35760450e-01
8.05001080e-01 -5.76689541e-01 -5.19304216e-01 6.30189657e-01
6.02729559e-01 5.34995496e-01 5.26097178e-01 -1.61983907e+00
-5.64569056e-01 3.26102614e-01 -5.14525831e-01 2.30676457e-01
7.64597118e-01 6.39058411e-01 8.62076044e-01 -5.09376764e-01
4.91889566e-01 1.38955212e+00 6.90045774e-01 6.04964256e-01
-5.30579209e-01 -4.77488190e-01 1.86195895e-01 7.47498035e-01
-1.03840864e+00 -5.32476485e-01 6.26267254e-01 -3.17126662e-01
9.07724261e-01 1.92678601e-01 9.12084460e-01 1.19471920e+00
1.45339444e-01 1.31907165e+00 1.37174869e+00 -1.23816282e-01
2.93646723e-01 -9.67963099e-01 -2.78858036e-01 6.29301608e-01
-4.72191393e-01 1.21343091e-01 -1.11015546e+00 2.55339712e-01
9.06663477e-01 1.57614261e-01 2.04570726e-01 -2.11409360e-01
-1.68620872e+00 2.33751759e-01 3.50173652e-01 2.64307588e-01
-4.56312865e-01 6.65643334e-01 6.88316524e-01 1.53884888e-01
1.69826776e-01 7.93172140e-03 -5.40601015e-01 -1.07410038e+00
-8.09923291e-01 3.54041636e-01 8.99553970e-02 5.99380851e-01
5.33175945e-01 -5.17954715e-02 -6.54119134e-01 8.60467494e-01
1.30309686e-01 8.19881678e-01 3.81310701e-01 -1.61413455e+00
7.18898535e-01 8.21993589e-01 3.10411565e-02 -9.70282674e-01
-6.72660649e-01 4.29893970e-01 -6.38672531e-01 5.22198856e-01
6.57664120e-01 1.83125496e-01 -9.20494556e-01 1.38855493e+00
4.72662598e-01 9.10043493e-02 -2.29645148e-02 1.12406695e+00
1.05679047e+00 9.94817540e-02 4.55674738e-01 1.60627559e-01
1.39556181e+00 -1.37296224e+00 -7.90880561e-01 9.39094722e-02
7.75139689e-01 -4.67572391e-01 1.51563835e+00 4.74142849e-01
-1.15558171e+00 -9.77148235e-01 -8.10852349e-01 -2.61450350e-01
-4.68831450e-01 4.37487423e-01 7.63963223e-01 4.23689187e-01
-6.73675835e-01 6.66506290e-01 -1.01683438e+00 -3.48830998e-01
7.88523555e-01 -1.69118673e-01 -8.13964605e-01 -1.84441477e-01
-1.20562840e+00 6.90390766e-01 3.45168889e-01 -3.45166065e-02
-1.06024981e+00 -4.95617062e-01 -9.01272476e-01 -4.77785826e-01
6.76502168e-01 -4.48136210e-01 1.17328870e+00 -8.95734310e-01
-1.72857618e+00 8.58723521e-01 1.75203368e-01 -1.77823335e-01
6.08018219e-01 -3.25547606e-01 -5.81540227e-01 6.02967381e-01
3.13434333e-01 6.57703757e-01 6.28394544e-01 -5.25851250e-01
-1.06234670e+00 -5.80889165e-01 5.85622013e-01 3.27555716e-01
-1.16934530e-01 5.89853041e-02 -5.28884768e-01 -8.81432772e-01
1.70296505e-01 -8.83158982e-01 7.53253847e-02 3.79363626e-01
-2.73144126e-01 -3.05847228e-01 6.23211861e-01 -4.70284671e-01
1.27372801e+00 -2.24551940e+00 3.03032041e-01 -4.73580033e-01
3.68957520e-02 4.32717651e-02 -2.23509803e-01 2.78981537e-01
4.10659313e-02 -9.05007944e-02 1.08947195e-01 -2.35006556e-01
2.20111057e-01 3.04263383e-01 2.53668308e-01 4.51016694e-01
5.81838936e-03 1.26616561e+00 -9.44548666e-01 -9.85235333e-01
5.79698205e-01 5.07098585e-02 -2.55344361e-01 -1.17398299e-01
-7.26469457e-02 4.75941867e-01 -7.36156225e-01 1.40515172e+00
3.40198666e-01 1.43541358e-02 -4.20805275e-01 -6.86562061e-01
3.04646976e-02 -2.14305297e-01 -1.55832720e+00 2.43404102e+00
-9.60898399e-02 1.54736549e-01 -1.75304934e-01 -5.48427403e-01
3.74004364e-01 4.23542224e-02 1.15617251e+00 -1.29051304e+00
-2.00386951e-03 -5.86085506e-02 -4.76434141e-01 -8.90605569e-01
5.59983134e-01 -7.24971592e-02 -6.58179104e-01 6.26880467e-01
5.35242334e-02 6.04285635e-02 4.60924029e-01 2.92840421e-01
1.54284215e+00 9.75659251e-01 3.55540216e-01 1.72728479e-01
5.17825246e-01 5.28467856e-02 6.46271288e-01 4.91401672e-01
-1.06108701e+00 6.60961926e-01 3.74759525e-01 -7.14108050e-01
-6.66652441e-01 -1.07788682e+00 1.47438586e-01 1.65059936e+00
2.92578816e-01 -6.14795625e-01 -5.10539174e-01 -9.02802706e-01
9.93094519e-02 -2.09220171e-01 -1.10902643e+00 -1.76925004e-01
-4.85126197e-01 -3.16721559e-01 1.03224289e+00 1.15702605e+00
1.19651806e+00 -1.32835269e+00 -7.70866394e-01 -7.13374317e-02
-5.96564054e-01 -1.15683627e+00 -4.83638674e-01 -2.49529153e-01
-7.72332668e-01 -1.58225501e+00 -8.67678702e-01 3.25719453e-02
-1.54603412e-02 2.07975879e-01 8.77172887e-01 -4.14168030e-01
-2.74333090e-01 8.84814978e-01 -9.60220635e-01 -1.17289469e-01
8.02707225e-02 -2.72766352e-01 2.15373278e-01 1.31263003e-01
3.01670313e-01 -3.73364329e-01 -8.30095232e-01 6.57208562e-01
-9.34786797e-01 8.87578577e-02 9.20045614e-01 3.32495362e-01
9.09410417e-01 5.08958139e-02 2.72712082e-01 -1.60802454e-01
1.98766753e-01 -4.36118357e-02 1.73220545e-01 4.55699503e-01
-1.03469677e-01 -1.31260291e-01 1.65858582e-01 -2.56120771e-01
-1.13854766e+00 1.68566853e-01 4.30835038e-02 -2.86023289e-01
-6.03355408e-01 4.46734913e-02 -3.56312126e-01 -1.59954488e-01
9.03673530e-01 1.36314705e-01 -1.82250693e-01 -5.76067746e-01
5.44824421e-01 5.14697134e-01 8.56335640e-01 -8.63678873e-01
3.38742644e-01 7.83622801e-01 2.18210936e-01 -3.43512952e-01
-1.01614463e+00 -6.59185231e-01 -1.27602518e+00 -1.05441380e+00
1.36399937e+00 -9.97989416e-01 -1.03487146e+00 1.13251209e+00
-5.03864706e-01 -7.20205545e-01 -5.12475431e-01 6.89579010e-01
-1.06283021e+00 5.50384760e-01 -6.49675548e-01 -7.16076076e-01
7.48598129e-02 -7.62297928e-01 1.64894938e+00 1.49036407e-01
-5.62260933e-02 -6.43736362e-01 1.76869825e-01 1.08508313e+00
2.26639763e-01 8.76244664e-01 3.76796722e-02 2.94309199e-01
-3.00998896e-01 -8.42115432e-02 -2.39381358e-01 3.18102956e-01
2.41193026e-01 -3.63575816e-02 -7.46013165e-01 2.25205228e-01
-7.26592124e-01 -1.12843895e+00 9.84857082e-01 4.09969479e-01
1.23246658e+00 1.57302260e-01 8.89692977e-02 5.19994497e-01
9.23174322e-01 -8.57036263e-02 9.57701862e-01 6.20168865e-01
9.52926040e-01 4.14192200e-01 1.32441533e+00 6.83791161e-01
5.93623519e-01 8.48137677e-01 3.76172304e-01 -5.07354960e-02
-6.07503533e-01 -2.10677743e-01 5.97294927e-01 3.97190839e-01
-9.80263352e-01 2.06073761e-01 -7.12007642e-01 3.97051454e-01
-2.18803310e+00 -1.43428254e+00 -2.27478325e-01 1.55654156e+00
7.14026511e-01 -6.46444038e-02 8.77935529e-01 5.06318748e-01
2.60303915e-01 4.38082516e-01 -7.49037445e-01 8.96156803e-02
-3.17302108e-01 2.98846513e-02 5.44658601e-01 -1.63187459e-01
-1.57449973e+00 6.88203990e-01 6.31553030e+00 1.14125168e+00
-5.54996848e-01 1.84107721e-01 1.55309841e-01 -5.84769666e-01
2.53681481e-01 -3.57018828e-01 -5.85152030e-01 4.68872488e-01
6.63894355e-01 1.09374680e-01 3.34079154e-02 8.62920702e-01
3.81953746e-01 -4.96098906e-01 -8.56316090e-01 1.22437143e+00
2.51776308e-01 -1.10256135e+00 -2.58037299e-01 -1.30725369e-01
7.03571677e-01 -1.68481544e-01 -2.04331487e-01 5.71367681e-01
2.82856464e-01 -8.99496138e-01 9.89163697e-01 1.00338399e+00
9.78226185e-01 -3.93885583e-01 4.10755873e-01 6.48253085e-03
-1.72200727e+00 -3.13645571e-01 6.51384667e-02 -3.09034914e-01
3.04504305e-01 1.60157040e-01 4.69732106e-01 7.94296563e-01
1.11674547e+00 1.43596900e+00 -9.47804451e-01 7.76495516e-01
-1.48691952e-01 2.11273640e-01 -3.77529934e-02 2.15288982e-01
3.93436551e-01 2.10789397e-01 -6.18373267e-02 9.55419362e-01
-2.28318125e-02 3.93851340e-01 4.73316222e-01 9.02775377e-02
3.95907760e-01 -4.57442701e-02 -1.03577919e-01 -1.61247998e-01
-1.32015362e-01 1.24368060e+00 -7.98703372e-01 -2.72152096e-01
-4.84693646e-01 1.11102128e+00 -6.86660260e-02 2.01512277e-01
-1.09272993e+00 -1.82461977e-01 6.87138557e-01 1.18648140e-02
5.20863719e-02 -2.12266013e-01 -2.44669721e-01 -1.21495914e+00
1.12494394e-01 -9.44455624e-01 9.18228507e-01 -1.00004673e+00
-1.02565455e+00 3.88979316e-02 1.89054281e-01 -1.69280112e+00
1.36338815e-01 -6.59521937e-01 -2.36535385e-01 2.23283648e-01
-1.25807846e+00 -1.71493661e+00 -8.84250581e-01 1.25204933e+00
7.06988692e-01 -5.65117560e-02 6.78134084e-01 5.95823705e-01
-4.64999020e-01 5.37481070e-01 -4.17035431e-01 2.50235230e-01
8.24269593e-01 -1.20228410e+00 -7.18401792e-03 6.76510692e-01
-3.11956853e-02 -1.09838583e-01 9.54932347e-02 -7.30552435e-01
-1.40790153e+00 -9.55618024e-01 3.67610544e-01 -7.94924438e-01
6.13691866e-01 1.57920390e-01 -3.24431032e-01 3.46406341e-01
-4.58121777e-01 4.82257932e-01 7.87516057e-01 3.92774306e-02
-2.66106844e-01 -2.34878004e-01 -1.14543247e+00 2.26585582e-01
1.64440441e+00 -6.04729772e-01 -4.67274159e-01 2.93769985e-01
2.24385470e-01 -3.63836378e-01 -1.40440476e+00 4.99242693e-01
1.13623440e+00 -1.23538625e+00 1.03769755e+00 -8.20628583e-01
8.45853031e-01 -4.77434486e-01 -5.22747695e-01 -8.39793265e-01
-4.81697321e-01 9.36684757e-03 -4.08253163e-01 9.16878521e-01
-1.81906238e-01 1.77682459e-01 9.86680448e-01 5.84422410e-01
-2.50589043e-01 -7.48628736e-01 -1.04026508e+00 -8.44059050e-01
-2.31635228e-01 -9.78116751e-01 4.74731714e-01 6.21439278e-01
7.51561448e-02 -2.43106723e-01 -7.51112044e-01 -4.11541075e-01
5.65529346e-01 2.37268224e-01 1.03208005e+00 -8.76710415e-01
-3.62596750e-01 -3.69037718e-01 -9.50797379e-01 -9.25507784e-01
-1.45686820e-01 -2.82394528e-01 -6.87846020e-02 -1.81864011e+00
2.96109885e-01 -1.20671675e-01 -4.78057534e-01 1.00345886e+00
-4.03187945e-02 7.88588285e-01 3.24096590e-01 2.70603895e-01
-1.56085682e+00 6.88329041e-01 1.71276462e+00 -1.25716731e-01
1.62620157e-01 -3.57607827e-02 -3.12742680e-01 8.34093451e-01
5.07277668e-01 -1.26871049e-01 -3.63428742e-01 -2.63618380e-01
2.11564302e-01 2.36871645e-01 6.97166204e-01 -1.43807411e+00
8.96077156e-02 -7.12460518e-01 5.91833174e-01 -7.27649629e-01
5.10776222e-01 -5.25314927e-01 5.94690181e-02 3.66550744e-01
-4.00908828e-01 -3.83206159e-01 -1.14244729e-01 7.03740954e-01
-3.20341378e-01 5.06195307e-01 7.72727013e-01 -4.43595022e-01
-1.58074951e+00 5.52308559e-01 -2.21286625e-01 2.44907156e-01
1.37294245e+00 -7.05764413e-01 -3.26251954e-01 -1.72246113e-01
-1.08788896e+00 2.39123493e-01 3.68537813e-01 7.61746407e-01
5.61956465e-01 -2.07273316e+00 -5.85528672e-01 -8.72294083e-02
6.68343723e-01 -3.87945950e-01 1.00028050e+00 1.24176311e+00
-3.84811938e-01 3.20263594e-01 -9.47340250e-01 -3.58832717e-01
-1.41606021e+00 2.18962222e-01 2.43439659e-01 -1.43393874e-01
-6.11348927e-01 6.56791389e-01 -2.04829156e-01 -3.11702400e-01
2.62370378e-01 -4.69968766e-01 -4.46991205e-01 1.52721569e-01
8.43604803e-01 1.13558674e+00 -2.55605966e-01 -1.03261089e+00
-6.92966759e-01 9.07155871e-01 6.48289263e-01 -1.45051349e-02
1.15839863e+00 -2.43968129e-01 3.50470752e-01 5.82676053e-01
9.61691558e-01 -3.83166373e-01 -1.59439039e+00 -1.09283544e-01
-3.25953245e-01 -8.70568573e-01 -6.24704175e-02 -1.04266369e+00
-1.09258401e+00 8.14973116e-01 6.70072198e-01 -1.36404797e-01
1.52879989e+00 5.43503538e-02 1.02043283e+00 1.75831735e-01
6.13508344e-01 -1.83729339e+00 8.16467464e-01 3.32432389e-01
8.81974399e-01 -1.42550266e+00 2.69441754e-01 -3.70363481e-02
-1.07746160e+00 1.11052179e+00 8.73035789e-01 1.27410680e-01
3.63869876e-01 4.65198234e-02 5.96636757e-02 -4.33695406e-01
-3.88421148e-01 -6.60671830e-01 4.71203178e-01 7.38475442e-01
4.89786044e-02 7.14621469e-02 -2.03579605e-01 7.45430589e-01
2.15886027e-01 5.32639742e-01 1.40617698e-01 1.20526314e+00
-5.22804320e-01 -8.14828753e-01 -3.03347677e-01 4.19882119e-01
-4.69084918e-01 4.77935880e-01 -5.46693027e-01 5.23908079e-01
5.80599487e-01 1.05239975e+00 -3.90690535e-01 -7.69353092e-01
8.49318862e-01 -1.53631926e-01 6.97675943e-01 -2.23734498e-01
-4.46167678e-01 -2.80248642e-01 2.28783891e-01 -1.69726467e+00
-1.11860740e+00 -7.73308992e-01 -1.17388082e+00 -3.65223080e-01
1.89873859e-01 -4.49283987e-01 4.07579154e-01 1.08911681e+00
3.35492224e-01 6.36142552e-01 4.21966642e-01 -1.05111921e+00
-1.99626312e-01 -9.78477120e-01 -1.04338610e+00 8.67014468e-01
4.91719060e-02 -1.18458521e+00 -2.26781726e-01 3.88070047e-01] | [7.906041145324707, 0.4084053039550781] |
59df01f0-601a-456c-9086-0debf34c3baf | local-implicit-ray-function-for-generalizable | 2304.12746 | null | https://arxiv.org/abs/2304.12746v1 | https://arxiv.org/pdf/2304.12746v1.pdf | Local Implicit Ray Function for Generalizable Radiance Field Representation | We propose LIRF (Local Implicit Ray Function), a generalizable neural rendering approach for novel view rendering. Current generalizable neural radiance fields (NeRF) methods sample a scene with a single ray per pixel and may therefore render blurred or aliased views when the input views and rendered views capture scene content with different resolutions. To solve this problem, we propose LIRF to aggregate the information from conical frustums to construct a ray. Given 3D positions within conical frustums, LIRF takes 3D coordinates and the features of conical frustums as inputs and predicts a local volumetric radiance field. Since the coordinates are continuous, LIRF renders high-quality novel views at a continuously-valued scale via volume rendering. Besides, we predict the visible weights for each input view via transformer-based feature matching to improve the performance in occluded areas. Experimental results on real-world scenes validate that our method outperforms state-of-the-art methods on novel view rendering of unseen scenes at arbitrary scales. | ['Qing Wang', 'Xuan Wang', 'Xiaoyu Li', 'Ying Feng', 'Qi Zhang', 'Xin Huang'] | 2023-04-25 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Huang_Local_Implicit_Ray_Function_for_Generalizable_Radiance_Field_Representation_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Huang_Local_Implicit_Ray_Function_for_Generalizable_Radiance_Field_Representation_CVPR_2023_paper.pdf | cvpr-2023-1 | ['neural-rendering'] | ['computer-vision'] | [ 3.08036000e-01 -2.22455248e-01 4.14969593e-01 -4.76354808e-01
-6.12945855e-01 -6.14919245e-01 5.80740452e-01 -4.71958816e-01
1.63580582e-01 4.54100460e-01 1.44673914e-01 -2.78003305e-01
-6.83154613e-02 -1.44330239e+00 -9.61726427e-01 -5.60681045e-01
2.72373319e-01 2.27892801e-01 1.93514094e-01 -2.84937263e-01
1.42190427e-01 9.91224289e-01 -1.70191145e+00 8.64709198e-01
7.98576653e-01 1.15252256e+00 3.81476343e-01 1.02749908e+00
-3.67433190e-01 8.47292900e-01 -5.11646152e-01 -3.53801310e-01
7.04081535e-01 -3.86132598e-02 -3.63918453e-01 -3.66800316e-02
1.22851670e+00 -9.74837899e-01 -5.93381405e-01 8.42615545e-01
3.96899700e-01 4.57746923e-01 5.22504389e-01 -6.75724030e-01
-1.03272033e+00 -6.59608543e-02 -6.68890238e-01 1.21768944e-01
4.73093301e-01 -9.46100131e-02 7.85029173e-01 -1.22085857e+00
6.20851219e-01 1.42769444e+00 6.52154922e-01 4.09142494e-01
-1.21948981e+00 -4.96929526e-01 5.71440518e-01 -4.51580673e-01
-9.81926918e-01 -1.32887274e-01 1.12608230e+00 -3.35484952e-01
9.21867907e-01 8.88162494e-01 8.69113088e-01 7.08476961e-01
4.11203891e-01 3.73245686e-01 1.09347367e+00 1.04079053e-01
1.48903355e-01 1.37732672e-02 -2.11495176e-01 8.69818807e-01
-2.53294110e-01 4.41165507e-01 -5.16876340e-01 -4.27884609e-01
1.66605330e+00 4.59630758e-01 -8.12418997e-01 -3.72213364e-01
-1.27696097e+00 7.30221391e-01 1.06847727e+00 -2.41095796e-01
-2.73679942e-01 2.92417735e-01 -4.18721810e-02 1.42076999e-01
1.16343105e+00 3.43338817e-01 -5.19855917e-01 4.22882050e-01
-5.65392554e-01 3.04172099e-01 2.94022113e-01 8.98229718e-01
7.81444013e-01 3.42119426e-01 -1.43220976e-01 8.65190744e-01
1.76305518e-01 6.83851659e-01 -5.33681288e-02 -1.29980552e+00
5.59682846e-01 5.04321575e-01 1.82104915e-01 -1.00355542e+00
-2.56183118e-01 -6.16996765e-01 -8.77199471e-01 7.67277658e-01
-6.89814538e-02 2.46994823e-01 -1.00431931e+00 1.12585831e+00
5.50433218e-01 3.81974816e-01 -1.08516766e-02 1.11023033e+00
1.38942695e+00 1.15302706e+00 -5.90413213e-01 5.03507406e-02
1.05800307e+00 -9.31440353e-01 -2.61467159e-01 5.96465133e-02
2.53295135e-02 -7.40831375e-01 1.32536530e+00 5.51657796e-01
-1.39513218e+00 -7.91773438e-01 -7.42122293e-01 -5.29934466e-01
-1.73553646e-01 5.28398119e-02 8.75520706e-01 3.73813987e-01
-1.16195655e+00 5.00083685e-01 -4.71406758e-01 4.16721195e-01
3.60024750e-01 2.32854802e-02 -1.28900027e-02 -2.76922047e-01
-6.02347374e-01 1.33744493e-01 -3.48623961e-01 3.42889756e-01
-7.45034158e-01 -1.28644013e+00 -6.94307208e-01 2.03795433e-01
8.23989064e-02 -1.14045668e+00 9.12974715e-01 -7.49256074e-01
-1.34116888e+00 6.66670442e-01 -2.01666653e-01 1.60551623e-01
6.09063864e-01 -2.35984460e-01 -2.51803905e-01 1.83764085e-01
-2.21981704e-01 3.45274925e-01 8.68662834e-01 -1.89374900e+00
-5.14783919e-01 -4.69843835e-01 6.28467441e-01 5.99245191e-01
3.32249641e-01 -4.75420833e-01 -1.50755301e-01 -6.90956831e-01
6.85035467e-01 -3.42332482e-01 -2.60652989e-01 6.83189988e-01
-2.23534003e-01 2.12885484e-01 8.21314752e-01 -6.19019568e-01
4.56423581e-01 -2.05845618e+00 -8.30788352e-03 9.11713392e-02
7.39859641e-01 -3.19770753e-01 -1.55657232e-01 -9.33709666e-02
-3.76781672e-02 -1.49765313e-01 -5.98088950e-02 -1.70612112e-01
-4.78710532e-01 1.42235225e-02 -7.95882285e-01 3.22833121e-01
-2.89455861e-01 8.13787222e-01 -8.87742579e-01 -6.91596139e-03
6.87050819e-01 1.20927918e+00 -8.08566630e-01 3.50143790e-01
-3.06454509e-01 7.04791427e-01 -6.49713755e-01 4.91558880e-01
1.37400270e+00 -3.04739743e-01 -4.40459043e-01 -5.21601558e-01
-2.77101815e-01 1.94222316e-01 -8.17839563e-01 1.79891562e+00
-9.04421866e-01 7.30980814e-01 -1.86907515e-01 -2.58461505e-01
1.10750520e+00 -6.36277124e-02 5.90496004e-01 -9.00967538e-01
-1.29503474e-01 -2.77419895e-01 -8.25959623e-01 -2.20856562e-01
7.38437533e-01 1.52228758e-01 3.42260957e-01 1.56606704e-01
-4.14086044e-01 -5.78082383e-01 -7.41449058e-01 8.12061653e-02
7.54380286e-01 4.06332046e-01 1.64759532e-02 3.89541760e-02
3.02679300e-01 -3.73043269e-01 3.45327854e-01 6.98912859e-01
4.96829152e-01 1.33599854e+00 -1.42065465e-01 -1.14001799e+00
-1.06072736e+00 -1.85356498e+00 -3.74334484e-01 1.05323386e+00
3.42933148e-01 -1.00130230e-01 -4.77668464e-01 -3.98932964e-01
-1.73374161e-01 6.44994617e-01 -6.37124121e-01 3.71681809e-01
-8.07077408e-01 -4.92188871e-01 -2.67147541e-01 4.15039510e-01
4.22831982e-01 -8.65629613e-01 -1.04243135e+00 6.11274242e-02
-2.65621930e-01 -9.74038184e-01 -2.30138361e-01 -5.29780760e-02
-1.06718493e+00 -9.81973231e-01 -9.65922356e-01 -3.63312691e-01
8.37663949e-01 8.91890585e-01 1.58565962e+00 4.26550470e-02
-5.33495009e-01 3.88353795e-01 -2.66946480e-02 -2.04408634e-02
-1.11849681e-01 -5.36984742e-01 -5.74637175e-01 -1.10512033e-01
-3.51769149e-01 -8.44399810e-01 -1.06131423e+00 1.83214188e-01
-8.02508414e-01 7.12311208e-01 -6.59095570e-02 5.64313471e-01
1.02726412e+00 -2.30078846e-01 -2.01971978e-01 -9.58508730e-01
2.93269426e-01 -1.57741472e-01 -1.04249752e+00 4.45227951e-01
-3.07914987e-02 -7.70466477e-02 8.05688858e-01 -3.04778576e-01
-1.58037889e+00 -3.26205462e-01 -4.48308438e-02 -6.64452493e-01
-6.18263036e-02 -5.70529141e-02 1.92818865e-01 -2.48974308e-01
8.14069152e-01 3.30454469e-01 -7.83473134e-01 -5.17382503e-01
5.68559706e-01 9.07951072e-02 6.15684807e-01 -6.22156143e-01
8.75238419e-01 1.08129990e+00 1.84445307e-01 -7.43478894e-01
-9.65266526e-01 -2.10898101e-01 -5.03184378e-01 -5.83198190e-01
9.01684344e-01 -1.22039938e+00 -9.02433693e-01 1.67504176e-01
-1.44746161e+00 -3.01662147e-01 -5.42083621e-01 2.99390405e-01
-6.47906542e-01 -1.04103005e-02 -6.28485322e-01 -9.19957697e-01
-4.00673270e-01 -1.16710258e+00 1.54722190e+00 3.48901898e-01
4.86562788e-01 -8.25444520e-01 5.84911257e-02 3.16541374e-01
3.58618289e-01 5.45957863e-01 9.85789359e-01 6.88202620e-01
-1.47372139e+00 3.81876439e-01 -6.15957856e-01 1.51792750e-01
9.38895270e-02 -3.90845612e-02 -1.50706184e+00 -2.60847628e-01
3.06242913e-01 -1.37704164e-01 8.05577874e-01 7.06529260e-01
1.78347719e+00 -3.26363534e-01 -6.68689534e-02 1.57090926e+00
1.79957950e+00 2.82726437e-01 6.28975987e-01 5.39473854e-02
1.06321883e+00 4.07484263e-01 3.13092262e-01 5.92500150e-01
3.38844597e-01 4.68393266e-01 8.69343221e-01 -5.37069917e-01
-2.61340201e-01 -3.22347850e-01 -1.14100583e-01 5.80891013e-01
-7.26889312e-01 -4.48508918e-01 -2.51217455e-01 -6.55163601e-02
-1.42070830e+00 -1.02964985e+00 -3.14530730e-01 2.28027725e+00
1.58020020e-01 -1.72350556e-01 -5.03057599e-01 -3.30629200e-01
1.63491592e-01 6.73634648e-01 -8.61785293e-01 -3.91856849e-01
-2.85352975e-01 2.36942396e-01 4.61878985e-01 6.96765244e-01
-5.70834160e-01 6.40225649e-01 6.01329613e+00 3.44186932e-01
-1.30070853e+00 8.67946222e-02 8.93782020e-01 -4.79192227e-01
-1.16630435e+00 -1.56261995e-01 -5.45606315e-01 2.47842893e-02
2.11852297e-01 2.40250036e-01 1.01964748e+00 7.55284905e-01
2.75084600e-02 8.67282227e-02 -8.31712723e-01 1.24216020e+00
2.22292587e-01 -1.66918254e+00 4.48823571e-01 5.65967010e-03
1.00633252e+00 4.25755054e-01 4.92539912e-01 -1.32060319e-01
3.24970692e-01 -9.13998067e-01 7.49895513e-01 1.07759118e+00
1.03730667e+00 -7.16768146e-01 -5.48266992e-02 1.39306799e-01
-1.35844302e+00 4.76920046e-02 -9.48608935e-01 2.98717946e-01
2.41943359e-01 6.41986728e-01 -3.95729095e-01 7.10099459e-01
1.05703509e+00 6.50090575e-01 -4.52002227e-01 8.73515725e-01
1.09313339e-01 3.89599800e-03 -2.28436545e-01 4.69355792e-01
1.54169261e-01 -5.56336761e-01 5.98109484e-01 4.49685752e-01
5.61729193e-01 2.42844909e-01 3.71132642e-02 1.32925916e+00
-9.78740980e-04 -1.63966164e-01 -6.92915857e-01 7.31670916e-01
-7.43926391e-02 1.09687829e+00 -5.01542509e-01 -2.48106226e-01
-4.73914891e-01 1.19896090e+00 4.51289147e-01 1.05469346e+00
-9.08641219e-01 4.39681821e-02 6.55738831e-01 4.30728763e-01
1.51123300e-01 -1.14088550e-01 -1.98495939e-01 -1.50473964e+00
1.48561403e-01 -2.29298428e-01 4.28563682e-03 -1.73641205e+00
-1.11362982e+00 1.18658948e+00 -4.49930094e-02 -1.57138205e+00
2.82753725e-02 -7.04885423e-01 -3.83989692e-01 1.09529138e+00
-1.84433305e+00 -1.17843843e+00 -8.71226549e-01 7.21399903e-01
8.74350190e-01 9.06684250e-02 7.36568332e-01 -1.96068119e-02
1.89622074e-01 1.77623183e-01 3.89560729e-01 -1.57827079e-01
2.65832812e-01 -9.71529365e-01 8.87939572e-01 4.39656049e-01
2.32716903e-01 4.61919278e-01 3.73040497e-01 -4.02790874e-01
-1.29426217e+00 -1.18738246e+00 -1.79436635e-02 -4.65212375e-01
-1.55963674e-01 -6.77352786e-01 -1.01767242e+00 5.78506589e-01
-1.27155319e-01 5.32902360e-01 3.41430932e-01 -8.59575421e-02
-6.08108640e-01 -2.22358078e-01 -1.22620559e+00 5.76531112e-01
1.44660568e+00 -6.89936042e-01 -1.70812443e-01 4.76829171e-01
1.00818968e+00 -9.87011075e-01 -8.21339309e-01 4.65637177e-01
7.64676929e-01 -1.67007256e+00 1.71631873e+00 -2.87169784e-01
6.64694548e-01 -2.94522196e-01 -5.59463859e-01 -1.39965618e+00
-4.40095097e-01 -2.71379441e-01 -1.33768901e-01 4.97809261e-01
6.86536403e-03 -8.51320624e-01 6.90383017e-01 4.54063773e-01
-2.07995474e-01 -9.30356264e-01 -7.44079947e-01 -2.43148461e-01
-1.48416981e-01 -3.70029926e-01 1.02870619e+00 7.27233589e-01
-9.96045768e-01 8.02148283e-02 -2.04775155e-01 5.40178895e-01
8.92169654e-01 7.70468771e-01 7.28776574e-01 -1.47449303e+00
-4.56383556e-01 -1.50025815e-01 -2.97979247e-02 -1.63821578e+00
-1.14271440e-01 -8.07717860e-01 -1.97249234e-01 -1.73367631e+00
7.57731646e-02 -8.03751051e-01 -1.08410083e-01 -2.23441377e-01
-8.28924254e-02 3.07954788e-01 1.99774593e-01 1.02244906e-01
-7.55412728e-02 6.69436812e-01 1.89296961e+00 -6.79453984e-02
-2.57948369e-01 6.73704147e-02 -4.38221633e-01 8.75065446e-01
4.45143908e-01 -9.66763869e-02 -7.75937021e-01 -1.18942916e+00
5.53780854e-01 5.59550047e-01 8.17434907e-01 -9.33001280e-01
-2.13090435e-01 -3.70779306e-01 1.08694363e+00 -1.24735963e+00
1.04239142e+00 -9.54405248e-01 3.16586614e-01 -2.07920626e-01
-3.45059454e-01 2.74426132e-01 -1.08934501e-02 5.92953682e-01
3.07861298e-01 2.72153974e-01 7.32609987e-01 -5.52705109e-01
-3.03803474e-01 8.07442605e-01 2.41385281e-01 -9.85018015e-02
5.85047901e-01 -4.54496384e-01 -5.57187498e-01 -4.24347371e-01
-4.24303383e-01 -3.87184203e-01 7.06340849e-01 3.84827197e-01
1.26852667e+00 -1.16748273e+00 -6.36877775e-01 6.89664245e-01
-1.01070735e-03 7.24640191e-01 7.98583329e-01 -1.14591800e-01
-7.96830237e-01 1.08978547e-01 -8.68283808e-02 -8.10445845e-01
-1.20273221e+00 4.85773385e-01 6.24391496e-01 1.73108011e-01
-1.28253150e+00 1.05100763e+00 1.42048132e+00 -7.25217462e-01
-5.33793718e-02 -7.84028888e-01 -1.09096309e-02 -6.55356944e-01
7.20957816e-01 2.12569922e-01 -4.05165032e-02 -6.16310716e-01
2.40174219e-01 1.02066457e+00 1.80855125e-01 2.07448918e-02
1.45556939e+00 -1.14722662e-01 4.64987159e-02 6.28178179e-01
1.31006682e+00 3.43331039e-01 -1.80300057e+00 -1.49767846e-01
-1.31221032e+00 -1.19602191e+00 4.67970550e-01 -7.66613424e-01
-1.47089946e+00 1.04076052e+00 7.18918085e-01 -1.39183803e-02
1.20153379e+00 -7.86464512e-02 7.45453179e-01 2.20609203e-01
6.25605226e-01 -3.25208455e-01 -1.57381669e-01 2.84274727e-01
1.23602057e+00 -9.14765656e-01 1.72996059e-01 -8.13607812e-01
-2.08935097e-01 1.27558053e+00 6.62643313e-01 -3.38077396e-01
6.19795740e-01 2.93379128e-01 1.34367943e-01 -3.35251600e-01
-5.93251348e-01 4.81670231e-01 6.22661829e-01 6.20259345e-01
4.51966792e-01 1.40524000e-01 5.59927702e-01 -1.48451045e-01
-5.13116598e-01 -3.26423526e-01 4.15698081e-01 5.00050962e-01
-2.23541230e-01 -3.43185961e-01 -4.77840126e-01 6.18261695e-01
-1.89388040e-02 -3.48930001e-01 6.82687834e-02 4.18410838e-01
2.44704504e-02 3.83567363e-01 4.61603582e-01 -8.03993493e-02
4.01233941e-01 -7.01452971e-01 7.79629707e-01 -3.85149151e-01
-5.43952107e-01 1.60156250e-01 -3.68822575e-01 -9.08940673e-01
-3.12493116e-01 -1.27711266e-01 -1.13878989e+00 -2.24561676e-01
-1.20445579e-01 -2.51654804e-01 6.61907077e-01 2.76955485e-01
2.37296283e-01 9.05563414e-01 1.02266657e+00 -1.31188512e+00
-3.74409296e-02 -2.71611154e-01 -5.04131556e-01 2.23171607e-01
7.37677574e-01 -5.13220727e-01 -4.47222859e-01 -1.25204146e-01] | [9.518056869506836, -3.073570966720581] |
5c53dd4d-02d1-4817-b8e8-5e53965b3ead | gipso-geometrically-informed-propagation-for | 2207.09763 | null | https://arxiv.org/abs/2207.09763v1 | https://arxiv.org/pdf/2207.09763v1.pdf | GIPSO: Geometrically Informed Propagation for Online Adaptation in 3D LiDAR Segmentation | 3D point cloud semantic segmentation is fundamental for autonomous driving. Most approaches in the literature neglect an important aspect, i.e., how to deal with domain shift when handling dynamic scenes. This can significantly hinder the navigation capabilities of self-driving vehicles. This paper advances the state of the art in this research field. Our first contribution consists in analysing a new unexplored scenario in point cloud segmentation, namely Source-Free Online Unsupervised Domain Adaptation (SF-OUDA). We experimentally show that state-of-the-art methods have a rather limited ability to adapt pre-trained deep network models to unseen domains in an online manner. Our second contribution is an approach that relies on adaptive self-training and geometric-feature propagation to adapt a pre-trained source model online without requiring either source data or target labels. Our third contribution is to study SF-OUDA in a challenging setup where source data is synthetic and target data is point clouds captured in the real world. We use the recent SynLiDAR dataset as a synthetic source and introduce two new synthetic (source) datasets, which can stimulate future synthetic-to-real autonomous driving research. Our experiments show the effectiveness of our segmentation approach on thousands of real-world point clouds. Code and synthetic datasets are available at https://github.com/saltoricristiano/gipso-sfouda. | ['Fabio Poiesi', 'Elisa Ricci', 'Giuseppe Fiameni', 'Fabio Galasso', 'Nicu Sebe', 'Stéphane Lathuilière', 'Evgeny Krivosheev', 'Cristiano Saltori'] | 2022-07-20 | null | null | null | null | ['point-cloud-segmentation'] | ['computer-vision'] | [ 5.81439026e-02 -1.17176659e-02 1.50225507e-02 -4.60712194e-01
-4.46897060e-01 -7.99400151e-01 6.90052450e-01 -1.05525598e-01
-5.73150337e-01 6.69887543e-01 -7.45710790e-01 -3.96885365e-01
1.54273376e-01 -8.88056159e-01 -1.22160494e+00 -4.72892851e-01
6.09069690e-02 1.07427394e+00 8.80107164e-01 -6.10047400e-01
2.82719851e-01 8.11310709e-01 -1.88257790e+00 -2.66676486e-01
1.15734804e+00 7.95878112e-01 3.47899109e-01 6.57453656e-01
-3.08905452e-01 5.25561832e-02 -3.22736114e-01 -2.21544668e-01
7.55923688e-01 1.69890951e-02 -5.92622042e-01 7.62497336e-02
5.09775221e-01 -3.35981399e-02 -1.40662611e-01 1.09805024e+00
3.34189296e-01 1.81522921e-01 6.18008554e-01 -1.54874408e+00
-8.22522864e-02 -1.76686011e-02 -3.06817442e-01 2.06662819e-01
-2.27470070e-01 5.18434763e-01 3.51347089e-01 -7.52198935e-01
8.11531246e-01 9.83157992e-01 6.12121165e-01 5.96299589e-01
-9.96494710e-01 -8.75554860e-01 1.64177164e-01 3.89449745e-01
-1.36231625e+00 -3.72310191e-01 9.67088163e-01 -6.80567086e-01
7.54201651e-01 -1.57294869e-01 6.63703740e-01 1.09004545e+00
4.39724140e-02 5.57717562e-01 1.07103062e+00 -2.14362457e-01
6.47146106e-01 3.45990717e-01 5.17239608e-03 3.43487531e-01
4.34387803e-01 5.78157067e-01 -2.83746302e-01 3.17870229e-01
5.63267171e-01 -3.19815099e-01 2.07494929e-01 -8.97053540e-01
-1.11573434e+00 9.05463696e-01 4.41788048e-01 1.91277280e-01
-2.08014652e-01 -1.77688841e-02 2.37810314e-01 4.29697484e-01
4.95645285e-01 2.50638366e-01 -6.59806848e-01 -1.41952813e-01
-8.52101743e-01 6.46361411e-01 7.85420716e-01 1.26700652e+00
1.17556858e+00 1.10349156e-01 4.31439042e-01 5.98521948e-01
1.94610327e-01 7.97475457e-01 2.37820581e-01 -1.07263029e+00
3.62706512e-01 4.18615401e-01 2.84189135e-01 -5.63829005e-01
-4.96589512e-01 -6.12089097e-01 -2.65178293e-01 6.21831000e-01
6.04772627e-01 -1.91868439e-01 -1.13734972e+00 1.41593504e+00
7.07157016e-01 4.07496423e-01 2.04013601e-01 8.85850489e-01
6.62357509e-01 3.86348814e-01 -9.58561450e-02 4.14475113e-01
9.39897835e-01 -1.09854162e+00 -1.14450730e-01 -7.48600841e-01
6.59729004e-01 -5.02008498e-01 1.12734652e+00 2.22809732e-01
-6.67073667e-01 -6.53066099e-01 -1.12560856e+00 1.38278976e-01
-8.77378106e-01 -1.81353882e-01 3.91897351e-01 5.08843124e-01
-9.77740407e-01 4.24445421e-01 -7.97938466e-01 -7.06141055e-01
6.73727155e-01 3.42249662e-01 -2.88965076e-01 -7.28388205e-02
-1.26660383e+00 8.10278952e-01 5.71441889e-01 -2.24684149e-01
-8.84366393e-01 -8.25098038e-01 -6.74424827e-01 -5.26818693e-01
6.24298453e-01 -7.43303180e-01 1.53063047e+00 -1.13251054e+00
-1.56664848e+00 1.08786452e+00 -1.61032781e-01 -7.98992217e-01
8.86032999e-01 -1.49121761e-01 -3.68018150e-01 1.41677290e-01
3.51358235e-01 1.14653504e+00 8.38521600e-01 -1.63583910e+00
-8.82814109e-01 -4.09543127e-01 1.82330720e-02 1.86608359e-01
2.53479928e-01 -6.03318393e-01 -4.72847819e-01 -2.58166164e-01
6.79498687e-02 -1.24251962e+00 -3.44768703e-01 4.29626219e-02
-1.42690048e-01 -1.56725392e-01 9.13339615e-01 -8.52464437e-02
4.46447283e-01 -2.10559130e+00 -8.55800062e-02 -1.27910087e-02
-1.35073122e-02 4.80000496e-01 -1.50501713e-01 3.63649368e-01
7.25464970e-02 2.22330596e-02 -7.04043090e-01 -3.00051987e-01
5.86910918e-02 4.84805346e-01 -3.54478776e-01 3.91245872e-01
2.83917487e-01 9.00715888e-01 -8.99613738e-01 -4.38588738e-01
5.17639935e-01 2.87716389e-01 -4.43641424e-01 -7.08226115e-02
-6.55388117e-01 9.38847125e-01 -4.35987264e-01 5.06025553e-01
1.12800443e+00 1.06808156e-01 -5.09955585e-01 2.83447027e-01
-4.63258505e-01 3.61107774e-02 -1.04921436e+00 1.83354259e+00
-3.87115359e-01 6.55904114e-01 -1.31404042e-01 -1.03027117e+00
1.05774772e+00 -2.56181329e-01 4.53949571e-01 -9.27434862e-01
2.62702465e-01 5.76967657e-01 -2.66365092e-02 -3.58117104e-01
5.49591482e-01 -9.14264023e-02 -1.87586639e-02 -8.82240236e-02
-7.13152513e-02 -6.57909930e-01 3.51668388e-01 6.14527501e-02
7.43637264e-01 4.32617724e-01 -1.08092174e-01 -3.38484228e-01
5.79845846e-01 8.12424839e-01 4.44877803e-01 7.61777222e-01
-4.91639435e-01 7.31379151e-01 3.05497199e-01 -3.55564654e-01
-1.22368586e+00 -1.07057917e+00 -2.86652267e-01 7.57730126e-01
6.00069106e-01 2.29460523e-01 -8.45434368e-01 -8.40697169e-01
2.27843404e-01 1.03339911e+00 -5.38500130e-01 -1.57791264e-02
-5.53353190e-01 -3.18493664e-01 4.04619545e-01 3.90312552e-01
7.02771723e-01 -8.94877493e-01 -9.13436055e-01 1.19314782e-01
4.52278666e-02 -1.44166636e+00 7.73211345e-02 1.05700992e-01
-9.81830001e-01 -9.68306661e-01 -5.20461142e-01 -5.66351533e-01
4.39975858e-01 4.66592491e-01 1.21474743e+00 -1.88613057e-01
1.53479204e-01 3.42590243e-01 -3.60074937e-01 -8.80197167e-01
-5.07344425e-01 5.76561093e-01 -9.66311991e-02 -7.43599609e-02
7.87344098e-01 -6.09585524e-01 -5.18787742e-01 6.03823781e-01
-9.80719745e-01 -3.23697291e-02 4.56433564e-01 3.26031387e-01
8.54634702e-01 -1.45563334e-01 5.12940466e-01 -9.39285755e-01
1.20590538e-01 -6.80357099e-01 -1.07413948e+00 -4.06372070e-01
-5.14605820e-01 -5.92796281e-02 5.64868152e-01 -3.02991152e-01
-8.77925992e-01 4.46500510e-01 -4.39353317e-01 -6.40756428e-01
-7.04476595e-01 9.97100919e-02 -5.79693466e-02 -3.78073901e-01
9.54215169e-01 2.26332009e-01 4.75728512e-02 -2.67410398e-01
3.15958709e-01 5.13534486e-01 6.52021945e-01 -5.27878463e-01
1.27166188e+00 1.01299083e+00 1.54396221e-01 -8.90168011e-01
-4.54329818e-01 -6.83262169e-01 -1.11364031e+00 -2.64223784e-01
7.80778944e-01 -9.83447731e-01 -3.01829100e-01 6.90579414e-01
-1.02989936e+00 -8.28015268e-01 -5.44996440e-01 2.44104624e-01
-8.25723886e-01 1.94658116e-01 1.72374472e-01 -5.24687767e-01
9.57646444e-02 -1.09543765e+00 1.14562333e+00 4.28009450e-01
1.77507028e-01 -1.03842235e+00 1.46050990e-01 3.29315662e-01
4.92627442e-01 3.53458822e-01 3.60192299e-01 -7.15360224e-01
-8.52696836e-01 -1.32442430e-01 -1.67339042e-01 2.42464826e-01
-2.31342629e-01 -8.99478868e-02 -1.01027262e+00 -1.08009912e-01
-8.34846795e-02 -1.46569386e-01 9.54143941e-01 2.19088003e-01
9.08024013e-01 2.74150372e-01 -4.34205770e-01 8.18519354e-01
1.45562804e+00 2.93677211e-01 6.39814019e-01 8.27864170e-01
5.97772062e-01 6.90710425e-01 1.04667461e+00 3.23221795e-02
7.53189504e-01 5.49398065e-01 7.89042890e-01 -1.38972402e-01
-1.73085704e-01 -2.58992970e-01 1.67541150e-02 4.15055037e-01
7.42498413e-02 -3.11369896e-01 -1.38070929e+00 9.79282677e-01
-1.86647809e+00 -4.77677375e-01 -4.11105484e-01 2.18861651e+00
4.00403619e-01 4.57146138e-01 1.51360393e-01 -7.48186633e-02
4.43911076e-01 -4.48283479e-02 -9.21407342e-01 -2.49771342e-01
-1.14509366e-01 2.36130714e-01 1.01025760e+00 5.33120096e-01
-1.30764592e+00 1.43864727e+00 4.63383579e+00 7.08916306e-01
-1.38369286e+00 4.11447167e-01 1.72068834e-01 1.05263777e-01
-2.10224107e-01 9.08125043e-02 -8.86020422e-01 4.89565164e-01
1.14526629e+00 -8.84974748e-02 1.54633164e-01 9.57724452e-01
1.80461347e-01 -3.77477795e-01 -8.60513985e-01 8.30362260e-01
-1.58507869e-01 -1.28389859e+00 -2.45837420e-01 9.96966213e-02
7.52985418e-01 8.20191205e-01 8.54612738e-02 3.39045554e-01
2.62541056e-01 -6.07434034e-01 8.69629145e-01 4.50126827e-01
7.18917549e-01 -6.91532016e-01 7.20358372e-01 6.86348915e-01
-9.93526816e-01 8.92149806e-02 -4.76111144e-01 6.82136929e-03
1.19857378e-01 4.42477137e-01 -1.08770597e+00 7.12754309e-01
8.28953922e-01 8.22644413e-01 -6.45158231e-01 1.24944818e+00
-1.89392954e-01 5.62506080e-01 -6.96419060e-01 1.09450482e-01
5.34476638e-01 -2.89262265e-01 9.07783866e-01 1.08726799e+00
2.76047885e-01 -1.82278797e-01 2.57141106e-02 9.39586699e-01
2.45418623e-01 -1.05718948e-01 -9.06462729e-01 2.43092656e-01
4.28554237e-01 9.66931522e-01 -9.01908755e-01 -2.42526531e-01
-2.05234483e-01 6.83094501e-01 2.75494270e-02 3.71989906e-01
-8.58174205e-01 -3.56174350e-01 8.99110317e-01 4.66218174e-01
4.23023283e-01 -6.65785491e-01 -5.74428260e-01 -9.80066895e-01
-2.17485130e-02 -5.00825047e-01 -1.69326141e-02 -6.22793972e-01
-1.06574070e+00 5.72208822e-01 2.71872640e-01 -1.55517602e+00
-1.76426098e-01 -5.69303870e-01 -3.92289042e-01 7.39491642e-01
-2.14461517e+00 -1.10829055e+00 -6.56178176e-01 4.98757154e-01
8.29523921e-01 -1.75474852e-01 2.95597881e-01 2.89329678e-01
-2.43717939e-01 2.72596419e-01 2.17184260e-01 -2.68327057e-01
6.83075190e-01 -1.11331367e+00 1.01515472e+00 8.85196209e-01
-1.10216983e-01 -4.58609201e-02 9.11640525e-01 -6.35457635e-01
-1.16340888e+00 -1.49782932e+00 6.17783427e-01 -7.75741875e-01
5.50580323e-01 -5.59844315e-01 -1.09897912e+00 5.67457199e-01
-3.06346864e-02 1.79495543e-01 1.48997724e-01 -4.38045502e-01
-5.51662501e-03 -2.69388527e-01 -1.28962481e+00 5.54687619e-01
1.32259965e+00 -1.17715776e-01 -4.28178877e-01 3.95930648e-01
9.26977754e-01 -7.55607188e-01 -3.54207218e-01 6.50998354e-01
1.66574195e-01 -1.19173622e+00 8.46093595e-01 -2.34706461e-01
2.39840522e-01 -6.14334702e-01 -8.97685662e-02 -1.33446264e+00
7.64933275e-03 -3.61395568e-01 1.14191227e-01 8.34382951e-01
4.18000281e-01 -1.19571888e+00 9.96590257e-01 2.12719351e-01
-4.53595668e-01 -3.44813704e-01 -1.12542200e+00 -1.15585053e+00
5.19294679e-01 -8.49639893e-01 8.40730309e-01 8.18342686e-01
-8.83596480e-01 5.78899048e-02 3.19029748e-01 3.95127416e-01
7.37674057e-01 -2.41191033e-02 1.40600717e+00 -1.59809673e+00
1.21664219e-01 -3.17881286e-01 -5.65360188e-01 -1.04876471e+00
2.46643230e-01 -9.08428729e-01 8.86443034e-02 -1.48284853e+00
-6.06677175e-01 -8.37754428e-01 1.07892416e-01 2.39851490e-01
4.12106812e-01 2.61619598e-01 2.18750089e-01 2.61472166e-01
-4.52495217e-01 6.31173313e-01 1.15956044e+00 -1.43822074e-01
-3.97870272e-01 3.04631293e-01 -3.31118315e-01 6.32530510e-01
1.20048523e+00 -7.00853467e-01 -5.51424742e-01 -5.93798518e-01
1.30854979e-01 -3.56589913e-01 6.39701128e-01 -1.47821248e+00
2.73399204e-01 -2.64561296e-01 -1.13040395e-02 -8.59310925e-01
3.44244689e-01 -9.61944461e-01 7.37753361e-02 3.41072768e-01
3.45960975e-01 -1.55296564e-01 7.47545063e-01 5.17340541e-01
-1.36026859e-01 -2.50865996e-01 9.23542202e-01 -1.93807662e-01
-1.27912939e+00 3.80307645e-01 -1.91609398e-01 2.10188195e-01
1.35646319e+00 -6.77630484e-01 -3.34982693e-01 -4.48806696e-02
-5.47891676e-01 4.78635401e-01 8.85971129e-01 7.15415359e-01
3.75862449e-01 -8.77892435e-01 -6.95872247e-01 5.57264209e-01
4.85537261e-01 5.89067101e-01 3.28398138e-01 6.56263292e-01
-8.33314538e-01 4.33606148e-01 -3.22758138e-01 -1.16820431e+00
-7.36924827e-01 4.93902624e-01 5.69044530e-01 1.85336024e-01
-6.00016534e-01 8.37879241e-01 2.09070846e-01 -9.15888131e-01
-1.55276939e-01 -2.34291703e-01 2.14552041e-02 -1.53564602e-01
-9.79464725e-02 1.67410806e-01 3.30168545e-01 -8.62697482e-01
-3.82604927e-01 7.99393415e-01 1.18297048e-01 -2.40738302e-01
1.09500062e+00 -1.98211476e-01 3.39969486e-01 6.56900644e-01
9.86622572e-01 -3.15895230e-01 -1.61189985e+00 -2.28831530e-01
-1.53136933e-02 -6.26114488e-01 -7.33721107e-02 -7.33652115e-01
-9.58600044e-01 1.09233332e+00 9.79013443e-01 1.38774645e-02
8.98860037e-01 1.36839315e-01 8.56664717e-01 3.49549711e-01
6.84133530e-01 -1.21671391e+00 -2.69390494e-01 8.32445085e-01
7.29355454e-01 -1.64844072e+00 -4.81338382e-01 -4.21557456e-01
-6.60049617e-01 8.83670330e-01 8.23333621e-01 -2.59663641e-01
7.83722222e-01 6.81228936e-02 3.50108355e-01 -1.99629128e-01
-4.80847329e-01 -6.51467085e-01 -4.43994254e-03 9.93450344e-01
-3.05709898e-01 4.81096916e-02 -3.99742909e-02 2.88010120e-01
-5.07648647e-01 2.05207855e-01 5.51532924e-01 9.23221707e-01
-5.09519756e-01 -1.11850619e+00 -3.84939939e-01 1.38048247e-01
1.67473763e-01 2.65850574e-01 -4.43353117e-01 1.03546715e+00
4.89393562e-01 8.75143588e-01 2.71372736e-01 -3.37288469e-01
6.22464597e-01 8.24826360e-02 2.47031569e-01 -5.98957419e-01
-2.32530311e-01 -2.86259919e-01 -2.11054161e-01 -4.53683436e-01
-3.17404360e-01 -8.45284700e-01 -1.38554037e+00 -1.92111880e-01
5.01921773e-02 4.85590249e-02 1.16290474e+00 8.27945232e-01
5.97516716e-01 4.56602365e-01 4.77004230e-01 -1.25558102e+00
-2.18672737e-01 -5.71997643e-01 -2.65419483e-01 3.14494997e-01
5.26678562e-01 -1.00684154e+00 -3.69540274e-01 -5.23270518e-02] | [8.146190643310547, -2.6270854473114014] |
997cc91c-0404-47e6-9027-37d3dd498862 | towards-understanding-distributional-1 | null | null | https://openreview.net/forum?id=nK7eZEURiJ4 | https://openreview.net/pdf?id=nK7eZEURiJ4 | Towards Understanding Distributional Reinforcement Learning: Regularization, Optimization, Acceleration and Sinkhorn Algorithm | Distributional reinforcement learning~(RL) is a class of state-of-the-art algorithms that estimate the whole distribution of the total return rather than only its expectation. Despite the remarkable performance of distributional RL, a theoretical understanding of its advantages over expectation-based RL remains elusive. In this paper, we interpret distributional RL as entropy-regularized maximum likelihood estimation in the \textit{neural Z-fitted iteration} framework and establish the connection of the resulting risk-aware regularization with maximum entropy RL. In addition, We shed light on the stability-promoting distributional loss with desirable smoothness properties in distributional RL, which can yield stable optimization and guaranteed generalization. We also analyze the acceleration behavior while optimizing distributional RL algorithms and show that an appropriate approximation to the true target distribution can speed up the convergence. From the perspective of representation, we find that distributional RL encourages state representation from the same action class classified by the policy in tighter clusters. Finally, we propose a class of \textit{Sinkhorn distributional RL} algorithm that interpolates between the Wasserstein distance and maximum mean discrepancy~(MMD). Experiments on a suite of Atari games reveal the competitive performance of our algorithm relative to existing state-of-the-art distributional RL algorithms. | ['Linglong Kong', 'Bei Jiang', 'Xiaodong Yan', 'Aref Sadeghi', 'Yafei Wang', 'Enze Shi', 'Yi Liu', 'Yingnan Zhao', 'Ke Sun'] | 2021-09-29 | null | null | null | null | ['distributional-reinforcement-learning'] | ['methodology'] | [-2.45523989e-01 3.70799690e-01 -5.03195703e-01 -3.83269310e-01
-1.19989729e+00 -4.85634834e-01 3.56886804e-01 9.37523842e-02
-8.15416694e-01 8.89251709e-01 1.50322556e-01 -5.31245291e-01
-5.68193436e-01 -6.75764084e-01 -7.93722451e-01 -1.21707988e+00
-1.75419778e-01 6.53598011e-01 -2.01263800e-01 -1.01659976e-01
3.66347134e-01 3.42481881e-01 -1.23692334e+00 -1.39639899e-01
1.03412831e+00 1.15173054e+00 1.02040417e-01 4.95609671e-01
-7.90397748e-02 1.12155879e+00 -5.42100608e-01 -3.53073955e-01
4.59469825e-01 -5.63492000e-01 -5.00858068e-01 -2.07763970e-01
-1.71736795e-02 -2.65704453e-01 -2.31424227e-01 1.37901068e+00
6.95619762e-01 5.23073554e-01 1.05155861e+00 -1.19968116e+00
-6.04875684e-01 8.05943191e-01 -9.80920911e-01 2.03582719e-01
-1.10346697e-01 1.71075061e-01 1.40295672e+00 -5.08899510e-01
5.05734622e-01 1.38423252e+00 5.89352310e-01 7.20670223e-01
-1.63147593e+00 -6.43975437e-01 4.20890242e-01 -2.33071834e-01
-1.31566370e+00 -1.51729897e-01 6.70799196e-01 -3.82711709e-01
7.52519786e-01 -6.25820458e-02 2.87389338e-01 9.12793636e-01
2.44766548e-01 1.31931269e+00 1.28754926e+00 -3.37298661e-01
7.38821507e-01 2.77656376e-01 -8.69038105e-02 6.63947105e-01
2.29889646e-01 4.31568265e-01 -3.84021431e-01 -3.32924604e-01
5.75355768e-01 -1.18192255e-01 -1.29946083e-01 -9.25676346e-01
-5.74600875e-01 1.32721508e+00 2.76865304e-01 -1.53010160e-01
-4.76992875e-01 7.47499108e-01 3.80548596e-01 2.89346129e-01
8.41690242e-01 4.25866216e-01 -3.32491904e-01 -4.00920957e-01
-8.67290974e-01 4.09491122e-01 6.27172410e-01 8.27998877e-01
7.08387077e-01 3.73541117e-01 -2.65900493e-01 5.74034333e-01
5.87692440e-01 8.04760754e-01 3.13490242e-01 -1.30999255e+00
5.50115705e-01 1.72246635e-01 3.01113069e-01 -4.37348187e-01
-2.09807992e-01 -5.83402991e-01 -7.14369833e-01 5.28420448e-01
3.92727762e-01 -5.69136739e-01 -3.08301061e-01 2.20854545e+00
1.73589900e-01 1.68899626e-01 1.97363943e-01 6.22912705e-01
-2.46576145e-01 5.69755971e-01 5.87082170e-02 -5.83939791e-01
4.10213500e-01 -4.28465307e-01 -5.27142942e-01 1.22871965e-01
7.45096266e-01 -1.49893925e-01 1.35494924e+00 3.44390422e-01
-1.12283432e+00 -4.03174199e-03 -6.62828803e-01 4.69724506e-01
1.41267106e-01 -1.62377685e-01 6.49297297e-01 8.39604139e-01
-1.15646088e+00 9.56672549e-01 -7.73862064e-01 -9.23588499e-03
4.94210482e-01 3.48566443e-01 3.02944958e-01 3.45120698e-01
-9.53238845e-01 8.22770953e-01 2.81421095e-01 -1.32324770e-01
-1.14165390e+00 -7.44475484e-01 -6.68468475e-01 -9.88519862e-02
4.83042270e-01 -4.40250129e-01 1.65882778e+00 -1.03165865e+00
-1.90296423e+00 3.72551531e-01 2.51397900e-02 -8.87602091e-01
7.02874184e-01 -2.60816991e-01 2.50126034e-01 2.66799964e-02
-2.03181654e-02 5.08498132e-01 8.37137818e-01 -1.06846821e+00
-6.58119500e-01 -2.34448090e-01 -1.50925025e-01 4.74852383e-01
-3.28402460e-01 -2.83814073e-01 3.14151943e-01 -5.29917002e-01
-5.07076263e-01 -7.40197837e-01 -4.48751628e-01 -3.28436404e-01
-2.20093831e-01 -6.63415372e-01 2.48021886e-01 -1.15825102e-01
1.45497644e+00 -2.11465263e+00 4.12195891e-01 5.93936622e-01
8.15957934e-02 -1.75439551e-01 5.15536144e-02 4.41554487e-01
-3.07410657e-02 1.32641450e-01 -1.94953680e-01 -4.96363878e-01
6.63238168e-01 2.59872675e-01 -7.36049414e-01 1.00260293e+00
-1.40700966e-01 7.07844973e-01 -1.05420017e+00 -3.05983037e-01
9.08221677e-02 1.15301060e-02 -8.57046187e-01 3.35177034e-01
-5.90472937e-01 1.09834142e-01 -7.31827319e-01 1.21382236e-01
2.67413557e-01 2.77672894e-02 3.39414179e-02 3.08473498e-01
-7.98390508e-02 3.94507535e-02 -1.10893786e+00 1.51899612e+00
-3.62570912e-01 4.30427521e-01 3.88211645e-02 -1.27020228e+00
7.92420387e-01 -8.04787278e-02 7.96033621e-01 -2.77125597e-01
3.08483005e-01 4.44948003e-02 -2.55965233e-01 -1.89229444e-01
6.11644506e-01 -4.59167033e-01 -3.23740870e-01 9.71105993e-01
-2.61428412e-02 -1.21744625e-01 -1.35923371e-01 1.55548856e-01
7.49908328e-01 3.46010834e-01 1.04986623e-01 -7.05488503e-01
-1.07894614e-01 -2.44414836e-01 5.20130992e-01 1.06860077e+00
-5.11142671e-01 -3.65388989e-02 7.82673895e-01 -4.01109681e-02
-8.99326622e-01 -1.48628056e+00 -7.67337531e-02 1.47648180e+00
-9.22468901e-02 -2.72594154e-01 -7.27654457e-01 -7.99265265e-01
2.45872274e-01 1.28637838e+00 -8.21461439e-01 -4.95753050e-01
-2.49850795e-01 -9.06535387e-01 7.80736327e-01 4.54980046e-01
3.29504788e-01 -8.44106913e-01 -6.37011766e-01 1.07365035e-01
7.89645186e-04 -9.04064253e-02 -6.36097491e-01 4.71530646e-01
-8.87993693e-01 -8.24443698e-01 -6.07344925e-01 -4.12040561e-01
5.01952231e-01 -3.08307618e-01 1.31241941e+00 -7.61568964e-01
1.26340777e-01 8.30918133e-01 3.86708714e-02 -6.98483288e-01
-2.90838927e-01 -4.04159874e-02 4.44595814e-01 -4.21493351e-02
4.02535200e-01 -4.34857011e-01 -6.61346793e-01 -4.06771414e-02
-5.98138690e-01 -5.77729821e-01 4.60962892e-01 7.01378286e-01
7.61230409e-01 8.01872239e-02 7.35312700e-01 -9.51160908e-01
1.20665061e+00 -5.86300790e-01 -7.96360433e-01 2.71126300e-01
-8.19573104e-01 6.97694778e-01 4.29742038e-01 -5.90064287e-01
-1.20521975e+00 -1.17061689e-01 -3.95130366e-03 -7.05187559e-01
2.57802010e-01 3.73111516e-01 2.56005436e-01 3.20461154e-01
7.40220129e-01 3.72552335e-01 2.26881564e-01 -1.92899421e-01
8.58101130e-01 5.28745294e-01 3.21562171e-01 -1.16696846e+00
5.24762213e-01 2.23959938e-01 -9.50938016e-02 -7.07532465e-01
-1.13571358e+00 -4.19974238e-01 -4.11903262e-02 -3.20992589e-01
8.52157354e-01 -7.54161716e-01 -1.16783464e+00 3.31452936e-02
-5.26689529e-01 -8.77786219e-01 -1.10312963e+00 5.02711952e-01
-1.48843753e+00 1.77040726e-01 -5.45617878e-01 -1.63437545e+00
-2.97082961e-01 -1.01007819e+00 9.10516202e-01 -2.14538798e-02
3.03516369e-02 -1.26271617e+00 5.14930427e-01 -2.24825561e-01
2.82117695e-01 2.05387190e-01 8.41156363e-01 -5.48123479e-01
-2.62803465e-01 3.08661014e-01 1.87579632e-01 4.87573326e-01
-2.40590930e-01 -1.54681906e-01 -6.84453666e-01 -6.08993828e-01
1.75667182e-01 -7.21657395e-01 1.04003954e+00 9.49456990e-01
1.20888150e+00 -3.99195790e-01 3.46042253e-02 5.35859585e-01
1.52148736e+00 7.46958554e-02 3.13120663e-01 1.78601086e-01
1.73162684e-01 3.77959907e-01 9.04870629e-01 9.24432456e-01
2.26306513e-01 1.27813742e-01 5.01401186e-01 3.35916877e-01
4.90229189e-01 -5.30109525e-01 9.79419231e-01 5.88865936e-01
1.80630893e-01 7.29819620e-03 -5.20817578e-01 3.62494051e-01
-2.15128660e+00 -1.29566550e+00 5.10737181e-01 2.44441032e+00
1.10504520e+00 1.68486074e-01 5.51121593e-01 -3.28804106e-01
6.50733173e-01 1.62272066e-01 -1.05158317e+00 -5.98296702e-01
9.49233621e-02 2.24767357e-01 8.81672978e-01 7.20228970e-01
-9.77894425e-01 9.94176209e-01 6.92414236e+00 1.35866046e+00
-6.96126580e-01 2.35814936e-02 8.58978033e-01 -4.06927288e-01
-5.09344161e-01 -4.00647551e-01 -9.66331244e-01 2.75176585e-01
1.09061730e+00 -5.60981452e-01 7.53623009e-01 1.17328608e+00
3.33397925e-01 -1.65250935e-02 -1.20052683e+00 8.57985437e-01
-1.38106480e-01 -1.16969335e+00 -1.98628873e-01 3.53974432e-01
7.97462165e-01 3.08153003e-01 4.84934449e-01 6.92792177e-01
1.24576437e+00 -9.50648248e-01 8.74591231e-01 6.67286158e-01
7.31553435e-01 -1.41146255e+00 4.61791515e-01 6.83079362e-01
-1.04164159e+00 -3.48929614e-01 -8.75758290e-01 7.10042343e-02
-2.78506905e-01 3.92574400e-01 -5.49747884e-01 1.80309638e-01
4.97615218e-01 7.45291412e-01 3.38164605e-02 6.07538402e-01
-1.65627271e-01 6.46255910e-01 -4.14364487e-01 -4.38797772e-01
6.59126341e-01 -6.30653858e-01 4.78323787e-01 9.89539146e-01
1.79163903e-01 -2.02303883e-02 5.18142581e-01 1.08195484e+00
-4.73667197e-02 7.26877451e-02 -1.00270712e+00 -4.03253995e-02
3.03224206e-01 7.35586405e-01 -2.84649640e-01 -3.75809968e-02
-6.46080375e-02 5.45298755e-01 8.67392480e-01 4.23936874e-01
-9.93145108e-01 -2.58972913e-01 9.50889885e-01 -2.31246978e-01
2.61401504e-01 -3.93622220e-02 -1.27872750e-01 -1.04757917e+00
-2.73494422e-01 -5.42989373e-01 6.41395628e-01 -2.36896768e-01
-1.66963041e+00 2.86180943e-01 8.27126503e-02 -1.12642741e+00
-5.19933820e-01 -5.01818120e-01 -4.74487931e-01 6.25946760e-01
-1.39305890e+00 -3.65346074e-01 6.77449584e-01 9.05348957e-01
4.73436952e-01 -2.22588167e-01 6.15892231e-01 -1.98911011e-01
-6.24145448e-01 8.23314548e-01 9.05566275e-01 -1.11206926e-01
6.26668453e-01 -1.87728798e+00 -2.81892627e-01 4.41088617e-01
-7.39111239e-03 4.75009590e-01 9.11551833e-01 -4.89079654e-01
-1.44523072e+00 -1.26981211e+00 6.70298710e-02 -4.70753133e-01
1.11103404e+00 -1.40459776e-01 -3.31828266e-01 8.13938916e-01
1.73634425e-01 -3.17297220e-01 6.99726701e-01 1.99770853e-01
-2.09235966e-01 -1.98238462e-01 -1.17579603e+00 7.37611830e-01
7.01159298e-01 -5.78909516e-01 -5.91665506e-01 4.17446941e-01
6.91119730e-01 2.33463217e-02 -8.48032832e-01 -1.76332623e-01
2.11212754e-01 -9.56776083e-01 9.62124646e-01 -8.30688119e-01
1.35660142e-01 1.61360987e-02 -3.28105330e-01 -1.73555624e+00
-1.42528579e-01 -9.94632840e-01 -2.18418837e-01 9.53525364e-01
2.98265725e-01 -6.64421380e-01 8.97044122e-01 4.73020107e-01
2.24565249e-02 -9.58610594e-01 -9.90405142e-01 -9.86587465e-01
9.59487677e-01 -6.20369554e-01 -4.19861730e-03 5.30128121e-01
5.11282206e-01 6.05093203e-02 -4.27759856e-01 -1.57284848e-02
1.18932474e+00 2.78449088e-01 4.83839512e-01 -1.01479530e+00
-5.87904155e-01 -7.77331471e-01 1.41654119e-01 -1.30562496e+00
8.12862813e-01 -9.89597023e-01 3.66903722e-01 -1.07656121e+00
2.24544927e-01 -5.35698891e-01 -6.08660281e-01 2.48280242e-01
2.05931868e-02 -3.65510672e-01 1.65165573e-01 6.20742030e-02
-1.11623895e+00 1.08761072e+00 8.84000480e-01 3.59258987e-02
-3.61349523e-01 3.30004871e-01 -8.21496904e-01 9.33266401e-01
1.00248754e+00 -5.91983140e-01 -9.10280228e-01 -1.33888543e-01
3.19458961e-01 2.45678455e-01 -1.78390041e-01 -4.92456198e-01
1.62983626e-01 -5.11389196e-01 1.47005208e-02 -5.63306987e-01
1.49532080e-01 -5.81022263e-01 -4.61400598e-01 5.91416001e-01
-1.04067159e+00 1.02685437e-01 -1.59214765e-01 1.07642984e+00
1.78760394e-01 -2.74631977e-01 1.09733677e+00 -8.68208557e-02
-1.58563346e-01 3.66148561e-01 -6.54970646e-01 7.32740223e-01
1.10524726e+00 2.58284837e-01 1.46232238e-02 -7.67196953e-01
-5.66701233e-01 4.23521549e-01 1.23293780e-01 -2.33413130e-01
6.54814363e-01 -1.24886537e+00 -7.50300825e-01 -1.09797724e-01
-1.59891084e-01 -1.33681074e-01 -2.75949508e-01 8.20663154e-01
4.64050211e-02 5.05926795e-02 1.73996180e-01 -5.31415701e-01
-6.31737292e-01 6.94108546e-01 5.10042906e-01 -6.20350838e-01
-5.20810187e-01 8.37468326e-01 3.74105126e-01 -4.74445224e-01
7.28271067e-01 -3.61095876e-01 2.40302190e-01 -6.51819333e-02
4.70135599e-01 4.78704959e-01 -5.50330400e-01 -8.62480402e-02
-1.02857232e-01 2.56171644e-01 2.32127272e-02 -5.11102796e-01
1.44966948e+00 -2.29120702e-01 1.39137998e-01 7.71535933e-01
1.03100955e+00 -1.80922255e-01 -1.96642005e+00 -1.78762272e-01
3.28141361e-01 -3.04071963e-01 8.20327699e-02 -4.78588641e-01
-1.11236191e+00 8.66071761e-01 5.55859208e-01 9.74175036e-02
8.69611382e-01 -7.18708262e-02 2.65608668e-01 7.71922648e-01
5.75180709e-01 -1.42617440e+00 2.18423322e-01 4.45399225e-01
5.31281710e-01 -9.10714328e-01 1.27698388e-02 3.92590135e-01
-1.05877995e+00 8.44933808e-01 3.17906111e-01 -6.74368382e-01
8.47657084e-01 3.52812797e-01 -3.15756083e-01 -2.31724530e-02
-1.03661394e+00 -2.96473265e-01 -1.80551618e-01 6.37352288e-01
2.06480339e-01 1.74030721e-01 -9.66408774e-02 5.73876560e-01
-1.66859254e-01 -3.04681689e-01 4.64082867e-01 7.91937530e-01
-7.88591862e-01 -9.39223468e-01 6.15625232e-02 5.11449754e-01
-5.53722024e-01 -1.04583152e-01 -1.01011224e-01 7.39517152e-01
-5.78228474e-01 9.30672646e-01 9.94406119e-02 -1.38901889e-01
1.81320071e-01 9.35851969e-03 4.93266642e-01 -3.31036299e-01
-5.23631155e-01 3.13046187e-01 -1.93954006e-01 -7.44982600e-01
-2.77090758e-01 -9.58470881e-01 -1.29230762e+00 -5.39135754e-01
-3.14202398e-01 5.68700492e-01 3.03865045e-01 8.18536222e-01
1.03684887e-01 2.52437681e-01 9.28290784e-01 -7.13874757e-01
-1.68176341e+00 -6.44230962e-01 -1.04131079e+00 3.10792983e-01
2.64420211e-01 -6.98131263e-01 -5.41448772e-01 -3.47610533e-01] | [4.091377258300781, 2.572293758392334] |
a475caa0-da23-44d5-832d-e8951f2941c8 | practical-and-consistent-estimation-of-f | 1905.11112 | null | https://arxiv.org/abs/1905.11112v2 | https://arxiv.org/pdf/1905.11112v2.pdf | Practical and Consistent Estimation of f-Divergences | The estimation of an f-divergence between two probability distributions based on samples is a fundamental problem in statistics and machine learning. Most works study this problem under very weak assumptions, in which case it is provably hard. We consider the case of stronger structural assumptions that are commonly satisfied in modern machine learning, including representation learning and generative modelling with autoencoder architectures. Under these assumptions we propose and study an estimator that can be easily implemented, works well in high dimensions, and enjoys faster rates of convergence. We verify the behavior of our estimator empirically in both synthetic and real-data experiments, and discuss its direct implications for total correlation, entropy, and mutual information estimation. | ['Paul K. Rubenstein', 'Josip Djolonga', 'Olivier Bousquet', 'Ilya Tolstikhin', 'Carlos Riquelme'] | 2019-05-27 | practical-and-consistent-estimation-of-f-1 | http://papers.nips.cc/paper/8661-practical-and-consistent-estimation-of-f-divergences | http://papers.nips.cc/paper/8661-practical-and-consistent-estimation-of-f-divergences.pdf | neurips-2019-12 | ['mutual-information-estimation'] | ['methodology'] | [ 3.93769518e-02 1.84491158e-01 -1.20985679e-01 -4.99425620e-01
-5.25610864e-01 -4.04802501e-01 5.61660767e-01 3.64722103e-01
-5.55042267e-01 9.51127887e-01 1.99659187e-02 -4.15084809e-01
-4.07409370e-01 -8.77255619e-01 -5.54466367e-01 -6.61871493e-01
-2.74996430e-01 6.30843580e-01 -8.79527107e-02 1.04113281e-01
1.33143455e-01 3.09684813e-01 -1.45854235e+00 -4.27022874e-01
8.73413563e-01 1.01179397e+00 -1.32810950e-01 7.74557471e-01
2.84972131e-01 5.96934497e-01 -5.78775346e-01 -5.92981279e-01
2.47213811e-01 -4.66518730e-01 -7.99119949e-01 -1.07417507e-02
5.08884005e-02 -1.46509916e-01 -9.11310688e-02 1.25642776e+00
3.74857575e-01 2.05214605e-01 1.20999002e+00 -1.22543001e+00
-4.75058109e-01 6.43985868e-01 -5.32485247e-01 1.90818563e-01
1.78341225e-01 -2.59775728e-01 1.12627697e+00 -4.02434707e-01
1.78171024e-01 1.09706688e+00 8.17750573e-01 2.88582027e-01
-1.49946880e+00 -5.04134655e-01 -2.60599166e-01 -2.35618046e-03
-1.59512818e+00 -1.72010973e-01 4.87768054e-01 -4.36338067e-01
6.38308704e-01 7.56161287e-02 4.86295938e-01 9.41356003e-01
3.69405687e-01 5.06909907e-01 1.16139352e+00 -5.50142646e-01
4.88468051e-01 4.12125170e-01 2.09030122e-01 4.07076865e-01
7.20668674e-01 1.37890369e-01 -4.33320254e-02 -3.97005945e-01
8.42790723e-01 -9.64010730e-02 -3.16712141e-01 -4.44541484e-01
-9.57121670e-01 1.23076653e+00 -7.61664212e-02 3.88030887e-01
-3.64623904e-01 6.78805560e-02 1.94646493e-01 5.54977179e-01
6.21890962e-01 2.80329198e-01 -3.32577080e-01 -2.73857296e-01
-7.55432487e-01 2.77267694e-01 1.38496578e+00 8.24954748e-01
6.01244628e-01 -9.33366120e-02 3.26137781e-01 7.63121247e-01
1.43075049e-01 6.51627719e-01 5.51402390e-01 -7.92588413e-01
1.21429987e-01 6.06043637e-02 1.08798653e-01 -9.57057476e-01
-3.19984049e-01 -6.60170376e-01 -1.10867333e+00 -1.02039859e-01
4.89163637e-01 -4.03312266e-01 -3.43460143e-01 1.88123322e+00
1.72003299e-01 1.06770582e-01 1.22722343e-01 6.35668695e-01
3.21636915e-01 4.86702293e-01 -3.31947088e-01 -6.05386198e-01
8.46591413e-01 -1.01675816e-01 -8.10638785e-01 -3.29654813e-02
4.56176311e-01 -5.73621571e-01 6.85684621e-01 4.02688593e-01
-8.90913248e-01 -1.63813531e-01 -8.67091715e-01 2.22710699e-01
-2.09051803e-01 -3.24064828e-02 6.66985571e-01 9.79468822e-01
-8.71169329e-01 8.94900799e-01 -7.71862507e-01 -3.01432431e-01
1.11123845e-01 3.63898247e-01 -3.35628301e-01 1.64569721e-01
-9.97029781e-01 8.39757562e-01 6.19849265e-01 -6.74303547e-02
-4.84697789e-01 -5.02249062e-01 -7.68135488e-01 3.29646170e-01
2.31398374e-01 -6.16997004e-01 1.23338616e+00 -8.03971648e-01
-1.57339299e+00 5.17304361e-01 -1.51732108e-02 -6.67503119e-01
5.48914194e-01 -2.44692937e-01 -3.10493320e-01 -1.95626795e-01
-3.35908890e-01 2.79068984e-02 6.71098948e-01 -7.45936215e-01
-7.81165808e-02 -2.80229032e-01 -1.42518803e-01 -1.41690329e-01
-3.98876518e-01 -3.58907670e-01 2.61917561e-01 -4.20330375e-01
8.26409683e-02 -9.07766879e-01 -2.88900137e-01 -1.85256258e-01
-2.76147515e-01 -1.34481519e-01 3.20858508e-01 -3.61629069e-01
1.03902698e+00 -1.95901871e+00 6.38535768e-02 5.99417210e-01
1.09784432e-01 -6.49928069e-03 1.29536942e-01 6.11959815e-01
-5.58885261e-02 -2.64791888e-03 -4.04084653e-01 -9.76041481e-02
1.92056999e-01 2.20303446e-01 -1.63839892e-01 8.52352202e-01
-7.03005120e-02 5.46545148e-01 -7.48659432e-01 -3.49872321e-01
8.70254114e-02 4.57564920e-01 -7.38238275e-01 3.43259484e-01
-5.34859486e-03 1.50684699e-01 -2.19828486e-01 7.47335106e-02
5.26607931e-01 -5.78030169e-01 4.52859402e-01 1.46713093e-01
2.04738185e-01 2.51529574e-01 -1.62503588e+00 1.06696749e+00
-6.17753208e-01 8.46097529e-01 -1.67637095e-01 -1.32347000e+00
9.88711774e-01 1.84102684e-01 3.57261717e-01 -1.46810964e-01
4.36219782e-01 1.06603101e-01 2.16476694e-01 -2.34719917e-01
2.34046772e-01 -4.72085893e-01 1.09304981e-02 6.76453471e-01
1.68007433e-01 -2.85220832e-01 3.55936624e-02 -8.32770839e-02
9.96092081e-01 -3.94851297e-01 8.02597404e-01 -5.80313742e-01
3.01591933e-01 -7.92700469e-01 3.46286774e-01 9.39162433e-01
9.35769081e-02 3.69803429e-01 6.80570364e-01 -7.63781667e-02
-1.19330430e+00 -1.19572198e+00 -4.43467945e-01 5.87077439e-01
2.64870236e-03 -3.98427337e-01 -6.42528296e-01 -4.29604053e-01
9.65112969e-02 7.81812191e-01 -7.80187070e-01 -2.81458408e-01
4.37913574e-02 -1.08598030e+00 3.03237349e-01 4.90044653e-01
2.60415345e-01 -4.75001097e-01 -3.49195659e-01 4.71222727e-03
1.10183194e-01 -1.14763618e+00 -2.45321989e-02 2.59613752e-01
-8.52038682e-01 -1.16289806e+00 -7.08962619e-01 -2.88872033e-01
4.53714997e-01 -1.86361820e-01 1.09894764e+00 -2.42483556e-01
-4.59128134e-02 3.71165037e-01 -5.62371686e-02 -4.04167801e-01
-6.43609464e-01 1.45384297e-01 3.08616161e-01 -2.31987983e-02
3.51291090e-01 -8.26538324e-01 -2.14821935e-01 3.19467396e-01
-1.02893436e+00 -2.93665767e-01 7.05606997e-01 1.04999852e+00
3.41350049e-01 3.53933007e-01 5.08634329e-01 -8.69804323e-01
9.06104505e-01 -7.21352339e-01 -8.68753433e-01 6.57164976e-02
-7.13054478e-01 5.46621859e-01 6.87913120e-01 -3.86230111e-01
-8.42157364e-01 -2.43037567e-01 -1.47674695e-01 -2.31122106e-01
-1.83254600e-01 5.42848706e-01 -5.68445139e-02 6.60275817e-02
6.05297804e-01 2.01757610e-01 1.78845748e-01 -4.85613555e-01
1.21498905e-01 8.45162988e-01 3.01131278e-01 -4.95089144e-01
7.65417755e-01 3.73849690e-01 2.84889936e-01 -1.15345860e+00
-8.77085268e-01 -2.61914194e-01 -6.24769092e-01 4.55157124e-02
4.27402586e-01 -6.24833167e-01 -9.70504761e-01 1.96586013e-01
-9.18742061e-01 -2.05594257e-01 -3.10894310e-01 9.21413779e-01
-7.91756034e-01 3.47362399e-01 -3.51864398e-01 -1.19961834e+00
-5.64845316e-02 -6.94595873e-01 6.37718439e-01 2.35498011e-01
-2.54824132e-01 -1.53880131e+00 3.80449057e-01 -3.44735026e-01
5.00425696e-01 1.66240662e-01 9.34533298e-01 -8.31517279e-01
-2.51165163e-02 -3.18831682e-01 -8.54717791e-02 7.00215161e-01
6.57216609e-02 6.57599941e-02 -6.93582892e-01 -2.78881252e-01
-2.68944167e-02 -1.84692815e-01 9.01788592e-01 5.21191537e-01
1.34928083e+00 -5.05437970e-01 -1.91719234e-01 3.53871912e-01
1.41770554e+00 -2.28362963e-01 4.75146234e-01 -3.06896806e-01
1.13020189e-01 6.73594713e-01 1.73996359e-01 7.04393506e-01
1.87488496e-01 3.50096166e-01 1.56061485e-01 3.89600694e-01
5.36050797e-01 -9.27011520e-02 2.33231321e-01 1.22883022e+00
-8.99483263e-02 -3.40429515e-01 -7.45639920e-01 5.46102583e-01
-1.74866307e+00 -1.12799358e+00 9.68473256e-02 2.53778982e+00
8.86301160e-01 1.99337631e-01 2.70773888e-01 3.76531839e-01
8.60159338e-01 -1.10217273e-01 -3.58446896e-01 -5.29427946e-01
1.79317519e-02 4.70546275e-01 5.58281720e-01 5.11601567e-01
-9.81222391e-01 3.36518288e-01 7.87676907e+00 7.05027580e-01
-9.45347607e-01 8.59517083e-02 5.29186368e-01 3.25776413e-02
-3.88870865e-01 -1.15332216e-01 -5.50017178e-01 5.32257080e-01
1.20178854e+00 -3.63535613e-01 2.40154937e-01 9.04545307e-01
-2.05788195e-01 -2.77314901e-01 -1.22410858e+00 1.05253410e+00
-7.09487796e-02 -8.92098248e-01 -4.16354477e-01 1.96438283e-01
6.65084064e-01 9.79219470e-03 1.37442902e-01 1.06792204e-01
5.19410312e-01 -1.22700417e+00 2.12261796e-01 5.72919786e-01
4.95569557e-01 -9.85685468e-01 1.02995253e+00 6.19752049e-01
-5.52218616e-01 -7.72827417e-02 -6.82081223e-01 -3.84215862e-01
-9.05217901e-02 1.41466999e+00 -8.09374988e-01 4.37848419e-01
2.62521178e-01 6.12386048e-01 -2.97436029e-01 1.23011816e+00
-1.12421416e-01 9.21925366e-01 -8.11708689e-01 -5.61570287e-01
-1.42549857e-01 -3.11012834e-01 3.71078432e-01 1.23045313e+00
5.09573936e-01 -1.86223641e-01 -1.44779071e-01 9.63884234e-01
-1.77963171e-02 7.57603496e-02 -7.29111373e-01 -1.07187271e-01
5.49254894e-01 9.26983416e-01 -6.26652420e-01 -2.55718112e-01
-3.03628981e-01 5.89161634e-01 3.34260494e-01 2.15019226e-01
-7.60427654e-01 -4.71422970e-01 7.00042725e-01 1.38999715e-01
4.74058330e-01 -3.70815277e-01 -1.29126944e-02 -1.34848559e+00
5.45760021e-02 -4.81251448e-01 3.19835097e-01 -4.47468050e-02
-1.57801569e+00 3.79834503e-01 4.07441795e-01 -8.93624902e-01
-7.77116120e-01 -6.96656287e-01 -4.44582999e-01 5.80102444e-01
-1.19795430e+00 -2.99498528e-01 -9.51431599e-03 3.64722073e-01
-1.91768840e-01 -7.53424093e-02 9.44113314e-01 1.74241617e-01
-4.85838383e-01 5.64509094e-01 7.03967154e-01 6.68821335e-02
3.44100356e-01 -1.39080334e+00 9.72587094e-02 5.58189631e-01
3.78055632e-01 5.52982032e-01 1.09703112e+00 -3.14076662e-01
-1.33135593e+00 -8.08700144e-01 6.61003232e-01 -5.91965020e-02
7.19853342e-01 -4.86054540e-01 -8.44605744e-01 5.82194507e-01
-2.76911119e-03 7.87026063e-02 8.79787922e-01 5.23917437e-01
-3.95253181e-01 7.70711303e-02 -1.10687518e+00 1.50779128e-01
7.30610371e-01 -4.85088348e-01 -5.39170563e-01 4.33933139e-01
1.35342062e-01 -1.36915654e-01 -1.15344083e+00 3.79960299e-01
8.51518214e-01 -1.13600993e+00 6.98159277e-01 -4.75634515e-01
2.68681943e-01 1.42642438e-01 -2.74436563e-01 -1.43908978e+00
-1.00281663e-01 -6.39995873e-01 -2.34834716e-01 9.13956344e-01
3.41228366e-01 -9.98230100e-01 5.07484853e-01 4.59136188e-01
6.98027968e-01 -7.54721582e-01 -1.01558399e+00 -1.29212213e+00
4.57244545e-01 -5.25303245e-01 4.88598049e-01 8.19998622e-01
1.10717513e-01 3.77945185e-01 -3.99004996e-01 -2.58281790e-02
7.04029739e-01 1.02450572e-01 7.81948447e-01 -1.62182391e+00
-7.28772700e-01 -4.76457357e-01 -8.40309918e-01 -1.15836024e+00
4.90837663e-01 -7.96468258e-01 2.24460904e-02 -1.02853465e+00
2.31210098e-01 -3.20019424e-01 -1.45425856e-01 -9.41892248e-03
1.30952373e-02 -2.00668294e-02 -2.10051507e-01 -5.42040281e-02
-3.81198466e-01 7.83202946e-01 6.89317584e-01 1.74198821e-01
8.78341049e-02 2.43756577e-01 -6.34141207e-01 8.19761693e-01
8.49899411e-01 -4.89213616e-01 -3.53469402e-01 4.97566583e-03
3.65339220e-01 -1.16073310e-01 3.20399046e-01 -1.13062239e+00
-7.95721412e-02 -1.76751390e-01 4.07329440e-01 -2.46329695e-01
2.59587586e-01 -6.91621125e-01 8.77239704e-02 4.19303030e-01
-4.70121384e-01 -2.85337120e-02 -2.75741499e-02 5.66566825e-01
-1.76791385e-01 -4.76303041e-01 9.29670632e-01 1.40244618e-01
-4.90890555e-02 1.67435676e-01 -3.87037545e-01 2.73750544e-01
9.72397506e-01 2.27495164e-01 -5.65359630e-02 -1.02755964e+00
-5.28547704e-01 -9.04555544e-02 3.53594005e-01 1.46660917e-02
5.45261919e-01 -1.24162781e+00 -8.98444653e-01 1.68619886e-01
-7.92920142e-02 -3.47292125e-01 -9.98819023e-02 1.05386639e+00
-2.63099879e-01 3.86108786e-01 2.25721430e-02 -5.05377293e-01
-1.01112318e+00 5.24861813e-01 3.10233891e-01 -1.72705710e-01
-3.87589425e-01 5.13701916e-01 1.51655585e-01 -2.44051814e-01
5.26814982e-02 -2.36289144e-01 -2.71118768e-02 -1.88547924e-01
5.76742172e-01 4.02361989e-01 -6.04596846e-02 -4.73235965e-01
-2.07856059e-01 2.97132552e-01 -9.43388510e-03 -3.01922619e-01
1.30468941e+00 -7.82158133e-03 3.87861626e-03 8.32565248e-01
1.47385287e+00 -6.85838377e-03 -7.30150402e-01 -3.65162134e-01
2.85897171e-04 -3.76164615e-01 4.31195162e-02 -3.22005272e-01
-8.91208172e-01 1.02404189e+00 3.75172108e-01 6.82139874e-01
9.58761096e-01 1.71660930e-01 4.92948353e-01 7.31390476e-01
1.71720952e-01 -1.08686507e+00 -7.67948851e-02 5.28688490e-01
6.37250781e-01 -1.15051472e+00 2.37427488e-01 -3.27887595e-01
-2.81123281e-01 1.22032356e+00 8.37885737e-02 -4.05863523e-01
1.13366985e+00 3.53295445e-01 -4.25940216e-01 1.35785058e-01
-9.34007406e-01 -2.35188469e-01 4.42100197e-01 5.18490851e-01
6.87776327e-01 2.38462001e-01 -4.83656496e-01 4.64479417e-01
-5.16962171e-01 -1.55859679e-01 6.49862885e-01 5.92451513e-01
-4.02147353e-01 -1.06345320e+00 -9.30382535e-02 7.52805531e-01
-6.32602930e-01 -4.39026505e-02 -2.87611216e-01 6.92021549e-01
-2.63096571e-01 7.95279145e-01 2.53603965e-01 -1.78462654e-01
-2.72085309e-01 -5.25204577e-02 6.90852106e-01 -2.90198922e-01
-4.66964878e-02 -1.47397354e-01 -3.00836638e-02 -3.25756788e-01
-6.32136464e-01 -9.69425678e-01 -7.87773013e-01 -5.41371167e-01
-7.68767297e-01 4.11918104e-01 7.72902310e-01 1.13070369e+00
2.40901351e-01 2.75630236e-01 5.93161941e-01 -2.25229442e-01
-9.31236327e-01 -1.21405327e+00 -9.00340676e-01 1.46804094e-01
3.19674253e-01 -7.61838377e-01 -6.02778375e-01 -2.67627448e-01] | [7.430787086486816, 4.031768798828125] |
fbfed48d-4aab-4215-818d-68416d97df48 | gentle-a-genre-diverse-multilayer-challenge | 2306.01966 | null | https://arxiv.org/abs/2306.01966v1 | https://arxiv.org/pdf/2306.01966v1.pdf | GENTLE: A Genre-Diverse Multilayer Challenge Set for English NLP and Linguistic Evaluation | We present GENTLE, a new mixed-genre English challenge corpus totaling 17K tokens and consisting of 8 unusual text types for out-of domain evaluation: dictionary entries, esports commentaries, legal documents, medical notes, poetry, mathematical proofs, syllabuses, and threat letters. GENTLE is manually annotated for a variety of popular NLP tasks, including syntactic dependency parsing, entity recognition, coreference resolution, and discourse parsing. We evaluate state-of-the-art NLP systems on GENTLE and find severe degradation for at least some genres in their performance on all tasks, which indicates GENTLE's utility as an evaluation dataset for NLP systems. | ['Amir Zeldes', 'YIlun Zhu', 'Siyao Peng', 'Yang Janet Liu', 'Jessica Lin', 'Lauren Levine', 'Luke Gessler', 'Shabnam Behzad', 'Tatsuya Aoyama'] | 2023-06-03 | null | null | null | null | ['mathematical-proofs', 'discourse-parsing', 'dependency-parsing', 'coreference-resolution'] | ['miscellaneous', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [-7.78588876e-02 4.41544801e-01 -6.77082360e-01 -2.11974174e-01
-1.12353158e+00 -1.07707238e+00 7.29149282e-01 5.69191337e-01
-7.87540793e-01 1.23192501e+00 8.57489765e-01 -5.21729469e-01
-6.56920299e-02 -1.61972448e-01 -3.76683295e-01 -3.20828021e-01
-2.37648010e-01 9.72757220e-01 1.17319122e-01 -5.41937113e-01
2.90845841e-01 9.61326286e-02 -4.86859262e-01 6.08597279e-01
6.11685574e-01 4.96255875e-01 -3.45418751e-01 6.40458226e-01
-1.30038217e-01 1.11764729e+00 -8.81313324e-01 -1.07946873e+00
-3.23556185e-01 8.79368410e-02 -1.32738197e+00 -5.72460175e-01
4.39937353e-01 1.48907050e-01 -3.43594462e-01 7.62182176e-01
8.53853226e-01 -4.31480557e-02 6.53112888e-01 -8.61992478e-01
-6.85383856e-01 1.21225941e+00 -2.74873495e-01 6.50608003e-01
1.03283584e+00 -1.06780827e-01 1.45121777e+00 -7.42228389e-01
1.53930771e+00 1.13212681e+00 7.77575910e-01 8.49564612e-01
-9.95832562e-01 -8.06770921e-01 -2.91954756e-01 2.01389432e-01
-1.04280269e+00 -8.57596934e-01 5.94190121e-01 -2.01951623e-01
1.84463298e+00 7.66936988e-02 2.43302993e-03 1.81810963e+00
3.46873850e-01 9.79529202e-01 6.90491199e-01 -5.32598615e-01
-2.10500099e-02 3.18270400e-02 5.35619974e-01 5.23626328e-01
3.60880122e-02 -1.71966180e-01 -8.62944722e-01 -4.54576284e-01
8.51708837e-03 -1.12249756e+00 -3.63311470e-01 4.37750220e-01
-1.32364559e+00 6.43098831e-01 -3.46443415e-01 2.91002125e-01
-4.40980047e-02 -3.85314196e-01 1.15278649e+00 3.95039499e-01
4.05353099e-01 9.37522650e-01 -9.62043762e-01 -7.11265683e-01
-3.91819417e-01 6.24250948e-01 1.57562613e+00 1.05271089e+00
-2.57178485e-01 -4.33756202e-01 -1.30809218e-01 1.08267522e+00
-2.42762025e-02 3.95535350e-01 5.42875946e-01 -9.19972897e-01
1.37086034e+00 4.53012526e-01 2.53140628e-02 -8.57388675e-01
-7.77986646e-01 1.44566596e-01 -6.21028125e-01 -6.91421628e-01
6.98497653e-01 -4.39811975e-01 -4.44293946e-01 1.64925086e+00
2.69611359e-01 -1.53994560e-01 7.65648484e-01 6.33615851e-01
1.81950736e+00 6.36316657e-01 5.90101480e-01 -3.85908812e-01
1.64296961e+00 -8.72450173e-01 -8.19040954e-01 -3.10254365e-01
7.28079021e-01 -1.09299946e+00 8.81156564e-01 3.93489718e-01
-1.31135833e+00 -3.77454199e-02 -6.38289511e-01 -5.69490314e-01
-3.18692893e-01 -1.93586908e-02 5.45514286e-01 2.11918637e-01
-3.61334473e-01 4.71252888e-01 -5.61010718e-01 -3.32914591e-01
3.59932333e-01 9.78166461e-02 -6.89608753e-01 1.05676815e-01
-1.45401716e+00 9.58072543e-01 7.79618680e-01 -4.31814015e-01
-2.49198183e-01 -9.53391790e-01 -1.18748450e+00 4.60886862e-03
4.53398168e-01 -2.44265199e-01 1.40466464e+00 -9.71863866e-02
-1.29262793e+00 1.58712530e+00 -3.27443369e-02 -5.35579801e-01
3.32428515e-01 -4.56712008e-01 -9.91892874e-01 1.50695622e-01
3.29232693e-01 3.45098943e-01 2.86837310e-01 -3.68508011e-01
-5.67665100e-01 4.93087694e-02 6.36636391e-02 2.62582719e-01
-6.25942275e-02 7.57147729e-01 -4.75390881e-01 -7.83128321e-01
-4.82467532e-01 -7.06113577e-01 1.11001551e-01 -7.76509166e-01
-9.52210128e-01 -8.88294697e-01 6.14781320e-01 -8.39391232e-01
1.46200776e+00 -2.31054735e+00 7.01516271e-02 -2.09362909e-01
5.28202690e-02 1.07777335e-01 -8.14961940e-02 4.68394399e-01
-1.45613477e-02 2.98872113e-01 4.98637296e-02 -4.59903240e-01
2.75534272e-01 3.94212842e-01 -4.00989681e-01 1.96556374e-01
3.21378946e-01 9.84097421e-01 -9.79070306e-01 -1.03360128e+00
-2.02784091e-01 1.32210195e-01 -4.07641232e-01 1.45295002e-02
-4.59634542e-01 3.29569936e-01 -6.76096201e-01 8.76755953e-01
1.10972010e-01 -2.20246479e-01 4.56859082e-01 -2.10228249e-01
-2.45476700e-02 1.43258274e+00 -8.29110026e-01 1.76732361e+00
-1.82571616e-02 7.16011405e-01 1.21137224e-01 -5.94508111e-01
4.45770979e-01 6.06259525e-01 2.18256712e-01 -6.37334049e-01
3.01318884e-01 2.32952356e-01 5.77769205e-02 -7.45557487e-01
8.68610084e-01 4.41326462e-02 -8.17724884e-01 3.92066896e-01
2.72718459e-01 4.96408157e-02 7.58508205e-01 4.13542449e-01
1.35615921e+00 -6.28830492e-02 7.92161763e-01 -3.78518343e-01
4.07998860e-01 6.47947788e-01 7.48920739e-01 5.64706504e-01
-2.26190567e-01 2.98119962e-01 7.63423920e-01 -3.15483034e-01
-1.02537155e+00 -9.29830790e-01 -5.95131695e-01 1.34151578e+00
-1.30095601e-01 -8.52711916e-01 -4.19022441e-01 -1.08573151e+00
-1.05219141e-01 5.63873947e-01 -4.02921498e-01 3.69203955e-01
-1.17091393e+00 -9.74823475e-01 1.37979269e+00 5.86673856e-01
5.27990580e-01 -1.56945980e+00 -3.50999922e-01 4.39087689e-01
-5.74496090e-01 -1.71891987e+00 -5.01036882e-01 3.49468499e-01
-3.38472426e-01 -1.63649035e+00 -2.99531817e-01 -1.13207531e+00
1.84982449e-01 -7.39313662e-01 1.63507748e+00 -3.34314883e-01
-5.42256981e-02 1.39699444e-01 -4.68695253e-01 -4.55182433e-01
-8.84467185e-01 4.13170099e-01 7.90341720e-02 -9.52593207e-01
7.08274066e-01 -1.88480660e-01 1.17035890e-02 1.71336774e-02
-2.15782359e-01 -3.18135023e-01 3.43995690e-01 1.05062664e+00
4.63149339e-01 -3.56033593e-01 4.41499770e-01 -1.41653979e+00
1.10411251e+00 -5.99976361e-01 -1.37264162e-01 2.65904486e-01
-2.43718317e-03 -8.73151049e-02 7.30487943e-01 -6.29021466e-01
-1.30499291e+00 -4.33002084e-01 -4.15596992e-01 1.36817291e-01
-2.83023477e-01 7.37354815e-01 -1.24415286e-01 3.65616173e-01
1.00382519e+00 -9.49555561e-02 -4.19898003e-01 -7.08956718e-01
9.46255326e-02 7.54002750e-01 1.23094249e+00 -1.04380810e+00
2.36757115e-01 -1.21707849e-01 -2.82815129e-01 -8.64699900e-01
-1.02874851e+00 -6.20418966e-01 -4.30325329e-01 5.67838490e-01
7.25499153e-01 -8.57123315e-01 -9.80834484e-01 2.35938802e-01
-1.46222150e+00 -3.62854898e-01 -2.83175796e-01 4.12485898e-01
-1.47159457e-01 3.97253931e-01 -1.22834086e+00 -1.53665125e-01
-7.83492446e-01 -6.61730826e-01 1.03062999e+00 1.85981721e-01
-7.43538022e-01 -1.21469891e+00 3.66971642e-01 4.64991003e-01
-2.25910723e-01 3.66552353e-01 1.13516140e+00 -1.33701146e+00
3.12129378e-01 -1.86057128e-02 -8.48027095e-02 -5.02096675e-02
-2.46595800e-01 -4.71791513e-02 -5.42891085e-01 -1.13686264e-01
-5.09103417e-01 -8.91146243e-01 5.50254881e-01 -1.12811103e-02
6.20295227e-01 -6.42033577e-01 -6.60133183e-01 4.29309458e-01
8.89934003e-01 1.44973874e-01 5.70967972e-01 5.94042599e-01
4.43475991e-01 3.14452946e-01 5.84242523e-01 3.71478319e-01
5.65154314e-01 5.37717938e-01 -3.29019189e-01 2.80069679e-01
-4.05724011e-02 -7.82806799e-02 3.86985958e-01 9.05130744e-01
-3.06892041e-02 -5.12187660e-01 -1.24157333e+00 7.94935822e-01
-1.65421891e+00 -1.06902409e+00 3.55974026e-02 1.35425842e+00
1.63537502e+00 5.56853890e-01 -1.31334085e-02 1.21059446e-02
3.63940448e-01 1.73343852e-01 -2.02708960e-01 -8.14434946e-01
-4.34675992e-01 5.08299172e-01 3.78519237e-01 2.41564527e-01
-1.54792666e+00 1.26166904e+00 7.03998375e+00 8.74918044e-01
-8.99304867e-01 9.16709900e-02 3.51918608e-01 1.66887820e-01
2.08862409e-01 -3.55522573e-01 -1.32816064e+00 4.64630812e-01
9.55572248e-01 -3.01705271e-01 -1.53473437e-01 9.03068304e-01
-3.80560875e-01 8.74289423e-02 -1.32567573e+00 7.03548670e-01
2.90119611e-02 -1.70421684e+00 -2.24087358e-01 -3.34539324e-01
4.64356124e-01 4.93937224e-01 -2.83998549e-01 7.38956988e-01
6.51567638e-01 -9.81474161e-01 4.80715245e-01 -1.94885701e-01
9.97202039e-01 -4.85294580e-01 9.19243872e-01 3.33753824e-01
-7.26351261e-01 2.20363498e-01 -2.18293533e-01 8.27787220e-02
4.51864362e-01 2.86482096e-01 -9.87409234e-01 4.11838949e-01
7.34918714e-01 8.25095177e-01 -2.44441852e-01 6.61583602e-01
-5.48654497e-01 8.83120418e-01 -6.40970886e-01 -1.48431629e-01
3.21806461e-01 5.49219847e-01 9.80901599e-01 1.99222970e+00
-4.55514014e-01 6.18638217e-01 3.30620259e-01 3.87381673e-01
-6.66133881e-01 3.35200191e-01 -3.06432486e-01 -3.02568197e-01
8.34083200e-01 1.08822834e+00 -4.34899449e-01 -3.87945175e-01
-4.28432375e-01 5.48683167e-01 4.28760976e-01 4.71934341e-02
-6.70650005e-01 -6.70068860e-01 5.84545135e-01 -2.07871541e-01
7.87531883e-02 -1.50223911e-01 -3.46905105e-02 -1.29478455e+00
-1.14914462e-01 -1.14211404e+00 9.44548428e-01 -3.25814664e-01
-1.60852277e+00 6.49983466e-01 1.91941306e-01 -9.30970371e-01
-6.02900684e-01 -8.19507539e-01 -6.94791019e-01 4.97088760e-01
-1.53073072e+00 -1.01937008e+00 2.33754471e-01 7.26885080e-01
6.27974212e-01 -5.40538907e-01 1.13943696e+00 5.62714279e-01
-6.83634937e-01 9.44130599e-01 -3.84888351e-02 9.39582705e-01
1.19283056e+00 -1.27125657e+00 6.86412334e-01 5.54738104e-01
1.42366216e-01 4.67299700e-01 8.57702076e-01 -6.60625875e-01
-1.21983540e+00 -7.47060120e-01 1.62611747e+00 -6.59511209e-01
9.76840734e-01 -2.37442359e-01 -8.39450657e-01 8.64322126e-01
3.69547725e-01 -1.51236534e-01 8.96304488e-01 6.77536488e-01
-5.28045297e-01 6.35877728e-01 -1.21044970e+00 5.40601969e-01
1.03578663e+00 -3.76843810e-01 -1.37025559e+00 5.39635599e-01
7.18249917e-01 -1.31112480e+00 -1.21721160e+00 5.11483431e-01
5.07123470e-01 -9.99951437e-02 9.39311206e-01 -1.12045300e+00
8.60280097e-01 3.64328623e-01 1.15271565e-02 -1.05053329e+00
-4.80875969e-02 -8.79164338e-01 -2.91559309e-01 1.64655590e+00
9.32310879e-01 -3.20237815e-01 7.43660688e-01 7.33412683e-01
-3.77525598e-01 -4.82138902e-01 -1.24619305e+00 -4.87153113e-01
3.39370638e-01 -4.88371909e-01 1.51652634e-01 1.56231153e+00
1.04283321e+00 1.07320201e+00 4.05597761e-02 -2.29303420e-01
4.47561026e-01 8.46152008e-02 5.11391044e-01 -9.88030374e-01
-3.92789721e-01 -4.26424265e-01 -2.57930625e-02 -1.00644207e+00
7.05137968e-01 -1.00083470e+00 -1.64289042e-01 -1.19048309e+00
1.38749510e-01 -5.53751945e-01 2.29953542e-01 9.35772181e-01
-3.66812199e-02 1.14825048e-01 9.97164994e-02 1.94633812e-01
-9.74292755e-01 -6.50994480e-02 9.58656192e-01 -3.21232647e-01
-3.28565687e-01 -3.85918617e-01 -7.52171278e-01 8.39244127e-01
5.46094120e-01 -5.84782243e-01 2.57589340e-01 -4.71273661e-01
3.34851831e-01 1.65868625e-01 -2.13296339e-01 -3.58789593e-01
2.81338453e-01 -2.34438911e-01 1.62801832e-01 -3.87244910e-01
1.10416718e-01 -2.28989422e-01 -4.52468127e-01 1.30782694e-01
-4.92103338e-01 6.11075871e-02 5.84681034e-01 3.02536637e-02
-3.54327738e-01 -2.09152490e-01 5.82252741e-01 -1.64107159e-01
-8.48519802e-01 -9.54295844e-02 -2.07595512e-01 1.22600937e+00
6.27068818e-01 1.04016207e-01 -1.02253973e+00 1.34991750e-01
-9.28852737e-01 6.26453102e-01 -6.35308474e-02 6.46326780e-01
3.37192774e-01 -8.45252335e-01 -1.10545206e+00 -1.75328255e-01
7.22133741e-02 9.59122479e-02 -2.60587692e-01 6.47395611e-01
-5.11621714e-01 5.48003018e-01 -1.06903836e-01 -1.18439816e-01
-1.64315259e+00 1.99961379e-01 -1.37751698e-01 -9.52365637e-01
-9.68904853e-01 9.66629684e-01 -2.77370840e-01 -4.37962770e-01
4.18741226e-01 -3.66720945e-01 -5.52476048e-01 1.35958895e-01
5.87049484e-01 2.58403838e-01 -7.12219393e-03 -6.75221860e-01
-7.96138942e-01 1.77949488e-01 -5.03590465e-01 1.81522265e-01
1.24907339e+00 5.49132116e-02 -3.47413472e-03 8.81670639e-02
1.00205922e+00 3.95177901e-01 -4.65159893e-01 -3.54825765e-01
4.91822302e-01 1.64462060e-01 -4.98197794e-01 -1.30285263e+00
-4.25117493e-01 3.54838699e-01 -4.03608263e-01 5.91581687e-03
7.17324913e-01 3.38418782e-01 1.08415270e+00 8.68166983e-01
2.74307132e-01 -1.32895923e+00 -2.76890606e-01 1.19155598e+00
8.56905282e-01 -1.30246329e+00 2.10574374e-01 -4.85037416e-01
-9.35790002e-01 1.07714558e+00 4.30253744e-01 -3.30742747e-02
6.78484678e-01 8.22995901e-01 -3.13069895e-02 -3.53688359e-01
-1.01858330e+00 2.33560666e-01 2.05121771e-01 4.12489951e-01
7.57952392e-01 -1.06745876e-01 -7.73970783e-01 1.34774709e+00
-5.94348848e-01 -2.61951476e-01 5.53704441e-01 9.06166553e-01
1.03842445e-01 -1.04356098e+00 -1.47978468e-02 1.88529834e-01
-1.32895529e+00 -5.40739298e-01 -6.36240780e-01 1.21994257e+00
-1.02486826e-01 6.19545221e-01 7.41110519e-02 3.12482435e-02
5.37664473e-01 1.41750425e-01 4.47157621e-01 -7.33926415e-01
-1.02888405e+00 -1.29218936e-01 1.25619495e+00 -4.21073675e-01
-6.84676468e-01 -8.33844960e-01 -1.55624974e+00 -5.64717472e-01
-1.78843945e-01 5.15523255e-01 2.34310508e-01 1.05864286e+00
2.26515174e-01 2.31150836e-01 -1.74798131e-01 -4.37803388e-01
-3.66879463e-01 -1.11084270e+00 -1.99591354e-01 7.79115558e-01
1.21876635e-01 -5.37376761e-01 -1.43573061e-01 2.00023293e-01] | [9.432639122009277, 9.410720825195312] |
ab7c4818-467f-49c1-88cc-473e0364b745 | joint-semantic-mining-for-weakly-supervised | null | null | http://proceedings.neurips.cc/paper/2021/hash/642e92efb79421734881b53e1e1b18b6-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/642e92efb79421734881b53e1e1b18b6-Paper.pdf | Joint Semantic Mining for Weakly Supervised RGB-D Salient Object Detection | Training saliency detection models with weak supervisions, e.g., image-level tags or captions, is appealing as it removes the costly demand of per-pixel annotations. Despite the rapid progress of RGB-D saliency detection in fully-supervised setting, it however remains an unexplored territory when only weak supervision signals are available. This paper is set to tackle the problem of weakly-supervised RGB-D salient object detection. The key insight in this effort is the idea of maintaining per-pixel pseudo-labels with iterative refinements by reconciling the multimodal input signals in our joint semantic mining (JSM). Considering the large variations in the raw depth map and the lack of explicit pixel-level supervisions, we propose spatial semantic modeling (SSM) to capture saliency-specific depth cues from the raw depth and produce depth-refined pseudo-labels. Moreover, tags and captions are incorporated via a fill-in-the-blank training in our textual semantic modeling (TSM) to estimate the confidences of competing pseudo-labels. At test time, our model involves only a light-weight sub-network of the training pipeline, i.e., it requires only an RGB image as input, thus allowing efficient inference. Extensive evaluations demonstrate the effectiveness of our approach under the weakly-supervised setting. Importantly, our method could also be adapted to work in both fully-supervised and unsupervised paradigms. In each of these scenarios, superior performance has been attained by our approach with comparing to the state-of-the-art dedicated methods. As a by-product, a CapS dataset is constructed by augmenting existing benchmark training set with additional image tags and captions. | ['Li Cheng', 'Huchuan Lu', 'Yongri Piao', 'Miao Zhang', 'Cheng Yan', 'Qi Bi', 'Wei Ji', 'Jingjing Li'] | 2021-12-01 | null | https://openreview.net/forum?id=mv-1sL8FMN5 | https://openreview.net/pdf?id=mv-1sL8FMN5 | neurips-2021-12 | ['rgb-d-salient-object-detection'] | ['computer-vision'] | [ 7.41076589e-01 6.01424456e-01 -2.12224215e-01 -5.63967466e-01
-8.70739818e-01 -3.11721265e-01 6.09956026e-01 2.13063046e-01
-5.08743942e-01 5.56580842e-01 2.06465293e-02 -1.23946562e-01
9.06571895e-02 -4.56911236e-01 -1.01027489e+00 -6.53198779e-01
3.22773159e-01 2.48554304e-01 7.72924781e-01 -1.13818318e-01
1.86709568e-01 1.94734976e-01 -1.97206759e+00 2.51592427e-01
8.65555346e-01 1.29798436e+00 7.36315489e-01 3.15537512e-01
-7.73140788e-02 7.08140850e-01 -6.71173558e-02 -3.18458140e-01
2.67724097e-01 -2.88357466e-01 -7.05857575e-01 4.60572422e-01
5.25898039e-01 -2.17972875e-01 -9.35625136e-02 1.03663945e+00
2.51784056e-01 -1.70145288e-01 2.72286355e-01 -1.21526921e+00
-2.39868999e-01 3.97318840e-01 -7.46767998e-01 -1.28218949e-01
1.67630270e-01 6.24134615e-02 1.04458058e+00 -1.05561471e+00
4.95965719e-01 8.40186417e-01 4.73722488e-01 4.73491132e-01
-1.34903514e+00 -2.62618959e-01 4.69791263e-01 2.31391713e-02
-1.13236141e+00 -4.29112881e-01 1.02271855e+00 -1.33022487e-01
7.42462873e-01 2.16441080e-01 5.66650271e-01 1.05532372e+00
-3.49114984e-01 1.26114595e+00 1.29245067e+00 -6.17861986e-01
4.40435261e-01 4.82766807e-01 -1.67014226e-01 8.49339664e-01
1.64944261e-01 -6.50731474e-02 -1.00114536e+00 6.55080453e-02
8.20988894e-01 2.56873630e-02 -1.46543905e-01 -9.20219898e-01
-1.29019070e+00 6.36229813e-01 5.88650763e-01 1.63948581e-01
-1.95428967e-01 2.01069936e-01 8.27954561e-02 -1.91709086e-01
6.73279881e-01 2.38410994e-01 -5.55169761e-01 1.84393376e-01
-1.45645273e+00 -2.83478294e-02 3.72248054e-01 9.74213719e-01
1.11109245e+00 -1.90527260e-01 -9.42731947e-02 5.71823120e-01
4.10189450e-01 5.11316061e-01 1.87243804e-01 -9.40398872e-01
4.21624750e-01 6.51197076e-01 2.81763136e-01 -7.46210098e-01
-4.76262540e-01 -8.07138443e-01 -3.94919813e-01 3.27345222e-01
4.33720320e-01 1.92707866e-01 -1.11224937e+00 1.86297429e+00
5.11813462e-01 3.86388123e-01 -8.23629722e-02 1.18656075e+00
5.69176197e-01 1.48201734e-01 4.68935855e-02 1.80676039e-02
1.32734263e+00 -1.09004736e+00 -3.71150643e-01 -6.82930827e-01
4.41983134e-01 -6.44110739e-01 1.18033409e+00 1.41932607e-01
-1.13282001e+00 -4.23682004e-01 -1.02113664e+00 -2.37005502e-01
-3.08709174e-01 3.05840552e-01 8.40282559e-01 5.42757630e-01
-1.20559406e+00 3.83496761e-01 -1.06541824e+00 -3.75598460e-01
5.60641110e-01 3.72512132e-01 -2.25024253e-01 -1.06229037e-02
-1.03110242e+00 7.84174502e-01 3.32411110e-01 7.59737641e-02
-9.54057753e-01 -6.56047642e-01 -1.07509625e+00 -1.55214116e-01
7.14504302e-01 -7.10578203e-01 1.19230044e+00 -1.14078784e+00
-1.36325014e+00 1.12160146e+00 -3.26115370e-01 -4.00893927e-01
5.90176702e-01 -1.65791288e-01 9.51352715e-02 4.02104765e-01
2.94497401e-01 1.12648797e+00 9.27160442e-01 -1.61122739e+00
-9.30852711e-01 -3.30532283e-01 3.12319994e-01 2.25386962e-01
-3.23820233e-01 -2.39567593e-01 -7.95072496e-01 -4.98091877e-01
3.95142317e-01 -9.10512745e-01 -2.82789022e-01 2.37674892e-01
-4.94500905e-01 1.48509234e-01 6.18565619e-01 -3.14129114e-01
6.87171936e-01 -2.13334060e+00 2.30826572e-01 8.80430490e-02
7.72212893e-02 5.14043532e-02 3.83561626e-02 4.10107970e-02
1.67494923e-01 -2.74696976e-01 -7.26057827e-01 -9.65255737e-01
1.07275516e-01 2.27186456e-01 -3.68012428e-01 4.81903881e-01
6.73953056e-01 1.06385744e+00 -1.16035557e+00 -7.38301337e-01
4.91268575e-01 4.30747360e-01 -5.74712396e-01 1.30831867e-01
-5.22000909e-01 5.28179407e-01 -4.83987957e-01 8.79072309e-01
6.28399670e-01 -5.17638981e-01 -4.77522686e-02 -2.42600277e-01
-2.88788706e-01 4.42661732e-01 -1.11790884e+00 2.41894245e+00
-4.40224379e-01 3.85349631e-01 1.22269824e-01 -1.00801158e+00
6.95975006e-01 -4.11629817e-03 3.58045250e-01 -8.19678068e-01
6.58369740e-04 4.89565909e-01 -4.68398541e-01 -1.71899542e-01
4.78672266e-01 -1.16365790e-01 -5.78252189e-02 4.18152750e-01
2.31669724e-01 -2.58812308e-01 -1.06334507e-01 2.48586684e-01
9.55049336e-01 6.46011114e-01 -6.43102406e-03 -1.49787158e-01
3.71524036e-01 1.03791252e-01 4.59053665e-01 8.61203551e-01
-8.72805342e-02 9.72199202e-01 3.57609212e-01 -2.32271454e-03
-9.77982581e-01 -1.15035117e+00 -1.90571725e-01 1.06510365e+00
5.45278549e-01 -1.81292117e-01 -6.20024741e-01 -7.68699527e-01
-6.54557198e-02 5.77786744e-01 -6.38820529e-01 -5.52146919e-02
-2.44801223e-01 -7.03896224e-01 2.33138815e-01 4.71537858e-01
5.00774324e-01 -9.89242256e-01 -1.07790315e+00 3.25231217e-02
-1.63351614e-02 -1.42509353e+00 -5.92802092e-02 6.80848956e-01
-9.18084741e-01 -7.82367110e-01 -7.45049179e-01 -8.29448938e-01
7.88698256e-01 5.90459228e-01 1.01699173e+00 -1.56817526e-01
2.15740548e-03 3.71867031e-01 -3.31569344e-01 -3.37550670e-01
7.28475675e-02 1.54707581e-01 -2.24748522e-01 1.91181093e-01
1.50184378e-01 -6.32871509e-01 -9.10364151e-01 3.67099166e-01
-9.90504980e-01 6.19110107e-01 8.38202715e-01 7.21321881e-01
7.56252170e-01 -3.21654409e-01 5.01694739e-01 -8.35702002e-01
-2.28766114e-01 -3.33463758e-01 -5.46913981e-01 1.10281318e-01
-5.64610898e-01 1.50470659e-01 2.15693489e-01 -1.18583038e-01
-1.11621940e+00 5.27390778e-01 5.35133742e-02 -4.77047741e-01
-2.28774264e-01 3.49660903e-01 -2.39593938e-01 -6.41135797e-02
3.37408751e-01 3.64401609e-01 -1.62871759e-02 -5.27446747e-01
5.27907968e-01 5.97795844e-01 7.37333298e-01 -3.88074756e-01
8.53461385e-01 8.94163609e-01 2.68506780e-02 -4.89466846e-01
-1.41297054e+00 -6.08739316e-01 -7.24362791e-01 -1.37685329e-01
6.87112808e-01 -1.27055955e+00 -2.90283471e-01 4.68963861e-01
-9.96638119e-01 -4.39910620e-01 -2.60087371e-01 2.57935375e-01
-7.17389703e-01 3.71337801e-01 -3.53654802e-01 -8.85116458e-01
-2.18527596e-02 -1.07279456e+00 1.69962180e+00 1.97308525e-01
-2.53071226e-02 -8.25272679e-01 -3.46287400e-01 5.42732060e-01
3.95489872e-01 2.62559503e-01 5.76585293e-01 -3.59823525e-01
-9.39438760e-01 -4.29730937e-02 -5.69538414e-01 2.81643897e-01
7.09298998e-02 -5.36936581e-01 -1.46424687e+00 -1.61981452e-02
1.60636485e-03 -4.95385170e-01 1.09071660e+00 2.08572164e-01
1.18305016e+00 9.81608927e-02 -2.71037161e-01 5.54126143e-01
1.29720914e+00 -6.08814538e-01 3.55269313e-01 4.96717930e-01
8.05050373e-01 7.50724018e-01 8.51926208e-01 4.03786868e-01
6.15485311e-01 8.29502583e-01 8.61683011e-01 -5.71573436e-01
-2.48570204e-01 -4.59455162e-01 2.26540864e-01 2.46206760e-01
2.03700840e-01 -5.48836216e-02 -7.76059866e-01 6.87000513e-01
-2.10598969e+00 -6.29762352e-01 -1.25839844e-01 2.16902065e+00
1.08981490e+00 4.23251688e-01 1.72068737e-02 2.94433653e-01
4.32479501e-01 2.31751159e-01 -7.18795955e-01 2.31289983e-01
-3.72471154e-01 2.30153248e-01 6.27479613e-01 5.13248205e-01
-1.23618221e+00 1.07771063e+00 4.88854504e+00 7.80641377e-01
-1.20948255e+00 3.12143862e-01 7.14947939e-01 -3.79125178e-01
-5.62428534e-01 2.30288982e-01 -7.05752850e-01 4.75544423e-01
6.43163204e-01 5.39634705e-01 1.88921943e-01 7.59770393e-01
2.64263690e-01 -6.15807712e-01 -1.14452124e+00 9.34118629e-01
2.42758542e-01 -1.13648427e+00 -1.68197379e-01 -6.60962388e-02
7.31062293e-01 2.95537740e-01 2.15471476e-01 -7.27934465e-02
-1.98481418e-02 -6.79386020e-01 1.26889634e+00 3.76124471e-01
8.46547365e-01 -3.93727690e-01 7.79353678e-01 3.65873396e-01
-8.94001842e-01 2.53246874e-02 -1.94805369e-01 -9.04662386e-02
3.10161322e-01 9.22076344e-01 -6.59820676e-01 5.20970345e-01
6.91642046e-01 8.33698928e-01 -7.34389782e-01 8.40445280e-01
-4.83933628e-01 4.36795980e-01 -5.20176768e-01 1.23119995e-01
4.17784572e-01 6.65288270e-02 3.49984825e-01 1.16702843e+00
8.11154917e-02 -1.88169643e-01 1.39129609e-01 1.03259301e+00
1.65415347e-01 -1.51958480e-01 -2.05246195e-01 2.70168573e-01
2.33203709e-01 1.34457898e+00 -9.48153853e-01 -1.30142063e-01
-5.08807778e-01 1.11270893e+00 4.16143268e-01 2.71947354e-01
-8.50099027e-01 5.09591177e-02 4.05032337e-01 3.46852392e-01
6.50628746e-01 -1.13460854e-01 -6.87909424e-01 -1.06226552e+00
2.68062234e-01 -3.50968957e-01 1.39317319e-01 -9.02126908e-01
-9.26548839e-01 3.48322302e-01 1.31910043e-02 -1.24449170e+00
-5.58524132e-02 -5.51625252e-01 -2.61963427e-01 8.22348475e-01
-2.24050760e+00 -1.55490291e+00 -5.57963312e-01 5.19028783e-01
4.24718201e-01 3.56839597e-01 5.12578309e-01 1.89261943e-01
-3.42235923e-01 4.31354076e-01 -2.45204851e-01 -3.09141546e-01
6.64422810e-01 -1.35576606e+00 1.23426817e-01 1.01207447e+00
2.73497224e-01 3.81527722e-01 9.19115841e-01 -4.26261276e-01
-1.23096621e+00 -1.10697377e+00 8.36876750e-01 -6.16487503e-01
7.29420245e-01 -7.80030012e-01 -7.04058588e-01 3.77317816e-01
-1.04055561e-01 2.50177145e-01 4.75825042e-01 -1.42891571e-01
-1.86075553e-01 -2.05870867e-02 -8.59630466e-01 4.20997977e-01
1.22800851e+00 -7.05220342e-01 -4.98914927e-01 1.93385854e-01
9.79824305e-01 -5.24045885e-01 -3.43693465e-01 5.19234359e-01
2.92376965e-01 -1.25965786e+00 1.03694630e+00 6.06737426e-03
5.73603034e-01 -7.68959343e-01 -2.49958783e-01 -8.14444661e-01
1.29529864e-01 -3.91042739e-01 -1.93747863e-01 1.32138348e+00
5.44101477e-01 -2.45204926e-01 9.69640315e-01 6.11147523e-01
-4.03063834e-01 -8.42288136e-01 -1.00348282e+00 -3.56103301e-01
-5.18309712e-01 -6.80138826e-01 2.86677539e-01 6.71252191e-01
-1.06482022e-01 1.91489592e-01 -4.00732636e-01 3.77441674e-01
8.63608479e-01 3.73068452e-01 6.03274763e-01 -1.09524679e+00
-4.51476961e-01 -2.32897252e-01 -3.65403980e-01 -1.30962849e+00
1.97500110e-01 -9.06247020e-01 2.40891621e-01 -1.44102764e+00
2.39388213e-01 -6.89150035e-01 -5.89739740e-01 7.76661694e-01
-3.04754466e-01 6.20304585e-01 4.72803451e-02 2.12748542e-01
-1.06493604e+00 7.81606138e-01 9.64457691e-01 -1.45858333e-01
-1.55607566e-01 -1.60757139e-01 -7.45866537e-01 7.67168760e-01
6.07527554e-01 -4.46623683e-01 -5.07888615e-01 -4.69272017e-01
2.58622080e-01 -1.53133735e-01 7.96059132e-01 -8.73545051e-01
3.96290898e-01 6.08832538e-02 2.76383787e-01 -6.62622273e-01
5.11609256e-01 -8.39900732e-01 -2.44837865e-01 1.14857912e-01
-2.79634684e-01 -5.83785355e-01 9.09201428e-02 7.28570938e-01
-1.88039109e-01 -1.68647513e-01 6.53478622e-01 -6.37433752e-02
-7.78302729e-01 1.26873136e-01 2.48531118e-01 -3.95410918e-02
8.17622125e-01 -4.93368924e-01 -1.37591943e-01 -1.16986491e-01
-5.46281278e-01 2.00790092e-01 9.09095109e-01 3.82121503e-01
4.88304675e-01 -1.02134073e+00 -3.46351802e-01 2.07442403e-01
4.52555925e-01 5.27062118e-01 1.31899074e-01 1.16613650e+00
-8.42999145e-02 4.33204383e-01 -3.58678028e-02 -9.49642241e-01
-8.85959923e-01 4.34745699e-01 4.43511754e-02 -3.99749689e-02
-5.17254531e-01 8.40887964e-01 4.22908455e-01 -1.74238071e-01
4.06116635e-01 -4.03666317e-01 9.97641310e-02 -9.19115320e-02
3.18269789e-01 -1.86898798e-01 2.07444966e-01 -5.23720801e-01
-5.35359979e-01 4.37116057e-01 2.89363526e-02 -2.62836605e-01
1.51434720e+00 -4.23985928e-01 -2.12895144e-02 6.84692681e-01
8.54838669e-01 -1.05159424e-01 -1.82803738e+00 -4.75741804e-01
1.60055041e-01 -4.72038269e-01 3.67249399e-01 -7.58743763e-01
-9.90157545e-01 8.44684482e-01 4.82499599e-01 -1.95584729e-01
1.13735616e+00 4.26432580e-01 5.43274403e-01 8.31429884e-02
6.22521341e-01 -1.03856134e+00 2.48367757e-01 1.03765704e-01
5.21394610e-01 -1.68718588e+00 -6.79752827e-02 -6.47841275e-01
-6.29588068e-01 6.44268870e-01 4.60658073e-01 3.56220156e-02
3.40826422e-01 9.55758914e-02 -3.92746851e-02 -2.15122715e-01
-6.19170487e-01 -5.86707234e-01 2.86743611e-01 4.71410781e-01
2.13610083e-01 -1.87717155e-01 2.60500181e-02 5.34320951e-01
6.15331307e-02 -4.16415930e-02 3.28175992e-01 1.06850410e+00
-5.37574410e-01 -9.51937437e-01 -1.23462744e-01 1.49375245e-01
-1.90264955e-01 -3.12872946e-01 -3.00227314e-01 5.59072673e-01
2.05221266e-01 9.74409938e-01 -1.30582482e-01 -1.48348659e-01
1.39630198e-01 1.36783405e-03 3.84431511e-01 -7.12379634e-01
-2.38646835e-01 1.60685331e-01 -1.51766658e-01 -7.14003980e-01
-9.22999680e-01 -7.54736423e-01 -1.29505610e+00 3.28647435e-01
-3.71540725e-01 -9.84125957e-02 7.84494221e-01 1.16208696e+00
4.42636997e-01 4.53034759e-01 4.62323666e-01 -1.17516494e+00
-2.28473306e-01 -7.32331514e-01 -4.83333737e-01 4.40060437e-01
5.11038065e-01 -9.29883361e-01 -4.26071703e-01 -2.84313038e-03] | [9.711653709411621, -0.6276580691337585] |
ca3a493f-9ac5-47d4-a1e1-bd4d33b34c01 | priorcvae-scalable-mcmc-parameter-inference | 2304.04307 | null | https://arxiv.org/abs/2304.04307v2 | https://arxiv.org/pdf/2304.04307v2.pdf | PriorCVAE: scalable MCMC parameter inference with Bayesian deep generative modelling | In applied fields where the speed of inference and model flexibility are crucial, the use of Bayesian inference for models with a stochastic process as their prior, e.g. Gaussian processes (GPs) is ubiquitous. Recent literature has demonstrated that the computational bottleneck caused by GP priors or their finite realizations can be encoded using deep generative models such as variational autoencoders (VAEs), and the learned generators can then be used instead of the original priors during Markov chain Monte Carlo (MCMC) inference in a drop-in manner. While this approach enables fast and highly efficient inference, it loses information about the stochastic process hyperparameters, and, as a consequence, makes inference over hyperparameters impossible and the learned priors indistinct. We propose to resolve this issue and disentangle the learned priors by conditioning the VAE on stochastic process hyperparameters. This way, the hyperparameters are encoded alongside GP realisations and can be explicitly estimated at the inference stage. We believe that the new method, termed PriorCVAE, will be a useful tool among approximate inference approaches and has the potential to have a large impact on spatial and spatiotemporal inference in crucial real-life applications. Code showcasing PriorCVAE can be found on GitHub: https://github.com/elizavetasemenova/PriorCVAE | ['Seth Flaxman', 'Max Cairney-Leeming', 'Elizaveta Semenova'] | 2023-04-09 | null | null | null | null | ['bayesian-inference'] | ['methodology'] | [-8.50734580e-03 8.64002779e-02 4.71398048e-02 -1.71824917e-02
-5.30201912e-01 -6.74803495e-01 1.07217658e+00 -1.52478412e-01
-2.73341596e-01 1.01356077e+00 5.53402863e-02 -5.38955927e-01
-3.21202189e-01 -1.07914734e+00 -7.99702346e-01 -1.03211486e+00
1.55907581e-02 8.16819727e-01 1.82916164e-01 1.74180403e-01
1.10724777e-01 5.60478210e-01 -1.57166910e+00 -2.79816926e-01
8.05254459e-01 4.71908152e-01 2.58843839e-01 8.77414227e-01
-1.27952158e-01 5.18030405e-01 -4.06400353e-01 -5.72494149e-01
-4.52822074e-02 -2.33669445e-01 -3.11838418e-01 -3.08906674e-01
-2.55025297e-01 -4.14272249e-01 -2.45877966e-01 9.91095841e-01
3.13746959e-01 2.48608455e-01 9.63694811e-01 -1.02757108e+00
-2.52467036e-01 4.14099813e-01 -4.36671793e-01 1.80759896e-02
9.24277976e-02 4.37729985e-01 7.63209641e-01 -7.31425703e-01
3.44692826e-01 1.34389615e+00 6.71551704e-01 9.78875309e-02
-1.75389576e+00 -6.27589762e-01 -5.78537695e-02 -5.55067994e-02
-1.82452476e+00 -3.97743672e-01 4.75409448e-01 -5.91363490e-01
6.92345083e-01 1.91588819e-01 7.42978215e-01 1.63332760e+00
5.28070390e-01 6.69789553e-01 9.49344158e-01 -1.72998533e-01
7.51318336e-01 -3.56715289e-03 -2.95405656e-01 4.28221166e-01
3.68304014e-01 2.95595706e-01 -4.33510810e-01 -5.39481878e-01
1.13203180e+00 2.46424511e-01 -2.48636395e-01 -4.69980314e-02
-9.00249422e-01 1.08675945e+00 2.69141849e-02 1.27507355e-02
-7.41883576e-01 6.87255859e-01 7.48889148e-02 -2.96508282e-01
4.35498297e-01 4.72244946e-03 -3.29349965e-01 -4.81135517e-01
-1.11270738e+00 5.63327610e-01 1.05705833e+00 7.73370981e-01
9.09294784e-01 -1.13309845e-01 -3.02481651e-01 5.37007570e-01
6.81249976e-01 9.25489426e-01 -2.00820386e-01 -1.23719907e+00
-3.36780958e-02 -9.59556103e-02 5.50719976e-01 -7.96385109e-01
5.35153458e-03 -2.52515882e-01 -9.40998614e-01 9.72649679e-02
4.08787757e-01 -4.07247692e-01 -1.11799228e+00 1.68995988e+00
4.51659113e-01 6.80942953e-01 -2.29301721e-01 5.18827081e-01
2.01816842e-01 9.43377376e-01 3.21017534e-01 5.63230626e-02
1.37938058e+00 -2.23673686e-01 -7.70941913e-01 -2.37242669e-01
-3.94345038e-02 -6.76769674e-01 8.59571457e-01 2.67778844e-01
-9.53596890e-01 -2.58455276e-01 -6.82131708e-01 2.81697303e-01
-2.54861772e-01 -1.40018657e-01 7.14738488e-01 6.58575833e-01
-1.09371901e+00 7.64451265e-01 -1.52701879e+00 -2.21911356e-01
5.30698061e-01 9.91052389e-02 2.68674582e-01 -4.38462794e-02
-1.19313443e+00 7.14622676e-01 3.60899299e-01 4.00991410e-01
-1.15800226e+00 -7.90037334e-01 -6.66727304e-01 2.57110476e-01
4.21075404e-01 -9.01433468e-01 1.14755094e+00 -2.76853859e-01
-1.78403986e+00 1.89986348e-01 -3.12251091e-01 -3.88430834e-01
6.76919043e-01 -2.74268836e-01 -1.23280242e-01 2.51461882e-02
-7.42477179e-02 5.34136772e-01 1.18026936e+00 -1.17133391e+00
-2.81751305e-01 -1.75708090e-03 -1.97819442e-01 -5.39475642e-02
3.87810498e-01 -2.67497092e-01 -6.05683506e-01 -3.12686443e-01
-5.66884875e-02 -1.14254463e+00 -2.33896270e-01 -6.20385483e-02
-3.23506832e-01 -2.29141176e-01 4.49681461e-01 -7.56991625e-01
1.12688673e+00 -1.99478924e+00 1.89071521e-01 3.93448800e-01
7.17537403e-02 6.73923492e-02 2.30385840e-01 8.10120761e-01
5.87641537e-01 1.45524874e-01 -6.58862829e-01 -3.90215665e-01
1.80399433e-01 5.80534935e-01 -4.59099323e-01 4.99121308e-01
3.78888458e-01 8.85373056e-01 -9.04892981e-01 -2.86750495e-01
5.11334896e-01 1.03067541e+00 -5.15512586e-01 8.48348290e-02
-4.15464073e-01 7.40468442e-01 -5.50308347e-01 1.69523150e-01
8.47343266e-01 -3.23271275e-01 2.63198137e-01 1.74053133e-01
-3.43343988e-02 1.18059717e-01 -1.48837328e+00 1.32819879e+00
-5.18018961e-01 5.30861676e-01 7.31007978e-02 -5.48542500e-01
6.39597595e-01 3.35285991e-01 3.46652046e-02 -3.15633118e-02
-8.84798616e-02 8.33455846e-02 -1.21919848e-01 -1.19977698e-01
4.13367659e-01 -2.22819194e-01 1.74871042e-01 5.29214799e-01
2.50851214e-02 -5.33711135e-01 1.33366182e-01 1.35794431e-01
8.56935978e-01 7.01527715e-01 1.62228435e-01 -3.31288338e-01
5.25396280e-02 -3.78760278e-01 4.46634948e-01 1.17352855e+00
2.36445665e-01 5.09395301e-01 6.23213887e-01 -2.41901249e-01
-1.16654336e+00 -1.53018367e+00 -4.69022602e-01 5.75841784e-01
-4.50851470e-01 -4.09907997e-01 -6.05290532e-01 -6.54892251e-02
-2.38063689e-02 1.05227292e+00 -6.13615036e-01 -5.39303385e-02
-3.26293081e-01 -1.08022439e+00 4.97469306e-01 4.73858207e-01
2.14451924e-01 -8.22039843e-01 -6.35451496e-01 4.77264047e-01
-9.33002755e-02 -8.29498768e-01 1.69123530e-01 7.31557384e-02
-8.78666341e-01 -6.69781804e-01 -7.43336260e-01 3.06956917e-01
3.91103029e-01 -4.24199104e-01 1.18559241e+00 -1.91508964e-01
1.64288599e-02 2.83562034e-01 -8.55458602e-02 -4.52578872e-01
-3.88833284e-01 -3.81315574e-02 -8.01789761e-02 2.30387654e-02
2.53259242e-01 -1.05713451e+00 -8.21513951e-01 -4.56770770e-02
-1.02111948e+00 1.45956025e-01 4.73900080e-01 7.12031245e-01
5.19832850e-01 8.15995038e-02 2.13614747e-01 -8.66028309e-01
5.33562303e-01 -6.74933791e-01 -1.00762093e+00 7.80739868e-03
-3.30641448e-01 4.11056697e-01 2.38588214e-01 -3.71726036e-01
-1.38075161e+00 -2.33671352e-01 -3.33086342e-01 -5.74199915e-01
-3.77622813e-01 7.08750844e-01 -8.23147967e-03 5.42227328e-01
4.30881947e-01 1.90981477e-01 -9.66008827e-02 -4.63861942e-01
3.79835397e-01 4.44005787e-01 3.45443010e-01 -8.79195929e-01
6.27371550e-01 8.49222720e-01 3.56870651e-01 -9.57556069e-01
-5.22245049e-01 -1.08301058e-01 -2.81947225e-01 -1.52844593e-01
8.99782896e-01 -1.01016736e+00 -6.35040045e-01 4.98640478e-01
-1.26271236e+00 -5.66150486e-01 -2.39720300e-01 6.61349177e-01
-5.93737304e-01 1.59561425e-01 -5.90136349e-01 -1.27229762e+00
4.25072424e-02 -1.03014457e+00 1.17557240e+00 3.85171890e-01
-4.84061271e-01 -1.33665967e+00 2.52745628e-01 -1.78432167e-01
4.65819389e-01 1.35304183e-01 8.64354491e-01 -3.03077877e-01
-8.38434577e-01 -2.57006168e-01 -1.17826760e-01 3.03765554e-02
-2.75300294e-02 5.17274320e-01 -1.20622754e+00 -1.66733582e-02
-3.66865993e-02 3.52149904e-01 7.40607858e-01 7.95636415e-01
1.06657755e+00 -2.25705236e-01 -3.39751631e-01 5.82507968e-01
1.51056242e+00 -1.17773861e-01 9.11966801e-01 -2.35042095e-01
6.04108691e-01 3.76003891e-01 -3.70020466e-03 7.82524645e-01
3.05780351e-01 3.89405251e-01 2.62475729e-01 3.59808385e-01
2.32756078e-01 -5.68449914e-01 1.29754856e-01 5.86394906e-01
-4.26662445e-01 -5.00512958e-01 -1.19034410e+00 3.53332579e-01
-1.88251352e+00 -1.08900976e+00 -2.62520790e-01 2.28926373e+00
8.03650677e-01 -1.50707802e-02 -2.51428366e-01 -2.21013650e-01
7.02613652e-01 1.31763086e-01 -4.82633680e-01 -2.22015277e-01
1.68626636e-01 4.64927346e-01 5.24539292e-01 7.70922542e-01
-8.42298269e-01 7.63710201e-01 6.20234203e+00 1.13019919e+00
-6.62471354e-01 2.85879493e-01 4.66569841e-01 -2.19934415e-02
-5.12174487e-01 3.46665114e-01 -8.84897709e-01 8.29353988e-01
1.39339554e+00 2.46878266e-01 6.30363107e-01 5.34373522e-01
5.12840509e-01 -6.32892668e-01 -9.40190792e-01 7.49745190e-01
-5.77234983e-01 -1.31964040e+00 3.89075466e-02 4.21859682e-01
5.80944061e-01 1.30118847e-01 -2.83336472e-02 2.61174053e-01
9.87289906e-01 -1.02224123e+00 6.27624810e-01 1.08303142e+00
4.28938687e-01 -9.18948591e-01 8.46120119e-01 4.81527001e-01
-8.31492722e-01 2.17638478e-01 -4.76822466e-01 -2.34952554e-01
4.89764601e-01 1.00567901e+00 -6.91239715e-01 3.71693701e-01
5.63587070e-01 3.95944238e-01 -1.34045810e-01 9.96807516e-01
-5.91686130e-01 1.14486945e+00 -7.75776923e-01 -4.33445238e-02
1.76659718e-01 -6.54637396e-01 6.70742691e-01 1.05268502e+00
5.86308479e-01 -1.23452768e-01 -2.89654851e-01 1.42357028e+00
4.14099306e-01 -5.17450154e-01 -5.26752710e-01 -2.18464807e-01
7.53809094e-01 9.49431121e-01 -8.65372300e-01 -2.72425681e-01
-2.57510185e-01 8.20882916e-01 1.27818316e-01 8.69320869e-01
-1.00329745e+00 -1.17216825e-01 7.18275070e-01 -3.07441000e-02
6.56535447e-01 -3.16374183e-01 -1.00524999e-01 -1.01663840e+00
-1.85757682e-01 -5.75730324e-01 6.60809055e-02 -8.65820706e-01
-1.31045985e+00 1.38806000e-01 5.45127273e-01 -6.97795928e-01
-6.85393214e-01 -5.23973167e-01 -7.61842668e-01 1.25257897e+00
-1.18050718e+00 -9.43584621e-01 1.70091659e-01 3.03696722e-01
1.70358613e-01 3.70466024e-01 8.31835270e-01 -1.15388453e-01
-5.56174219e-01 -6.87716901e-02 5.41158974e-01 -2.20853120e-01
2.06354439e-01 -1.11900878e+00 4.38053370e-01 7.36202240e-01
1.72406048e-01 8.31953645e-01 1.11184871e+00 -8.68019760e-01
-1.26098609e+00 -8.65863502e-01 5.52577436e-01 -7.44555354e-01
8.24147880e-01 -5.11733353e-01 -1.02054930e+00 7.55122304e-01
2.16572687e-01 -2.68915832e-01 5.99368215e-01 2.72379756e-01
9.72154364e-03 3.44514430e-01 -8.29610109e-01 7.49400258e-01
5.44300914e-01 -6.20588183e-01 -2.22971246e-01 1.24340527e-01
4.80965674e-01 -1.65988624e-01 -9.46095884e-01 1.22810148e-01
6.41282737e-01 -8.95294547e-01 9.29532409e-01 -1.87960789e-01
2.42041573e-01 -3.98371100e-01 -1.74172327e-01 -1.27741253e+00
-1.47270501e-01 -7.65243769e-01 -6.27141476e-01 1.29804361e+00
2.97951341e-01 -8.31673145e-01 4.93548721e-01 7.87506402e-01
2.92244643e-01 -5.00431001e-01 -1.02042878e+00 -6.54602647e-01
2.47909799e-01 -7.72064805e-01 8.86536181e-01 4.50271726e-01
-8.06862354e-01 4.71616276e-02 -3.20450515e-01 5.49443007e-01
8.12400222e-01 -1.29624605e-01 6.70965970e-01 -1.39939952e+00
-6.44010723e-01 -2.16958523e-01 3.57204932e-03 -9.36798930e-01
9.11794528e-02 -4.23662424e-01 1.66962951e-01 -1.39553964e+00
1.09326527e-01 -3.31702560e-01 1.83565274e-01 9.85343158e-02
-3.51270199e-01 -1.17578864e-01 -6.31419867e-02 2.65566200e-01
-4.08020727e-02 7.00194538e-01 1.01894927e+00 2.29198769e-01
-7.88652524e-02 1.89357132e-01 -1.98313564e-01 8.37235093e-01
8.04947257e-01 -6.73279464e-01 -4.39424276e-01 -2.71286696e-01
5.05532086e-01 1.21104613e-01 8.43763888e-01 -7.92436659e-01
2.15571165e-01 -3.75375688e-01 4.84755844e-01 -6.37740552e-01
5.34590662e-01 -5.13387680e-01 8.07776153e-01 2.49483138e-01
2.11797625e-01 -1.53747156e-01 2.91041315e-01 8.52172196e-01
1.25312954e-01 -4.78583038e-01 3.90059352e-01 -2.53477097e-01
-2.30687037e-01 3.22609246e-01 -9.22336400e-01 -8.89215469e-02
4.87379104e-01 -3.23843770e-02 -4.45972057e-03 -6.29262924e-01
-7.86884964e-01 1.65741835e-02 5.24226487e-01 -2.20138714e-01
3.01678568e-01 -1.03905094e+00 -7.41143763e-01 9.48319361e-02
-3.01003218e-01 3.46299291e-01 4.62751359e-01 8.70313227e-01
-5.30162096e-01 1.94816753e-01 2.54282445e-01 -7.84211576e-01
-3.63074183e-01 2.30456278e-01 2.20143333e-01 -3.51898521e-01
-6.05587423e-01 6.81542277e-01 1.50044605e-01 -4.63077188e-01
-1.52278736e-01 -1.41643494e-01 3.57564837e-01 1.99206509e-02
3.99587840e-01 4.22502309e-01 -2.69368172e-01 -2.72773474e-01
-1.22233845e-01 2.58961976e-01 2.55737871e-01 -6.03672922e-01
1.41329205e+00 -4.35240388e-01 -9.42138135e-02 7.35846281e-01
7.35232711e-01 2.53871232e-02 -1.83308160e+00 1.36053292e-02
-1.14624329e-01 -3.84279817e-01 2.57436275e-01 -5.11120677e-01
-5.75301051e-01 1.08831465e+00 4.03025210e-01 1.56288996e-01
6.57690585e-01 1.02096319e-01 3.14517945e-01 3.00499678e-01
3.22218090e-01 -8.62583756e-01 -4.35878307e-01 4.03738141e-01
6.84119701e-01 -8.78507257e-01 6.34142160e-02 -1.80715173e-01
-4.29803759e-01 7.65815258e-01 -8.19169953e-02 -2.23366767e-01
9.05094802e-01 4.73378003e-01 -3.67354423e-01 -2.57313281e-01
-7.17705369e-01 -1.52242839e-01 6.49223477e-02 7.17765391e-01
2.30514184e-01 2.68815786e-01 1.35846749e-01 4.16987211e-01
-1.87708393e-01 1.77329853e-01 3.60990435e-01 8.15972149e-01
-7.01957643e-02 -1.02103233e+00 -4.16774750e-01 4.68127519e-01
-3.95969361e-01 -4.13475811e-01 3.44860524e-01 7.18531489e-01
-2.96894927e-02 6.67056978e-01 4.33492899e-01 2.17278793e-01
-1.91207603e-01 2.94460416e-01 4.20467317e-01 -4.45121258e-01
-3.70697193e-02 2.69352138e-01 9.91032831e-03 -4.50259864e-01
-1.47139803e-01 -9.81555283e-01 -7.96120763e-01 -6.51496887e-01
-2.04860568e-01 2.07483381e-01 6.51640236e-01 8.99678946e-01
4.74163175e-01 4.39505100e-01 5.03454208e-02 -1.22542810e+00
-5.08002698e-01 -9.94964361e-01 -5.37426949e-01 -4.49453816e-02
2.16093943e-01 -8.77784193e-01 -5.46517670e-01 -7.30151543e-03] | [6.864471912384033, 3.869149923324585] |
aaaf8124-9f38-4bc9-b429-4ce7c88ed0c5 | count-and-similarity-aware-r-cnn-for | null | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/2678_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123620086.pdf | Count- and Similarity-aware R-CNN for Pedestrian Detection | Recent pedestrian detection methods generally rely on additional supervision, such as visible bounding-box annotations, to handle heavy occlusions. We propose an approach that leverages pedestrian count and proposal similarity information within a two-stage pedestrian detection framework. Both pedestrian count and proposal similarity are derived from standard full-body annotations commonly used to train pedestrian detectors. We introduce a count-weighted detection loss function that assigns higher weights to the detection errors occurring at highly overlapping pedestrians. The proposed loss function is utilized at both stages of the two-stage detector. We further introduce a count-and-similarity branch within the two-stage detection framework, which predicts pedestrian count as well as proposal similarity. Lastly, we introduce a count and similarity-aware NMS strategy to identify distinct proposals. Our approach requires neither part information nor visible bounding-box annotations. Experiments are performed on the CityPersons and CrowdHuman datasets. Our method sets a new state-of-the-art on both datasets. Further, it achieves an absolute gain of 2.4\% over the current state-of-the-art, in terms of log-average miss rate, on the heavily occluded ( extbf{HO}) set of CityPersons test set. Finally, we demonstrate the applicability of our approach for the problem of human instance segmentation. Code and models are available at: https://github.com/Leotju/CaSe . | ['Rao Muhammad Anwer', 'Hisham Cholakkal', 'Fahad Shahbaz Khan', 'Yanwei Pang', 'Mubarak Shah', 'Ling Shao', 'Jin Xie'] | null | null | null | null | eccv-2020-8 | ['human-instance-segmentation'] | ['computer-vision'] | [-7.66316056e-02 -1.45457089e-01 -5.50978407e-02 -5.21145225e-01
-9.07489836e-01 -3.11311483e-01 4.97999460e-01 2.87156135e-01
-8.32939506e-01 5.26066005e-01 -2.29750529e-01 -7.24064931e-02
8.16031933e-01 -6.82187676e-01 -7.08182871e-01 -5.90506136e-01
-3.88206504e-02 3.68614107e-01 1.01905608e+00 -5.02373204e-02
-9.54488888e-02 2.49050051e-01 -1.52522326e+00 2.12028056e-01
7.08986700e-01 8.76032770e-01 -3.40619753e-03 8.41006577e-01
3.36400896e-01 2.37309381e-01 -5.33378005e-01 -8.40939879e-01
5.84200144e-01 -1.08317342e-02 -4.27388817e-01 2.49741733e-01
9.43719685e-01 -5.32357812e-01 -4.19870883e-01 8.65915239e-01
6.65090621e-01 1.60729036e-01 6.57808304e-01 -1.27353680e+00
-7.15995207e-02 6.69311509e-02 -1.23246837e+00 6.09121799e-01
3.63178343e-01 6.79026544e-01 9.45801020e-01 -1.11885154e+00
2.79079705e-01 1.30556977e+00 8.92712951e-01 5.11173606e-01
-1.23978448e+00 -6.68591022e-01 4.57320154e-01 7.81457871e-02
-1.37648010e+00 -4.59346741e-01 4.46330816e-01 -6.99959576e-01
8.13764691e-01 1.08648285e-01 8.08522284e-01 9.91599441e-01
-1.59955531e-01 1.26313293e+00 8.62522244e-01 -3.47721815e-01
-1.19054981e-03 8.19166843e-03 4.28540558e-01 9.83361900e-01
6.77866697e-01 3.16869617e-01 -2.71695405e-01 -1.14915632e-01
6.01122320e-01 -4.55209427e-02 1.29895031e-01 -5.48049092e-01
-7.48541176e-01 8.86757493e-01 5.45842469e-01 -3.12588066e-01
-1.38325781e-01 1.47421226e-01 5.24702430e-01 -4.66493398e-01
5.03263950e-01 -2.57178456e-01 -3.23851734e-01 9.76730809e-02
-1.18145061e+00 4.59935278e-01 6.32700086e-01 1.15677416e+00
4.36343700e-01 -2.72823215e-01 -5.75545192e-01 6.45676672e-01
5.64176559e-01 6.09046698e-01 -1.40888244e-01 -8.06520402e-01
6.48860097e-01 5.92364430e-01 3.09554726e-01 -7.48133600e-01
-3.73928219e-01 -4.69093770e-01 -3.88664484e-01 2.13359609e-01
8.30741167e-01 -1.83606789e-01 -1.11667943e+00 1.51065767e+00
8.32149386e-01 2.38601893e-01 -4.67612296e-01 1.14089882e+00
9.56495047e-01 2.18826964e-01 4.59810257e-01 3.16709369e-01
1.79290879e+00 -1.20384932e+00 -1.15947723e-01 -5.02032936e-01
4.18365955e-01 -6.99305356e-01 8.21760535e-01 6.02928139e-02
-1.18119514e+00 -6.85530424e-01 -9.76395428e-01 -1.73233211e-01
-2.49180362e-01 5.40122211e-01 5.39226949e-01 9.54973817e-01
-6.65146112e-01 3.35700870e-01 -1.06128442e+00 -6.24411166e-01
7.62157083e-01 2.65958756e-01 -6.89256564e-02 -9.59583372e-02
-7.63576746e-01 7.34020829e-01 2.60537952e-01 -3.68089904e-03
-8.31487954e-01 -6.27553165e-01 -1.01005745e+00 -7.17985854e-02
4.83644068e-01 -1.07271147e+00 1.30762136e+00 -2.96944112e-01
-1.07182026e+00 1.27577209e+00 -3.01886559e-01 -6.14085376e-01
1.06436801e+00 -5.16053319e-01 -5.43805473e-02 1.76595032e-01
6.05685353e-01 9.39328969e-01 6.07943118e-01 -1.26403701e+00
-1.20734978e+00 -3.69723082e-01 -3.24580185e-02 -5.49451336e-02
1.17258407e-01 2.13014528e-01 -9.14259255e-01 -5.81865609e-01
-6.23010546e-02 -8.30066144e-01 -5.22077620e-01 4.29961771e-01
-7.36284137e-01 -2.37808973e-01 6.31419957e-01 -7.42143750e-01
1.08129942e+00 -1.87942660e+00 -2.30253115e-01 1.29133925e-01
4.04882193e-01 4.69141126e-01 2.17923298e-02 -1.05798237e-01
2.97285229e-01 -1.33111790e-01 -3.14734071e-01 -1.00571203e+00
1.44966543e-01 -7.62720332e-02 2.19181180e-01 7.35406876e-01
5.18940270e-01 7.03985393e-01 -8.51946712e-01 -8.40327919e-01
4.23247367e-01 5.31134427e-01 -5.36570966e-01 1.29734904e-01
1.42031416e-01 3.55409086e-01 -2.85615981e-01 9.56106842e-01
7.88131773e-01 -1.76624075e-01 -2.10354701e-01 -1.71563223e-01
-3.18284214e-01 1.08070165e-01 -1.35797369e+00 1.32236743e+00
1.38308480e-01 4.21433777e-01 3.64144333e-02 -6.66825056e-01
5.39651811e-01 -1.90189127e-02 3.05487454e-01 -3.17311257e-01
3.48686606e-01 1.47723049e-01 -1.51786609e-02 -1.76870167e-01
4.93188441e-01 2.70500541e-01 -8.90105888e-02 -5.28754555e-02
-2.01075301e-02 3.25863361e-01 7.35642910e-01 2.67586112e-01
1.12475884e+00 3.66164714e-01 4.33806241e-01 -1.65528744e-01
5.91958046e-01 1.89872794e-02 7.47913837e-01 1.01628494e+00
-9.06607747e-01 7.04627991e-01 2.86988586e-01 -4.64488000e-01
-1.14819670e+00 -1.21106946e+00 -2.14619651e-01 1.20372927e+00
2.23611951e-01 -3.51828039e-01 -9.15869772e-01 -8.08192372e-01
1.51816189e-01 3.45415682e-01 -5.20828068e-01 2.60166049e-01
-8.50002110e-01 -9.25737560e-01 6.43199563e-01 9.49674487e-01
6.03827477e-01 -6.65416658e-01 -9.30278003e-01 1.76029220e-01
-1.48602605e-01 -1.51884258e+00 -6.79105759e-01 -7.85978362e-02
-5.39366901e-01 -1.30168140e+00 -9.82386410e-01 -6.03685737e-01
5.57660758e-01 4.14262414e-01 1.23238039e+00 2.07996383e-01
-7.44372010e-01 4.81676459e-01 -1.71759620e-01 -3.46955448e-01
1.33038297e-01 4.78672236e-03 3.27783637e-02 -1.56534001e-01
5.94513893e-01 -1.44502401e-01 -9.65225399e-01 5.29785454e-01
-1.59443602e-01 -7.07462952e-02 5.30970156e-01 6.85205579e-01
7.04470217e-01 -4.21310812e-01 3.55204605e-02 -5.42220891e-01
-1.42915234e-01 -2.77448982e-01 -8.04807782e-01 3.76516022e-02
-9.49781835e-02 -3.37734073e-01 6.27989843e-02 -4.09190953e-01
-9.79230940e-01 4.70151633e-01 -1.60808295e-01 -2.18317538e-01
-4.02847946e-01 -4.26860511e-01 -4.01020736e-01 -1.55951500e-01
4.79820371e-01 -7.52600208e-02 -4.94596958e-01 -5.06840646e-01
4.17500705e-01 3.02858740e-01 8.00687790e-01 -7.43280590e-01
8.76382113e-01 7.89696276e-01 -1.05139859e-01 -6.99471235e-01
-8.86686206e-01 -1.15407288e+00 -9.64808404e-01 -2.07378134e-01
1.01441586e+00 -1.13188159e+00 -8.17017853e-01 4.82716799e-01
-1.36047685e+00 -1.86922640e-01 -1.79735973e-01 3.13438952e-01
-2.99961746e-01 6.13199651e-01 -8.28078628e-01 -1.12578070e+00
-3.95049363e-01 -1.10890615e+00 1.47623265e+00 4.20541823e-01
-1.03327304e-01 -6.01553619e-01 -2.47047052e-01 4.84230012e-01
-1.12269618e-01 4.67682093e-01 8.30054730e-02 -6.04685664e-01
-6.47853494e-01 -3.74523759e-01 -6.26923859e-01 7.69209862e-02
-4.42756891e-01 -1.74330901e-02 -9.49253798e-01 -3.63584220e-01
-6.35911644e-01 -9.53850672e-02 1.35080600e+00 6.05704367e-01
7.47671306e-01 4.20879833e-02 -6.87245727e-01 5.23381650e-01
1.05497491e+00 -3.46495539e-01 3.08071107e-01 4.67698872e-01
7.94052899e-01 6.01190984e-01 7.31325269e-01 6.29162669e-01
7.24979520e-01 9.85913575e-01 2.63868660e-01 -3.31234843e-01
-4.92756575e-01 -2.95132607e-01 2.80293554e-01 4.51567471e-02
-2.55411655e-01 -5.48239648e-02 -9.65953171e-01 5.82002759e-01
-2.02701974e+00 -1.04737723e+00 -6.30182385e-01 2.25475693e+00
5.62474370e-01 5.92898428e-01 8.39171827e-01 1.40286297e-01
1.08985198e+00 -1.16740339e-01 -5.03340960e-01 1.73666418e-01
8.16403627e-02 6.40730411e-02 6.54001474e-01 5.15857756e-01
-1.85894334e+00 1.03484893e+00 5.67327547e+00 6.87488675e-01
-3.67173165e-01 3.56883496e-01 6.44459844e-01 -1.04900636e-01
6.24821901e-01 -1.32713005e-01 -1.52101254e+00 5.95992684e-01
4.68693018e-01 4.13428485e-01 -3.63961101e-01 1.07800543e+00
2.74157166e-01 -4.01160002e-01 -1.14586461e+00 9.43278491e-01
-8.69683363e-03 -7.38692343e-01 -3.59104872e-01 5.06458431e-02
4.40312475e-01 -1.26592457e-01 3.75197865e-02 4.48537588e-01
4.06119913e-01 -5.16980886e-01 9.49035227e-01 1.05159648e-01
5.28679252e-01 -6.35000229e-01 8.14527810e-01 2.33024493e-01
-1.99433994e+00 1.56623889e-02 -3.64702106e-01 -4.57313545e-02
5.15262723e-01 5.79412937e-01 -6.78749800e-01 2.99067050e-01
8.91095519e-01 4.84300673e-01 -7.85350204e-01 1.57026184e+00
-5.28056264e-01 4.91605073e-01 -6.03024364e-01 1.95264533e-01
1.05711244e-01 1.25513151e-01 7.15222478e-01 1.82533216e+00
-1.66954011e-01 2.71290064e-01 8.43313575e-01 8.04662943e-01
1.62371501e-01 -7.90800154e-02 -2.81910393e-02 8.07435572e-01
4.74916875e-01 1.31402063e+00 -9.81488526e-01 -5.74771523e-01
-4.66464341e-01 9.81776059e-01 3.51643831e-01 1.98997781e-01
-1.19654322e+00 -7.49202818e-02 6.97385252e-01 4.41451758e-01
6.58894658e-01 -1.39573485e-01 -3.90276015e-01 -1.08064127e+00
3.04591864e-01 -4.25370336e-01 5.78065574e-01 -1.15784325e-01
-1.53560519e+00 3.05242866e-01 1.68491304e-01 -1.05350649e+00
4.23812419e-02 -6.81135893e-01 -7.61318624e-01 6.77439630e-01
-1.57849562e+00 -1.55337346e+00 -4.70626354e-01 4.31110054e-01
7.20095336e-01 1.22216091e-01 2.30518162e-01 7.01576650e-01
-9.75267351e-01 9.58998919e-01 -5.71147442e-01 6.82340860e-01
6.13228261e-01 -1.25906456e+00 8.36015344e-01 1.19436193e+00
-3.13679010e-01 4.83384818e-01 7.53161490e-01 -7.60330319e-01
-6.91758037e-01 -1.31475663e+00 6.85269058e-01 -7.76382506e-01
4.16436374e-01 -4.71284240e-01 -5.17589509e-01 5.91624737e-01
-2.81790912e-01 5.52227318e-01 7.14104772e-01 8.12214464e-02
-4.70824569e-01 2.52562523e-01 -1.35255086e+00 4.79775608e-01
1.29029715e+00 9.39974785e-02 -5.03072977e-01 3.12396944e-01
5.22746980e-01 -6.18676543e-01 -5.33135414e-01 3.75360519e-01
7.09589005e-01 -1.04229832e+00 1.40971315e+00 -4.84071940e-01
8.02440941e-02 -5.47579646e-01 -1.58917055e-01 -6.27962768e-01
-3.77494454e-01 -2.39261135e-01 -3.65147203e-01 1.17612171e+00
2.46527433e-01 -3.17016274e-01 1.04086614e+00 7.35799193e-01
-1.08665682e-01 -7.77800381e-01 -1.11006093e+00 -9.36344564e-01
-8.73182938e-02 -4.05443132e-01 8.77454802e-02 3.18896383e-01
-4.78957981e-01 2.07845196e-01 -1.96466967e-01 3.87396008e-01
1.31557596e+00 -2.69168735e-01 1.11889493e+00 -1.22473395e+00
-3.95294517e-01 -4.81271297e-01 -6.95416152e-01 -1.26065910e+00
-9.69454348e-02 -5.20900786e-01 2.34109506e-01 -1.48456240e+00
6.12353206e-01 -3.03836972e-01 -3.50306258e-02 4.57861125e-01
-6.20136261e-01 5.43609858e-01 5.41633606e-01 1.38860986e-01
-1.09820700e+00 4.12805855e-01 7.68936515e-01 -1.09579198e-01
-1.75490201e-01 3.14645350e-01 -2.67451167e-01 1.17482579e+00
6.85523987e-01 -4.58922386e-01 2.75309563e-01 -1.92514539e-01
-6.23870134e-01 -4.63454366e-01 9.59396541e-01 -1.46584582e+00
2.48749256e-01 2.06410170e-01 7.86718071e-01 -9.31428790e-01
5.93197107e-01 -4.36183691e-01 -5.01250088e-01 6.82600200e-01
3.99570614e-02 -1.35926222e-02 2.21182749e-01 6.24890208e-01
3.18276376e-01 9.72777456e-02 1.20379698e+00 -1.73059016e-01
-7.38658845e-01 4.55281019e-01 -3.01553315e-04 2.28696823e-01
1.22178710e+00 -3.77633214e-01 -2.07408682e-01 1.75979435e-01
-7.09849477e-01 5.11000216e-01 3.56703907e-01 2.61102438e-01
5.98366261e-01 -1.09052682e+00 -9.41590726e-01 2.08568480e-02
1.80511683e-01 9.29744076e-03 8.49060249e-03 9.75848854e-01
-2.41634995e-01 2.17542410e-01 1.04959264e-01 -9.70886886e-01
-1.64521909e+00 3.54435891e-01 3.71288449e-01 -3.68345529e-01
-7.13263571e-01 1.07820821e+00 2.84011036e-01 -2.12565273e-01
3.99349600e-01 -2.16860741e-01 1.14716977e-01 -1.42141849e-01
7.19804466e-01 8.14745545e-01 -3.16345841e-01 -9.44565535e-01
-6.52804375e-01 5.10985911e-01 -1.56353071e-01 -1.33366615e-01
8.46616328e-01 -1.08053252e-01 4.11241472e-01 -6.07311688e-02
8.42453718e-01 -2.67292470e-01 -1.62115943e+00 -2.25844383e-01
7.76884630e-02 -6.23938799e-01 -3.01467061e-01 -5.60422122e-01
-9.24312294e-01 8.22103083e-01 9.26696777e-01 -3.44210356e-01
5.84313571e-01 1.78867683e-01 9.88812387e-01 1.23339087e-01
3.82466733e-01 -1.07826662e+00 1.26388654e-01 4.38837618e-01
2.91251272e-01 -1.73317671e+00 2.09787622e-01 -8.33656430e-01
-3.97062153e-01 6.86410010e-01 1.03103495e+00 -1.83892772e-01
4.26556259e-01 4.11087453e-01 -2.46535569e-01 -1.20651396e-02
-2.67861664e-01 -8.00215662e-01 4.11636293e-01 7.81507969e-01
4.67272967e-01 2.57777333e-01 -3.45915318e-01 7.66771019e-01
-6.34703925e-03 -2.17902794e-01 1.55995399e-01 1.04172349e+00
-7.75148809e-01 -9.16908979e-01 -7.85742342e-01 3.73325348e-01
-4.03153092e-01 -7.06840754e-02 -1.91339090e-01 8.14742804e-01
5.56631565e-01 1.16051519e+00 -1.26731128e-01 -6.31467551e-02
6.11528575e-01 -2.14836642e-01 5.11131227e-01 -6.93366885e-01
-5.69356501e-01 8.29519108e-02 1.50431424e-01 -7.31720686e-01
-3.35508496e-01 -9.06391382e-01 -1.27976561e+00 -3.87593776e-01
-4.65651661e-01 -2.62892783e-01 3.11112642e-01 7.95026124e-01
5.51402643e-02 3.80043179e-01 9.17678326e-02 -1.39913487e+00
-4.40998793e-01 -8.23822856e-01 -1.99514702e-01 4.89536703e-01
1.37249753e-01 -1.03901076e+00 -6.55219629e-02 3.08544636e-02] | [7.917999267578125, -0.5806717872619629] |
9863d609-be65-42d3-b64f-59c4c36a67ed | occluded-video-instance-segmentation-dataset | 2111.07950 | null | https://arxiv.org/abs/2111.07950v1 | https://arxiv.org/pdf/2111.07950v1.pdf | Occluded Video Instance Segmentation: Dataset and ICCV 2021 Challenge | Although deep learning methods have achieved advanced video object recognition performance in recent years, perceiving heavily occluded objects in a video is still a very challenging task. To promote the development of occlusion understanding, we collect a large-scale dataset called OVIS for video instance segmentation in the occluded scenario. OVIS consists of 296k high-quality instance masks and 901 occluded scenes. While our human vision systems can perceive those occluded objects by contextual reasoning and association, our experiments suggest that current video understanding systems cannot. On the OVIS dataset, all baseline methods encounter a significant performance degradation of about 80% in the heavily occluded object group, which demonstrates that there is still a long way to go in understanding obscured objects and videos in a complex real-world scenario. To facilitate the research on new paradigms for video understanding systems, we launched a challenge based on the OVIS dataset. The submitted top-performing algorithms have achieved much higher performance than our baselines. In this paper, we will introduce the OVIS dataset and further dissect it by analyzing the results of baselines and submitted methods. The OVIS dataset and challenge information can be found at http://songbai.site/ovis . | ['Song Bai', 'Philip H. S. Torr', 'Alan Yuille', 'Serge Belongie', 'Xiang Bai', 'Xiaoyu Liu', 'Xinggang Wang', 'Yao Hu', 'Yan Gao', 'Jiyang Qi'] | 2021-11-15 | null | null | null | null | ['video-instance-segmentation'] | ['computer-vision'] | [ 2.09864527e-01 -1.08348101e-01 -4.47608232e-01 -4.27959830e-01
-5.77561140e-01 -4.75869894e-01 3.48377883e-01 -4.51454639e-01
-2.27142468e-01 4.89996761e-01 1.96293175e-01 -1.27096027e-01
2.85952777e-01 -3.19131255e-01 -1.12313199e+00 -4.78251070e-01
-1.35904491e-01 3.63033354e-01 4.88145798e-01 1.79373756e-01
1.24520659e-01 1.68075189e-01 -1.58831120e+00 8.62892330e-01
6.32985830e-01 1.12502062e+00 2.15084374e-01 7.80286610e-01
1.78812042e-01 8.16579103e-01 -8.39978993e-01 -2.58929163e-01
4.79774028e-01 -1.86094299e-01 -8.73702466e-01 2.93158889e-01
1.21336055e+00 -6.83947742e-01 -8.57249796e-01 8.08127344e-01
8.21971074e-02 -2.37072501e-02 4.45621938e-01 -1.45857131e+00
-6.24950588e-01 4.07288134e-01 -5.47166288e-01 7.42383599e-01
3.99550766e-01 3.74072701e-01 8.99368525e-01 -1.12422311e+00
6.94801629e-01 1.31394279e+00 4.08117980e-01 4.59995836e-01
-8.61102819e-01 -7.34183967e-01 6.77706003e-01 7.79910684e-01
-1.36672127e+00 -5.08531690e-01 4.51470703e-01 -5.25126278e-01
9.31218147e-01 3.55682462e-01 7.52049327e-01 1.30396664e+00
-5.21801552e-03 1.49624038e+00 8.26920688e-01 1.03079453e-01
1.95448101e-02 -1.68189868e-01 5.08589268e-01 4.59963888e-01
3.89426798e-01 1.10015161e-01 -7.03986526e-01 3.77007753e-01
6.14918709e-01 1.45227000e-01 -6.26509905e-01 -1.36112422e-01
-1.35061800e+00 6.09727442e-01 5.69694459e-01 1.29586786e-01
-5.70498817e-02 2.63040811e-01 2.98702866e-01 2.69356012e-01
4.80463535e-01 2.87471801e-01 -4.90027398e-01 -1.91775367e-01
-9.31642294e-01 3.35782200e-01 7.76347935e-01 1.13782334e+00
4.79280323e-01 5.45478687e-02 -2.99761444e-01 5.74672639e-01
1.69389829e-01 4.13215637e-01 -6.44681826e-02 -1.23585629e+00
5.87125182e-01 3.21952790e-01 4.99825962e-02 -9.26957369e-01
-6.28293818e-03 -2.50602096e-01 -6.70058370e-01 -1.82164386e-02
4.77774858e-01 1.01152964e-01 -1.41140521e+00 1.32416117e+00
1.43845573e-01 6.25890791e-01 -1.64240021e-02 1.31809461e+00
1.32205307e+00 9.11652863e-01 9.19913203e-02 -1.32207379e-01
1.31963694e+00 -1.44133067e+00 -8.69958282e-01 -4.25608724e-01
3.08294863e-01 -6.03137851e-01 1.01632488e+00 6.21483028e-01
-9.43281949e-01 -8.22259963e-01 -1.19787204e+00 -2.22336978e-01
-2.15166882e-01 -7.26429839e-03 8.46176684e-01 4.08095568e-01
-1.01330471e+00 1.87312454e-01 -7.49828815e-01 -3.70539635e-01
1.07045317e+00 1.23519428e-01 -4.65671450e-01 -5.59620798e-01
-1.07083964e+00 5.41877151e-01 4.41121608e-01 2.50123024e-01
-1.63198185e+00 -8.32843125e-01 -8.36106896e-01 1.18812904e-01
9.68256593e-01 -3.96059662e-01 9.71743345e-01 -1.12621546e+00
-8.35120022e-01 8.43088508e-01 -2.97158122e-01 -5.98229527e-01
6.46928132e-01 -7.55068958e-01 -4.72534567e-01 4.45229679e-01
-3.16845253e-02 1.05302989e+00 9.83838737e-01 -1.38480854e+00
-6.13046527e-01 -2.89837867e-01 3.69103611e-01 1.70256034e-01
7.90457241e-03 -3.23327966e-02 -1.10866797e+00 -6.16852582e-01
1.70312911e-01 -9.65352356e-01 5.35390489e-02 9.14207101e-02
-2.98897594e-01 -2.82286435e-01 1.29543149e+00 -6.75844312e-01
1.01911473e+00 -2.25410032e+00 -9.05792043e-02 -4.31551903e-01
5.68517387e-01 6.68800235e-01 -4.05258596e-01 8.50088969e-02
-8.32929015e-02 3.39888424e-01 4.03796695e-02 -2.96852887e-01
-1.86079830e-01 4.35682058e-01 -6.07486844e-01 2.72501200e-01
3.04006636e-01 1.07029235e+00 -6.81730032e-01 -4.20036107e-01
3.32345605e-01 3.67376357e-01 -6.99063838e-01 3.56502593e-01
-4.32579875e-01 4.06455278e-01 -2.43511245e-01 1.06544077e+00
8.63969207e-01 -3.08545202e-01 -1.70412421e-01 -4.42217231e-01
3.07576627e-01 -7.19929338e-02 -9.91420686e-01 1.70677662e+00
1.90732911e-01 1.41411412e+00 -8.85467976e-02 -1.16404140e+00
4.87613201e-01 2.77297288e-01 3.66625965e-01 -6.76829517e-01
-1.38466418e-01 -1.74717322e-01 -2.08022469e-03 -8.30496728e-01
3.72257948e-01 4.31958497e-01 3.01282763e-01 -5.44418246e-02
1.95505116e-02 8.70833099e-02 4.21037972e-01 5.86976528e-01
1.07056546e+00 2.22727895e-01 -9.34010595e-02 -2.91926026e-01
2.78854042e-01 1.61385924e-01 8.17818344e-01 1.10256720e+00
-6.22161388e-01 8.74703109e-01 5.09685695e-01 -1.03235769e+00
-8.84762943e-01 -1.28099287e+00 -2.03398764e-01 9.87931848e-01
6.41148150e-01 -6.22451961e-01 -8.56235683e-01 -7.31861055e-01
-2.06241459e-02 2.59532094e-01 -6.16095126e-01 1.13872416e-01
-4.88595158e-01 -6.57745004e-01 3.44289482e-01 8.68314385e-01
9.15497959e-01 -1.10465574e+00 -3.97768140e-01 -2.36185491e-01
-6.41753435e-01 -1.73413587e+00 -4.04072821e-01 -3.16028863e-01
-6.50639415e-01 -1.27775383e+00 -4.92294312e-01 -7.38264918e-01
4.87215489e-01 8.13012123e-01 1.51300001e+00 3.96178067e-01
-6.32955253e-01 4.12481308e-01 -4.80070144e-01 -5.91351867e-01
8.10544789e-02 -1.60218135e-01 -4.64840271e-02 -1.00370564e-01
6.80348098e-01 -1.14008352e-01 -8.36642444e-01 6.99720740e-01
-9.85864937e-01 2.94086218e-01 3.45647186e-01 5.81623733e-01
5.83950102e-01 -7.77935535e-02 4.63656366e-01 -6.91839576e-01
-8.24827999e-02 -4.69305575e-01 -5.07748187e-01 2.43256956e-01
-4.77922289e-03 -4.36493099e-01 1.23009965e-01 -4.82897043e-01
-1.00283718e+00 -1.01254858e-01 -3.09831481e-02 -6.12598062e-01
-4.52350974e-01 2.60144383e-01 -4.09670323e-01 7.78835788e-02
3.97039205e-01 -6.83398768e-02 -3.52284580e-01 -4.54290718e-01
2.37630755e-01 6.47126198e-01 5.89651167e-01 -4.34426963e-01
5.41514158e-01 6.65758431e-01 -6.54738963e-01 -1.02263474e+00
-1.26561379e+00 -4.74131137e-01 -4.41658616e-01 -4.54699010e-01
1.04527342e+00 -1.31225097e+00 -7.05693662e-01 4.75153267e-01
-1.14464080e+00 -4.96296197e-01 -9.91122872e-02 3.99253994e-01
-4.35370684e-01 4.07967269e-01 -5.91274321e-01 -5.08596003e-01
6.22186773e-02 -1.42473149e+00 1.13049686e+00 2.80794680e-01
-1.36789344e-02 -5.15264392e-01 -3.19017947e-01 1.09399641e+00
2.57960439e-01 4.03394736e-02 4.57215488e-01 -5.83002567e-01
-1.38561511e+00 1.17883189e-02 -6.23496592e-01 4.10124123e-01
-2.12641329e-01 9.49489325e-02 -1.15957451e+00 -4.48801547e-01
1.95252057e-02 -5.12554467e-01 1.40164232e+00 4.65249270e-01
1.51934683e+00 -1.37994578e-02 -4.68577564e-01 8.86828363e-01
1.28404999e+00 3.91526133e-01 9.93857384e-01 5.67053668e-02
1.00410259e+00 3.66561025e-01 8.67873609e-01 1.69543445e-01
3.01928192e-01 6.71625435e-01 6.30202293e-01 -2.55711406e-01
-3.73826504e-01 -1.32804215e-01 3.86271715e-01 6.05051875e-01
-2.32766420e-01 -6.03292704e-01 -9.14271712e-01 5.43904603e-01
-1.96989727e+00 -1.00570393e+00 -1.23535752e-01 1.96917081e+00
3.74019861e-01 2.28197485e-01 -1.64969236e-01 -8.43096823e-02
6.22859776e-01 4.54777390e-01 -7.21064925e-01 3.41698751e-02
-2.86806136e-01 -1.58221439e-01 2.37609565e-01 4.52136129e-01
-1.44598913e+00 1.28109074e+00 6.82247448e+00 7.77403295e-01
-9.42069411e-01 5.29291108e-02 9.25458133e-01 -2.32136041e-01
1.11002356e-01 -3.43592614e-02 -9.29767549e-01 5.39071560e-01
5.14627516e-01 9.82505083e-02 2.79269993e-01 8.46412957e-01
1.44710213e-01 -4.03293818e-01 -1.29272318e+00 1.25155473e+00
4.11336511e-01 -1.43878984e+00 1.69306561e-01 -1.84196997e-02
9.49116170e-01 3.35820377e-01 5.68230785e-02 4.85785097e-01
-1.17635662e-02 -1.34458196e+00 4.94083643e-01 3.84200215e-01
6.06831431e-01 -2.05197394e-01 7.29516923e-01 2.06271652e-02
-1.10269940e+00 1.71767380e-02 -5.08620143e-01 -2.36067563e-01
1.54182747e-01 3.75331849e-01 -3.81182760e-01 2.63487816e-01
1.22676504e+00 1.09059596e+00 -6.93688273e-01 1.21712863e+00
-1.24836132e-01 8.83552313e-01 -3.84214044e-01 5.01793027e-01
2.40386963e-01 -9.14218128e-02 4.48664904e-01 1.14262795e+00
-1.44589096e-01 3.36265534e-01 5.02343595e-01 6.95744514e-01
-2.72523850e-01 -3.18035901e-01 -5.73411226e-01 -1.61953300e-01
1.48495927e-01 1.02594674e+00 -7.90711939e-01 -8.36347282e-01
-5.61081767e-01 9.65699792e-01 4.14726511e-02 8.45093668e-01
-1.29838574e+00 2.17028216e-01 9.74425852e-01 7.06669735e-03
4.97939914e-01 -3.49716216e-01 -1.21476807e-01 -1.62618458e+00
2.64306873e-01 -1.22864640e+00 4.16275889e-01 -9.87225592e-01
-9.20043826e-01 6.47735178e-01 8.87215436e-02 -9.77211297e-01
2.18404889e-01 -7.92207718e-01 -4.18879509e-01 1.55038998e-01
-1.48684776e+00 -9.69070613e-01 -9.61459458e-01 6.21335030e-01
1.31470275e+00 -9.11835060e-02 3.83886695e-01 4.52171206e-01
-6.31180465e-01 3.45903248e-01 -1.08076923e-01 3.54372531e-01
7.92560697e-01 -8.34135234e-01 4.90224987e-01 9.15078223e-01
7.71735549e-01 2.08971187e-01 6.89213395e-01 -6.03434622e-01
-1.43150902e+00 -9.47935641e-01 2.83132166e-01 -6.69796050e-01
3.46993327e-01 -5.24686575e-01 -1.19117868e+00 1.02982771e+00
3.92901838e-01 3.42756033e-01 4.05531734e-01 8.34834278e-02
-4.71199751e-01 -3.70783955e-02 -6.32575691e-01 7.11581886e-01
1.61928606e+00 -4.10941660e-01 -5.61249077e-01 6.42659605e-01
8.51400495e-01 -4.52882677e-01 -6.44169092e-01 9.12397325e-01
5.40197492e-01 -1.12396061e+00 1.18220830e+00 -9.43214953e-01
4.58571732e-01 -4.36434448e-01 -2.23878682e-01 -8.24562728e-01
-5.78772370e-03 -3.73188108e-01 -3.28741312e-01 8.68244290e-01
6.75761849e-02 -2.51523077e-01 1.07408309e+00 6.61989570e-01
-7.80025199e-02 -7.69558609e-01 -6.23882115e-01 -6.85802281e-01
-2.17272416e-01 -7.72314489e-01 1.89476356e-01 6.80891871e-01
-3.90890628e-01 1.96665853e-01 -5.38610756e-01 2.68836617e-01
6.09565258e-01 1.98353603e-01 1.03888631e+00 -9.07769501e-01
-1.80921018e-01 -1.17604561e-01 -7.00514078e-01 -1.50095165e+00
1.26420110e-01 -4.64274704e-01 -1.30850077e-01 -1.44520450e+00
5.08143723e-01 -1.12625815e-01 -2.16096222e-01 3.55559170e-01
-4.62040663e-01 7.85890818e-01 6.38901949e-01 2.73959070e-01
-1.27655578e+00 4.56147313e-01 1.29451489e+00 -5.28731585e-01
1.74218267e-01 -1.96112812e-01 -5.72012305e-01 9.24438000e-01
6.76156223e-01 -1.04927935e-01 -5.14833570e-01 -8.58128846e-01
-3.10186177e-01 -1.23102672e-01 5.80023885e-01 -1.16952646e+00
8.32252800e-02 -1.46303177e-02 6.26657486e-01 -9.55829263e-01
6.26836836e-01 -6.70406878e-01 6.42435998e-02 2.84842908e-01
-1.97161511e-01 -2.28927583e-01 2.68451780e-01 5.78464329e-01
-5.63834310e-01 7.09528476e-02 5.18711686e-01 -3.65578532e-02
-1.46980596e+00 6.15617037e-01 -3.98872972e-01 3.68573725e-01
1.19130695e+00 -4.48618621e-01 -6.62634075e-01 -5.83594739e-01
-9.26776409e-01 5.03463447e-01 3.87830436e-01 8.81498873e-01
8.63287807e-01 -9.47334170e-01 -7.64227509e-01 2.55862266e-01
1.44378603e-01 6.95541725e-02 5.91605604e-01 6.06931746e-01
-7.09660113e-01 4.62820143e-01 -1.96003124e-01 -1.02812958e+00
-1.54449987e+00 7.40239084e-01 2.07125545e-01 1.56863019e-01
-8.00668836e-01 1.03062749e+00 9.92523968e-01 2.97707081e-01
5.91270864e-01 -3.12433183e-01 -1.88437343e-01 -2.26108372e-01
8.67473185e-01 1.53390914e-01 -2.60855198e-01 -4.99268442e-01
-2.81971812e-01 4.97700959e-01 -3.17095250e-01 3.79682958e-01
1.06689346e+00 -2.25528434e-01 2.54602790e-01 2.23776460e-01
1.12124860e+00 -4.44131434e-01 -1.71140242e+00 -2.76363701e-01
-1.25922069e-01 -9.86186981e-01 -1.60724506e-01 -7.55999207e-01
-1.37305808e+00 1.00362766e+00 6.08306468e-01 2.96058878e-03
1.16430283e+00 2.60547429e-01 6.80661798e-01 4.50680882e-01
3.01259428e-01 -1.02918506e+00 2.95978010e-01 5.56471586e-01
1.01465881e+00 -1.80297101e+00 2.23677233e-02 -8.87950778e-01
-8.62411201e-01 8.16469908e-01 1.04219699e+00 -5.53639196e-02
4.71285164e-01 3.19446951e-01 3.72827649e-01 -2.36845523e-01
-9.39303398e-01 -2.61927843e-01 3.26613814e-01 5.90420723e-01
2.32005119e-01 -2.23829839e-02 1.91802960e-02 4.09944713e-01
1.47686541e-01 3.45919058e-02 4.47222799e-01 4.83668894e-01
-3.15827668e-01 -6.48787975e-01 -3.95958871e-01 3.97589803e-01
-5.26790142e-01 -9.19538215e-02 -2.57093191e-01 9.62081194e-01
1.24582894e-01 1.03159571e+00 2.76947975e-01 -1.99053794e-01
8.05947408e-02 -1.35972172e-01 3.63210052e-01 -3.97640914e-01
-9.93036702e-02 -2.19820403e-02 5.68838455e-02 -1.19775212e+00
-5.11708021e-01 -4.31376964e-01 -9.45331216e-01 -1.17079966e-01
-2.09012315e-01 -6.22377172e-02 5.38262784e-01 1.02048850e+00
3.03858459e-01 6.61018193e-01 8.45669135e-02 -1.11081588e+00
5.26143424e-02 -7.25655019e-01 -1.85452923e-01 7.08804429e-01
2.83269674e-01 -8.01520586e-01 -3.66430283e-01 3.42811167e-01] | [9.27838134765625, 0.11483034491539001] |
a9ac4b59-752a-47f2-8160-48e63301ba33 | flame-few-shot-learning-from-natural-language | 2306.08042 | null | https://arxiv.org/abs/2306.08042v1 | https://arxiv.org/pdf/2306.08042v1.pdf | FLamE: Few-shot Learning from Natural Language Explanations | Natural language explanations have the potential to provide rich information that in principle guides model reasoning. Yet, recent work by Lampinen et al. (2022) has shown limited utility of natural language explanations in improving classification. To effectively learn from explanations, we present FLamE, a two-stage few-shot learning framework that first generates explanations using GPT-3, and then finetunes a smaller model (e.g., RoBERTa) with generated explanations. Our experiments on natural language inference demonstrate effectiveness over strong baselines, increasing accuracy by 17.6% over GPT-3 Babbage and 5.7% over GPT-3 Davinci in e-SNLI. Despite improving classification performance, human evaluation surprisingly reveals that the majority of generated explanations does not adequately justify classification decisions. Additional analyses point to the important role of label-specific cues (e.g., "not know" for the neutral label) in generated explanations. | ['Chenhao Tan', 'Yiming Zhang', 'Yangqiaoyu Zhou'] | 2023-06-13 | null | null | null | null | ['natural-language-inference'] | ['natural-language-processing'] | [ 2.74574369e-01 1.06607497e+00 -7.19857693e-01 -6.25381649e-01
-9.58895862e-01 -3.36670011e-01 1.12433577e+00 1.61638036e-01
3.32739390e-02 1.02887571e+00 8.02062809e-01 -8.50340486e-01
-1.58124808e-02 -5.99504590e-01 -5.53647518e-01 -3.05317581e-01
3.75965983e-01 7.17343986e-01 9.87833273e-03 -2.48529553e-01
7.86873758e-01 9.43854004e-02 -1.56816936e+00 8.39509964e-01
9.50033426e-01 1.53757572e-01 -1.87016532e-01 7.88423181e-01
-6.19230449e-01 1.34820914e+00 -6.18683755e-01 -6.51300251e-01
-2.53058761e-01 -7.58302212e-01 -1.13124216e+00 -4.29960974e-02
6.82747185e-01 -5.09431779e-01 -1.65434688e-01 6.36088967e-01
3.32046226e-02 4.08241361e-01 1.00353575e+00 -1.41702962e+00
-1.30664504e+00 1.30110824e+00 -1.05535388e-01 2.61934489e-01
5.01126766e-01 5.92734456e-01 1.42735243e+00 -1.02812243e+00
6.30843699e-01 1.60047650e+00 5.03075719e-01 1.21831894e+00
-1.48665380e+00 -7.63391495e-01 3.29860240e-01 5.79582453e-01
-6.47883177e-01 -4.44517463e-01 3.95173550e-01 -3.57861459e-01
1.20798397e+00 5.11130512e-01 3.68869662e-01 1.50618184e+00
1.85288593e-01 8.57420385e-01 1.31872594e+00 -5.33530772e-01
4.19661701e-01 1.57267943e-01 7.75771141e-01 7.11914361e-01
3.27609599e-01 3.70830059e-01 -7.43515909e-01 -3.06331843e-01
3.46876860e-01 -2.28572220e-01 -5.88647723e-02 -4.55356808e-03
-1.01572263e+00 1.27380276e+00 6.69476688e-01 -9.27340984e-02
-2.16011986e-01 3.51797044e-01 1.97699681e-01 7.24054351e-02
5.08546650e-01 9.04436111e-01 -4.02266026e-01 -2.07582220e-01
-6.04399681e-01 5.53918481e-01 8.78679276e-01 9.65130031e-01
6.20551884e-01 3.42188478e-01 -5.51598191e-01 7.00466514e-01
2.53030986e-01 3.13127726e-01 4.35219407e-01 -1.34574151e+00
1.44587129e-01 3.59362811e-01 4.18903202e-01 -6.35202646e-01
-2.16823533e-01 -2.34918714e-01 -2.58320063e-01 2.44146839e-01
4.96113420e-01 -9.07550007e-02 -1.00770533e+00 1.68659425e+00
-9.37128365e-02 9.40693840e-02 2.86904067e-01 9.54498351e-01
1.05524337e+00 5.71247280e-01 5.22282600e-01 4.36010212e-02
1.13871872e+00 -1.30325139e+00 -4.69501048e-01 -8.66049230e-01
5.57819426e-01 -4.98253822e-01 1.55913734e+00 2.17039987e-01
-9.98487115e-01 -6.34694219e-01 -8.91580820e-01 -3.33461732e-01
-1.77233651e-01 -1.97994560e-01 1.14847064e+00 4.57840174e-01
-8.26179266e-01 4.79963094e-01 -3.88611466e-01 -5.58466673e-01
3.63139153e-01 8.95086676e-02 -5.50966673e-02 -4.06486243e-01
-1.18125176e+00 1.31083858e+00 3.34825456e-01 -5.78048706e-01
-9.43909407e-01 -7.65728831e-01 -1.04474604e+00 3.84230196e-01
5.21279931e-01 -8.37026238e-01 1.76100683e+00 -7.94865251e-01
-1.12957311e+00 5.72163403e-01 -5.64062476e-01 -7.92114317e-01
3.82071823e-01 -4.16323990e-01 -2.96422899e-01 4.25109081e-02
5.68970680e-01 1.28217161e+00 4.42930311e-01 -1.39535880e+00
-5.34187138e-01 1.80317312e-01 5.34360588e-01 5.33983521e-02
9.11272168e-02 -1.57987371e-01 2.22219244e-01 -4.58667636e-01
-4.68990058e-02 -8.58584285e-01 -2.04940483e-01 -2.18427360e-01
-4.29633230e-01 -6.88720763e-01 3.60133022e-01 -4.88371074e-01
8.25333834e-01 -1.60160494e+00 -3.46928149e-01 -1.78668067e-01
2.42038682e-01 -4.97431355e-03 -2.57239461e-01 3.14525753e-01
-3.63269597e-01 6.35407209e-01 -1.36513412e-01 -8.56385157e-02
2.61048257e-01 5.07757783e-01 -7.81933784e-01 -2.26511657e-02
3.88046116e-01 1.04346108e+00 -1.24889958e+00 -3.92576367e-01
4.41430479e-01 2.44126827e-01 -8.52004707e-01 -6.44452404e-03
-6.98839486e-01 1.63077101e-01 -2.50665396e-01 3.87138575e-01
2.87751760e-02 -4.29900885e-01 -2.16951385e-01 2.23195150e-01
-6.92115873e-02 9.33592200e-01 -5.38887501e-01 1.39826035e+00
-3.64452511e-01 7.96206892e-01 -5.89002371e-01 -4.12963092e-01
7.06925154e-01 3.33438545e-01 -4.39285934e-01 -4.01419967e-01
3.35445143e-02 1.70079991e-01 2.52267808e-01 -6.38875842e-01
4.63975638e-01 -5.25202036e-01 -3.99490818e-02 7.26625443e-01
-1.79749519e-01 -4.16444361e-01 1.30488738e-01 6.99628115e-01
1.01617169e+00 9.16058272e-02 6.19092584e-01 -3.61841112e-01
2.41656229e-01 6.48832977e-01 3.59335035e-01 1.29081690e+00
-1.49517372e-01 6.49782717e-01 5.22211671e-01 -5.88713050e-01
-1.06784439e+00 -8.62362266e-01 1.99899063e-01 1.07680917e+00
4.19903994e-02 -4.93811697e-01 -5.53865671e-01 -9.09592390e-01
2.06232727e-01 2.33946657e+00 -8.45781922e-01 -4.38343644e-01
-1.75756067e-01 -2.06371695e-01 4.16761816e-01 7.01881647e-01
1.60622820e-01 -1.20007026e+00 -7.28582025e-01 6.67055324e-02
-3.74259979e-01 -6.79376483e-01 -2.69055009e-01 2.57560879e-01
-8.09435487e-01 -1.00993836e+00 -3.96797992e-02 -1.74025863e-01
6.50912642e-01 5.10400295e-01 1.48837090e+00 7.10806906e-01
-3.86433862e-03 4.05754715e-01 -4.38139975e-01 -5.37061930e-01
-6.31176353e-01 -1.98389858e-01 -6.68443963e-02 -8.20125759e-01
7.72814929e-01 -2.15815350e-01 -2.05716118e-01 8.69146511e-02
-4.07806307e-01 4.34923083e-01 3.90035450e-01 1.18843102e+00
2.32633366e-03 -3.58003587e-01 5.99349320e-01 -1.61083889e+00
7.32467949e-01 -6.08326852e-01 8.00044462e-02 1.68531984e-01
-1.01809764e+00 4.26640928e-01 5.06168544e-01 -2.72547245e-01
-1.37437987e+00 -4.72637028e-01 1.11008622e-02 -3.96228433e-02
-4.17037189e-01 2.23066315e-01 3.12575787e-01 3.73737991e-01
1.21616447e+00 -1.51981294e-01 -1.82014033e-01 4.09275889e-02
8.45838308e-01 4.95714992e-01 6.64761961e-01 -7.98659205e-01
7.89144695e-01 1.70253679e-01 -5.50473809e-01 -3.61738682e-01
-1.52091157e+00 -1.61326930e-01 -1.73861966e-01 -7.71775022e-02
8.07062387e-01 -6.95285916e-01 -5.39980769e-01 -5.90416968e-01
-1.40155959e+00 -4.71838325e-01 -3.58461201e-01 5.31726778e-01
-5.86965561e-01 -5.49984723e-02 -7.02163994e-01 -9.59344506e-01
-2.15645820e-01 -8.58051121e-01 7.13436544e-01 3.30211252e-01
-1.35436404e+00 -8.92558634e-01 -2.43532538e-01 8.36239219e-01
3.13017815e-01 -1.01372562e-01 1.39251518e+00 -1.08074546e+00
-5.09790778e-01 1.13166399e-01 -2.96405733e-01 -1.71225891e-01
-2.41286159e-02 1.45866245e-01 -1.18624961e+00 2.79717565e-01
-1.25091359e-01 -6.86328351e-01 9.52205062e-01 2.04489484e-01
9.41793203e-01 -5.13364136e-01 -2.86197126e-01 -6.53970838e-02
1.03144002e+00 1.66404456e-01 3.65979075e-01 1.99918494e-01
4.38818693e-01 7.53828704e-01 7.74524093e-01 1.82187900e-01
4.80946809e-01 1.25222757e-01 3.42603236e-01 8.06756988e-02
-4.26318794e-01 -5.33095241e-01 4.00831014e-01 1.77374914e-01
1.12021081e-01 -3.64951313e-01 -1.03883493e+00 5.83786547e-01
-1.93993402e+00 -1.28640878e+00 -4.76509482e-01 1.62039435e+00
6.22447848e-01 5.50382853e-01 -4.08781767e-01 -1.56197771e-01
5.26836514e-01 9.91234928e-02 -8.70610476e-01 -8.08147550e-01
6.36224449e-02 2.35954784e-02 1.02012120e-01 1.07049036e+00
-4.42091048e-01 1.23741198e+00 7.31737328e+00 4.67358708e-01
-4.57588851e-01 -8.08292627e-02 6.13333344e-01 -1.58053815e-01
-1.04874730e+00 3.86852294e-01 -5.34762263e-01 1.51173756e-01
1.00340140e+00 -5.41662455e-01 4.33277845e-01 1.07271421e+00
2.68682182e-01 -1.59318760e-01 -1.38325953e+00 5.32195270e-01
3.53807986e-01 -1.59695280e+00 5.13281584e-01 -1.96157768e-01
9.09252286e-01 -2.50211567e-01 -1.39808953e-01 8.81451249e-01
9.73907411e-01 -1.19202161e+00 9.34961081e-01 2.31748760e-01
3.93594921e-01 -7.04343498e-01 7.39984334e-01 6.02036595e-01
-5.37404239e-01 -1.69183478e-01 -6.18493140e-01 -7.91049838e-01
9.90757123e-02 3.13328505e-01 -1.38484740e+00 7.61100799e-02
1.68289229e-01 5.59905350e-01 -6.29380465e-01 4.36100155e-01
-1.22703922e+00 9.52151954e-01 3.36091340e-01 -2.35787481e-01
3.82534921e-01 2.75788635e-01 3.90739828e-01 1.21473098e+00
1.14775084e-01 7.38642216e-01 1.64620996e-01 1.30892813e+00
-1.12109840e-01 -3.20184797e-01 -7.63673663e-01 -1.68374345e-01
6.04662180e-01 1.05249822e+00 -7.06017077e-01 -9.81418908e-01
-3.53684098e-01 7.61197746e-01 4.34296966e-01 3.27487469e-01
-9.68703449e-01 6.38147667e-02 6.45672560e-01 6.56960811e-03
-1.94491953e-01 2.23853350e-01 -8.09822798e-01 -1.03377402e+00
-6.78559303e-01 -1.12455797e+00 3.03279042e-01 -1.55512261e+00
-1.35241127e+00 4.77853984e-01 1.49429008e-01 -5.41559100e-01
-7.35268533e-01 -4.96816188e-01 -1.00974989e+00 8.79896104e-01
-1.14613140e+00 -8.92155826e-01 -3.02522093e-01 3.26245576e-02
1.32071137e+00 -4.56484184e-02 9.87522066e-01 -4.79071558e-01
-1.32705659e-01 2.76410580e-01 -5.86806357e-01 -2.72754997e-01
9.11837995e-01 -1.65087843e+00 9.53273654e-01 8.51300240e-01
2.61556417e-01 1.20147991e+00 1.31380773e+00 -9.33348835e-01
-8.49668205e-01 -7.44323432e-01 1.28050387e+00 -8.76414657e-01
5.48655272e-01 -1.74708664e-02 -1.09039390e+00 1.01571286e+00
5.55708170e-01 -6.11709356e-01 8.18006873e-01 5.94377995e-01
-8.49143565e-01 5.47010839e-01 -1.20871270e+00 1.10430706e+00
1.03244770e+00 -6.49957538e-01 -1.36707091e+00 4.23082948e-01
9.57029700e-01 -2.02138022e-01 -7.73240477e-02 6.37819394e-02
5.47159731e-01 -1.16116250e+00 8.98356438e-01 -1.09276569e+00
1.03496861e+00 -1.22755229e-01 -5.86141013e-02 -1.59281147e+00
-6.44141376e-01 -2.83059686e-01 1.08752795e-01 1.07068169e+00
6.69422925e-01 -3.54227126e-01 7.46699870e-01 1.45154989e+00
-2.91719288e-01 -3.77198607e-01 -3.31372529e-01 -4.52989489e-01
7.94100836e-02 -6.30962431e-01 5.35655320e-01 9.09032345e-01
4.72936898e-01 7.94648767e-01 -2.81237245e-01 7.54074194e-03
5.88086724e-01 1.47551239e-01 7.96900153e-01 -1.19694853e+00
-2.31212437e-01 -3.05720419e-01 2.66152084e-01 -8.35394919e-01
6.10717058e-01 -1.05732393e+00 3.83426577e-01 -1.96366549e+00
5.80216408e-01 -9.74393114e-02 2.98109297e-02 8.18725586e-01
-6.70326173e-01 -8.21897238e-02 2.87542045e-01 8.22515711e-02
-4.47717935e-01 3.30649823e-01 1.08244264e+00 -1.60529912e-01
2.36300364e-01 -4.33136255e-01 -1.45238483e+00 9.18506742e-01
9.26281035e-01 -5.91026425e-01 -7.73442149e-01 -6.02096677e-01
-3.48214954e-02 2.81476434e-02 5.89462936e-01 -6.62131548e-01
6.08337149e-02 -5.88986099e-01 3.99073303e-01 -2.84795374e-01
2.92449534e-01 -2.69476980e-01 -4.21810150e-02 5.49060762e-01
-1.15184772e+00 -5.87753057e-02 2.28042126e-01 5.71447432e-01
9.84145850e-02 -4.56970215e-01 5.65272868e-01 -3.48878264e-01
-8.47182751e-01 -2.50671297e-01 -5.34368277e-01 8.38110596e-02
6.23714864e-01 -2.14249253e-01 -8.91129196e-01 -6.63412452e-01
-5.19745409e-01 4.06996936e-01 4.82993633e-01 5.03377795e-01
6.07222974e-01 -1.28217697e+00 -7.98167884e-01 -1.73514411e-01
3.53021145e-01 -4.67923105e-01 5.57770673e-03 2.58695394e-01
-9.42941681e-02 5.83567083e-01 -1.60714269e-01 -1.28968939e-01
-1.18734956e+00 5.30732751e-01 5.15608042e-02 1.13435313e-01
-7.46582031e-01 1.17722619e+00 3.21082234e-01 -4.27327514e-01
2.88610552e-02 -3.79968315e-01 -1.87456816e-01 -1.79653585e-01
7.46272445e-01 4.29450363e-01 -4.42806274e-01 -1.07584648e-01
-2.73098409e-01 7.06309872e-03 -2.03263864e-01 -4.27189589e-01
1.06848323e+00 2.97993161e-02 1.23756431e-01 8.37484777e-01
5.54320872e-01 -1.19941011e-02 -1.21218348e+00 5.61741218e-02
2.72338957e-01 -5.64791083e-01 -1.95280015e-01 -1.36089003e+00
-2.12191194e-01 1.10627270e+00 -1.85051993e-01 1.92975271e-02
2.58828402e-01 1.23383023e-01 3.88889581e-01 5.98385036e-01
4.14738297e-01 -6.21103644e-01 1.14564523e-01 5.36531687e-01
1.06716287e+00 -1.46473610e+00 -7.62972748e-03 -3.08730155e-01
-9.35069442e-01 1.18639886e+00 1.00090659e+00 2.85944305e-02
-3.02652448e-01 -1.22233354e-01 2.32268661e-01 -3.30263585e-01
-1.49738145e+00 5.76969124e-02 1.54570356e-01 4.94934946e-01
9.11925316e-01 3.11648875e-01 -3.36592138e-01 7.34545290e-01
-6.16046965e-01 -2.11704150e-01 9.30736244e-01 4.21736032e-01
-8.23356926e-01 -8.34429383e-01 -4.63979155e-01 6.79381847e-01
8.52000341e-02 -4.14853990e-01 -8.17856610e-01 9.22442198e-01
-6.31885603e-02 1.43471467e+00 -1.05865717e-01 -2.45110735e-01
1.50984460e-02 5.88072360e-01 1.31513894e-01 -1.26421273e+00
-7.70901561e-01 -2.78472543e-01 5.13621569e-01 -5.06871343e-01
-2.33322218e-01 -5.87744474e-01 -1.70317769e+00 -5.74948967e-01
-1.48890421e-01 6.09404147e-01 1.34350762e-01 1.29351342e+00
4.27796096e-02 7.33351707e-01 -1.35573581e-01 -4.21005726e-01
-7.36475527e-01 -9.71802413e-01 -1.01115495e-01 5.02938509e-01
2.93790877e-01 -7.15979040e-01 -7.15364516e-01 4.42712903e-02] | [9.581384658813477, 6.870041370391846] |
772afc58-d0e7-46bc-99a0-74792ba4fe20 | aligning-bird-eye-view-representation-of | 2305.02909 | null | https://arxiv.org/abs/2305.02909v1 | https://arxiv.org/pdf/2305.02909v1.pdf | Aligning Bird-Eye View Representation of Point Cloud Sequences using Scene Flow | Low-resolution point clouds are challenging for object detection methods due to their sparsity. Densifying the present point cloud by concatenating it with its predecessors is a popular solution to this challenge. Such concatenation is possible thanks to the removal of ego vehicle motion using its odometry. This method is called Ego Motion Compensation (EMC). Thanks to the added points, EMC significantly improves the performance of single-frame detectors. However, it suffers from the shadow effect that manifests in dynamic objects' points scattering along their trajectories. This effect results in a misalignment between feature maps and objects' locations, thus limiting performance improvement to stationary and slow-moving objects only. Scene flow allows aligning point clouds in 3D space, thus naturally resolving the misalignment in feature spaces. By observing that scene flow computation shares several components with 3D object detection pipelines, we develop a plug-in module that enables single-frame detectors to compute scene flow to rectify their Bird-Eye View representation. Experiments on the NuScenes dataset show that our module leads to a significant increase (up to 16%) in the Average Precision of large vehicles, which interestingly demonstrates the most severe shadow effect. The code is published at https://github.com/quan-dao/pc-corrector. | ['Elwan Héry', 'Vincent Frémont', 'Minh-Quan Dao'] | 2023-05-04 | null | null | null | null | ['motion-compensation'] | ['computer-vision'] | [-1.24769464e-01 -4.01563734e-01 2.05833763e-01 -6.41227737e-02
-4.60603297e-01 -9.08631742e-01 8.57876658e-01 2.66927741e-02
-4.00840491e-01 1.19226061e-01 -4.65187989e-02 -1.27795786e-01
3.68868262e-01 -8.64286423e-01 -8.19771707e-01 -5.67230463e-01
2.31922753e-02 5.08780956e-01 8.49386811e-01 -2.06613749e-01
2.74285078e-01 9.44312274e-01 -1.80822897e+00 1.14850990e-01
5.96224844e-01 6.23031497e-01 3.09513539e-01 8.57566118e-01
-2.91727602e-01 4.76624399e-01 -4.14589971e-01 -1.78783923e-01
5.56465149e-01 1.17370978e-01 -3.90374511e-01 6.69213235e-02
1.03966022e+00 -6.39028668e-01 -2.79380172e-01 1.07915473e+00
1.05213851e-01 -8.35430250e-03 4.04684216e-01 -1.48491013e+00
5.15213534e-02 -2.50390828e-01 -6.95166826e-01 3.13551068e-01
3.01167697e-01 4.84836072e-01 8.57238054e-01 -1.23975682e+00
7.56559014e-01 1.45448899e+00 8.06152701e-01 1.72688663e-01
-1.31441009e+00 -6.44010305e-01 -9.96850654e-02 1.32850945e-01
-1.57408011e+00 -5.89277089e-01 4.92695391e-01 -7.92950928e-01
1.13453305e+00 3.47110659e-01 5.55638134e-01 6.47923946e-01
-1.50407121e-01 5.87497115e-01 6.31789386e-01 -5.25398403e-02
6.00994490e-02 1.51697755e-01 9.89477783e-02 3.98366213e-01
6.52274132e-01 2.78300226e-01 -4.99954402e-01 1.80804785e-02
6.11456335e-01 1.32773131e-01 -1.98752478e-01 -7.64152706e-01
-1.25912893e+00 6.14957035e-01 7.07877398e-01 -2.50059515e-02
-3.53160501e-01 2.01670989e-01 4.73677255e-02 -8.96537863e-03
4.85265493e-01 2.62923956e-01 -3.39771926e-01 -8.57503191e-02
-8.27688098e-01 5.40873885e-01 5.53295672e-01 1.06693864e+00
1.13464677e+00 -1.28864363e-01 1.08595528e-01 3.13329041e-01
2.69642979e-01 9.29857254e-01 -1.75649881e-01 -1.14337587e+00
5.50659180e-01 8.48503828e-01 3.27809989e-01 -1.25559950e+00
-5.64821064e-01 -4.08464968e-01 -4.08893138e-01 6.82087600e-01
5.50912142e-01 4.25075293e-02 -7.82568514e-01 1.30673146e+00
8.77268314e-01 5.05836070e-01 -1.62019730e-01 1.16571522e+00
7.94548392e-01 6.06000483e-01 -2.07883060e-01 8.14671293e-02
1.49619591e+00 -6.62744045e-01 -3.91542077e-01 -4.78977352e-01
6.55025840e-01 -8.74394596e-01 6.00748420e-01 -2.39001289e-02
-8.17822039e-01 -4.29450333e-01 -1.00810575e+00 -2.20730007e-01
-2.69336760e-01 -9.61886719e-02 4.52636629e-01 5.01021862e-01
-9.38622177e-01 5.37499249e-01 -1.01942933e+00 -5.23293197e-01
6.29383802e-01 3.26476544e-01 -4.54414040e-01 -4.85302694e-02
-4.88494307e-01 9.64262068e-01 1.35268986e-01 -6.45548552e-02
-5.17024279e-01 -1.25344980e+00 -7.58265615e-01 -7.19445432e-03
4.76982027e-01 -7.07324624e-01 1.09564984e+00 -4.83077258e-01
-1.03195965e+00 8.60235751e-01 -5.17491102e-01 -4.21973526e-01
7.32477903e-01 -4.23694611e-01 -1.38682097e-01 6.55470118e-02
2.48161808e-01 8.30404878e-01 8.08015168e-01 -1.25840366e+00
-1.11654353e+00 -4.75016862e-01 -1.48293912e-01 2.45596588e-01
5.49426265e-02 -1.07213929e-01 -8.15789044e-01 -5.68394214e-02
3.67486089e-01 -1.22037411e+00 -1.22296669e-01 1.78658485e-01
-1.75851196e-01 6.75491020e-02 1.27793050e+00 -2.18488306e-01
7.66280472e-01 -2.21865392e+00 -2.12266266e-01 1.10187903e-02
3.37726563e-01 2.72662669e-01 -8.21310878e-02 2.41222739e-01
1.24340184e-01 -1.55325875e-01 -4.65109833e-02 -6.09824061e-01
-2.51850814e-01 3.34304303e-01 -3.40342253e-01 8.30552161e-01
5.14241278e-01 7.56642640e-01 -9.49273527e-01 -1.86346471e-01
7.22958207e-01 7.24241078e-01 -7.26918876e-01 -1.58804983e-01
-1.02022722e-01 5.56219280e-01 -2.70873398e-01 5.74263513e-01
1.43274748e+00 -1.23651482e-01 -2.11053133e-01 -2.22506747e-01
-7.25530744e-01 4.17863101e-01 -1.55652821e+00 1.46024561e+00
-5.96123114e-02 9.32443500e-01 2.05305010e-01 -2.73673564e-01
6.30730391e-01 -1.30517274e-01 5.80125988e-01 -5.02304971e-01
9.11031589e-02 1.98867947e-01 -2.19256371e-01 -2.19408244e-01
9.07290637e-01 2.13441700e-01 3.66285682e-01 -6.94645122e-02
-2.48512447e-01 -2.74742275e-01 3.22097540e-01 3.55650753e-01
1.21747100e+00 4.20933515e-02 -6.49416680e-03 -1.51175827e-01
4.75841969e-01 5.18154204e-01 5.79991996e-01 6.44690037e-01
-1.69925138e-01 8.03137362e-01 3.58776927e-01 -3.64772439e-01
-1.09143794e+00 -1.01296902e+00 -3.66485536e-01 5.40160298e-01
4.95243877e-01 -5.94772279e-01 -4.42146897e-01 -5.23673177e-01
4.69661713e-01 5.76627553e-01 -3.35678965e-01 6.83138296e-02
-7.26939797e-01 -5.88840663e-01 1.90421194e-01 3.10731649e-01
2.84629762e-01 -4.59162265e-01 -1.09320760e+00 2.94681918e-02
-1.77196488e-01 -1.40934193e+00 -2.37112358e-01 -1.17816493e-01
-1.01084793e+00 -1.06975150e+00 -3.77267212e-01 -1.35805428e-01
6.05013072e-01 1.09690297e+00 1.08858871e+00 9.58788320e-02
-3.07847232e-01 2.24212721e-01 -2.66170144e-01 -4.80808437e-01
-3.81325722e-01 -1.06276698e-01 1.18844278e-01 -1.05531156e-01
6.12794459e-01 -3.30701202e-01 -7.79313028e-01 3.74723762e-01
-6.02934778e-01 1.04198143e-01 2.44636342e-01 3.54628712e-01
4.90369350e-01 -2.55212247e-01 -2.89695591e-01 -4.97717530e-01
-2.56960988e-01 -4.17535871e-01 -1.30804992e+00 -5.65052390e-01
-4.74788994e-01 -3.34408954e-02 1.02361798e-01 -8.97895843e-02
-7.72451341e-01 5.05262852e-01 1.47109836e-01 -8.19009483e-01
-2.45480165e-01 -2.39230394e-01 6.46340027e-02 -1.02322578e-01
7.51256108e-01 -1.27009615e-01 1.35976091e-01 -6.52297258e-01
4.32223380e-01 3.94443125e-01 6.90315127e-01 9.92325321e-02
1.21553731e+00 1.22142839e+00 2.56992012e-01 -1.04875648e+00
-5.83424449e-01 -1.03848946e+00 -7.81697333e-01 -3.38782459e-01
8.74849796e-01 -1.14063501e+00 -8.09601963e-01 3.39304268e-01
-1.44002366e+00 4.79785353e-02 -2.45968759e-01 6.10981047e-01
-1.75788566e-01 2.55103528e-01 -2.62480974e-01 -6.39455259e-01
4.81507834e-03 -1.20575500e+00 1.26150739e+00 1.74653962e-01
-2.10454874e-02 -5.67303956e-01 1.41985878e-01 8.38833526e-02
2.42378771e-01 1.57473415e-01 1.69283703e-01 -1.49385482e-01
-1.20235622e+00 -1.36166692e-01 -4.74222124e-01 1.96201280e-02
-9.99236852e-02 2.23625839e-01 -1.24946094e+00 -3.24127853e-01
-8.96434113e-02 6.72820210e-01 9.75382745e-01 3.86341780e-01
2.32377991e-01 1.45437717e-01 -6.31383061e-01 7.88970709e-01
1.50997388e+00 -1.60148278e-01 5.04424691e-01 5.26612818e-01
1.01632750e+00 6.07111275e-01 7.20656633e-01 4.32519585e-01
4.68567371e-01 1.08923316e+00 9.36848998e-01 -7.48166740e-02
-5.08302152e-01 -6.98960125e-02 3.47949028e-01 2.54337907e-01
7.16399448e-03 1.50039628e-01 -1.16723764e+00 6.02294683e-01
-1.88384295e+00 -1.00753236e+00 -1.10792518e+00 2.34559655e+00
2.22031847e-01 4.14932184e-02 1.20930076e-01 -9.12270322e-02
6.16766989e-01 1.40870940e-02 -3.35740149e-01 1.33189589e-01
-8.62365961e-02 -2.58338779e-01 1.17380142e+00 7.03881383e-01
-1.10872078e+00 1.08432567e+00 5.27631474e+00 3.37430805e-01
-1.22467589e+00 1.56291515e-01 -2.16174453e-01 -3.63419890e-01
4.27503213e-02 3.31768036e-01 -1.30615270e+00 4.65187937e-01
7.95906544e-01 -7.74544105e-03 2.08298817e-01 8.80359292e-01
3.76102984e-01 -2.93610752e-01 -1.01726270e+00 1.18550992e+00
-1.26219094e-01 -1.56440485e+00 -1.87468827e-01 2.12933332e-01
5.45973003e-01 8.09700131e-01 -1.47424564e-01 -4.75299209e-02
1.59153461e-01 -4.73681271e-01 9.37919915e-01 1.15352087e-01
5.71547151e-01 -6.49577618e-01 4.95471239e-01 4.40385967e-01
-1.34859407e+00 3.05289365e-02 -5.10236263e-01 -1.60610616e-01
4.34931099e-01 7.02235878e-01 -1.30905282e+00 3.46963644e-01
8.25220466e-01 8.39901328e-01 -7.41567433e-01 1.26231515e+00
3.82998511e-02 2.01076716e-01 -7.54370391e-01 4.39549714e-01
8.92486051e-02 -2.83339381e-01 1.21909058e+00 1.38990867e+00
4.34585989e-01 -1.04044609e-01 -4.51739430e-02 1.03362536e+00
1.20538279e-01 -3.01604927e-01 -6.49265230e-01 5.28788030e-01
6.06846869e-01 1.43778622e+00 -7.44791806e-01 -4.13181841e-01
-5.07714868e-01 8.66899252e-01 -4.56705783e-03 2.25371972e-01
-8.55319262e-01 -9.92207453e-02 1.34607363e+00 5.57912409e-01
5.94554305e-01 -5.05091488e-01 -3.40076506e-01 -1.14548922e+00
8.75035152e-02 -3.75849038e-01 -6.27313703e-02 -6.98603153e-01
-7.15436637e-01 2.21846655e-01 3.31122875e-02 -1.69833517e+00
-5.29192351e-02 -5.48879325e-01 -3.27744812e-01 9.50616658e-01
-1.63849401e+00 -8.70161176e-01 -7.20253468e-01 5.33853054e-01
4.34666783e-01 3.03840339e-01 3.06055874e-01 6.39525890e-01
-3.72647166e-01 -2.09728745e-03 -4.44616610e-03 -6.88554198e-02
6.61964536e-01 -1.14067435e+00 7.28657722e-01 1.14794695e+00
3.07868749e-01 4.07913715e-01 1.01994455e+00 -6.74218774e-01
-1.69957447e+00 -1.23038328e+00 8.95284235e-01 -1.03631353e+00
4.82740879e-01 -4.20578301e-01 -1.08971632e+00 6.26835823e-01
-2.13193864e-01 2.97124237e-01 -4.08522412e-02 -2.39787325e-01
-4.08694625e-01 -1.12822257e-01 -8.88717413e-01 5.59913278e-01
1.04563498e+00 -2.74484634e-01 -3.18120658e-01 2.68305719e-01
6.91461504e-01 -7.39460588e-01 -4.22161698e-01 5.46604335e-01
3.40641320e-01 -1.24440694e+00 1.19607508e+00 -8.58978927e-02
7.22860247e-02 -9.98652756e-01 -2.12267227e-02 -1.08015335e+00
-3.56320679e-01 -4.16482419e-01 -1.75091505e-01 1.14377654e+00
6.70365617e-02 -6.58762515e-01 8.84284556e-01 4.62987036e-01
-1.95516244e-01 7.73715153e-02 -9.17539895e-01 -7.87520945e-01
-3.81456584e-01 -6.53955460e-01 5.27483284e-01 9.02875483e-01
-4.72074181e-01 3.45716029e-01 3.25130373e-02 7.64329672e-01
7.97562361e-01 1.14272989e-01 1.34978008e+00 -1.35803175e+00
6.55515417e-02 -4.43383217e-01 -6.93147779e-01 -1.26412773e+00
-2.77532548e-01 -8.65279615e-01 8.32515955e-02 -1.35675800e+00
-9.94809344e-02 -5.77792764e-01 2.80057907e-01 1.70151159e-01
-9.38278511e-02 4.55968380e-01 6.38352215e-01 5.43482959e-01
-4.48848724e-01 1.67191982e-01 8.03589642e-01 1.88372627e-01
-4.06673878e-01 2.86825430e-02 -2.06810966e-01 8.67147088e-01
6.41586244e-01 -7.43448019e-01 9.07071605e-02 -6.70789719e-01
1.30720899e-01 -2.72664487e-01 8.16844106e-01 -1.27179873e+00
3.06781530e-01 9.12420601e-02 3.77613664e-01 -1.21470916e+00
5.60097337e-01 -9.87585723e-01 4.27951038e-01 6.04391038e-01
3.98196816e-01 8.22532997e-02 4.14426267e-01 4.00870144e-01
-2.10540928e-02 -3.92076746e-02 8.68904591e-01 4.73351311e-03
-8.97629797e-01 2.34777823e-01 -3.43496919e-01 -2.46343493e-01
9.56990480e-01 -4.45004404e-01 -5.28712332e-01 -2.11191531e-02
-2.27511153e-01 2.78070450e-01 9.55877662e-01 6.69893861e-01
3.28466237e-01 -1.14479935e+00 -7.01777935e-01 3.69006544e-01
2.14748994e-01 3.83502334e-01 2.44842350e-01 1.15075684e+00
-7.43023694e-01 5.31089306e-01 -6.82534352e-02 -1.30665004e+00
-1.53164983e+00 4.36267853e-01 3.16460580e-01 2.25907579e-01
-8.83073807e-01 7.19942033e-01 3.92408282e-01 -1.85851097e-01
-1.09128214e-01 -5.49186230e-01 -1.52979009e-02 1.59191057e-01
6.48919880e-01 6.85149550e-01 2.65856802e-01 -1.04666960e+00
-6.92488074e-01 7.69928575e-01 8.05339590e-02 8.85446817e-02
1.11094344e+00 -3.10186535e-01 1.31508410e-02 1.14614926e-01
1.28387678e+00 3.52923304e-01 -1.59379208e+00 -1.22565791e-01
4.25822362e-02 -7.95027852e-01 1.69614896e-01 -1.09930173e-01
-9.84319687e-01 8.03445339e-01 7.31184125e-01 4.85276580e-02
6.34318709e-01 1.32929876e-01 5.44826984e-01 6.03857562e-02
3.02031785e-01 -5.71250558e-01 -3.18821311e-01 7.84028530e-01
5.43845534e-01 -1.36108887e+00 2.73206890e-01 -6.66127503e-01
-3.48436981e-01 9.48828042e-01 5.47903955e-01 -3.09052438e-01
3.74897957e-01 3.21081430e-01 6.26330497e-03 -3.45976263e-01
-5.90093732e-01 -5.25874615e-01 1.00333422e-01 5.51771224e-01
-1.11120775e-01 -8.77503827e-02 9.65914652e-02 -1.77692249e-01
-2.15919852e-01 -2.16033533e-01 5.64044058e-01 7.38197029e-01
-5.75650215e-01 -8.09995472e-01 -8.10822308e-01 2.87653469e-02
-1.50646791e-01 5.43904230e-02 -1.19965315e-01 8.62274170e-01
1.98750272e-01 9.12168562e-01 6.69057786e-01 -2.58974850e-01
6.77037716e-01 -3.23074728e-01 1.85391128e-01 -4.88739252e-01
-4.18277115e-01 1.71971068e-01 -1.09004848e-01 -1.05747449e+00
-2.90736169e-01 -8.90051901e-01 -1.40959728e+00 -4.33296382e-01
-3.46643388e-01 -3.25614452e-01 9.76460516e-01 4.55559641e-01
7.28562593e-01 3.56773913e-01 4.08960491e-01 -1.40387785e+00
-3.00689697e-01 -4.98682141e-01 -2.19072476e-01 5.94330370e-01
7.49641895e-01 -7.90163875e-01 -5.72944403e-01 7.05760792e-02] | [7.709722995758057, -2.4682867527008057] |
a8054300-14f1-459a-aa2b-e5e7dc5866ea | instance-segmentation-based-semantic-matting | 1904.05457 | null | http://arxiv.org/abs/1904.05457v1 | http://arxiv.org/pdf/1904.05457v1.pdf | Instance Segmentation based Semantic Matting for Compositing Applications | Image compositing is a key step in film making and image editing that aims to
segment a foreground object and combine it with a new background. Automatic
image compositing can be done easily in a studio using chroma-keying when the
background is pure blue or green. However, image compositing in natural scenes
with complex backgrounds remains a tedious task, requiring experienced artists
to hand-segment. In order to achieve automatic compositing in natural scenes,
we propose a fully automated method that integrates instance segmentation and
image matting processes to generate high-quality semantic mattes that can be
used for image editing task. Our approach can be seen both as a refinement of
existing instance segmentation algorithms and as a fully automated semantic
image matting method. It extends automatic image compositing techniques such as
chroma-keying to scenes with complex natural backgrounds without the need for
any kind of user interaction. The output of our approach can be considered as
both refined instance segmentations and alpha mattes with semantic meanings. We
provide experimental results which show improved performance results as
compared to existing approaches. | ['Guanqing Hu', 'James J. Clark'] | 2019-04-10 | null | null | null | null | ['semantic-image-matting'] | ['computer-vision'] | [ 1.18880594e+00 8.55870843e-02 4.06253397e-01 -5.09040475e-01
-5.46502113e-01 -6.44673705e-01 6.76239073e-01 1.50252357e-01
-2.82453924e-01 4.66803372e-01 -3.97216678e-01 -4.32787061e-01
2.07794651e-01 -1.05170488e+00 -8.66373658e-01 -4.08517241e-01
6.79274082e-01 6.12998486e-01 6.79815173e-01 -3.53944212e-01
2.74535865e-01 5.73923469e-01 -1.60074568e+00 5.43154836e-01
1.00749433e+00 5.93240499e-01 5.63403666e-01 1.00009227e+00
-9.48408008e-01 8.92126858e-01 -6.65974855e-01 -5.89968324e-01
3.76735926e-01 -9.53090847e-01 -9.92993355e-01 8.55590343e-01
5.42014658e-01 6.47717938e-02 4.03217494e-01 1.34536505e+00
-1.17376409e-01 1.01504680e-02 3.40867132e-01 -1.05237412e+00
-2.84754485e-01 5.46911299e-01 -6.71127081e-01 -2.71583050e-01
3.18153352e-01 5.33768088e-02 5.27376175e-01 -6.35335684e-01
5.66989839e-01 1.06187737e+00 3.70823026e-01 4.74336684e-01
-1.35354602e+00 -4.09535557e-01 2.19907209e-01 -2.33225822e-02
-1.06037700e+00 -3.00283641e-01 8.44988823e-01 -4.02209878e-01
4.78699267e-01 8.08644354e-01 1.09921920e+00 2.12549880e-01
-3.63341063e-01 9.07010972e-01 1.39948893e+00 -8.53706241e-01
3.54725420e-01 4.42650288e-01 -9.86739621e-02 7.37851501e-01
-6.53954670e-02 -6.65432692e-01 -2.62559682e-01 2.01306209e-01
1.14871943e+00 -5.31499870e-02 -1.12273097e-01 -1.92721173e-01
-1.31827569e+00 3.48230541e-01 2.89081067e-01 3.70853782e-01
-4.82336998e-01 2.87011981e-01 -3.97740211e-03 3.20655018e-01
5.80917418e-01 6.06927335e-01 -1.32491603e-01 -1.16324894e-01
-1.55504870e+00 3.09302062e-01 7.20787942e-01 1.18758130e+00
1.10148776e+00 3.57063450e-02 -3.46415080e-02 9.18880761e-01
1.39119215e-02 4.81455863e-01 -5.57414182e-02 -1.29453123e+00
1.08350255e-01 8.74176383e-01 3.10703605e-01 -8.32131565e-01
2.67158691e-02 1.54731259e-01 -4.34052289e-01 6.54403448e-01
5.17715096e-01 1.47328570e-01 -1.33181584e+00 1.13295710e+00
5.72170973e-01 1.64341331e-01 -1.74159706e-01 7.17510760e-01
7.11503804e-01 7.47413754e-01 -3.20188478e-02 -2.11172119e-01
1.48286259e+00 -1.36231172e+00 -8.86232734e-01 -1.78214133e-01
1.74281850e-01 -1.18569684e+00 1.45659864e+00 6.89075053e-01
-1.61703908e+00 -5.70061028e-01 -1.01152992e+00 -1.75205663e-01
-3.60670656e-01 -2.33644992e-02 6.71471953e-01 8.83216977e-01
-1.08281362e+00 5.38882971e-01 -6.97107434e-01 -2.63343692e-01
3.72838795e-01 1.83517292e-01 -1.89992294e-01 8.24358687e-02
-7.00344801e-01 7.56888807e-01 6.25359476e-01 4.20833081e-02
-5.98646462e-01 -6.66277885e-01 -7.12523878e-01 -2.53700495e-01
8.14424872e-01 -7.70016134e-01 1.33584750e+00 -1.92591441e+00
-1.76173103e+00 1.16686869e+00 -2.75426984e-01 -1.54687524e-01
8.00991833e-01 -2.94178784e-01 -1.90234125e-01 3.36543351e-01
3.39869745e-02 5.51732659e-01 1.19400072e+00 -1.80728626e+00
-8.67528200e-01 -9.82850343e-02 2.97453135e-01 4.79599059e-01
1.88483730e-01 1.58257082e-01 -9.86510098e-01 -8.34499180e-01
2.64236063e-01 -7.78520644e-01 -3.62369269e-01 5.41222952e-02
-4.89694268e-01 3.36866647e-01 9.39913630e-01 -8.90898407e-01
1.09476626e+00 -1.83940411e+00 2.73864388e-01 5.33258140e-01
7.33834878e-02 3.56680959e-01 8.15451145e-02 1.61325410e-01
5.81680872e-02 -3.26781496e-02 -7.31823742e-01 -3.47368687e-01
-2.52278239e-01 1.25958726e-01 2.92610619e-02 -3.99560891e-02
2.15233281e-01 7.38201380e-01 -1.05949342e+00 -6.17450655e-01
5.45892835e-01 3.40992033e-01 -4.01820809e-01 2.31206715e-01
-7.31799543e-01 6.03715718e-01 -1.35702118e-01 5.32962561e-01
7.01189518e-01 -2.90292557e-02 9.70317498e-02 -2.05859959e-01
-1.82663679e-01 -4.60887142e-02 -1.49940610e+00 2.01537561e+00
-5.19944251e-01 4.88871276e-01 2.68068224e-01 -8.52393091e-01
9.76381242e-01 1.40082566e-02 3.65246773e-01 -5.45548975e-01
8.65750760e-02 3.00249875e-01 -5.08236170e-01 -4.99197781e-01
9.28625286e-01 -3.46904397e-01 1.42083749e-01 6.40189886e-01
-3.32908958e-01 -9.79086101e-01 5.06038189e-01 4.12815154e-01
6.65643275e-01 5.24806142e-01 -7.60508999e-02 -2.01771393e-01
5.49951375e-01 4.85417604e-01 1.66600659e-01 6.66093946e-01
5.86622357e-01 8.36931407e-01 3.03874940e-01 -3.71172935e-01
-1.38162923e+00 -1.01800299e+00 2.11560458e-01 9.79736984e-01
4.10148799e-01 -4.21966136e-01 -1.44050312e+00 -3.51008117e-01
-5.66957593e-01 8.61485660e-01 -3.51693183e-01 2.81551868e-01
-6.13079309e-01 -6.64713025e-01 2.25399286e-02 3.41932207e-01
7.87619174e-01 -1.28039753e+00 -6.01489961e-01 2.56728560e-01
-2.78456688e-01 -1.16520023e+00 -4.82048303e-01 -2.95245387e-02
-7.41213739e-01 -1.04361868e+00 -8.82902443e-01 -9.62266922e-01
9.25884545e-01 3.96584809e-01 1.28444159e+00 4.46508020e-01
-5.04033029e-01 4.91635203e-01 -4.78103846e-01 -4.57840115e-01
-8.96431327e-01 -1.99701294e-01 -6.39295280e-01 4.60912466e-01
-3.88158076e-02 -5.28182209e-01 -5.70051193e-01 1.90669432e-01
-1.59414184e+00 9.65414047e-01 3.29681456e-01 2.65310496e-01
8.32701325e-01 1.32356554e-01 1.20400690e-01 -1.63276565e+00
3.78063560e-01 1.49926215e-01 -6.06382430e-01 4.81710345e-01
-1.20114788e-01 -1.00431234e-01 3.20348620e-01 -3.06309193e-01
-1.40456605e+00 3.17492932e-01 -9.48365107e-02 -5.88248633e-02
-3.12311888e-01 1.65454492e-01 -2.22638667e-01 -9.44647044e-02
3.63388479e-01 2.55128592e-01 -5.35317101e-02 -4.37553525e-01
7.98590124e-01 5.82594991e-01 6.62436903e-01 -5.31703055e-01
8.90824258e-01 6.38949037e-01 -2.22986534e-01 -1.12568426e+00
-6.55679047e-01 -2.96359986e-01 -9.25638676e-01 -5.01359701e-01
1.15483570e+00 -6.65650666e-01 -2.02847436e-01 6.54559016e-01
-1.04004979e+00 -8.32733870e-01 -4.86400127e-01 -6.90187886e-02
-7.01073587e-01 3.70944828e-01 -4.12415415e-01 -5.80542386e-01
-1.79673001e-01 -1.21207070e+00 1.15590107e+00 3.70146006e-01
-1.69512004e-01 -8.25940192e-01 -2.37303302e-01 6.76322877e-01
3.81760895e-01 4.69039649e-01 8.36991668e-01 1.60528034e-01
-8.04617584e-01 -6.14202134e-02 -3.38252544e-01 3.68933499e-01
3.51592451e-01 3.38670522e-01 -9.90463614e-01 1.47911176e-01
-2.07467124e-01 -2.63236975e-03 6.86077476e-01 1.56157359e-01
1.16962862e+00 -2.26721525e-01 1.91826536e-03 4.96674627e-01
1.49349034e+00 4.17331219e-01 8.70786607e-01 4.90137130e-01
1.08236146e+00 6.12471461e-01 6.29389167e-01 1.88659996e-01
2.07142875e-01 7.76578844e-01 1.60286158e-01 -6.95398808e-01
-3.71692359e-01 -1.70622021e-01 -7.24484166e-03 6.19925797e-01
-3.11993510e-01 6.18366078e-02 -6.55224979e-01 4.27459568e-01
-1.83164752e+00 -9.56533849e-01 -4.84845877e-01 2.02632546e+00
1.10770178e+00 -1.74870025e-02 3.60109180e-01 4.06163603e-01
8.25484872e-01 -1.67056724e-01 -2.54673868e-01 -6.25178456e-01
6.06554039e-02 5.67841470e-01 2.72263974e-01 7.64682114e-01
-9.74515855e-01 1.34805858e+00 5.98618126e+00 8.40993166e-01
-1.02341926e+00 1.41352206e-01 7.42168307e-01 1.43769562e-01
-5.71477950e-01 1.12025611e-01 -3.71269196e-01 3.08388740e-01
2.68461168e-01 1.22041151e-01 6.65825486e-01 5.63673615e-01
2.40601435e-01 -6.19100392e-01 -7.85890698e-01 9.87351120e-01
1.07367337e-01 -1.30539465e+00 2.64373124e-01 -4.53900427e-01
1.01928341e+00 -8.04621220e-01 -2.31268749e-01 -2.56901801e-01
3.78686517e-01 -8.72772932e-01 1.15594959e+00 5.01389563e-01
7.19308853e-01 -5.98703563e-01 1.81021348e-01 1.55881330e-01
-1.15519369e+00 4.76033896e-01 -8.16031769e-02 -2.97829770e-02
4.06350046e-01 6.37676597e-01 -7.65632749e-01 3.59324932e-01
5.68386436e-01 3.21842939e-01 -6.93775535e-01 9.90530133e-01
-2.94053316e-01 3.67533594e-01 -3.70915711e-01 1.45845383e-01
1.58135742e-01 -6.44926548e-01 2.88643301e-01 1.33571219e+00
8.90716463e-02 2.55639981e-02 3.43719423e-01 1.07181871e+00
9.38281417e-02 4.23323035e-01 -3.02916497e-01 -7.32849911e-02
-7.42181242e-02 1.39605415e+00 -1.63760138e+00 -7.86877632e-01
-1.74948394e-01 1.79838467e+00 -1.34607330e-01 4.08483714e-01
-9.34922695e-01 -4.23354298e-01 9.40650478e-02 4.53588516e-01
1.81981385e-01 -1.85365841e-01 -4.35533434e-01 -1.05746508e+00
4.12435047e-02 -9.39850748e-01 -1.13007680e-01 -1.07756639e+00
-6.42927408e-01 7.22719967e-01 1.19105898e-01 -9.34855461e-01
1.83584079e-01 -3.73180270e-01 -5.97101450e-01 5.84428012e-01
-1.17507124e+00 -1.36725843e+00 -6.94234252e-01 7.65030980e-01
9.41549122e-01 2.30942905e-01 5.59959710e-01 3.33892554e-01
-2.02900007e-01 7.86830857e-02 -2.52906054e-01 -1.76592037e-01
3.21768761e-01 -1.61013472e+00 4.07655776e-01 1.07551432e+00
4.00400847e-01 2.36973807e-01 8.44834149e-01 -7.06423104e-01
-1.11192119e+00 -1.08296585e+00 4.62206781e-01 -1.67376816e-01
2.69347876e-01 -4.58280504e-01 -8.57629001e-01 5.28254330e-01
4.88918871e-01 -3.96652400e-01 3.83789808e-01 -3.39701265e-01
6.36846498e-02 -5.90898283e-02 -8.90287578e-01 8.28180671e-01
8.65664303e-01 -3.36374164e-01 -3.52860659e-01 5.35110056e-01
5.52500129e-01 -5.53248048e-01 -5.38846016e-01 2.32366309e-01
2.82802671e-01 -1.01179779e+00 9.19703066e-01 -1.92061499e-01
3.07925642e-01 -7.81684995e-01 2.57115275e-01 -9.76626635e-01
9.11194384e-02 -9.03956056e-01 6.72200143e-01 1.17938459e+00
4.98892605e-01 -1.66948944e-01 5.96981227e-01 8.68931353e-01
-8.05859491e-02 -8.37190002e-02 -2.53987253e-01 -2.88489074e-01
-4.44358081e-01 -6.79899096e-01 4.92961764e-01 9.47327077e-01
-3.48341495e-01 1.55764669e-01 -3.19677472e-01 -1.15916431e-02
6.02146566e-01 3.79929155e-01 1.19037592e+00 -9.32444513e-01
-4.03953582e-01 -4.77143139e-01 -1.51524156e-01 -7.51553416e-01
-3.05681199e-01 -7.63592005e-01 1.83268368e-01 -1.73183477e+00
1.48820922e-01 -7.30617523e-01 3.17716897e-01 3.71684968e-01
-3.24732631e-01 8.38068128e-01 5.13795555e-01 -2.14792974e-02
-6.64454162e-01 -1.12941032e-02 1.56745946e+00 -2.38330930e-01
-4.63328689e-01 1.67352147e-02 -5.69688797e-01 9.34024572e-01
7.62592435e-01 -3.46560806e-01 -5.09910285e-01 -3.55992913e-01
3.76971185e-01 -4.37100790e-02 3.28379363e-01 -1.01363754e+00
-8.38201568e-02 -4.06183898e-01 2.86425918e-01 -3.57575893e-01
3.30343515e-01 -8.52237105e-01 7.69098520e-01 1.96553245e-01
-2.51509249e-01 -8.09573084e-02 2.06980377e-01 2.91811645e-01
-1.20765671e-01 -5.81473112e-01 8.77996862e-01 -6.82816267e-01
-7.74611294e-01 -1.45641148e-01 -3.93306792e-01 -2.55981058e-01
1.24698281e+00 -6.43836498e-01 1.45431295e-01 -3.92067015e-01
-1.12283182e+00 -2.47470140e-01 9.26578999e-01 2.87627935e-01
5.73457897e-01 -1.09421325e+00 -4.62427318e-01 1.86768368e-01
-2.43840054e-01 3.14832270e-01 2.10706994e-01 6.57280684e-01
-1.46765816e+00 -3.65585059e-01 -1.96061060e-01 -5.05814075e-01
-1.58280504e+00 4.49352980e-01 1.68476656e-01 6.51396886e-02
-7.88195491e-01 7.05149293e-01 3.90048742e-01 -2.80195959e-02
8.88312012e-02 -5.68991005e-01 2.95972656e-02 -7.45347738e-02
7.67134607e-01 2.36413389e-01 1.09394178e-01 -4.81387973e-01
1.40717655e-01 6.95422292e-01 5.94117977e-02 -4.39920306e-01
1.25389981e+00 -2.69144654e-01 -4.69071388e-01 5.98029792e-01
5.78029335e-01 8.67985412e-02 -1.07269764e+00 -6.88483715e-02
-3.14426608e-02 -7.23426878e-01 -2.22735517e-02 -7.29343414e-01
-1.17577851e+00 7.98388600e-01 3.55003148e-01 4.97291535e-01
1.43019867e+00 -1.11812189e-01 7.40212858e-01 8.34694430e-02
4.17680830e-01 -1.23426843e+00 2.78807074e-01 8.38433206e-03
9.16563690e-01 -1.00943804e+00 4.73634666e-03 -8.41574192e-01
-8.06600690e-01 1.16046298e+00 4.05445069e-01 -2.34678596e-01
3.48672211e-01 3.90122294e-01 3.15417439e-01 -2.82756537e-01
-8.80265310e-02 -3.61976594e-01 3.64903331e-01 3.30923766e-01
4.12825823e-01 8.46900791e-02 -3.02564979e-01 6.05688319e-02
-1.10092729e-01 1.92907825e-02 5.74214220e-01 1.08979309e+00
-6.24360740e-01 -1.33253098e+00 -5.93536317e-01 2.85272956e-01
-4.27590668e-01 -1.31135032e-01 -5.89269400e-01 5.48754454e-01
2.20431417e-01 7.70451367e-01 9.25975144e-02 -7.33593032e-02
3.31791788e-01 6.59951344e-02 7.80523121e-01 -7.94339538e-01
-8.28540385e-01 4.44787264e-01 6.24145120e-02 -5.36906064e-01
-9.69326854e-01 -3.89169186e-01 -1.26444566e+00 -1.71955332e-01
-3.08961481e-01 -8.23186859e-02 8.17185760e-01 9.40601051e-01
-1.47670865e-01 7.14884818e-01 2.21604049e-01 -1.02829206e+00
6.17821991e-01 -7.34557331e-01 -5.66961348e-01 8.42812598e-01
-1.68876722e-02 -1.47378057e-01 2.22389609e-01 9.99987960e-01] | [10.757508277893066, -0.7810491919517517] |
af6b2958-7c88-42fc-ae21-48067f6b6db7 | data-audit-identifying-attribute-utility-and | 2304.03218 | null | https://arxiv.org/abs/2304.03218v1 | https://arxiv.org/pdf/2304.03218v1.pdf | Data AUDIT: Identifying Attribute Utility- and Detectability-Induced Bias in Task Models | To safely deploy deep learning-based computer vision models for computer-aided detection and diagnosis, we must ensure that they are robust and reliable. Towards that goal, algorithmic auditing has received substantial attention. To guide their audit procedures, existing methods rely on heuristic approaches or high-level objectives (e.g., non-discrimination in regards to protected attributes, such as sex, gender, or race). However, algorithms may show bias with respect to various attributes beyond the more obvious ones, and integrity issues related to these more subtle attributes can have serious consequences. To enable the generation of actionable, data-driven hypotheses which identify specific dataset attributes likely to induce model bias, we contribute a first technique for the rigorous, quantitative screening of medical image datasets. Drawing from literature in the causal inference and information theory domains, our procedure decomposes the risks associated with dataset attributes in terms of their detectability and utility (defined as the amount of information knowing the attribute gives about a task label). To demonstrate the effectiveness and sensitivity of our method, we develop a variety of datasets with synthetically inserted artifacts with different degrees of association to the target label that allow evaluation of inherited model biases via comparison of performance against true counterfactual examples. Using these datasets and results from hundreds of trained models, we show our screening method reliably identifies nearly imperceptible bias-inducing artifacts. Lastly, we apply our method to the natural attributes of a popular skin-lesion dataset and demonstrate its success. Our approach provides a means to perform more systematic algorithmic audits and guide future data collection efforts in pursuit of safer and more reliable models. | ['Mathias Unberath', 'Mohammad Mehdi Farhangi', 'Nicholas Petrick', 'Nathan Drenkow', 'Mitchell Pavlak'] | 2023-04-06 | null | null | null | null | ['causal-inference', 'causal-inference'] | ['knowledge-base', 'miscellaneous'] | [ 7.86190450e-01 3.58706176e-01 -4.96194333e-01 -5.94477236e-01
-9.48351741e-01 -4.85664457e-01 6.44260287e-01 4.07054961e-01
-4.79184747e-01 7.69688904e-01 2.70885646e-01 -7.93265820e-01
-3.08858305e-01 -6.64613366e-01 -9.04267848e-01 -5.91113985e-01
-8.04931298e-02 1.12498216e-01 -3.98225248e-01 3.78391445e-01
3.06339145e-01 4.47889447e-01 -1.33233416e+00 2.97843993e-01
8.99642110e-01 8.90617549e-01 -4.98813123e-01 2.39743620e-01
4.70063210e-01 6.90150678e-01 -5.56333899e-01 -8.89868379e-01
1.44578844e-01 -3.43309522e-01 -6.22296631e-01 2.88117630e-03
6.41983271e-01 -6.29659116e-01 8.35219547e-02 1.14938760e+00
5.09375870e-01 -4.58343267e-01 1.06313026e+00 -1.38727939e+00
-6.23368084e-01 5.08560836e-01 -5.88757336e-01 -4.06753272e-02
1.86704665e-01 5.82553446e-01 1.12293851e+00 -5.13679147e-01
6.20114744e-01 1.24642158e+00 8.39047730e-01 5.36564589e-01
-1.57375705e+00 -1.08610272e+00 -7.46024251e-02 3.25040184e-02
-9.42693055e-01 -5.65745056e-01 6.70776188e-01 -6.84622586e-01
2.33882383e-01 3.37097913e-01 4.56361830e-01 1.64401317e+00
3.14967215e-01 6.10793769e-01 1.32021081e+00 -5.03819346e-01
4.28226143e-01 4.18676764e-01 2.08312467e-01 7.07960904e-01
1.01696718e+00 5.60734212e-01 -4.34584051e-01 -6.33103430e-01
3.31084222e-01 -2.06585914e-01 -8.33072215e-02 -5.29126942e-01
-1.15842259e+00 9.44615602e-01 2.32587814e-01 -1.79707885e-01
-4.46796358e-01 1.47752896e-01 2.52489775e-01 -1.60545468e-01
3.32921594e-01 8.18978965e-01 -6.07258201e-01 1.96768060e-01
-8.26428235e-01 3.35498810e-01 5.72864950e-01 5.41426361e-01
3.76625448e-01 -2.34336048e-01 -4.00936753e-01 6.31512284e-01
3.53601187e-01 5.14048815e-01 1.84142396e-01 -1.04540813e+00
1.74649626e-01 5.73404849e-01 3.15049022e-01 -1.10950673e+00
-3.04988205e-01 -3.11773181e-01 -7.28693366e-01 3.64385337e-01
6.03050768e-01 -1.96091622e-01 -1.03423512e+00 1.92840469e+00
3.59638691e-01 -1.35472760e-01 -6.28011227e-02 5.59296727e-01
4.37520534e-01 -1.58980995e-01 5.79683185e-01 -2.78716862e-01
1.51374757e+00 1.38906226e-01 -6.04732394e-01 -1.20484598e-01
8.81901562e-01 -4.03278202e-01 1.24921405e+00 3.52863699e-01
-7.51521230e-01 6.93777297e-03 -9.68272448e-01 1.51194066e-01
-1.39970049e-01 -4.09972332e-02 8.39612961e-01 1.04603028e+00
-4.63298470e-01 5.22625506e-01 -6.35692239e-01 -1.73514068e-01
1.09063494e+00 2.82999724e-01 -2.83768654e-01 -1.25623703e-01
-1.14951599e+00 6.84378147e-01 1.26493633e-01 -1.18815443e-02
-1.12985098e+00 -1.10853410e+00 -7.69372165e-01 -3.69927064e-02
3.84840041e-01 -7.67201900e-01 1.15431106e+00 -1.09649134e+00
-6.48221254e-01 1.08372498e+00 5.11459820e-02 -4.42407012e-01
7.03777015e-01 -9.45830494e-02 -2.06985608e-01 -1.21297859e-01
1.34642735e-01 4.95491207e-01 8.04634750e-01 -1.32112586e+00
-5.93187332e-01 -5.53665280e-01 -5.41714728e-02 -2.78783441e-01
-5.34038365e-01 1.91614926e-01 2.80508757e-01 -7.20822573e-01
-2.82651722e-01 -9.22068179e-01 -2.95997947e-01 2.57432878e-01
-8.58415604e-01 1.64579272e-01 3.04628998e-01 -5.68123400e-01
1.02397907e+00 -2.04461503e+00 -5.49390435e-01 3.42849135e-01
3.68861824e-01 9.18877274e-02 6.34162128e-02 -1.64398327e-01
-1.78690150e-01 5.82183242e-01 -4.51621205e-01 8.50564912e-02
-1.87317915e-02 -1.15050681e-01 -2.41114914e-01 6.28909171e-01
4.97623861e-01 7.23197222e-01 -8.31528842e-01 -6.48570120e-01
-7.30248615e-02 2.84369916e-01 -8.22239935e-01 -3.35590467e-02
-4.58736382e-02 2.46180128e-02 -4.42940384e-01 9.29492593e-01
3.73981386e-01 -6.55415654e-02 2.44134694e-01 -2.96428621e-01
3.82322729e-01 2.66109079e-01 -8.20148408e-01 8.68427336e-01
-2.71822453e-01 5.10232925e-01 -7.28648603e-02 -7.67168641e-01
6.17810488e-01 1.12889275e-01 3.06117237e-01 -4.10713911e-01
2.35453144e-01 1.45782128e-01 2.65670508e-01 -5.83654165e-01
1.92844316e-01 -3.54813010e-01 -5.96588850e-02 5.02268434e-01
-2.97822684e-01 -5.19415997e-02 -2.92960525e-01 1.28855675e-01
1.25445497e+00 -2.77806640e-01 4.38376635e-01 -2.34237537e-01
-3.24153714e-02 2.51912475e-01 7.45318592e-01 9.84969735e-01
-4.93242204e-01 4.73934114e-01 7.76485264e-01 -4.21430767e-01
-1.07051122e+00 -1.08723640e+00 -6.08607113e-01 5.73872328e-01
-3.18248749e-01 1.40273407e-01 -5.91487944e-01 -1.04558551e+00
4.92634088e-01 1.11169183e+00 -1.09968412e+00 -5.92718244e-01
-1.73519962e-02 -1.27059221e+00 7.33103156e-01 5.49318433e-01
2.22096622e-01 -9.01690066e-01 -8.90167534e-01 -1.96354926e-01
8.73721689e-02 -8.59800816e-01 -1.85364023e-01 -5.22503741e-02
-7.63143480e-01 -1.57166851e+00 -3.66605937e-01 -2.27633089e-01
9.29814398e-01 -2.51791060e-01 1.18708706e+00 1.42139405e-01
-5.48096359e-01 3.45401436e-01 1.46669708e-02 -1.01070476e+00
-7.27514207e-01 -4.32012439e-01 4.95847836e-02 7.15371817e-02
6.69001937e-01 -3.32288712e-01 -7.25037873e-01 1.84423357e-01
-7.91170120e-01 -1.14110664e-01 7.96641409e-01 1.07282031e+00
4.34851050e-01 -1.81538612e-01 5.93452752e-01 -1.32650280e+00
8.32256913e-01 -6.00767434e-01 -5.73188245e-01 3.73102486e-01
-1.22730911e+00 5.41988201e-02 1.81543425e-01 -6.67964339e-01
-1.03610909e+00 1.32752135e-01 4.69451070e-01 -1.17725044e-01
-3.17110777e-01 5.65146089e-01 -2.90859967e-01 1.78950012e-01
1.12258422e+00 -1.28118426e-01 7.67509565e-02 -1.40228212e-01
3.16990137e-01 6.94503129e-01 3.34799916e-01 -7.31957912e-01
7.43827760e-01 5.00546753e-01 1.88652247e-01 -3.52453828e-01
-8.62580836e-01 1.85098171e-01 -2.66930044e-01 -1.41895831e-01
6.20921195e-01 -7.57275224e-01 -8.71458650e-01 3.48412186e-01
-8.17183316e-01 -3.56695235e-01 -1.91507787e-01 4.86781120e-01
-4.39019203e-01 1.20192938e-01 -3.13902259e-01 -1.11515164e+00
-2.31448084e-01 -1.12103426e+00 9.46238458e-01 -1.15057908e-01
-7.57489502e-01 -8.42424929e-01 -2.18947843e-01 5.10726154e-01
2.22792730e-01 6.25086308e-01 1.37847638e+00 -9.15382743e-01
-4.34164912e-01 -4.67001498e-01 -2.91351825e-01 2.81865478e-01
3.25037628e-01 3.31963301e-01 -1.17138422e+00 -1.19675532e-01
-1.19592175e-01 -5.85596085e-01 8.21447313e-01 6.40110493e-01
1.44909739e+00 -6.27377570e-01 -4.99354750e-01 4.88427430e-01
1.09687734e+00 2.25383341e-01 5.03280401e-01 3.93710107e-01
6.20103896e-01 9.13388848e-01 6.02671564e-01 3.82418573e-01
1.61374241e-01 4.98145431e-01 4.13100541e-01 -2.76389301e-01
1.12398818e-01 -3.40691566e-01 1.37821689e-01 -4.51814532e-01
1.21161059e-01 4.18374985e-02 -9.78637397e-01 7.08549798e-01
-1.36778522e+00 -8.97074580e-01 5.44340648e-02 2.46246743e+00
1.14954925e+00 4.08753961e-01 1.53698683e-01 1.80940226e-01
6.27427816e-01 -1.81590721e-01 -7.71738946e-01 -3.61308336e-01
1.54444380e-02 -1.13928497e-01 7.16941595e-01 1.49662837e-01
-1.01024854e+00 4.09909248e-01 6.95029831e+00 5.82897782e-01
-8.37317586e-01 -2.57505119e-01 1.34125698e+00 -2.84795791e-01
-7.86358178e-01 -1.55417532e-01 -4.18672860e-01 4.49316621e-01
8.70978355e-01 -3.64643157e-01 1.55565500e-01 8.20752382e-01
4.34372514e-01 -1.31792098e-01 -1.47216475e+00 5.16963422e-01
-1.61807731e-01 -1.34502578e+00 1.15240589e-01 3.79411638e-01
6.60197616e-01 -5.19227743e-01 4.22288120e-01 7.53859477e-03
6.84205592e-01 -1.16336811e+00 7.82854378e-01 2.62076139e-01
1.09988606e+00 -5.85888028e-01 7.55307913e-01 -8.15673694e-02
-1.32397071e-01 -3.02324682e-01 -4.91802348e-04 -1.55385036e-03
-4.20035839e-01 9.04320538e-01 -1.33313239e+00 9.18093175e-02
5.81319153e-01 3.82695645e-01 -5.41342556e-01 8.09894681e-01
-2.85496026e-01 1.02093506e+00 -1.42311797e-01 -6.43317997e-02
-3.10252339e-01 2.98085183e-01 3.01509023e-01 1.09198070e+00
1.30969554e-01 7.36602247e-02 -5.31040251e-01 1.19135439e+00
-2.95373589e-01 -6.80822805e-02 -8.02027404e-01 -2.12430626e-01
7.42489636e-01 1.07470953e+00 -4.56084073e-01 -1.52562127e-01
-2.97540098e-01 2.28989020e-01 -2.20679622e-02 4.70532358e-01
-5.06228685e-01 -8.18485394e-02 8.33455861e-01 1.40799880e-01
-2.57067680e-01 3.95530760e-01 -9.76892173e-01 -7.22655177e-01
5.27335284e-03 -1.31573427e+00 6.19140506e-01 -5.45951962e-01
-1.41921902e+00 4.38388251e-02 1.22926615e-01 -9.24940050e-01
-2.11689696e-01 -6.76943481e-01 -3.69094372e-01 9.42757785e-01
-1.23871100e+00 -1.16889751e+00 -1.21214323e-01 4.16528434e-01
1.78370532e-02 -1.77549005e-01 8.37628782e-01 -7.95579050e-03
-7.09789991e-01 1.13927448e+00 -2.54919678e-01 2.50214607e-01
9.18877602e-01 -1.23682213e+00 2.23388359e-01 7.71500111e-01
-8.52278545e-02 8.12698722e-01 9.21787143e-01 -9.09401238e-01
-1.08272576e+00 -1.03529251e+00 6.58605516e-01 -8.83741379e-01
6.57955229e-01 -2.16977999e-01 -7.41924524e-01 6.77424073e-01
-5.71216822e-01 -1.72278821e-01 8.97129655e-01 5.40927947e-01
-5.81554294e-01 -1.53093532e-01 -1.53344119e+00 7.34104574e-01
9.04929280e-01 -4.53662157e-01 -5.48608541e-01 1.77024484e-01
5.39419830e-01 -1.29906870e-02 -6.70597196e-01 6.81036055e-01
8.65453362e-01 -9.39057350e-01 8.40172827e-01 -1.10460520e+00
8.52355778e-01 1.38350025e-01 -1.85160205e-01 -1.11160767e+00
-2.06474558e-01 -3.49008322e-01 2.80547947e-01 1.20210719e+00
8.18607926e-01 -5.25804520e-01 9.04583693e-01 1.35654020e+00
1.71238467e-01 -8.24386418e-01 -8.48265469e-01 -4.72449362e-01
1.08662307e-01 -6.99059129e-01 6.66364133e-01 1.25117564e+00
-1.43330768e-01 -8.81403238e-02 -3.41436446e-01 3.65320772e-01
9.56028402e-01 -1.02236927e-01 7.23324120e-01 -1.31107366e+00
-8.07753578e-02 -4.08030659e-01 -3.57086033e-01 9.99856964e-02
1.09186843e-01 -5.70605636e-01 4.77817543e-02 -1.11309707e+00
6.84449255e-01 -6.21302843e-01 -4.03579712e-01 7.35654950e-01
-5.66501379e-01 5.12312800e-02 -9.32424217e-02 1.50774820e-02
9.33620110e-02 1.64077714e-01 9.38916624e-01 -3.53557944e-01
-5.49000874e-02 5.66492304e-02 -1.30173707e+00 7.89936781e-01
6.02179348e-01 -6.94408953e-01 -3.44928354e-01 -2.19013207e-02
2.48607263e-01 -1.84337705e-01 8.49009752e-01 -4.27230835e-01
-2.31158316e-01 -4.71534103e-01 5.82552016e-01 3.50156985e-02
1.71721932e-02 -8.14767182e-01 4.30014729e-02 6.97441876e-01
-9.37419116e-01 -2.84457773e-01 2.85563946e-01 6.32811964e-01
2.17480958e-01 1.15617290e-02 5.80335736e-01 5.52885309e-02
-1.95187226e-01 4.53022271e-02 -1.29234120e-01 3.13779861e-02
1.02223182e+00 -9.44646746e-02 -4.85602617e-01 -2.77631462e-01
-3.55956703e-01 8.94518495e-02 5.04180372e-01 2.50770658e-01
5.34244537e-01 -1.09169292e+00 -8.98708105e-01 2.05746263e-01
4.36094165e-01 -3.23154360e-01 -1.12509258e-01 7.12958336e-01
-8.24937150e-02 1.32551312e-01 -8.40094984e-02 -4.06035542e-01
-1.29741883e+00 7.80768633e-01 4.28816527e-01 2.98480392e-02
-1.39683500e-01 6.31060779e-01 4.62506384e-01 2.33921614e-02
3.19826573e-01 -3.14210534e-01 1.05754219e-01 6.44521117e-02
5.39227784e-01 3.88405651e-01 -3.56915183e-02 -1.16600424e-01
-4.94591177e-01 -1.47449151e-02 -1.76838979e-01 -5.96436039e-02
1.19689155e+00 2.34018207e-01 1.90148100e-01 1.00484051e-01
7.72190869e-01 2.99893785e-02 -1.31751549e+00 1.10045731e-01
1.87020808e-01 -6.29484475e-01 1.20664947e-01 -1.28705394e+00
-8.20451260e-01 6.38085067e-01 7.58596838e-01 8.46129507e-02
1.10408747e+00 -7.81621933e-02 1.95126921e-01 7.66166821e-02
1.84325501e-01 -8.55570316e-01 -9.74988267e-02 -4.69492972e-01
7.81316221e-01 -1.61876559e+00 1.89097777e-01 -2.85753608e-01
-4.77672428e-01 6.40110791e-01 4.78517711e-01 4.24511760e-01
3.38843167e-01 1.24816038e-01 2.05415517e-01 -2.85724819e-01
-6.71667039e-01 1.98683932e-01 3.88183802e-01 7.68851936e-01
4.17591333e-01 3.36409122e-01 -3.85972708e-01 6.65336788e-01
-1.37916371e-01 1.77997336e-01 5.91490448e-01 4.56750482e-01
6.89744800e-02 -7.91883707e-01 -5.62108159e-01 1.09343433e+00
-8.03140163e-01 -1.59496501e-01 -6.28907502e-01 9.17732716e-01
2.86043361e-02 9.76831913e-01 -1.00460358e-01 -4.18074161e-01
4.11989570e-01 6.55174553e-02 2.58261114e-01 -4.94487762e-01
-2.97595829e-01 -2.96593189e-01 5.20110071e-01 -5.83504498e-01
-3.57830703e-01 -1.01848495e+00 -6.46970451e-01 -1.85271904e-01
-3.17105323e-01 -3.27306598e-01 4.96435374e-01 9.71447587e-01
2.86384761e-01 3.41499418e-01 5.00718832e-01 -3.37093323e-01
-9.70906198e-01 -7.23354518e-01 -4.52506691e-01 8.37820530e-01
4.72103864e-01 -6.75448120e-01 -6.11853719e-01 5.89668602e-02] | [8.752418518066406, 5.0290727615356445] |
9a857eb5-88ba-4f85-856e-207708801152 | multiple-tasks-integration-tagging-syntactic | null | null | https://aclanthology.org/2021.eacl-main.66 | https://aclanthology.org/2021.eacl-main.66.pdf | Multiple Tasks Integration: Tagging, Syntactic and Semantic Parsing as a Single Task | Departing from both sequential pipelines and monotask systems, we propose Multiple Tasks Integration (MTI), a multitask paradigm orthogonal to weight sharing. The essence of MTI is to process the input iteratively but concurrently at multiple levels of analysis, where each decision is based on all of the structures that are already inferred and free from usual ordering constraints. We illustrate MTI with a system that performs part-of-speech tagging, syntactic dependency parsing and semantic dependency parsing. We observe that both the use of reinforcement learning and the release from sequential constraints are beneficial to the quality of the syntactic and semantic parses. We also observe that our model adopts an easy-first strategy that consists, on average, of predicting shorter dependencies before longer ones, but that syntax is not always tackled before semantics. | ["Timoth{\\'e}e Bernard"] | 2021-04-01 | null | null | null | eacl-2021-2 | ['semantic-dependency-parsing'] | ['natural-language-processing'] | [ 2.62946606e-01 6.07586384e-01 1.34947971e-01 -5.59210122e-01
-7.68708467e-01 -7.97445118e-01 5.50084412e-01 3.96294624e-01
-7.06612885e-01 7.47803748e-01 4.87386018e-01 -6.04480147e-01
-6.99415430e-02 -5.14919221e-01 -5.28147459e-01 -3.50864679e-01
-1.41784161e-01 7.91302323e-01 5.20873845e-01 -2.91359127e-01
7.06265345e-02 1.66974902e-01 -1.31686974e+00 6.12254322e-01
6.18205607e-01 5.45759857e-01 5.25074124e-01 1.04082596e+00
-6.24320447e-01 9.70795631e-01 -5.09697080e-01 -6.85165405e-01
-1.70079861e-02 -6.61577284e-02 -1.25322771e+00 -9.07481089e-02
-4.66659293e-02 7.28036538e-02 1.69262528e-01 8.70355129e-01
1.19598910e-01 1.04154974e-01 3.53537053e-01 -1.03083479e+00
-3.15563977e-01 1.05988026e+00 -3.08409929e-01 3.12434822e-01
3.26440960e-01 -1.65805835e-02 1.59485602e+00 -8.50057542e-01
5.75403035e-01 1.35805869e+00 4.16958600e-01 4.34170008e-01
-1.29801035e+00 -2.64291823e-01 7.37833261e-01 -5.97315319e-02
-5.30453742e-01 -4.70382512e-01 5.04203022e-01 -4.34797794e-01
1.59143972e+00 4.63390388e-02 2.18451604e-01 8.63945663e-01
3.16245049e-01 9.64250743e-01 1.05569959e+00 -6.96074963e-01
1.97173670e-01 -2.08592162e-01 5.98631382e-01 7.00084567e-01
7.67366812e-02 -1.12396784e-01 -6.24090970e-01 -6.99495822e-02
2.69145966e-01 -3.67803335e-01 2.51580149e-01 -2.24988624e-01
-1.27612591e+00 6.27804697e-01 -1.93344951e-01 5.66218734e-01
-2.87827969e-01 2.64707506e-01 5.78016818e-01 5.41282773e-01
4.60500002e-01 4.07731950e-01 -1.07918584e+00 -2.84473926e-01
-7.97410309e-01 1.01716816e-01 1.06675136e+00 9.93289232e-01
9.69078898e-01 -2.84880966e-01 -1.44943431e-01 6.71709716e-01
3.36550981e-01 2.56851375e-01 4.60658222e-01 -1.11249721e+00
5.73577404e-01 5.27570844e-01 2.61135131e-01 -1.79998115e-01
-6.75572872e-01 -1.36829186e-02 -1.69702277e-01 3.39257985e-01
8.56546164e-01 -5.22361636e-01 -8.86945724e-01 1.95855558e+00
2.07441866e-01 -8.72909352e-02 1.55306518e-01 5.71914315e-01
3.48539352e-01 3.57498080e-01 5.76820076e-01 -2.01635569e-01
1.57882988e+00 -1.01676250e+00 -6.59853399e-01 -6.44191384e-01
9.24431801e-01 -8.00732076e-01 8.89373660e-01 4.34793204e-01
-1.31981552e+00 -5.15893877e-01 -7.28725016e-01 -2.76925057e-01
-2.36710578e-01 -7.45698065e-02 5.99347591e-01 4.16226417e-01
-1.17787910e+00 7.04540193e-01 -1.06790662e+00 -1.02263190e-01
-5.43660019e-03 4.07976031e-01 -5.00093639e-01 1.33489653e-01
-1.17648649e+00 1.20250261e+00 5.69862366e-01 -6.29226305e-03
-7.33045518e-01 -3.62876564e-01 -1.00945711e+00 2.42512405e-01
8.08540225e-01 -4.36379880e-01 1.70135534e+00 -1.24559951e+00
-1.56967497e+00 8.17402005e-01 -5.18851459e-01 -2.49832764e-01
2.13515058e-01 -5.16424417e-01 -1.55266719e-02 -7.66050667e-02
3.27558741e-02 4.45298523e-01 8.21700394e-01 -1.03640199e+00
-1.01483226e+00 -4.49284911e-01 3.33368927e-01 2.91375995e-01
9.84547660e-02 4.86389905e-01 -3.77863944e-01 -1.30051360e-01
1.01292515e-02 -8.00773144e-01 -4.20059592e-01 -4.64811534e-01
-4.00765747e-01 -5.11144996e-01 2.91193843e-01 -7.54290760e-01
1.12661147e+00 -2.14123058e+00 5.40951371e-01 -3.81630845e-02
3.63102794e-01 8.64301324e-02 -3.62282604e-01 4.65329051e-01
-2.05054790e-01 1.32355437e-01 -3.04526955e-01 -6.51866853e-01
1.51655272e-01 7.03246713e-01 2.85494188e-03 6.04220247e-03
5.32307208e-01 9.88820195e-01 -1.15713167e+00 -5.49455702e-01
2.62924880e-02 -1.63244188e-01 -5.82105637e-01 4.36165333e-01
-5.04686832e-01 4.30662036e-01 -4.78806674e-01 5.58385193e-01
3.18932086e-01 -2.07878307e-01 8.06268036e-01 1.39060691e-01
-4.78732556e-01 7.93517292e-01 -1.13298714e+00 1.77737045e+00
-6.21668220e-01 8.00763443e-02 4.62144643e-01 -1.12615454e+00
4.07393932e-01 5.12646377e-01 3.10857505e-01 -7.18691349e-01
-1.28121480e-01 2.74724185e-01 4.07235920e-01 -4.48543280e-01
2.81338274e-01 -2.46231910e-02 -3.55950534e-01 6.31142437e-01
4.79839474e-01 1.49674177e-01 5.16606927e-01 2.53575087e-01
1.25465131e+00 7.70588756e-01 5.35074890e-01 -5.34505129e-01
3.67490679e-01 1.09742224e-01 9.50635374e-01 8.56986344e-01
-1.69117749e-01 -5.20396121e-02 9.31589484e-01 -5.32212257e-01
-1.05994868e+00 -1.00432503e+00 2.46153325e-01 1.87656915e+00
-1.94945663e-01 -6.15698636e-01 -6.40553355e-01 -1.02127492e+00
-1.47709981e-01 7.54934311e-01 -5.43456912e-01 2.37978011e-01
-1.03235471e+00 -7.00810373e-01 4.36078429e-01 6.55567527e-01
-1.86581641e-01 -1.37336648e+00 -1.03308547e+00 6.13348484e-01
-2.33391691e-02 -1.13521290e+00 -1.99537754e-01 1.02406013e+00
-8.24723959e-01 -1.12653434e+00 -1.07248291e-01 -6.17860317e-01
2.63039857e-01 -9.47896987e-02 1.54399657e+00 3.10008615e-01
1.10394880e-01 6.93136081e-02 -5.63495755e-01 -3.53059173e-01
-6.05052054e-01 2.38257095e-01 -2.85162628e-01 -2.45234832e-01
3.13273638e-01 -6.48302674e-01 -1.09545767e-01 -3.34367529e-02
-6.75773263e-01 2.62028188e-01 7.63879001e-01 8.28497708e-01
3.04248720e-01 -3.75921041e-01 4.37727481e-01 -1.44409156e+00
5.59755325e-01 -4.64376748e-01 -5.42880118e-01 5.82610905e-01
-4.76587653e-01 5.17704427e-01 6.08650029e-01 -7.86715746e-02
-1.30600035e+00 1.95371568e-01 -2.54712164e-01 -8.83138087e-03
-4.21014965e-01 4.52045202e-01 -2.68652588e-01 4.44693714e-01
2.85022169e-01 -1.19564012e-02 -7.67011344e-02 -6.10666633e-01
6.11007452e-01 2.03711659e-01 2.76206881e-01 -1.08584404e+00
4.68116283e-01 5.62341176e-02 -2.54247606e-01 -6.25400841e-01
-9.42464173e-01 -3.81998718e-01 -1.19554520e+00 1.44752664e-02
1.22789598e+00 -6.58467770e-01 -7.23400652e-01 1.97741151e-01
-1.43780088e+00 -8.64107490e-01 -2.93495357e-01 1.68992639e-01
-3.40158015e-01 4.91792738e-01 -9.32791412e-01 -8.29448760e-01
-1.82297662e-01 -1.30981493e+00 1.04417408e+00 -5.11070229e-02
-5.28894782e-01 -1.07349825e+00 -4.67012636e-03 8.44473615e-02
3.28059196e-01 -3.31427962e-01 1.35181999e+00 -1.24695563e+00
-2.25025311e-01 2.91627705e-01 -1.35853499e-01 1.33301198e-01
-2.31839940e-02 -7.06030801e-02 -1.08088911e+00 1.53174708e-02
-3.52290571e-02 -4.68070328e-01 7.82465994e-01 1.73334330e-01
7.32111335e-01 -1.23704016e-01 -2.47038126e-01 1.62729219e-01
1.23048294e+00 3.29035461e-01 2.25425079e-01 3.53238344e-01
6.99016750e-01 1.05470133e+00 4.52455223e-01 1.58069700e-01
6.31287456e-01 4.86319512e-01 4.00245100e-01 7.71961883e-02
1.02636516e-01 -3.91021669e-02 4.35469180e-01 8.22311938e-01
-4.56592254e-02 -5.84767312e-02 -1.03602874e+00 5.60796142e-01
-2.06656837e+00 -7.04473436e-01 -2.90804923e-01 2.13352752e+00
8.97130489e-01 4.89202142e-01 1.10731736e-01 1.13400575e-02
4.56488520e-01 1.18142396e-01 -2.45111257e-01 -9.60086405e-01
2.72232860e-01 5.23160636e-01 5.34922481e-01 1.01764202e+00
-1.01737285e+00 1.26714396e+00 7.09198380e+00 2.87610501e-01
-6.80393934e-01 3.27065170e-01 3.76306474e-01 1.27950832e-02
-4.20908660e-01 4.72684652e-01 -8.65973473e-01 2.95729995e-01
1.19214022e+00 1.37329638e-01 3.86746258e-01 7.47696459e-01
-9.94427055e-02 -3.75725746e-01 -1.36453903e+00 2.35196110e-03
-4.37939584e-01 -8.52513075e-01 -2.27692351e-01 -1.75576523e-01
1.06318392e-01 2.56320864e-01 -6.25459373e-01 4.86306280e-01
1.12767756e+00 -7.72808015e-01 1.02899492e+00 1.89576313e-01
3.36549193e-01 -5.96267223e-01 6.58486068e-01 5.67583919e-01
-1.29321575e+00 -2.39927530e-01 -9.73502472e-02 -2.67602414e-01
4.75549489e-01 6.71824157e-01 -4.40662324e-01 7.35066235e-01
5.53028882e-01 3.66508156e-01 -4.66050506e-01 4.32823449e-01
-7.54700243e-01 6.52218997e-01 -2.40362570e-01 5.90121001e-02
3.57079297e-01 -2.52158791e-01 4.68668818e-01 1.69128704e+00
-2.13600039e-01 -6.02789558e-02 7.21799493e-01 3.91222268e-01
2.80548245e-01 -5.44716977e-03 -4.08033907e-01 1.83395639e-01
4.24091101e-01 1.33413887e+00 -7.50891149e-01 -6.49620056e-01
-8.55098784e-01 1.00695419e+00 7.44267821e-01 2.31527373e-01
-5.35729885e-01 -1.93131313e-01 6.21728241e-01 -2.78861403e-01
5.49436510e-01 -5.41619062e-01 -3.49190801e-01 -1.16358697e+00
1.84029471e-02 -8.11798275e-01 6.33835196e-01 -4.09122407e-01
-1.11661613e+00 6.17628217e-01 3.76115665e-02 -3.94237518e-01
-4.53380018e-01 -7.69878507e-01 -7.16750085e-01 1.04557180e+00
-1.92326784e+00 -1.01447248e+00 3.49704564e-01 3.50041747e-01
8.90782595e-01 3.08716327e-01 1.08718312e+00 4.86496426e-02
-6.15194082e-01 1.64021686e-01 -5.14994800e-01 2.02643171e-01
6.09469950e-01 -1.73309517e+00 7.31274784e-01 9.57301736e-01
2.95166701e-01 6.01997018e-01 6.14799976e-01 -6.69359982e-01
-1.26560688e+00 -7.30836868e-01 1.46645617e+00 -5.25450468e-01
1.06772614e+00 -5.06405354e-01 -1.10054314e+00 9.66311634e-01
6.16154552e-01 -2.37847298e-01 7.37720788e-01 7.06619442e-01
-4.70040500e-01 1.39891803e-01 -6.15017474e-01 2.96886206e-01
9.91647065e-01 -4.58254963e-01 -1.07528627e+00 2.03096792e-01
8.42840970e-01 -3.40702742e-01 -6.88738942e-01 1.09136663e-01
4.26264077e-01 -8.44382286e-01 6.35482311e-01 -9.71372247e-01
4.18444961e-01 -5.53199388e-02 -1.67851597e-01 -1.29856765e+00
-5.87094665e-01 -5.31263888e-01 -1.28518492e-01 1.10541260e+00
9.94173050e-01 -4.51471627e-01 5.32472849e-01 8.82181883e-01
-4.58780110e-01 -4.57989246e-01 -9.85283673e-01 -5.79913199e-01
1.60740674e-01 -5.44515610e-01 1.96827233e-01 7.41840899e-01
1.78380325e-01 8.34733486e-01 -1.72195405e-01 1.68634728e-01
5.20812273e-01 2.17681602e-01 3.19750011e-01 -1.48211670e+00
-6.94711685e-01 -3.96342725e-01 3.76725286e-01 -7.56179750e-01
4.34358001e-01 -8.77013624e-01 3.95604163e-01 -1.47182512e+00
-4.23784629e-02 -4.18121517e-01 -4.88757819e-01 9.30007637e-01
-3.70937407e-01 -3.66050959e-01 2.44887650e-01 2.31697127e-01
-8.80110264e-01 3.94495204e-02 1.01661074e+00 2.20210388e-01
-1.17666528e-01 -9.56753120e-02 -6.57921791e-01 9.76907730e-01
6.88581824e-01 -6.69403255e-01 -3.57877016e-01 -9.45086837e-01
4.51880217e-01 4.13395405e-01 -1.46278784e-01 -5.55625021e-01
3.03967595e-01 -2.97731668e-01 1.08602345e-01 -1.07239276e-01
1.44046009e-01 -6.59676850e-01 -3.36052328e-01 3.77194613e-01
-4.69566196e-01 5.36875963e-01 6.34909943e-02 4.29535478e-01
6.03648834e-03 -4.90106195e-01 5.37499309e-01 -6.42297268e-01
-8.54116917e-01 -2.82210894e-02 -7.69484282e-01 2.14467227e-01
8.35498035e-01 1.67164356e-01 -1.53397042e-02 2.82048792e-01
-1.27529776e+00 4.27821308e-01 1.45362154e-01 3.02260429e-01
1.12663776e-01 -6.88617587e-01 -6.42905116e-01 1.68136600e-02
-3.73992100e-02 2.22948834e-01 -3.05504501e-01 7.96002448e-01
-8.82991925e-02 3.33025247e-01 -7.54788518e-02 -3.04811448e-01
-1.23853171e+00 5.33633173e-01 1.50734643e-02 -9.15201247e-01
-7.39424884e-01 8.24314356e-01 1.29962772e-01 -3.71934921e-01
1.42983869e-01 -6.27300680e-01 -3.69703799e-01 3.02018493e-01
3.89059186e-01 -1.40596703e-02 2.15973601e-01 -1.87064633e-01
-5.02099097e-01 1.39149159e-01 -3.51761699e-01 -4.06015903e-01
1.39090252e+00 -1.01970911e-01 -4.12661225e-01 6.24783814e-01
3.97959352e-01 1.11546293e-01 -1.36235344e+00 -2.84547418e-01
1.00716770e+00 1.77847724e-02 -2.95132726e-01 -1.01067102e+00
-5.68324804e-01 9.66862619e-01 -1.40680626e-01 4.05917168e-01
8.09922934e-01 5.34294099e-02 5.46230912e-01 6.16689444e-01
3.92054021e-01 -1.33654404e+00 -2.80611720e-02 9.91505921e-01
3.36577803e-01 -1.08539677e+00 -3.50876361e-01 -3.51123691e-01
-8.01216602e-01 1.11826158e+00 6.35349810e-01 -1.77878633e-01
4.45690960e-01 8.20743501e-01 2.02431567e-02 -1.40303716e-01
-1.28895676e+00 -5.04690409e-01 -1.34138957e-01 3.83306891e-01
7.93506980e-01 7.52163455e-02 -4.36691016e-01 7.00291932e-01
1.34787381e-01 -2.92927057e-01 3.55196625e-01 1.19281375e+00
-7.46440530e-01 -1.73210728e+00 -4.30836268e-02 1.71127230e-01
-8.04643393e-01 -3.71504754e-01 -3.35961848e-01 7.18033254e-01
1.15002729e-01 8.73395026e-01 5.23300394e-02 3.52150798e-02
3.61031294e-01 8.04892838e-01 4.34746444e-01 -1.11759818e+00
-8.57097685e-01 1.98322460e-01 6.10170305e-01 -7.28119850e-01
-3.51648599e-01 -6.89461291e-01 -1.47517931e+00 2.38584653e-01
-8.30558464e-02 4.68930274e-01 8.02434862e-01 1.42931736e+00
1.76403448e-01 7.39666879e-01 2.69861162e-01 -7.33077645e-01
-5.87382019e-01 -9.94010448e-01 -2.08950460e-01 3.13950211e-01
2.29790330e-01 -6.13261580e-01 -7.15879723e-02 8.57114345e-02] | [10.354477882385254, 9.538748741149902] |
7b2e06ab-45f9-4dda-8e0e-a35304ffaada | compressive-change-retrieval-for-moving | 1608.02051 | null | http://arxiv.org/abs/1608.02051v1 | http://arxiv.org/pdf/1608.02051v1.pdf | Compressive Change Retrieval for Moving Object Detection | Change detection, or anomaly detection, from street-view images acquired by
an autonomous robot at multiple different times, is a major problem in robotic
mapping and autonomous driving. Formulation as an image comparison task, which
operates on a given pair of query and reference images is common to many
existing approaches to this problem. Unfortunately, providing relevant
reference images is not straightforward. In this paper, we propose a novel
formulation for change detection, termed compressive change retrieval, which
can operate on a query image and similar reference images retrieved from the
web. Compared to previous formulations, there are two sources of difficulty.
First, the retrieved reference images may frequently contain non-relevant
reference images, because even state-of-the-art place-recognition techniques
suffer from retrieval noise. Second, image comparison needs to be conducted in
a compressed domain to minimize the storage cost of large collections of
street-view images. To address the above issues, we also present a practical
change detection algorithm that uses compressed bag-of-words (BoW) image
representation as a scalable solution. The results of experiments conducted on
a practical change detection task, "moving object detection (MOD)," using the
publicly available Malaga dataset validate the effectiveness of the proposed
approach. | ['Tomoya Murase', 'Kanji Tanaka'] | 2016-08-06 | null | null | null | null | ['moving-object-detection'] | ['computer-vision'] | [ 5.82168937e-01 -8.32727849e-01 2.61974260e-02 -3.91582698e-01
-8.76413226e-01 -4.73581135e-01 7.72471368e-01 2.57516146e-01
-4.61020917e-01 4.73130375e-01 -2.63454348e-01 2.57925875e-02
-2.37256095e-01 -7.14602113e-01 -8.58736455e-01 -6.73837364e-01
1.76226512e-01 2.30951086e-02 6.32271767e-01 -4.19557452e-01
7.58578181e-01 6.25364602e-01 -2.28417516e+00 -2.01671451e-01
6.39281571e-01 1.08242464e+00 8.26444089e-01 6.27793908e-01
-4.23579663e-02 4.61642563e-01 -3.29816729e-01 -8.42392668e-02
5.58249831e-01 -1.73380777e-01 -4.41568941e-01 2.60293603e-01
9.41397786e-01 -5.74519217e-01 -4.13942307e-01 1.49824631e+00
3.46837759e-01 3.39238107e-01 4.59735066e-01 -1.50054014e+00
-5.67940414e-01 -2.75716811e-01 -7.01168537e-01 8.84576559e-01
3.20014834e-01 -1.94437757e-01 6.19394720e-01 -1.06023371e+00
8.34258437e-01 1.13117671e+00 4.44051802e-01 3.90170664e-02
-7.86792696e-01 -6.26564026e-01 1.98047385e-01 8.62652123e-01
-1.48604786e+00 -6.47062957e-01 9.65976357e-01 -4.16004777e-01
6.53500617e-01 3.50548029e-01 5.90493977e-01 4.60239887e-01
5.54122686e-01 5.31179965e-01 8.81432056e-01 -4.30708826e-01
2.31414706e-01 9.20082256e-02 6.88910112e-02 4.90744680e-01
6.31371617e-01 -3.64975072e-02 -5.54175198e-01 -1.31097004e-01
1.68164611e-01 3.53874445e-01 -3.41030955e-01 -7.98759878e-01
-1.11741984e+00 7.01580882e-01 2.33810022e-01 6.54432252e-02
-5.52702963e-01 9.74223986e-02 3.09169233e-01 4.64077532e-01
4.48998183e-01 -9.71761197e-02 8.79939739e-03 4.24939878e-02
-8.98216546e-01 4.02213484e-01 3.10124755e-01 1.22001004e+00
1.04922950e+00 -1.95955113e-01 2.80794621e-01 8.28374207e-01
4.19549972e-01 1.03310823e+00 7.89469779e-01 -7.56389439e-01
5.44016719e-01 3.29094440e-01 3.18641990e-01 -1.74120843e+00
7.44616613e-02 -5.97054847e-02 -7.24493980e-01 2.09102169e-01
-8.84807706e-02 4.74152714e-01 -7.76284575e-01 1.27643490e+00
6.81489170e-01 1.59184322e-01 3.35235298e-01 8.90335143e-01
5.67896545e-01 6.53779685e-01 -5.52394509e-01 -1.90677553e-01
1.21857929e+00 -1.00622225e+00 -9.81027246e-01 -5.67571223e-01
2.16199785e-01 -8.70849550e-01 6.18193686e-01 1.81626141e-01
-7.90125370e-01 -7.21576095e-01 -1.42476976e+00 5.96238673e-02
-8.01420867e-01 -1.30528852e-01 5.51612489e-02 2.82646090e-01
-1.07814944e+00 4.89296652e-02 -5.23813069e-01 -8.88903618e-01
-2.66950857e-02 1.74091477e-02 -5.76925635e-01 -6.92813933e-01
-1.05672348e+00 9.90296304e-01 1.79126799e-01 1.07349485e-01
-8.22998941e-01 -2.51323760e-01 -9.40215588e-01 -2.16067061e-01
4.31954801e-01 -1.65307119e-01 1.16033363e+00 -8.77902865e-01
-9.57501292e-01 8.72978747e-01 -6.17878973e-01 -5.64817250e-01
3.58122528e-01 -1.55434713e-01 -8.71381283e-01 3.96296144e-01
4.67215061e-01 6.26939178e-01 1.18895221e+00 -1.37571859e+00
-1.16375709e+00 -5.41930795e-01 -2.45150626e-01 5.10518312e-01
2.03374282e-01 -1.96360171e-01 -6.32504106e-01 -4.16607022e-01
7.85250187e-01 -9.81680930e-01 2.35770233e-02 2.00801224e-01
1.57638088e-01 2.08900012e-02 1.39088261e+00 -7.09774315e-01
9.62306201e-01 -2.42109036e+00 -1.55590951e-01 2.22811759e-01
-2.14880198e-01 7.46794119e-02 -2.19556212e-01 5.63494384e-01
1.16620056e-01 -3.43491584e-01 -4.08504844e-01 -1.48602081e-02
-2.44928539e-01 2.24569723e-01 -4.56763476e-01 8.03524911e-01
7.41167814e-02 4.34605688e-01 -9.14281607e-01 -4.74343598e-01
4.69859481e-01 6.82817847e-02 -1.63681522e-01 -5.20821894e-03
2.39874929e-01 2.84427524e-01 -4.06163037e-01 7.57442713e-01
1.00712359e+00 1.58866286e-01 -3.06965619e-01 -7.66263828e-02
-3.26817989e-01 -1.42659560e-01 -1.48000658e+00 1.68171680e+00
-4.29404788e-02 1.01539111e+00 9.06587467e-02 -1.20596492e+00
1.06023598e+00 -9.38101858e-02 5.16327441e-01 -1.09447885e+00
-2.46050477e-01 3.28201473e-01 -2.97235996e-01 -7.98000455e-01
1.13412452e+00 3.51572156e-01 -1.53424088e-02 1.20630577e-01
-6.42879725e-01 -4.51670438e-01 4.07545328e-01 1.11572206e-01
1.09672475e+00 -9.65637863e-02 4.55366373e-01 -8.63661915e-02
8.25454354e-01 5.07625759e-01 4.92133111e-01 8.82128298e-01
-6.49715662e-01 5.90437233e-01 -3.17176372e-01 -3.83009017e-01
-9.72801507e-01 -8.65230799e-01 -2.59874821e-01 6.56200707e-01
1.05445135e+00 -7.63085186e-02 -3.22447628e-01 -2.29218200e-01
2.35440016e-01 3.92298430e-01 -2.99445808e-01 -6.12793900e-02
-4.31541562e-01 -4.00131583e-01 1.21178754e-01 -5.65621220e-02
9.85660791e-01 -6.68901563e-01 -1.08201981e+00 2.25127563e-01
-4.65973079e-01 -1.18715656e+00 -4.19260204e-01 -1.49928451e-01
-7.02601016e-01 -1.22601116e+00 -5.39105535e-01 -9.33087349e-01
7.28370965e-01 1.46741021e+00 7.06856787e-01 -1.13866337e-01
-3.80663544e-01 7.72855937e-01 -4.81650561e-01 -4.06103373e-01
-1.64529130e-01 -4.62285161e-01 1.85488015e-01 2.33875766e-01
5.59792101e-01 -1.61193222e-01 -5.99043369e-01 4.08378601e-01
-1.30511081e+00 -1.70811027e-01 5.95292628e-01 7.56376922e-01
9.43390608e-01 3.57824177e-01 3.82352948e-01 -3.72576088e-01
2.98142225e-01 -6.32136166e-01 -8.91646862e-01 2.80049056e-01
-7.68586636e-01 -1.75794899e-01 1.14506491e-01 -1.39911965e-01
-8.54100943e-01 1.82376817e-01 3.54835749e-01 -4.89076883e-01
-1.99699789e-01 5.22936940e-01 3.61361541e-02 -2.56695122e-01
3.66115898e-01 7.24735796e-01 3.29249233e-01 -1.77254274e-01
3.42561126e-01 1.04023004e+00 8.13500285e-01 1.35283545e-01
9.13997233e-01 8.58965755e-01 -4.03660499e-02 -1.10988271e+00
-2.24996954e-01 -1.16833937e+00 -6.52215958e-01 -3.16065133e-01
5.46702504e-01 -1.30327129e+00 -4.41464968e-03 5.38045228e-01
-9.96042490e-01 2.84091860e-01 -5.15668765e-02 4.97382879e-01
-5.75360298e-01 6.24318957e-01 1.49126634e-01 -7.29269445e-01
-1.99198991e-01 -1.10585856e+00 1.09410131e+00 2.79111356e-01
2.83247232e-01 -4.57922637e-01 1.44226596e-01 3.25996131e-01
5.88564694e-01 2.18629882e-01 4.55319226e-01 -2.55880803e-01
-9.38819468e-01 -4.23634231e-01 -2.35892281e-01 1.15271203e-01
5.24705708e-01 -2.01419279e-01 -6.45714283e-01 -5.61016142e-01
1.87234372e-01 1.80199295e-01 7.31050730e-01 2.05335602e-01
7.62760341e-01 -1.19606853e-01 -5.07418752e-01 3.32367867e-01
1.75645411e+00 6.25639915e-01 5.98813295e-01 7.71459341e-01
3.59886080e-01 4.10104305e-01 1.29564893e+00 4.96110111e-01
4.91164416e-01 7.25127459e-01 5.47033072e-01 2.72353262e-01
-6.19279891e-02 -5.76779358e-02 3.57127607e-01 7.31517732e-01
6.64484680e-01 -2.52047837e-01 -7.75652766e-01 9.22531366e-01
-1.91534448e+00 -1.13039279e+00 -1.72460213e-01 2.30175972e+00
2.56506473e-01 -1.03400312e-01 -3.99375260e-01 2.46756583e-01
1.00555861e+00 3.60948533e-01 -9.36293840e-01 -4.80122641e-02
-1.09938271e-01 -5.04445195e-01 9.02495742e-01 3.17595601e-01
-1.12150490e+00 6.95078909e-01 5.94192314e+00 5.69186389e-01
-1.35773528e+00 2.76327670e-01 9.95147526e-02 6.18228950e-02
-5.71488179e-02 -7.01460168e-02 -8.48220587e-01 5.72392046e-01
6.83166087e-01 -2.83536971e-01 4.35802609e-01 8.23617220e-01
2.68161505e-01 -8.29063773e-01 -7.84275591e-01 1.46837711e+00
8.19399178e-01 -1.02852809e+00 3.47185843e-02 -1.75442606e-01
8.31060410e-01 4.14075315e-01 1.33404545e-02 2.89056934e-02
-2.05650076e-01 -3.24701399e-01 8.84642541e-01 6.28078878e-01
5.87143838e-01 -6.09226048e-01 6.61883295e-01 2.57271767e-01
-1.29227746e+00 -2.82378849e-02 -6.46768332e-01 2.25133356e-02
9.80020091e-02 4.99146730e-01 -5.55760145e-01 8.11503232e-01
1.13894176e+00 1.02691638e+00 -8.76572549e-01 1.24696434e+00
3.50564748e-01 -6.46666586e-02 -4.37689722e-01 1.17935091e-01
2.76269436e-01 -2.46367574e-01 6.85631752e-01 7.99454868e-01
7.29861259e-01 3.64948879e-03 5.80288582e-02 4.50318545e-01
1.94924012e-01 7.26428181e-02 -1.24412107e+00 3.04203510e-01
6.44343913e-01 1.05116189e+00 -5.69078386e-01 -4.39534962e-01
-6.34948909e-01 1.17445636e+00 -1.34965345e-01 3.31186712e-01
-5.73265493e-01 -6.55753374e-01 6.27457023e-01 -5.04496582e-02
6.49681568e-01 -3.88481021e-01 3.05170506e-01 -9.23574746e-01
4.27672029e-01 -7.43047953e-01 1.32161930e-01 -1.03792131e+00
-9.90991294e-01 2.66661972e-01 2.85613894e-01 -1.78397870e+00
-1.34028956e-01 -1.52537555e-01 -2.53241390e-01 6.86331809e-01
-2.15048623e+00 -7.01388776e-01 -9.07259464e-01 6.61731482e-01
9.54519749e-01 -1.91842213e-01 2.88659096e-01 4.40085799e-01
-3.43957096e-01 1.61141753e-01 8.78690541e-01 -3.51559967e-01
8.49601805e-01 -7.71292031e-01 4.12770092e-01 1.36102223e+00
4.31388505e-02 4.52829786e-02 7.38700032e-01 -7.90448964e-01
-1.71453118e+00 -1.33368039e+00 6.82712734e-01 -4.98651788e-02
3.35832506e-01 1.69604719e-01 -8.76712680e-01 5.18181860e-01
1.45140171e-01 9.98611748e-02 3.25839251e-01 -7.12813973e-01
-3.93095380e-03 -4.43996996e-01 -1.28014827e+00 4.14982915e-01
8.93049121e-01 -5.63853204e-01 -7.52254426e-01 2.09301680e-01
4.19041842e-01 -4.74024296e-01 -3.86238456e-01 3.00126374e-01
4.60408330e-01 -8.60885441e-01 8.71597528e-01 1.33544087e-01
2.37314794e-02 -8.34317207e-01 -7.66278744e-01 -1.20144582e+00
-1.88149124e-01 -1.46113962e-01 1.23052299e-01 1.11751199e+00
-3.26752126e-01 -7.93994427e-01 3.39023918e-01 3.61824512e-01
-9.07656178e-02 -1.56333089e-01 -1.07330465e+00 -9.16118503e-01
-6.01397276e-01 -3.41628700e-01 5.81005216e-01 7.77209401e-01
-4.32318300e-01 1.09511845e-01 -2.57063657e-01 7.12602437e-01
6.18050992e-01 2.32135132e-01 1.05722487e+00 -1.14706588e+00
2.83414930e-01 -7.93838874e-02 -8.67505550e-01 -9.93691981e-01
-1.00844361e-01 -7.50562489e-01 5.75390279e-01 -1.65863204e+00
1.20885961e-01 -4.70304072e-01 -1.76252633e-01 -4.40220647e-02
7.23967794e-03 2.31924996e-01 8.17400366e-02 7.15330660e-01
-6.23394251e-01 6.46283269e-01 8.27506721e-01 -4.98498946e-01
5.17714880e-02 -3.63923967e-01 -2.32580110e-01 4.64629322e-01
6.41229272e-01 -5.47741175e-01 -5.27896106e-01 -5.40745676e-01
-7.74949640e-02 -4.95471582e-02 3.63812089e-01 -1.30028105e+00
5.69628537e-01 -2.93410659e-01 9.74396840e-02 -1.45340908e+00
4.38091606e-01 -1.10642576e+00 4.37490791e-01 6.05028689e-01
6.59721568e-02 7.16731846e-01 1.45347074e-01 1.12082350e+00
-5.87033272e-01 -5.09430766e-01 7.53917933e-01 -1.18000135e-01
-1.54804969e+00 1.32385716e-01 -5.05725861e-01 -2.84844548e-01
1.39947259e+00 -4.21161681e-01 -3.96683633e-01 -3.03335041e-01
2.61665806e-02 3.13891083e-01 6.38952494e-01 7.88557947e-01
9.44742441e-01 -1.47159588e+00 -5.83232582e-01 5.70045352e-01
7.46835947e-01 1.07809767e-01 3.11680824e-01 7.17559695e-01
-8.94680381e-01 4.79950011e-01 -2.83295125e-01 -9.60877120e-01
-1.45633149e+00 7.82271862e-01 5.10411598e-02 3.42023134e-01
-7.43815601e-01 3.52898389e-01 -9.54037160e-02 -1.50580630e-01
1.45831406e-01 -2.66195536e-01 -1.97877437e-01 1.14002377e-01
6.93982601e-01 3.82530630e-01 3.02444011e-01 -1.06317401e+00
-5.03313959e-01 9.11082566e-01 -1.65500924e-01 -1.95785016e-01
9.66872692e-01 -9.50275004e-01 -3.87947373e-02 5.90325832e-01
1.18373692e+00 -1.65215671e-01 -1.02611375e+00 -6.52742684e-01
2.55792774e-02 -8.82883191e-01 1.84729502e-01 -2.03396186e-01
-8.34673107e-01 5.44611752e-01 1.40188360e+00 1.97253555e-01
1.05677259e+00 -3.02596122e-01 7.41805971e-01 8.86761546e-01
5.85400164e-01 -1.37034512e+00 -6.14745393e-02 4.60498005e-01
9.88557398e-01 -1.75201488e+00 5.49155921e-02 -2.80547857e-01
-3.20642918e-01 8.35018098e-01 5.06853700e-01 -1.27039567e-01
6.62383378e-01 -3.33981365e-01 1.26274124e-01 -1.12057231e-01
-5.17547727e-01 -4.17532891e-01 4.29888442e-03 4.14486885e-01
-5.04037201e-01 -1.95662141e-01 -3.36609006e-01 -3.12156975e-01
4.72136475e-02 -1.17368609e-01 7.88835526e-01 1.52333057e+00
-7.42908418e-01 -7.20793188e-01 -9.03211653e-01 4.88913387e-01
-1.59254968e-01 4.79067229e-02 -7.04814587e-03 6.80300951e-01
8.92339572e-02 1.27288306e+00 4.25238401e-01 -2.86506921e-01
4.41977084e-01 -4.84828278e-02 1.52452946e-01 -3.31123501e-01
2.03725100e-01 -2.33190373e-01 -3.94637704e-01 -5.93898714e-01
-7.81382859e-01 -9.71493363e-01 -9.38717842e-01 -2.08052814e-01
-2.85898715e-01 -4.29915413e-02 1.10427952e+00 8.35884094e-01
4.98131990e-01 1.97190687e-01 8.48731399e-01 -8.95804703e-01
-4.51028079e-01 -8.55550051e-01 -7.72590280e-01 5.85924208e-01
7.30639994e-01 -8.95329058e-01 -4.86517131e-01 2.59984702e-01] | [7.690834999084473, -1.8791528940200806] |
d2a609b1-13ce-4772-bb33-d7f4940fab91 | combining-multi-level-contexts-of-superpixel | 1803.05200 | null | http://arxiv.org/abs/1803.05200v1 | http://arxiv.org/pdf/1803.05200v1.pdf | Combining Multi-level Contexts of Superpixel using Convolutional Neural Networks to perform Natural Scene Labeling | Modern deep learning algorithms have triggered various image segmentation
approaches. However most of them deal with pixel based segmentation. However,
superpixels provide a certain degree of contextual information while reducing
computation cost. In our approach, we have performed superpixel level semantic
segmentation considering 3 various levels as neighbours for semantic contexts.
Furthermore, we have enlisted a number of ensemble approaches like max-voting
and weighted-average. We have also used the Dempster-Shafer theory of
uncertainty to analyze confusion among various classes. Our method has proved
to be superior to a number of different modern approaches on the same dataset. | ['Swarnendu Ghosh', 'Sandipan Choudhuri', 'Nibaran Das', 'Mita Nasipuri', 'Ritesh Sarkhel', 'Aritra Das'] | 2018-03-14 | null | null | null | null | ['scene-labeling'] | ['computer-vision'] | [ 2.42114693e-01 2.10100368e-01 -9.09593776e-02 -6.49738491e-01
-8.97810161e-01 -3.91902477e-01 6.40953064e-01 3.07189256e-01
-7.08189666e-01 1.08811033e+00 -2.18358964e-01 2.13307049e-02
-2.75253862e-01 -9.39607084e-01 -5.22220790e-01 -8.76566052e-01
1.90098822e-01 5.47608912e-01 6.95400536e-01 1.87708214e-01
5.27506471e-01 4.58687186e-01 -1.72691548e+00 1.98608086e-01
1.09039903e+00 1.15265810e+00 3.07826642e-02 3.58793914e-01
-1.79665521e-01 4.02557194e-01 -6.81116998e-01 -3.70872915e-01
2.01265305e-01 -8.21270049e-02 -1.01232743e+00 3.19537938e-01
5.49199939e-01 5.51347770e-02 1.20531693e-01 1.43032861e+00
2.95259297e-01 2.91748136e-01 9.45967138e-01 -9.88611937e-01
-2.45290965e-01 7.88687408e-01 -6.58506036e-01 1.68280512e-01
1.58682428e-02 -2.74668366e-01 9.90095317e-01 -6.52621388e-01
6.65553212e-01 1.11760342e+00 5.91355145e-01 1.11190282e-01
-1.23479295e+00 -2.39471242e-01 -3.97268496e-03 4.91948843e-01
-1.36752808e+00 1.07563585e-01 7.02278912e-01 -3.15301210e-01
6.37890399e-01 6.79389313e-02 6.59006417e-01 6.39420331e-01
8.13135430e-02 9.52830672e-01 1.64997768e+00 -5.22580087e-01
5.12557507e-01 1.04034416e-01 4.54463720e-01 4.17208910e-01
4.37712699e-01 -2.39524439e-01 -1.73654214e-01 6.37196526e-02
5.71474910e-01 -2.28693008e-01 -2.24755593e-02 -1.47025734e-01
-8.75196695e-01 8.42015147e-01 5.16448259e-01 7.29768336e-01
-4.49306875e-01 2.69477367e-01 1.80565327e-01 -4.32340354e-01
6.40808105e-01 1.43712655e-01 -3.95296365e-01 3.97118866e-01
-1.35508502e+00 7.76862726e-02 6.99567080e-01 5.74555755e-01
1.01528096e+00 -1.88023269e-01 -5.00480607e-02 6.59501612e-01
3.97341639e-01 1.96834743e-01 1.48538336e-01 -1.38394487e+00
3.74403819e-02 5.10699630e-01 2.90374085e-03 -1.00518084e+00
-3.49583745e-01 -4.84354794e-01 -7.39645302e-01 5.82704008e-01
5.31249762e-01 -9.78300720e-02 -1.43098700e+00 1.30084598e+00
2.45183736e-01 2.36172915e-01 -8.67410004e-02 6.91536844e-01
5.81499040e-01 4.34611708e-01 3.56853783e-01 -5.80726750e-02
9.59687769e-01 -6.91556334e-01 -6.55123115e-01 3.06142598e-01
1.28972799e-01 -7.72292316e-01 3.15440297e-01 8.45813334e-01
-8.84386301e-01 -6.58749461e-01 -1.06391537e+00 2.92022467e-01
-7.81085491e-01 -1.37778640e-01 6.20219707e-01 1.09406614e+00
-1.32428849e+00 9.63232338e-01 -6.76886797e-01 -2.70117253e-01
7.31567979e-01 4.34580475e-01 -2.01268747e-01 2.36702159e-01
-1.11653614e+00 8.29726934e-01 1.10151720e+00 6.61442056e-02
-6.21718228e-01 -1.42673984e-01 -5.04796267e-01 -1.78961173e-01
3.79687399e-01 -5.52219689e-01 8.83177161e-01 -1.13189697e+00
-1.20638263e+00 1.10624075e+00 1.14285927e-02 -7.00575531e-01
6.48655891e-01 -2.16574997e-01 -7.33903646e-02 4.23032045e-01
-6.13755779e-03 1.00907779e+00 5.25045037e-01 -1.55868649e+00
-8.27326119e-01 -5.15304327e-01 1.15224905e-01 1.15568750e-01
1.09191716e-01 -2.87094623e-01 -2.40033120e-01 -5.13412833e-01
6.99982285e-01 -7.82446086e-01 -5.54798186e-01 -4.05180097e-01
-5.36636233e-01 -4.95441735e-01 7.48617351e-01 -3.77141356e-01
9.15637016e-01 -1.66557205e+00 1.07031450e-01 3.86329442e-01
7.55561665e-02 2.96904236e-01 3.81785035e-01 1.23905823e-01
1.60733551e-01 3.89612556e-01 -6.35020792e-01 -1.84142217e-01
-1.32341295e-01 4.02966797e-01 1.83690250e-01 4.00652915e-01
-5.04702739e-02 6.15347087e-01 -7.49137282e-01 -1.08202982e+00
7.94950962e-01 5.78254104e-01 -1.35523811e-01 -3.98969144e-01
-3.66592288e-01 1.76481470e-01 -6.36495233e-01 6.91564500e-01
9.30578649e-01 -2.61391583e-03 -1.53544873e-01 -1.41200140e-01
-1.43320272e-02 -2.79753834e-01 -1.37993085e+00 1.89590371e+00
3.22501995e-02 5.55468738e-01 -1.53033495e-01 -1.14938676e+00
6.80273592e-01 2.02564582e-01 7.54587531e-01 -2.03260481e-01
4.35705483e-01 2.84509182e-01 -2.90845275e-01 -2.39639342e-01
7.49663293e-01 -1.80684645e-02 1.52874306e-01 3.93350013e-02
2.73084372e-01 -5.17561376e-01 4.01769847e-01 1.92125160e-02
6.83357418e-01 5.36107123e-01 2.44417563e-01 -6.41790330e-01
6.11790776e-01 3.52805257e-01 4.46779042e-01 9.90614593e-01
-6.18707120e-01 8.98830235e-01 2.59995222e-01 -2.34126762e-01
-8.92985225e-01 -1.22510886e+00 -5.40490806e-01 5.08079469e-01
4.53161865e-01 2.28182718e-01 -1.31859517e+00 -4.93325561e-01
-1.94949836e-01 8.64467621e-01 -5.40244162e-01 4.89311635e-01
-4.09199327e-01 -1.12783909e+00 5.20126283e-01 3.17119479e-01
8.82395387e-01 -9.68883216e-01 -7.43218184e-01 4.05190378e-01
-6.09473046e-03 -1.13308990e+00 2.18174353e-01 2.30674177e-01
-1.17634404e+00 -1.24858594e+00 -8.16737175e-01 -5.21156788e-01
4.80234981e-01 5.57970628e-03 1.05496442e+00 -1.78695530e-01
-4.76580948e-01 3.66708994e-01 -4.96312827e-01 -2.59615630e-01
-2.23317251e-01 1.84670351e-02 -3.48785907e-01 -4.31098565e-02
4.19759154e-01 -5.04662097e-01 -7.28354096e-01 9.84852910e-02
-9.64759052e-01 -3.08224201e-01 4.90267992e-01 4.58189666e-01
8.96094739e-01 5.18903911e-01 2.75008291e-01 -1.08640802e+00
3.78614902e-01 -1.54489771e-01 -6.11932695e-01 3.25813413e-01
-7.64225602e-01 9.74953547e-02 1.23803638e-01 7.02676922e-02
-1.37572420e+00 9.00226831e-02 -1.39057457e-01 -1.11801207e-01
-6.24988973e-01 1.97849274e-01 -3.46891358e-02 -1.07545868e-01
5.02370596e-01 -1.58830211e-01 -5.64078510e-01 -2.14709997e-01
4.87433672e-01 7.31294453e-01 3.71323884e-01 -5.22130132e-01
2.08167881e-01 7.30769932e-01 2.04243705e-01 -7.22775757e-01
-8.60361457e-01 -4.75922078e-01 -9.76445794e-01 -5.35332143e-01
1.41352880e+00 -5.55384398e-01 -3.36322606e-01 8.73842120e-01
-1.09768927e+00 3.92470658e-02 -1.36278167e-01 3.80631745e-01
-5.37308633e-01 5.59221864e-01 -4.87776577e-01 -9.86966252e-01
-1.21698365e-01 -1.35465598e+00 1.03005600e+00 7.51639426e-01
1.00983351e-01 -1.03328168e+00 -1.10464536e-01 5.86645722e-01
1.46723926e-01 7.10981727e-01 4.55222726e-01 -6.13952398e-01
-7.34090090e-01 -2.63420761e-01 -4.11588252e-01 5.93673944e-01
-2.40560323e-02 2.30058953e-01 -1.24799180e+00 2.81458139e-01
-9.10622478e-02 5.02067469e-02 1.31610584e+00 8.18412006e-01
1.22998798e+00 4.74232405e-01 -3.38075340e-01 2.25086018e-01
1.81747806e+00 3.85763824e-01 7.06185877e-01 3.72451007e-01
5.83007574e-01 5.91099024e-01 7.02746153e-01 1.48158461e-01
1.52413726e-01 2.74976820e-01 5.32265186e-01 1.11255221e-01
-2.05387566e-02 2.02103063e-01 -2.70037711e-01 2.12048158e-01
-4.49993819e-01 -4.21390533e-01 -1.02141058e+00 6.23159468e-01
-1.94391370e+00 -9.48951602e-01 -5.40739119e-01 1.86088407e+00
4.93684232e-01 4.80826467e-01 -1.53835431e-01 3.06333691e-01
9.98950779e-01 2.81154007e-01 -3.41256410e-01 -2.56563544e-01
-4.10366595e-01 4.18429852e-01 8.88318658e-01 5.78544438e-01
-1.46646261e+00 1.12944257e+00 6.69927120e+00 1.22723246e+00
-6.60843730e-01 2.27257460e-01 9.47045743e-01 4.28141356e-01
-2.24041834e-01 6.61031604e-02 -5.88413179e-01 3.37460876e-01
5.74108183e-01 9.04326066e-02 -4.36668657e-02 7.64277875e-01
-1.97741866e-01 -1.12526965e+00 -5.60475647e-01 7.09305584e-01
2.82409322e-02 -1.04405355e+00 -1.08350784e-01 1.20612048e-03
1.26632035e+00 1.15186587e-01 -1.21412888e-01 -3.22149068e-01
4.36904132e-01 -1.02727807e+00 5.36170244e-01 7.44238436e-01
1.32272601e-01 -8.74414027e-01 8.81060004e-01 1.98209479e-01
-8.48462760e-01 1.63999185e-01 -3.42741042e-01 1.76972479e-01
2.21425891e-01 1.10171413e+00 -4.72971171e-01 7.83304691e-01
6.82848513e-01 2.89050609e-01 -4.88813102e-01 1.13938034e+00
-3.25968266e-01 3.75404358e-01 -5.85766554e-01 1.61730051e-02
6.49626911e-01 -5.41341245e-01 4.36897725e-01 1.03401136e+00
1.39484435e-01 5.66243269e-02 6.08506729e-04 7.66664207e-01
2.32001856e-01 7.65904710e-02 -2.69179225e-01 1.33713588e-01
3.33807647e-01 1.18875802e+00 -1.60896134e+00 -5.71949422e-01
-1.21177159e-01 8.53860199e-01 -1.99180931e-01 1.93431571e-01
-5.68922222e-01 -1.72897935e-01 3.85629445e-01 -2.50761032e-01
2.05902696e-01 -2.61824131e-01 -7.71508396e-01 -8.37952137e-01
-3.92654419e-01 -1.51477665e-01 4.38018709e-01 -8.78834724e-01
-9.70375776e-01 6.26840293e-01 4.55257803e-01 -8.73472989e-01
-1.33442320e-02 -9.22389269e-01 -3.65123957e-01 6.84319556e-01
-1.25661635e+00 -1.02241969e+00 -5.86147308e-02 4.12881732e-01
6.89661920e-01 3.82343978e-02 4.91149783e-01 1.19207427e-01
-4.10882175e-01 -2.37035200e-01 1.51970163e-01 -8.87963846e-02
3.61448228e-01 -1.51977432e+00 -2.13221878e-01 9.76468921e-01
3.22741270e-01 2.24993721e-01 9.74713564e-01 -5.68773985e-01
-4.13735807e-01 -7.55047321e-01 5.42641580e-01 -3.17939103e-01
5.53084314e-01 2.84631044e-01 -7.80357122e-01 4.43823874e-01
5.82604825e-01 -1.55684844e-01 4.05385345e-01 -1.97930597e-02
1.05704732e-01 -2.99015194e-02 -1.42373455e+00 3.74227732e-01
7.47149587e-01 -3.08054566e-01 -7.72187889e-01 1.74976379e-01
3.37877870e-01 -2.76315510e-01 -8.63307893e-01 7.22990215e-01
3.58647704e-01 -1.57458150e+00 9.87791240e-01 1.90737881e-02
3.99709433e-01 -4.32386935e-01 -2.69085437e-01 -1.10462737e+00
2.69470103e-02 2.54654995e-04 5.87805152e-01 1.22251403e+00
3.46221060e-01 -5.90975106e-01 1.01877964e+00 6.33468986e-01
-1.33014008e-01 -5.03350854e-01 -1.06189787e+00 -5.23960829e-01
2.70242304e-01 -7.91694164e-01 3.26168031e-01 7.32239544e-01
-4.16581541e-01 -2.31312960e-01 5.82313910e-02 5.25072739e-02
1.21926129e+00 2.42409199e-01 9.47313383e-02 -1.39480639e+00
7.58659616e-02 -7.84701884e-01 -6.86520100e-01 -2.15235382e-01
1.27675459e-01 -7.42810428e-01 4.52555567e-02 -1.81137431e+00
2.92382479e-01 -3.01489413e-01 -8.85653615e-01 9.33999270e-02
-2.36450508e-01 5.27518988e-01 -4.68639471e-02 5.91406785e-02
-6.61212564e-01 2.34920438e-02 1.12146342e+00 -8.86128321e-02
6.32925555e-02 3.45417410e-02 -3.63739461e-01 1.11645734e+00
1.06467032e+00 -4.84590471e-01 -2.56048232e-01 -2.76615441e-01
7.74714649e-02 -3.00299197e-01 3.52681756e-01 -1.38874662e+00
2.88302302e-01 -1.86547805e-02 5.09253204e-01 -9.52046394e-01
2.65944391e-01 -8.48930001e-01 2.48605832e-01 2.60160416e-01
-1.37761012e-01 -5.23680627e-01 -6.13963306e-02 6.70615375e-01
-4.31058615e-01 -6.37617648e-01 1.02588058e+00 -5.10028660e-01
-1.27458096e+00 2.98221689e-02 -4.41275418e-01 -1.14788584e-01
1.34901524e+00 -5.92768908e-01 1.07666822e-02 1.00607656e-01
-1.10472310e+00 1.02711685e-01 4.95088637e-01 -7.51566738e-02
3.10157150e-01 -8.55603456e-01 -4.10234362e-01 -4.25908953e-01
-2.54282385e-01 1.57504901e-01 3.72045010e-01 6.91649020e-01
-8.92897725e-01 4.80630815e-01 -4.00804043e-01 -9.63463366e-01
-1.29689622e+00 2.66983837e-01 3.60056758e-01 -1.06074169e-01
-2.81849712e-01 1.18724871e+00 -8.32380056e-02 -1.53543338e-01
8.36678296e-02 -1.38309732e-01 -4.05807555e-01 5.10698915e-01
-1.09431341e-01 7.48116016e-01 -1.39294401e-01 -7.88596332e-01
-3.93057466e-01 8.77166867e-01 2.66527593e-01 -4.24586356e-01
1.11590350e+00 -1.95851147e-01 -2.43495911e-01 5.66350222e-01
8.24368775e-01 -3.85732174e-01 -1.23749542e+00 -1.67707890e-01
4.29333776e-01 -2.59273469e-01 5.02162039e-01 -7.86666751e-01
-9.50015068e-01 9.06679332e-01 8.12691271e-01 4.58597332e-01
1.39305031e+00 1.88462045e-02 4.92393464e-01 2.68025637e-01
8.68706107e-01 -1.69718540e+00 -5.19472003e-01 1.45922661e-01
3.37236613e-01 -1.56841671e+00 3.21906477e-01 -5.20907164e-01
-4.83663529e-01 1.20472372e+00 2.53714383e-01 -3.87251586e-01
7.85745859e-01 2.66912967e-01 1.20342746e-01 1.89841837e-02
-2.33234003e-01 -8.82074594e-01 2.79089391e-01 5.07888556e-01
3.32873911e-01 3.46940845e-01 -8.04364860e-01 3.88467275e-02
4.36996892e-02 2.08037272e-01 4.68547434e-01 7.26332366e-01
-8.56787264e-01 -1.14746642e+00 -5.40211439e-01 3.40214849e-01
-7.66877174e-01 8.70963559e-02 -2.74690211e-01 1.01932752e+00
6.72980547e-01 1.10809505e+00 -1.09417550e-01 -2.55186677e-01
-1.06195129e-01 1.50923803e-01 7.80657768e-01 -4.27927732e-01
-4.17150050e-01 2.47227028e-01 -7.71134272e-02 -5.15440345e-01
-1.18592525e+00 -8.75213563e-01 -1.29115045e+00 1.77330449e-01
-4.70287800e-01 1.46192268e-01 9.46431398e-01 1.16585946e+00
-2.95530051e-01 4.23073858e-01 2.24112198e-01 -9.05164242e-01
-1.43665910e-01 -1.01496410e+00 -9.97946441e-01 1.69890687e-01
-1.37171060e-01 -5.96748948e-01 -3.21255237e-01 1.45463079e-01] | [9.508308410644531, 0.23753029108047485] |
6d7fc980-2744-4cb1-ba57-92012f774047 | opd-single-view-3d-openable-part-detection | 2203.16421 | null | https://arxiv.org/abs/2203.16421v1 | https://arxiv.org/pdf/2203.16421v1.pdf | OPD: Single-view 3D Openable Part Detection | We address the task of predicting what parts of an object can open and how they move when they do so. The input is a single image of an object, and as output we detect what parts of the object can open, and the motion parameters describing the articulation of each openable part. To tackle this task, we create two datasets of 3D objects: OPDSynth based on existing synthetic objects, and OPDReal based on RGBD reconstructions of real objects. We then design OPDRCNN, a neural architecture that detects openable parts and predicts their motion parameters. Our experiments show that this is a challenging task especially when considering generalization across object categories, and the limited amount of information in a single image. Our architecture outperforms baselines and prior work especially for RGB image inputs. Short video summary at https://www.youtube.com/watch?v=P85iCaD0rfc | ['Angel X. Chang', 'Manolis Savva', 'Yongsen Mao', 'Hanxiao Jiang'] | 2022-03-30 | null | null | null | null | ['opd-single-view-3d-openable-part-detection'] | ['computer-vision'] | [ 9.71038714e-02 3.45857978e-01 -1.82159021e-01 -3.30160285e-04
-2.70728827e-01 -9.57602620e-01 5.11599541e-01 -5.42812526e-01
-2.02756315e-01 2.95051664e-01 2.48811558e-01 -1.92644462e-01
1.84091344e-01 -6.59032464e-01 -1.14587319e+00 -4.95736510e-01
1.02346979e-01 7.16706157e-01 4.81083155e-01 -4.52622958e-03
1.31330818e-01 6.28944874e-01 -1.54320872e+00 4.04382497e-01
3.99649590e-01 1.32109797e+00 5.38838267e-01 9.10529017e-01
-1.03063755e-01 6.16827726e-01 -2.06229061e-01 -5.82050532e-02
6.29348338e-01 -5.15831374e-02 -1.17906106e+00 2.03826502e-01
5.61984479e-01 -4.80546474e-01 -7.65602767e-01 6.44481122e-01
2.68406779e-01 1.56114027e-01 9.06345427e-01 -1.31677890e+00
-8.74025881e-01 4.07810330e-01 -2.45830491e-01 2.84813851e-01
5.06061494e-01 6.12847507e-01 8.13975215e-01 -9.46580470e-01
1.06056583e+00 1.33631277e+00 3.99158955e-01 8.26126933e-01
-1.14514637e+00 -3.08708549e-01 3.96359354e-01 2.34776005e-01
-1.15736425e+00 -4.36228096e-01 7.22013593e-01 -6.97112858e-01
9.97093379e-01 2.45898470e-01 1.07211745e+00 1.49141955e+00
2.12637335e-02 9.89766359e-01 6.70614481e-01 6.10018298e-02
4.60825115e-01 -4.65101212e-01 -3.27286534e-02 5.16197383e-01
1.47305980e-01 -1.78170919e-01 -2.65578747e-01 2.95104563e-01
1.00824416e+00 3.10069829e-01 -3.29199016e-01 -6.37732089e-01
-1.60292149e+00 2.54006177e-01 8.41815889e-01 -1.27415791e-01
-3.48414123e-01 5.18359661e-01 -8.84866714e-02 -9.69778746e-03
1.29264653e-01 2.42390871e-01 -7.41410613e-01 -2.77459949e-01
-5.29729307e-01 4.03999090e-01 7.34763622e-01 1.30087590e+00
6.39662981e-01 -2.68902034e-01 1.90109909e-01 3.59465390e-01
3.46520960e-01 6.00912154e-01 3.40770245e-01 -1.53261125e+00
5.09880602e-01 8.63909781e-01 2.84647018e-01 -4.85348105e-01
-3.68408054e-01 2.16874629e-01 -3.24310511e-01 3.31133753e-01
5.18245161e-01 4.08700034e-02 -1.22095466e+00 1.39861977e+00
3.74505639e-01 3.77152897e-02 -2.47504516e-03 1.02913678e+00
1.02852285e+00 9.05409157e-01 -2.00474232e-01 2.91689396e-01
1.25248575e+00 -1.21939814e+00 -4.08256978e-01 -4.67226148e-01
4.20740724e-01 -4.05461222e-01 1.07101631e+00 2.57762045e-01
-1.17334533e+00 -6.48435414e-01 -7.27827013e-01 -4.95008826e-01
-5.25181532e-01 9.76592973e-02 4.36358511e-01 -6.36696666e-02
-8.54457259e-01 6.62464559e-01 -1.20494115e+00 -2.73679376e-01
7.48364866e-01 3.27764869e-01 -5.00338316e-01 -1.04839303e-01
-4.95914817e-01 6.33890510e-01 3.22108537e-01 3.49547237e-01
-1.13368261e+00 -4.74311799e-01 -8.34018409e-01 -1.30364046e-01
4.87709969e-01 -8.36878836e-01 1.50258493e+00 -9.69322026e-01
-1.18023384e+00 8.48018467e-01 5.54517545e-02 -1.04344711e-01
6.25047266e-01 -4.49800462e-01 4.88806283e-04 2.88494736e-01
-1.37227356e-01 1.36154175e+00 8.32296014e-01 -1.27899218e+00
-4.03906792e-01 -3.78577530e-01 2.76677340e-01 2.27178529e-01
1.18542977e-01 -3.70998025e-01 -7.64368236e-01 -4.83430326e-01
3.58587235e-01 -1.24544168e+00 -9.71134752e-02 6.52271509e-01
-6.79818749e-01 -2.83980161e-01 9.35772836e-01 -4.16769058e-01
6.12241268e-01 -2.06635356e+00 6.80119276e-01 -3.31764460e-01
8.81501660e-02 -1.20683126e-02 -2.88390160e-01 2.39621341e-01
2.53466945e-02 2.73229718e-01 -1.72354534e-01 -2.31930077e-01
9.25429538e-02 4.84737635e-01 -4.02277321e-01 3.83075863e-01
2.96279758e-01 1.20216405e+00 -6.52528644e-01 -2.48135224e-01
1.57257125e-01 5.04605651e-01 -5.74126959e-01 2.65731156e-01
-7.77473927e-01 5.06480753e-01 -4.76091594e-01 9.09959853e-01
5.36661983e-01 -4.54567313e-01 -1.62879840e-01 -2.42886469e-01
6.56140409e-03 2.23317072e-01 -1.25588346e+00 1.90605462e+00
-4.73786108e-02 6.60289705e-01 -6.67095184e-02 -6.25617445e-01
5.49957693e-01 1.27802119e-01 4.15861547e-01 -4.07629460e-01
1.94349408e-01 2.05540985e-01 2.51750648e-02 -8.14683557e-01
4.16194260e-01 3.71868819e-01 -7.92806298e-02 5.11488914e-01
-3.82240774e-04 -4.21457320e-01 1.01232059e-01 2.14508325e-01
1.30321002e+00 7.57053435e-01 -5.41401319e-02 1.56042978e-01
-1.27740745e-02 -1.67189911e-02 4.55647945e-01 4.99033511e-01
-1.31782651e-01 1.10167396e+00 4.84000415e-01 -7.46924639e-01
-1.16284299e+00 -1.16138935e+00 -4.48226221e-02 5.77231407e-01
5.40830433e-01 -1.83761716e-01 -6.30819201e-01 -6.70321703e-01
2.02243805e-01 2.66879171e-01 -8.23539734e-01 -1.26979817e-02
-7.60096431e-01 -2.85167955e-02 1.62215217e-03 8.51924479e-01
1.33555740e-01 -1.39972508e+00 -1.09315073e+00 -1.54406652e-01
-2.83602238e-01 -1.25358295e+00 -5.00957251e-01 2.09110588e-01
-9.80558515e-01 -1.20819592e+00 -7.97344208e-01 -7.99162030e-01
8.66126180e-01 3.14216077e-01 1.13055742e+00 -8.47924501e-02
-4.81179625e-01 5.23840964e-01 -4.37144578e-01 -2.52568185e-01
-2.94923425e-01 1.63961306e-01 1.37980953e-01 -1.65341273e-01
4.64755371e-02 -5.77119470e-01 -9.81341660e-01 4.81137693e-01
-8.19660604e-01 4.47268754e-01 5.21766245e-01 1.56157658e-01
7.75072634e-01 -5.80987811e-01 -1.56402171e-01 -3.15102667e-01
-2.17391118e-01 -5.20353734e-01 -4.01046008e-01 5.10466248e-02
2.59238407e-02 2.53446084e-02 4.11448628e-01 -7.18961358e-01
-6.21300876e-01 5.02457321e-01 2.57719625e-02 -8.36928666e-01
-4.73950565e-01 -1.71730280e-01 -1.51155278e-01 1.73301861e-01
5.76401055e-01 -6.04020897e-03 1.23283342e-02 -6.73687935e-01
6.61265612e-01 5.50947309e-01 6.96732283e-01 -4.44559395e-01
7.72720575e-01 7.72023618e-01 -1.10432260e-01 -6.08951747e-01
-7.05462694e-01 -1.87136382e-01 -1.13043249e+00 -3.31044406e-01
1.02234483e+00 -1.00761008e+00 -8.41160893e-01 3.74091983e-01
-1.25071216e+00 -8.75896871e-01 -7.52975643e-01 2.77347028e-01
-9.20536101e-01 3.70091572e-02 -5.84929287e-01 -4.13991719e-01
-8.20667297e-02 -1.27039874e+00 1.39463317e+00 3.66598576e-01
-2.06952646e-01 -6.52541161e-01 6.60417543e-04 3.20738763e-01
1.99155789e-02 3.13670903e-01 5.30187368e-01 -3.56721789e-01
-1.17251408e+00 -6.41300157e-02 1.97248403e-02 8.09547752e-02
1.63342850e-03 2.68058538e-01 -7.64690161e-01 -8.57566893e-02
-4.80387777e-01 -6.66984558e-01 8.66339684e-01 4.55105811e-01
1.41441202e+00 -5.37797511e-01 -5.11809647e-01 8.59353900e-01
1.15791202e+00 1.68017372e-01 5.16022146e-01 2.98124284e-01
8.84834468e-01 4.63279575e-01 4.23765481e-01 2.46703386e-01
4.39551413e-01 7.14319766e-01 9.38419640e-01 1.37391821e-01
-3.68724734e-01 -3.19446266e-01 4.65569854e-01 5.66246450e-01
-3.05967301e-01 -4.77556556e-01 -9.98644233e-01 3.69447112e-01
-1.75385737e+00 -7.21578062e-01 -1.11828990e-01 1.69263697e+00
5.31388402e-01 1.25041246e-01 1.62964389e-01 -5.57002351e-02
2.62676507e-01 7.46187419e-02 -1.06276119e+00 -2.48341784e-01
-1.52604312e-01 -6.36789277e-02 3.77138346e-01 2.28397608e-01
-9.64903951e-01 7.71914244e-01 6.20405245e+00 2.10736215e-01
-9.67599928e-01 -6.01633452e-02 3.79668534e-01 -5.57546020e-01
-2.81646848e-01 9.63683948e-02 -7.03241229e-01 4.93629605e-01
7.70239472e-01 4.42502052e-01 4.73256141e-01 7.10779905e-01
1.44358706e-02 -1.67121843e-01 -1.39918518e+00 8.76929104e-01
6.94615096e-02 -1.32355642e+00 1.01832576e-01 2.01305062e-01
6.61597252e-01 3.06721658e-01 5.05053438e-02 6.88881576e-02
2.03713119e-01 -9.15080786e-01 1.42978454e+00 7.66702950e-01
5.69575846e-01 5.43442518e-02 1.75809473e-01 4.22328949e-01
-1.13025594e+00 -3.24698716e-01 -4.10123020e-01 -1.71633869e-01
2.91373163e-01 -6.88173994e-02 -5.52506089e-01 5.62752262e-02
9.66829658e-01 1.00720716e+00 -6.84125841e-01 1.03707802e+00
-5.15010595e-01 2.43568495e-01 -4.61578220e-01 9.64217708e-02
2.35953592e-02 -1.32176131e-01 5.83695114e-01 6.83408439e-01
4.98938173e-01 3.17290992e-01 -4.90008369e-02 1.16885340e+00
-3.49113911e-01 -4.97044027e-01 -6.17530584e-01 -1.90108009e-02
4.51998770e-01 1.31130183e+00 -1.00706172e+00 -2.57460177e-01
-3.15380841e-01 1.29669607e+00 3.64616185e-01 3.29576999e-01
-7.71988630e-01 -9.59860012e-02 7.87976146e-01 2.97866702e-01
8.23778570e-01 -4.26828474e-01 2.80685008e-01 -1.26599216e+00
3.03849131e-01 -4.00713325e-01 1.13477983e-01 -1.53998208e+00
-7.49640942e-01 4.82786655e-01 -3.57138626e-02 -1.44707727e+00
-4.38101254e-02 -1.05782461e+00 -4.14135873e-01 3.09440523e-01
-9.94858086e-01 -1.14912355e+00 -5.71306646e-01 3.66216511e-01
8.67683470e-01 3.12703013e-01 4.88337487e-01 -2.06689060e-01
-4.92637873e-01 2.29235869e-02 2.47751959e-02 2.35323280e-01
1.88813105e-01 -1.06376243e+00 8.14319789e-01 6.06891632e-01
3.09625983e-01 1.35772809e-01 4.73046839e-01 -5.97404897e-01
-1.89091146e+00 -1.13353646e+00 2.42121935e-01 -1.10287607e+00
4.24342096e-01 -7.53182292e-01 -7.55205691e-01 9.86931503e-01
-9.02715921e-02 6.74072266e-01 7.64413103e-02 -4.56192613e-01
-3.35723966e-01 2.62346476e-01 -7.83394158e-01 5.14272690e-01
1.64492452e+00 -3.86162847e-01 -7.22361326e-01 3.45278382e-01
1.11063528e+00 -8.42695296e-01 -8.35976303e-01 1.86018005e-01
7.92743623e-01 -8.68222415e-01 1.15784633e+00 -7.52199352e-01
7.15208828e-01 -4.05392468e-01 -2.40581349e-01 -1.01144397e+00
-1.79878756e-01 -3.90826255e-01 -5.87153733e-01 6.48878574e-01
2.61915296e-01 -1.47672147e-01 8.75185966e-01 6.06236458e-01
-3.26862127e-01 -9.86005247e-01 -9.61888909e-01 -8.03897560e-01
-1.34446695e-02 -3.73004109e-01 2.83495992e-01 4.82494831e-01
-2.64615774e-01 1.66306272e-02 -6.13196492e-02 1.10247776e-01
3.53764832e-01 3.21564883e-01 8.79136264e-01 -1.09654689e+00
-1.31798565e-01 -3.72230738e-01 -6.40913606e-01 -1.41295862e+00
6.66558370e-02 -9.01739836e-01 2.00961344e-02 -1.73441005e+00
2.43670523e-01 -3.79523546e-01 2.87541956e-01 7.58415401e-01
1.35759979e-01 3.04696620e-01 3.91934931e-01 4.98106122e-01
-8.63746226e-01 5.19407690e-01 1.40362775e+00 -2.35164255e-01
-2.91334450e-01 -1.06395841e-01 -3.76297146e-01 8.44721854e-01
6.40842855e-01 -3.35019320e-01 -1.43304303e-01 -8.77297103e-01
4.10820156e-01 8.09555799e-02 7.87168920e-01 -1.03453362e+00
-5.99264773e-03 -1.51220784e-01 7.33398139e-01 -7.65816569e-01
6.85103536e-01 -8.50033283e-01 5.36632776e-01 6.44184411e-01
-3.00766081e-01 2.62428224e-01 1.75649524e-01 7.00257957e-01
3.21773618e-01 3.06076594e-02 4.51882541e-01 -3.41168255e-01
-9.45886493e-01 6.96063578e-01 -2.81149328e-01 1.94890484e-01
1.08866811e+00 -5.19361496e-01 -4.49116528e-01 -8.23487714e-02
-9.88891363e-01 3.31875175e-01 8.89543593e-01 9.03908849e-01
8.10402393e-01 -1.37865412e+00 -2.38504738e-01 2.41100192e-01
1.20885096e-01 8.04005742e-01 2.33738914e-01 6.16000533e-01
-7.67745018e-01 2.49589190e-01 -3.47244114e-01 -8.86662245e-01
-8.89528394e-01 5.92914402e-01 4.21000659e-01 6.36836410e-01
-9.72507715e-01 1.11371839e+00 3.09540212e-01 -4.18036610e-01
3.02690953e-01 -9.83276606e-01 -1.02108335e-02 -1.22049659e-01
2.46805713e-01 3.09694886e-01 -1.88747957e-01 -8.20631564e-01
-3.34898084e-01 9.60571766e-01 1.24618754e-01 -2.50752661e-02
1.52253497e+00 -1.09156571e-01 3.71575505e-02 7.44278610e-01
1.18821490e+00 -3.15363020e-01 -1.95917070e+00 -2.66057383e-02
-4.53100324e-01 -4.34948951e-01 -4.17573392e-01 -5.70267141e-01
-8.81469429e-01 1.03593314e+00 5.35466731e-01 -2.28291880e-02
8.67338479e-01 6.63850665e-01 6.59217417e-01 5.08759677e-01
3.29012007e-01 -8.87503445e-01 4.73927766e-01 4.69478905e-01
1.15005481e+00 -1.15448058e+00 -9.02393647e-03 -1.89399704e-01
-5.34959376e-01 1.12745988e+00 6.60989165e-01 -1.69915318e-01
5.18731475e-01 7.04275891e-02 -1.81757584e-01 -2.74487436e-01
-9.14692640e-01 -1.26698136e-01 3.07329774e-01 4.71714914e-01
-6.98642582e-02 -4.57971636e-03 5.48653781e-01 3.63926709e-01
-3.16035300e-01 -1.10620521e-01 6.42097414e-01 1.14859259e+00
-3.88975590e-01 -6.99900866e-01 -2.83843398e-01 3.44988406e-01
-1.34132594e-01 3.57650906e-01 -5.60287297e-01 8.11397791e-01
3.90571684e-01 3.52325022e-01 5.03849506e-01 -4.10387576e-01
4.64282662e-01 -6.90591931e-02 7.85164714e-01 -5.78366041e-01
-1.36073981e-03 -1.23523042e-01 -3.31237406e-01 -9.07074869e-01
-5.26290476e-01 -9.41307068e-01 -1.38185966e+00 2.44990233e-02
-5.82400486e-02 -3.50338727e-01 7.12694049e-01 6.18823409e-01
4.84557360e-01 5.65784276e-01 4.67641592e-01 -1.54450166e+00
-3.39685351e-01 -7.65695930e-01 -3.21602106e-01 5.58967829e-01
6.82390094e-01 -6.80327535e-01 -4.37136292e-01 2.70876855e-01] | [7.07502555847168, -1.9152122735977173] |
3b558ead-3ba7-4732-a4e4-41729a0f78a9 | simple-questions-generate-named-entity | 2112.08808 | null | https://arxiv.org/abs/2112.08808v4 | https://arxiv.org/pdf/2112.08808v4.pdf | Simple Questions Generate Named Entity Recognition Datasets | Recent named entity recognition (NER) models often rely on human-annotated datasets, requiring the significant engagement of professional knowledge on the target domain and entities. This research introduces an ask-to-generate approach that automatically generates NER datasets by asking questions in simple natural language to an open-domain question answering system (e.g., "Which disease?"). Despite using fewer in-domain resources, our models, solely trained on the generated datasets, largely outperform strong low-resource models by an average F1 score of 19.4 for six popular NER benchmarks. Furthermore, our models provide competitive performance with rich-resource models that additionally leverage in-domain dictionaries provided by domain experts. In few-shot NER, we outperform the previous best model by an F1 score of 5.2 on three benchmarks and achieve new state-of-the-art performance. | ['Jaewoo Kang', 'Jinhyuk Lee', 'Seunghyun Yoon', 'Jaehyo Yoo', 'Hyunjae Kim'] | 2021-12-16 | null | null | null | null | ['few-shot-ner'] | ['natural-language-processing'] | [-6.62103668e-02 4.33867365e-01 -3.33858058e-02 -3.55734855e-01
-1.53091967e+00 -7.93514431e-01 7.04985082e-01 2.02977151e-01
-8.14973116e-01 1.10141945e+00 5.83855212e-01 4.79458421e-02
2.92179346e-01 -9.44100678e-01 -4.83345777e-01 1.42972782e-01
4.93451148e-01 6.63005650e-01 1.53219476e-01 -5.75365782e-01
-1.11012004e-01 -7.17242807e-02 -1.06935251e+00 1.90369904e-01
1.25960231e+00 6.92467928e-01 -8.94702673e-02 5.70981205e-01
-6.25240266e-01 9.64172721e-01 -8.85504425e-01 -9.91548538e-01
-7.45614916e-02 -4.55483228e-01 -1.09762180e+00 -5.89241147e-01
2.14100420e-01 -6.33727685e-02 -3.70385289e-01 7.85847127e-01
6.94215178e-01 3.80572766e-01 6.00190699e-01 -6.75683677e-01
-1.30696344e+00 7.16755211e-01 2.27719069e-01 2.89744437e-01
5.93840837e-01 2.74374317e-02 1.28177083e+00 -1.17990899e+00
1.06022024e+00 5.40274262e-01 7.94080138e-01 9.77817595e-01
-1.26532054e+00 -5.26776075e-01 -3.36825669e-01 -1.11880815e-02
-1.49607897e+00 -5.27204573e-01 2.98407137e-01 -1.10778883e-01
1.26426530e+00 1.19597927e-01 -1.98468134e-01 1.34358633e+00
-1.48958713e-01 4.11309838e-01 8.60351384e-01 -2.64144540e-01
3.20238620e-01 1.41084880e-01 2.33208045e-01 4.57407922e-01
3.26204598e-01 -2.44943738e-01 -2.32384324e-01 -3.74176621e-01
4.76772577e-01 -1.23348810e-01 -1.81960523e-01 3.72656047e-01
-1.15276217e+00 8.42215478e-01 2.51033396e-01 5.80783904e-01
-7.80177474e-01 -2.65268475e-01 2.45349899e-01 3.11840326e-03
5.83001971e-01 1.27362704e+00 -8.55738223e-01 -1.46591038e-01
-9.54192936e-01 1.19477287e-01 1.42100823e+00 1.12204790e+00
8.28046381e-01 -1.01759337e-01 -7.48873532e-01 1.02997935e+00
-2.51173884e-01 5.21650434e-01 5.61672688e-01 -7.74140596e-01
4.14964825e-01 7.68424749e-01 5.73804080e-01 -6.22972548e-01
-2.90630996e-01 -4.21828628e-01 -7.26176381e-01 -7.18164861e-01
2.96684414e-01 -7.33617842e-01 -9.43583846e-01 1.63196599e+00
3.92818213e-01 2.02797532e-01 5.08713186e-01 5.76142073e-01
1.52796292e+00 7.65678942e-01 6.43562257e-01 1.90922618e-01
1.73418260e+00 -1.10861599e+00 -7.88612664e-01 -3.41068953e-01
6.23300374e-01 -6.41323388e-01 9.76616800e-01 -1.80265486e-01
-7.59008288e-01 -4.47334558e-01 -4.41329598e-01 -4.19744730e-01
-7.67985702e-01 3.51297587e-01 3.35976839e-01 5.29326439e-01
-9.20616686e-01 3.42053473e-01 -4.08787787e-01 -7.07564533e-01
1.69936657e-01 -2.54070133e-01 -5.20315826e-01 -1.89796492e-01
-1.57159626e+00 1.08279312e+00 3.90175164e-01 -3.36304843e-01
-1.01033485e+00 -1.23616886e+00 -1.03225422e+00 2.26295725e-01
4.72875059e-01 -8.85828078e-01 1.41880155e+00 -3.11236441e-01
-1.22141850e+00 9.82580006e-01 -3.58705819e-01 -6.01213932e-01
1.93576425e-01 -5.25037169e-01 -9.47134912e-01 2.43110791e-01
6.38857007e-01 7.21336603e-01 1.78038716e-01 -8.02420259e-01
-3.85307997e-01 1.45147115e-01 2.83532053e-01 -1.55241609e-01
-4.64090377e-01 3.26741457e-01 -2.39082128e-01 -6.31160259e-01
-6.39495909e-01 -6.43038273e-01 -6.38022900e-01 -4.67855930e-01
-6.01453900e-01 -5.62862933e-01 1.47058323e-01 -9.05888796e-01
1.44529033e+00 -1.63694823e+00 -5.00910640e-01 -4.57840174e-01
2.56513745e-01 4.22079742e-01 -4.07822162e-01 6.84574306e-01
-7.88791403e-02 5.12272894e-01 -1.35709494e-01 -2.97112484e-02
1.39342532e-01 1.03749938e-01 -3.61590922e-01 -2.81774431e-01
7.72326708e-01 1.15591550e+00 -1.27221322e+00 -5.98478138e-01
-3.47658962e-01 4.21316445e-01 -5.23371041e-01 5.43921828e-01
-3.08324695e-01 2.92108923e-01 -6.46086454e-01 3.85442883e-01
2.48330623e-01 -6.55748367e-01 -8.61048251e-02 -2.06708536e-01
5.03318990e-03 6.73772693e-01 -9.35641766e-01 1.68028367e+00
-6.00696862e-01 2.40199238e-01 -3.06037754e-01 -5.30645847e-01
1.09697282e+00 7.50610352e-01 2.46992126e-01 -5.50425231e-01
-1.03571340e-01 2.12640390e-01 -4.19419110e-01 -5.72674870e-01
5.65388203e-01 -1.85250193e-01 -6.16856277e-01 4.11946714e-01
5.41500390e-01 1.91564143e-01 6.24186814e-01 4.16949630e-01
1.75268626e+00 -7.97225460e-02 5.72800279e-01 -1.24753632e-01
4.90719557e-01 4.40979987e-01 8.55111778e-01 7.89591312e-01
-1.45332381e-01 6.29666090e-01 2.59077311e-01 -2.93347925e-01
-1.05125177e+00 -9.26740587e-01 -8.07578042e-02 1.26587272e+00
-3.56084079e-01 -5.11436701e-01 -9.34840739e-01 -1.13540256e+00
-3.34567070e-01 1.19580340e+00 -6.65066421e-01 1.82574421e-01
-4.00531352e-01 -4.83554482e-01 1.24653554e+00 6.08848333e-01
4.84328628e-01 -1.28302538e+00 -3.11081409e-01 5.94150901e-01
-7.24076033e-01 -1.68272269e+00 -6.85343504e-01 -9.81246158e-02
-5.00121653e-01 -9.91832852e-01 -1.07016575e+00 -7.63136029e-01
6.23708248e-01 -2.66102135e-01 1.77483928e+00 -3.10094982e-01
-3.10290754e-01 3.54988545e-01 -6.72834158e-01 -4.24117088e-01
-3.24842334e-01 6.11275911e-01 -2.74195343e-01 -1.21761248e-01
8.55738640e-01 -2.51288712e-01 -4.69779462e-01 3.53806078e-01
-9.55254614e-01 -2.90088624e-01 6.67148232e-01 9.07808185e-01
5.63974023e-01 -4.87654269e-01 1.19505203e+00 -1.13531625e+00
7.90924430e-01 -9.46431994e-01 -2.27469146e-01 6.03718519e-01
-5.31627059e-01 3.35550785e-01 9.97434139e-01 -3.97750586e-01
-1.46627617e+00 -4.16864902e-02 -3.80301625e-01 -7.28107169e-02
-6.89838648e-01 5.43615222e-01 -9.08583924e-02 4.19599205e-01
1.28819120e+00 1.84301689e-01 -7.76718199e-01 -7.58901417e-01
6.13059640e-01 5.03298461e-01 7.65292168e-01 -4.96526092e-01
8.74750018e-01 1.72785982e-01 -6.43040001e-01 -5.75917304e-01
-1.62579381e+00 -6.67442560e-01 -3.48692775e-01 2.85667926e-01
1.23498964e+00 -1.08426511e+00 -1.58139467e-01 1.14800269e-02
-1.35480309e+00 -1.88066270e-02 -5.14339983e-01 2.76800454e-01
-1.71660520e-02 3.06729227e-02 -7.87992835e-01 -5.35888910e-01
-8.52392673e-01 -2.55186230e-01 1.04111040e+00 6.12129450e-01
-4.44624901e-01 -1.01898122e+00 4.83490586e-01 6.14028156e-01
6.22895300e-01 2.34579310e-01 6.42335951e-01 -1.59029448e+00
-9.68316942e-02 -2.42247030e-01 -2.67167956e-01 1.77919745e-01
1.64730325e-01 -6.43762469e-01 -7.94031501e-01 3.97442073e-01
-3.00481677e-01 -4.78531152e-01 5.74889779e-01 -3.20035249e-01
7.29055464e-01 -4.47048128e-01 -2.65807897e-01 9.81678218e-02
1.36797678e+00 -2.07868159e-01 5.44410467e-01 1.61859781e-01
5.14582932e-01 7.24262834e-01 4.64959651e-01 4.44545418e-01
8.67378533e-01 2.41619676e-01 -1.79780558e-01 -3.03899068e-02
-1.92623332e-01 -7.06138134e-01 2.11331144e-01 7.80900776e-01
1.35874689e-01 -1.18424855e-01 -1.19171476e+00 1.15954697e+00
-1.51615357e+00 -1.05055523e+00 -2.47573555e-01 1.68322408e+00
1.35989475e+00 -1.13897465e-01 -2.71349609e-01 -7.95213401e-01
8.50790739e-01 -2.47823875e-02 -7.50115871e-01 -1.76047832e-01
-2.29422599e-01 8.39842021e-01 4.31792527e-01 1.12245061e-01
-1.14428401e+00 1.15924621e+00 6.39487743e+00 6.98905110e-01
-5.01590967e-01 3.18851143e-01 4.29055333e-01 2.43527576e-01
-4.25941348e-01 1.51286751e-01 -1.23925829e+00 3.57945502e-01
1.60294974e+00 -6.95814252e-01 1.45767465e-01 1.07128143e+00
-9.04578269e-02 3.41046035e-01 -1.07511640e+00 6.76015198e-01
3.41729045e-01 -1.44761014e+00 1.27654850e-01 -1.27030805e-01
1.05626428e+00 2.87646115e-01 -6.30667031e-01 9.58111286e-01
1.02452767e+00 -1.17086899e+00 3.39728408e-02 8.32677662e-01
8.65942299e-01 -4.03928608e-01 9.83435512e-01 5.73085129e-01
-1.19471514e+00 2.12078735e-01 -3.69689077e-01 2.19785854e-01
5.33376515e-01 7.66586900e-01 -9.75692987e-01 7.86555409e-01
6.38237536e-01 2.35145807e-01 -5.19914865e-01 1.05277717e+00
-6.36371434e-01 9.34966862e-01 -2.86022842e-01 -2.47154236e-01
2.90698260e-01 3.17546427e-01 1.65938124e-01 1.53823733e+00
2.55432218e-01 5.87268531e-01 7.75474608e-02 1.10025394e+00
-7.76964962e-01 4.38820630e-01 -5.54476619e-01 -4.53603119e-01
6.26815557e-01 1.62998641e+00 -2.06043959e-01 -6.47959590e-01
-4.64139044e-01 1.03428495e+00 4.72924352e-01 2.37964615e-01
-7.98603475e-01 -9.85973358e-01 4.05575216e-01 -4.36936654e-02
5.65534532e-01 1.58045486e-01 1.26132858e-04 -1.46031511e+00
-1.41177833e-01 -9.19834137e-01 6.63530231e-01 -8.01491678e-01
-1.91417611e+00 9.22348082e-01 -4.53011245e-01 -9.65147376e-01
-5.45620441e-01 -2.42374629e-01 -7.15328813e-01 1.02149045e+00
-1.62541437e+00 -1.03650284e+00 -3.35619658e-01 5.33761263e-01
5.78339458e-01 -1.50823994e-02 1.41839731e+00 6.85228705e-01
-5.57581127e-01 7.20392883e-01 -7.74881318e-02 9.59331870e-01
1.10538518e+00 -1.23088348e+00 7.84248590e-01 8.08929205e-01
3.29953671e-01 8.67214024e-01 4.95419592e-01 -7.67718494e-01
-9.82883453e-01 -1.34075952e+00 1.81228065e+00 -1.06453085e+00
8.04883897e-01 -1.96833491e-01 -9.64537978e-01 5.37079096e-01
4.02375013e-01 1.50095701e-01 1.35114253e+00 2.05561444e-01
-6.18019581e-01 2.40887716e-01 -1.38529611e+00 4.32905257e-01
1.05632532e+00 -6.94414556e-01 -1.32670343e+00 3.03195596e-01
9.16449666e-01 -1.44775063e-01 -1.36916602e+00 1.40437305e-01
6.52299002e-02 -2.02886805e-01 7.44234562e-01 -1.28252590e+00
6.31335378e-01 -2.81112164e-01 5.49192056e-02 -1.23616934e+00
-2.34071404e-01 -6.13214016e-01 1.83606683e-03 1.62799692e+00
1.00834203e+00 -4.36008602e-01 5.86466491e-01 1.04422772e+00
4.47492581e-03 -4.21800852e-01 -6.87889040e-01 -7.57587612e-01
7.87299946e-02 -1.06928073e-01 5.44340312e-01 1.31135881e+00
2.29072850e-03 1.01298916e+00 -1.10385194e-01 1.80578440e-01
4.60890681e-01 -1.29169375e-01 5.35172462e-01 -1.44855011e+00
-1.25697320e-02 7.84338489e-02 1.34601280e-01 -9.19117928e-01
3.84833992e-01 -8.31492424e-01 1.86503306e-01 -1.94522631e+00
2.53913254e-01 -1.91784590e-01 -3.81528914e-01 7.99087107e-01
-6.59054101e-01 1.85708493e-01 -8.17439929e-02 2.86623929e-02
-1.12805068e+00 6.01920426e-01 7.28412449e-01 1.35197416e-01
-6.73300400e-02 -4.80489701e-01 -1.22519374e+00 6.89311802e-01
6.07391417e-01 -8.47622037e-01 -4.47460227e-02 -6.38530433e-01
2.33108908e-01 2.11752653e-02 8.47123191e-02 -9.31254923e-01
6.20159090e-01 -1.12919144e-01 1.54317632e-01 -2.08287507e-01
4.18105461e-02 -3.63190055e-01 -8.94861147e-02 5.57652488e-02
-6.64550364e-01 -2.19364092e-03 9.25060511e-02 5.63807189e-01
-3.27311397e-01 -4.79807496e-01 4.41353053e-01 -3.59222382e-01
-8.97769034e-01 2.79804111e-01 -1.60564855e-01 1.09567463e+00
8.25473785e-01 2.39227727e-01 -6.32561564e-01 -4.87993240e-01
-6.74317777e-01 4.03219521e-01 4.67885025e-02 4.96054441e-01
2.46699065e-01 -1.18923986e+00 -1.16699755e+00 -4.89628971e-01
4.93367434e-01 -1.75559193e-01 3.58963370e-01 2.99002737e-01
-1.69239342e-01 7.11596966e-01 5.94491586e-02 1.35915965e-01
-5.21633923e-01 2.94799834e-01 9.18354839e-02 -9.25070465e-01
-1.99027330e-01 7.40240574e-01 -5.86593226e-02 -9.31313097e-01
-3.11792523e-01 1.20157540e-01 -3.98284316e-01 3.83160599e-02
7.98492610e-01 3.61449182e-01 1.55750511e-03 -5.30210376e-01
-3.85628641e-01 3.23902480e-02 -6.96373954e-02 -1.27749294e-01
1.46868980e+00 3.00383657e-01 1.12139754e-01 1.04340976e-02
1.05017817e+00 2.60195166e-01 -6.70106709e-01 -5.04617512e-01
5.51685035e-01 1.43452644e-01 -2.75396883e-01 -1.28599906e+00
-4.45235938e-01 4.96477455e-01 1.31192682e-02 9.18470249e-02
8.19916189e-01 4.09486413e-01 1.15939105e+00 9.15475965e-01
3.66811216e-01 -1.13340080e+00 3.20104696e-03 1.00672174e+00
6.09088898e-01 -1.33395696e+00 -4.48291242e-01 -1.64200827e-01
-9.70870256e-01 6.79021776e-01 7.83079684e-01 -1.49604514e-01
4.50991094e-01 1.15640692e-01 1.80633098e-01 -4.65920568e-02
-8.74765515e-01 -5.52929878e-01 4.41656500e-01 6.35075867e-01
7.53615379e-01 -7.72611573e-02 -5.24272680e-01 1.37404585e+00
-2.63597339e-01 1.55577973e-01 3.48968238e-01 5.81378937e-01
-3.21864009e-01 -1.22836602e+00 6.33886233e-02 5.60395181e-01
-9.94060278e-01 -7.07191169e-01 -5.09811580e-01 4.53406990e-01
-1.19616687e-01 1.24240017e+00 -1.53672844e-01 -5.29496558e-02
6.93124056e-01 4.54922765e-01 -4.07309830e-01 -1.12240219e+00
-1.01589966e+00 -6.20158970e-01 7.50310063e-01 -3.81829470e-01
-2.31896445e-01 -2.68320322e-01 -1.31280327e+00 3.33704725e-02
-2.43402511e-01 7.28203893e-01 4.25234437e-01 9.59340215e-01
1.05700445e+00 2.55180061e-01 3.12257171e-01 5.41165769e-02
-6.49146020e-01 -1.16325998e+00 -7.55984336e-02 6.10317528e-01
-1.52233601e-01 -2.13890135e-01 -1.33107007e-01 2.40444288e-01] | [9.692569732666016, 9.411559104919434] |
71c56c8e-4b15-4015-8c82-f4e780b6a723 | unsupervised-induction-of-tree-substitution | null | null | https://aclanthology.org/D10-1117 | https://aclanthology.org/D10-1117.pdf | Unsupervised Induction of Tree Substitution Grammars for Dependency Parsing | Inducing a grammar directly from text is one of the oldest and most challenging tasks in Computational Linguistics. Significant progress has been made for inducing dependency grammars, however the models employed are overly simplistic, particularly in comparison to supervised parsing models. In this paper we present an approach to dependency grammar induction using tree substitution grammar which is capable of learning large dependency fragments and thereby better modelling the text. We define a hierarchical non-parametric Pitman-Yor Process prior which biases towards a small grammar with simple productions. This approach significantly improves the state-of-the-art, when measured by head attachment accuracy. | ['Trevor Cohn', 'Phil Blunsom'] | 2010-10-01 | null | null | null | null | ['dependency-grammar-induction', 'unsupervised-dependency-parsing'] | ['natural-language-processing', 'natural-language-processing'] | [ 1.32587090e-01 7.74202228e-01 -1.92136437e-01 -8.02594364e-01
-9.42462146e-01 -5.46085715e-01 5.73157012e-01 3.41440797e-01
-4.67817664e-01 8.99046719e-01 3.72677356e-01 -6.26893818e-01
2.26671621e-01 -5.70708632e-01 -4.17692155e-01 -5.84861219e-01
-8.57846290e-02 1.12893760e+00 1.20781839e-01 -2.67973751e-01
1.03137188e-01 2.28637293e-01 -1.26565552e+00 1.09448537e-01
7.25292742e-01 5.99213541e-02 4.68855590e-01 6.84988618e-01
-2.64221966e-01 9.62757885e-01 -3.77529502e-01 -6.63995683e-01
-4.54950869e-01 -5.52669406e-01 -1.18782163e+00 3.18110711e-03
1.25784799e-01 4.65795621e-02 3.22839439e-01 9.10662591e-01
1.55911714e-01 -1.48907825e-01 6.51616395e-01 -5.80602586e-01
-6.63042665e-01 1.23475301e+00 -3.15987408e-01 3.40886414e-01
5.95711350e-01 -5.58782756e-01 1.31000745e+00 -7.24883795e-01
7.62637079e-01 1.71660852e+00 3.85285527e-01 6.42644048e-01
-1.51267457e+00 -4.96653765e-01 3.66055459e-01 -2.70864367e-01
-1.16773868e+00 -4.14554834e-01 7.18743861e-01 -2.80079752e-01
1.48508179e+00 -8.21539834e-02 1.42069638e-01 1.00131392e+00
1.91229552e-01 7.76534855e-01 1.24624503e+00 -1.17722189e+00
-2.69122375e-03 -7.32672587e-02 3.94288838e-01 6.41394854e-01
2.62793779e-01 1.61371380e-02 -2.35135108e-01 -3.07912063e-02
8.27945590e-01 -6.86049402e-01 3.36416453e-01 1.68785863e-02
-7.16440499e-01 1.19578755e+00 -1.66386113e-01 6.79652452e-01
-1.34431079e-01 -1.05312631e-01 2.96783686e-01 3.21632147e-01
5.89571893e-01 1.35513425e-01 -6.86040699e-01 -1.42776519e-01
-7.11614549e-01 2.55234867e-01 1.05212772e+00 9.90707517e-01
5.61727524e-01 7.19068572e-02 2.21413046e-01 8.72716725e-01
5.43369532e-01 3.85176450e-01 2.69691974e-01 -5.77324629e-01
5.34541249e-01 4.98481303e-01 -2.44215593e-01 -3.88558477e-01
-6.83525443e-01 -1.08500369e-01 -2.98985213e-01 3.68831940e-02
4.95545357e-01 -1.79963455e-01 -1.08222401e+00 1.75926888e+00
3.08383524e-01 -4.93145138e-01 1.92774087e-01 2.64000535e-01
7.41785586e-01 5.14492631e-01 8.31401110e-01 -5.25572002e-01
1.41057289e+00 -5.97085059e-01 -8.30223978e-01 -5.68493366e-01
1.19151258e+00 -7.61297464e-01 1.17307150e+00 3.20210755e-01
-1.25534749e+00 -4.84033167e-01 -5.70039809e-01 -3.76965314e-01
-3.02082002e-01 -9.02072340e-02 9.84611869e-01 8.79419684e-01
-1.07591558e+00 4.10913974e-01 -9.48155701e-01 -4.51135904e-01
-2.16427222e-02 7.50426412e-01 -6.55124962e-01 -1.69046268e-01
-1.11883819e+00 1.11553288e+00 8.75708044e-01 -2.26480052e-01
-4.88042474e-01 -9.01211947e-02 -1.20705497e+00 -1.60256431e-01
3.49325746e-01 -3.25343341e-01 1.59790552e+00 -6.89946830e-01
-1.47876430e+00 1.35956454e+00 -2.92957038e-01 -3.47598672e-01
-7.29408562e-02 -2.68207848e-01 -2.71125913e-01 -2.64615715e-01
1.25749335e-01 6.83904707e-01 5.95773399e-01 -1.22413707e+00
-8.35517228e-01 -5.82754076e-01 5.19453324e-02 1.63268909e-01
2.31254280e-01 9.74297583e-01 -1.86766341e-01 -6.59076571e-01
1.00499295e-01 -1.09571517e+00 -2.52796143e-01 -1.28555071e+00
-1.74837098e-01 -9.09115493e-01 3.84888023e-01 -8.60502303e-01
1.50792551e+00 -1.76632512e+00 2.82854408e-01 -1.02692574e-01
-1.58018425e-01 1.71699494e-01 1.69590577e-01 5.04683256e-01
-2.85558939e-01 5.06910443e-01 -4.93254513e-01 -7.79982209e-01
-8.26180205e-02 1.15861225e+00 -8.62407386e-02 3.04395892e-02
3.43804598e-01 8.96734774e-01 -9.08971548e-01 -7.43798733e-01
2.84144487e-02 4.32664782e-01 -4.25007731e-01 3.35743934e-01
-1.83471471e-01 6.01593018e-01 -3.37932020e-01 6.37356639e-01
2.08222419e-01 2.84498334e-01 6.73313498e-01 6.21184051e-01
-3.73995394e-01 1.10373986e+00 -8.67274046e-01 1.68759167e+00
-5.91138721e-01 1.51992694e-01 1.35355089e-02 -9.14959967e-01
1.04039133e+00 6.43794000e-01 -1.60613805e-01 -1.73129603e-01
3.07140291e-01 2.44050950e-01 4.61639822e-01 -4.29873079e-01
1.59355104e-01 -7.68977344e-01 -4.53324974e-01 3.54358941e-01
4.26883817e-01 -4.18282390e-01 5.50497711e-01 -1.80222150e-02
8.32598269e-01 4.41179335e-01 7.36781955e-01 -5.82269967e-01
5.25023758e-01 -1.23980217e-01 6.69063687e-01 5.58933556e-01
1.95819646e-01 4.32181388e-01 7.00547516e-01 -5.05036473e-01
-8.13538313e-01 -7.73950994e-01 -4.07430530e-01 1.67438114e+00
-6.03778124e-01 -4.07962978e-01 -1.04229379e+00 -1.01301527e+00
-4.08485651e-01 1.26932418e+00 -6.89175248e-01 4.12781119e-01
-1.23696411e+00 -9.19879675e-01 6.29572988e-01 7.85149038e-01
-1.48877710e-01 -1.84731388e+00 -3.19207519e-01 6.82181656e-01
-8.57196450e-02 -1.36693633e+00 -5.80764608e-04 7.89058030e-01
-1.17978299e+00 -6.50133312e-01 -3.47118154e-02 -1.34335721e+00
5.96547008e-01 -5.21638632e-01 1.56157911e+00 2.12461814e-01
2.99398098e-02 -1.60210356e-01 -6.27919316e-01 -8.12277257e-01
-9.14480448e-01 4.40441847e-01 -2.02625096e-01 -5.11586428e-01
8.33841026e-01 -5.76248884e-01 3.08237880e-01 -3.59909654e-01
-8.56379986e-01 -2.27903262e-01 7.39651322e-01 6.04383528e-01
3.83414239e-01 -2.19581977e-01 3.43460798e-01 -1.68281853e+00
6.84299648e-01 -2.33764723e-01 -4.18365568e-01 1.41862053e-02
-6.58079565e-01 4.83298302e-01 6.12881541e-01 -3.71766500e-02
-1.45317817e+00 2.29290977e-01 -7.17256725e-01 5.29859662e-01
-8.21456194e-01 5.94185889e-01 -3.84675413e-01 1.83506072e-01
4.45202678e-01 -3.10168296e-01 -3.99606138e-01 -8.28498244e-01
3.39003295e-01 4.45370376e-01 5.21035492e-01 -9.56539154e-01
4.97851789e-01 6.55003125e-03 9.28613171e-02 -9.03917372e-01
-1.09833992e+00 -4.66243863e-01 -1.41977966e+00 2.43789092e-01
9.14145768e-01 -4.67568487e-01 -8.72927010e-02 1.85213566e-01
-1.28299963e+00 -3.12209845e-01 -1.39752299e-01 3.73916328e-01
-4.74152327e-01 3.98749679e-01 -1.06736851e+00 -8.37398291e-01
-2.83903539e-01 -7.96218514e-01 1.13874984e+00 -2.96534933e-02
-4.69319195e-01 -1.46015143e+00 4.88730997e-01 3.83042991e-02
-1.38710931e-01 2.82679647e-01 1.34107971e+00 -7.45445728e-01
1.51096195e-01 -1.11483991e-01 7.12666214e-02 3.36860240e-01
-4.64291088e-02 3.81414257e-02 -9.18707192e-01 -2.51456499e-02
1.61049992e-01 -3.45921308e-01 4.46909785e-01 4.23719585e-01
3.60274464e-01 -2.21054796e-02 -3.48208457e-01 1.99036390e-01
1.42922747e+00 3.71423095e-01 6.40743017e-01 3.57183307e-01
7.04986334e-01 9.30559695e-01 5.81290483e-01 -8.05839449e-02
6.24762297e-01 3.10357600e-01 1.23122551e-01 -1.42914981e-01
3.21055837e-02 -3.64356458e-01 3.72051805e-01 1.07553864e+00
-2.76947081e-01 -1.14596404e-01 -1.21293783e+00 7.24832177e-01
-1.71297550e+00 -4.80107099e-01 -4.74904001e-01 1.84472382e+00
1.06577647e+00 5.76071620e-01 1.02459431e-01 3.62039566e-01
6.09589159e-01 -5.69951348e-02 3.34661603e-01 -1.17898214e+00
-8.81550387e-02 8.37521195e-01 2.00978339e-01 1.03132725e+00
-1.10568976e+00 1.58486342e+00 7.38997412e+00 2.45212764e-01
-5.76887727e-01 2.28422493e-01 2.97449172e-01 5.88344097e-01
-1.09899253e-01 4.41725492e-01 -1.37194085e+00 9.19080749e-02
1.32765615e+00 2.06227049e-01 5.91137297e-02 7.72780478e-01
5.37661500e-02 -1.34499237e-01 -1.22987950e+00 5.36642134e-01
1.57972246e-01 -5.40914476e-01 6.75406381e-02 9.05064121e-02
3.48036230e-01 9.76003036e-02 -4.68238175e-01 5.68591475e-01
7.74506629e-01 -1.15985394e+00 4.08131927e-01 -2.82127298e-02
6.33704960e-01 -9.33641374e-01 9.42533672e-01 5.72303295e-01
-1.02836192e+00 1.84366792e-01 -4.05226767e-01 -5.59723914e-01
5.17417073e-01 4.33553487e-01 -7.60330558e-01 3.36827993e-01
5.55396974e-01 1.81176826e-01 -5.07960856e-01 4.51776743e-01
-9.71240461e-01 1.12084067e+00 -4.82170433e-01 1.62489098e-02
4.24017370e-01 -3.32177758e-01 3.32777441e-01 1.84120345e+00
-2.17826754e-01 3.51998121e-01 5.35150647e-01 3.01853210e-01
4.43482399e-02 5.67577839e-01 -7.65373588e-01 1.79746881e-01
2.30515972e-01 1.18020713e+00 -1.00429344e+00 -4.00305629e-01
-5.98167539e-01 5.74874878e-01 8.53009880e-01 -1.78449124e-01
-5.06266356e-01 -1.48714647e-01 1.37582794e-01 4.04112637e-02
3.58489335e-01 -3.96473050e-01 -3.42383951e-01 -8.73789489e-01
-8.23602304e-02 -7.96119928e-01 6.92825556e-01 -3.07721853e-01
-1.05827403e+00 9.71028090e-01 4.42351192e-01 -2.36679763e-01
-7.56821454e-01 -8.74631107e-01 -4.48389113e-01 1.10638893e+00
-1.49550498e+00 -1.44533908e+00 3.81375164e-01 1.12376474e-01
7.36886740e-01 2.02624351e-01 1.51335394e+00 -8.92299637e-02
-5.18956482e-01 3.67506087e-01 -3.67947489e-01 3.58416438e-01
4.63091582e-01 -1.73234999e+00 8.49381566e-01 9.60767806e-01
5.51544905e-01 6.79256678e-01 8.15019131e-01 -6.91782415e-01
-8.12637568e-01 -6.67714596e-01 1.75877500e+00 -8.41077507e-01
8.38817596e-01 -5.89609563e-01 -1.22970057e+00 1.10024333e+00
4.12387431e-01 -2.16972321e-01 7.57457614e-01 6.62858963e-01
-2.49259874e-01 4.16025221e-01 -8.81681621e-01 3.48396242e-01
9.91856635e-01 -3.95296901e-01 -1.25068867e+00 2.68259794e-01
4.26164091e-01 -2.51586944e-01 -1.00289977e+00 4.75747943e-01
3.41249526e-01 -7.14556217e-01 3.90511215e-01 -8.48490179e-01
1.92302853e-01 -1.81527529e-03 4.75844555e-02 -1.14551342e+00
-6.72689736e-01 -8.35222900e-01 9.26644877e-02 1.52849817e+00
7.54990339e-01 -3.15393418e-01 7.51592338e-01 7.17661202e-01
-1.55727312e-01 -3.60802829e-01 -7.54129171e-01 -7.30850101e-01
5.58018386e-01 -4.39076871e-01 1.38784632e-01 7.81597614e-01
3.30565721e-01 1.20825219e+00 -1.07847191e-01 -3.20712738e-02
6.94195390e-01 -1.36166245e-01 4.22550946e-01 -1.63685858e+00
-2.57074624e-01 -2.49683231e-01 -1.43625870e-01 -7.68418849e-01
5.74170411e-01 -8.56728852e-01 4.23319727e-01 -1.46908498e+00
-5.48210964e-02 -4.02172774e-01 3.94988507e-02 6.61679804e-01
-4.07131076e-01 1.32075399e-01 1.85951591e-03 5.40382527e-02
-2.41684973e-01 -8.82211626e-02 8.06384265e-01 4.68320966e-01
-2.80227363e-01 -9.11626145e-02 -6.22305095e-01 1.07385230e+00
1.04135621e+00 -1.11588359e+00 -2.42559314e-01 -5.73251724e-01
2.80617565e-01 -3.41453888e-02 -3.30569059e-01 -5.01549184e-01
-1.89874440e-01 7.54072219e-02 1.55389346e-02 -4.37910408e-01
-2.63407916e-01 -3.98891389e-01 -1.53373882e-01 1.54611692e-01
-1.92660660e-01 4.15560335e-01 2.17195839e-01 1.49200857e-01
-2.98304409e-01 -8.35913777e-01 8.23221743e-01 -5.59378326e-01
-5.33314884e-01 -1.28381655e-01 -5.44525146e-01 2.58820415e-01
6.31123245e-01 3.14053372e-02 3.47184837e-01 2.12401282e-02
-1.14612043e+00 -1.61383942e-01 2.34350547e-01 4.36113358e-01
6.09951317e-02 -6.48716569e-01 -9.07405972e-01 2.35251084e-01
7.58110499e-03 2.14661583e-01 -5.86832821e-01 3.96230698e-01
-4.96760130e-01 7.32698679e-01 1.61323994e-02 -3.94677877e-01
-1.47140849e+00 7.15440810e-01 -1.88923031e-01 -8.01863492e-01
-7.33558536e-01 8.42035472e-01 2.16988042e-01 -5.43120682e-01
1.20543227e-01 -4.54023123e-01 -6.67331219e-01 -1.23080708e-01
4.37272280e-01 2.33982150e-02 2.93073416e-01 -9.40556824e-01
-2.63443440e-01 4.02574033e-01 -4.66135800e-01 -2.45053038e-01
1.29370058e+00 -1.21551260e-01 -4.47563231e-01 6.67952895e-01
7.16588616e-01 6.67865053e-02 -8.39561820e-01 1.33597450e-02
7.81170428e-01 8.80160034e-02 -2.42875610e-02 -6.99122667e-01
-2.58284658e-01 8.70355368e-01 -9.78564098e-02 5.35716474e-01
7.89141476e-01 3.17295253e-01 4.52966511e-01 4.05610770e-01
4.31776047e-01 -1.12887776e+00 -7.18707323e-01 9.83521581e-01
6.18784606e-01 -1.19177949e+00 -1.25844061e-01 -8.73088121e-01
-5.53445816e-01 1.07799304e+00 2.60118425e-01 -4.13155735e-01
7.27884352e-01 7.72920191e-01 3.58908504e-01 -2.60804266e-01
-7.26594269e-01 -6.30823433e-01 -2.63916906e-02 7.87130356e-01
1.23535907e+00 2.94673234e-01 -8.87290657e-01 5.01460135e-01
-5.10200799e-01 -4.24986929e-01 4.17030096e-01 1.13115621e+00
-4.81030881e-01 -1.86164534e+00 -3.56595963e-01 9.48908776e-02
-1.12038398e+00 -5.65245807e-01 -5.86861014e-01 1.23394370e+00
7.49149173e-02 1.01906657e+00 -1.35176301e-01 1.98620498e-01
2.92133212e-01 5.99711001e-01 8.29788387e-01 -1.38049185e+00
-6.89065874e-01 4.47039932e-01 6.39156461e-01 1.20165730e-02
-6.63294435e-01 -1.04545355e+00 -1.45402825e+00 6.99563622e-02
-4.70314652e-01 5.69032371e-01 6.70024276e-01 1.31406748e+00
-2.95792788e-01 1.53233156e-01 2.83239484e-01 -8.65968645e-01
-3.08364064e-01 -1.37582493e+00 -5.42780995e-01 3.87414157e-01
-2.10499853e-01 -6.27389908e-01 -2.97819644e-01 5.11388958e-01] | [10.332280158996582, 9.751289367675781] |
c81b5ac0-c443-4300-b06a-a73e8fc5c62f | refinegan-universally-generating-waveform | 2111.00962 | null | https://arxiv.org/abs/2111.00962v3 | https://arxiv.org/pdf/2111.00962v3.pdf | RefineGAN: Universally Generating Waveform Better than Ground Truth with Highly Accurate Pitch and Intensity Responses | Most GAN(Generative Adversarial Network)-based approaches towards high-fidelity waveform generation heavily rely on discriminators to improve their performance. However, GAN methods introduce much uncertainty into the generation process and often result in mismatches of pitch and intensity, which is fatal when it comes to sensitive use cases such as singing voice synthesis(SVS). To address this problem, we propose RefineGAN, a high-fidelity neural vocoder focused on the robustness, pitch and intensity accuracy, and high-speed full-band audio generation. We applyed a pitch-guided refine architecture with a multi-scale spectrogram-based loss function to help stabilize the training process and maintain the robustness of the neural vocoder while using the GAN-based training method. Audio generated using this method shows a better performance in subjective tests when compared with the ground-truth audio. This result shows that the fidelity is even improved during the waveform reconstruction by eliminating defects produced by recording procedures. Moreover, it shows that models trained on a specified type of data can perform on totally unseen language and unseen speaker identically well. Generated sample pairs are provided onhttps://timedomain-tech.github.io/refinegan/. | ['Jing Guo', 'Wenxiao Zhao', 'Shengyuan Xu'] | 2021-11-01 | null | null | null | null | ['audio-generation', 'singing-voice-synthesis'] | ['audio', 'speech'] | [-4.71940860e-02 2.34151870e-01 2.95288116e-01 6.79454654e-02
-1.33058894e+00 -7.06158817e-01 3.13449442e-01 -2.86718518e-01
1.00029796e-01 7.95681119e-01 2.65316993e-01 3.06951609e-02
2.72350907e-01 -7.60351121e-01 -7.87964523e-01 -8.35732758e-01
2.40238652e-01 1.49763033e-01 -9.46638584e-02 -3.01088065e-01
-3.43232870e-01 1.72949731e-01 -1.60134017e+00 2.82832146e-01
8.10865760e-01 8.64084482e-01 -8.16791970e-03 8.57591510e-01
2.48945788e-01 4.63734180e-01 -1.22606730e+00 -5.70633471e-01
4.51888055e-01 -9.55169261e-01 -1.89027995e-01 -3.97530794e-01
4.15826023e-01 -2.24250391e-01 -1.71887398e-01 9.88286793e-01
9.64392662e-01 3.21766995e-02 5.33038139e-01 -1.19886243e+00
-3.82288456e-01 7.88194656e-01 3.05885822e-02 -1.44785270e-01
5.04431605e-01 4.15560901e-01 8.37756693e-01 -6.99339449e-01
3.49849731e-01 9.85643923e-01 8.03581119e-01 7.68895447e-01
-1.39950550e+00 -9.48461533e-01 -6.13764048e-01 -6.25654906e-02
-1.30000019e+00 -6.25108957e-01 1.11185443e+00 -3.24700296e-01
3.21583480e-01 5.47351956e-01 6.03338420e-01 1.59326506e+00
-9.01882257e-03 3.78150374e-01 9.81046140e-01 -4.47457522e-01
1.62326619e-01 1.99120790e-01 -6.30367458e-01 2.18786135e-01
-2.35472202e-01 5.54801285e-01 -7.10790813e-01 4.18192297e-02
6.73165143e-01 -5.81505597e-01 -7.00961471e-01 1.44266188e-01
-1.06716537e+00 7.39324152e-01 2.90293127e-01 5.20747244e-01
-3.52917910e-01 2.74596065e-01 3.44503909e-01 5.23494005e-01
3.37157577e-01 9.07177687e-01 -6.59551769e-02 -2.74953902e-01
-1.27729607e+00 3.41137469e-01 7.06543624e-01 5.74420154e-01
3.07641268e-01 1.01023483e+00 -4.72193986e-01 9.10705686e-01
5.93266413e-02 8.17576826e-01 6.76091731e-01 -8.22564423e-01
3.11281145e-01 -8.50079879e-02 -1.20364636e-01 -9.97179031e-01
-4.34872955e-02 -7.30616689e-01 -8.59631360e-01 4.19348687e-01
4.31171983e-01 -4.60370988e-01 -9.96958971e-01 1.95617819e+00
2.32808486e-01 3.30209196e-01 1.73030689e-01 9.52116787e-01
9.38048780e-01 9.35654759e-01 -2.60708034e-01 -1.31667286e-01
1.06397855e+00 -8.69493067e-01 -1.04678249e+00 4.17718627e-02
-1.21066673e-02 -1.06686437e+00 1.30148852e+00 5.62647641e-01
-1.38496006e+00 -9.81094420e-01 -1.14679098e+00 2.91278720e-01
-4.06898595e-02 1.03260681e-01 4.81210500e-02 1.07654428e+00
-1.01063669e+00 5.77469289e-01 -4.78859216e-01 2.74175406e-01
-1.02270722e-01 4.98083122e-02 -2.85285711e-01 4.30902660e-01
-1.50747812e+00 5.93463480e-01 1.86598346e-01 1.50210271e-02
-1.17990613e+00 -1.01758790e+00 -7.52924383e-01 2.21720114e-01
-5.49440123e-02 -5.61476409e-01 1.28108203e+00 -1.05107188e+00
-1.96768081e+00 2.88746089e-01 2.45722592e-01 -5.51266193e-01
6.60563707e-01 -1.53903291e-01 -7.51065254e-01 1.83488727e-01
-1.60434544e-01 4.76996064e-01 1.47730708e+00 -1.17363071e+00
-2.23732308e-01 2.40867153e-01 -3.66294324e-01 -7.48475119e-02
-2.77398974e-01 -3.28631699e-01 -3.65837701e-02 -1.15081394e+00
-1.04937255e-01 -8.04277539e-01 7.54812285e-02 -5.18351257e-01
-5.71433723e-01 3.59639615e-01 7.20923305e-01 -9.91020024e-01
1.14264405e+00 -2.26309657e+00 -4.70595136e-02 1.53755739e-01
-2.51843661e-01 5.02808750e-01 -2.52798021e-01 4.45833921e-01
-3.27323437e-01 2.25566611e-01 -2.23333508e-01 -4.32872385e-01
-3.63985747e-02 -2.70771652e-01 -5.97646832e-01 1.24417298e-01
2.70309657e-01 6.47270620e-01 -5.86731672e-01 -1.24151148e-01
2.72848636e-01 7.95522809e-01 -6.98968351e-01 7.19443381e-01
-9.19919834e-02 8.75408530e-01 4.10906896e-02 4.32896167e-01
3.67790937e-01 5.65207422e-01 -1.70541584e-01 -4.01502848e-01
2.20247164e-01 3.48905981e-01 -1.21404982e+00 1.69104362e+00
-8.16865265e-01 6.10866725e-01 2.11523399e-01 -5.81835032e-01
1.25038302e+00 8.60243499e-01 2.17877522e-01 -6.18910789e-01
1.15070648e-01 3.44940335e-01 1.60860881e-01 -1.97198480e-01
3.34618151e-01 -3.35099220e-01 1.67217292e-02 2.39883929e-01
2.45659113e-01 -8.39693904e-01 -2.35885203e-01 -1.57945856e-01
8.08287621e-01 5.02389446e-02 -1.91132203e-01 4.17835601e-02
3.86449099e-01 -4.39280987e-01 4.75104660e-01 3.91942829e-01
1.85629100e-01 1.15068984e+00 3.12008083e-01 1.95838600e-01
-1.30277658e+00 -1.04820395e+00 2.82593817e-02 5.74177742e-01
-4.60023373e-01 -3.29952687e-01 -1.16755331e+00 -2.65420049e-01
-3.49449068e-01 9.44515765e-01 -4.26247776e-01 -5.08050382e-01
-5.24627686e-01 -3.07531923e-01 1.08732247e+00 1.60246208e-01
2.99487710e-01 -1.17244351e+00 -1.62938103e-01 4.15780455e-01
-3.18669051e-01 -9.38470721e-01 -7.26909459e-01 7.69592151e-02
-5.39389610e-01 -5.89223266e-01 -1.03426933e+00 -5.48818588e-01
3.01552683e-01 -4.60919142e-01 9.46914732e-01 -2.54180163e-01
-1.63886305e-02 8.96835923e-02 -4.43081081e-01 -5.11046469e-01
-1.01798189e+00 1.62639245e-01 2.11206540e-01 1.85044140e-01
-4.91251290e-01 -1.00123370e+00 -5.65224588e-01 3.71705025e-01
-8.57109547e-01 -1.60206378e-01 1.76885575e-01 9.15719450e-01
4.30510044e-01 1.92975387e-01 1.05654109e+00 -4.55843717e-01
9.27686870e-01 -2.78286010e-01 -5.90955853e-01 -1.52703628e-01
-5.23208559e-01 -1.50186270e-01 1.12158751e+00 -7.11171269e-01
-9.21471119e-01 -1.59804806e-01 -6.97345912e-01 -6.66234672e-01
-6.42357916e-02 1.59497440e-01 -3.93669158e-01 1.67569399e-01
9.72389400e-01 2.36012265e-01 8.15522578e-03 -3.64275515e-01
2.94288039e-01 7.63364792e-01 9.32071924e-01 -4.04016167e-01
1.14453435e+00 -3.52624282e-02 -2.73979485e-01 -7.53843606e-01
-4.49755102e-01 2.29107752e-01 4.60399985e-02 -2.11050093e-01
7.00235605e-01 -9.90889132e-01 -5.04674196e-01 5.82627296e-01
-1.05930698e+00 -5.88507473e-01 -6.79224133e-01 6.51999116e-01
-6.45055950e-01 6.07577860e-02 -5.75325370e-01 -9.82653916e-01
-6.98385894e-01 -1.03168297e+00 8.88778746e-01 2.06095874e-01
-2.64416605e-01 -7.79858291e-01 2.37923905e-01 3.50060731e-01
7.24405527e-01 6.35659516e-01 6.51465654e-01 -4.38001364e-01
-2.48125449e-01 -2.36008346e-01 5.69286883e-01 8.27517986e-01
1.70247063e-01 9.55753550e-02 -1.45864344e+00 -2.77998656e-01
3.07148129e-01 -2.83053637e-01 5.45140445e-01 3.25338930e-01
8.86562347e-01 -4.80204344e-01 3.31377834e-01 7.18347967e-01
1.07018685e+00 2.97631145e-01 8.49795818e-01 -1.92271560e-01
5.09170771e-01 4.99434561e-01 4.44483578e-01 2.10041746e-01
-3.00516605e-01 8.76167953e-01 3.46429497e-01 -2.89005548e-01
-6.79341614e-01 -7.48521447e-01 5.94845414e-01 9.86151934e-01
6.24310188e-02 -5.37910879e-01 -5.55295944e-01 4.91862625e-01
-1.17982757e+00 -1.09588671e+00 -9.28321630e-02 2.38273883e+00
1.25617695e+00 1.75384402e-01 2.56170601e-01 6.96281791e-01
8.40442836e-01 1.19001701e-01 -3.75118047e-01 -5.58145761e-01
-1.29370630e-01 6.38820529e-01 1.75543770e-01 5.20169020e-01
-5.85113406e-01 7.05885410e-01 6.07946825e+00 1.12231493e+00
-1.64875352e+00 3.09221745e-01 5.26033223e-01 -2.26500884e-01
-6.33224189e-01 -4.56637025e-01 -3.67252052e-01 6.98115349e-01
1.30019343e+00 -9.22683403e-02 6.54216170e-01 7.68050253e-01
4.78266627e-01 2.85806775e-01 -8.47367525e-01 1.02119005e+00
1.65373203e-03 -1.13590229e+00 -1.22969225e-01 -1.63856536e-01
6.80750430e-01 -4.50155765e-01 3.66706043e-01 3.40131074e-01
-6.21434003e-02 -1.31574965e+00 1.16232264e+00 4.41103339e-01
1.31415379e+00 -9.63186443e-01 7.01084793e-01 2.26256639e-01
-1.05722427e+00 2.73177445e-01 1.64246932e-02 2.94216752e-01
4.17375863e-01 7.97481656e-01 -1.20298564e+00 4.55668747e-01
3.97718310e-01 -5.14315851e-02 -1.77512541e-01 9.12159443e-01
-6.46035850e-01 1.11532795e+00 -1.79211408e-01 3.08252931e-01
-5.50942868e-02 3.63707356e-03 8.52566004e-01 9.64716613e-01
5.54245412e-01 -2.66694456e-01 -3.59566271e-01 1.13687146e+00
-2.03552693e-01 -2.98934132e-02 -6.06539607e-01 -1.73117474e-01
4.81254458e-01 1.13121963e+00 -1.86944813e-01 9.27104130e-02
2.81115413e-01 9.10803974e-01 -2.97532260e-01 3.50209624e-01
-1.18669844e+00 -5.59496403e-01 4.88567531e-01 2.83915341e-01
1.85660169e-01 -4.06051315e-02 -1.49255157e-01 -8.05974543e-01
2.88457721e-02 -1.23228943e+00 -9.89020616e-02 -8.99115562e-01
-1.00444400e+00 1.09274554e+00 -4.70464408e-01 -1.43089390e+00
-7.20015526e-01 -5.07548265e-02 -8.24565291e-01 1.08650446e+00
-1.18164146e+00 -1.06999195e+00 -2.21464455e-01 6.14600062e-01
5.66617846e-01 -3.00502688e-01 1.03114974e+00 5.25708020e-01
-1.64196104e-01 1.01806164e+00 -9.24374014e-02 5.30938357e-02
8.87938082e-01 -1.14681327e+00 5.23402333e-01 9.50089753e-01
2.65707403e-01 1.99411243e-01 1.01253855e+00 -3.62965435e-01
-1.08706224e+00 -1.21698713e+00 5.43972731e-01 -2.03487813e-01
2.18504503e-01 -5.19717336e-01 -8.84080648e-01 9.63150039e-02
3.83970737e-01 -1.76048502e-01 5.87544918e-01 -3.18107873e-01
-2.00422794e-01 -2.97012627e-01 -1.41350257e+00 3.89641613e-01
4.97392535e-01 -6.35499477e-01 -5.16339183e-01 1.71326641e-02
9.47530508e-01 -5.61777771e-01 -8.77531230e-01 3.00908506e-01
5.31208038e-01 -1.12670898e+00 7.68926799e-01 -2.87957311e-01
4.79128689e-01 -3.79293203e-01 -1.31578922e-01 -1.74989069e+00
2.53270119e-02 -1.09983003e+00 4.55892459e-02 1.79533482e+00
6.72477305e-01 -5.59689105e-01 5.36722183e-01 2.19287470e-01
-2.22342730e-01 -3.16831470e-01 -8.11306596e-01 -1.08194172e+00
3.69843002e-03 -5.23392856e-01 7.78457046e-01 7.80039012e-01
-1.71213940e-01 4.20309126e-01 -7.57869780e-01 2.71260500e-01
3.66324991e-01 -1.59695983e-01 9.67802465e-01 -8.17160308e-01
-5.13857961e-01 -2.01267809e-01 -1.76957771e-01 -4.14695412e-01
-4.22821678e-02 -6.57913625e-01 2.38524839e-01 -1.06563580e+00
-6.61473453e-01 -4.36872631e-01 -7.82836229e-02 3.04277301e-01
-6.18770458e-02 5.08566022e-01 4.65723425e-01 -8.03745016e-02
2.99021810e-01 7.42367625e-01 1.30072463e+00 -1.20877825e-01
-3.13266575e-01 3.67507249e-01 -4.12609458e-01 4.61359739e-01
9.41297054e-01 -7.15845287e-01 -4.63520557e-01 -8.88116956e-02
-1.39901534e-01 4.79516149e-01 2.89131910e-01 -1.47124434e+00
-9.75518301e-02 2.04942092e-01 1.47858799e-01 -1.30789161e-01
6.83524489e-01 -7.12465167e-01 7.34611988e-01 4.49480563e-01
-2.95374870e-01 -2.51562893e-01 2.83433586e-01 1.63439140e-01
-5.94191849e-01 -1.51378065e-01 1.13632822e+00 1.53334066e-01
1.88505948e-01 6.91478923e-02 -1.20036572e-01 1.90593243e-01
6.32705152e-01 -4.84908186e-02 -4.44084546e-03 -8.89686882e-01
-6.53748274e-01 -4.73395616e-01 2.91187227e-01 3.48474413e-01
3.69211167e-01 -1.51566756e+00 -8.98083866e-01 4.28137213e-01
-2.25053519e-01 -6.58868477e-02 3.77264500e-01 6.52650595e-01
-2.74375468e-01 1.69988781e-01 -1.87670663e-01 -4.41640764e-01
-1.02671015e+00 3.71673375e-01 5.71900964e-01 -8.89272615e-02
-3.14568520e-01 7.68168569e-01 -8.69192332e-02 -2.24306732e-01
2.89229244e-01 -2.94685245e-01 8.73720180e-03 1.00490414e-01
4.75941658e-01 2.39566937e-01 3.97688210e-01 -3.69067818e-01
-1.33427233e-01 2.76225477e-01 6.77960992e-01 -5.33054352e-01
1.04543853e+00 2.47137994e-01 3.38936538e-01 4.48737442e-01
1.03203666e+00 7.77393818e-01 -1.14428425e+00 2.96791703e-01
-6.07014239e-01 -4.23187524e-01 1.16960011e-01 -9.37868536e-01
-1.43918133e+00 8.73822808e-01 6.88273132e-01 4.55230653e-01
1.21319330e+00 -4.27866578e-01 1.04727077e+00 -2.65523732e-01
2.67082423e-01 -8.44714224e-01 1.85945839e-01 1.88582793e-01
1.25973129e+00 -7.87352622e-01 -5.74626625e-01 -1.64066359e-01
-6.70356512e-01 9.84817505e-01 3.65928680e-01 -1.05042197e-01
3.50624382e-01 4.99478877e-01 4.30631757e-01 3.74274403e-01
-4.50397193e-01 3.73584516e-02 4.24975753e-01 8.12448859e-01
4.28724796e-01 1.14299601e-03 -1.29029155e-01 7.08326817e-01
-9.11638081e-01 -3.60145658e-01 4.71355408e-01 1.70193121e-01
-6.70649186e-02 -1.18262696e+00 -7.10243583e-01 9.50919986e-02
-7.05279410e-01 -2.77753025e-01 -2.34729990e-01 5.38753867e-01
1.63574621e-01 1.13493955e+00 -1.71446487e-01 -6.36484981e-01
5.19738078e-01 2.16435641e-01 2.18541384e-01 -4.46024716e-01
-1.00435221e+00 3.39358985e-01 2.38468096e-01 -4.06456202e-01
5.45249786e-05 -4.01463121e-01 -1.07741010e+00 -2.49906376e-01
-5.04952371e-01 4.79430139e-01 8.77129078e-01 4.25499767e-01
2.25994036e-01 9.60494220e-01 1.03108108e+00 -7.20700145e-01
-6.37290180e-01 -1.17339540e+00 -6.77289367e-01 4.72214192e-01
5.45009732e-01 -2.32426167e-01 -6.82994008e-01 1.95853531e-01] | [15.45550537109375, 6.07197904586792] |
fefc8a2d-2796-4f0c-b011-a4b90b1f349d | meeqa-natural-questions-in-meeting | 2305.08502 | null | https://arxiv.org/abs/2305.08502v1 | https://arxiv.org/pdf/2305.08502v1.pdf | MeeQA: Natural Questions in Meeting Transcripts | We present MeeQA, a dataset for natural-language question answering over meeting transcripts. It includes real questions asked during meetings by its participants. The dataset contains 48K question-answer pairs, extracted from 422 meeting transcripts, spanning multiple domains. Questions in transcripts pose a special challenge as they are not always clear, and considerable context may be required in order to provide an answer. Further, many questions asked during meetings are left unanswered. To improve baseline model performance on this type of questions, we also propose a novel loss function, \emph{Flat Hierarchical Loss}, designed to enhance performance over questions with no answer in the text. Our experiments demonstrate the advantage of using our approach over standard QA models. | ['Eyal Kolman', 'Amir Kantor', 'Tom Braude', 'Reut Apel'] | 2023-05-15 | null | null | null | null | ['natural-questions'] | ['miscellaneous'] | [ 2.67932475e-01 5.09399652e-01 4.10616279e-01 -7.26293087e-01
-1.90694356e+00 -9.23364103e-01 3.13996822e-01 1.56008422e-01
-2.84666270e-01 9.23104048e-01 7.77829707e-01 -3.04710239e-01
-2.10383475e-01 -3.34843099e-01 -4.16773558e-01 -2.04680279e-01
3.53966296e-01 7.26919532e-01 3.48929048e-01 -5.64620554e-01
-2.55590808e-02 -2.95614332e-01 -9.23016667e-01 9.10642982e-01
1.01395977e+00 9.00609255e-01 2.01250657e-01 7.98150897e-01
-5.64635932e-01 1.04105651e+00 -1.11398399e+00 -7.90609777e-01
-9.11647528e-02 -7.49925494e-01 -1.68506742e+00 3.17307934e-02
8.53780150e-01 -2.27719977e-01 -2.52215177e-01 5.60655475e-01
5.68248093e-01 1.76368564e-01 6.51706100e-01 -9.35374022e-01
-4.83169228e-01 6.31701469e-01 -3.28668989e-02 2.61927962e-01
1.07435691e+00 -5.99931963e-02 1.56339800e+00 -9.08636034e-01
2.75130033e-01 1.44383705e+00 4.61780012e-01 8.53781581e-01
-9.92182076e-01 -5.58282018e-01 1.29504994e-01 3.42996001e-01
-1.14127922e+00 -5.83704591e-01 6.64221644e-01 -1.15474463e-01
7.11201966e-01 5.97631037e-01 -5.35186417e-02 1.07936275e+00
8.59154016e-02 8.65381658e-01 1.01968777e+00 -3.86280030e-01
-1.08425260e-01 7.24689886e-02 4.78520215e-01 4.25074428e-01
-3.15237582e-01 -7.40079284e-01 -7.99553037e-01 -5.63259184e-01
1.60918742e-01 -5.10915995e-01 -7.68333673e-01 1.89282879e-01
-1.01378322e+00 7.41957128e-01 1.73613504e-01 2.59065509e-01
-1.82368398e-01 -8.51127729e-02 1.99041694e-01 7.89706528e-01
1.55915871e-01 6.89337432e-01 -8.41450989e-01 -3.43243778e-01
-7.30684102e-01 7.14974344e-01 1.34917665e+00 8.53253007e-01
6.75661266e-01 -7.06852138e-01 -7.04833329e-01 1.13227344e+00
1.20870143e-01 4.11888272e-01 2.16017634e-01 -1.50740373e+00
1.03703701e+00 6.21263683e-01 4.41021949e-01 -9.80622590e-01
-3.80100310e-01 -1.12543784e-01 -6.41167402e-01 -9.18740869e-01
7.89018452e-01 -4.13297862e-01 -4.13127214e-01 1.71402192e+00
2.81666368e-01 -2.70948857e-01 1.85351208e-01 6.87715590e-01
1.44714880e+00 7.38322854e-01 -2.26991162e-01 -1.53591409e-01
1.73381662e+00 -1.08907878e+00 -1.30862641e+00 -4.30850983e-01
4.24661726e-01 -9.25788879e-01 1.42877305e+00 2.06616268e-01
-1.26511097e+00 -5.20274282e-01 -2.72614092e-01 -4.10307884e-01
1.70424655e-01 -4.77485321e-02 -4.10109721e-02 3.37827384e-01
-1.05226719e+00 5.14753498e-02 -1.32185087e-01 -1.62509471e-01
-9.42289457e-02 2.55723506e-01 -2.72012558e-02 -3.36502433e-01
-1.68125153e+00 5.27079105e-01 -3.03592950e-01 1.93614095e-01
-5.19769967e-01 -5.21337807e-01 -9.68650520e-01 2.67505378e-01
5.53215683e-01 -7.01082230e-01 2.03759027e+00 -3.42466682e-01
-1.43150675e+00 8.22568834e-01 -8.18136632e-01 -2.15887755e-01
2.92823255e-01 -3.81968081e-01 -4.03861880e-01 4.64404076e-01
3.45591664e-01 6.13000035e-01 6.51589155e-01 -1.09879100e+00
-4.81292844e-01 -2.60834754e-01 7.10380375e-01 3.31336826e-01
-2.22727373e-01 4.42686230e-02 -4.49859887e-01 -3.84047836e-01
1.87394500e-01 -7.34540343e-01 -7.61038363e-02 -3.87343824e-01
-2.54448831e-01 -7.29461014e-01 6.42149329e-01 -1.23270047e+00
1.46934867e+00 -1.75883090e+00 -2.47282580e-01 -1.33056358e-01
1.36509314e-01 3.23463418e-02 -2.82534659e-01 8.72613251e-01
4.16232586e-01 4.84559238e-02 -4.28452820e-01 -9.00450572e-02
2.19632894e-01 2.35873178e-01 -5.79596639e-01 -3.37029129e-01
3.55882525e-01 1.11814761e+00 -7.86942780e-01 -7.20844328e-01
-5.06925404e-01 1.12586757e-02 -5.42162240e-01 5.99758208e-01
-5.57745337e-01 6.28248632e-01 -5.52135646e-01 5.71316957e-01
4.58807647e-01 -6.53446317e-01 1.68929532e-01 1.42692626e-01
6.45409167e-01 1.06698990e+00 -1.02106929e+00 1.64176226e+00
-5.44695079e-01 5.63049912e-01 6.00736439e-01 -7.35519230e-01
9.83447552e-01 6.99748278e-01 4.89719808e-02 -5.54888964e-01
5.43667376e-03 1.77481219e-01 -7.57350028e-02 -8.84956479e-01
5.75863004e-01 -2.32521221e-01 -3.91667932e-01 3.99396807e-01
-1.05296932e-01 -5.98084450e-01 1.76443562e-01 3.79926056e-01
1.47943342e+00 -3.74777436e-01 -2.60500312e-02 -1.49386361e-01
1.07066369e+00 2.62495759e-03 4.99179691e-01 8.61761093e-01
-4.35213178e-01 7.51542389e-01 6.08488083e-01 5.27496710e-02
-2.64713138e-01 -9.37300980e-01 -3.57043706e-02 1.20447528e+00
-4.08006348e-02 -5.28590977e-01 -8.62617433e-01 -1.00865650e+00
-3.31582487e-01 4.47858095e-01 -2.26298168e-01 2.09776819e-01
-9.06781971e-01 -7.11179078e-02 6.25272274e-01 3.04374158e-01
6.80817485e-01 -1.12246501e+00 -3.32359523e-02 4.71863061e-01
-1.46213603e+00 -1.61328697e+00 -1.06660235e+00 -1.57934889e-01
-7.92257011e-01 -1.20133197e+00 -9.75478470e-01 -1.16968918e+00
3.57989222e-01 2.30434150e-01 1.85252881e+00 3.38662229e-02
3.37436795e-01 9.31538284e-01 -5.02407432e-01 -2.84618646e-01
-4.07904238e-01 3.65378052e-01 -4.34003085e-01 -2.25977376e-02
6.51492000e-01 -2.61450976e-01 -6.64295375e-01 5.53645909e-01
-7.88708150e-01 -6.54741943e-01 1.33263528e-01 9.06504512e-01
1.22456893e-01 -3.01235527e-01 1.14232659e+00 -8.99010658e-01
1.26797032e+00 -3.75382513e-01 -1.72411669e-02 4.31914210e-01
1.45120427e-01 -1.34885043e-01 4.59189951e-01 -2.23462403e-01
-1.15706158e+00 -3.37731600e-01 -6.27419770e-01 3.72533172e-01
-8.72263536e-02 4.39405680e-01 -4.08357084e-01 1.86337098e-01
5.99314153e-01 8.23527277e-02 -8.75948966e-02 -5.59857965e-01
7.96831474e-02 9.95943367e-01 4.78402555e-01 -6.05536520e-01
6.22807205e-01 5.64328358e-02 -6.12283707e-01 -9.08022344e-01
-1.45337498e+00 -9.85444188e-01 -3.26072007e-01 -7.37020746e-02
7.21682847e-01 -8.89515698e-01 -9.50807750e-01 -1.66355632e-02
-1.56071234e+00 -2.33129993e-01 -1.31851077e-01 1.48753330e-01
-4.77882713e-01 6.42559826e-01 -8.46783519e-01 -9.45384204e-01
-4.80228275e-01 -7.28249550e-01 1.09473050e+00 2.39215270e-01
-7.50863492e-01 -1.02634132e+00 -2.89086699e-02 1.43236589e+00
2.36107141e-01 -2.42467538e-01 9.43735600e-01 -7.80968428e-01
-4.74415302e-01 -6.92153955e-03 6.95135891e-02 5.08268952e-01
4.31688160e-01 -6.92279637e-01 -1.00000215e+00 -1.63333625e-01
3.39626998e-01 -8.83591235e-01 8.04477215e-01 1.43065825e-01
1.17890763e+00 -5.90142429e-01 1.41358944e-02 -3.08928668e-01
6.75060809e-01 -1.61832333e-01 4.00098085e-01 -9.71561112e-03
2.25679740e-01 1.18715811e+00 7.49814391e-01 1.44944891e-01
1.12043893e+00 5.00631928e-01 1.66590109e-01 2.00032294e-01
6.17607310e-02 -2.01449931e-01 5.02769768e-01 1.36350143e+00
6.43707395e-01 -5.65093875e-01 -8.41339350e-01 9.65071797e-01
-1.60525227e+00 -7.51223087e-01 -5.56073308e-01 1.76566517e+00
1.28229117e+00 -3.97924744e-02 -9.23768654e-02 2.18000576e-01
4.67636108e-01 3.49961281e-01 -3.59604508e-01 -4.24145073e-01
-1.61380693e-01 3.99106026e-01 -3.43417585e-01 8.03471744e-01
-8.59769940e-01 6.12099946e-01 6.74345160e+00 7.14279413e-01
-3.24062854e-01 1.37805939e-01 4.05607343e-01 2.78744817e-01
-6.39582217e-01 -1.71844214e-02 -9.16825831e-01 2.69211143e-01
1.04110420e+00 9.56732035e-03 -6.12430722e-02 3.63975197e-01
2.92359591e-01 -5.28064780e-02 -1.09178090e+00 8.14499557e-01
2.47985899e-01 -1.07863295e+00 -2.94356216e-02 -3.10456008e-01
6.30530238e-01 -4.94896501e-01 -1.75340593e-01 6.53396785e-01
2.88067907e-01 -9.63278651e-01 8.57875496e-02 3.54052603e-01
5.07499456e-01 -4.51074690e-01 9.16455805e-01 7.56896019e-01
-1.11919951e+00 -6.38538301e-02 -2.93738425e-01 -3.21611524e-01
4.21818346e-01 4.56546247e-01 -1.09818292e+00 4.68564481e-01
8.17961097e-01 3.69319580e-02 -4.61807311e-01 7.96886384e-01
-4.12627131e-01 1.00875878e+00 -2.05725104e-01 -2.47528136e-01
4.72606659e-01 8.59594997e-03 3.45331132e-01 1.16155565e+00
5.48337400e-02 5.35129607e-01 2.21033543e-01 5.41530132e-01
-5.00405490e-01 2.34654889e-01 -3.49888504e-01 1.35788456e-01
5.16963243e-01 1.14354837e+00 -4.91631292e-02 -2.14239702e-01
-4.20663774e-01 1.14744282e+00 2.20940158e-01 5.95164359e-01
-4.82689440e-01 -5.32475948e-01 5.16251445e-01 1.88181952e-01
2.31044386e-02 -2.00819939e-01 2.06332788e-01 -1.07395458e+00
5.45366943e-01 -1.58306491e+00 7.68927455e-01 -5.98102987e-01
-1.65562308e+00 4.71593678e-01 -3.24752927e-01 -1.00149703e+00
-5.02212644e-01 -3.60991359e-01 -4.48058039e-01 8.32769275e-01
-1.76504183e+00 -7.29866266e-01 -3.60889256e-01 6.58610702e-01
1.00521505e+00 3.28308344e-01 8.44852865e-01 5.15597522e-01
-2.97893886e-03 8.61728311e-01 -1.61880925e-01 2.65461773e-01
1.15956843e+00 -1.33920336e+00 2.20276088e-01 9.70408395e-02
2.63781756e-01 5.81291378e-01 6.35751665e-01 -2.53303051e-01
-1.20541906e+00 -8.15011501e-01 1.73252237e+00 -1.01941895e+00
4.58971739e-01 -3.74308795e-01 -1.42774785e+00 6.06258869e-01
6.20679796e-01 -6.36916280e-01 1.10644293e+00 1.81219473e-01
-8.17578584e-02 -8.13014135e-02 -1.05822325e+00 4.36700284e-01
6.60960317e-01 -7.50950098e-01 -1.38359702e+00 4.35178071e-01
1.27307200e+00 -4.59788859e-01 -7.33872414e-01 4.99440640e-01
9.43294689e-02 -5.64852655e-01 7.44451582e-01 -5.68683505e-01
2.82482773e-01 1.89738730e-05 -1.26426116e-01 -9.15629566e-01
3.25326920e-02 -9.78550732e-01 -1.74139798e-01 1.51428723e+00
8.33437383e-01 -3.50110739e-01 9.53441262e-01 8.36754262e-01
-1.05687074e-01 -6.62681818e-01 -1.17632473e+00 -5.82730412e-01
2.37645969e-01 -2.74490088e-01 6.23412609e-01 8.44768524e-01
2.13471040e-01 1.05465353e+00 -3.12673777e-01 1.16661698e-01
3.45155746e-01 1.45017639e-01 7.54531980e-01 -1.03764296e+00
-2.38211304e-01 5.75818494e-03 3.77000540e-01 -2.00438690e+00
2.92175502e-01 -6.05466723e-01 3.20698023e-01 -2.11473513e+00
-2.22138558e-02 -1.16528943e-01 8.72477069e-02 2.09108531e-01
-5.94378412e-01 7.16977194e-02 1.35221809e-01 -1.72207376e-03
-9.05412555e-01 7.47345448e-01 1.63063478e+00 -4.06065285e-01
-1.90750752e-02 4.22099799e-01 -9.32790041e-01 7.98864186e-01
6.11119330e-01 -3.94596636e-01 -3.41195256e-01 -6.47505760e-01
1.17465422e-01 6.90884352e-01 9.02509838e-02 -7.29567468e-01
3.25085521e-01 -2.52204053e-02 -1.18004434e-01 -9.54277337e-01
6.91093326e-01 -7.64497221e-01 -6.32408202e-01 1.56397045e-01
-7.14780807e-01 8.33796784e-02 5.07118329e-02 5.06300569e-01
-8.01123202e-01 -4.99503940e-01 2.62736946e-01 -2.61875987e-01
-1.15612499e-01 7.21970424e-02 -4.24399763e-01 9.63117719e-01
2.20357269e-01 6.27779216e-02 -5.82903586e-02 -1.19197309e+00
-7.57473409e-01 1.10925210e+00 -2.29906812e-01 4.70244884e-01
7.10946321e-01 -1.20445085e+00 -1.08659101e+00 -5.01581669e-01
1.53112590e-01 1.21116452e-01 4.86354887e-01 4.95570719e-01
-8.96410346e-02 7.32761264e-01 6.44333363e-01 -4.70719486e-01
-1.40396380e+00 -1.71784639e-01 3.69619608e-01 -7.86604226e-01
-2.58266926e-01 1.17305803e+00 2.92367071e-01 -1.08259523e+00
4.66702819e-01 -4.99875724e-01 -5.92294753e-01 3.55364442e-01
6.87369585e-01 2.50118792e-01 1.54403523e-01 -2.84492373e-01
-3.73075098e-01 5.46699286e-01 -2.28479449e-02 -3.47271353e-01
8.47770751e-01 -7.33255208e-01 -9.45730805e-02 4.56572741e-01
1.05155134e+00 1.90349519e-01 -8.54382217e-01 -8.57921362e-01
3.57240975e-01 -2.30906963e-01 -6.05891943e-01 -1.01073503e+00
-3.70206177e-01 1.02224672e+00 5.86679913e-02 4.53061104e-01
1.05321240e+00 3.87597114e-01 1.42416990e+00 1.05669034e+00
9.75936502e-02 -1.09777808e+00 7.50902176e-01 1.14017475e+00
1.29564643e+00 -1.49364901e+00 -6.11025572e-01 -4.63563591e-01
-7.00885773e-01 9.36637998e-01 7.65568137e-01 4.53971028e-01
2.58896679e-01 -1.86376452e-01 6.57706559e-01 -2.99161315e-01
-9.91214514e-01 -2.02633604e-01 4.46237326e-01 3.49126071e-01
6.93793595e-01 -2.38273799e-01 -4.65405285e-01 9.60840821e-01
-6.41115487e-01 -2.36631423e-01 6.06000721e-01 8.76791775e-01
-6.19199276e-01 -1.17427325e+00 -6.07344389e-01 5.02583027e-01
-9.48897660e-01 -2.04639271e-01 -7.72354782e-01 3.15823764e-01
-3.77027512e-01 1.92014718e+00 -2.70985305e-01 -2.71384809e-02
6.75491452e-01 4.91234303e-01 2.18661144e-01 -9.39058304e-01
-9.88990486e-01 -2.30697393e-01 5.67875862e-01 -1.79146916e-01
-2.78611332e-01 -4.38192457e-01 -1.22560298e+00 8.98653418e-02
-4.29636627e-01 8.46567035e-01 1.63943395e-01 1.02091324e+00
4.19973910e-01 5.28025389e-01 7.01977551e-01 9.70752761e-02
-9.01260793e-01 -1.13785732e+00 -2.27258861e-01 3.09372991e-01
6.74640954e-01 -4.00072373e-02 -4.58446562e-01 1.57439068e-01] | [11.642705917358398, 8.02055835723877] |
8001af62-95d7-4765-bf0a-250b87446769 | deception-detection-in-videos-using-the | 2105.13659 | null | https://arxiv.org/abs/2105.13659v1 | https://arxiv.org/pdf/2105.13659v1.pdf | Deception Detection in Videos using the Facial Action Coding System | Facts are important in decision making in every situation, which is why it is important to catch deceptive information before they are accepted as facts. Deception detection in videos has gained traction in recent times for its various real-life application. In our approach, we extract facial action units using the facial action coding system which we use as parameters for training a deep learning model. We specifically use long short-term memory (LSTM) which we trained using the real-life trial dataset and it provided one of the best facial only approaches to deception detection. We also tested cross-dataset validation using the Real-life trial dataset, the Silesian Deception Dataset, and the Bag-of-lies Deception Dataset which has not yet been attempted by anyone else for a deception detection system. We tested and compared all datasets amongst each other individually and collectively using the same deep learning training model. The results show that adding different datasets for training worsen the accuracy of the model. One of the primary reasons is that the nature of these datasets vastly differs from one another. | ['Muhammad Waqas Anwar', 'Fan Zhang', 'Usama Ijaz Bajwa', 'Hammad Ud Din Ahmed'] | 2021-05-28 | null | null | null | null | ['deception-detection-in-videos', 'deception-detection-in-videos', 'deception-detection'] | ['computer-vision', 'miscellaneous', 'miscellaneous'] | [ 1.11254618e-01 1.11645773e-01 2.36239493e-01 -5.58387995e-01
-3.79795045e-01 -3.49130988e-01 1.01923263e+00 -4.80471492e-01
-6.16750002e-01 9.11285520e-01 1.90965787e-01 -1.79130554e-01
-9.07584503e-02 -4.41179335e-01 -5.28900683e-01 -8.63085330e-01
9.39178318e-02 6.19370081e-02 -1.09604530e-01 -3.12845826e-01
6.99843943e-01 6.85563445e-01 -1.72316897e+00 7.49420524e-01
4.80787486e-01 1.36529541e+00 -6.65372372e-01 7.91929662e-01
2.25629777e-01 1.27938640e+00 -9.98736262e-01 -1.11614931e+00
3.46704692e-01 -6.27159297e-01 -6.82473302e-01 4.77696881e-02
8.69024754e-01 -6.83046639e-01 -5.84976554e-01 9.38968122e-01
4.84503746e-01 -5.09399036e-03 7.63048172e-01 -1.38985765e+00
-7.06619978e-01 1.27993505e-02 -4.41626400e-01 3.71119797e-01
3.91847312e-01 1.61188111e-01 4.25093561e-01 -6.59888625e-01
4.69434470e-01 1.28923357e+00 5.31836033e-01 1.00025523e+00
-5.19713640e-01 -9.54610229e-01 -3.73249203e-01 8.01432431e-01
-1.17595208e+00 -1.07440317e+00 6.91740811e-01 -3.56162280e-01
9.35737550e-01 -4.27833665e-03 7.44183362e-01 1.66430056e+00
5.31279743e-01 7.95800686e-01 1.22886586e+00 -4.05799985e-01
1.47829041e-01 -2.31388360e-02 -1.75836489e-01 8.07577729e-01
1.25366271e-01 4.89272833e-01 -6.97083950e-01 -1.90828182e-02
4.15361464e-01 -2.41862923e-01 -9.02489275e-02 6.63556829e-02
-6.37766004e-01 1.04363370e+00 2.76723504e-01 4.72075790e-01
-2.92867184e-01 3.10463548e-01 5.88334501e-01 6.56087697e-01
4.19067502e-01 1.82938606e-01 -1.39311373e-01 -4.49334502e-01
-1.19755495e+00 3.24156731e-02 7.49364257e-01 3.08686662e-02
1.68663803e-02 5.02974391e-01 2.65083313e-01 7.56091654e-01
6.71397448e-02 3.28041196e-01 8.48582387e-01 -8.47939193e-01
1.71846658e-01 5.28963864e-01 8.15412437e-04 -1.47292316e+00
-2.32861280e-01 -2.08403766e-02 -7.76113272e-01 7.97062457e-01
6.11466527e-01 -2.32545644e-01 -1.18030715e+00 1.28574800e+00
-1.06138147e-01 2.67357886e-01 2.82037675e-01 1.02163100e+00
9.64895666e-01 5.35604537e-01 6.45175055e-02 3.46750766e-02
1.08415794e+00 -5.95296085e-01 -6.26418829e-01 -9.55748409e-02
6.03532612e-01 -5.97842038e-01 3.96893263e-01 8.72129738e-01
-7.99852490e-01 -2.76388377e-01 -1.26300454e+00 -5.87850250e-02
-8.13765347e-01 1.97648495e-01 6.56022131e-01 1.11433268e+00
-8.87667537e-01 5.72326183e-01 -3.93177837e-01 -3.81706983e-01
8.91254783e-01 3.43965024e-01 -8.81382167e-01 -4.35154401e-02
-1.28915024e+00 1.37977731e+00 2.87238538e-01 5.59849739e-01
-1.01550353e+00 -3.32856774e-02 -8.04468930e-01 -8.07504058e-02
2.82044947e-01 -2.07894117e-01 9.95333791e-01 -1.93608749e+00
-1.51656330e+00 1.34381211e+00 2.24907190e-01 -6.86875165e-01
6.72146559e-01 -5.43852746e-02 -1.05097449e+00 3.31002206e-01
-4.04363573e-01 6.35379374e-01 1.62710893e+00 -1.00946832e+00
-4.42289442e-01 -6.48852348e-01 1.70530602e-01 -1.55047238e-01
-1.15873083e-01 2.76958197e-01 6.14619076e-01 -3.50559950e-01
-4.52834666e-01 -6.67419195e-01 7.09053159e-01 5.32878079e-02
2.83898320e-02 -4.24238630e-02 1.10948431e+00 -9.33734953e-01
1.03818703e+00 -2.19299030e+00 1.22927567e-02 -6.82194903e-02
2.23099738e-01 8.83038759e-01 -1.48021534e-01 3.08620751e-01
-3.71330589e-01 2.25574017e-01 -2.54013129e-02 -2.04866618e-01
-1.29479125e-01 4.55275536e-01 -2.13515282e-01 7.84808874e-01
3.32904935e-01 9.39154267e-01 -6.54388964e-01 -4.28655654e-01
2.19796017e-01 5.51646471e-01 7.03182817e-02 -6.76559880e-02
2.25048393e-01 5.16547337e-02 7.63683915e-02 7.06509411e-01
7.49316096e-01 5.99936366e-01 -3.76760483e-01 1.31138638e-01
2.14585692e-01 -1.57884687e-01 -6.25154436e-01 1.24287844e+00
-2.46780321e-01 1.16305828e+00 -1.07965082e-01 -1.10791147e+00
8.81298065e-01 6.18562639e-01 -9.88782868e-02 -8.75955939e-01
5.16386330e-01 4.22492325e-01 3.91915530e-01 -9.12875116e-01
3.29106688e-01 -6.92046285e-01 9.32735652e-02 1.86271891e-01
3.70837420e-01 3.28730159e-02 -2.23350283e-02 -1.05884308e-02
8.22236121e-01 2.35758945e-02 4.36531097e-01 7.69043639e-02
9.04354274e-01 -2.61090398e-01 1.83736160e-01 4.15870428e-01
-6.57793581e-01 6.20353639e-01 9.13461030e-01 -9.06395018e-01
-6.95943236e-01 -5.93438447e-01 -2.07931697e-02 7.43469715e-01
-4.90724474e-01 1.08523056e-01 -6.54345930e-01 -1.18347502e+00
7.58229345e-02 1.01619780e+00 -1.14090621e+00 -5.63265383e-01
-3.28052014e-01 -4.40372735e-01 1.07171834e+00 6.66435212e-02
8.42371106e-01 -1.37794662e+00 -9.70035136e-01 -2.09961161e-01
2.54229158e-01 -8.90859425e-01 1.33064225e-01 -1.59640610e-01
-4.11332756e-01 -1.45344317e+00 -7.06097007e-01 -2.97527730e-01
3.93838853e-01 -9.95678827e-02 7.95176983e-01 3.80830199e-01
-2.33049139e-01 2.90542334e-01 -3.89635056e-01 -7.83702850e-01
-6.89489841e-01 -3.00164878e-01 3.09618637e-02 5.40500104e-01
7.41979659e-01 -2.95447350e-01 -3.70662153e-01 3.51557583e-02
-1.01789272e+00 -4.47817057e-01 5.28094649e-01 8.74764562e-01
-5.15133440e-01 8.47058650e-03 7.12382495e-01 -5.48934162e-01
7.61860251e-01 -4.71827239e-01 -3.67430538e-01 8.78075734e-02
8.21360480e-03 -2.47024134e-01 4.90120262e-01 -1.64972529e-01
-1.00858748e+00 -4.34931964e-01 -2.42727757e-01 -5.28108001e-01
-5.01942158e-01 2.76751578e-01 -3.42147760e-02 -4.94008392e-01
4.99144256e-01 2.73253769e-01 2.40977958e-01 -7.92795867e-02
-7.35335872e-02 1.02219868e+00 4.25526768e-01 1.10311367e-01
3.53701353e-01 3.69169742e-01 2.99020857e-01 -1.01318836e+00
-8.12911451e-01 -1.14307813e-01 -8.67018878e-01 -5.21216094e-01
6.33690774e-01 -4.54776496e-01 -1.06765413e+00 1.14497733e+00
-1.18077290e+00 -1.05438521e-02 4.05034982e-02 1.92595631e-01
-3.57882380e-01 2.90934175e-01 -2.61462033e-01 -1.26239002e+00
-1.31315738e-01 -7.28758514e-01 8.21720362e-01 2.82908455e-02
-5.54547124e-02 -1.08004546e+00 -2.06374034e-01 6.14557743e-01
4.85889077e-01 6.58755124e-01 6.97944164e-01 -7.65354872e-01
1.53550759e-01 -5.74133217e-01 -3.43370020e-01 8.99250627e-01
8.02550390e-02 2.70620316e-01 -1.25126505e+00 3.62331495e-02
3.65787596e-01 -8.59387636e-01 1.11446571e+00 9.59104300e-02
9.57180321e-01 -4.69176114e-01 2.12061003e-01 2.66585976e-01
1.14632773e+00 3.00526500e-01 1.22939932e+00 2.74532974e-01
4.59932894e-01 6.59227729e-01 2.59727716e-01 1.58656031e-01
3.96880060e-02 5.50098181e-01 7.24750757e-01 1.34694412e-01
-5.64027354e-02 -2.15394069e-02 7.17024148e-01 9.46493298e-02
-2.84894735e-01 -3.31209242e-01 -8.26873660e-01 5.13702512e-01
-1.59822643e+00 -1.56065202e+00 -9.01293680e-02 2.05061197e+00
4.72232342e-01 -1.14955800e-02 1.91470414e-01 4.20567244e-01
2.28166103e-01 3.05594027e-01 -1.93483517e-01 -1.26205730e+00
-8.57337937e-02 1.34289652e-01 1.22241318e-01 4.19408411e-01
-1.18040359e+00 8.69515777e-01 6.07174683e+00 8.42865348e-01
-1.56766546e+00 1.24516021e-02 8.68008316e-01 -4.15311247e-01
5.23838162e-01 -6.18436396e-01 -4.28712338e-01 7.07397759e-01
1.43167269e+00 1.75958559e-01 2.83705950e-01 5.55104911e-01
3.48475158e-01 -6.13181770e-01 -9.64879453e-01 1.21615374e+00
8.47079635e-01 -1.00067973e+00 2.07704604e-01 3.18475477e-02
3.97324026e-01 -4.68307227e-01 1.29435718e-01 3.29662830e-01
-1.59283832e-01 -1.64151216e+00 7.48237491e-01 7.25574851e-01
7.37111032e-01 -9.96323347e-01 1.22247875e+00 3.19720656e-01
-6.29920140e-02 -2.41458923e-01 -4.33148444e-01 -3.86455983e-01
-6.94401562e-02 3.31241518e-01 -5.09650886e-01 3.53949994e-01
3.49144518e-01 6.24433517e-01 -7.04897821e-01 7.30532289e-01
-3.57626140e-01 5.84720612e-01 -8.24595541e-02 -4.82722968e-02
5.35467565e-01 4.98080347e-03 4.50032562e-01 1.04776883e+00
2.06895813e-01 -8.71651098e-02 -6.34234369e-01 4.87039864e-01
-6.59759417e-02 -3.69506121e-01 -8.12561750e-01 -2.70799547e-01
-3.09621781e-01 1.14365482e+00 -2.32487425e-01 -2.03583807e-01
-5.39420843e-01 1.34525168e+00 1.14496715e-01 1.05644189e-01
-8.59137058e-01 -3.09047222e-01 6.45645916e-01 1.54329076e-01
1.78454071e-01 3.79205830e-02 6.22368651e-03 -1.21651423e+00
-8.43786970e-02 -1.03400862e+00 3.79590631e-01 -9.41081762e-01
-1.24158967e+00 6.58781469e-01 7.21040070e-02 -7.62052178e-01
-5.03758550e-01 -1.01850343e+00 -4.68315482e-01 7.16677129e-01
-1.52010572e+00 -1.39039958e+00 -2.90383190e-01 7.26543367e-01
3.42027098e-01 -5.33898056e-01 8.95648420e-01 1.68121055e-01
-5.08419812e-01 6.19739294e-01 -1.73672482e-01 4.37189937e-01
6.85131490e-01 -8.37147117e-01 9.50192586e-02 8.71813536e-01
2.45363563e-01 2.71109343e-01 4.60912615e-01 -3.85574907e-01
-9.95401859e-01 -5.61632633e-01 1.04299867e+00 -5.14592111e-01
7.05899835e-01 -4.30314720e-01 -7.95410156e-01 6.20933175e-01
3.85071337e-01 5.98225892e-02 7.21346736e-01 -4.21400666e-01
-4.25181001e-01 -1.13557473e-01 -1.69611895e+00 2.44372129e-01
7.37739384e-01 -4.61023301e-01 -8.12901378e-01 -5.64200319e-02
-2.77127564e-01 -1.62290931e-01 -3.62158090e-01 -1.08233951e-01
9.87299740e-01 -1.54646409e+00 5.86374521e-01 -9.15357232e-01
9.04865265e-01 1.25962347e-01 2.74241921e-02 -1.48691571e+00
1.05981484e-01 -4.19779480e-01 -2.88763732e-01 1.13364661e+00
1.28532439e-01 -9.41807747e-01 8.11932981e-01 5.97503781e-01
2.87178755e-01 -6.48086488e-01 -1.33725321e+00 -7.53145337e-01
2.88642317e-01 -1.93728015e-01 2.86139339e-01 8.64411652e-01
-1.04844995e-01 1.41413972e-01 -7.57154047e-01 -4.62173194e-01
4.08003628e-01 -1.81676418e-01 6.45655334e-01 -1.11787522e+00
1.63749591e-01 -3.81369054e-01 -1.22706258e+00 -2.99277395e-01
4.56817240e-01 -5.67048192e-01 -3.30891758e-01 -1.09675539e+00
6.19549304e-02 1.78430721e-01 -7.73932040e-02 8.09930027e-01
1.80108801e-01 5.84705532e-01 2.55562752e-01 -1.93493500e-01
-1.47512943e-01 5.56082606e-01 9.96262372e-01 1.85119994e-02
3.20177138e-01 -2.60377433e-02 -6.92811728e-01 7.86028326e-01
8.61913204e-01 -4.15604770e-01 -1.82885155e-01 -3.13298315e-01
2.08385885e-01 -1.25219136e-01 9.63857889e-01 -1.11130154e+00
4.39056708e-03 1.76001471e-02 8.89889896e-01 -9.71309096e-02
6.16111517e-01 -9.50852275e-01 -1.90058067e-01 3.47826689e-01
-4.79456931e-02 -2.62879401e-01 4.09191072e-01 2.87443638e-01
-3.44991595e-01 -5.49066722e-01 9.93743420e-01 -3.21435809e-01
-9.91321564e-01 -5.25152497e-02 -5.49162686e-01 -3.23199719e-01
1.09850538e+00 -6.84224367e-01 -3.98337871e-01 -8.47988307e-01
-6.11433923e-01 -1.68373048e-01 3.09334904e-01 6.04022384e-01
7.54372537e-01 -1.14808798e+00 -9.96311307e-01 3.61457467e-01
3.08161844e-02 -9.05940890e-01 1.04817696e-01 8.62800717e-01
-5.40840924e-01 5.73285878e-01 -8.32032800e-01 -8.49136338e-02
-1.54895329e+00 4.46777970e-01 7.47542620e-01 2.52316356e-01
-3.14483583e-01 8.25419605e-01 -2.56771266e-01 2.04710010e-02
4.71494952e-03 1.12896912e-01 -4.00066048e-01 2.92930245e-01
7.82022476e-01 5.45930982e-01 3.20060223e-01 -1.23151588e+00
-4.30814773e-01 2.51068532e-01 -3.98829319e-02 -2.83260667e-03
1.42470121e+00 1.88560352e-01 -1.12484679e-01 3.85632426e-01
1.27097666e+00 -1.23610064e-01 -9.89278495e-01 5.33848345e-01
-2.02138335e-01 -9.07189548e-01 3.60534936e-01 -1.22715378e+00
-1.35265970e+00 1.14696670e+00 5.36985338e-01 2.83212841e-01
1.12756777e+00 -5.16142666e-01 5.05494297e-01 2.11339355e-01
3.33092690e-01 -9.51751232e-01 -4.85495478e-02 3.64980727e-01
1.34319627e+00 -1.41418481e+00 1.18099056e-01 1.44786641e-01
-7.27669239e-01 1.49831748e+00 3.73734027e-01 -3.03336382e-01
3.37554395e-01 -1.50819167e-01 1.54009283e-01 -3.45020771e-01
-5.80134630e-01 -1.13527523e-02 2.21159384e-01 7.86834836e-01
1.68553591e-01 -1.66467622e-01 -3.72859210e-01 3.94216299e-01
-2.69361436e-01 4.49986070e-01 6.97110355e-01 6.56985164e-01
-1.26607522e-01 -9.48634684e-01 -3.83041710e-01 5.41409552e-01
-9.36342657e-01 6.24257699e-02 -1.14816570e+00 1.06160092e+00
2.86835223e-01 1.13406777e+00 -1.13844700e-01 -4.46142614e-01
6.22444553e-03 4.78310794e-01 7.89081991e-01 -1.34837255e-01
-8.15060675e-01 -8.19210112e-01 4.81021911e-01 -5.91542125e-01
-4.48316157e-01 -7.44973719e-01 -7.86024094e-01 -8.19043279e-01
-2.35994816e-01 -1.84247032e-01 7.43220329e-01 1.09351623e+00
2.86503196e-01 1.06181994e-01 3.24233025e-01 -8.84147465e-01
-4.73749667e-01 -1.15045941e+00 -5.96061885e-01 5.45738339e-01
6.13318563e-01 -8.26636374e-01 -6.79529130e-01 3.86269949e-02] | [13.093595504760742, 1.904586911201477] |
c6c50372-fd41-42f6-85f4-f170ca172ddc | conversational-information-seeking | 2201.08808 | null | https://arxiv.org/abs/2201.08808v2 | https://arxiv.org/pdf/2201.08808v2.pdf | Conversational Information Seeking | Conversational information seeking (CIS) is concerned with a sequence of interactions between one or more users and an information system. Interactions in CIS are primarily based on natural language dialogue, while they may include other types of interactions, such as click, touch, and body gestures. This monograph provides a thorough overview of CIS definitions, applications, interactions, interfaces, design, implementation, and evaluation. This monograph views CIS applications as including conversational search, conversational question answering, and conversational recommendation. Our aim is to provide an overview of past research related to CIS, introduce the current state-of-the-art in CIS, highlight the challenges still being faced in the community. and suggest future directions. | ['Filip Radlinski', 'Jeff Dalton', 'Johanne R. Trippas', 'Hamed Zamani'] | 2022-01-21 | null | null | null | null | ['conversational-search'] | ['natural-language-processing'] | [ 9.91870165e-02 1.54611722e-01 -3.97031307e-01 -4.00646001e-01
-4.03733611e-01 -1.00159752e+00 1.06669939e+00 2.24717140e-01
-4.12013173e-01 3.30127776e-01 4.94988769e-01 -6.97713494e-01
-3.12862456e-01 -1.20588213e-01 2.59161055e-01 -1.58927411e-01
-6.36591613e-02 3.33973676e-01 1.01459093e-01 -5.67014039e-01
9.80031013e-01 1.89712077e-01 -1.37489974e+00 5.96703768e-01
5.99252522e-01 7.02551544e-01 2.82042295e-01 1.02726603e+00
-9.06537354e-01 9.68849838e-01 -8.49385262e-01 -4.30339307e-01
-3.94405246e-01 -3.79791200e-01 -1.42532229e+00 -1.93612874e-01
1.45893842e-01 -7.75961220e-01 -1.08677400e-02 5.33234835e-01
5.84159553e-01 3.19771081e-01 5.26248395e-01 -1.52613592e+00
-3.21859360e-01 5.99868536e-01 -1.24078169e-01 2.62182653e-01
1.50353944e+00 -7.48525336e-02 8.31692338e-01 -7.63224244e-01
5.80498397e-01 1.53024483e+00 3.99913192e-01 9.08787191e-01
-8.28260243e-01 -2.53004611e-01 4.42926854e-01 5.57443574e-02
-1.02511132e+00 -4.53632683e-01 4.13211524e-01 -5.90961695e-01
1.09652710e+00 9.05463874e-01 4.80814815e-01 9.73539114e-01
3.57105676e-03 1.20581985e+00 5.91226935e-01 -8.20227444e-01
1.31237417e-01 5.07363081e-01 9.15049851e-01 2.43354008e-01
-1.83243215e-01 -3.70719761e-01 -7.96271980e-01 -9.04170871e-01
4.89576966e-01 -3.58511731e-02 -2.97100782e-01 -2.85340726e-01
-1.14378285e+00 6.23025537e-01 -7.67138824e-02 7.12129056e-01
-3.98915768e-01 -1.80895254e-01 2.64094383e-01 3.92277688e-01
-7.96668902e-02 5.47051430e-01 -1.53728068e-01 -9.25953329e-01
-1.28384754e-01 5.64186931e-01 1.89842391e+00 1.01267040e+00
1.72080621e-01 -8.06188285e-01 -5.15335441e-01 1.13802195e+00
7.52307773e-01 2.60675520e-01 1.99324876e-01 -1.03490281e+00
3.60671729e-01 5.60371637e-01 6.97847009e-01 -9.80965912e-01
-3.78578126e-01 5.96709788e-01 1.04152694e-01 1.48064299e-02
4.59405690e-01 -4.43568438e-01 -1.22604162e-01 9.58752930e-01
2.77986407e-01 -8.93582046e-01 -9.33742300e-02 6.40686870e-01
1.40501189e+00 1.82325572e-01 1.30410612e-01 -2.72314012e-01
1.43227267e+00 -1.00271654e+00 -1.19469619e+00 -1.01607092e-01
5.04707456e-01 -1.12326467e+00 1.11149728e+00 2.10451558e-01
-1.23756361e+00 -3.50275473e-03 -5.01299381e-01 -1.95071980e-01
-6.64005756e-01 -1.00682817e-01 6.76229119e-01 1.04119265e+00
-9.55565751e-01 1.04120433e-01 -7.00623512e-01 -9.85139549e-01
-4.44459081e-01 3.57490987e-01 2.77319431e-01 1.77038714e-01
-1.00716782e+00 1.04913163e+00 -3.66534352e-01 6.98013380e-02
1.06343910e-01 -3.59445482e-01 -4.19351876e-01 -2.15643197e-01
2.97719032e-01 -3.90381873e-01 2.27930498e+00 -1.71943903e-01
-1.88759947e+00 5.61207116e-01 -4.84462410e-01 1.86016858e-01
3.11755180e-01 -5.13473392e-01 -3.93060982e-01 1.08437978e-01
6.98045548e-03 4.70119476e-01 1.27097312e-02 -1.24645102e+00
-8.05611849e-01 -1.09545805e-01 5.98497570e-01 8.75641406e-01
-1.23335324e-01 7.09170163e-01 -7.95629740e-01 5.03453873e-02
-2.07703248e-01 -9.19243276e-01 -5.42974733e-02 4.60566059e-02
-2.93388605e-01 -8.09191287e-01 1.08021188e+00 -5.19278824e-01
1.73640192e+00 -1.88666785e+00 -2.21761674e-01 1.72629714e-01
2.30542868e-01 3.56512398e-01 2.03116745e-01 1.50631452e+00
4.88672137e-01 2.32562721e-01 2.82339722e-01 -3.31312269e-01
2.87186056e-01 -2.61091322e-01 7.17311278e-02 -4.51577045e-02
-7.57197678e-01 1.01178777e+00 -9.48774099e-01 -5.23568690e-01
5.02921760e-01 4.83658463e-01 -2.42451087e-01 4.68856961e-01
-2.11976588e-01 4.98242915e-01 -6.86902583e-01 7.04572797e-01
2.69586563e-01 -3.63591790e-01 4.43031162e-01 4.10927832e-01
-5.78549504e-01 9.57620859e-01 -9.49222267e-01 1.42366099e+00
-5.18359363e-01 7.59483457e-01 5.31944990e-01 -1.61025256e-01
4.99263763e-01 5.10011315e-01 5.36420584e-01 -4.85404432e-01
-6.46226630e-02 -8.82318020e-02 -7.79300928e-02 -9.87200201e-01
5.45050025e-01 7.24832773e-01 -7.22589791e-02 1.21464479e+00
-6.47263646e-01 4.74740611e-03 2.16065347e-01 4.51675951e-01
9.92507815e-01 -2.91477948e-01 6.26492918e-01 -7.41102472e-02
6.28793538e-01 -7.08474815e-02 -4.13504869e-01 1.23468792e+00
-2.91170746e-01 -1.59927696e-01 3.57362896e-01 -1.42971992e-01
1.42520629e-02 -7.24477291e-01 -8.99025332e-03 1.67997217e+00
5.35318851e-01 -8.85647058e-01 -9.80203867e-01 -7.49572217e-01
-1.04984075e-01 7.27470994e-01 -3.06628585e-01 5.07566333e-01
-3.41247022e-01 -7.69738629e-02 2.42855340e-01 2.38635302e-01
7.55446672e-01 -1.50720823e+00 -8.91304016e-01 8.50027204e-02
-6.53942823e-01 -8.27055633e-01 -7.78080702e-01 -6.51099086e-01
-7.31324494e-01 -1.27285445e+00 -7.62286127e-01 -8.30466390e-01
2.74230897e-01 7.70182610e-01 1.04910254e+00 4.05875713e-01
3.22664692e-03 1.29688311e+00 -5.18542945e-01 -2.70229608e-01
-1.09521002e-01 -5.40386140e-02 -2.09332019e-01 -1.70175314e-01
9.69900787e-01 -1.78250208e-01 -9.73358393e-01 7.19559431e-01
-5.30870318e-01 -1.74647272e-01 1.20697103e-01 4.51341003e-01
-5.90209544e-01 -5.67509890e-01 1.74815252e-01 -1.03135490e+00
1.68489361e+00 -3.40577841e-01 1.63843155e-01 4.60531443e-01
-5.45553923e-01 -5.22451937e-01 -3.04050922e-01 -5.15519738e-01
-1.55773664e+00 -4.04141933e-01 -1.95381984e-01 8.08276176e-01
-5.87278128e-01 5.32260656e-01 6.56697080e-02 -5.08956015e-01
8.55818331e-01 -1.10662736e-01 1.08774185e-01 -7.40522265e-01
3.96834493e-01 1.59023225e+00 -8.56249779e-03 -4.62753892e-01
7.14368299e-02 1.61050573e-01 -9.34046865e-01 -1.32453239e+00
-1.88189477e-01 -1.06235540e+00 -3.69877070e-01 -6.23106301e-01
5.40784061e-01 -8.89978483e-02 -1.70240402e+00 4.19627815e-01
-1.34864998e+00 -5.23200572e-01 3.73709738e-01 4.19282287e-01
-2.36544982e-01 7.58787870e-01 -9.90198791e-01 -1.52499521e+00
-6.87722862e-01 -9.81889486e-01 8.00153017e-01 5.53768158e-01
-1.27580678e+00 -1.14094675e+00 1.01093568e-01 8.56705427e-01
8.81415248e-01 -2.54704684e-01 6.89616740e-01 -8.52746487e-01
-2.24058047e-01 -3.89657795e-01 -1.95863517e-03 -4.02634680e-01
3.79590571e-01 -7.03909248e-02 -4.72564161e-01 -1.63778484e-01
-1.38157219e-01 -1.67648494e-01 -1.38422623e-01 9.39232931e-02
6.96143985e-01 -5.55235207e-01 -9.46816325e-01 -5.60008049e-01
6.53147817e-01 1.16545188e+00 3.66330504e-01 1.65273383e-01
1.16695911e-01 1.04936659e+00 7.46116400e-01 3.81706595e-01
5.87265849e-01 8.69610667e-01 -6.89246356e-02 3.94046187e-01
3.02682314e-02 -5.40817492e-02 1.14297479e-01 5.28187037e-01
-3.45006436e-02 -3.86911720e-01 -1.05939162e+00 3.11848789e-01
-1.94979918e+00 -8.08360755e-01 -8.58766213e-02 2.09817171e+00
6.07197464e-01 -3.48781794e-01 3.79915804e-01 -1.72025070e-01
7.37213552e-01 -1.20904790e-02 -5.32500923e-01 -7.05583274e-01
7.88414419e-01 -2.60063827e-01 -2.50478417e-01 9.61940467e-01
-7.19603121e-01 7.22526133e-01 7.99406099e+00 -2.45863106e-03
-8.71520579e-01 -6.79321662e-02 6.58154860e-02 1.58622533e-01
-8.11981931e-02 -1.22471467e-01 -8.02798092e-01 3.44713390e-01
6.39627695e-01 -9.24473926e-02 7.14795828e-01 8.98568451e-01
1.21746764e-01 -5.64048529e-01 -1.29746246e+00 1.20640922e+00
2.21037492e-01 -1.20565820e+00 -2.19640136e-01 -8.77416413e-03
1.50898099e-01 -2.89259762e-01 -2.81087339e-01 2.95203686e-01
2.84469128e-01 -5.88915825e-01 3.93165089e-03 4.13430601e-01
5.17720997e-01 -1.88220158e-01 4.17941451e-01 5.45967638e-01
-1.06957746e+00 -9.41088051e-02 5.26448250e-01 -3.82002950e-01
4.60792691e-01 -2.18045920e-01 -6.77623093e-01 3.08375154e-02
9.25910532e-01 3.12465519e-01 -1.57321870e-01 1.14596868e+00
1.27757236e-01 9.34869498e-02 -2.63034940e-01 -1.00713861e+00
-1.28336728e-01 -3.17900181e-01 5.82573354e-01 1.34997725e+00
-2.75635391e-01 7.62589216e-01 9.94445235e-02 3.66507053e-01
4.63137299e-01 2.92135984e-01 -6.29407763e-01 -3.05478573e-01
8.73027265e-01 1.09466958e+00 -4.02394116e-01 -3.21359575e-01
-6.28127992e-01 1.06693006e+00 -8.88943374e-02 6.14896238e-01
-1.28383145e-01 -8.67939115e-01 8.71425867e-01 -1.28439050e-02
-2.67548025e-01 -3.85868847e-01 -2.76820838e-01 -8.46470833e-01
8.47409666e-02 -1.22171974e+00 4.84739631e-01 -6.69057727e-01
-9.53115344e-01 4.10640657e-01 3.02939087e-01 -8.82012248e-01
-8.40338707e-01 -2.79589206e-01 -8.83436382e-01 8.72852623e-01
-6.31163239e-01 -5.65459251e-01 -7.34360278e-01 3.89552861e-01
7.18051314e-01 2.11158574e-01 1.12891150e+00 1.79159135e-01
-1.93369836e-01 4.19582009e-01 -1.12738661e-01 -1.09658480e-01
7.34609425e-01 -1.04654884e+00 4.47362423e-01 -3.85122672e-02
-3.44891280e-01 1.45953321e+00 6.61010206e-01 -4.70081270e-01
-1.80377173e+00 1.12610072e-01 1.30521524e+00 -6.14059150e-01
2.93289542e-01 -4.48530823e-01 -5.17752588e-01 7.44102895e-01
7.67856121e-01 -1.09456229e+00 1.00153589e+00 6.45245373e-01
4.12972756e-02 4.13646758e-01 -1.24223650e+00 1.25593102e+00
1.04158187e+00 -8.34084213e-01 -5.14431953e-01 5.88529646e-01
4.39896047e-01 -6.58334434e-01 -7.10074604e-01 -2.33884677e-02
1.15848684e+00 -9.51343775e-01 9.07931447e-01 -4.53954101e-01
-2.02752396e-01 2.02633366e-01 2.22381413e-01 -9.34112787e-01
-7.98519235e-03 -1.38545597e+00 -2.57397383e-01 1.28433967e+00
3.72276872e-01 -8.49952698e-01 6.87789619e-01 1.79948342e+00
3.02991986e-01 -4.79741514e-01 -2.43268520e-01 -1.95423737e-01
-3.00475061e-01 -3.06109935e-01 3.69624048e-01 9.29522932e-01
1.55781484e+00 8.32708478e-01 -7.52647445e-02 -3.15069854e-01
2.19085142e-01 3.63560058e-02 1.07594621e+00 -1.22484815e+00
-6.79766312e-02 -6.47548676e-01 3.27179700e-01 -1.98042238e+00
-4.16653007e-01 -1.83330595e-01 2.42585931e-02 -1.86769700e+00
-6.72501773e-02 -2.99906045e-01 4.23299462e-01 2.60036319e-01
6.94140494e-02 -3.18785787e-01 2.74869829e-01 2.77482033e-01
-7.57212460e-01 -2.48332066e-03 1.04096293e+00 2.00505331e-01
-8.36512446e-01 5.20069420e-01 -1.02881742e+00 5.28746545e-01
8.76712263e-01 2.44648084e-01 -4.04986620e-01 -2.48350408e-02
2.01040283e-01 5.97477555e-01 -2.37695858e-01 -3.30720544e-01
8.38845015e-01 -3.14514548e-01 -1.78842947e-01 -7.32347965e-01
5.63445330e-01 -6.57499909e-01 -1.99453175e-01 4.23985183e-01
-8.94838214e-01 2.97100339e-02 7.64131034e-03 3.91518742e-01
-1.13327980e-01 -4.01215911e-01 -4.08482999e-02 -7.89615735e-02
-5.31251431e-01 -3.70061040e-01 -1.30374765e+00 -3.19997877e-01
7.44827390e-01 -2.28268385e-01 -4.47825432e-01 -1.23343337e+00
-7.82320321e-01 6.81729078e-01 -7.54362643e-02 8.32080603e-01
4.42882121e-01 -9.45163965e-01 7.50872418e-02 -1.74756482e-01
2.97871321e-01 -5.65673113e-01 -4.28542942e-02 6.06426418e-01
-3.19062591e-01 1.08599198e+00 2.36421511e-01 -3.76224995e-01
-1.83953774e+00 1.06049396e-01 2.20824629e-01 6.54062862e-03
-2.62053251e-01 5.04481077e-01 8.58144090e-02 -7.67992795e-01
1.02618933e+00 -2.09082873e-03 -7.78932869e-01 -1.29047811e-01
1.10026228e+00 5.92958450e-01 -1.36488184e-01 -1.60383452e-02
-6.15956008e-01 2.49406889e-01 -2.81657219e-01 -7.60484338e-01
3.77965093e-01 -7.06743777e-01 -1.90453976e-01 7.19269514e-01
8.99377406e-01 -1.54934585e-01 -4.25739080e-01 -3.33938271e-01
2.42530987e-01 -4.56669956e-01 -4.06019568e-01 -1.25272131e+00
-1.91024974e-01 7.04973638e-01 3.87004852e-01 7.68122673e-01
4.55930024e-01 1.87169835e-01 7.18476892e-01 9.53402221e-01
3.80890340e-01 -1.13591063e+00 3.84146631e-01 5.92183948e-01
1.16630304e+00 -1.09378541e+00 -3.79808843e-01 -7.12286294e-01
-8.77878487e-01 1.13805532e+00 6.05461538e-01 6.21007264e-01
1.18798494e+00 6.07373714e-02 5.10541797e-01 -5.26818931e-01
-7.31126785e-01 -9.61614996e-02 2.25312069e-01 6.08679652e-01
1.20776033e+00 -1.60701856e-01 -9.64117050e-01 2.91366994e-01
-1.39530897e-01 1.35163710e-01 1.55993015e-01 1.63297606e+00
-2.64151871e-01 -1.24884558e+00 -2.85423696e-01 5.11796594e-01
-3.21839809e-01 -7.45440573e-02 -1.23892224e+00 5.42260289e-01
-7.93529153e-01 2.09304976e+00 -5.10832071e-02 -4.48541760e-01
5.20897329e-01 3.21674228e-01 4.00401540e-02 -5.74119210e-01
-1.24401891e+00 -1.79307118e-01 8.24076056e-01 -7.87986875e-01
-4.19579417e-01 -8.55360031e-01 -9.50740159e-01 -4.86031055e-01
-4.58117962e-01 6.44465148e-01 9.79369044e-01 9.33536291e-01
5.71799695e-01 -2.09280953e-01 2.47738153e-01 -7.36664474e-01
-3.17074090e-01 -1.33273077e+00 -2.26341143e-01 2.06238851e-01
4.15793955e-01 -4.21012163e-01 -4.37715113e-01 -1.80953085e-01] | [12.249733924865723, 7.774843692779541] |
6819881e-9b7f-4475-aa47-38dedf6843b1 | learning-to-cluster-faces-via-transformer | 2104.11502 | null | https://arxiv.org/abs/2104.11502v1 | https://arxiv.org/pdf/2104.11502v1.pdf | Learning to Cluster Faces via Transformer | Face clustering is a useful tool for applications like automatic face annotation and retrieval. The main challenge is that it is difficult to cluster images from the same identity with different face poses, occlusions, and image quality. Traditional clustering methods usually ignore the relationship between individual images and their neighbors which may contain useful context information. In this paper, we repurpose the well-known Transformer and introduce a Face Transformer for supervised face clustering. In Face Transformer, we decompose the face clustering into two steps: relation encoding and linkage predicting. Specifically, given a face image, a \textbf{relation encoder} module aggregates local context information from its neighbors and a \textbf{linkage predictor} module judges whether a pair of images belong to the same cluster or not. In the local linkage graph view, Face Transformer can generate more robust node and edge representations compared to existing methods. Experiments on both MS-Celeb-1M and DeepFashion show that our method achieves state-of-the-art performance, e.g., 91.12\% in pairwise F-score on MS-Celeb-1M. | ['Hanqing Wu', 'Hao Li', 'Xiuyu Sun', 'Kai Wang', 'Baigui Sun', 'Xioajiang Peng', 'Jinxing Ye'] | 2021-04-23 | null | null | null | null | ['face-clustering'] | ['computer-vision'] | [-9.93765667e-02 8.70367661e-02 -1.68361828e-01 -8.67081881e-01
-5.17752588e-01 -3.46282214e-01 3.27551395e-01 -2.32830435e-01
2.10292712e-01 2.62176305e-01 7.70756081e-02 2.45393187e-01
-2.81660527e-01 -5.76010942e-01 -5.56092978e-01 -8.60642731e-01
-5.35483211e-02 7.10830450e-01 -1.81955218e-01 1.92672044e-01
-1.28415659e-01 4.14338559e-01 -1.63595629e+00 4.28251296e-01
4.74643320e-01 1.20316517e+00 5.49129769e-02 8.26759115e-02
-5.95259108e-02 6.50052190e-01 -4.37008083e-01 -8.22289944e-01
1.02013782e-01 -5.30054271e-01 -8.23686242e-01 2.42035151e-01
9.67527807e-01 -2.20455408e-01 -4.80946362e-01 1.32289600e+00
5.87040067e-01 -2.08491430e-01 6.92808032e-01 -1.73446679e+00
-7.40767717e-01 6.36363268e-01 -1.06686366e+00 -6.35490119e-02
3.38192761e-01 -2.59171277e-01 7.43496478e-01 -1.11605525e+00
7.55151153e-01 1.83101869e+00 8.06367993e-01 6.28997386e-01
-1.32497931e+00 -1.10974658e+00 2.29227796e-01 4.54933941e-01
-1.96682227e+00 -9.57744539e-01 6.43946528e-01 -4.07450199e-01
4.26974952e-01 1.21363834e-01 2.37070024e-01 6.84586644e-01
-1.94670856e-01 4.43444103e-01 8.97633016e-01 -9.72543433e-02
5.75787649e-02 -1.11183785e-01 -9.54344422e-02 9.83523309e-01
1.10073641e-01 -2.76979685e-01 -9.43254411e-01 -1.09386025e-03
5.05960166e-01 -7.39854947e-02 -4.34922606e-01 -4.52791989e-01
-7.26092398e-01 7.21257865e-01 6.79812491e-01 1.98810715e-02
-6.70472383e-02 1.99710146e-01 8.90197456e-02 9.57028940e-02
3.27945441e-01 -2.96000838e-01 7.16257654e-03 3.63968104e-01
-1.07025099e+00 -2.38196254e-01 7.02723682e-01 1.24086773e+00
8.31236184e-01 -3.60599458e-01 -9.04951096e-02 1.07246530e+00
7.18363702e-01 4.92750049e-01 1.05509102e-01 -1.16582477e+00
2.92076617e-01 6.69291019e-01 -5.90283155e-01 -1.47843194e+00
-1.00744382e-01 -6.74377978e-02 -1.13135517e+00 -4.19443958e-02
-1.33313229e-02 1.51600748e-01 -7.98637211e-01 1.93889737e+00
5.03494084e-01 6.46946669e-01 -1.03910781e-01 7.17335761e-01
1.26104379e+00 3.60670686e-01 -2.02111825e-01 -4.86849874e-01
1.43512440e+00 -8.60692501e-01 -8.45580816e-01 -8.41021817e-03
2.55073696e-01 -9.42622960e-01 4.04732555e-01 7.35062957e-02
-1.00446022e+00 -6.30696774e-01 -7.48931825e-01 -7.71148726e-02
-1.13870390e-01 4.01581824e-01 4.99643326e-01 7.35285997e-01
-1.45869899e+00 3.40845972e-01 -5.36445975e-01 -4.19224679e-01
9.11754012e-01 7.01363802e-01 -7.96727777e-01 -6.32883787e-01
-6.83209062e-01 2.57262141e-01 2.57781055e-02 2.65833855e-01
-8.69579375e-01 -4.72934842e-01 -9.02124882e-01 -3.83050181e-02
3.93202335e-01 -3.62506658e-01 7.63294280e-01 -7.80556262e-01
-1.01340640e+00 1.09800780e+00 -7.35380650e-01 1.31510347e-01
9.54548195e-02 3.26665007e-02 -5.59148014e-01 2.97930598e-01
3.73387784e-01 1.05469739e+00 9.47012782e-01 -1.43166029e+00
-4.87085789e-01 -8.60893905e-01 -3.87334287e-01 1.47819191e-01
-4.78204370e-01 2.03500926e-01 -1.32304883e+00 -5.63667715e-01
3.61580640e-01 -9.80667651e-01 3.41682971e-01 3.64532650e-01
-5.97890913e-01 -5.15196025e-01 1.25422716e+00 -4.46937293e-01
1.08038223e+00 -2.26625752e+00 2.15275869e-01 5.46810091e-01
3.70135874e-01 1.66540205e-01 -3.11924458e-01 1.26920536e-01
-3.78953934e-01 2.17565134e-01 -5.34857213e-02 -5.67697883e-01
-1.18260264e-01 1.66677937e-01 1.63634449e-01 6.03468060e-01
1.63417682e-01 6.64807796e-01 -6.96430147e-01 -8.28742027e-01
-6.55006543e-02 8.62250090e-01 -5.22386968e-01 1.27628550e-01
1.04460284e-01 4.15305287e-01 -2.20984034e-02 8.27405274e-01
1.03472602e+00 -2.87817508e-01 7.05608249e-01 -7.08858848e-01
4.29169774e-01 -8.82054586e-03 -1.42493665e+00 1.71349859e+00
-2.58801482e-03 7.07670510e-01 4.91086245e-01 -8.15572083e-01
7.86176443e-01 3.32165629e-01 5.97504437e-01 -3.99561226e-01
1.18047461e-01 -1.73716798e-01 -1.38512507e-01 -3.50845426e-01
3.21282595e-02 3.83361787e-01 4.69395816e-01 5.07301390e-01
2.81799912e-01 4.75626081e-01 1.89051583e-01 4.01679933e-01
8.53187442e-01 -1.88611355e-02 -1.07780606e-01 -4.43274081e-01
5.99604964e-01 -6.30938232e-01 8.33067954e-01 1.00174852e-01
-1.88414022e-01 6.93338335e-01 7.02639282e-01 -1.97635770e-01
-4.79892939e-01 -1.08632159e+00 -2.32105196e-01 1.07414579e+00
2.25510612e-01 -1.03782344e+00 -1.03695881e+00 -7.98561037e-01
4.06438485e-02 -4.70508225e-02 -6.67436481e-01 -2.17828661e-01
-3.79779160e-01 -8.19182694e-01 5.69726825e-01 4.77883130e-01
6.79959357e-01 -6.48368061e-01 2.43140787e-01 -3.14780056e-01
-3.88037205e-01 -1.12257195e+00 -8.49107981e-01 -4.32940632e-01
-5.27615309e-01 -1.44009244e+00 -2.89804757e-01 -1.31789863e+00
1.07197654e+00 5.15993416e-01 1.18486822e+00 3.87544543e-01
-4.32812393e-01 3.09472740e-01 -1.59190461e-01 1.30674586e-01
7.01955520e-03 -1.64746180e-01 2.44095370e-01 3.63945276e-01
7.12855279e-01 -3.98498267e-01 -7.62943089e-01 7.44125366e-01
-4.82152224e-01 -1.83451876e-01 3.06937903e-01 6.73690379e-01
7.73183763e-01 4.42099065e-01 5.13681769e-01 -9.43793297e-01
2.05446839e-01 -4.54014152e-01 -2.92353719e-01 6.32402658e-01
-4.76509988e-01 -2.59797305e-01 2.16609687e-02 -1.09507762e-01
-8.83834183e-01 5.36308825e-01 1.32088751e-01 -6.59990013e-01
3.99660990e-02 1.41118899e-01 -7.98074424e-01 -1.36404097e-01
2.58038372e-01 -1.14976995e-01 2.13207483e-01 -3.88130933e-01
4.25364375e-01 7.76795208e-01 7.20128000e-01 -6.16294205e-01
7.17334926e-01 6.18922174e-01 8.23401660e-02 -8.55940580e-01
-5.26185632e-01 -3.82916123e-01 -9.03106153e-01 -3.86700869e-01
9.18860316e-01 -1.25082040e+00 -1.24401963e+00 3.23374718e-01
-1.03447783e+00 8.36041793e-02 2.87669688e-01 4.27642465e-02
-1.29773796e-01 3.28399330e-01 -5.81335902e-01 -5.11642277e-01
-6.86430261e-02 -1.41429710e+00 1.29030001e+00 3.17602158e-01
1.44356759e-02 -6.51885211e-01 -4.89256352e-01 6.14928663e-01
-1.00847498e-01 3.63608673e-02 8.19454551e-01 -2.45126739e-01
-5.96694410e-01 1.24381870e-01 -4.80381489e-01 2.34486446e-01
5.12884617e-01 2.04754472e-01 -1.21006644e+00 -5.99801958e-01
-2.38628566e-01 -4.00689095e-01 8.02998483e-01 1.56077385e-01
1.40249133e+00 -3.52744132e-01 -8.21922958e-01 7.93087602e-01
1.12605238e+00 1.57655388e-01 6.42455101e-01 -5.48275769e-01
1.00840235e+00 9.81696725e-01 3.54019254e-01 2.09770098e-01
4.78641123e-01 8.89928222e-01 2.23557517e-01 1.10751957e-01
-4.67264682e-01 -2.62307376e-01 3.94442201e-01 7.33490586e-01
1.27815798e-01 -1.68677270e-01 -8.44224632e-01 3.50059032e-01
-1.75281858e+00 -9.27023888e-01 8.71901680e-03 1.98145974e+00
8.14673841e-01 -4.19228494e-01 -7.60183483e-02 -6.00935966e-02
1.30743623e+00 -1.20480657e-01 -4.46148187e-01 3.64835560e-01
-1.45879969e-01 9.45279300e-02 -1.35918818e-02 4.55479234e-01
-1.09179640e+00 1.01776242e+00 5.45759773e+00 9.82747555e-01
-8.89694273e-01 1.72941640e-01 1.09363794e+00 -2.45687380e-01
7.03099817e-02 -1.89069286e-01 -1.09334874e+00 5.94496727e-01
5.11943400e-01 1.90041631e-01 4.88162816e-01 6.02800608e-01
-1.83381036e-01 2.08272897e-02 -1.44025707e+00 1.44460905e+00
6.31905377e-01 -1.17575026e+00 2.25588173e-01 3.41698736e-01
7.19208837e-01 -3.17247421e-01 1.97708786e-01 -1.31005228e-01
3.04487318e-01 -1.51668561e+00 3.83394271e-01 4.20081109e-01
1.13133240e+00 -9.82359111e-01 6.14314675e-01 -1.04437947e-01
-1.68532109e+00 7.87906051e-02 -4.55842137e-01 5.18761575e-01
-1.15431912e-01 5.79693317e-01 -7.48396337e-01 5.96878588e-01
1.18496716e+00 8.89321923e-01 -8.30024064e-01 5.96053004e-01
-1.63564444e-01 2.96297073e-01 -3.68373066e-01 6.08751118e-01
-3.83970559e-01 -4.10925508e-01 1.66060925e-02 8.37028742e-01
3.41070533e-01 2.87795424e-01 2.94020772e-01 5.74020445e-01
-6.43261373e-01 -4.76147085e-02 -6.56954587e-01 2.19315112e-01
9.23103213e-01 1.43166697e+00 -8.32298040e-01 -7.07711205e-02
-4.85158592e-01 1.02752423e+00 5.10039151e-01 2.63929069e-01
-7.16551304e-01 8.15096200e-02 9.69287932e-01 1.25541240e-01
4.90627676e-01 1.46645501e-01 1.21602818e-01 -8.18382680e-01
7.99015723e-03 -1.00964200e+00 5.50933301e-01 -7.61785448e-01
-1.43473899e+00 8.89127314e-01 -2.54064828e-01 -8.18200886e-01
8.96731205e-03 -4.75283235e-01 -3.92276287e-01 5.89859486e-01
-1.09458899e+00 -1.45965719e+00 -6.01368427e-01 9.17142510e-01
2.69259423e-01 -4.49351668e-01 8.02689016e-01 6.91786468e-01
-9.44940209e-01 9.71402466e-01 7.56315961e-02 5.63873827e-01
1.07905686e+00 -9.10488367e-01 5.47081046e-02 7.34139979e-01
4.61040974e-01 8.29951048e-01 5.68130314e-02 -7.98023403e-01
-1.39726436e+00 -1.52936876e+00 6.88721716e-01 -3.37806225e-01
4.35796380e-02 -5.81009448e-01 -7.08644271e-01 6.87640965e-01
3.39465052e-01 3.38248402e-01 9.60131466e-01 1.92203090e-01
-7.47018099e-01 -6.24425530e-01 -1.20631635e+00 5.44993937e-01
1.45005906e+00 -7.57522464e-01 -7.72049576e-02 4.22828466e-01
3.20479542e-01 -8.74824226e-02 -9.97336030e-01 4.21712101e-01
4.05854046e-01 -1.10179746e+00 1.04750490e+00 -2.98285693e-01
2.22823873e-01 -6.56398296e-01 -4.24046099e-01 -1.02757180e+00
-4.22047168e-01 -5.66551626e-01 -1.09327249e-02 1.93277919e+00
1.01139143e-01 -2.14838699e-01 8.86666954e-01 3.76619995e-01
2.54556835e-01 -6.35771573e-01 -1.02969623e+00 -5.52158713e-01
-3.08099419e-01 -8.54875520e-02 8.69014382e-01 1.16695762e+00
-2.58495569e-01 6.30918145e-01 -1.73081309e-01 3.97824973e-01
9.12505388e-01 2.14940429e-01 7.65661836e-01 -1.53339934e+00
1.86448351e-01 -3.65279138e-01 -5.33908188e-01 -9.38371599e-01
5.88688254e-01 -1.16853797e+00 5.24360500e-02 -1.32644784e+00
5.08865535e-01 -6.09742939e-01 -6.62994534e-02 6.96513355e-01
-1.92993358e-02 8.65224957e-01 1.17672734e-01 3.21147352e-01
-7.45856702e-01 5.16648293e-01 8.80186975e-01 -2.83932269e-01
2.57919341e-01 -4.13469762e-01 -7.70585895e-01 7.43652761e-01
4.85777408e-01 -4.82235402e-01 -7.12886095e-01 -5.22309780e-01
1.00663677e-02 -1.37021556e-01 2.71643609e-01 -1.03851080e+00
6.72403991e-01 2.45517299e-01 8.19008291e-01 -5.97256601e-01
5.50212681e-01 -9.09707725e-01 4.41818535e-01 1.46825507e-01
-1.00678876e-01 2.76241690e-01 -4.98141311e-02 6.82333767e-01
-2.96890706e-01 2.57260799e-02 7.60000169e-01 5.22278771e-02
-5.76023459e-01 7.14285493e-01 1.91439077e-01 4.13242988e-02
1.17678320e+00 -9.38886926e-02 -4.33722317e-01 -4.12312865e-01
-6.87838912e-01 4.32338595e-01 5.97467661e-01 5.22562683e-01
7.43184030e-01 -1.65091193e+00 -6.92991138e-01 4.69810814e-01
1.72024980e-01 9.49022174e-02 2.53838509e-01 7.12739825e-01
-2.42947698e-01 7.29584172e-02 -4.76617478e-02 -9.06419456e-01
-1.92004371e+00 4.82158870e-01 2.88433671e-01 4.65341479e-01
-2.80420840e-01 1.19696963e+00 4.62094098e-01 -2.76739419e-01
4.49483663e-01 3.66071343e-01 -2.25894168e-01 4.43999946e-01
7.09297180e-01 1.82453394e-01 3.59107815e-02 -1.32953179e+00
-7.00980723e-01 9.41639841e-01 -1.93819478e-01 2.03035459e-01
1.11648989e+00 -1.50323138e-01 -8.54130685e-01 -7.49819428e-02
1.38425994e+00 -1.02876730e-01 -1.08627129e+00 -1.20333269e-01
-8.46916363e-02 -6.66517735e-01 -6.89279214e-02 -5.83821893e-01
-1.88506341e+00 8.45288575e-01 7.59977400e-01 -2.41204515e-01
1.29847312e+00 2.95908868e-01 4.90299821e-01 2.68196881e-01
2.56721526e-01 -9.51362193e-01 2.03735784e-01 2.14623138e-02
8.53521466e-01 -1.21978557e+00 9.21184495e-02 -1.03383756e+00
-4.44654644e-01 7.90602982e-01 8.90344381e-01 3.95213753e-01
1.09402835e+00 2.88644016e-01 8.17862749e-02 -4.88103032e-01
-7.13693321e-01 -7.32611865e-02 3.40310812e-01 7.93353975e-01
5.96960187e-01 3.35617014e-03 2.44029716e-01 4.09212381e-01
-1.71206534e-01 -3.02727789e-01 -7.62432665e-02 4.98200893e-01
-1.82913527e-01 -1.15038812e+00 -4.58293587e-01 4.97175932e-01
-3.56738925e-01 -1.93558745e-02 -6.94528162e-01 5.50517440e-01
4.82900888e-01 1.35748696e+00 4.12583739e-01 -6.58405483e-01
-1.83438584e-01 5.69454171e-02 5.92376828e-01 -6.62564039e-01
-2.23545179e-01 2.03517005e-01 -8.67472515e-02 -6.95790887e-01
-7.54868031e-01 -7.96775401e-01 -1.25139260e+00 -6.62458479e-01
-2.65502512e-01 1.39542669e-01 4.34516579e-01 6.25066638e-01
6.60528243e-01 2.87731946e-01 6.97340965e-01 -5.22074223e-01
1.90728411e-01 -8.16416264e-01 -5.19008636e-01 5.75970352e-01
-7.90350214e-02 -8.60448658e-01 9.09166485e-02 4.23038423e-01] | [13.485862731933594, 1.01964271068573] |
cf1fb002-c3dd-4110-abf3-970c3e102254 | id-pose-sparse-view-camera-pose-estimation-by | 2306.17140 | null | https://arxiv.org/abs/2306.17140v1 | https://arxiv.org/pdf/2306.17140v1.pdf | ID-Pose: Sparse-view Camera Pose Estimation by Inverting Diffusion Models | Given sparse views of an object, estimating their camera poses is a long-standing and intractable problem. We harness the pre-trained diffusion model of novel views conditioned on viewpoints (Zero-1-to-3). We present ID-Pose which inverses the denoising diffusion process to estimate the relative pose given two input images. ID-Pose adds a noise on one image, and predicts the noise conditioned on the other image and a decision variable for the pose. The prediction error is used as the objective to find the optimal pose with the gradient descent method. ID-Pose can handle more than two images and estimate each of the poses with multiple image pairs from triangular relationships. ID-Pose requires no training and generalizes to real-world images. We conduct experiments using high-quality real-scanned 3D objects, where ID-Pose significantly outperforms state-of-the-art methods. | ['Ying Shan', 'Yan-Pei Cao', 'Weihao Cheng'] | 2023-06-29 | null | null | null | null | ['pose-estimation'] | ['computer-vision'] | [ 1.36429891e-01 1.35360926e-01 1.09578736e-01 -3.81496251e-01
-1.02280974e+00 -5.50574481e-01 3.80429357e-01 -7.45025277e-01
-8.03412423e-02 1.03398904e-01 1.59886897e-01 3.11089516e-01
-3.41859683e-02 -5.75345457e-01 -7.40555882e-01 -5.95254600e-01
7.17940331e-02 1.07825077e+00 2.73365378e-01 1.84292912e-01
3.29852015e-01 5.98709941e-01 -1.12654066e+00 2.06919953e-01
2.87919074e-01 1.02566934e+00 5.74776173e-01 7.29668140e-01
2.93714017e-01 5.26179969e-01 -1.52697548e-01 -4.45952445e-01
7.78574824e-01 -2.77049690e-01 -4.97483730e-01 7.04005361e-01
1.01479650e+00 -8.02999854e-01 -4.55030143e-01 1.18959856e+00
2.92454541e-01 4.81877811e-02 7.38476634e-01 -8.59523177e-01
-6.08443141e-01 -9.66851115e-02 -9.69788611e-01 2.21624002e-02
5.14543116e-01 1.29267797e-01 8.79602909e-01 -1.39994621e+00
1.12823367e+00 1.46778452e+00 7.22417653e-01 3.91457140e-01
-1.49844515e+00 -3.19833934e-01 2.98553497e-01 1.23893656e-01
-1.42838454e+00 -3.87803227e-01 9.70537066e-01 -7.08998859e-01
7.32734382e-01 -2.03473717e-01 7.46575296e-01 1.13646376e+00
2.00529531e-01 5.18145025e-01 1.10685575e+00 -1.22323222e-01
3.19915950e-01 -1.62684619e-01 -2.41419718e-01 8.61352324e-01
1.31458953e-01 1.59316082e-02 -8.39743793e-01 -2.13729054e-01
1.14453173e+00 2.53376096e-01 -4.19714838e-01 -8.90638828e-01
-1.24577844e+00 6.99848235e-01 3.33297044e-01 -2.31942654e-01
-6.02065861e-01 1.84279844e-01 -4.20337111e-01 5.73097952e-02
5.50015211e-01 2.45149225e-01 -3.57873499e-01 4.72242245e-03
-9.19499338e-01 1.75829634e-01 8.92901421e-01 1.16196728e+00
9.00007188e-01 -2.99620125e-02 3.66359591e-01 7.19241500e-01
5.66100597e-01 8.95416915e-01 -1.94254920e-01 -1.60977626e+00
5.31028092e-01 2.10847914e-01 1.90241948e-01 -1.34174550e+00
-2.66198248e-01 -4.31755394e-01 -7.00537920e-01 3.44414532e-01
4.20542777e-01 -5.46266511e-03 -1.01802671e+00 1.52922153e+00
5.40306628e-01 2.49457464e-01 -2.82831699e-01 1.12238979e+00
4.29673463e-01 7.17073917e-01 -8.76751542e-01 -2.62233615e-01
9.29766774e-01 -9.88617718e-01 -4.77194101e-01 -7.17013717e-01
-1.47638366e-01 -9.81066287e-01 4.18159604e-01 8.77644181e-01
-1.42044997e+00 -3.97518426e-01 -9.72321212e-01 -2.23322406e-01
1.72270402e-01 1.66720584e-01 1.96263269e-01 2.25098819e-01
-1.01619840e+00 6.00313902e-01 -1.11683142e+00 -1.67709157e-01
3.60745758e-01 3.05817455e-01 -6.99466765e-01 -5.46426058e-01
-2.76856631e-01 9.95814025e-01 -2.58790851e-01 2.70141631e-01
-1.19163609e+00 -4.65484619e-01 -7.83371389e-01 -2.53377199e-01
6.31427467e-01 -8.76486778e-01 1.10140121e+00 -7.06873536e-01
-1.43723619e+00 1.04527843e+00 -3.09230953e-01 9.24286470e-02
5.82563400e-01 -3.93725961e-01 6.12709159e-03 4.29598868e-01
3.67215157e-01 7.46018887e-01 1.32127273e+00 -1.83849251e+00
-2.17691660e-01 -7.91673422e-01 -9.37510878e-02 4.32107657e-01
1.23554692e-01 -5.21432161e-01 -9.80404198e-01 -5.40366113e-01
1.14130306e+00 -1.19140673e+00 -2.14000136e-01 6.28564477e-01
-2.77153939e-01 2.97633499e-01 8.60124648e-01 -8.30110013e-01
6.35705471e-01 -2.08286190e+00 8.37666571e-01 2.32247725e-01
3.65824312e-01 -3.99937868e-01 -2.53244221e-01 1.88891411e-01
1.29433304e-01 -3.09358805e-01 -3.22136916e-02 -6.48408473e-01
-3.03072274e-01 4.74889308e-01 1.49409864e-02 8.49411488e-01
7.67005309e-02 4.42027301e-01 -8.46102715e-01 -1.51686668e-01
2.25499675e-01 4.76092339e-01 -8.95983934e-01 3.45054150e-01
-1.07920587e-01 5.19831419e-01 -4.01827902e-01 5.31913221e-01
1.14180553e+00 -5.66551268e-01 3.06814849e-01 -7.81879246e-01
1.35725376e-03 -3.72865126e-02 -1.59444261e+00 2.06913090e+00
-1.75195143e-01 3.14796537e-01 4.48337317e-01 -6.85059428e-01
8.23042452e-01 -1.03605494e-01 5.52471340e-01 -1.78259894e-01
7.00279474e-02 1.86157137e-01 -2.78226048e-01 -7.28766859e-01
2.01750427e-01 -1.93691701e-01 2.93434769e-01 4.35408652e-01
3.89634222e-01 -7.09740162e-01 -6.08019531e-02 3.44502598e-01
1.02688503e+00 3.66033912e-01 6.03277162e-02 -2.44532023e-02
2.67918169e-01 -2.52862722e-01 7.00709462e-01 5.06275415e-01
1.08196214e-01 1.25070584e+00 2.86142856e-01 -3.91242057e-01
-1.00245214e+00 -1.46051121e+00 2.03051895e-01 4.06058460e-01
5.64538717e-01 -1.16096310e-01 -4.58190173e-01 -8.36047590e-01
4.97539677e-02 6.25280499e-01 -5.89470685e-01 1.79583743e-01
-5.02754390e-01 -1.78144470e-01 -3.50814581e-01 4.34902042e-01
2.79689729e-01 -2.93060333e-01 -2.33717442e-01 -1.74062559e-03
-3.65521044e-01 -1.17837608e+00 -8.76732767e-01 2.42695212e-01
-9.70889688e-01 -9.97925997e-01 -7.42612362e-01 -8.42006981e-01
1.11865366e+00 5.82004786e-01 1.34757757e+00 -2.85942376e-01
-2.27693439e-01 6.74485743e-01 7.59457052e-02 -2.27397457e-02
-1.64203405e-01 -3.72285664e-01 1.75928801e-01 1.22812353e-01
1.88340455e-01 -8.81619811e-01 -7.83852398e-01 6.24638438e-01
-6.04228675e-01 1.02713369e-01 6.13722920e-01 8.67260337e-01
8.68643045e-01 -3.17984939e-01 -1.14642434e-01 -6.59848094e-01
-7.62339830e-02 -3.17327499e-01 -8.08080077e-01 1.14890002e-01
-3.36221278e-01 1.78529933e-01 -2.58292817e-02 -6.54362798e-01
-1.05980384e+00 6.31954193e-01 3.13109271e-02 -9.79208410e-01
1.68504313e-01 2.81794339e-01 -1.37751475e-01 -1.00282677e-01
5.26086211e-01 -8.52926821e-02 1.31212249e-01 -5.96387565e-01
5.21889627e-01 1.42702863e-01 7.25914061e-01 -5.36366045e-01
1.07834554e+00 9.64195967e-01 1.09461337e-01 -7.19658136e-01
-1.22262156e+00 -5.53789616e-01 -1.02927887e+00 -4.24120486e-01
1.03738976e+00 -1.08313799e+00 -5.26163340e-01 4.14963961e-01
-1.45704174e+00 -8.98441896e-02 -2.10696757e-01 5.82859337e-01
-6.01291120e-01 4.06686574e-01 -5.31780183e-01 -5.01780391e-01
3.97651829e-02 -1.20543420e+00 1.47187793e+00 -1.45592541e-01
-7.49974232e-03 -7.60118544e-01 -1.02165118e-02 5.50021231e-01
-2.03824472e-02 -1.23934090e-01 5.70805907e-01 -4.56710979e-02
-1.21150136e+00 -9.45490301e-02 -4.31556553e-02 5.48565507e-01
-3.74049544e-02 -1.86921731e-01 -7.55680144e-01 -3.89696151e-01
6.05746031e-01 -3.91589016e-01 6.64459705e-01 5.17029881e-01
8.62469435e-01 -1.81070253e-01 -2.94504821e-01 8.21231902e-01
1.45728040e+00 -2.62783188e-03 5.41660011e-01 4.15443555e-02
7.66642869e-01 5.37852407e-01 4.49115694e-01 4.13796961e-01
3.72534811e-01 6.72191024e-01 8.33252609e-01 1.27733678e-01
-1.68277964e-01 -4.04720575e-01 3.47373366e-01 1.08704782e+00
-1.22599311e-01 -3.21498334e-01 -7.44842589e-01 2.80855745e-01
-1.53007853e+00 -9.65277970e-01 -5.99317327e-02 2.15956473e+00
4.10248220e-01 4.55230698e-02 -3.63400549e-01 -4.73872811e-01
5.21406233e-01 3.12070549e-01 -8.90224576e-01 3.88847113e-01
-7.18523040e-02 -2.44302124e-01 4.64859098e-01 8.24852049e-01
-7.21180081e-01 5.64171791e-01 6.84082031e+00 3.96321833e-01
-9.60268080e-01 2.02263802e-01 5.14841795e-01 -2.19271123e-01
-3.23294103e-01 2.10532293e-01 -7.57373452e-01 2.47870646e-02
4.67511006e-02 2.41061315e-01 8.26122820e-01 7.42024779e-01
-1.24952495e-01 -3.82958949e-01 -1.32029307e+00 1.22918141e+00
6.16584301e-01 -1.18988717e+00 3.56535017e-02 2.46113494e-01
1.10733795e+00 2.68955261e-01 1.60751134e-01 -2.56164223e-01
3.88583034e-01 -5.17007589e-01 9.99307692e-01 8.10303211e-01
6.37834072e-01 -2.74586052e-01 4.26194429e-01 5.86694181e-01
-8.72067213e-01 1.11786582e-01 -4.83301014e-01 2.32154429e-01
6.16986454e-01 6.38631403e-01 -5.83554268e-01 3.92673135e-01
8.77153575e-01 1.09528780e+00 -3.49423498e-01 5.73383629e-01
-3.01935226e-01 1.25779986e-01 -6.45014584e-01 5.04709423e-01
-4.01688702e-02 -6.97493255e-01 7.63397276e-01 4.58566904e-01
7.19403446e-01 2.92860448e-01 4.52400416e-01 9.99920011e-01
-1.14541024e-01 -6.28157377e-01 -6.11586988e-01 3.46663445e-01
3.45288724e-01 1.30964088e+00 -7.51327991e-01 -2.49631435e-01
-5.88776112e-01 1.41716933e+00 2.79145688e-01 4.61850584e-01
-4.71462935e-01 4.80663478e-01 4.73991811e-01 4.19407219e-01
7.37145245e-01 -5.92200339e-01 -1.48920864e-01 -1.26828003e+00
9.34852064e-02 -9.12631273e-01 7.99767114e-03 -1.35448527e+00
-1.80878413e+00 3.92883211e-01 4.54730950e-02 -1.56176925e+00
-2.25177065e-01 -7.61773467e-01 -1.27464309e-01 8.48493218e-01
-9.51737106e-01 -9.38632011e-01 -4.72997129e-01 3.01848978e-01
7.27430820e-01 7.07056671e-02 4.98858869e-01 7.16403127e-02
-6.30885586e-02 -2.00329989e-01 1.08253449e-01 -7.68327340e-02
7.99565613e-01 -9.97259498e-01 3.54067922e-01 7.35124886e-01
2.19813943e-01 4.59809691e-01 6.34178877e-01 -7.22355962e-01
-1.84184694e+00 -6.29494548e-01 3.91940236e-01 -6.88410878e-01
5.27553916e-01 -4.01312649e-01 -5.94483018e-01 8.47856581e-01
1.73293605e-01 3.36564749e-01 1.01992615e-01 -2.12787867e-01
-3.58790070e-01 -8.00402611e-02 -1.03286338e+00 4.00684416e-01
1.30878222e+00 -6.64051831e-01 -5.63596725e-01 3.37372303e-01
2.85152525e-01 -9.16287780e-01 -7.73411214e-01 2.03110665e-01
5.68920195e-01 -1.16268909e+00 1.27019548e+00 -9.33671091e-03
5.63595951e-01 -4.05554414e-01 -4.27925229e-01 -1.32151830e+00
-4.91553843e-01 -4.45925802e-01 -1.04965918e-01 7.65683651e-01
2.50728041e-01 -4.17231947e-01 9.51018453e-01 6.04257822e-01
3.08835749e-02 -8.34997416e-01 -8.90079916e-01 -7.98312366e-01
-2.20173076e-01 -2.35217229e-01 1.25002312e-02 7.71623552e-01
-7.58615732e-01 6.84266925e-01 -6.47998810e-01 7.23081052e-01
1.09329915e+00 1.63704902e-01 1.04880667e+00 -1.19929457e+00
-7.17425108e-01 5.38933948e-02 -4.64657664e-01 -1.91781735e+00
-3.62926559e-03 -6.58686280e-01 2.90868044e-01 -1.41472042e+00
3.25633168e-01 -1.89460553e-02 4.98061448e-01 1.09450854e-01
4.35507037e-02 3.08853835e-01 2.10537657e-01 3.33881199e-01
-4.69031394e-01 6.33051097e-01 1.42755544e+00 -3.95946890e-01
8.46406817e-02 -9.92071256e-02 -3.36110741e-01 1.23280752e+00
1.40643986e-02 -7.75361538e-01 -3.68319631e-01 -9.19485867e-01
3.98685604e-01 3.75338316e-01 3.75657052e-01 -8.45824182e-01
4.12321359e-01 -1.37380838e-01 7.27904201e-01 -9.93319333e-01
9.69467342e-01 -1.21258879e+00 4.96772289e-01 2.07881585e-01
-1.43570036e-01 2.73445666e-01 -3.80996019e-01 1.06895852e+00
-5.52198924e-02 -4.07942712e-01 6.94839060e-01 -4.06716317e-01
-4.78999346e-01 5.60080707e-01 -4.95120184e-03 1.39770165e-01
8.42772603e-01 -3.16261828e-01 -5.94741665e-02 -7.14739561e-01
-1.03832269e+00 1.81070298e-01 7.88066387e-01 1.66198179e-01
8.95130157e-01 -1.32356048e+00 -5.82204223e-01 4.41127390e-01
-1.40891925e-01 4.52317119e-01 2.62997657e-01 7.45599806e-01
-6.48531854e-01 -3.96290928e-01 1.38544619e-01 -1.23868358e+00
-1.41333246e+00 4.51965153e-01 3.14785242e-01 -1.00941442e-01
-6.23514473e-01 1.07707405e+00 3.64159554e-01 -5.34434140e-01
1.04366004e-01 -8.48910511e-02 3.98312360e-01 -8.36320743e-02
2.93125570e-01 4.41159159e-01 -1.24818087e-01 -8.25958312e-01
-3.16895455e-01 1.15816832e+00 -1.91289380e-01 -5.00820339e-01
1.65124309e+00 -3.88287961e-01 -2.14360744e-01 5.20378351e-01
1.45991337e+00 9.80222598e-02 -1.94666922e+00 -3.93653035e-01
-2.89917529e-01 -8.60022128e-01 2.11776942e-01 -5.23482144e-01
-1.22588694e+00 8.52899194e-01 5.98286092e-01 -3.17584723e-01
8.86276007e-01 2.52025366e-01 3.94027740e-01 4.83986914e-01
6.60057783e-01 -8.81160021e-01 7.61646092e-01 5.72131574e-01
1.23991549e+00 -1.28945935e+00 5.68878114e-01 -7.15728760e-01
-5.74664116e-01 1.16797042e+00 7.36871541e-01 -2.02010304e-01
8.80367696e-01 3.90304685e-01 2.16982976e-01 -4.21498507e-01
-6.20698035e-01 1.99789315e-01 4.21993941e-01 5.10914385e-01
-9.30974036e-02 -4.34478790e-01 3.46902907e-01 3.30989398e-02
1.55367523e-01 -1.67047620e-01 2.82538027e-01 9.21113551e-01
-1.97727501e-01 -8.59854341e-01 -7.05630124e-01 3.62270832e-01
-3.41982730e-02 1.87033236e-01 -3.61828923e-01 4.36923116e-01
-9.67699196e-03 7.32945919e-01 9.16139260e-02 -2.73702979e-01
4.05670345e-01 -2.11908415e-01 9.28668201e-01 -6.64661884e-01
-7.21548160e-04 4.51144218e-01 -1.51136726e-01 -8.62422526e-01
-6.23221576e-01 -9.30152833e-01 -8.75568807e-01 -6.94844648e-02
-6.30800366e-01 -3.36555600e-01 6.11061633e-01 7.80206680e-01
3.97047937e-01 1.19207591e-01 7.48138905e-01 -1.68976223e+00
-8.85493338e-01 -7.20821261e-01 -6.28310323e-01 5.76660454e-01
3.16190064e-01 -8.73025119e-01 -6.54527724e-01 4.00801420e-01] | [8.4166841506958, -2.822279930114746] |
39e840af-b98a-4c13-9414-9101bf7d19e8 | cubeslam-monocular-3d-object-detection-and | 1806.00557 | null | http://arxiv.org/abs/1806.00557v2 | http://arxiv.org/pdf/1806.00557v2.pdf | CubeSLAM: Monocular 3D Object SLAM | We present a method for single image 3D cuboid object detection and
multi-view object SLAM in both static and dynamic environments, and demonstrate
that the two parts can improve each other. Firstly for single image object
detection, we generate high-quality cuboid proposals from 2D bounding boxes and
vanishing points sampling. The proposals are further scored and selected based
on the alignment with image edges. Secondly, multi-view bundle adjustment with
new object measurements is proposed to jointly optimize poses of cameras,
objects and points. Objects can provide long-range geometric and scale
constraints to improve camera pose estimation and reduce monocular drift.
Instead of treating dynamic regions as outliers, we utilize object
representation and motion model constraints to improve the camera pose
estimation. The 3D detection experiments on SUN RGBD and KITTI show better
accuracy and robustness over existing approaches. On the public TUM, KITTI
odometry and our own collected datasets, our SLAM method achieves the
state-of-the-art monocular camera pose estimation and at the same time,
improves the 3D object detection accuracy. | ['Shichao Yang', 'Sebastian Scherer'] | 2018-06-01 | null | null | null | null | ['object-slam'] | ['computer-vision'] | [-3.60538214e-01 -3.10378999e-01 -2.64042675e-01 -3.02761912e-01
-5.93421221e-01 -7.97514260e-01 2.65033364e-01 -2.82919824e-01
-4.34648752e-01 1.95472464e-01 -2.08070755e-01 2.47149467e-01
3.74175668e-01 -3.50493193e-01 -8.89008999e-01 -5.03293633e-01
3.86129111e-01 9.65966105e-01 7.11564660e-01 6.50875717e-02
4.41093445e-01 8.66772532e-01 -1.47046459e+00 -2.14070454e-01
5.83763540e-01 9.20417428e-01 3.87839854e-01 6.84701264e-01
2.38774836e-01 3.46420884e-01 -2.48720631e-01 -1.70194596e-01
6.50940776e-01 2.50022769e-01 -1.84431151e-01 7.63562143e-01
8.71357501e-01 -6.29477978e-01 -3.43725801e-01 1.12593520e+00
4.70178038e-01 -2.02134386e-01 3.13016087e-01 -1.28775048e+00
-2.79002786e-01 -1.91659823e-01 -1.16743565e+00 -2.03600630e-01
8.64885807e-01 2.19129696e-01 6.85452044e-01 -1.44377792e+00
8.43504190e-01 1.36491418e+00 7.75915205e-01 2.84207880e-01
-1.13817537e+00 -6.84628546e-01 3.22295964e-01 -2.05381885e-02
-1.66952538e+00 -4.34739292e-01 6.46842062e-01 -4.17741030e-01
9.16080773e-01 6.84398264e-02 8.47825766e-01 7.28764296e-01
-3.52574550e-02 7.99947381e-01 8.57746542e-01 -3.46397907e-01
-1.36296764e-01 4.13796246e-01 -6.97900727e-02 9.73250270e-01
6.58208489e-01 2.86210328e-02 -6.71872199e-01 -1.52992621e-01
8.20757389e-01 4.23985153e-01 -2.12133333e-01 -1.34246194e+00
-1.59158003e+00 5.22908747e-01 3.98693025e-01 -4.96102750e-01
-1.72416806e-01 7.04665929e-02 6.65254071e-02 -2.94271499e-01
2.94230461e-01 6.55793697e-02 -3.06911290e-01 1.73152670e-01
-5.53138256e-01 1.81728721e-01 4.58869219e-01 1.77365291e+00
1.03983247e+00 -1.71750948e-01 3.74854416e-01 5.93034625e-01
5.42834461e-01 1.12554145e+00 2.32053734e-02 -1.01199365e+00
5.66908836e-01 9.68949437e-01 4.06469852e-01 -1.07145822e+00
-3.84207547e-01 -2.38264054e-02 -3.82999212e-01 2.32646152e-01
2.00227529e-01 2.92935818e-01 -9.34328377e-01 9.26684916e-01
8.82826924e-01 -1.32709384e-01 -1.87321663e-01 1.28483486e+00
6.65534317e-01 3.75269771e-01 -8.33856702e-01 -7.25709870e-02
1.33430815e+00 -8.78014863e-01 -3.90105247e-01 -6.65037513e-01
4.57875311e-01 -9.98242557e-01 7.21071720e-01 3.97823125e-01
-1.00845706e+00 -3.90571445e-01 -1.12845480e+00 -2.10151136e-01
2.46526133e-02 5.26095510e-01 5.00469744e-01 5.14192045e-01
-6.57773077e-01 -1.69960707e-01 -8.93202484e-01 -4.78271991e-01
1.51370361e-01 3.63344073e-01 -5.36457002e-01 -4.19249862e-01
-3.44351441e-01 1.01221931e+00 4.14201379e-01 2.53174510e-02
-1.00346959e+00 -4.06471610e-01 -1.02929556e+00 -4.63470966e-01
6.91134393e-01 -7.55573690e-01 9.75045860e-01 -2.75874734e-01
-1.29557776e+00 1.31096721e+00 -3.06260496e-01 -1.08375832e-01
6.46872938e-01 -4.00426984e-01 1.01879179e-01 1.88557938e-01
2.87882477e-01 7.28156626e-01 7.01542258e-01 -1.46426952e+00
-9.82108176e-01 -8.48186910e-01 -2.87667274e-01 8.11041534e-01
1.82958513e-01 -9.68869254e-02 -1.11431885e+00 7.16367811e-02
1.13913119e+00 -1.30268526e+00 -3.16677123e-01 2.02158824e-01
-3.44720900e-01 1.53333485e-01 8.92037451e-01 -2.48209104e-01
4.82544661e-01 -2.06716180e+00 2.87380189e-01 9.80123058e-02
9.84171852e-02 -3.13760608e-01 2.73191661e-01 5.11731301e-03
4.77169394e-01 -3.69277179e-01 4.04425651e-01 -6.87679350e-01
-2.51454681e-01 2.09342584e-01 -1.54859480e-02 1.19474006e+00
-1.44171923e-01 5.51157832e-01 -7.35690653e-01 -4.73126084e-01
5.25999725e-01 1.65318653e-01 -5.61352253e-01 1.25359222e-01
-9.44021195e-02 2.93541044e-01 -2.51107067e-01 1.16499960e+00
1.03968751e+00 -1.20547980e-01 -6.95003197e-02 -3.87195230e-01
-2.36657530e-01 1.35372296e-01 -1.84716547e+00 1.81243992e+00
-8.29588771e-02 3.70396733e-01 2.03424737e-01 -2.23017484e-01
1.01564944e+00 -1.65920675e-01 5.03726006e-01 -1.95584863e-01
-1.07537135e-02 2.89353430e-01 -4.39772010e-01 -2.17870221e-01
8.61231089e-01 3.61413717e-01 -3.49129587e-02 1.82999864e-01
-5.74132167e-02 -7.33944654e-01 -8.31751525e-02 2.36123264e-01
5.33386648e-01 4.26120758e-01 5.48423171e-01 -3.01584844e-02
3.93101156e-01 3.29064548e-01 7.93958306e-01 5.73143899e-01
-1.80178463e-01 8.71320784e-01 -7.81875625e-02 -5.47328711e-01
-1.32838893e+00 -1.04336941e+00 -2.36752138e-01 5.07701874e-01
9.82883990e-01 -1.99285641e-01 -3.27297777e-01 -6.30242884e-01
4.82249260e-01 1.45418227e-01 -1.58448040e-01 9.45935920e-02
-5.71328878e-01 -5.97426295e-01 -2.29964219e-02 3.50445002e-01
3.07107657e-01 -3.05667907e-01 -8.27050745e-01 -3.28867994e-02
-1.21101856e-01 -1.55276036e+00 -7.27308333e-01 1.07283801e-01
-8.79117072e-01 -1.26343024e+00 -4.61618453e-01 -5.69488466e-01
9.41494882e-01 1.22935259e+00 8.55167687e-01 -1.60127848e-01
-4.10621375e-01 6.18657351e-01 -2.50089109e-01 -2.91845053e-01
-7.39099681e-02 -2.54154444e-01 6.78867042e-01 -8.43928754e-02
4.76836652e-01 -1.37296394e-01 -6.70220196e-01 8.73137295e-01
-2.27476299e-01 1.48439765e-01 5.01887202e-01 6.33604109e-01
8.65816176e-01 -4.10913736e-01 -3.67004126e-01 -3.17979306e-01
-4.45125043e-01 1.55476540e-01 -1.32910860e+00 -1.54824322e-02
-6.68127596e-01 -3.35800320e-01 -2.24136487e-01 -4.51591909e-01
-7.85309076e-01 8.59085202e-01 6.21626675e-01 -9.42007542e-01
-7.65654519e-02 -2.56643474e-01 -2.25751132e-01 -5.86753309e-01
6.05388224e-01 2.59588748e-01 3.35332118e-02 -3.84771764e-01
4.47738141e-01 6.62813187e-01 4.42836702e-01 -2.31558099e-01
1.03290355e+00 1.08183658e+00 -2.12359037e-02 -7.62726307e-01
-8.37179303e-01 -1.08845651e+00 -1.15800607e+00 -4.09802318e-01
9.32912111e-01 -1.54847813e+00 -8.60996127e-01 3.92410278e-01
-1.33842397e+00 2.77206630e-01 9.31240693e-02 8.28128278e-01
-5.14747381e-01 5.65310538e-01 -3.40785056e-01 -1.02811337e+00
-9.86739546e-02 -1.46760190e+00 1.67958462e+00 6.90166131e-02
3.64583045e-01 -4.15373117e-01 -2.14662358e-01 4.50776279e-01
-4.31096673e-01 1.23039156e-01 -1.70859899e-02 -1.14629250e-02
-1.38733757e+00 -4.96548593e-01 -1.85698226e-01 -5.77694252e-02
-2.24925607e-01 2.20006611e-02 -7.97391891e-01 -6.38238251e-01
1.10648699e-01 -1.39426455e-01 6.10118568e-01 3.34946513e-01
4.80823636e-01 1.58665404e-01 -7.18299866e-01 1.01164198e+00
1.48805296e+00 1.36634782e-01 1.99771300e-01 5.95099688e-01
9.61082339e-01 2.66861498e-01 1.20750487e+00 5.42347670e-01
5.65896273e-01 1.02252805e+00 8.60079885e-01 -5.50943203e-02
1.65325418e-01 -3.19102973e-01 5.27335763e-01 5.59847116e-01
1.01000719e-01 2.95476824e-01 -1.07209086e+00 4.04646546e-01
-1.86908495e+00 -5.78105688e-01 -5.09901524e-01 2.38015032e+00
2.22618207e-01 2.25860953e-01 2.50096917e-01 -3.30815345e-01
8.14067602e-01 -8.60061795e-02 -6.63822174e-01 4.79321152e-01
-2.51219541e-01 -9.06165957e-01 1.28000176e+00 6.17775440e-01
-9.60591376e-01 1.21802354e+00 5.84623432e+00 1.06660508e-01
-7.59087741e-01 7.32167885e-02 -2.44763158e-02 -3.93661857e-01
1.46298975e-01 3.41204196e-01 -1.76239491e+00 5.56470007e-02
-1.06306061e-01 2.73963690e-01 4.00314957e-01 1.26137877e+00
-7.50118420e-02 -4.54873264e-01 -1.05435133e+00 1.56170988e+00
5.61113477e-01 -1.20160604e+00 -2.35749424e-01 3.75358939e-01
9.08993959e-01 4.66804296e-01 -2.32847884e-01 -1.39283687e-01
2.78724402e-01 -2.41761133e-01 1.19249117e+00 2.00244486e-01
4.54700351e-01 -6.69944406e-01 5.76858938e-01 6.35708690e-01
-1.33952081e+00 -1.49797708e-01 -8.61000359e-01 1.04521448e-02
3.81938487e-01 5.38758099e-01 -1.21814191e+00 5.58261752e-01
9.45505440e-01 9.10647511e-01 -8.18871975e-01 1.14887261e+00
-1.52578846e-01 -2.86483496e-01 -6.78318501e-01 1.41163338e-02
7.33071789e-02 -5.39275467e-01 9.99326468e-01 7.07902968e-01
4.12571371e-01 1.27441036e-02 6.36719048e-01 7.08870828e-01
1.89956769e-01 -1.04692087e-01 -8.71510565e-01 7.65784204e-01
6.99102879e-01 1.30544937e+00 -6.67462289e-01 -2.85506815e-01
-4.78929460e-01 9.97120619e-01 2.15277165e-01 1.34692594e-01
-9.04021502e-01 8.16105008e-02 7.13039637e-01 2.34475702e-01
3.52014691e-01 -7.10313439e-01 -2.19065458e-01 -1.72896278e+00
3.57818991e-01 -6.75006032e-01 2.18897164e-01 -1.05880177e+00
-7.75690317e-01 2.79198855e-01 6.11597523e-02 -1.63746643e+00
2.00438853e-02 -7.89151251e-01 -2.58312933e-02 4.46891755e-01
-1.18355381e+00 -1.34736574e+00 -7.13077068e-01 4.85507995e-01
6.02973700e-01 -5.95016032e-02 2.70530790e-01 1.62831515e-01
-3.28596652e-01 8.60718414e-02 1.20587647e-01 -1.30116865e-01
9.06902313e-01 -1.08930588e+00 3.08534473e-01 9.78542745e-01
3.86457026e-01 4.17124987e-01 4.86206263e-01 -8.07419956e-01
-1.94668555e+00 -8.78206730e-01 3.93008858e-01 -1.09226000e+00
2.78065771e-01 -9.89904642e-01 -4.81137812e-01 8.55070472e-01
-3.59089404e-01 1.70103684e-01 -2.63008811e-02 -2.37359595e-03
-2.09319398e-01 -2.86630452e-01 -9.53933716e-01 4.58740294e-01
1.27282035e+00 -2.78431922e-01 -3.23885113e-01 6.27008915e-01
6.88345373e-01 -1.16524780e+00 -5.72106898e-01 5.34749866e-01
6.56511724e-01 -1.02042174e+00 1.28861296e+00 -8.26843902e-02
-4.99503404e-01 -9.05228496e-01 -4.29772645e-01 -8.17790151e-01
-1.14338145e-01 -5.92948079e-01 -1.81398727e-02 1.03746569e+00
7.05601126e-02 -4.20127511e-01 1.11181891e+00 4.02181119e-01
-9.94614139e-02 -2.34193802e-01 -1.09044075e+00 -9.03995097e-01
-7.47148275e-01 -3.10890228e-01 3.87949973e-01 6.22198761e-01
-3.48696530e-01 3.95295024e-01 -6.02364123e-01 7.87895739e-01
1.12041700e+00 5.12741327e-01 1.63890946e+00 -1.26801443e+00
-5.79941981e-02 6.33702129e-02 -6.93705380e-01 -1.55929768e+00
-1.37054190e-01 -6.18715525e-01 2.26065040e-01 -1.10915649e+00
4.45817113e-01 -3.42997998e-01 4.82371181e-01 1.41182011e-02
-7.02275038e-02 4.29622054e-01 3.83156717e-01 5.19008219e-01
-8.95963907e-01 4.21357423e-01 1.03804851e+00 1.86340019e-01
-2.45838553e-01 8.17804709e-02 -5.78669161e-02 8.93513024e-01
2.37615585e-01 -6.16827130e-01 3.15300971e-02 -5.80546319e-01
1.89605340e-01 1.09674416e-01 6.00313663e-01 -1.05256438e+00
3.84132415e-01 -1.90329239e-01 7.54773617e-01 -1.40680611e+00
7.11010814e-01 -1.01870811e+00 1.59302577e-01 6.18227363e-01
3.88119727e-01 2.39306703e-01 -2.66160034e-02 8.26829791e-01
1.92674994e-01 -1.29869655e-01 1.02350104e+00 -4.18753088e-01
-8.15067768e-01 4.94484961e-01 4.20563519e-02 -3.12638283e-01
1.24979639e+00 -6.00974262e-01 -8.26147273e-02 -2.71171838e-01
-4.78205442e-01 5.01844645e-01 1.21757138e+00 5.33878922e-01
8.75146091e-01 -1.45248210e+00 -5.06107390e-01 5.56821525e-01
5.40778399e-01 6.41960263e-01 -1.39058962e-01 1.11931050e+00
-6.94295228e-01 4.07669604e-01 1.62575692e-01 -1.49653542e+00
-1.56597912e+00 6.36665404e-01 4.07482505e-01 3.86804819e-01
-6.49133146e-01 6.66925788e-01 2.49008104e-01 -6.31895781e-01
3.61096263e-01 -4.69286233e-01 3.67899925e-01 -1.40638754e-01
3.19666475e-01 4.02528554e-01 -7.66204372e-02 -8.24965596e-01
-6.24483585e-01 1.14148188e+00 -9.52179879e-02 -4.75483462e-02
1.15936983e+00 -7.69533098e-01 -7.77935684e-02 4.25899625e-01
8.21753919e-01 1.21185847e-01 -1.54007506e+00 -2.86857814e-01
-9.01839063e-02 -1.04342413e+00 -4.39588949e-02 -2.81061530e-01
-5.84050119e-01 7.15828240e-01 7.48927236e-01 -3.12108934e-01
6.09498620e-01 1.99474916e-01 2.85419106e-01 6.57202661e-01
8.49761963e-01 -1.06771195e+00 2.91909069e-01 6.19331300e-01
7.12936401e-01 -1.64386094e+00 4.52123851e-01 -6.56906247e-01
-6.98961020e-01 1.18382215e+00 9.42226112e-01 -2.94356465e-01
1.19658262e-01 1.75296143e-01 5.70882410e-02 -1.63431093e-01
-2.69569039e-01 -2.62969255e-01 3.02498996e-01 5.83578527e-01
-2.29258537e-01 -2.55080044e-01 3.69511276e-01 -2.48721600e-01
1.85471982e-01 -4.90046382e-01 5.91634393e-01 7.46412694e-01
-7.39780128e-01 -7.13518620e-01 -9.90613878e-01 3.24727409e-02
-2.07968019e-02 2.73382813e-01 -4.74872082e-01 1.10662270e+00
-1.88710883e-01 5.85910499e-01 5.33884726e-02 -3.12367201e-01
5.55200338e-01 -1.38312280e-01 6.68261588e-01 -7.40807235e-01
-9.35514569e-02 5.92783451e-01 -5.95410801e-02 -7.48024285e-01
-2.68246412e-01 -9.53621030e-01 -1.01177227e+00 -8.48499835e-02
-9.92759287e-01 -2.93141723e-01 1.00289369e+00 5.34331739e-01
3.59441757e-01 -2.01799974e-01 7.91327596e-01 -1.44819188e+00
-6.97855830e-01 -7.15765238e-01 -6.33516312e-01 3.18691820e-01
4.33278233e-01 -8.36426497e-01 -4.90055144e-01 -5.42982593e-02] | [7.3649678230285645, -2.3587968349456787] |
1d028ff0-e2a5-479a-bf0e-67b813e959d9 | hybrid-curriculum-learning-for-emotion | 2112.11718 | null | https://arxiv.org/abs/2112.11718v2 | https://arxiv.org/pdf/2112.11718v2.pdf | Hybrid Curriculum Learning for Emotion Recognition in Conversation | Emotion recognition in conversation (ERC) aims to detect the emotion label for each utterance. Motivated by recent studies which have proven that feeding training examples in a meaningful order rather than considering them randomly can boost the performance of models, we propose an ERC-oriented hybrid curriculum learning framework. Our framework consists of two curricula: (1) conversation-level curriculum (CC); and (2) utterance-level curriculum (UC). In CC, we construct a difficulty measurer based on "emotion shift" frequency within a conversation, then the conversations are scheduled in an "easy to hard" schema according to the difficulty score returned by the difficulty measurer. For UC, it is implemented from an emotion-similarity perspective, which progressively strengthens the model's ability in identifying the confusing emotions. With the proposed model-agnostic hybrid curriculum learning strategy, we observe significant performance boosts over a wide range of existing ERC models and we are able to achieve new state-of-the-art results on four public ERC datasets. | ['Longjun Cai', 'Yue Mao', 'Yi Shen', 'Lin Yang'] | 2021-12-22 | null | null | null | null | ['emotion-recognition-in-conversation'] | ['natural-language-processing'] | [ 2.66145378e-01 1.46058947e-01 4.95500714e-02 -7.61457503e-01
-8.26194704e-01 -4.02174413e-01 4.54491407e-01 3.09964806e-01
-3.92560571e-01 2.06845716e-01 4.35292572e-01 -2.11167857e-01
4.55971062e-02 -3.79441798e-01 -3.85899067e-01 -6.27364874e-01
1.13913171e-01 4.15790975e-01 -2.22983152e-01 -4.80056107e-01
2.62885123e-01 1.45713031e-01 -1.86041689e+00 7.63072670e-01
1.16865814e+00 1.22807062e+00 7.06497952e-02 6.30684197e-01
-4.48609531e-01 1.03378022e+00 -7.02461541e-01 -5.19975662e-01
-4.12274152e-01 -6.33157909e-01 -1.00205481e+00 2.07656980e-01
-3.82133685e-02 1.14130564e-01 2.26166129e-01 7.97562659e-01
4.10575211e-01 3.89188200e-01 6.51151538e-01 -1.31596899e+00
-2.69012421e-01 8.09800744e-01 -2.44957343e-01 -1.89245015e-01
5.47052443e-01 -3.31640482e-01 1.02306652e+00 -7.47869313e-01
5.74807763e-01 1.26516342e+00 4.17582452e-01 1.00413704e+00
-1.06764984e+00 -7.43130684e-01 7.22949982e-01 3.59223127e-01
-8.06908309e-01 -1.94827557e-01 1.20860589e+00 -3.12507987e-01
1.00224066e+00 3.97624731e-01 6.23513639e-01 1.41721022e+00
-1.07428618e-01 1.09419692e+00 1.40756285e+00 -6.24386668e-01
4.52401876e-01 6.17424488e-01 3.29995632e-01 1.60505190e-01
-7.48710394e-01 -9.46664251e-03 -6.32139802e-01 9.05721486e-02
-1.71373710e-01 -1.84708625e-01 -3.76641393e-01 -2.45611414e-01
-6.56456292e-01 1.05023718e+00 1.15790837e-01 2.68106401e-01
-2.95226544e-01 -1.40694082e-01 8.22015047e-01 6.55128717e-01
7.22820818e-01 5.47747552e-01 -7.07002163e-01 -6.98549032e-01
-4.24286067e-01 2.38327935e-01 1.02795911e+00 8.95135880e-01
4.30120975e-01 -1.59708723e-01 -1.92048013e-01 1.35788310e+00
5.48668429e-02 -5.16994968e-02 5.92045367e-01 -8.59143913e-01
2.23121107e-01 7.08575904e-01 -2.09228814e-01 -8.55338454e-01
-4.27789450e-01 -3.97662818e-01 -6.42565131e-01 -1.57230809e-01
-1.03364453e-01 -4.71001506e-01 -4.39312607e-01 2.08327150e+00
3.73674035e-01 4.81394231e-01 4.16216135e-01 6.77989066e-01
9.86995637e-01 5.38665473e-01 5.14614284e-01 -4.63270605e-01
1.54112077e+00 -1.16050851e+00 -9.22201812e-01 -8.11732113e-02
8.58112752e-01 -7.36307085e-01 1.22780287e+00 7.37956464e-01
-8.28466713e-01 -5.73140979e-01 -8.66755545e-01 9.70541909e-02
-5.50516486e-01 -8.37784186e-02 6.85392141e-01 8.59403431e-01
-9.16461349e-01 2.37380341e-01 -2.04966754e-01 -2.00099424e-01
-1.74916759e-01 -5.42957149e-02 2.06621289e-02 -8.71004164e-02
-1.38804138e+00 9.43573415e-01 1.54857576e-01 -9.97579396e-02
-7.85090923e-01 -7.49775231e-01 -9.76246953e-01 3.25966805e-01
3.24653774e-01 -1.68403149e-01 1.43085170e+00 -1.34265769e+00
-2.00353956e+00 7.20018387e-01 -4.27864715e-02 -2.10173413e-01
2.84475952e-01 -1.52733535e-01 -5.60596347e-01 2.15148002e-01
-3.93529624e-01 8.86158288e-01 6.10374570e-01 -1.59887362e+00
-8.66526186e-01 7.26326322e-03 2.33444944e-01 5.53634882e-01
-6.58522785e-01 2.29315922e-01 -3.48098338e-01 -4.20167118e-01
-9.92633477e-02 -8.44909966e-01 -1.87595174e-01 -7.24737346e-01
6.07076399e-02 -8.07166755e-01 7.53742993e-01 -3.62854213e-01
1.50028086e+00 -2.17682719e+00 3.80325019e-01 2.59438992e-01
1.06821563e-02 1.91480343e-04 -3.56496394e-01 3.53280932e-01
-4.61961269e-01 7.92714208e-02 7.94332251e-02 -7.62292683e-01
2.96156555e-01 6.13767020e-02 -1.27087727e-01 -1.75252452e-01
3.55730236e-01 3.58038038e-01 -1.00972998e+00 -3.43220919e-01
1.31138176e-01 3.92508179e-01 -9.40460503e-01 6.95547521e-01
-2.01815650e-01 3.91142070e-01 -2.35216439e-01 3.81320655e-01
4.18263525e-01 6.06871955e-02 4.48321283e-01 2.54389830e-02
1.61836758e-01 3.67386013e-01 -1.06062376e+00 1.49526954e+00
-8.73064518e-01 3.27405244e-01 6.65998906e-02 -1.18448448e+00
1.22473919e+00 6.93770230e-01 4.50248033e-01 -7.16733038e-01
1.72419444e-01 -1.87300101e-01 2.14937497e-02 -5.90389907e-01
7.68769383e-01 -1.76834360e-01 -5.11600494e-01 3.62318099e-01
2.68840313e-01 -1.78478792e-01 1.42969280e-01 2.50803292e-01
9.90863502e-01 -3.57385688e-02 -2.23374795e-02 -2.91935951e-01
7.20539093e-01 -3.11258942e-01 6.28590405e-01 5.13885200e-01
-4.95101005e-01 6.11701980e-02 7.24201858e-01 -1.62565574e-01
-4.24830824e-01 -5.64629555e-01 -9.98629443e-03 1.66722488e+00
-4.75034416e-02 -5.14591753e-01 -9.68142807e-01 -8.59893084e-01
-3.54636818e-01 9.29355979e-01 -7.65739620e-01 -5.51179647e-01
-2.62554139e-01 -4.21536148e-01 2.19654739e-01 4.92559701e-01
2.94767231e-01 -1.30457139e+00 -4.05826539e-01 3.59687060e-01
-6.05154455e-01 -1.23563397e+00 -4.21098888e-01 5.13855338e-01
-4.74558681e-01 -7.31542408e-01 -2.01773360e-01 -9.09496605e-01
5.10347903e-01 1.93062529e-01 1.41629434e+00 1.75370663e-01
-5.84592950e-03 7.01015472e-01 -9.57804084e-01 -6.32539392e-01
-4.34821248e-01 -8.43124911e-02 -1.22702122e-01 2.37098411e-01
6.96959317e-01 -4.50716257e-01 -5.08972645e-01 2.53101557e-01
-6.67282879e-01 7.91899785e-02 2.63346225e-01 7.65883029e-01
2.16719151e-01 1.95869371e-01 1.04159558e+00 -9.44491327e-01
1.11982989e+00 -6.17117643e-01 -1.60517171e-02 2.34051734e-01
-6.80232584e-01 -2.64051855e-01 6.05804563e-01 -7.95752108e-01
-1.44961333e+00 -6.28785565e-02 -4.05255944e-01 -3.94248933e-01
-4.65205699e-01 5.70486128e-01 -6.54426217e-02 2.20392570e-01
2.95981973e-01 1.74715325e-01 -2.48182014e-01 -2.40344405e-01
3.12829673e-01 8.50112140e-01 3.10471177e-01 -1.28490376e+00
1.21233128e-01 -2.63002217e-01 -7.97571301e-01 -5.93682528e-01
-1.00284469e+00 -7.17364073e-01 -7.70710483e-02 -8.27115774e-01
1.04860234e+00 -9.98075962e-01 -1.14074326e+00 2.35022634e-01
-1.00675893e+00 -5.04645765e-01 -1.77396424e-02 4.15258765e-01
-5.14995873e-01 -7.10417405e-02 -7.96455324e-01 -1.17872858e+00
-2.15820208e-01 -1.38874567e+00 9.82072830e-01 5.19512534e-01
-5.15468776e-01 -1.08879042e+00 1.03689365e-01 4.99958873e-01
4.81143147e-01 1.29027933e-01 1.27947855e+00 -9.75306749e-01
1.60883084e-01 -3.12275197e-02 1.71197593e-01 5.25821030e-01
-3.23774099e-01 -1.90037683e-01 -1.36831594e+00 -8.82384256e-02
1.88623846e-01 -8.70832145e-01 5.11887133e-01 -2.19803788e-02
1.60714126e+00 7.87688270e-02 6.17317855e-03 1.17329724e-01
1.16672599e+00 5.34324288e-01 7.74773240e-01 1.06211126e-01
1.58444226e-01 1.00832832e+00 6.91785276e-01 4.00145084e-01
7.22753108e-01 5.94815016e-01 1.52989775e-01 -1.61615804e-01
3.11806977e-01 -1.52768061e-01 5.19686282e-01 1.19225550e+00
9.71127525e-02 -1.56867415e-01 -6.11195505e-01 4.79792982e-01
-1.76290679e+00 -7.85032868e-01 2.66077936e-01 1.77630103e+00
1.16887856e+00 1.32809341e-01 -3.43591608e-02 6.43804073e-02
4.07886982e-01 7.19750151e-02 -2.57533997e-01 -1.18318462e+00
2.80223370e-01 3.03205967e-01 -4.37203944e-01 4.95715410e-01
-9.31886792e-01 9.55904663e-01 5.90193701e+00 7.68966556e-01
-1.01960671e+00 -3.29233520e-02 1.02713168e+00 3.62970643e-02
-4.61027980e-01 -3.50039214e-01 -7.44924188e-01 4.04094636e-01
1.25914872e+00 -3.58723216e-02 2.97670186e-01 1.12566733e+00
1.30072728e-01 4.69331257e-02 -1.26023376e+00 8.94784570e-01
2.57160246e-01 -9.38761830e-01 -1.65976793e-01 -2.35829800e-01
6.11418307e-01 -5.15969276e-01 -8.02058578e-02 1.39162421e+00
4.63599861e-01 -8.41199279e-01 4.73613471e-01 2.98463643e-01
2.83519715e-01 -1.17708182e+00 9.45051312e-01 2.56470472e-01
-1.05409110e+00 -4.45371680e-02 -1.99058011e-01 1.71955600e-02
-2.30944276e-01 2.28759512e-01 -6.21596813e-01 5.62889636e-01
8.23150635e-01 2.97457337e-01 -3.44105899e-01 4.25380439e-01
-1.68784305e-01 7.52884388e-01 2.09916696e-01 -3.22968662e-01
2.61526078e-01 -3.28431100e-01 6.71681091e-02 1.63671041e+00
2.11439319e-02 3.86066228e-01 4.78923529e-01 4.79070514e-01
-2.04253107e-01 4.39065635e-01 -1.61119431e-01 1.25201344e-01
7.10830390e-01 1.60837650e+00 -5.17904878e-01 -4.54510182e-01
-4.36460584e-01 1.01056099e+00 5.54713666e-01 2.70667970e-01
-7.99340904e-01 -3.76613349e-01 6.96166575e-01 -5.95752835e-01
2.13728249e-01 4.46880490e-01 -3.20151374e-02 -8.88200402e-01
-2.39651531e-01 -1.40308332e+00 4.45806652e-01 -6.88983619e-01
-1.37274170e+00 7.44851887e-01 -1.14643909e-01 -1.06165171e+00
-2.86826164e-01 -5.60438633e-01 -9.81117666e-01 5.53441286e-01
-1.58334875e+00 -7.43187428e-01 -4.38190937e-01 5.92597842e-01
9.26320851e-01 -7.23306015e-02 1.29943025e+00 3.05483460e-01
-6.60018802e-01 8.13025117e-01 -1.91692561e-01 -2.41523166e-03
8.66322994e-01 -1.62301087e+00 -3.97112131e-01 3.67281377e-01
-2.72436261e-01 4.83020514e-01 8.52882922e-01 -1.38704628e-01
-1.22154546e+00 -7.42475688e-01 9.83964860e-01 -1.34572640e-01
5.48603594e-01 -5.08866608e-01 -9.63597834e-01 4.29003924e-01
4.79552001e-01 -5.12192070e-01 1.25370729e+00 7.79546261e-01
-5.95349252e-01 -1.30672574e-01 -1.17035282e+00 6.34072006e-01
6.79755151e-01 -4.82011467e-01 -5.96765101e-01 2.32134819e-01
1.08358645e+00 -4.48705971e-01 -1.11797237e+00 3.80232394e-01
4.12799239e-01 -8.92960370e-01 6.67410851e-01 -8.29238176e-01
9.29830611e-01 3.58273119e-01 -2.11027488e-01 -1.70360243e+00
-7.64940754e-02 -4.41202432e-01 -9.89325047e-02 1.56708658e+00
3.03915679e-01 -2.69575119e-01 5.19559503e-01 6.47811651e-01
-5.05120218e-01 -1.09046519e+00 -5.59715688e-01 -3.24918061e-01
2.56552637e-01 -6.46074653e-01 5.87694883e-01 1.22497356e+00
6.03753448e-01 6.17889702e-01 -2.45759144e-01 8.04356486e-02
1.44921601e-01 9.39291120e-02 5.71850479e-01 -1.29334545e+00
-3.21179777e-01 -7.06928909e-01 1.43309981e-01 -1.00392282e+00
5.74498177e-01 -6.44072771e-01 4.17865545e-01 -1.05312800e+00
2.21196383e-01 -4.10414249e-01 -5.73148608e-01 2.02362955e-01
-4.71457332e-01 -2.31723070e-01 1.58329427e-01 -5.45138180e-01
-1.03056073e+00 6.17871881e-01 9.88534868e-01 1.23719737e-01
-1.84257120e-01 -1.39319107e-01 -9.82093513e-01 7.10414588e-01
7.10067749e-01 -1.97352454e-01 -5.76755941e-01 3.56888212e-02
1.72411408e-02 2.48692289e-01 -2.18208149e-01 -7.62416542e-01
1.28775150e-01 -1.75366595e-01 1.21311896e-01 -3.85742337e-01
3.19308817e-01 -9.03641105e-01 -3.91109347e-01 6.77988455e-02
-8.78509521e-01 -2.07375735e-02 3.45982164e-01 4.41505760e-01
-3.79677057e-01 -3.49936903e-01 7.57903576e-01 1.59741431e-01
-8.25528145e-01 -1.21513866e-01 -5.45147657e-01 -4.69314419e-02
8.52471352e-01 2.29038119e-01 -3.13073158e-01 -6.98588789e-01
-7.75481224e-01 5.95402181e-01 -1.38550892e-01 8.22890639e-01
5.39643288e-01 -1.32119191e+00 -5.48587799e-01 5.75000681e-02
6.32618964e-01 -2.41509959e-01 6.28200173e-01 6.96092486e-01
2.93647707e-01 2.59285390e-01 1.11453228e-01 -5.65954924e-01
-1.55147457e+00 4.63681281e-01 3.91243577e-01 -7.18340397e-01
-1.85010701e-01 9.19500530e-01 1.32810131e-01 -1.01840484e+00
7.67001152e-01 -2.27637500e-01 -4.76518124e-01 3.17725986e-01
6.93823040e-01 -3.29924040e-02 1.13670394e-01 -3.16811681e-01
-8.45522881e-02 2.41589904e-01 -4.21574622e-01 8.80516320e-02
1.27451110e+00 -1.57240167e-01 1.79192245e-01 5.12151837e-01
1.25600755e+00 -2.09855601e-01 -1.08411539e+00 -2.83611059e-01
2.98840970e-01 5.07479906e-02 6.95039481e-02 -1.22646046e+00
-8.74393463e-01 7.99892545e-01 7.74658084e-01 2.80533046e-01
1.45297885e+00 -7.91584179e-02 5.17851949e-01 2.57595658e-01
1.00043580e-01 -1.55330753e+00 5.33430636e-01 7.92709827e-01
8.52440476e-01 -1.23544908e+00 -5.91865361e-01 -3.71772826e-01
-1.10843873e+00 9.36413407e-01 9.41850483e-01 2.14847088e-01
5.17070234e-01 4.27405298e-01 3.95158798e-01 -2.13881761e-01
-1.33930337e+00 -2.45172560e-01 2.04855919e-01 4.47195739e-01
8.48156154e-01 3.00553739e-01 -4.98953164e-01 9.98654485e-01
-2.67028093e-01 -3.80984038e-01 3.77951503e-01 7.79668808e-01
-5.24300456e-01 -1.16388762e+00 -6.24080263e-02 6.20572157e-02
-2.04237640e-01 1.01313749e-02 -4.13656145e-01 6.25900269e-01
3.40427719e-02 1.36645520e+00 2.31211200e-01 -6.53356373e-01
6.19433880e-01 7.12962687e-01 1.46484569e-01 -5.44583976e-01
-1.05316222e+00 2.42948651e-01 4.70153153e-01 -6.87348247e-01
-5.13418853e-01 -4.32676941e-01 -1.18076122e+00 -1.50574416e-01
-1.66347399e-01 6.24528646e-01 7.08612978e-01 1.13112724e+00
2.01755598e-01 1.04825103e+00 1.02457571e+00 -5.64974904e-01
-5.31623185e-01 -1.18934250e+00 -4.21400547e-01 8.18966448e-01
-1.15688093e-01 -6.17169678e-01 -4.54058319e-01 -1.92976706e-02] | [13.081714630126953, 6.0574140548706055] |
28ce24b0-da5c-4a4d-8367-406f7bd6c5bb | dore-document-ordered-relation-extraction | 2210.16064 | null | https://arxiv.org/abs/2210.16064v2 | https://arxiv.org/pdf/2210.16064v2.pdf | DORE: Document Ordered Relation Extraction based on Generative Framework | In recent years, there is a surge of generation-based information extraction work, which allows a more direct use of pre-trained language models and efficiently captures output dependencies. However, previous generative methods using lexical representation do not naturally fit document-level relation extraction (DocRE) where there are multiple entities and relational facts. In this paper, we investigate the root cause of the underwhelming performance of the existing generative DocRE models and discover that the culprit is the inadequacy of the training paradigm, instead of the capacities of the models. We propose to generate a symbolic and ordered sequence from the relation matrix which is deterministic and easier for model to learn. Moreover, we design a parallel row generation method to process overlong target sequences. Besides, we introduce several negative sampling strategies to improve the performance with balanced signals. Experimental results on four datasets show that our proposed method can improve the performance of the generative DocRE models. We have released our code at https://github.com/ayyyq/DORE. | ['Zheng Zhang', 'Xipeng Qiu', 'Hang Yan', 'Yuqing Yang', 'Qipeng Guo'] | 2022-10-28 | null | null | null | null | ['document-level-relation-extraction'] | ['natural-language-processing'] | [ 7.59808123e-02 3.00542593e-01 -3.01971436e-01 -1.65558770e-01
-4.06869560e-01 -4.55141693e-01 7.01895356e-01 -8.20580423e-02
1.07087158e-01 8.93378079e-01 3.11162651e-01 -4.91454005e-01
-5.37438281e-02 -1.04158342e+00 -6.64640605e-01 -4.65706766e-01
3.57523002e-02 5.44919193e-01 1.77944452e-01 -3.49376947e-01
1.37072606e-02 9.51393098e-02 -1.34830308e+00 3.24734032e-01
9.70459580e-01 4.99604285e-01 2.40360275e-01 5.28691888e-01
-4.08575088e-01 1.05727184e+00 -6.11374795e-01 -7.24065363e-01
7.51943737e-02 -9.19387460e-01 -7.08422482e-01 -9.40169543e-02
-3.52965236e-01 -1.80235580e-01 -2.69014984e-01 1.06839943e+00
3.78162622e-01 -3.21106255e-01 7.89887249e-01 -1.34712517e+00
-8.02297413e-01 1.37727773e+00 -5.84458768e-01 1.57211751e-01
3.42025697e-01 -3.32576811e-01 1.19240522e+00 -8.95342350e-01
6.86368227e-01 1.19864321e+00 4.70839828e-01 3.75850022e-01
-9.50650752e-01 -8.76023293e-01 1.66777566e-01 1.70111492e-01
-1.63845730e+00 -4.63822871e-01 9.49831247e-01 -2.93284446e-01
9.76343036e-01 2.85863459e-01 6.54384494e-01 1.38908732e+00
1.55215546e-01 8.70004058e-01 9.56508100e-01 -6.62942588e-01
-1.89850889e-02 3.31325322e-01 1.04099177e-01 6.61664605e-01
6.36038661e-01 -3.06048952e-02 -5.17549872e-01 -1.92679748e-01
8.45788896e-01 -9.87063441e-03 -2.88290352e-01 -2.44806573e-01
-9.98638570e-01 8.66250038e-01 -1.15601614e-01 5.52312911e-01
-2.45454863e-01 -1.05704613e-01 1.80961579e-01 5.12126014e-02
3.75628561e-01 2.98688918e-01 -2.78952986e-01 -1.32630110e-01
-8.70637238e-01 2.17335343e-01 1.06473637e+00 1.36991501e+00
4.70471054e-01 5.33740222e-02 -1.20345272e-01 7.56629169e-01
3.64976138e-01 4.20423895e-01 6.08547330e-01 -2.14964524e-01
6.71114087e-01 6.31855905e-01 -1.47519022e-01 -1.12067437e+00
-8.23534206e-02 -6.62622809e-01 -9.91072893e-01 -5.17328680e-01
2.52539098e-01 -3.05900961e-01 -8.08623791e-01 1.49463272e+00
1.09678656e-01 3.69203053e-02 2.81790674e-01 5.86593568e-01
7.05319285e-01 7.48807549e-01 -4.23985235e-02 -3.43892097e-01
1.28904307e+00 -8.87068331e-01 -9.69802201e-01 -4.05411333e-01
7.61178017e-01 -7.89983928e-01 7.02593744e-01 3.89035344e-01
-8.88587117e-01 -3.58382821e-01 -1.13865912e+00 1.31550118e-01
-3.37493032e-01 4.68313158e-01 9.34093297e-01 6.17210686e-01
-3.65830421e-01 3.26909065e-01 -8.57711673e-01 -2.18587160e-01
6.90773949e-02 1.59160838e-01 -7.18303444e-03 5.16246073e-02
-1.47455418e+00 7.66572773e-01 9.06316757e-01 3.81807804e-01
-5.07931888e-01 -1.75004214e-01 -6.83780491e-01 6.33728132e-02
7.92440116e-01 -5.90937197e-01 1.00300252e+00 -6.96830213e-01
-1.40639961e+00 2.08629042e-01 -1.93230420e-01 -4.38757390e-01
3.51847678e-01 -4.60049808e-01 -5.07600367e-01 -2.00352892e-01
-7.19565973e-02 9.46145952e-02 5.01864195e-01 -1.35903931e+00
-5.05504251e-01 -1.03611045e-01 -2.15696573e-01 3.22504826e-02
-2.73113102e-01 -1.41847715e-01 -3.84867162e-01 -6.70366168e-01
1.72339812e-01 -8.56108308e-01 -1.62479386e-01 -1.13736713e+00
-9.51470554e-01 -2.71751702e-01 3.25840414e-01 -5.38529575e-01
1.87746692e+00 -1.89083874e+00 -4.91709262e-02 4.91632968e-01
1.08888291e-01 3.12834471e-01 1.09125167e-01 1.04739797e+00
-2.78543472e-01 4.82283503e-01 -5.21085970e-02 6.53813127e-03
-4.03965265e-02 3.57182473e-01 -7.52532005e-01 -1.01302966e-01
3.49807531e-01 1.12958860e+00 -8.23567748e-01 -7.00331569e-01
-3.08256239e-01 2.78656214e-01 -3.80829364e-01 3.56902868e-01
-3.23780507e-01 9.27894041e-02 -7.28327870e-01 4.60447192e-01
5.17334223e-01 -3.97794306e-01 6.83825731e-01 -1.49446607e-01
4.71958779e-02 6.38708174e-01 -1.35960150e+00 1.39734972e+00
-2.06924617e-01 2.86837909e-02 -6.50676250e-01 -8.18340838e-01
1.22584689e+00 3.95762086e-01 1.76631734e-01 -3.22332948e-01
2.18809113e-01 3.04093719e-01 3.25017869e-01 -5.38681805e-01
5.49405515e-01 -1.08149000e-01 -2.06561610e-02 4.38155413e-01
1.09800071e-01 8.98359567e-02 4.36170399e-01 5.07307768e-01
8.87224972e-01 2.92012691e-01 7.15579569e-01 7.20241815e-02
3.22321087e-01 -1.18140997e-02 7.15027809e-01 6.77543879e-01
5.69096565e-01 5.22479415e-01 7.17802823e-01 -1.37583032e-01
-9.72273350e-01 -8.00102353e-01 7.30312318e-02 6.47941411e-01
-4.70383605e-03 -9.60056841e-01 -4.78470355e-01 -6.87640071e-01
-3.42694640e-01 8.55900407e-01 -3.35383624e-01 -9.84481126e-02
-6.12505794e-01 -1.17560244e+00 6.66041076e-01 5.98663747e-01
3.16090882e-01 -8.95385146e-01 -2.07720861e-01 4.21392351e-01
-4.57304239e-01 -1.08170867e+00 -8.48397091e-02 3.04259688e-01
-8.69102955e-01 -8.33622217e-01 -3.71642888e-01 -6.11043751e-01
6.79617763e-01 3.15516181e-02 1.24091876e+00 1.55627593e-01
1.27109155e-01 -3.45020920e-01 -7.77244270e-01 -3.66714418e-01
-5.81826508e-01 6.50696099e-01 -1.63263381e-01 -1.30595326e-01
5.71603656e-01 -7.78522968e-01 -2.22131595e-01 1.20668888e-01
-9.16684628e-01 5.51966071e-01 1.04835606e+00 9.17234659e-01
4.59551662e-01 2.70054281e-01 7.33039677e-01 -1.38673639e+00
8.81831288e-01 -6.87440455e-01 -4.09118503e-01 4.09168661e-01
-1.02648282e+00 4.53843981e-01 5.92936158e-01 -6.07775450e-01
-1.04847491e+00 -2.18384508e-02 -5.88105060e-02 -1.19995460e-01
-3.93694192e-02 9.01706040e-01 -3.05389345e-01 7.15164244e-01
4.77561146e-01 5.07227838e-01 -4.24566776e-01 -3.45520496e-01
4.48529243e-01 7.22248316e-01 2.19036013e-01 -8.40216696e-01
1.01300752e+00 2.27390211e-02 -1.48719028e-01 -4.10433918e-01
-8.38689446e-01 -4.92456071e-02 -4.06711161e-01 1.26008868e-01
5.20815611e-01 -8.62244785e-01 -2.94010282e-01 9.67655629e-02
-1.27671981e+00 3.12844403e-02 -5.98925650e-02 4.85448986e-01
-1.59196392e-01 3.32455426e-01 -5.80597222e-01 -1.14698994e+00
-4.39588994e-01 -7.88382173e-01 7.55629122e-01 1.87741652e-01
-3.45140308e-01 -7.08588719e-01 2.15210244e-01 1.47310477e-02
6.22152910e-02 1.49209470e-01 9.82416093e-01 -9.11555946e-01
-8.76958787e-01 -1.51398063e-01 -5.27713597e-02 9.70930457e-02
2.12007880e-01 1.65274411e-01 -7.73128808e-01 9.78096575e-02
-1.33277416e-01 -8.68025348e-02 6.54763341e-01 -6.48759231e-02
8.14303279e-01 -5.83001077e-01 -5.40691912e-01 3.59847516e-01
1.46327853e+00 5.38169980e-01 8.86762023e-01 1.23981044e-01
7.16153741e-01 5.03094196e-01 4.84692037e-01 3.40564251e-01
5.80001950e-01 5.32303929e-01 -1.09621705e-02 8.85777622e-02
2.62820031e-02 -8.93297970e-01 3.38540971e-01 1.35992348e+00
-3.97790849e-01 -5.04742146e-01 -9.52174008e-01 6.37896299e-01
-2.01544046e+00 -9.68900561e-01 -2.66989321e-01 1.88273430e+00
1.18931520e+00 3.91782671e-01 -4.94833440e-02 3.01663339e-01
5.22426307e-01 -1.08638547e-01 -4.60527502e-02 -1.94568574e-01
-4.49870601e-02 2.68809617e-01 3.36299866e-01 3.88404369e-01
-7.99138665e-01 1.11814582e+00 5.65601921e+00 1.06842983e+00
-8.16894472e-01 -2.68227272e-02 4.26796824e-01 1.06835149e-01
-6.01148188e-01 3.22476178e-01 -1.21325946e+00 4.80182499e-01
9.95848238e-01 -2.49507472e-01 1.42700538e-01 7.47333229e-01
-1.16318420e-01 2.56080832e-02 -1.20328665e+00 7.67266572e-01
9.91762877e-02 -1.04661250e+00 2.04279184e-01 2.00279325e-01
5.55116653e-01 -3.99594843e-01 -2.88899332e-01 4.40652698e-01
5.53989947e-01 -1.06483555e+00 6.21939719e-01 6.00894511e-01
3.32370281e-01 -8.53010297e-01 9.39233661e-01 5.71323752e-01
-1.14639127e+00 1.16926812e-01 -3.40707719e-01 -6.86151236e-02
1.96450889e-01 8.97880971e-01 -1.13137531e+00 1.03829455e+00
1.79812387e-01 4.61734414e-01 -6.52000546e-01 6.10692084e-01
-6.72087610e-01 7.62170315e-01 -2.08767474e-01 -3.29962373e-01
-1.66396484e-01 -2.66330361e-01 3.31868619e-01 1.32715869e+00
5.02738714e-01 2.52391070e-01 4.84617427e-02 9.59463537e-01
1.34543702e-01 2.75064141e-01 -8.51713181e-01 -4.60024297e-01
6.81032777e-01 1.24486947e+00 -7.31923878e-01 -2.47263253e-01
-3.48874867e-01 6.49067402e-01 2.70928591e-01 4.50603247e-01
-9.28478122e-01 -4.20827508e-01 -6.41573370e-02 4.98085693e-02
4.15338814e-01 -4.01233763e-01 -3.29590052e-01 -1.43498969e+00
1.17245987e-01 -1.08827639e+00 1.16316736e-01 -5.74016988e-01
-1.09848332e+00 8.14441264e-01 1.71565309e-01 -1.14093018e+00
-8.21061492e-01 -1.31847784e-01 -2.75349200e-01 8.78021657e-01
-1.19394135e+00 -1.22040975e+00 1.10421725e-01 2.44007781e-01
3.33002329e-01 -1.91409573e-01 8.34039092e-01 3.11618507e-01
-7.36468792e-01 5.93757451e-01 -3.39298189e-01 4.40668017e-01
4.22993571e-01 -1.09768772e+00 4.34608579e-01 1.20886552e+00
5.69099724e-01 1.01665366e+00 7.69050360e-01 -8.94228816e-01
-1.41348815e+00 -8.20498765e-01 1.30957079e+00 -4.75033671e-01
7.07713246e-01 -7.10433066e-01 -8.25049281e-01 9.15476143e-01
3.69083852e-01 -5.93448043e-01 7.01940477e-01 2.79465407e-01
-3.47298354e-01 -1.13349564e-01 -5.88153183e-01 6.89240813e-01
1.07616162e+00 -2.33270720e-01 -6.78694606e-01 7.52825513e-02
5.76260209e-01 -4.01583165e-01 -7.15921581e-01 4.10898060e-01
5.41314602e-01 -7.83160865e-01 7.54932106e-01 -4.60036874e-01
7.89330244e-01 -3.52381736e-01 -5.85958846e-02 -1.17745578e+00
-2.61156917e-01 -5.57392240e-01 -5.10864675e-01 1.68316102e+00
7.81270444e-01 -6.21270776e-01 4.91582632e-01 3.11584771e-01
2.00073868e-01 -9.13371146e-01 -3.43030095e-01 -7.92155087e-01
-1.61098644e-01 -3.16227704e-01 8.48633111e-01 8.83561134e-01
1.38335705e-01 9.30581093e-01 -5.93562424e-01 2.00640455e-01
3.13254297e-01 4.25577074e-01 8.25999618e-01 -1.06789637e+00
-6.53499901e-01 -6.10282682e-02 -2.07891781e-02 -1.21068227e+00
-1.77403197e-01 -9.61379230e-01 2.58302037e-02 -1.52919388e+00
3.24460924e-01 -5.69540262e-01 -1.03406228e-01 6.68162405e-01
-2.63755471e-01 -1.53006583e-01 6.97535723e-02 3.61246556e-01
-2.70161480e-01 5.61713994e-01 1.17702746e+00 1.98419809e-01
-2.33068839e-01 -6.40254989e-02 -1.12542546e+00 4.17888254e-01
8.77160966e-01 -7.54964888e-01 -7.32228935e-01 -2.57096201e-01
7.42297709e-01 -3.32404114e-02 1.59596398e-01 -7.55311668e-01
1.91100284e-01 -1.63440034e-01 2.95952171e-01 -7.02067971e-01
7.93538764e-02 -6.26314640e-01 5.60993135e-01 1.82342023e-01
-1.52196079e-01 1.21666707e-01 -1.48742318e-01 3.35223675e-01
-2.46229663e-01 -3.51715297e-01 2.31270969e-01 -2.62947045e-02
-2.54404843e-01 2.42674686e-02 -1.45222574e-01 1.06844030e-01
6.90877318e-01 2.13357717e-01 -3.89937192e-01 -2.83434063e-01
-3.16012204e-01 -4.55451794e-02 -6.11782894e-02 4.35979873e-01
6.11813486e-01 -1.35611713e+00 -6.79554701e-01 1.69674039e-01
-5.17011173e-02 2.65367106e-02 -1.57214969e-01 7.19617605e-01
-2.28768975e-01 6.42449379e-01 9.75784659e-02 -1.03936739e-01
-9.04687107e-01 4.77832586e-01 -3.50523070e-02 -7.05281198e-01
-5.85850835e-01 5.32128453e-01 -3.97505797e-02 -1.14209503e-01
-1.52473271e-01 -5.12483299e-01 -3.90625715e-01 1.98367685e-02
3.72204244e-01 -1.65728908e-02 -1.07647069e-02 -2.75295138e-01
-2.60950446e-01 1.35481238e-01 -3.19248348e-01 -1.88024953e-01
1.30196452e+00 -4.86130528e-02 -1.05889060e-01 5.57466328e-01
6.49247468e-01 3.10951710e-01 -5.67745090e-01 -1.75349265e-01
2.18029141e-01 -1.67572513e-01 -2.26383269e-01 -7.36789525e-01
-9.84453619e-01 6.79897666e-01 1.28517509e-01 5.52458823e-01
1.02382958e+00 6.76687881e-02 8.83753717e-01 1.66683540e-01
3.74656439e-01 -8.16324949e-01 -2.41527483e-01 4.73764986e-01
6.52826846e-01 -8.58399570e-01 1.02142096e-01 -8.88566434e-01
-6.99506640e-01 1.02848744e+00 4.82401013e-01 -3.10426056e-02
5.21328330e-01 5.72462738e-01 -6.40198737e-02 -1.17215686e-01
-9.34880555e-01 -3.74015629e-01 2.05263346e-01 3.37465554e-01
7.69191921e-01 1.40568733e-01 -8.03175628e-01 1.17494118e+00
-6.15094662e-01 1.73914433e-01 5.48829257e-01 8.63610268e-01
-2.47150049e-01 -1.72042215e+00 -7.12613314e-02 3.19774657e-01
-5.61153293e-01 -4.95298535e-01 -4.38130051e-01 8.62434685e-01
1.94064617e-01 8.99378359e-01 -3.93372297e-01 -6.05791926e-01
1.44580975e-01 4.54936773e-01 3.83191437e-01 -7.59715736e-01
-2.89793104e-01 3.24835360e-01 4.20908123e-01 -8.46056715e-02
-3.03424448e-01 -5.46207905e-01 -1.17223048e+00 -1.50795572e-03
-5.52388370e-01 5.65400958e-01 3.55766684e-01 8.95423174e-01
3.39706481e-01 6.03936851e-01 5.14231324e-01 -1.24899028e-02
-3.92713070e-01 -1.22240448e+00 -5.06176174e-01 6.77926540e-02
-1.95024863e-01 -6.27113163e-01 -1.53015330e-01 2.21840322e-01] | [9.431200981140137, 8.711225509643555] |
31f65c50-fa1f-4e1e-9ddd-5f2ae1bfb65f | bridging-the-gap-between-local-semantic | 2210.08875 | null | https://arxiv.org/abs/2210.08875v1 | https://arxiv.org/pdf/2210.08875v1.pdf | Bridging the Gap between Local Semantic Concepts and Bag of Visual Words for Natural Scene Image Retrieval | This paper addresses the problem of semantic-based image retrieval of natural scenes. A typical content-based image retrieval system deals with the query image and images in the dataset as a collection of low-level features and retrieves a ranked list of images based on the similarities between features of the query image and features of images in the image dataset. However, top ranked images in the retrieved list, which have high similarities to the query image, may be different from the query image in terms of the semantic interpretation of the user which is known as the semantic gap. In order to reduce the semantic gap, this paper investigates how natural scene retrieval can be performed using the bag of visual word model and the distribution of local semantic concepts. The paper studies the efficiency of using different approaches for representing the semantic information, depicted in natural scene images, for image retrieval. An extensive experimental work has been conducted to study the efficiency of using semantic information as well as the bag of visual words model for natural and urban scene image retrieval. | ['Yousef Alqasrawi'] | 2022-10-17 | null | null | null | null | ['content-based-image-retrieval'] | ['computer-vision'] | [ 3.18319976e-01 -3.10364932e-01 -1.53691038e-01 -5.11311710e-01
-6.09252095e-01 -4.21801686e-01 6.77555025e-01 6.18985593e-01
-5.83457470e-01 1.48698762e-01 3.11842233e-01 2.09542334e-01
-5.06445527e-01 -9.86222148e-01 -3.27538162e-01 -4.53177452e-01
2.15764731e-01 4.56375360e-01 6.33292079e-01 -3.62428814e-01
7.83751547e-01 4.90574658e-01 -2.23286676e+00 4.45373625e-01
1.75639227e-01 8.21336269e-01 7.98034370e-01 4.27532315e-01
-6.13228738e-01 7.22894967e-01 -7.98545122e-01 3.12281728e-01
1.56359032e-01 -5.55489779e-01 -1.15558743e+00 4.54707265e-01
5.22439837e-01 -1.86703220e-01 -2.43296877e-01 1.35075855e+00
3.87861431e-01 3.90967906e-01 8.33560765e-01 -1.21290946e+00
-6.42018437e-01 -1.73804432e-01 -1.17928416e-01 3.85072082e-01
5.23620844e-01 -6.36614740e-01 9.31585312e-01 -7.60888636e-01
8.13392997e-01 1.56077123e+00 -3.23062301e-01 1.30617753e-01
-6.62462950e-01 -1.53512821e-01 -1.05148656e-02 6.14099205e-01
-1.66693115e+00 -1.07990585e-01 6.76794171e-01 -4.80709821e-01
7.39191532e-01 3.41500491e-01 6.52106345e-01 -7.11900517e-02
1.30729795e-01 5.89945555e-01 8.26578856e-01 -8.53324950e-01
2.48731494e-01 3.91581357e-01 5.02456903e-01 5.59393644e-01
4.32220921e-02 -1.58684760e-01 -4.76972789e-01 -2.55828738e-01
5.77847302e-01 3.17610085e-01 -1.46030158e-01 -5.76401532e-01
-7.53875732e-01 9.16699052e-01 9.17760968e-01 7.55619645e-01
-4.49230164e-01 1.03390887e-01 4.85206604e-01 2.20908254e-01
2.71412641e-01 3.10359776e-01 -6.57288299e-04 5.67671418e-01
-7.35663950e-01 2.50202566e-01 4.50678468e-01 8.32689941e-01
1.13888085e+00 -4.13023770e-01 -1.83388349e-02 1.05603266e+00
6.09129667e-01 7.33316898e-01 6.94978714e-01 -6.41821027e-01
-5.18982597e-02 8.99549961e-01 -9.66510400e-02 -1.65161872e+00
5.85627519e-02 1.16519041e-01 -2.64347911e-01 -3.54294479e-02
-1.57264143e-01 8.59386265e-01 -1.21394336e+00 1.32741213e+00
1.63601980e-01 -2.72413880e-01 3.00512016e-01 1.10783291e+00
1.24907494e+00 9.79837120e-01 2.22028241e-01 -1.08392417e-01
1.50291681e+00 -8.98471773e-01 -5.90692878e-01 -3.31725091e-01
5.21795094e-01 -1.11936986e+00 8.62222910e-01 -2.92325348e-01
-7.57673800e-01 -7.19545126e-01 -8.65177274e-01 -1.23949833e-01
-9.81432378e-01 -3.91990170e-02 1.71607390e-01 2.19481602e-01
-1.22117162e+00 1.10321127e-01 -5.90016507e-02 -1.08698583e+00
-8.08664039e-02 9.48598422e-03 -3.69999319e-01 -5.98894715e-01
-1.14223599e+00 8.91175210e-01 9.54078734e-01 -3.43988121e-01
-9.44476783e-01 -2.46839285e-01 -9.38500524e-01 1.62947863e-01
2.56817758e-01 -3.92972022e-01 7.35705376e-01 -1.21405375e+00
-4.67898369e-01 1.37894082e+00 -3.97797465e-01 -1.06502727e-01
-2.12424204e-01 -1.53346574e-02 -3.98504257e-01 7.44407594e-01
5.15843451e-01 9.19275582e-01 7.83334970e-01 -1.57388592e+00
-6.68772876e-01 -5.06267011e-01 1.65291637e-01 5.89175463e-01
-8.87212306e-02 2.29219750e-01 -7.91678309e-01 -3.94134790e-01
3.49276185e-01 -7.29419589e-01 -1.86203346e-01 -1.05424888e-01
9.42009389e-02 -2.90770173e-01 9.94132817e-01 -3.54471475e-01
9.71073151e-01 -2.53650331e+00 -3.73914063e-01 7.28589594e-01
-9.16945934e-02 2.75514215e-01 -3.86905164e-01 8.60128522e-01
-6.18499741e-02 2.19179645e-01 8.32806081e-02 2.81198621e-01
-4.34119046e-01 5.39324164e-01 -2.76758105e-01 1.49226114e-01
-3.11628997e-01 6.55686259e-01 -1.04516709e+00 -1.01500428e+00
4.97004837e-01 3.03808123e-01 -1.11235902e-01 9.31120515e-02
4.84010428e-02 3.33254449e-02 -9.35384810e-01 4.04520035e-01
4.71251607e-01 -1.07784092e-01 -9.57179293e-02 -1.95734441e-01
5.52749410e-02 -8.63085911e-02 -1.01171625e+00 1.50047779e+00
-2.83379138e-01 7.58752227e-01 -4.85281706e-01 -1.04020762e+00
1.08010018e+00 2.37572730e-01 4.85934317e-01 -1.17505324e+00
-4.47823144e-02 1.74397737e-01 -3.65162522e-01 -6.21317089e-01
9.44531739e-01 -1.70402750e-01 -8.87813568e-02 3.65674138e-01
-7.42167309e-02 -4.08021003e-01 3.67732227e-01 4.71832365e-01
5.18650770e-01 -3.54528159e-01 4.08721775e-01 -4.90189791e-01
7.11452067e-01 4.28649873e-01 -1.27785325e-01 9.27690506e-01
-9.69759449e-02 5.36510289e-01 -1.05414346e-01 -4.91438240e-01
-1.06547725e+00 -9.49761808e-01 3.94205898e-02 1.18626964e+00
8.74772191e-01 -5.33124447e-01 -6.02919936e-01 -1.00459807e-01
-1.48258507e-01 3.78691316e-01 -3.42645049e-01 -3.32367510e-01
-8.90604779e-02 -3.15408260e-01 8.36297590e-03 1.03431344e-01
7.50737906e-01 -1.23751068e+00 -7.85845876e-01 -7.63431042e-02
-3.06024194e-01 -8.54453564e-01 -3.06711614e-01 -2.71444678e-01
-6.45863473e-01 -1.20843208e+00 -5.38540900e-01 -1.27282953e+00
9.45203483e-01 1.17780793e+00 1.13652360e+00 6.40075028e-01
-6.40723944e-01 9.13942218e-01 -9.33808148e-01 -3.87137771e-01
-3.11314911e-01 -4.42628384e-01 -4.44760323e-01 -2.96184868e-02
6.36320591e-01 1.58814579e-01 -6.94364190e-01 3.49382341e-01
-1.49538684e+00 -3.14710915e-01 4.27199841e-01 5.83761930e-01
7.11960077e-01 6.05742812e-01 1.69890419e-01 -3.66314977e-01
5.44317663e-01 -3.45228136e-01 -2.75036246e-01 6.07929111e-01
-3.99834841e-01 1.47350803e-01 -2.35427879e-02 1.54243428e-02
-7.78696418e-01 -1.96892008e-01 4.03697282e-01 -3.70629311e-01
-2.43864313e-01 7.14985967e-01 1.30449478e-02 -2.94989850e-02
5.12448847e-01 5.68185687e-01 1.20138540e-03 -3.24020773e-01
3.51830781e-01 7.32950568e-01 3.02881151e-01 -2.52455920e-01
4.10883605e-01 6.24395192e-01 1.35907501e-01 -1.29755199e+00
-8.30081820e-01 -1.43935597e+00 -5.43642998e-01 -5.05722284e-01
9.94558990e-01 -1.01168430e+00 -6.50292858e-02 1.61282301e-01
-1.03500617e+00 5.46371758e-01 -2.51630992e-01 4.05899286e-01
-4.75818396e-01 6.37029111e-01 2.34185369e-03 -6.72776401e-01
-2.22008422e-01 -1.08637977e+00 1.37337041e+00 8.24571252e-02
-5.56276925e-02 -9.12473798e-01 -1.55793950e-01 3.82336855e-01
2.05860570e-01 -4.71982434e-02 1.21301389e+00 -7.93754101e-01
-6.31144702e-01 -4.30440754e-01 -4.87253308e-01 3.44149888e-01
2.25474223e-01 -2.62707859e-01 -8.35395575e-01 -3.17943037e-01
5.83205447e-02 -2.38989621e-01 9.07434940e-01 4.50694084e-01
7.46864080e-01 -8.83065388e-02 -4.21996295e-01 -2.02953577e-01
2.12731266e+00 5.67493379e-01 9.02087331e-01 3.86656970e-01
2.94030130e-01 8.65947723e-01 1.02676368e+00 2.37799615e-01
-1.07269712e-01 5.52682817e-01 3.48084927e-01 -1.37267530e-01
7.68642826e-03 -1.67026013e-01 -2.23335117e-01 6.65995955e-01
3.61400008e-01 -4.27922994e-01 -9.32023644e-01 8.32406640e-01
-1.88849843e+00 -1.03883803e+00 -1.80968121e-02 2.23347139e+00
1.73074782e-01 -2.96904534e-01 -9.59762856e-02 1.42547548e-01
9.57832694e-01 3.28469485e-01 -2.28284821e-02 -4.67880785e-01
5.50045585e-03 -5.95365874e-02 3.95918310e-01 5.47172189e-01
-9.63865995e-01 1.08268404e+00 6.77254534e+00 1.28347743e+00
-9.47642684e-01 -5.04402183e-02 3.15907925e-01 4.56185013e-01
-2.17862770e-01 3.31207961e-01 -5.14188349e-01 2.03462303e-01
4.10529137e-01 -4.09915298e-01 7.40567744e-02 8.95730257e-01
1.64540976e-01 -6.86407089e-01 -8.20531130e-01 1.11598277e+00
6.99883223e-01 -9.53603685e-01 9.08911765e-01 -1.05806850e-01
5.39839745e-01 -1.71335682e-01 -1.48502871e-01 -2.11618885e-01
-1.18258975e-01 -9.23763335e-01 5.75681508e-01 7.50953138e-01
2.51715362e-01 -8.37046683e-01 8.91467035e-01 2.34673128e-01
-1.30468965e+00 1.69177666e-01 -6.87091172e-01 9.95830297e-02
-2.13629663e-01 1.12257160e-01 -6.31315231e-01 6.14025056e-01
8.31896126e-01 7.56038547e-01 -9.05392587e-01 1.19077575e+00
2.89625168e-01 -9.74115264e-03 -1.96020231e-01 -1.58890188e-01
5.89551866e-01 -2.04368085e-01 4.09779817e-01 1.01111877e+00
2.16605082e-01 9.61816311e-02 5.67047417e-01 5.35015285e-01
3.20597917e-01 6.62408769e-01 -9.67971325e-01 -9.76569057e-02
3.86988610e-01 8.04153502e-01 -9.12354290e-01 -6.62174821e-01
-1.97955206e-01 8.96260619e-01 -3.01243275e-01 3.80015284e-01
-1.47228360e-01 -3.95042002e-01 2.06846982e-01 1.28590807e-01
8.27920884e-02 -6.30278885e-02 4.55706447e-01 -6.48944855e-01
-5.37514612e-02 -5.38608670e-01 4.52496022e-01 -1.33063078e+00
-1.07286370e+00 5.67235589e-01 5.60121536e-01 -1.46784317e+00
-1.36687547e-01 -3.83603483e-01 -2.73265868e-01 7.78686345e-01
-1.44442189e+00 -1.08156109e+00 -4.53671575e-01 8.42257798e-01
7.44620800e-01 -1.36980951e-01 7.82205462e-01 5.97256832e-02
4.27804887e-01 -2.42222339e-01 3.95920962e-01 8.74541327e-02
6.13661230e-01 -8.46792042e-01 -4.62326914e-01 4.13071662e-01
2.34577805e-01 4.79419798e-01 6.91654682e-01 -5.43480575e-01
-8.07240903e-01 -9.19487238e-01 1.18651199e+00 3.46017815e-02
3.59034270e-01 3.34794670e-01 -7.85004735e-01 9.18407738e-02
3.06391060e-01 -2.76965499e-01 7.41283596e-01 -5.12557507e-01
-3.19889516e-01 2.92116776e-04 -1.07508612e+00 5.74462950e-01
5.98333001e-01 -8.44249725e-01 -9.44535553e-01 4.71549571e-01
5.13089240e-01 4.18122530e-01 -6.12256467e-01 1.94502801e-01
3.08881402e-01 -7.19341874e-01 1.32515860e+00 -4.53180552e-01
2.30659708e-01 -5.08215249e-01 -6.80447519e-01 -1.15037119e+00
-2.32299894e-01 4.94508952e-01 1.10507452e+00 1.01942348e+00
-1.01898924e-01 -2.12723777e-01 3.39934289e-01 1.96702659e-01
3.96907926e-01 -2.99961604e-02 -6.65335834e-01 -8.25800419e-01
-1.87028259e-01 -3.13075669e-02 3.61681700e-01 5.62490821e-01
-2.74824291e-01 5.04350483e-01 -8.96498933e-02 1.78179350e-02
5.05637228e-01 4.68992859e-01 5.09302855e-01 -1.28227210e+00
2.80858785e-01 -3.66837591e-01 -1.13485658e+00 -7.05071032e-01
1.25872657e-01 -1.00142527e+00 2.06832439e-01 -1.86213088e+00
6.96417570e-01 -1.52993277e-01 -3.00272077e-01 2.21339181e-01
-3.28536481e-02 3.30447078e-01 6.58104062e-01 6.53069019e-01
-9.17759061e-01 3.28723252e-01 1.14916658e+00 -3.42134386e-01
8.60269219e-02 -4.39291060e-01 -2.26864025e-01 5.84898472e-01
5.99942744e-01 -5.53409219e-01 -6.81980550e-01 -2.11051062e-01
1.77843928e-01 6.48773788e-03 6.48890913e-01 -8.58564198e-01
2.12876797e-01 -2.06187814e-01 1.31449476e-01 -7.79510736e-01
4.60141420e-01 -1.20007443e+00 1.14885375e-01 4.74215418e-01
-5.61071515e-01 8.93524438e-02 -7.36325467e-03 7.11378455e-01
-9.58197236e-01 -7.75003672e-01 6.88106537e-01 -6.07783735e-01
-1.50337470e+00 -6.67791143e-02 -6.93054736e-01 -1.94406703e-01
9.29344893e-01 -5.65307975e-01 -2.30370477e-01 -6.06364667e-01
-6.49776876e-01 1.75228864e-01 5.29851675e-01 6.08150601e-01
1.03162491e+00 -1.30701077e+00 -4.12805796e-01 3.24777551e-02
8.87721896e-01 -4.27836895e-01 2.48282075e-01 3.49829383e-02
-8.55449080e-01 7.45661974e-01 -2.53292501e-01 -6.46440327e-01
-1.78790581e+00 8.42517436e-01 2.21250668e-01 -3.86534408e-02
-2.66262949e-01 3.54398519e-01 5.76933563e-01 -4.40088660e-02
1.58195626e-02 5.11119552e-02 -6.56858563e-01 1.89474076e-01
5.11344254e-01 1.35830347e-03 -1.53816313e-01 -1.33073330e+00
-4.41206723e-01 1.11140227e+00 9.26181749e-02 -1.03074983e-01
8.44646335e-01 -4.80890870e-01 -6.60168886e-01 4.91390049e-01
1.77345145e+00 -4.13539618e-01 -3.94986756e-02 -6.05610907e-01
2.42651463e-01 -9.26040471e-01 1.94777712e-01 -4.19119507e-01
-8.57948065e-01 6.24172330e-01 1.00559175e+00 3.04034084e-01
1.27780449e+00 4.67465967e-01 3.03829551e-01 5.51050365e-01
5.16812623e-01 -1.19964206e+00 3.24970186e-01 4.68812019e-01
8.93744886e-01 -1.40488231e+00 1.19627461e-01 -6.20796919e-01
-5.18290699e-01 9.96167064e-01 1.80876464e-01 -4.02122319e-01
1.01218545e+00 -6.79628789e-01 1.40366599e-01 -7.28714645e-01
-3.27965200e-01 -7.76375234e-01 5.39069057e-01 4.06008631e-01
8.24757218e-02 -4.37802374e-02 -6.21647716e-01 -4.88532305e-01
1.44629449e-01 -2.85048485e-01 1.55981109e-01 1.35964227e+00
-1.08567023e+00 -1.00537157e+00 -5.01107633e-01 1.62518188e-01
-2.81370372e-01 -1.83994681e-01 -6.48076117e-01 6.94525361e-01
1.38577288e-02 1.17651761e+00 3.71901780e-01 -1.00846924e-01
9.22022313e-02 1.90690272e-02 2.51750857e-01 -6.97667599e-01
-2.10731089e-01 6.92463070e-02 -1.84492886e-01 -3.51070076e-01
-9.63764727e-01 -2.25916147e-01 -1.19709969e+00 1.14941798e-01
-3.67838770e-01 5.28105855e-01 8.91176224e-01 1.02901578e+00
9.65502635e-02 1.23904668e-01 5.96907020e-01 -5.03459930e-01
1.25964135e-01 -7.39340067e-01 -8.68991494e-01 1.05296278e+00
1.34466514e-01 -4.58471358e-01 -3.39211911e-01 3.35171193e-01] | [10.814258575439453, 0.3817130923271179] |
4056fb9c-b51d-4ad5-9f2a-57bc2b67453d | learning-so3-equivariant-representations-with | 1711.06721 | null | http://arxiv.org/abs/1711.06721v3 | http://arxiv.org/pdf/1711.06721v3.pdf | Learning SO(3) Equivariant Representations with Spherical CNNs | We address the problem of 3D rotation equivariance in convolutional neural
networks. 3D rotations have been a challenging nuisance in 3D classification
tasks requiring higher capacity and extended data augmentation in order to
tackle it. We model 3D data with multi-valued spherical functions and we
propose a novel spherical convolutional network that implements exact
convolutions on the sphere by realizing them in the spherical harmonic domain.
Resulting filters have local symmetry and are localized by enforcing smooth
spectra. We apply a novel pooling on the spectral domain and our operations are
independent of the underlying spherical resolution throughout the network. We
show that networks with much lower capacity and without requiring data
augmentation can exhibit performance comparable to the state of the art in
standard retrieval and classification benchmarks. | ['Ameesh Makadia', 'Christine Allen-Blanchette', 'Kostas Daniilidis', 'Carlos Esteves'] | 2017-11-17 | learning-so3-equivariant-representations-with-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Carlos_Esteves_Learning_SO3_Equivariant_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Carlos_Esteves_Learning_SO3_Equivariant_ECCV_2018_paper.pdf | eccv-2018-9 | ['3d-classification'] | ['computer-vision'] | [ 5.68882637e-02 1.53031096e-01 2.80543983e-01 -2.33059496e-01
-4.14848119e-01 -9.50941443e-01 9.56188560e-01 -5.47357380e-01
-6.18654847e-01 1.82660744e-01 4.71896261e-01 -2.83492744e-01
-1.06111594e-01 -6.79221392e-01 -1.21621668e+00 -6.90058112e-01
-1.94533363e-01 2.18136296e-01 1.47602558e-01 -3.78815383e-01
1.75188091e-02 1.45882428e+00 -1.29645014e+00 2.93032289e-01
3.15095276e-01 8.03886831e-01 -3.03393871e-01 7.59090424e-01
2.30726212e-01 2.32159838e-01 -5.59199929e-01 -1.77905813e-01
8.78012896e-01 2.33989671e-01 -8.01754832e-01 -9.24681574e-02
7.28219628e-01 -1.20262295e-01 -7.86128402e-01 9.53840852e-01
7.53125489e-01 3.44347090e-01 1.02150309e+00 -7.79387891e-01
-1.06278253e+00 2.54677832e-01 -2.53953099e-01 2.39443585e-01
1.50588509e-02 -6.04598701e-01 8.92037988e-01 -1.47918594e+00
7.89571464e-01 1.44045615e+00 7.72017777e-01 4.90166128e-01
-1.36730778e+00 -2.43991897e-01 -1.13050446e-01 -2.91008592e-01
-1.58879316e+00 -4.71488863e-01 5.00793874e-01 -2.57260263e-01
1.35906029e+00 2.06124470e-01 2.07379878e-01 9.39078152e-01
1.61819696e-01 4.00563717e-01 7.35017478e-01 -3.45488280e-01
1.64101329e-02 -1.60538957e-01 -2.76460737e-01 6.26331985e-01
2.49608263e-01 -2.31488705e-01 -3.52395058e-01 2.95595437e-01
1.56596398e+00 1.61908492e-01 -3.04888397e-01 -7.20836699e-01
-1.40920305e+00 8.60653579e-01 9.18341696e-01 6.47141635e-02
-1.90465197e-01 1.68945566e-01 6.18021898e-02 4.59657758e-01
7.13496983e-01 7.97834456e-01 -4.80719715e-01 6.00741684e-01
-7.46443152e-01 3.22194546e-01 8.02270830e-01 1.13799274e+00
5.78577995e-01 7.58461133e-02 -1.04152821e-01 5.77796221e-01
-1.13891527e-01 6.12154841e-01 3.38623673e-01 -8.69236648e-01
1.66807473e-01 3.50046933e-01 3.10627613e-02 -7.22288251e-01
-8.49378705e-01 -8.00559103e-01 -1.00837135e+00 4.80107725e-01
3.89440894e-01 1.52733639e-01 -1.19526327e+00 1.55209982e+00
8.47390592e-02 2.69715279e-01 2.45917872e-01 8.54051948e-01
1.01382506e+00 4.62274581e-01 -7.68589437e-01 2.62314290e-01
1.10686016e+00 -6.17823660e-01 -2.13451788e-01 3.40590626e-01
4.18947011e-01 -1.02403295e+00 8.88791382e-01 1.59980863e-01
-1.41272056e+00 -4.15499777e-01 -1.34965134e+00 -5.74169636e-01
-6.83498919e-01 1.48466393e-01 4.49974835e-01 5.49697697e-01
-1.49615633e+00 6.80197656e-01 -9.16337907e-01 -2.35203937e-01
4.48365271e-01 4.93487388e-01 -8.05599332e-01 9.43411887e-02
-1.14645708e+00 1.11225343e+00 -1.22823536e-01 4.51673903e-02
-6.10452712e-01 -9.12696004e-01 -1.11774886e+00 2.41192684e-01
-2.89875478e-01 -5.62400341e-01 1.26469696e+00 -2.09256619e-01
-1.51699555e+00 1.12984002e+00 -8.01387802e-02 -6.69763982e-01
4.58267421e-01 -1.96801916e-01 -5.29962592e-02 2.48068154e-01
-1.81592047e-01 5.42046189e-01 1.04111075e+00 -6.92055464e-01
2.63357982e-02 -3.49402696e-01 4.02566642e-01 4.52858269e-01
-2.29086146e-01 1.98994819e-02 -2.94557452e-01 -7.64767289e-01
5.97659826e-01 -1.06326759e+00 -5.76525815e-02 9.59084649e-03
-2.18954057e-01 -2.05172598e-01 6.45849407e-01 -3.35601896e-01
4.96457428e-01 -2.13414574e+00 4.36499625e-01 4.38302726e-01
3.59470874e-01 8.78167227e-02 -3.07596982e-01 8.92070681e-03
-4.88709301e-01 3.14708382e-01 1.15822770e-01 -4.29655463e-01
2.74283797e-01 -5.39210699e-02 -5.69690168e-01 1.06734228e+00
6.58740759e-01 9.67756152e-01 -4.18983012e-01 1.37941718e-01
1.88388780e-01 1.28617859e+00 -8.58018994e-01 -3.44305485e-02
2.21092150e-01 7.16845244e-02 -1.23189948e-01 3.83240700e-01
8.61442864e-01 -4.39374179e-01 -4.11224961e-01 -4.77022558e-01
-2.57523537e-01 4.36981857e-01 -1.41240168e+00 1.78491414e+00
-5.82179427e-01 6.89997733e-01 5.90451173e-02 -1.28338587e+00
9.97920811e-01 2.05251411e-01 3.02957058e-01 -3.43860775e-01
7.81906545e-02 3.28468293e-01 -1.21093944e-01 9.76948291e-02
4.25489336e-01 7.34337121e-02 -9.28461850e-02 1.84455320e-01
5.13828337e-01 -5.87489009e-01 -1.72895446e-01 -4.44985218e-02
9.23200667e-01 1.25825450e-01 -1.11183673e-02 -7.10567892e-01
5.27940273e-01 -5.81298351e-01 1.31232277e-01 7.13902712e-01
3.34772557e-01 1.07701135e+00 5.81362486e-01 -7.13902056e-01
-1.37276816e+00 -1.47080898e+00 -4.87823278e-01 1.01037109e+00
-8.62628222e-03 -2.48891842e-02 -5.98638952e-01 -2.96474874e-01
-1.93046089e-02 9.89609882e-02 -7.95046866e-01 -9.32187662e-02
-6.23217583e-01 -7.17442513e-01 6.80066705e-01 6.38581812e-01
6.66434348e-01 -6.27226293e-01 -2.64351487e-01 -9.78832040e-03
4.27352458e-01 -1.54118454e+00 -6.16703272e-01 5.10727823e-01
-6.56640708e-01 -1.01200509e+00 -1.21303034e+00 -9.47003365e-01
7.56963789e-01 5.92162132e-01 1.11377931e+00 -4.77445692e-01
-3.70648354e-01 3.98821205e-01 -8.65390375e-02 -5.18069685e-01
1.69851974e-01 4.62296337e-01 4.24594790e-01 -2.70603240e-01
3.52342486e-01 -7.07807899e-01 -5.22759080e-01 4.73728538e-01
-8.78430665e-01 -3.65250558e-01 5.25330424e-01 7.03949809e-01
6.37276828e-01 -1.54055983e-01 2.77093530e-01 -4.44265455e-01
6.24695063e-01 -5.26176095e-02 -7.55311847e-01 -1.00014314e-01
1.47673726e-01 5.61307132e-01 5.72414458e-01 -2.50924468e-01
-6.95989132e-01 1.73262477e-01 7.79740959e-02 -6.16578281e-01
2.69593969e-02 6.00204803e-02 1.29525304e-01 -6.64804399e-01
1.16241086e+00 2.07902472e-02 7.95026571e-02 -5.14315486e-01
6.00969732e-01 3.01308453e-01 7.23630488e-01 -4.14202064e-01
1.34808969e+00 9.15382802e-01 5.70293605e-01 -1.12456822e+00
-9.35796380e-01 -4.09967631e-01 -9.50900555e-01 3.61127049e-01
1.13583755e+00 -1.24269664e+00 -6.82471037e-01 4.29761350e-01
-1.32892478e+00 -1.42290652e-01 -4.71351683e-01 6.53028131e-01
-6.57813370e-01 -1.03690606e-02 -3.97019863e-01 -2.88486868e-01
-3.60272229e-01 -1.21945727e+00 1.26045430e+00 2.06733480e-01
2.19592839e-01 -6.80733562e-01 -1.77501500e-01 -4.35266167e-01
7.62840807e-01 2.35696658e-01 6.42931819e-01 -8.35026860e-01
-6.92542493e-01 -2.56011099e-01 -5.50133646e-01 5.57594121e-01
-7.75702894e-02 -1.56222612e-01 -1.21897352e+00 -3.44707280e-01
-9.23643112e-02 -2.18161970e-01 1.17805946e+00 5.48413396e-01
1.23597395e+00 2.84089334e-02 1.80421963e-01 1.28944159e+00
1.02501178e+00 -4.71172541e-01 4.39288288e-01 2.20938921e-01
4.56498474e-01 3.26322943e-01 -2.76616156e-01 1.25271305e-01
-5.10851247e-03 3.81015986e-01 5.28479934e-01 -3.68789762e-01
-2.31144607e-01 1.34185329e-01 3.01483244e-01 4.75272775e-01
-5.78759253e-01 6.19534813e-02 -5.98488986e-01 4.02857214e-01
-1.56812990e+00 -8.18519473e-01 2.54085451e-01 2.33604813e+00
5.17555416e-01 2.83511877e-01 -1.99115440e-01 1.25124931e-01
4.89450485e-01 3.41421217e-01 -4.65306401e-01 -5.74711025e-01
-6.68160081e-01 1.10152602e+00 9.95480537e-01 6.60481036e-01
-1.61604631e+00 9.41896856e-01 6.85082197e+00 3.28768700e-01
-1.24518609e+00 -2.07004040e-01 4.90339585e-02 -2.15999603e-01
-2.08910570e-01 -6.71659112e-01 -9.30915117e-01 -3.26492488e-01
5.20833135e-01 -1.22469235e-02 6.42608225e-01 7.75071561e-01
-1.25186950e-01 7.74996936e-01 -1.13526773e+00 8.01352859e-01
2.43178502e-01 -1.69129074e+00 2.57929295e-01 1.56056816e-02
9.64506030e-01 6.58504546e-01 4.88822222e-01 -5.44927455e-02
3.59860927e-01 -1.63762522e+00 5.43450654e-01 4.86891985e-01
1.06394792e+00 -8.91865253e-01 5.70254266e-01 -1.00552954e-01
-1.18399286e+00 2.83069760e-01 -8.05818677e-01 1.71721876e-02
-2.57192880e-01 1.55779764e-01 -9.80103552e-01 3.75235677e-01
9.06204045e-01 6.86607897e-01 -3.53619367e-01 1.05615664e+00
-1.84619516e-01 -1.79148182e-01 -6.50272846e-01 2.21867338e-02
4.55328763e-01 -1.91709492e-02 6.40052676e-01 1.31164992e+00
6.97923183e-01 -1.30466893e-01 -2.00393707e-01 8.66143107e-01
-6.94481075e-01 -1.47737369e-01 -8.40250552e-01 2.70903081e-01
2.36971006e-01 1.13949001e+00 -5.49631774e-01 5.81708401e-02
-8.00548434e-01 9.92462873e-01 1.92395583e-01 7.60785878e-01
-4.81679887e-01 -8.46185684e-01 1.24629581e+00 1.13185130e-01
7.00652242e-01 -7.23329663e-01 -1.68131590e-01 -1.38089228e+00
-1.30487055e-01 -2.66471595e-01 4.96251918e-02 -7.49438822e-01
-1.08943486e+00 5.70181429e-01 -1.52566522e-01 -1.00914800e+00
-5.53605705e-02 -1.57276857e+00 -4.50226992e-01 1.22852778e+00
-1.79902756e+00 -1.02976954e+00 -3.34427804e-01 1.07045817e+00
-1.13720922e-02 -3.10319960e-01 1.00340188e+00 8.08616877e-02
1.99284106e-01 6.34808660e-01 1.61961317e-01 3.54931593e-01
8.68958950e-01 -1.58507252e+00 8.88672888e-01 7.98646867e-01
2.46027634e-01 9.00942385e-01 4.43394572e-01 1.20072775e-01
-1.45941138e+00 -1.07022119e+00 8.20731282e-01 -5.36005735e-01
5.51184952e-01 -7.02813506e-01 -8.80242109e-01 8.81682754e-01
5.16567789e-02 5.04425466e-01 4.73269552e-01 -5.76997511e-02
-9.62029636e-01 1.44375339e-01 -9.35963750e-01 6.32154882e-01
1.30147159e+00 -9.15071845e-01 -8.06253135e-01 4.66964811e-01
1.14923608e+00 -8.38509381e-01 -9.18635428e-01 5.40545940e-01
4.17354733e-01 -4.92553294e-01 1.58303571e+00 -9.82565522e-01
2.02610984e-01 -3.91478598e-01 -3.41120392e-01 -1.34378922e+00
-4.55551922e-01 -7.98165798e-01 1.16289824e-01 1.42737076e-01
6.14747941e-01 -6.38562202e-01 7.82148361e-01 1.89938083e-01
-5.55306256e-01 -3.65747958e-01 -1.21602309e+00 -7.08619297e-01
5.81261277e-01 -1.28080711e-01 5.71700156e-01 8.10886562e-01
-2.70336926e-01 2.55835131e-02 -1.06129140e-01 5.14377952e-01
6.07755840e-01 -2.94190161e-02 7.02459335e-01 -1.38339806e+00
1.87507853e-01 -7.22117186e-01 -6.67057574e-01 -1.32261419e+00
3.71144682e-01 -1.41455984e+00 -3.67975444e-01 -1.18418908e+00
-1.74463227e-01 1.66140318e-01 -2.65069872e-01 2.52323091e-01
4.53704476e-01 6.14924908e-01 -1.15328722e-01 -1.85695946e-01
-3.83903354e-01 3.55001986e-01 1.44342303e+00 2.15176009e-02
-2.17689723e-01 1.75728984e-02 -5.81107259e-01 7.71405995e-01
6.22451901e-01 1.99322745e-01 -2.58487135e-01 -6.79919600e-01
5.70950091e-01 -6.20885670e-01 6.14819944e-01 -7.98672259e-01
4.14575160e-01 4.57107961e-01 8.20659518e-01 -5.24699628e-01
5.50814033e-01 -6.12516463e-01 -4.14090514e-01 4.46784087e-02
-3.47352654e-01 1.25829786e-01 2.20104069e-01 3.00423324e-01
-1.52316913e-01 1.87719420e-01 9.15212214e-01 -1.75957695e-01
-2.29973093e-01 6.46676779e-01 -6.92759678e-02 3.46153267e-02
4.35372680e-01 1.02823414e-01 -4.09903646e-01 -3.29116762e-01
-1.12707758e+00 -1.30387023e-01 5.65673649e-01 3.40572178e-01
6.90551519e-01 -1.58584821e+00 -8.87017846e-01 6.26938045e-01
-1.57379955e-01 4.29066479e-01 -1.39704928e-01 5.01140356e-01
-8.44646871e-01 6.48257732e-01 -3.22871864e-01 -5.25149405e-01
-8.52202773e-01 2.27049127e-01 8.70999455e-01 3.41722034e-02
-5.48360109e-01 1.05056632e+00 3.33239377e-01 -8.09141338e-01
4.22977149e-01 -7.01240122e-01 -1.67788669e-01 5.35522476e-02
3.78312558e-01 7.59714916e-02 5.30178726e-01 -6.87653482e-01
-3.74506176e-01 7.99884140e-01 5.78287244e-02 3.99271324e-02
1.82363605e+00 3.57795715e-01 -7.26499036e-02 -1.16344124e-01
1.71216393e+00 -9.81423929e-02 -1.19316721e+00 -5.75316370e-01
-4.21426296e-01 -1.45679474e-01 1.21783599e-01 -1.38763472e-01
-7.82327771e-01 1.07508397e+00 2.05202207e-01 3.73834819e-01
6.75915658e-01 1.86802804e-01 1.61811307e-01 9.11601543e-01
-2.04224810e-01 -9.31754947e-01 8.82653370e-02 1.18490207e+00
1.33826149e+00 -9.80435669e-01 -6.53388500e-02 -3.09592068e-01
5.77464625e-02 1.43560112e+00 1.62061423e-01 -9.38341856e-01
1.08275783e+00 2.18520150e-01 -2.65808523e-01 -9.65862647e-02
-1.64142445e-01 -3.40108097e-01 8.46808255e-01 5.69362223e-01
6.22337520e-01 -1.43675432e-01 6.76880404e-02 4.83290344e-01
-4.70279634e-01 -6.46803617e-01 5.15975595e-01 4.48104769e-01
-4.36074018e-01 -6.61021590e-01 -3.92760307e-01 1.15591966e-01
-6.90208852e-01 -3.46059173e-01 -3.44867468e-01 8.96486223e-01
-3.37611586e-01 2.54169106e-01 3.50079626e-01 -6.85221329e-02
6.91697299e-01 1.21756390e-01 6.98176146e-01 -4.86380398e-01
1.03610987e-03 1.67350084e-01 -2.01185584e-01 -7.64640093e-01
-5.80723941e-01 -6.05583072e-01 -1.23464584e+00 -1.58161566e-01
3.71869467e-02 -2.43461579e-01 8.06618452e-01 5.10816336e-01
4.98850822e-01 5.29328287e-01 7.51516819e-01 -1.35965621e+00
-8.95704627e-01 -8.54916096e-01 -5.58726549e-01 2.09779263e-01
7.70904481e-01 -7.50383615e-01 -6.65552139e-01 -1.53151199e-01] | [8.885322570800781, 2.3575592041015625] |
152ba743-14cd-4772-80d2-3f6111862e03 | a-restarted-large-scale-spectral-clustering | 2306.15138 | null | https://arxiv.org/abs/2306.15138v2 | https://arxiv.org/pdf/2306.15138v2.pdf | A Restarted Large-Scale Spectral Clustering with Self-Guiding and Block Diagonal Representation | Spectral clustering is one of the most popular unsupervised machine learning methods. Constructing similarity matrix is crucial to this type of method. In most existing works, the similarity matrix is computed once for all or is updated alternatively. However, the former is difficult to reflect comprehensive relationships among data points, and the latter is time-consuming and is even infeasible for large-scale problems. In this work, we propose a restarted clustering framework with self-guiding and block diagonal representation. An advantage of the strategy is that some useful clustering information obtained from previous cycles could be preserved as much as possible. To the best of our knowledge, this is the first work that applies restarting strategy to spectral clustering. The key difference is that we reclassify the samples in each cycle of our method, while they are classified only once in existing methods. To further release the overhead, we introduce a block diagonal representation with Nystr\"{o}m approximation for constructing the similarity matrix. Theoretical results are established to show the rationality of inexact computations in spectral clustering. Comprehensive experiments are performed on some benchmark databases, which show the superiority of our proposed algorithms over many state-of-the-art algorithms for large-scale problems. Specifically, our framework has a potential boost for clustering algorithms and works well even using an initial guess chosen randomly. | ['Gang Wu', 'Yongyan Guo'] | 2023-06-27 | null | null | null | null | ['clustering'] | ['methodology'] | [ 1.16813146e-01 -4.89952087e-01 -2.11564332e-01 -2.55812500e-02
-4.85155374e-01 -4.55362648e-01 2.09333479e-01 9.03085619e-02
-3.60637963e-01 4.54600543e-01 -1.77355036e-01 -1.44622177e-01
-4.65929300e-01 -6.55063450e-01 -2.73173124e-01 -1.25942659e+00
-9.23759714e-02 4.49977547e-01 4.16094303e-01 -3.73436846e-02
4.48432922e-01 3.38379025e-01 -1.53799260e+00 1.18977517e-01
1.10028601e+00 8.80939424e-01 3.98315489e-01 1.94175974e-01
-1.93331152e-01 4.15141612e-01 -2.80505896e-01 -1.10498384e-01
4.70205903e-01 -4.76883948e-01 -7.22305536e-01 6.69248760e-01
-2.56824464e-01 1.45444512e-01 -3.43969315e-01 1.24988174e+00
4.16825503e-01 3.56813520e-01 5.91172636e-01 -1.33578575e+00
-3.45979840e-01 7.04273105e-01 -1.12365067e+00 3.24434377e-02
1.89559013e-01 -1.11420192e-01 9.43307996e-01 -8.13670814e-01
3.61886501e-01 9.27745402e-01 7.13910103e-01 1.62563529e-02
-1.13349545e+00 -7.78958142e-01 1.93645611e-01 4.82078224e-01
-1.93127286e+00 -3.72465998e-01 1.07305753e+00 -3.79662871e-01
2.86278367e-01 3.06900084e-01 6.61561966e-01 3.44535887e-01
-4.09291506e-01 7.72434056e-01 1.10282147e+00 -4.38037336e-01
3.75994623e-01 -1.06466003e-01 3.02041084e-01 5.08465230e-01
4.46297407e-01 -4.25255775e-01 -1.43588111e-01 -3.84287715e-01
5.36492169e-01 4.96806592e-01 -3.97711575e-01 -7.70080805e-01
-1.31412375e+00 7.93699920e-01 3.12308878e-01 5.00277519e-01
-1.65191904e-01 -2.18964249e-01 3.55203718e-01 1.48545757e-01
1.64922327e-01 2.49620769e-02 -9.91339907e-02 -8.17444250e-02
-1.09703100e+00 -1.40101537e-01 6.17189229e-01 8.94242287e-01
1.04521060e+00 -2.63378918e-01 3.94475341e-01 1.16136324e+00
1.79689154e-01 2.34003186e-01 7.76422858e-01 -7.76932001e-01
3.76050442e-01 7.70714283e-01 1.40383961e-02 -1.53805292e+00
-4.29905385e-01 -4.44709957e-01 -1.21997440e+00 -2.72166580e-01
3.78648639e-01 -4.58771735e-02 -4.98956770e-01 1.40077531e+00
4.17856753e-01 5.44899464e-01 -6.91676652e-03 8.45272243e-01
3.78060818e-01 7.75260091e-01 -4.24637377e-01 -8.18753421e-01
1.10462630e+00 -9.58824515e-01 -6.81151688e-01 6.23102486e-02
4.94661063e-01 -8.92749608e-01 7.32223392e-01 4.92925614e-01
-8.09072018e-01 -5.92466116e-01 -1.08588064e+00 4.06056792e-01
-1.32607624e-01 3.30425501e-01 8.67115200e-01 7.25397944e-01
-8.82784486e-01 4.93873030e-01 -9.02885914e-01 -4.82237577e-01
-3.09313415e-04 4.08033639e-01 -2.01583743e-01 -5.47936037e-02
-9.38559830e-01 1.83891580e-01 6.97203577e-01 3.96601647e-01
-2.01339737e-01 -2.48043314e-01 -4.65685368e-01 -4.60825823e-02
6.98755860e-01 -1.54667109e-01 8.90396059e-01 -8.01745713e-01
-1.19474924e+00 5.05874515e-01 -4.08045352e-01 -1.42183214e-01
3.46393734e-01 2.02785075e-01 -3.99626315e-01 2.79893786e-01
1.92086488e-01 5.23022525e-02 6.31264985e-01 -1.46798086e+00
-6.63014829e-01 -4.34070528e-01 -2.88226366e-01 3.43889713e-01
-7.20434427e-01 -2.75567830e-01 -1.09642160e+00 -6.25904143e-01
8.54985297e-01 -1.25762355e+00 -5.85456371e-01 -4.35805082e-01
-3.50177318e-01 -1.68067142e-01 7.58090317e-01 -3.23497266e-01
1.89185166e+00 -2.30184603e+00 7.89260268e-02 7.19159365e-01
1.54645994e-01 2.75268883e-01 2.27925226e-01 9.38262522e-01
-1.85304269e-01 -5.74010238e-02 -5.12248755e-01 -5.42675480e-02
-1.42812192e-01 1.40100509e-01 -1.60808891e-01 6.79249406e-01
-5.60029924e-01 2.94129044e-01 -9.11609173e-01 -7.20278502e-01
2.19664708e-01 6.64868429e-02 -4.01358366e-01 -1.16240546e-01
3.96135539e-01 1.67989537e-01 -4.11839455e-01 4.67850924e-01
1.01238382e+00 -4.64263052e-01 5.81954718e-01 -5.09421170e-01
-1.77470595e-01 -3.23938638e-01 -2.02721739e+00 1.67674005e+00
-4.05746931e-03 2.13267282e-02 1.10627994e-01 -1.45992565e+00
8.61181736e-01 2.17206359e-01 9.85057294e-01 -1.62552789e-01
5.36074527e-02 2.31020197e-01 6.06756061e-02 -2.95461565e-01
3.27388138e-01 -2.70822505e-03 8.23121741e-02 6.24419391e-01
-4.63394225e-01 1.67111441e-01 6.65224016e-01 3.28019828e-01
8.84997606e-01 -3.76205921e-01 5.43208718e-01 -3.35277379e-01
9.71041441e-01 8.83723944e-02 8.31224203e-01 5.84478378e-01
-1.41689017e-01 5.75939298e-01 2.85653591e-01 -1.62135690e-01
-8.45635116e-01 -7.81351089e-01 -8.22264478e-02 8.36865902e-01
5.41326284e-01 -6.23633206e-01 -7.94438124e-01 -4.04228359e-01
-1.56607255e-01 1.60460085e-01 -3.94331396e-01 -4.93908450e-02
-4.71490592e-01 -1.19937408e+00 1.91071182e-01 3.19503069e-01
6.22820139e-01 -5.62416553e-01 -1.39658004e-01 2.69271225e-01
-3.31843376e-01 -8.16022575e-01 -6.35807335e-01 -7.22550973e-02
-1.14745641e+00 -1.35417295e+00 -5.49686074e-01 -1.04120409e+00
1.08223891e+00 1.10021377e+00 4.81421739e-01 2.65971065e-01
-8.26513991e-02 4.52826209e-02 -6.97428405e-01 1.64250210e-01
3.22394189e-03 1.97389409e-01 1.99377418e-01 4.81223375e-01
4.01736528e-01 -7.67923474e-01 -7.94300914e-01 7.16063857e-01
-1.08832717e+00 1.33586787e-02 6.84821367e-01 8.95534456e-01
7.94218838e-01 8.86690140e-01 4.19144988e-01 -1.01699138e+00
6.12006128e-01 -5.18100917e-01 -4.71722901e-01 2.22121432e-01
-7.64639080e-01 -9.95661169e-02 1.01992464e+00 -3.86153221e-01
-9.22311187e-01 5.01515388e-01 1.91766858e-01 -4.00924921e-01
8.46356601e-02 5.38613975e-01 -2.13371009e-01 1.75630450e-02
3.02489936e-01 5.59400558e-01 4.88648638e-02 -5.43810189e-01
3.48918289e-01 9.11382198e-01 4.84792173e-01 -7.43138433e-01
1.04176259e+00 8.18817079e-01 6.74065053e-02 -8.28609467e-01
-5.26154935e-01 -1.12743783e+00 -8.23366344e-01 -5.18575646e-02
2.88028479e-01 -7.86604226e-01 -8.75815451e-01 4.36572105e-01
-6.04324102e-01 2.72415668e-01 2.37591267e-01 5.36393225e-01
-3.21785301e-01 1.01607406e+00 -4.29535985e-01 -8.38116169e-01
-1.66501150e-01 -1.21781516e+00 6.98888958e-01 3.94872986e-02
2.18219385e-01 -8.25953305e-01 1.26121834e-01 4.20545876e-01
-7.07903653e-02 1.19058125e-01 4.93463784e-01 -6.39137685e-01
-4.18108404e-01 -1.56505004e-01 -2.02028736e-01 4.97963512e-03
4.62606490e-01 1.23865686e-01 -5.07393122e-01 -6.41977131e-01
1.37406737e-01 1.41495019e-02 7.92117834e-01 1.09185360e-01
1.27630961e+00 -1.70920283e-01 -6.78307176e-01 4.59927827e-01
1.46065176e+00 6.10250831e-01 5.66562355e-01 3.00120980e-01
7.03445733e-01 3.82730961e-01 7.74408937e-01 6.87009811e-01
3.65494698e-01 6.06396914e-01 8.20742846e-02 -1.28768548e-01
3.49332809e-01 4.29642200e-03 5.05080596e-02 1.50319290e+00
-1.52300909e-01 1.48922741e-01 -9.20054555e-01 6.27021551e-01
-2.24681306e+00 -1.12993586e+00 -4.40314472e-01 2.26926565e+00
8.05310607e-01 -8.11128989e-02 2.30427146e-01 7.25085974e-01
1.20664120e+00 9.87690911e-02 -4.31540817e-01 1.16663747e-01
6.71074819e-03 -7.02120960e-02 3.64891350e-01 2.91686594e-01
-1.17796326e+00 6.06579959e-01 5.43945646e+00 1.21476614e+00
-8.06872368e-01 -6.48672581e-02 4.01567101e-01 1.40529722e-02
1.41937658e-01 3.30500245e-01 -5.71109593e-01 7.54584670e-01
4.18179989e-01 -2.71970153e-01 5.24828553e-01 8.22902739e-01
3.20163816e-01 -2.01474264e-01 -7.97150910e-01 1.34567714e+00
9.65522379e-02 -9.78683233e-01 -8.53231102e-02 4.85528586e-03
9.45799410e-01 -3.60906184e-01 -1.46464780e-01 -8.01394135e-02
2.24730894e-01 -5.50457001e-01 2.69554913e-01 1.94901079e-01
4.59753811e-01 -1.15593684e+00 6.87514901e-01 5.18553555e-01
-1.75995266e+00 -4.18463210e-03 -6.80457056e-01 5.92672564e-02
-1.05643291e-02 9.66797471e-01 -5.46369076e-01 1.01306379e+00
6.70768380e-01 7.58686244e-01 -5.43403447e-01 1.22461319e+00
1.73697948e-01 6.19033217e-01 -4.60029662e-01 -3.93446945e-02
3.60806227e-01 -8.48519385e-01 2.51907974e-01 1.19321454e+00
4.15349364e-01 3.58876079e-01 6.39759600e-01 2.30180711e-01
1.11428179e-01 6.02140844e-01 -5.09206653e-01 1.72039226e-01
7.66394556e-01 1.38085270e+00 -1.19401693e+00 -3.54192108e-01
-5.77041328e-01 8.58587265e-01 2.07374662e-01 2.57803321e-01
-6.87232554e-01 -7.26627171e-01 1.88229889e-01 4.52515483e-02
4.25049901e-01 -3.51650238e-01 -1.35389417e-01 -1.28334248e+00
2.37857364e-02 -9.55992639e-01 7.54877806e-01 -4.01878148e-01
-1.33841443e+00 3.15173537e-01 -7.90642202e-03 -1.60866892e+00
-5.26134036e-02 -2.00970262e-01 -4.02195305e-01 4.45357621e-01
-1.05006301e+00 -7.77577460e-01 -3.28492135e-01 8.67909789e-01
4.09659415e-01 -1.02411643e-01 4.09501374e-01 4.53933984e-01
-7.86978245e-01 4.99636054e-01 9.04448569e-01 1.53559506e-01
7.03093827e-01 -1.06600833e+00 -1.32309452e-01 1.01402771e+00
8.06222484e-02 1.11580813e+00 5.44178963e-01 -6.44417465e-01
-1.45424914e+00 -7.80146360e-01 2.88898408e-01 1.83610916e-01
8.87759447e-01 -1.77642003e-01 -9.47628558e-01 3.81036043e-01
4.38737124e-02 -2.19754249e-01 9.77328181e-01 1.02981180e-01
-1.05153255e-01 -4.46884930e-01 -8.25751424e-01 7.05979228e-01
1.06621885e+00 -2.39132240e-01 -5.11058092e-01 3.56152534e-01
2.24115402e-01 -1.16814055e-01 -9.10029829e-01 3.93085927e-01
3.90896410e-01 -1.19185507e+00 9.71746743e-01 3.99829596e-02
2.75854138e-03 -9.93210971e-01 -7.63861313e-02 -1.17178941e+00
-3.95557404e-01 -7.54653573e-01 3.38565521e-02 1.38021374e+00
9.93851274e-02 -7.20061839e-01 9.61237192e-01 2.08493218e-01
7.63485432e-02 -6.54480636e-01 -6.66925609e-01 -9.22547400e-01
-2.46202409e-01 -2.62475133e-01 6.01089299e-01 1.30407834e+00
2.78087437e-01 1.93370283e-01 -4.60960090e-01 2.58550406e-01
8.29410553e-01 6.70526922e-01 7.87486315e-01 -1.29847038e+00
-3.83836597e-01 -4.24632579e-01 -3.32473576e-01 -1.15818346e+00
-1.62029237e-01 -8.74338210e-01 -1.17827982e-01 -1.33875787e+00
6.08211040e-01 -6.30992055e-01 -3.17878157e-01 2.73524672e-01
-4.24236357e-01 1.60919070e-01 1.15078844e-01 7.70259738e-01
-7.68985450e-01 3.44155043e-01 1.12266135e+00 8.90374854e-02
-4.12878990e-01 4.62056249e-02 -7.29189098e-01 7.71063387e-01
9.05723453e-01 -3.75624120e-01 -8.03407609e-01 -5.50916605e-02
-1.06407665e-01 -1.00573497e-02 -2.47388616e-01 -1.06655383e+00
7.51131892e-01 -3.26128036e-01 1.08115852e-01 -1.00259638e+00
7.13976622e-02 -1.08318138e+00 5.49822688e-01 4.67401773e-01
9.76411924e-02 1.43596891e-03 -1.37243658e-01 8.06319296e-01
-3.73257488e-01 -3.71931434e-01 8.13744664e-01 -1.75955176e-01
-5.64428627e-01 4.01252627e-01 -2.05113336e-01 -9.17259678e-02
1.15541816e+00 -4.98214602e-01 6.12600259e-02 -2.21466437e-01
-5.14214873e-01 3.31704617e-01 7.40745783e-01 3.53221521e-02
4.52058017e-01 -1.53014362e+00 -4.54267085e-01 1.77462567e-02
1.15495212e-01 7.20074773e-02 3.57107878e-01 9.37159777e-01
-6.23318017e-01 4.57944721e-01 1.06071033e-01 -7.01641083e-01
-1.22778821e+00 9.97394860e-01 -1.66215867e-01 -1.87732488e-01
-5.63802779e-01 3.24925423e-01 1.89736202e-01 -2.73926705e-01
1.23808302e-01 1.14423372e-01 -1.95034817e-01 2.26127967e-01
4.91670340e-01 5.72370887e-01 -5.60044637e-03 -6.57421112e-01
-3.17179680e-01 1.02240121e+00 -1.71963140e-01 -5.02356626e-02
1.18411720e+00 -4.41118151e-01 -5.46151042e-01 4.61915642e-01
1.14859009e+00 1.91690609e-01 -8.34816933e-01 -4.19202834e-01
-7.07619824e-03 -5.77179611e-01 -2.28020176e-01 1.63642503e-02
-1.25194633e+00 5.95684469e-01 3.57148409e-01 4.53460127e-01
1.45779574e+00 -2.97888875e-01 8.82709086e-01 7.41077125e-01
4.93802756e-01 -1.31945872e+00 -6.74971715e-02 7.55480900e-02
3.90857249e-01 -1.07075727e+00 3.31382513e-01 -6.97071135e-01
-5.65203667e-01 9.99308228e-01 5.36726654e-01 -5.80758341e-02
5.80338299e-01 -4.83848453e-02 -1.79233581e-01 -3.61231379e-02
-3.66651386e-01 -3.41413766e-01 2.40844232e-03 3.20707321e-01
2.39438012e-01 -3.06805354e-02 -8.09333146e-01 4.88032430e-01
5.84772155e-02 -1.92946255e-01 3.95523220e-01 1.00737953e+00
-6.57183111e-01 -1.29146218e+00 -8.19216609e-01 3.37026775e-01
-1.93186626e-01 2.29457431e-02 -2.75750160e-01 6.53861642e-01
6.72943443e-02 1.24627185e+00 -2.61064231e-01 -3.82131428e-01
1.48741543e-01 -6.35253564e-02 7.81163350e-02 -3.75914633e-01
-2.29728520e-01 5.29631674e-01 -3.25733513e-01 -3.32415015e-01
-7.30979800e-01 -8.98272276e-01 -1.43431139e+00 -4.45540369e-01
-4.43019122e-01 7.17258394e-01 3.15733850e-01 5.41847348e-01
2.66589195e-01 1.79317310e-01 1.11032760e+00 -5.92263401e-01
-3.51248950e-01 -6.46539211e-01 -8.27491224e-01 5.56386948e-01
-1.62616700e-01 -8.00726354e-01 -4.98674035e-01 2.47103795e-01] | [7.576948165893555, 4.727859020233154] |
de9f7847-8660-43fc-8f97-1e46ff293843 | domain-specific-chatbots-for-science-using | 2306.10067 | null | https://arxiv.org/abs/2306.10067v1 | https://arxiv.org/pdf/2306.10067v1.pdf | Domain-specific ChatBots for Science using Embeddings | Large language models (LLMs) have emerged as powerful machine-learning systems capable of handling a myriad of tasks. Tuned versions of these systems have been turned into chatbots that can respond to user queries on a vast diversity of topics, providing informative and creative replies. However, their application to physical science research remains limited owing to their incomplete knowledge in these areas, contrasted with the needs of rigor and sourcing in science domains. Here, we demonstrate how existing methods and software tools can be easily combined to yield a domain-specific chatbot. The system ingests scientific documents in existing formats, and uses text embedding lookup to provide the LLM with domain-specific contextual information when composing its reply. We similarly demonstrate that existing image embedding methods can be used for search and retrieval across publication figures. These results confirm that LLMs are already suitable for use by physical scientists in accelerating their research efforts. | ['Kevin G. Yager'] | 2023-06-15 | null | null | null | null | ['chatbot', 'chatbot'] | ['methodology', 'natural-language-processing'] | [-2.17533223e-02 -1.73537377e-02 -2.28107721e-01 -2.68579304e-01
-1.10815823e+00 -8.20995271e-01 8.44884932e-01 3.36383224e-01
-4.43212777e-01 6.24257505e-01 3.62018287e-01 -6.33000076e-01
-1.24344088e-01 -5.68929076e-01 -4.60556597e-01 -3.82742882e-01
2.19867975e-01 5.85055053e-01 1.77072749e-01 -2.53806114e-01
7.01151609e-01 6.36346579e-01 -1.04782736e+00 4.95585352e-01
3.29895824e-01 4.16951925e-01 3.50190461e-01 7.88114130e-01
-7.84942627e-01 9.58337903e-01 -9.44375873e-01 -5.80466688e-01
-2.51355499e-01 -1.57954425e-01 -1.09906256e+00 -6.74341559e-01
5.17076194e-01 -2.50756621e-01 -4.68411267e-01 6.81567371e-01
5.86438715e-01 -6.38497919e-02 4.48631316e-01 -1.19294608e+00
-9.57759857e-01 3.90137881e-01 -2.00057194e-01 3.86933357e-01
7.42867112e-01 1.55289456e-01 1.13773108e+00 -8.93217027e-01
9.56663251e-01 1.41217542e+00 4.06091690e-01 5.59583187e-01
-1.19397259e+00 -6.53499603e-01 -3.44948977e-01 3.66336137e-01
-7.08325922e-01 -3.95299971e-01 8.54133904e-01 -6.82422161e-01
1.08143461e+00 3.22710365e-01 3.94955397e-01 1.21384645e+00
2.91364551e-01 4.49347943e-01 1.05073082e+00 -4.89956081e-01
1.78457782e-01 5.35785496e-01 2.03291610e-01 6.65386081e-01
4.47694249e-02 -3.72463286e-01 -9.35227573e-01 -5.00800192e-01
8.51618409e-01 -3.87318172e-02 -5.63999191e-02 1.91055849e-01
-1.61120379e+00 8.77381682e-01 4.25721467e-01 6.00174010e-01
-1.20582893e-01 2.73531139e-01 6.22336984e-01 6.35942876e-01
4.56592530e-01 1.12094939e+00 -2.43181989e-01 -3.58972341e-01
-7.69069552e-01 3.94371063e-01 1.06492853e+00 8.16013813e-01
6.92293048e-01 -2.21170574e-01 -1.09738097e-01 6.70419872e-01
3.27360451e-01 5.31618372e-02 2.36666337e-01 -1.30251181e+00
1.70206755e-01 7.09368885e-01 3.08294237e-01 -1.00340152e+00
-2.36324996e-01 -9.67101231e-02 -3.38492841e-01 3.61360937e-01
3.61505181e-01 2.04048902e-01 -3.63307655e-01 1.35347319e+00
2.88223922e-01 4.09769826e-02 -6.88643679e-02 7.48528242e-01
1.20518851e+00 7.81440675e-01 2.54050106e-01 3.10879290e-01
1.68995130e+00 -5.54233432e-01 -7.39580572e-01 1.29005248e-02
6.82181180e-01 -1.15148234e+00 1.38555443e+00 2.54951298e-01
-1.27109206e+00 -3.76229197e-01 -8.62029195e-01 -7.33599663e-01
-8.05754364e-01 -1.89750344e-01 5.92140794e-01 3.57610792e-01
-1.27163243e+00 6.63509429e-01 -4.24393356e-01 -4.30408925e-01
3.94863427e-01 1.85008161e-02 -4.00106639e-01 -6.99494034e-02
-9.69033897e-01 1.03417063e+00 -2.55887061e-01 -2.89042324e-01
-5.67603111e-01 -1.07085586e+00 -4.48310614e-01 5.83337210e-02
1.93162695e-01 -6.99845254e-01 1.36135507e+00 -2.09698692e-01
-1.47213924e+00 1.13913274e+00 -9.87308845e-02 -3.10306698e-02
3.31358343e-01 -1.35172337e-01 -2.68214226e-01 5.06374538e-01
2.33925924e-01 6.18231773e-01 5.86192846e-01 -8.94662082e-01
-5.85204251e-02 -1.18458435e-01 2.85387963e-01 -2.15735257e-01
-4.51126397e-01 6.31353259e-01 -3.71374428e-01 -4.09612060e-01
-2.90734798e-01 -4.82500285e-01 -1.29476443e-01 4.73945111e-01
-1.79071739e-01 -3.73681486e-01 1.17528212e+00 -5.76363921e-01
9.11746979e-01 -1.85782540e+00 1.20598480e-01 -2.45554984e-01
5.56672215e-01 2.81782180e-01 -2.25621074e-01 9.92965102e-01
3.43850523e-01 4.85405266e-01 1.47546917e-01 -2.79883206e-01
3.33971351e-01 8.16876888e-02 -2.55018294e-01 3.36669475e-01
1.97401226e-01 1.16948676e+00 -9.59791303e-01 -5.94497323e-01
2.39965260e-01 5.18831968e-01 -3.62473786e-01 3.82956833e-01
-5.66947460e-01 4.21709299e-01 -6.82872713e-01 5.38490355e-01
2.13723958e-01 -6.88869953e-01 6.02704175e-02 -5.45511162e-03
-3.41556519e-01 6.37955785e-01 -6.01357400e-01 1.64876187e+00
-6.53140783e-01 1.03331089e+00 3.83430600e-01 -9.09483075e-01
8.72494698e-01 4.22277838e-01 3.88304770e-01 -3.90588135e-01
-1.38430431e-01 1.29716426e-01 -3.11213821e-01 -8.14031005e-01
5.36465228e-01 3.99560221e-02 -1.20142609e-01 1.01424348e+00
1.66016400e-01 -3.76345754e-01 3.34832259e-02 6.62904739e-01
1.26587248e+00 7.95812011e-02 -4.49703075e-02 -1.42410800e-01
2.87234932e-01 2.24029154e-01 -1.44530535e-01 8.21471572e-01
-1.42173082e-01 3.61660510e-01 5.53839624e-01 -7.33228028e-01
-1.51226079e+00 -7.97484279e-01 -6.91978484e-02 1.25478923e+00
-2.43749008e-01 -6.17627859e-01 -4.45348591e-01 -3.93560618e-01
-1.73309557e-02 5.10014534e-01 -3.53525609e-01 1.55136168e-01
-4.25407410e-01 -2.87592590e-01 5.01956999e-01 2.77214289e-01
1.13671750e-01 -1.39105129e+00 -7.52094805e-01 2.91002452e-01
-7.15550557e-02 -1.13503397e+00 -2.22699672e-01 -1.04595862e-01
-4.63863879e-01 -1.03195572e+00 -8.60099971e-01 -8.47266555e-01
3.08013618e-01 3.39212149e-01 1.26537240e+00 2.69702822e-01
-7.35701621e-01 7.85706162e-01 -1.90303057e-01 -3.96682918e-01
-7.92917013e-01 1.05881952e-01 -1.88916147e-01 -6.65391147e-01
4.66780871e-01 -5.87947309e-01 -4.98528808e-01 1.21906675e-01
-1.06646371e+00 -1.76260456e-01 4.06931460e-01 7.04135716e-01
1.45902596e-02 -7.31105328e-01 8.81864786e-01 -9.15673018e-01
9.87465799e-01 -5.53734958e-01 -5.17267942e-01 2.87198961e-01
-5.30625522e-01 3.30465771e-02 4.69311535e-01 -3.83299619e-01
-7.96615243e-01 -5.92536449e-01 -2.98330486e-01 -1.41530275e-01
-2.44568035e-01 7.12077856e-01 2.03334749e-01 -3.72422904e-01
6.44371450e-01 2.70314724e-03 1.64052561e-01 -7.42164433e-01
4.20046955e-01 7.87490368e-01 4.58809406e-01 -6.36723757e-01
5.31803071e-01 2.43231386e-01 -7.82296583e-02 -1.03843653e+00
-5.70450962e-01 -6.38021767e-01 -2.81611323e-01 -2.66623616e-01
7.88459420e-01 -5.68195879e-01 -1.09065127e+00 -1.79430153e-02
-1.45841587e+00 -1.66437820e-01 -1.77346930e-01 4.44998592e-01
-4.16902989e-01 5.45611680e-01 -8.87757480e-01 -5.14026225e-01
-4.43298519e-01 -9.55649078e-01 1.08698463e+00 2.23555699e-01
-5.06883442e-01 -1.08817232e+00 1.76665708e-01 4.79713798e-01
7.95174778e-01 5.31391129e-02 1.27775085e+00 -6.32115781e-01
-7.92557359e-01 -4.66485143e-01 -4.01056826e-01 -2.15435952e-01
-1.93957478e-01 1.71347722e-01 -1.04735303e+00 -7.42641836e-03
-2.77137130e-01 -6.10326827e-01 4.80435938e-01 7.38919303e-02
1.07157660e+00 -5.34159422e-01 -3.81374180e-01 2.25094691e-01
1.20387840e+00 -1.78890765e-01 6.33190870e-01 5.91482341e-01
4.33477521e-01 6.79756165e-01 4.89335449e-04 3.05718482e-01
1.99227735e-01 7.46186137e-01 1.21603690e-01 -3.14316377e-02
8.80601071e-03 -1.81751683e-01 2.00120673e-01 7.30251968e-01
1.97325870e-01 -2.41795763e-01 -8.72512817e-01 3.45402002e-01
-1.79115546e+00 -1.17855155e+00 -1.94714189e-01 1.79326701e+00
9.92137492e-01 -1.05335619e-02 2.29804292e-02 -4.18793023e-01
2.10707009e-01 2.86109746e-01 -4.92430121e-01 -6.18182182e-01
7.10712448e-02 2.70286262e-01 1.94680616e-01 4.60935235e-01
-5.87376833e-01 9.75033641e-01 7.76383018e+00 6.31109357e-01
-1.05241120e+00 3.00254524e-01 3.39245856e-01 1.54979425e-02
-7.12524831e-01 4.03365567e-02 -4.88272846e-01 2.98924655e-01
1.28302550e+00 -3.68790507e-01 3.15983385e-01 7.37247467e-01
3.44396323e-01 1.47679180e-01 -1.08906579e+00 9.04065430e-01
-2.25142941e-01 -2.30195618e+00 6.18601358e-03 -2.24255681e-01
2.68421650e-01 6.42296746e-02 2.83172037e-02 2.74407744e-01
4.12171900e-01 -1.24051237e+00 2.62953132e-01 5.68595648e-01
8.62163484e-01 -2.26007342e-01 5.16651273e-01 2.03523949e-01
-5.69245875e-01 2.86178052e-01 -5.12515068e-01 -2.59737581e-01
1.63929954e-01 4.69443858e-01 -9.93361592e-01 2.85086930e-01
7.77164459e-01 4.75506842e-01 -4.32447612e-01 9.22855854e-01
-6.59184158e-02 6.86226904e-01 -1.89245567e-01 -5.62443554e-01
2.45923147e-01 -7.01758787e-02 5.04642487e-01 1.53150010e+00
5.65400980e-02 -1.30104706e-01 4.31539603e-02 1.26216841e+00
-3.80772352e-01 1.41417861e-01 -7.82168329e-01 -6.65465653e-01
5.39925277e-01 1.41773701e+00 -5.59883714e-01 -4.04649973e-01
-4.78013575e-01 8.13653052e-01 3.70082945e-01 4.51350778e-01
-4.15029854e-01 -6.48979366e-01 6.71196461e-01 1.70568705e-01
-1.26911223e-01 -5.69109380e-01 -3.41100246e-01 -9.73227203e-01
-1.58173695e-01 -8.85137141e-01 1.83293521e-01 -1.20934558e+00
-1.38730633e+00 4.13876921e-01 -1.05087169e-01 -6.66821301e-01
-3.47072855e-02 -7.74816096e-01 -8.17479372e-01 9.62699950e-01
-1.31686401e+00 -1.22068906e+00 -1.23212136e-01 2.89253473e-01
4.85237479e-01 -2.17924669e-01 1.08567822e+00 2.44263396e-01
-2.74158210e-01 3.26201767e-01 1.97792262e-01 2.54817568e-02
8.92586768e-01 -1.14211118e+00 4.25539374e-01 8.84280205e-02
3.48561257e-01 9.31278348e-01 8.93056512e-01 -4.13163066e-01
-1.80873871e+00 -4.91175771e-01 1.37734866e+00 -7.66785920e-01
1.01660597e+00 -6.14244878e-01 -1.15826797e+00 5.67432880e-01
5.25631130e-01 -3.48619938e-01 8.83710563e-01 2.60777086e-01
-6.33871317e-01 3.46875668e-01 -1.11177099e+00 6.95227563e-01
5.75879276e-01 -9.33355451e-01 -5.65811634e-01 9.51612890e-01
8.63602102e-01 -1.89081490e-01 -8.91269326e-01 -3.99479210e-01
4.97759432e-01 -4.86275226e-01 1.12792361e+00 -1.08616531e+00
8.80525768e-01 -7.04680160e-02 1.13282725e-01 -8.69681299e-01
-4.14464146e-01 -1.06077397e+00 1.67657018e-01 1.06855416e+00
3.71714085e-01 -4.73251760e-01 6.82720959e-01 8.23170602e-01
-1.65127724e-01 -3.75477463e-01 -9.30484831e-01 -4.69603717e-01
2.31436491e-01 -2.98097491e-01 3.53807837e-01 1.07259011e+00
5.74870646e-01 4.99088019e-01 -3.18474501e-01 -2.48048112e-01
2.72448272e-01 2.14160547e-01 8.61978054e-01 -1.30373621e+00
-2.32951745e-01 -5.96998632e-01 -3.56310964e-01 -7.54080474e-01
1.39882043e-01 -1.02904654e+00 -2.26510003e-01 -1.73645592e+00
1.98614582e-01 -4.03010279e-01 -1.03955202e-01 5.11036694e-01
1.46018445e-01 2.18410432e-01 6.96759298e-02 4.49373186e-01
-5.48964560e-01 1.38032421e-01 9.69446123e-01 -1.90828845e-01
6.40242845e-02 -3.43657076e-01 -8.07623625e-01 4.76844013e-01
7.69958079e-01 -4.47647840e-01 5.21760769e-02 -2.49239534e-01
5.80825627e-01 -1.69377699e-01 7.78294802e-01 -7.61607349e-01
3.41138870e-01 -2.87097156e-01 2.25758582e-01 -2.15999961e-01
2.88621932e-01 -4.65790302e-01 -4.28755134e-02 1.36036173e-01
-8.63015234e-01 2.16003165e-01 3.13401371e-01 4.67477173e-01
-2.22903974e-02 -3.65967691e-01 4.97063398e-01 -4.02699322e-01
-5.10481477e-01 7.63762370e-02 -6.37286305e-01 -5.37510030e-03
6.21281207e-01 8.44985992e-02 -7.46035933e-01 -6.74134135e-01
-3.98981541e-01 5.20736389e-02 5.87054253e-01 3.24866623e-01
6.04263961e-01 -9.96722817e-01 -6.25318468e-01 -1.62623614e-01
2.32125819e-02 -4.88951892e-01 1.92276910e-01 6.37496889e-01
-6.60198450e-01 5.63875318e-01 -2.51417041e-01 -3.97380203e-01
-1.13201344e+00 6.65809929e-01 -1.74558148e-01 -1.68070465e-01
-8.23140562e-01 8.21454525e-01 -1.45688787e-01 -4.35487747e-01
2.89421052e-01 1.72679033e-03 -1.66088548e-02 -1.16535299e-01
8.60924780e-01 2.42889985e-01 4.01716009e-02 -1.30748391e-01
-3.18440616e-01 3.18194777e-01 -1.51768357e-01 -3.17874610e-01
1.67846382e+00 1.25136241e-01 -1.46737605e-01 5.80182731e-01
1.46347034e+00 1.76587656e-01 -6.92997456e-01 -2.84054905e-01
1.21069483e-01 -2.84034401e-01 -2.03636847e-02 -9.43323433e-01
-3.46713632e-01 1.17675304e+00 1.14661880e-01 3.48118395e-01
3.56343985e-01 5.13008177e-01 8.46560597e-01 6.28826022e-01
8.62728655e-02 -9.46210682e-01 4.84225631e-01 3.30167234e-01
9.80088174e-01 -1.18605328e+00 8.56577605e-02 2.23776847e-02
-4.77995574e-01 1.49867153e+00 1.34611547e-01 1.53173894e-01
4.64083999e-01 4.29462790e-01 2.97755837e-01 -7.02933371e-01
-9.02009368e-01 3.44854474e-01 4.75300029e-02 6.38042212e-01
8.07360351e-01 -2.05743492e-01 -3.38466883e-01 1.91830337e-01
-7.61443302e-02 1.35339335e-01 7.19822049e-01 1.14382911e+00
-6.83458209e-01 -1.31872642e+00 -3.61727744e-01 3.41278285e-01
-6.21411383e-01 -1.09431066e-01 -6.75179005e-01 5.05863249e-01
-5.04708588e-01 7.81166613e-01 -8.48773122e-02 -1.61699206e-01
-4.33617570e-02 3.71077329e-01 2.83566564e-01 -7.72315800e-01
-7.85550833e-01 -2.45901555e-01 1.41515225e-01 -5.71555495e-01
-3.20879668e-01 -5.02796173e-01 -1.03656769e+00 -5.36266327e-01
1.31641805e-01 3.71252775e-01 9.46407020e-01 7.35363960e-01
7.72511244e-01 2.82191306e-01 9.26461071e-02 -8.40881526e-01
-5.92756093e-01 -7.36389160e-01 -1.23296008e-01 2.76664913e-01
3.01074386e-01 -3.76472592e-01 -2.37288028e-01 8.97974819e-02] | [10.473663330078125, 8.328375816345215] |
adf1c6a5-499a-4865-b160-6b7c54299b95 | time-perception-machine-temporal-point | 1808.04063 | null | http://arxiv.org/abs/1808.04063v2 | http://arxiv.org/pdf/1808.04063v2.pdf | Time Perception Machine: Temporal Point Processes for the When, Where and What of Activity Prediction | Numerous powerful point process models have been developed to understand
temporal patterns in sequential data from fields such as health-care,
electronic commerce, social networks, and natural disaster forecasting. In this
paper, we develop novel models for learning the temporal distribution of human
activities in streaming data (e.g., videos and person trajectories). We propose
an integrated framework of neural networks and temporal point processes for
predicting when the next activity will happen. Because point processes are
limited to taking event frames as input, we propose a simple yet effective
mechanism to extract features at frames of interest while also preserving the
rich information in the remaining frames. We evaluate our model on two
challenging datasets. The results show that our model outperforms traditional
statistical point process approaches significantly, demonstrating its
effectiveness in capturing the underlying temporal dynamics as well as the
correlation within sequential activities. Furthermore, we also extend our model
to a joint estimation framework for predicting the timing, spatial location,
and category of the activity simultaneously, to answer the when, where, and
what of activity prediction. | ['Guang-Tong Zhou', 'Yatao Zhong', 'Luke Bornn', 'Greg Mori', 'Bicheng Xu'] | 2018-08-13 | null | null | null | null | ['activity-prediction', 'activity-prediction'] | ['computer-vision', 'time-series'] | [ 5.85095249e-02 -3.61743867e-01 -4.00091231e-01 -3.03325176e-01
-1.68137431e-01 -2.74472564e-01 7.40716517e-01 4.79351819e-01
-3.78309220e-01 4.75461304e-01 7.86724627e-01 -6.82151467e-02
-3.57391238e-01 -8.93969119e-01 -7.09033430e-01 -4.54523116e-01
-5.44422984e-01 3.06506455e-01 2.92703658e-01 1.85619727e-01
-5.80370193e-03 3.58966559e-01 -1.20610261e+00 1.94029614e-01
4.40370381e-01 9.02246416e-01 1.71983112e-02 6.41516149e-01
4.01322246e-02 1.25897884e+00 -3.77032042e-01 -4.36110348e-02
8.97078440e-02 -2.76022583e-01 -4.53952461e-01 2.61342287e-01
-9.16579813e-02 -6.24366105e-01 -9.34558451e-01 4.18428898e-01
4.68799472e-02 4.50527757e-01 6.98345542e-01 -1.40205550e+00
-3.77132982e-01 3.29986721e-01 -6.90622151e-01 6.36893332e-01
5.19652605e-01 1.21039324e-01 7.60411978e-01 -6.03107989e-01
2.76444256e-01 1.18218732e+00 8.73913348e-01 2.88085222e-01
-8.50921631e-01 -4.48439121e-01 6.45477891e-01 4.03456390e-01
-1.08680987e+00 -2.22476736e-01 6.58509493e-01 -5.94145417e-01
6.76783562e-01 -8.76033157e-02 7.18584180e-01 1.44092703e+00
3.73515368e-01 1.13945210e+00 5.38914382e-01 5.50549962e-02
2.69012451e-01 -5.89219451e-01 1.30849347e-01 4.35767233e-01
-2.83268914e-02 -1.23917811e-01 -8.70087683e-01 -5.09916723e-01
9.11211312e-01 1.01957166e+00 -1.17564350e-01 -3.26693468e-02
-1.37635601e+00 4.56030518e-01 5.15095368e-02 8.11045989e-02
-9.55574751e-01 4.07138079e-01 2.49280781e-01 -1.67820841e-01
7.34914124e-01 -2.29239672e-01 -5.07244587e-01 -4.69838113e-01
-9.86799836e-01 4.19039041e-01 7.32702732e-01 7.81739891e-01
4.63545352e-01 -3.95824105e-01 -7.41319001e-01 4.35060680e-01
1.85766533e-01 5.18982828e-01 1.70016348e-01 -9.82449114e-01
6.74400866e-01 5.18761873e-01 4.84985322e-01 -1.30384409e+00
-4.45400566e-01 2.07305700e-02 -9.96937275e-01 -7.42766440e-01
5.07716715e-01 -5.28870761e-01 -6.47996306e-01 1.62742782e+00
2.73167163e-01 1.00073123e+00 -4.39251810e-01 4.58717704e-01
3.17544758e-01 1.11013710e+00 3.75693083e-01 -3.20061952e-01
1.23519361e+00 -7.47781873e-01 -7.92462707e-01 -2.44973507e-02
3.27870846e-01 -8.53489935e-02 5.98204732e-01 1.68440789e-01
-1.02579117e+00 -5.46686947e-01 -2.12225795e-01 1.50533751e-01
7.33519346e-02 -1.62689276e-02 5.73677182e-01 1.00696437e-01
-7.10558355e-01 7.41166770e-01 -1.41506517e+00 -4.63670790e-01
7.02810347e-01 8.25818405e-02 -7.51862377e-02 -4.75729518e-02
-9.67447102e-01 2.81585902e-01 1.41246364e-01 -4.92093246e-03
-9.44089115e-01 -7.48689830e-01 -6.45711064e-01 4.42112982e-01
5.57179511e-01 -6.29330993e-01 1.23545325e+00 -6.51028216e-01
-1.07723880e+00 1.29567310e-01 -7.73298264e-01 -8.04617107e-01
5.07218122e-01 -4.40339535e-01 -4.12794918e-01 3.04087043e-01
1.91422626e-01 3.60810578e-01 6.05179191e-01 -6.08026385e-01
-1.15759158e+00 -2.39609346e-01 -1.44391865e-01 2.63222843e-01
-5.89384556e-01 1.48355007e-01 -6.57126129e-01 -7.20596313e-01
-1.81773588e-01 -8.45579505e-01 -3.76401871e-01 1.11888036e-01
-1.57848880e-01 -5.38776278e-01 7.44329035e-01 -9.82729554e-01
1.30162597e+00 -2.14117050e+00 2.94169262e-02 1.00954391e-01
2.59473532e-01 -1.95252433e-01 1.66123584e-01 7.42763638e-01
5.16665280e-01 5.45839109e-02 -6.89663440e-02 -4.74039555e-01
-1.22618906e-01 3.73468727e-01 -4.77162272e-01 5.09820461e-01
1.10669732e-01 8.84820044e-01 -1.05017436e+00 -3.61163676e-01
1.16216928e-01 6.49573028e-01 -3.15227032e-01 1.90576226e-01
-1.93016693e-01 6.64205372e-01 -6.79439366e-01 4.38880712e-01
1.66190431e-01 -5.60944438e-01 1.69999674e-01 2.83222526e-01
4.88565043e-02 1.68532968e-01 -1.04514360e+00 1.27425051e+00
-1.84122875e-01 6.25674367e-01 -3.67164761e-01 -8.84827256e-01
5.79692721e-01 5.58194339e-01 1.14078164e+00 -5.31625152e-01
-3.15882117e-01 -4.62088197e-01 -5.26097596e-01 -5.62579274e-01
4.86405343e-01 1.04358243e-02 -2.07806453e-01 6.66679025e-01
-1.16978996e-01 8.56959343e-01 2.18826890e-01 3.50975841e-02
1.43117189e+00 2.20487684e-01 2.70416081e-01 1.44466996e-01
1.91376254e-01 -2.17485338e-01 8.62543225e-01 9.02921438e-01
-4.46374625e-01 4.74538416e-01 6.45659685e-01 -9.40959454e-01
-8.78598690e-01 -1.00038159e+00 5.00452697e-01 1.05784345e+00
1.59734219e-01 -3.35056126e-01 -4.53775227e-01 -6.95149422e-01
-8.64997730e-02 4.20941323e-01 -7.10831881e-01 1.82450432e-02
-8.63973439e-01 -8.29234600e-01 3.09562504e-01 8.23924601e-01
5.49735546e-01 -1.12415290e+00 -4.46688831e-01 4.26241308e-01
-7.40379453e-01 -1.29399467e+00 -7.46469676e-01 -4.10852998e-01
-9.91311371e-01 -1.01047051e+00 -6.18573844e-01 -4.29879278e-01
2.91028261e-01 4.28532153e-01 8.99711370e-01 -1.87724128e-01
1.62654489e-01 5.76326013e-01 -1.65490821e-01 -5.15215039e-01
1.51416108e-01 1.33230938e-02 1.21260263e-01 5.06087959e-01
5.67758203e-01 -6.35321975e-01 -9.29404557e-01 3.61068368e-01
-8.60321105e-01 -1.19033501e-01 1.73350900e-01 3.50499421e-01
5.83874524e-01 3.96449089e-01 4.72818345e-01 -5.38116217e-01
6.23479187e-01 -9.81478333e-01 -1.01593092e-01 2.32943058e-01
-2.68281639e-01 -2.59958029e-01 5.02994716e-01 -6.06008053e-01
-1.17411518e+00 1.11583341e-02 2.66830385e-01 -5.17530322e-01
-5.26250243e-01 4.57166940e-01 -3.82771082e-02 8.36049855e-01
1.85630977e-01 4.53346759e-01 -2.07509384e-01 -3.67792159e-01
-1.15904517e-01 2.36209646e-01 4.21119273e-01 -3.81204933e-01
5.66965878e-01 1.12850749e+00 -9.46491770e-03 -8.58256221e-01
-9.08432543e-01 -1.01805186e+00 -5.89748859e-01 -1.72383875e-01
7.70774305e-01 -1.14381051e+00 -8.32850754e-01 6.29254520e-01
-1.14748406e+00 -3.60740602e-01 -3.53630990e-01 5.35193861e-01
-6.91031992e-01 4.73010749e-01 -9.89406168e-01 -9.17563558e-01
-8.93958565e-03 -4.15279835e-01 1.16744673e+00 1.92901284e-01
-3.64367157e-01 -1.38857234e+00 3.07938248e-01 2.22921550e-01
-1.09296050e-02 3.41770440e-01 5.69050670e-01 -5.11844993e-01
-6.54266477e-01 -2.29449525e-01 -6.68732524e-02 -2.05060318e-01
4.43337977e-01 -1.21039934e-01 -4.11262453e-01 -8.14024452e-03
-9.84202474e-02 4.67738301e-01 8.33028376e-01 9.17762697e-01
1.34206796e+00 -6.28522098e-01 -7.85018265e-01 4.98905808e-01
9.27658379e-01 3.74599040e-01 7.12094188e-01 2.09013775e-01
7.71798134e-01 7.96419740e-01 5.62185764e-01 7.24018276e-01
7.01520085e-01 5.42487144e-01 1.11302003e-01 3.61924581e-02
3.68232757e-01 -6.63476527e-01 4.71052200e-01 3.11213970e-01
-3.65479559e-01 -6.04874969e-01 -9.38812852e-01 8.22263539e-01
-2.41785765e+00 -1.68297541e+00 -1.06647104e-01 2.07405257e+00
2.44664550e-01 -3.21280472e-02 6.00352883e-01 -4.27435338e-02
6.91807151e-01 3.83966506e-01 -4.49301004e-01 2.45194688e-01
1.50517141e-02 -3.29835087e-01 5.10350287e-01 1.28620967e-01
-1.42372739e+00 5.66765904e-01 6.49795151e+00 4.82020110e-01
-6.75804377e-01 1.93075567e-01 8.55242968e-01 -4.94088650e-01
8.16130787e-02 -2.24291310e-01 -8.45821559e-01 8.04928601e-01
1.08405030e+00 4.35281433e-02 1.92659527e-01 4.34147537e-01
1.06496537e+00 8.51248056e-02 -1.03470647e+00 7.93827891e-01
-1.01392165e-01 -1.32451785e+00 -2.10519438e-03 3.99985313e-01
6.64565623e-01 -6.63890988e-02 -4.96760719e-02 2.32549727e-01
3.43586445e-01 -8.91196609e-01 5.46878576e-01 1.02876103e+00
8.62464756e-02 -8.25106800e-01 4.42626804e-01 7.31580794e-01
-1.23028028e+00 -3.63779604e-01 3.22394446e-02 -3.91145945e-01
7.27238655e-01 5.76950312e-01 -6.04580462e-01 2.93700516e-01
8.95006061e-01 1.32222116e+00 -1.99701399e-01 1.05104065e+00
-1.12539947e-01 9.96541321e-01 -3.57912481e-01 1.58298790e-01
1.59686193e-01 -8.01158249e-02 4.36590463e-01 1.16877556e+00
6.43458307e-01 1.65293455e-01 4.63545054e-01 4.59645748e-01
-3.55040208e-02 -1.35964006e-01 -6.00398481e-01 5.18057644e-02
3.58885676e-01 8.21777940e-01 -8.23860407e-01 -4.16786045e-01
-6.47499323e-01 9.08937633e-01 1.99836195e-01 6.15580380e-01
-9.02018726e-01 2.26199254e-01 7.62577415e-01 5.14881372e-01
4.67984617e-01 -5.09745300e-01 -2.14660708e-02 -1.28512156e+00
2.36974001e-01 -4.51779127e-01 6.42851591e-01 -5.61106205e-01
-1.27749062e+00 4.06157710e-02 3.41499478e-01 -1.17122102e+00
-3.77373308e-01 -2.33703762e-01 -9.27398384e-01 5.59451342e-01
-1.29246092e+00 -1.05786562e+00 -1.10177122e-01 7.37329066e-01
6.62940443e-01 3.57774347e-02 2.56172001e-01 2.15944901e-01
-5.72389841e-01 -2.40549609e-01 1.43799663e-01 4.07283843e-01
3.17627996e-01 -1.00463092e+00 7.06805170e-01 9.68218505e-01
3.66654158e-01 3.81996959e-01 5.19057810e-01 -1.00257242e+00
-1.06145251e+00 -1.18190110e+00 1.18345761e+00 -6.18702888e-01
7.62224674e-01 -1.99844494e-01 -1.02313840e+00 8.97672772e-01
-1.82248294e-01 -7.60546997e-02 7.75270283e-01 3.41757923e-01
6.73127994e-02 2.53153332e-02 -7.78016925e-01 6.18583977e-01
1.17867005e+00 -2.96786487e-01 -4.83378917e-01 3.63885224e-01
6.51154399e-01 4.42728475e-02 -6.13067567e-01 1.34532183e-01
5.08666158e-01 -7.07285583e-01 1.08551776e+00 -7.27330327e-01
5.71784377e-01 -5.43721020e-02 3.92036475e-02 -1.11703551e+00
-5.52353561e-01 -6.35223508e-01 -9.02254760e-01 1.07186699e+00
1.00367084e-01 -2.67623335e-01 1.02888262e+00 7.66398132e-01
3.40445071e-01 -5.51778674e-01 -8.80142510e-01 -5.45801640e-01
-4.49968368e-01 -5.67229211e-01 8.25964630e-01 6.72902405e-01
-1.59339473e-01 6.64595440e-02 -9.96136963e-01 4.52108800e-01
6.08681500e-01 -1.49339633e-02 6.89468265e-01 -1.33014274e+00
-4.05839711e-01 -1.36797622e-01 -2.75317401e-01 -1.34151185e+00
4.94423173e-02 -2.27640718e-02 -3.59345190e-02 -1.68869281e+00
5.04625201e-01 -6.51286403e-03 -4.46001530e-01 2.99939543e-01
-4.01210189e-01 -2.49587983e-01 7.11565092e-02 6.58243299e-01
-9.18258369e-01 6.40475869e-01 8.49610567e-01 2.60992069e-02
-4.04497862e-01 5.95969141e-01 -3.70788515e-01 8.54131401e-01
8.00705850e-01 -5.68163455e-01 -3.93925786e-01 -3.91836375e-01
1.78529352e-01 4.20418829e-01 6.34908974e-01 -1.00122976e+00
4.11063552e-01 -6.31994605e-01 4.51229423e-01 -6.93931878e-01
4.37392175e-01 -9.18908954e-01 9.23880413e-02 3.64162922e-01
-5.11907995e-01 1.99203163e-01 -1.37525916e-01 1.40559137e+00
-2.38479823e-01 2.98118591e-01 2.01454815e-02 -1.03645310e-01
-7.17908263e-01 1.06004596e+00 -7.69881248e-01 -1.15258507e-01
1.02693260e+00 -3.00149173e-01 6.70911148e-02 -1.03223503e+00
-8.37922335e-01 5.58928609e-01 1.44120589e-01 5.97910345e-01
5.21179855e-01 -1.23249006e+00 -4.63016927e-01 -8.62899646e-02
-2.59752180e-02 -4.51716734e-03 5.52964687e-01 8.12077343e-01
-1.63292974e-01 3.28837484e-01 -2.00688257e-04 -5.67305207e-01
-1.20527077e+00 7.34399199e-01 1.28702596e-01 -6.29408956e-01
-7.20773697e-01 1.94220528e-01 3.55400085e-01 -1.29212320e-01
2.09158584e-01 -4.99107212e-01 -4.14848387e-01 1.45815769e-02
7.75914729e-01 7.58436799e-01 -5.06234646e-01 -6.72455251e-01
-1.14906386e-01 1.43460721e-01 1.80798352e-01 -1.47049159e-01
1.56870675e+00 -3.96364897e-01 1.82564855e-01 7.92518914e-01
7.82446265e-01 -3.42613250e-01 -1.94436884e+00 -5.62700212e-01
1.89362049e-01 -2.97588646e-01 -2.45692596e-01 -1.71302363e-01
-7.87876308e-01 8.95593464e-01 2.18380421e-01 3.64946008e-01
1.01019371e+00 1.32048965e-01 1.04826677e+00 2.12189406e-01
3.27233464e-01 -1.07843983e+00 8.08003247e-02 5.02140403e-01
3.05343688e-01 -1.05224502e+00 -7.53378719e-02 -2.68996775e-01
-6.70806527e-01 7.54745185e-01 2.21085280e-01 -1.51680216e-01
9.41999614e-01 -5.93227558e-02 -5.77463448e-01 -2.28035748e-01
-1.00418413e+00 -9.83060002e-02 2.66118973e-01 5.98841429e-01
1.13557123e-01 2.35800460e-01 1.43001646e-01 4.28153664e-01
3.32042098e-01 4.20406640e-01 3.56902719e-01 1.02535748e+00
-3.94883305e-01 -6.41557634e-01 -3.55752856e-01 5.31103611e-01
-8.20336878e-01 1.44727349e-01 -5.59665263e-02 4.08790618e-01
9.56330355e-03 9.64169264e-01 5.51142633e-01 2.29376759e-02
3.50800514e-01 9.86149982e-02 1.61435395e-01 -4.14569080e-01
-3.66649598e-01 -1.79373771e-02 -6.91153780e-02 -5.88089228e-01
-6.31751180e-01 -1.06304634e+00 -8.92868638e-01 -3.57410759e-01
4.12693173e-01 -9.58452839e-03 2.21659899e-01 1.24377239e+00
6.04165554e-01 3.93014550e-01 5.63463211e-01 -8.25904489e-01
-1.90965638e-01 -8.58434141e-01 -5.45770466e-01 6.28335297e-01
4.72570777e-01 -5.32993376e-01 1.88098836e-03 3.93375039e-01] | [7.083510398864746, 3.165496587753296] |
72ea918a-0ac2-4965-8019-cbdf700d1225 | brits-bidirectional-recurrent-imputation-for | 1805.10572 | null | http://arxiv.org/abs/1805.10572v1 | http://arxiv.org/pdf/1805.10572v1.pdf | BRITS: Bidirectional Recurrent Imputation for Time Series | Time series are widely used as signals in many classification/regression
tasks. It is ubiquitous that time series contains many missing values. Given
multiple correlated time series data, how to fill in missing values and to
predict their class labels? Existing imputation methods often impose strong
assumptions of the underlying data generating process, such as linear dynamics
in the state space. In this paper, we propose BRITS, a novel method based on
recurrent neural networks for missing value imputation in time series data. Our
proposed method directly learns the missing values in a bidirectional recurrent
dynamical system, without any specific assumption. The imputed values are
treated as variables of RNN graph and can be effectively updated during the
backpropagation.BRITS has three advantages: (a) it can handle multiple
correlated missing values in time series; (b) it generalizes to time series
with nonlinear dynamics underlying; (c) it provides a data-driven imputation
procedure and applies to general settings with missing data.We evaluate our
model on three real-world datasets, including an air quality dataset, a
health-care data, and a localization data for human activity. Experiments show
that our model outperforms the state-of-the-art methods in both imputation and
classification/regression accuracies. | ['Lei LI', 'Hao Zhou', 'Jian Li', 'Dong Wang', 'Yitan Li', 'Wei Cao'] | 2018-05-27 | brits-bidirectional-recurrent-imputation-for-1 | http://papers.nips.cc/paper/7911-brits-bidirectional-recurrent-imputation-for-time-series | http://papers.nips.cc/paper/7911-brits-bidirectional-recurrent-imputation-for-time-series.pdf | neurips-2018-12 | ['multivariate-time-series-imputation', 'traffic-data-imputation'] | ['time-series', 'time-series'] | [ 3.95398259e-01 -3.64812255e-01 -2.78497219e-01 -5.54195344e-01
-6.24212861e-01 -2.73456037e-01 1.81659952e-01 -1.71971709e-01
-1.14000835e-01 1.04225600e+00 4.41527247e-01 -3.85018289e-01
-5.26492178e-01 -7.84084141e-01 -1.07779491e+00 -8.18040073e-01
-1.22705527e-01 4.85012412e-01 -4.93458986e-01 -3.56266767e-01
-1.80798560e-01 2.69311935e-01 -1.20478904e+00 8.42984170e-02
9.41583872e-01 8.47127676e-01 -1.68922737e-01 3.16804886e-01
4.70896177e-02 1.38981819e+00 -4.93114889e-01 -6.41298816e-02
1.74680695e-01 -5.67133069e-01 -3.55194390e-01 -4.84555811e-01
-2.17914417e-01 -1.40380159e-01 -5.37289619e-01 5.82462072e-01
3.89842838e-01 1.78214133e-01 6.12389445e-01 -1.60250616e+00
-7.62036443e-01 7.43370891e-01 -5.93488932e-01 9.93592441e-02
-8.06008950e-02 7.69881578e-03 3.31166923e-01 -6.20189726e-01
3.75341922e-01 8.04051578e-01 1.23397028e+00 5.27186692e-01
-1.45083892e+00 -9.41912591e-01 1.19922377e-01 1.16506487e-01
-1.19016278e+00 -3.19403648e-01 1.12754595e+00 -6.00664079e-01
9.10748005e-01 2.37900972e-01 5.62208772e-01 1.59842694e+00
4.19920951e-01 4.15034831e-01 1.14013970e+00 1.73573270e-01
2.72284061e-01 -4.84453022e-01 2.58966953e-01 7.32491910e-02
6.82695135e-02 3.27796906e-01 -4.40405577e-01 -3.60792518e-01
8.05754483e-01 8.03958356e-01 -8.03238600e-02 -5.56347147e-03
-1.69323874e+00 5.58514833e-01 2.41484359e-01 -3.33405733e-02
-7.24760592e-01 3.85850132e-01 3.21725011e-01 7.63125896e-01
8.61850858e-01 -8.20398852e-02 -5.97623825e-01 -2.66190708e-01
-7.64376521e-01 3.56459208e-02 8.86161327e-01 7.50367880e-01
4.13920939e-01 4.38931972e-01 -1.81510806e-01 8.15947294e-01
-8.51085037e-02 9.28564131e-01 5.96791506e-01 -8.19786668e-01
8.80952775e-01 4.47259754e-01 3.44403654e-01 -9.66746032e-01
-5.83774805e-01 -5.45395076e-01 -1.83075023e+00 -1.68635324e-01
4.32273597e-01 -6.44139349e-01 -9.06452775e-01 2.23364401e+00
1.55762866e-01 8.95590842e-01 2.25227416e-01 8.90007794e-01
5.02695739e-01 8.67465377e-01 -2.73229927e-01 -5.49629152e-01
6.23531699e-01 -4.77836281e-01 -1.08016086e+00 3.86475548e-02
2.52699167e-01 -3.49618495e-01 5.56011975e-01 2.77387559e-01
-7.78219759e-01 -4.60422993e-01 -5.71880162e-01 1.35235682e-01
-2.97038704e-01 -2.45063901e-02 4.62750703e-01 2.01359212e-01
-6.45536602e-01 7.01742172e-01 -1.22909606e+00 1.58744752e-01
-4.68653552e-02 5.95046222e-01 -3.68538409e-01 2.93173909e-01
-1.33803880e+00 7.83481836e-01 -4.30293307e-02 6.80574358e-01
-7.08441198e-01 -8.34729671e-01 -7.82082617e-01 -1.64473906e-01
1.65691048e-01 -9.63011503e-01 9.93776739e-01 -9.03803766e-01
-1.27170086e+00 2.18637567e-03 -4.68806893e-01 -5.12761176e-01
5.40570915e-01 -9.06402469e-02 -8.32956910e-01 -6.21064603e-01
-2.49072313e-02 -1.00573137e-01 9.25579846e-01 -7.30032504e-01
-8.04344043e-02 -5.28847039e-01 -5.66718638e-01 -1.20538101e-01
-6.19297214e-02 -4.62151706e-01 3.29134732e-01 -1.04881179e+00
4.68427122e-01 -1.03242743e+00 -2.77464271e-01 -3.99934769e-01
-5.53925276e-01 9.38203260e-02 7.30609596e-01 -9.48969662e-01
1.13828337e+00 -1.95593131e+00 3.51307303e-01 2.75124699e-01
5.07233292e-03 -3.69638145e-01 -9.89301205e-02 7.37509668e-01
-4.17292684e-01 -2.02346772e-01 -6.93970919e-01 -3.44639450e-01
-9.88626406e-02 4.00401056e-01 -8.98009539e-01 6.10053182e-01
4.64488417e-02 9.34922576e-01 -6.89035356e-01 1.02364741e-01
1.82715654e-01 8.74942064e-01 6.63982779e-02 2.11804003e-01
4.21278253e-02 1.03454518e+00 -2.27533221e-01 4.69137788e-01
5.86089611e-01 -3.27941328e-01 -5.96534796e-02 -6.56388029e-02
-1.34540141e-01 1.67178243e-01 -1.20327473e+00 1.47597909e+00
-5.57727039e-01 3.70693058e-01 -3.37187856e-01 -1.09566855e+00
1.06414580e+00 4.51467842e-01 7.89523005e-01 -6.73226118e-01
-6.54747263e-02 7.94513300e-02 -1.50864139e-01 -5.81239581e-01
2.45683223e-01 -1.01426065e-01 -2.03729346e-01 5.89698255e-01
-2.67033249e-01 4.65516359e-01 -2.61137545e-01 -2.64298469e-01
1.21932828e+00 2.89625406e-01 -1.52195379e-01 1.43742606e-01
2.06080437e-01 -2.16681138e-01 1.19203341e+00 8.34889710e-01
3.20377290e-01 6.94795310e-01 4.82774734e-01 -6.41318560e-01
-1.06177247e+00 -1.04944742e+00 -7.20086843e-02 6.97482407e-01
-4.07621115e-02 2.11203575e-01 -2.48847827e-01 -1.89211428e-01
1.08827889e-01 6.07549608e-01 -7.71798551e-01 -2.20489219e-01
-7.17720926e-01 -1.24027371e+00 5.12550831e-01 7.67154336e-01
2.89981961e-01 -1.21399760e+00 -1.96606174e-01 5.67682803e-01
-7.09541738e-01 -8.13825011e-01 -5.31804740e-01 3.56640786e-01
-1.27527225e+00 -8.20784748e-01 -5.94971597e-01 -5.39815485e-01
8.15017641e-01 5.51932454e-02 7.86846161e-01 -2.50064403e-01
7.79390559e-02 1.12456702e-01 -4.00729328e-02 -3.83216083e-01
-2.63759553e-01 1.00665435e-01 2.44341090e-01 3.45711648e-01
2.55226165e-01 -9.30812061e-01 -5.44596791e-01 3.78078967e-01
-7.55148590e-01 2.41856232e-01 3.37827981e-01 1.23330057e+00
7.37765908e-01 1.43883296e-03 1.13793087e+00 -8.89411747e-01
6.00377440e-01 -1.00754774e+00 -5.86274683e-01 2.10729852e-01
-5.65530717e-01 9.54046920e-02 8.62496912e-01 -5.92937648e-01
-9.51389194e-01 2.79678732e-01 -6.39954209e-02 -5.08453429e-01
3.19476798e-02 8.19504321e-01 2.60521859e-01 4.53091830e-01
3.36002946e-01 4.35234755e-01 2.02734724e-01 -6.65278554e-01
-5.21489978e-02 5.93127549e-01 7.43126452e-01 -3.24915558e-01
7.80374646e-01 4.66910303e-01 1.23654053e-01 -6.47838116e-01
-6.68780088e-01 -3.41827750e-01 -6.06378078e-01 -1.92973331e-01
4.88420427e-01 -1.21456659e+00 -9.25968349e-01 8.19291472e-01
-1.11195552e+00 -4.38348174e-01 -3.12408447e-01 7.96904802e-01
-5.79561055e-01 -7.19690770e-02 -8.09279978e-01 -1.03919399e+00
-4.05049860e-01 -6.22422934e-01 7.52848446e-01 -4.06368822e-02
-1.04395129e-01 -1.18929255e+00 3.44993442e-01 6.87464178e-02
5.57443023e-01 8.19357932e-01 8.99180114e-01 -2.67826229e-01
-2.97681808e-01 -2.05259249e-01 1.39597818e-01 2.28624791e-01
3.62701863e-01 -2.41890386e-01 -8.31764877e-01 -3.06427002e-01
2.64637411e-01 1.14776604e-01 8.43980789e-01 5.35143793e-01
1.51218915e+00 -6.69261336e-01 -1.50008544e-01 7.70529091e-01
1.14143360e+00 1.37559995e-01 6.04302227e-01 -8.72279629e-02
7.72331715e-01 5.69385588e-01 1.74679250e-01 4.36391711e-01
6.92178309e-01 4.14147615e-01 6.72550321e-01 -2.48544887e-01
4.12719488e-01 -4.99615997e-01 6.32617652e-01 1.61745214e+00
-1.83264852e-01 -1.50824949e-01 -9.19013798e-01 7.11516380e-01
-2.50708938e+00 -1.38809383e+00 -5.64472258e-01 2.42832017e+00
8.09648991e-01 -2.44509086e-01 2.51707733e-01 1.45492002e-01
9.26094949e-01 -5.01154959e-02 -1.23647940e+00 1.28703029e-03
-3.21846843e-01 -4.56828363e-02 6.00546002e-01 1.95651561e-01
-9.47760522e-01 1.25443801e-01 6.14592218e+00 1.07864335e-01
-1.09295094e+00 2.21974671e-01 5.75246453e-01 -1.63472742e-01
-2.42002293e-01 -2.96152830e-01 -4.06060278e-01 8.45631182e-01
1.31689119e+00 -9.15875733e-02 7.03712583e-01 2.51895040e-01
5.75380683e-01 5.16565204e-01 -9.50277627e-01 9.36573625e-01
-1.09823592e-01 -1.19289005e+00 -2.99801230e-01 -4.17895839e-02
8.03385973e-01 4.08910692e-01 1.99105829e-01 3.31372350e-01
4.87075210e-01 -1.13146269e+00 3.07462275e-01 1.14317763e+00
6.15489364e-01 -7.68043041e-01 8.29397619e-01 7.71959782e-01
-8.88385236e-01 -2.50968128e-01 -2.87849873e-01 -3.92597497e-01
2.71109849e-01 8.84327114e-01 -2.61024117e-01 6.00271821e-01
6.57386959e-01 1.23614645e+00 -2.21592113e-01 9.48012114e-01
1.73417758e-03 1.03957248e+00 -3.97697240e-01 2.46764287e-01
-2.39504889e-01 -6.35855734e-01 3.99471343e-01 7.78324962e-01
7.46858299e-01 -1.03144586e-01 -5.99549785e-02 8.74815285e-01
-1.75475851e-02 -3.35426599e-01 -8.26551378e-01 2.29201987e-01
5.59238911e-01 9.25717235e-01 -3.39501023e-01 1.64186023e-02
-3.91686916e-01 7.35929728e-01 6.34092838e-02 7.08715856e-01
-1.11101854e+00 -2.94013947e-01 5.14327049e-01 -2.20940318e-02
2.76194047e-02 -2.93478906e-01 -3.60896379e-01 -1.48125935e+00
2.45807186e-01 -6.98134005e-01 4.32622433e-01 -9.75819111e-01
-1.88832760e+00 6.42554641e-01 -1.48601383e-01 -1.69909096e+00
-6.32281959e-01 -6.72183037e-02 -5.46368361e-01 9.48157609e-01
-1.42147136e+00 -1.15981376e+00 -3.64255965e-01 9.89235580e-01
5.69655076e-02 -1.32410422e-01 8.62975955e-01 4.76253927e-01
-7.84965098e-01 7.42638931e-02 6.91839159e-01 3.38144332e-01
5.69686890e-01 -1.08962178e+00 5.92110991e-01 8.39002252e-01
2.52592433e-02 6.88935697e-01 6.72132313e-01 -8.42709601e-01
-1.83500254e+00 -1.55405295e+00 9.16618824e-01 -5.03770947e-01
7.37448514e-01 -4.68130529e-01 -1.19492137e+00 9.77094412e-01
-4.01660837e-02 1.05259448e-01 7.33307362e-01 2.78608024e-01
-2.71916837e-01 -4.56285805e-01 -1.01188719e+00 2.30911523e-01
9.92810011e-01 -5.68524182e-01 -7.18846262e-01 3.81111920e-01
6.24269307e-01 -6.79786205e-01 -1.01333463e+00 4.81597394e-01
5.64155877e-01 -4.65508133e-01 9.26173389e-01 -6.55281126e-01
4.57484990e-01 -4.49686795e-01 -6.89807069e-03 -1.58079743e+00
-2.60107934e-01 -7.29964495e-01 -5.78845739e-01 1.16542721e+00
5.33766150e-01 -1.03098583e+00 3.72578472e-01 7.99178779e-01
3.90736833e-02 -4.32422757e-01 -1.23906338e+00 -9.12896395e-01
3.91942412e-02 -4.98656422e-01 1.14773250e+00 1.30184734e+00
-2.94659138e-01 1.78166583e-01 -1.18728411e+00 2.39569634e-01
8.10703874e-01 3.04843247e-01 6.18275166e-01 -1.46859050e+00
-2.67594755e-01 2.49715433e-01 -7.82917589e-02 -6.41244113e-01
3.15172583e-01 -9.44242597e-01 -1.55018300e-01 -1.33345985e+00
-2.14321576e-02 -4.99484330e-01 -8.45109940e-01 8.70214581e-01
3.01455796e-01 1.98520184e-01 -9.43456069e-02 5.05615354e-01
-1.32396936e-01 7.44080842e-01 8.00797284e-01 -2.17562199e-01
-3.99398118e-01 4.45990056e-01 -4.54652578e-01 5.15449107e-01
9.56692934e-01 -8.71796370e-01 -5.12709856e-01 -6.22600079e-01
5.40906489e-01 9.80450332e-01 8.40811193e-01 -7.43917882e-01
4.35227066e-01 -3.81189257e-01 6.24666333e-01 -8.66725564e-01
3.61034155e-01 -1.18264401e+00 1.11887896e+00 4.70167071e-01
-3.39751005e-01 6.38841867e-01 -1.51321078e-02 9.03354585e-01
-2.18277037e-01 3.13628227e-01 1.95397094e-01 3.23220432e-01
-2.32558981e-01 4.61503744e-01 -2.22030312e-01 -3.23167294e-02
7.08992720e-01 -6.90811872e-02 -5.77122450e-01 -6.61010146e-01
-9.47973907e-01 4.13899720e-01 8.20042789e-02 7.31191039e-01
6.19426787e-01 -1.68669581e+00 -7.92808056e-01 4.70372409e-01
-8.94927830e-02 -1.84844196e-01 4.31937277e-01 1.19638515e+00
1.41440183e-01 -5.74212335e-02 -1.76295415e-01 -7.44666636e-01
-7.15212941e-01 4.36864436e-01 3.77794951e-01 3.73014547e-02
-9.50676560e-01 -1.21843986e-01 -3.03302050e-01 -8.81690264e-01
1.64755732e-01 -4.62955624e-01 -4.77594696e-02 5.25408275e-02
3.49259198e-01 5.97882867e-01 8.42490867e-02 -5.94555855e-01
-2.01565176e-01 4.84491289e-01 3.00151139e-01 7.45145082e-02
1.83252323e+00 -1.78184688e-01 -3.99323612e-01 1.17675507e+00
1.06758106e+00 -4.80397522e-01 -1.32866538e+00 -2.35890538e-01
-1.06909096e-01 -1.23978809e-01 -1.63798094e-01 -1.04986989e+00
-1.18075228e+00 6.77685261e-01 5.30728579e-01 3.42753321e-01
1.27670026e+00 -6.45839155e-01 9.34684992e-01 3.76596689e-01
4.55325782e-01 -8.36674631e-01 -5.60626507e-01 6.81266010e-01
9.41607118e-01 -1.12257266e+00 -4.19406623e-01 6.48684949e-02
-5.58698416e-01 9.56257224e-01 3.28284830e-01 -1.43430978e-01
7.58776069e-01 4.73118633e-01 8.86256695e-02 1.40841588e-01
-1.02535212e+00 2.80251324e-01 9.51859504e-02 5.95978677e-01
4.26834673e-01 2.47630961e-02 -6.19328860e-03 8.21415961e-01
-2.88049579e-01 4.80195582e-01 7.48340666e-01 5.48255742e-01
2.82590300e-01 -9.08861399e-01 -4.44283634e-01 7.83345819e-01
-3.92620146e-01 1.03901379e-01 -7.91602805e-02 5.60673833e-01
-2.37318665e-01 1.02070832e+00 8.75655189e-02 -2.21093893e-01
4.30387914e-01 1.51706338e-01 -5.65370545e-03 3.17228325e-02
-6.89717054e-01 -1.22652993e-01 -2.56535262e-01 -3.64115000e-01
-5.06500542e-01 -8.32230568e-01 -1.12789679e+00 -5.41781604e-01
-1.32011116e-01 1.69689775e-01 6.75930858e-01 8.73071611e-01
7.48423100e-01 7.59823859e-01 7.55865812e-01 -5.99795759e-01
-6.22029305e-01 -8.05461109e-01 -4.88927871e-01 5.50906420e-01
9.92393196e-01 -4.55824822e-01 -4.82023954e-01 1.06921703e-01] | [7.041081428527832, 3.1626834869384766] |
ed274fc4-9a7a-4fce-9365-52390abe92c7 | a-tip-attribute-aware-text-infilling-via-pre | null | null | https://aclanthology.org/2022.coling-1.511 | https://aclanthology.org/2022.coling-1.511.pdf | A-TIP: Attribute-aware Text Infilling via Pre-trained Language Model | Text infilling aims to restore incomplete texts by filling in blanks, which has attracted more attention recently because of its wide application in ancient text restoration and text rewriting. However, attribute- aware text infilling is yet to be explored, and existing methods seldom focus on the infilling length of each blank or the number/location of blanks. In this paper, we propose an Attribute-aware Text Infilling method via a Pre-trained language model (A-TIP), which contains a text infilling component and a plug- and-play discriminator. Specifically, we first design a unified text infilling component with modified attention mechanisms and intra- and inter-blank positional encoding to better perceive the number of blanks and the infilling length for each blank. Then, we propose a plug-and-play discriminator to guide generation towards the direction of improving attribute relevance without decreasing text fluency. Finally, automatic and human evaluations on three open-source datasets indicate that A-TIP achieves state-of- the-art performance compared with all baselines. | ['Manabu Okumura', 'Kotaro Funakoshi', 'Jingyi You', 'Dongyuan Li'] | null | null | null | null | coling-2022-10 | ['ancient-tex-restoration', 'text-infilling'] | ['miscellaneous', 'natural-language-processing'] | [ 5.28209269e-01 -4.51182164e-02 -5.96885122e-02 -2.05854446e-01
-5.90165615e-01 -4.74169791e-01 7.52889931e-01 4.12824631e-01
-2.71843553e-01 5.94635487e-01 7.99268425e-01 -5.78082979e-01
1.51105091e-01 -7.17805743e-01 -6.24058485e-01 -3.29872578e-01
7.38640785e-01 3.84993315e-01 1.59641847e-01 -4.61830586e-01
6.43162549e-01 -2.73813203e-04 -1.35901868e+00 4.15508509e-01
1.51591432e+00 3.56025457e-01 6.30025327e-01 5.29939175e-01
-5.23623645e-01 4.61434871e-01 -8.65885317e-01 -2.99165845e-01
-1.37460217e-01 -5.62924683e-01 -6.68015897e-01 2.97865743e-04
4.91626084e-01 -3.77772897e-01 -4.44032669e-01 7.42809832e-01
8.28011572e-01 4.32625003e-02 7.12906539e-01 -6.03040993e-01
-1.32052088e+00 1.03290582e+00 -5.60384333e-01 1.80061489e-01
4.50759113e-01 1.41994283e-01 1.08202386e+00 -8.87692273e-01
4.20361906e-01 1.31757355e+00 4.20001090e-01 7.12384105e-01
-1.13929212e+00 -3.30933928e-01 3.76900584e-01 9.36443955e-02
-1.02499962e+00 -4.88908559e-01 7.14029551e-01 -3.02639544e-01
1.03104472e+00 3.61871749e-01 5.53453326e-01 8.06207716e-01
1.29810780e-01 1.12658691e+00 6.49392605e-01 -8.56045246e-01
6.61107758e-03 -1.65442497e-01 -2.03762949e-01 5.84227920e-01
1.70530841e-01 -3.34818721e-01 -5.70785522e-01 2.84258157e-01
7.18229473e-01 -2.72840083e-01 -3.82494479e-01 -5.09890653e-02
-1.15703917e+00 5.67111731e-01 2.64714301e-01 2.35370263e-01
-1.52072966e-01 -4.66248393e-02 5.20843208e-01 -1.30464047e-01
5.85702896e-01 6.39668047e-01 -2.49351278e-01 -1.69784442e-01
-1.07516050e+00 3.18535417e-01 3.42531085e-01 9.17470515e-01
5.08939922e-01 -1.48214772e-02 -8.60880435e-01 1.10843706e+00
1.21730119e-01 4.63077009e-01 7.37675250e-01 -5.59006095e-01
8.39602292e-01 1.00006557e+00 3.89845646e-03 -6.67725861e-01
-2.22725838e-01 -5.27659237e-01 -8.14755619e-01 -1.74078971e-01
1.58309624e-01 4.02238145e-02 -7.89733112e-01 1.70890522e+00
1.07316859e-01 -2.63696194e-01 -1.67571053e-01 6.82843864e-01
8.00206780e-01 7.07321942e-01 -7.95059204e-02 5.78347221e-02
1.46345437e+00 -1.29654169e+00 -8.59147489e-01 -6.81649685e-01
8.05482030e-01 -9.88120079e-01 1.85944033e+00 7.09012225e-02
-1.20625365e+00 -6.20294988e-01 -1.23315370e+00 -7.89874673e-01
-2.99613595e-01 6.84641719e-01 3.53759021e-01 6.37691915e-01
-8.80663097e-01 5.64735770e-01 -4.49677408e-01 -3.00361663e-01
2.14243367e-01 -9.18246955e-02 1.12998290e-02 2.04920657e-02
-1.08937550e+00 8.32199812e-01 2.40778551e-01 1.65305898e-01
-3.80764872e-01 -7.58376241e-01 -8.86457384e-01 2.52019107e-01
2.67859250e-01 -9.58924949e-01 1.19977236e+00 -4.64767396e-01
-1.58754122e+00 6.76055074e-01 -4.66787308e-01 -1.96759880e-01
5.96137047e-01 -5.45034170e-01 -2.01761886e-01 -2.62482941e-01
3.68064821e-01 7.05765545e-01 8.67197931e-01 -1.11192513e+00
-5.41271389e-01 -3.30438644e-01 -8.97637233e-02 5.38034320e-01
-6.39730632e-01 -3.21393877e-01 -6.48313344e-01 -1.26099002e+00
1.77229822e-01 -3.98703396e-01 6.47554025e-02 -1.09469958e-01
-7.08345056e-01 -3.21393222e-01 7.62856543e-01 -9.33527589e-01
2.02441907e+00 -1.90746760e+00 1.49057195e-01 -3.83611530e-01
1.55837923e-01 4.59912211e-01 -3.38478476e-01 5.91315031e-01
2.07091138e-01 3.49720508e-01 -3.44623297e-01 -8.00579548e-01
9.39732641e-02 6.88912645e-02 -5.15837669e-01 6.42273249e-03
1.92648068e-01 1.02922893e+00 -1.04392374e+00 -4.00628507e-01
3.34823042e-01 4.48088169e-01 -6.77674949e-01 6.66738302e-02
-4.84737664e-01 2.89482832e-01 -2.86066532e-01 5.53858638e-01
5.39748788e-01 -2.34212074e-02 -4.71778251e-02 -3.80417034e-02
-3.37342083e-01 7.78543591e-01 -7.89974570e-01 1.82081997e+00
-4.41190928e-01 7.63333619e-01 -3.66731435e-01 -3.80531430e-01
9.80827153e-01 -2.75442842e-02 -1.24308527e-01 -1.05736184e+00
1.14440739e-01 2.50689745e-01 -1.79413006e-01 -3.43620837e-01
1.37587154e+00 2.69163489e-01 -7.66717419e-02 6.20081425e-01
-5.80763936e-01 -3.43592227e-01 6.84144795e-01 4.24907297e-01
9.67823565e-01 3.59840333e-01 8.99839699e-02 -1.69579580e-01
6.94831908e-01 -3.31926078e-01 4.24514711e-01 7.57804394e-01
1.86831772e-01 8.78429472e-01 6.65220618e-01 -2.37569585e-01
-1.20922101e+00 -6.61586940e-01 1.96179673e-01 1.25286782e+00
1.69441864e-01 -8.29549134e-01 -1.07891214e+00 -5.04697919e-01
-1.86897680e-01 1.07673907e+00 -6.64636791e-01 -3.57570499e-01
-8.53142321e-01 -5.04956007e-01 5.74075699e-01 6.08790636e-01
4.91111040e-01 -1.30100858e+00 -5.14110386e-01 3.50950360e-01
-6.42892718e-01 -8.06706488e-01 -1.12913418e+00 -1.19921587e-01
-8.28336120e-01 -7.61548221e-01 -8.50721359e-01 -8.64474118e-01
8.51828098e-01 3.03556621e-01 1.10729969e+00 3.26371372e-01
-1.07727706e-01 -1.63154006e-01 -5.93201935e-01 -3.08846146e-01
-3.88659656e-01 5.78069508e-01 -3.50229234e-01 -3.01847726e-01
9.90451736e-05 -3.59576702e-01 -7.02281892e-01 3.23574603e-01
-9.95449066e-01 5.93198299e-01 6.74509466e-01 7.67096281e-01
5.06027818e-01 -3.05984348e-01 4.68907535e-01 -9.02836204e-01
1.17716968e+00 4.41024415e-02 -2.83004463e-01 4.74646151e-01
-6.74772263e-01 2.76277542e-01 8.13985169e-01 -3.16684246e-01
-1.05472946e+00 -3.18200082e-01 -2.89603114e-01 9.74435657e-02
2.29843825e-01 4.66439396e-01 -6.35380089e-01 6.49892569e-01
5.11108637e-01 5.08826911e-01 -4.84827042e-01 -5.93285084e-01
6.77894413e-01 8.28906655e-01 7.08416760e-01 -7.69881427e-01
4.15700465e-01 1.41699493e-01 -3.52586925e-01 -4.88903761e-01
-8.17269146e-01 -7.13187233e-02 -5.25331140e-01 -2.66818032e-02
4.95430589e-01 -7.10919023e-01 -4.06719923e-01 6.21566176e-01
-1.29458892e+00 -5.25929034e-01 -4.78082776e-01 -1.45689532e-01
-2.96562850e-01 5.42471111e-01 -5.39605618e-01 -6.08830631e-01
-7.64039695e-01 -1.18982065e+00 1.29965389e+00 3.82500082e-01
-4.09626514e-01 -6.61024868e-01 -8.04326236e-02 4.97718126e-01
4.94926631e-01 -2.80968845e-01 1.45167732e+00 -2.59056091e-01
-3.40933055e-01 -4.71384302e-02 -4.22432065e-01 1.12574920e-01
2.15935424e-01 -1.69032127e-01 -7.25036740e-01 -1.06648982e-01
-5.41394949e-01 1.45087048e-01 1.07727849e+00 3.52547556e-01
1.19509566e+00 -3.44985813e-01 -1.83138713e-01 6.59612775e-01
7.29550958e-01 1.40110761e-01 1.07517791e+00 5.82997501e-01
7.99816132e-01 3.02875042e-01 3.80997986e-01 6.34335399e-01
6.85409069e-01 5.54183125e-01 3.91273171e-01 -1.57862499e-01
-6.12416267e-01 -7.79934049e-01 2.30843171e-01 6.76707804e-01
3.54082555e-01 -6.97481096e-01 -6.21518433e-01 7.22384930e-01
-1.67272949e+00 -6.20840073e-01 -1.78291947e-01 2.20531011e+00
1.09313273e+00 2.80480951e-01 -1.72109276e-01 4.21103150e-01
9.57947373e-01 3.17916423e-01 -6.96774602e-01 -4.57967252e-01
-4.53263402e-01 -4.01442777e-03 1.08431295e-01 6.40463769e-01
-6.37984097e-01 1.44371712e+00 5.45364714e+00 1.02603734e+00
-9.89452899e-01 -1.97685435e-01 6.71831787e-01 1.55029982e-01
-7.92891562e-01 2.13091597e-01 -9.69155669e-01 5.90135932e-01
1.72214806e-01 -2.08789874e-02 4.97785121e-01 4.25138831e-01
3.65434200e-01 -1.52151868e-01 -1.04044247e+00 6.96900964e-01
2.46277764e-01 -1.22352958e+00 5.04530430e-01 -1.67066440e-01
7.19098449e-01 -5.41385531e-01 1.81907248e-02 3.37629318e-01
-1.32892877e-01 -7.82952011e-01 1.14879620e+00 3.88367295e-01
9.92772281e-01 -4.89109993e-01 3.08087438e-01 1.80679217e-01
-1.22034490e+00 7.45147914e-02 -2.33903214e-01 -1.89576477e-01
2.69355655e-01 6.55040205e-01 -7.71435201e-01 3.01412016e-01
2.40149707e-01 6.67655349e-01 -8.24016809e-01 1.01704884e+00
-9.22061801e-01 5.12392759e-01 5.20381369e-02 -2.34575808e-01
-1.41283602e-01 -1.05892695e-01 5.89562058e-01 1.19216406e+00
4.92375493e-01 -1.40074775e-01 -5.64491451e-02 9.88172650e-01
-3.55550796e-01 2.98011243e-01 1.73826918e-01 -3.62903178e-01
9.15993214e-01 9.91205275e-01 -6.42122090e-01 -2.47351170e-01
-9.81111731e-03 1.15341115e+00 3.08994949e-01 3.46816510e-01
-5.72515070e-01 -8.39550257e-01 3.77126724e-01 4.04699564e-01
3.40102166e-01 -9.25499052e-02 -8.12961340e-01 -1.06585515e+00
3.01644266e-01 -7.95790553e-01 1.41827866e-01 -1.15078044e+00
-8.97282779e-01 6.47342622e-01 -3.53833377e-01 -9.24879193e-01
5.66298589e-02 -2.33045220e-01 -7.44073510e-01 1.11007369e+00
-1.43936169e+00 -1.27610826e+00 -2.36026570e-01 1.17795140e-01
9.69567537e-01 -5.95773831e-02 5.24472356e-01 2.41466895e-01
-7.24741220e-01 8.12715352e-01 3.94048057e-02 1.15180224e-01
8.36386442e-01 -1.21415961e+00 9.95784938e-01 1.24384964e+00
-1.34235352e-01 8.03531945e-01 8.18484008e-01 -1.01726496e+00
-1.14178073e+00 -9.75473404e-01 1.40817261e+00 -1.54834166e-01
2.89565712e-01 -4.63132173e-01 -1.05284512e+00 4.34371173e-01
3.12410444e-01 -5.02792954e-01 1.93572462e-01 1.62218779e-01
-3.67197990e-01 -9.20321867e-02 -8.01250279e-01 1.09788489e+00
1.06663275e+00 -3.44774246e-01 -6.59263730e-01 8.94195214e-02
8.18171859e-01 -5.71602523e-01 -1.05181977e-01 1.96152791e-01
4.65153694e-01 -8.62056971e-01 7.48541117e-01 -8.20691660e-02
8.14792871e-01 -4.46356654e-01 1.26696572e-01 -1.47427225e+00
-4.55484778e-01 -8.43800724e-01 -5.79952076e-02 1.58336806e+00
3.34488273e-01 -4.39442754e-01 6.76999092e-01 2.86570519e-01
-7.64278829e-01 -7.95408010e-01 -6.75836265e-01 -2.82426476e-01
1.02220081e-01 -2.14608908e-01 9.29807365e-01 5.04586160e-01
1.16782434e-01 6.38887882e-01 -4.37911361e-01 -3.07954311e-01
1.05476648e-01 -1.27558978e-02 6.27837360e-01 -9.01723325e-01
-1.63487285e-01 -7.96450853e-01 4.03979778e-01 -1.69561267e+00
-5.70853688e-02 -7.93140352e-01 2.50897408e-01 -2.13694859e+00
2.37475157e-01 -5.29523671e-01 3.91652316e-01 7.62383163e-01
-7.50299513e-01 -9.77264568e-02 2.91153550e-01 1.51681915e-01
-3.51444870e-01 9.61596489e-01 1.48966300e+00 -2.90016800e-01
-4.74303424e-01 -2.67149806e-01 -1.08845067e+00 4.66470867e-01
7.52178371e-01 -2.88450807e-01 -3.72224003e-01 -1.00164199e+00
5.04858494e-01 -8.66937116e-02 6.41179159e-02 -6.61051035e-01
2.53941387e-01 -7.37225413e-02 3.53998095e-01 -1.02072418e+00
1.33725598e-01 -9.18477103e-02 -4.37418818e-01 2.34521568e-01
-6.44116640e-01 3.76118422e-01 2.43478417e-01 2.43230566e-01
2.48786435e-01 -4.36556488e-01 4.38732058e-01 -1.51069285e-02
-3.15034747e-01 -8.50878507e-02 -4.72750068e-01 3.43636096e-01
3.90575707e-01 -2.47057006e-01 -6.51011527e-01 -1.76357538e-01
-1.18746974e-01 3.00089508e-01 4.95713770e-01 7.71274507e-01
6.67598724e-01 -1.16634095e+00 -8.76883149e-01 4.10257608e-01
1.14665791e-01 8.10165331e-02 2.80637860e-01 2.70005465e-01
-5.05353689e-01 3.65385473e-01 1.51551887e-01 -2.75773346e-01
-1.16012347e+00 3.11267972e-01 8.10176954e-02 -5.31223953e-01
-6.42059743e-01 8.32658648e-01 1.22436464e-01 -4.30229723e-01
3.44294697e-01 -3.73045355e-01 -1.62581533e-01 -5.28895669e-02
7.30309069e-01 4.66836393e-01 2.08745658e-01 -3.39617461e-01
-2.13064328e-02 5.40378630e-01 -3.31285775e-01 -1.88848734e-01
9.75599289e-01 -5.87604821e-01 -2.41434321e-01 1.54201388e-01
5.61984062e-01 2.47981444e-01 -1.18010271e+00 -2.60499746e-01
-7.82585740e-02 -5.86140931e-01 2.79037971e-02 -1.17837989e+00
-7.26605535e-01 1.06494749e+00 1.95797428e-01 -1.33162692e-01
1.18396616e+00 -2.36567706e-01 1.10949612e+00 1.64756745e-01
-2.13737026e-01 -1.25295055e+00 2.77000636e-01 8.93544972e-01
1.09073997e+00 -8.51378560e-01 6.47101551e-02 -3.58749509e-01
-6.28680646e-01 9.41841304e-01 7.73918808e-01 4.36381072e-01
-1.35050952e-01 1.99836195e-01 -1.00746728e-01 2.20077664e-01
-6.22121215e-01 -7.28301704e-02 4.31961298e-01 4.47659373e-01
8.85740221e-01 -1.62003804e-02 -7.60395706e-01 7.29194224e-01
-5.13849378e-01 -9.00365487e-02 4.91460532e-01 7.90428698e-01
-6.87240839e-01 -1.50676441e+00 -2.63855934e-01 4.08701569e-01
-1.86671764e-01 -6.64472163e-01 -3.74159068e-01 3.30312550e-01
2.53491074e-01 1.02789462e+00 5.41489199e-03 -2.97604501e-01
4.16035593e-01 -1.01069845e-01 5.14631748e-01 -6.83749735e-01
-7.68749833e-01 3.08592319e-01 -6.00084178e-02 3.81337814e-02
3.27438712e-01 -6.23467147e-01 -1.35185778e+00 -2.97491729e-01
-4.78457600e-01 1.13714710e-02 6.15474343e-01 9.42767859e-01
5.43599963e-01 9.77302253e-01 2.88607508e-01 -7.14662015e-01
-3.54877144e-01 -1.25134206e+00 -3.25135410e-01 1.99592397e-01
9.11308452e-02 -2.95704901e-01 -9.17804986e-02 -1.52255436e-02] | [11.8569974899292, 9.047163963317871] |
086097f6-fdee-4a62-9f84-ff6809fd4bc1 | graph-collaborative-signals-denoising-and | 2304.03344 | null | https://arxiv.org/abs/2304.03344v2 | https://arxiv.org/pdf/2304.03344v2.pdf | Graph Collaborative Signals Denoising and Augmentation for Recommendation | Graph collaborative filtering (GCF) is a popular technique for capturing high-order collaborative signals in recommendation systems. However, GCF's bipartite adjacency matrix, which defines the neighbors being aggregated based on user-item interactions, can be noisy for users/items with abundant interactions and insufficient for users/items with scarce interactions. Additionally, the adjacency matrix ignores user-user and item-item correlations, which can limit the scope of beneficial neighbors being aggregated. In this work, we propose a new graph adjacency matrix that incorporates user-user and item-item correlations, as well as a properly designed user-item interaction matrix that balances the number of interactions across all users. To achieve this, we pre-train a graph-based recommendation method to obtain users/items embeddings, and then enhance the user-item interaction matrix via top-K sampling. We also augment the symmetric user-user and item-item correlation components to the adjacency matrix. Our experiments demonstrate that the enhanced user-item interaction matrix with improved neighbors and lower density leads to significant benefits in graph-based recommendation. Moreover, we show that the inclusion of user-user and item-item correlations can improve recommendations for users with both abundant and insufficient interactions. The code is in \url{https://github.com/zfan20/GraphDA}. | ['Zhang Dong', 'Philip S. Yu', 'Jiawei Zhang', 'Hao Peng', 'Ke Xu', 'Ziwei Fan'] | 2023-04-06 | null | null | null | null | ['collaborative-filtering'] | ['miscellaneous'] | [-4.58014160e-01 -2.08058104e-01 -4.73071158e-01 -1.51731342e-01
-1.37775511e-01 -6.14312291e-01 2.54412413e-01 1.82104230e-01
-5.87133616e-02 2.64070749e-01 7.65901625e-01 -3.75246614e-01
-7.82792449e-01 -1.11765516e+00 -2.78543741e-01 -4.52362180e-01
-4.95159924e-01 1.33417070e-01 9.43940952e-02 -5.31591654e-01
-8.42713863e-02 1.42576590e-01 -1.07929182e+00 1.67131647e-01
1.01586533e+00 7.14524567e-01 1.88327730e-01 5.30804694e-01
-2.27344394e-01 3.28311086e-01 -2.24149570e-01 -3.92521471e-01
4.22433913e-01 -5.57373941e-01 -1.47404581e-01 4.41143662e-02
1.98335186e-01 -9.20516774e-02 -8.27592194e-01 1.14784646e+00
4.21718150e-01 5.67103148e-01 5.52742600e-01 -1.11427176e+00
-1.46202421e+00 1.05980313e+00 -6.27730072e-01 2.61814356e-01
5.07045984e-01 -2.07725838e-01 1.68876851e+00 -1.01796699e+00
2.75156081e-01 1.03241479e+00 5.09596348e-01 2.40560502e-01
-1.04764903e+00 -6.60514832e-01 5.65472305e-01 4.39572893e-02
-1.36153221e+00 8.92345384e-02 8.96502316e-01 -3.23441446e-01
7.33548701e-01 5.91585040e-01 9.78053033e-01 7.48901248e-01
-1.40198231e-01 6.68474197e-01 2.81361789e-01 -1.55664384e-01
-1.87581759e-02 -7.45558552e-03 7.03836739e-01 4.83814478e-01
5.47332883e-01 -7.67726302e-02 -3.66479546e-01 -4.45966631e-01
1.01110899e+00 6.98952317e-01 -5.03520906e-01 -3.76469821e-01
-9.95099664e-01 9.86395419e-01 8.09262455e-01 5.10557830e-01
-3.76227319e-01 -1.63972214e-01 2.12880932e-02 5.01672685e-01
5.48207521e-01 5.03857672e-01 -3.38849515e-01 2.01299548e-01
-4.42283660e-01 -1.09952323e-01 8.79930794e-01 1.00865281e+00
8.32787812e-01 -5.10399528e-02 -2.34849602e-01 1.13234890e+00
4.78895247e-01 3.53776664e-01 2.71354347e-01 -4.92528141e-01
3.47952902e-01 8.41288626e-01 4.00146656e-02 -1.58795834e+00
-4.03363645e-01 -9.20990288e-01 -1.22379386e+00 -6.13245785e-01
3.03294569e-01 -2.43194461e-01 -3.85406077e-01 1.67241204e+00
3.40969145e-01 5.01841605e-01 -3.85285497e-01 1.14419591e+00
8.74226451e-01 5.19687414e-01 -3.03741395e-01 -3.22368592e-01
1.14496613e+00 -1.04262054e+00 -8.48036587e-01 1.03044376e-01
6.33129954e-01 -6.28288746e-01 1.40157604e+00 1.08106472e-01
-7.37694681e-01 -4.82138455e-01 -6.63949609e-01 2.12095022e-01
-3.00381303e-01 3.63513529e-02 8.99406612e-01 6.18309021e-01
-8.37488055e-01 4.95767921e-01 -3.62380922e-01 -2.06197709e-01
2.62923867e-01 4.76610750e-01 -1.73756957e-01 -2.44146824e-01
-1.60785699e+00 2.31711209e-01 -9.00195241e-02 1.33945560e-02
5.67900948e-03 -7.12265611e-01 -6.44733191e-01 3.23420942e-01
5.18493295e-01 -8.18040192e-01 5.91351449e-01 -7.32918024e-01
-1.20503414e+00 -2.70536035e-01 -1.16564641e-02 -6.35086894e-02
-7.44884685e-02 -1.90036580e-01 -8.41026843e-01 -2.96829283e-01
-2.39554271e-01 -1.60164237e-01 4.44424391e-01 -1.15152955e+00
-4.39744622e-01 -5.42293787e-01 3.76427770e-01 3.70473772e-01
-9.15265441e-01 -4.83776987e-01 -9.35557663e-01 -9.32717621e-01
2.52831519e-01 -8.98013532e-01 -3.42428446e-01 -4.05538380e-01
-2.16813102e-01 -2.59817004e-01 4.65770006e-01 -4.38869238e-01
1.97520936e+00 -2.22928309e+00 7.40969628e-02 8.86132181e-01
5.17676950e-01 4.46776301e-02 -5.92874289e-01 9.45951998e-01
8.25749263e-02 1.72121301e-01 3.31806302e-01 -7.80551732e-02
1.19994976e-01 2.80105144e-01 -3.92749393e-03 2.93714970e-01
-4.15214658e-01 1.01107883e+00 -1.28901255e+00 1.61519811e-01
2.32604206e-01 7.62391925e-01 -9.94626880e-01 6.73030168e-02
2.24292185e-02 3.49528939e-01 -6.28364921e-01 1.44912019e-01
5.17808616e-01 -6.18284345e-01 5.16741097e-01 -4.14125770e-01
3.22804064e-01 3.59028816e-01 -1.47837460e+00 1.31984580e+00
-5.28962612e-01 8.28998908e-02 4.07428741e-02 -6.15036547e-01
7.16557860e-01 1.82202309e-02 7.07674682e-01 -5.08793294e-01
7.91272968e-02 -4.57193889e-02 3.40497255e-01 -2.02666789e-01
6.02400303e-01 5.42193592e-01 4.29881774e-02 6.30890250e-01
7.21001849e-02 5.04276812e-01 3.88151526e-01 9.35682774e-01
1.18663371e+00 -5.45764029e-01 1.52528718e-01 -3.82894874e-01
3.21444392e-01 -7.04913020e-01 3.65240097e-01 6.22316837e-01
2.34749854e-01 4.60764468e-01 3.60406309e-01 -1.79782417e-02
-6.20922327e-01 -9.45804238e-01 1.84828565e-01 1.43738127e+00
4.27197188e-01 -1.06956840e+00 -1.49323583e-01 -7.27743030e-01
3.16720456e-01 4.91608441e-01 -7.54662454e-01 -3.27483177e-01
-1.00243874e-01 -7.96835303e-01 -2.93757856e-01 2.98418164e-01
-6.87648430e-02 -4.97320354e-01 8.23420584e-01 2.43442506e-01
2.05700044e-02 -4.54723924e-01 -1.51268387e+00 -3.12606573e-01
-5.94201505e-01 -1.20971370e+00 -5.70192456e-01 -5.24758399e-01
9.54843402e-01 9.15478408e-01 9.95132446e-01 3.83500785e-01
3.72679681e-01 6.70124173e-01 -9.05431211e-01 1.83444455e-01
2.42508441e-01 2.14854442e-02 3.40538740e-01 4.13930118e-01
3.42569798e-01 -8.93757463e-01 -1.04103398e+00 5.97876132e-01
-9.25465822e-01 -1.25238732e-01 2.74412364e-01 1.01114333e+00
3.42957884e-01 1.92351401e-01 7.99961805e-01 -1.25559366e+00
1.08842540e+00 -7.97890484e-01 -1.88542739e-01 8.67475197e-02
-7.97540605e-01 -3.00863951e-01 8.86043787e-01 -6.94861650e-01
-4.62263346e-01 -2.74418950e-01 -2.12415248e-01 -3.08974653e-01
4.31189865e-01 9.42445278e-01 -3.10981035e-01 -4.60757762e-02
5.87511063e-01 -8.85668769e-02 -2.50657797e-01 -6.32115245e-01
7.63525486e-01 6.37483835e-01 -6.58139139e-02 -1.77629694e-01
6.36438370e-01 9.95326936e-02 -4.44942862e-01 -7.26902902e-01
-9.02627945e-01 -1.01399744e+00 -3.89791340e-01 -1.31119475e-01
3.61018628e-01 -8.78373623e-01 -5.27653515e-01 -3.09471916e-02
-6.04265094e-01 -9.49935913e-02 -3.86242032e-01 9.39257443e-01
2.09263742e-01 6.05452776e-01 -8.38374734e-01 -7.09825158e-01
-3.12154770e-01 -7.40937054e-01 4.50610608e-01 1.67449892e-01
-1.04490697e-01 -1.33873034e+00 6.80605471e-02 1.69056684e-01
4.74236041e-01 -5.09461701e-01 5.47532141e-01 -8.56036484e-01
-3.85300547e-01 -4.59357083e-01 -3.17817807e-01 1.55556798e-01
6.49169862e-01 -2.16113657e-01 -5.72215877e-02 -5.39484680e-01
-5.39839745e-01 4.01523143e-01 8.13357532e-01 4.60932851e-01
9.72759604e-01 -5.10376871e-01 -2.75674820e-01 3.23463053e-01
1.15460598e+00 -1.98344350e-01 3.57017130e-01 -4.03785110e-01
1.28073072e+00 2.42467269e-01 3.32031757e-01 6.39302313e-01
5.44188023e-01 5.99501073e-01 1.40909076e-01 -2.71614864e-02
-3.01018246e-02 -5.64215600e-01 2.81258196e-01 1.44226813e+00
-1.22736521e-01 -7.78933823e-01 -4.00243998e-01 3.34007710e-01
-2.04515696e+00 -7.98085690e-01 -6.48560703e-01 2.36479139e+00
4.52503443e-01 -7.18913227e-02 5.05056024e-01 4.77816425e-02
7.09938943e-01 5.04756942e-02 -3.51225734e-01 8.76886845e-02
-1.02727957e-01 -2.80910973e-02 6.60105228e-01 7.96229422e-01
-7.32645154e-01 7.15105832e-01 5.03332710e+00 8.46535981e-01
-5.82419515e-01 2.01730892e-01 1.86074793e-01 -3.68195772e-01
-9.74880457e-01 -2.64184624e-01 -5.73517263e-01 6.78387523e-01
6.91645145e-01 -2.89403141e-01 8.55240464e-01 4.93520737e-01
4.10565197e-01 2.98197776e-01 -7.49832928e-01 8.71895492e-01
1.18295588e-01 -1.20638168e+00 1.56768963e-01 3.39466751e-01
9.62072253e-01 7.31560290e-02 3.82881574e-02 3.39063764e-01
7.47114122e-01 -5.46456397e-01 2.94654574e-02 5.23790002e-01
3.87880772e-01 -8.02884758e-01 7.05994070e-01 3.12261283e-01
-1.48706317e+00 -2.08684146e-01 -4.61611390e-01 -2.18766749e-01
2.33233035e-01 1.28126335e+00 -1.76737458e-01 8.98620546e-01
5.60926020e-01 1.02947867e+00 -2.87969470e-01 1.07609963e+00
-1.91315204e-01 8.38264763e-01 -4.21911895e-01 -1.81347400e-01
1.07962482e-01 -1.01705062e+00 5.22736073e-01 1.03863013e+00
6.06048346e-01 4.20928925e-01 5.64149976e-01 3.07047725e-01
-2.76702046e-01 6.36983752e-01 -5.83500504e-01 -1.82738006e-01
7.91586459e-01 1.38177717e+00 -5.14264286e-01 -3.31039429e-01
-7.41519094e-01 1.00305545e+00 4.21348929e-01 6.36314452e-01
-6.01517081e-01 -2.98318088e-01 9.60781336e-01 3.98689449e-01
2.76417851e-01 -4.83508795e-01 3.96764912e-02 -1.34671760e+00
-2.59244919e-01 -5.65409005e-01 6.32457793e-01 -3.50520253e-01
-1.84112620e+00 3.63911092e-01 -4.78054881e-01 -1.24478006e+00
2.88556904e-01 -2.85372674e-01 -6.21510625e-01 8.71589422e-01
-8.51058960e-01 -1.04418206e+00 -9.17854160e-02 7.61076987e-01
3.44595313e-02 -2.29593933e-01 5.92952371e-01 8.98023248e-01
-7.26802588e-01 9.61479545e-01 3.99530232e-01 9.92614925e-02
5.83655536e-01 -1.12889254e+00 1.89855888e-01 5.01474857e-01
6.67157114e-01 1.23382366e+00 3.41445357e-01 -7.85123706e-01
-1.72096205e+00 -1.30678618e+00 6.77963853e-01 -5.11237621e-01
9.47290301e-01 -5.23937106e-01 -9.02927876e-01 7.14935780e-01
5.75038418e-03 7.96415210e-02 1.31253040e+00 8.81927848e-01
-4.36060607e-01 2.28494853e-02 -7.97677755e-01 9.16244745e-01
1.46925282e+00 -6.35158658e-01 -2.52234697e-01 5.16629815e-01
8.18002462e-01 3.87898609e-02 -1.15988493e+00 3.26668844e-02
5.86067736e-01 -6.69085026e-01 9.43741977e-01 -5.81634820e-01
-4.99025807e-02 -4.34440374e-01 -1.86235085e-01 -1.76006031e+00
-1.25497973e+00 -6.10923231e-01 -4.70629960e-01 1.15806556e+00
6.39897287e-01 -8.03727627e-01 7.38757551e-01 5.07423401e-01
-1.02030508e-01 -8.10780823e-01 -3.53401452e-01 -6.25162840e-01
-2.72769541e-01 -3.91330093e-01 7.46928811e-01 1.23958766e+00
4.39533055e-01 6.88496172e-01 -8.27250719e-01 2.32549191e-01
2.73743331e-01 2.34914377e-01 7.60818779e-01 -1.21073997e+00
-6.79398596e-01 -4.69960272e-01 -1.27772242e-01 -1.51148200e+00
-2.73377776e-01 -1.12860370e+00 -5.59319377e-01 -1.78431821e+00
2.17557251e-01 -5.77077866e-01 -7.59640157e-01 1.55509487e-01
-3.18821818e-01 3.84474039e-01 1.04046553e-01 3.27133805e-01
-8.14740241e-01 4.95367289e-01 1.55854201e+00 7.46633783e-02
-6.69243574e-01 2.25145787e-01 -1.05284953e+00 3.01556081e-01
6.90953791e-01 -1.33388937e-01 -7.15197504e-01 -3.33793133e-01
6.13607347e-01 -1.90534502e-01 -2.58860171e-01 -5.04720688e-01
2.89516956e-01 -1.20527588e-01 2.07418352e-01 -3.01638722e-01
1.87337756e-01 -9.08132255e-01 5.06348312e-01 4.62081619e-02
-3.70453447e-01 -2.33371988e-01 -3.14596653e-01 1.04861963e+00
-5.05136549e-02 2.40930721e-01 3.11415792e-01 3.35452147e-02
-2.11167201e-01 9.19356108e-01 -1.37272477e-01 -3.06811094e-01
6.45768225e-01 -6.96388073e-03 -1.72881871e-01 -8.16807330e-01
-1.00087643e+00 5.26421726e-01 3.51716161e-01 5.72286487e-01
5.41163206e-01 -1.81186390e+00 -5.52695394e-01 2.59276807e-01
2.25802392e-01 -5.89033544e-01 5.80513537e-01 9.12229538e-01
1.75836027e-01 1.00831956e-01 2.64096767e-01 4.09800373e-02
-1.22234285e+00 6.56172812e-01 -8.81909132e-02 -2.99786508e-01
-3.38809013e-01 1.06338859e+00 2.30383456e-01 -6.59610629e-01
8.63178745e-02 -1.10651173e-01 -4.85141575e-01 3.03794235e-01
4.98793870e-01 4.21318054e-01 -2.49012813e-01 -6.99229360e-01
-3.78612950e-02 3.68453801e-01 -1.52293900e-02 3.12302023e-01
1.28072464e+00 -4.89335567e-01 -7.18043670e-02 3.44220698e-01
1.22862422e+00 6.97702527e-01 -5.42826951e-01 -4.37491655e-01
-4.52344716e-01 -8.79758775e-01 3.09583813e-01 -4.78329480e-01
-1.38905334e+00 4.65448767e-01 1.99435055e-01 8.14069271e-01
9.40159976e-01 -5.68390917e-03 9.63449121e-01 1.05937235e-01
4.59480911e-01 -7.55781353e-01 2.58994028e-02 7.58413792e-01
7.27017224e-01 -9.78607953e-01 6.83217123e-03 -6.65447712e-01
-4.91552591e-01 6.10725045e-01 4.87455130e-01 -3.01236987e-01
1.40039718e+00 -1.77700073e-01 -1.73158482e-01 -1.86832964e-01
-5.95512867e-01 -5.02063394e-01 8.33265305e-01 4.11307991e-01
7.42668271e-01 4.60530192e-01 -6.60323322e-01 1.17016506e+00
2.20360402e-02 -1.84008256e-01 1.96984261e-01 2.59491652e-01
-4.29557443e-01 -1.43537354e+00 7.78740048e-02 1.29276383e+00
-2.58989960e-01 -4.40449983e-01 -4.55261230e-01 1.69369981e-01
-2.24277768e-02 1.18915021e+00 2.92138718e-02 -1.02576101e+00
4.50705826e-01 -4.31673974e-01 2.45134264e-01 -9.76553619e-01
-6.41122937e-01 3.41870755e-01 6.85209036e-02 -5.29263377e-01
-6.69677928e-03 -2.68821716e-01 -1.06216836e+00 -6.24664545e-01
-9.14194882e-01 6.25115633e-01 9.18410793e-02 4.45836306e-01
7.67103553e-01 5.52990496e-01 9.21259165e-01 -7.71066904e-01
-1.63692594e-01 -9.15509641e-01 -1.20203328e+00 7.73280740e-01
-5.30146398e-02 -5.88672101e-01 -6.24271929e-01 -5.27548611e-01] | [10.146980285644531, 5.593122482299805] |
0bdfe3aa-455f-4086-a933-9f41c6b6750a | learning-sensorimotor-primitives-of | 2203.03797 | null | https://arxiv.org/abs/2203.03797v1 | https://arxiv.org/pdf/2203.03797v1.pdf | Learning Sensorimotor Primitives of Sequential Manipulation Tasks from Visual Demonstrations | This work aims to learn how to perform complex robot manipulation tasks that are composed of several, consecutively executed low-level sub-tasks, given as input a few visual demonstrations of the tasks performed by a person. The sub-tasks consist of moving the robot's end-effector until it reaches a sub-goal region in the task space, performing an action, and triggering the next sub-task when a pre-condition is met. Most prior work in this domain has been concerned with learning only low-level tasks, such as hitting a ball or reaching an object and grasping it. This paper describes a new neural network-based framework for learning simultaneously low-level policies as well as high-level policies, such as deciding which object to pick next or where to place it relative to other objects in the scene. A key feature of the proposed approach is that the policies are learned directly from raw videos of task demonstrations, without any manual annotation or post-processing of the data. Empirical results on object manipulation tasks with a robotic arm show that the proposed network can efficiently learn from real visual demonstrations to perform the tasks, and outperforms popular imitation learning algorithms. | ['Abdeslam Boularias', 'Kostas Bekris', 'Bowen Wen', 'Junchi Liang'] | 2022-03-08 | null | null | null | null | ['robot-manipulation'] | ['robots'] | [ 4.18263465e-01 1.09139256e-01 -2.46471077e-01 -2.34352246e-01
-1.26459956e-01 -5.07778943e-01 6.90831780e-01 7.79530481e-02
-8.11277211e-01 8.62624168e-01 -3.37724090e-01 -8.39201435e-02
-2.19507620e-01 -2.07042217e-01 -9.48991537e-01 -8.17406416e-01
-3.91529560e-01 7.84559488e-01 4.59670961e-01 4.70835231e-02
5.21729171e-01 8.23495567e-01 -1.55941117e+00 2.48420853e-02
5.13857365e-01 8.48796308e-01 9.22892570e-01 8.10192704e-01
3.31262052e-01 1.04530358e+00 -5.63732922e-01 4.40981567e-01
4.23591495e-01 -2.70287812e-01 -8.89327645e-01 3.26519310e-01
3.93892825e-02 -7.29276597e-01 -3.13279033e-01 1.10933220e+00
3.59899886e-02 5.32940149e-01 9.14076924e-01 -1.68391657e+00
-1.81888100e-02 4.79169577e-01 -1.53148457e-01 -8.21177214e-02
3.41164172e-01 7.13927150e-01 5.74184299e-01 -4.32363719e-01
5.70921659e-01 1.39501727e+00 8.04154724e-02 5.74080884e-01
-9.84061599e-01 -4.46172595e-01 2.31734678e-01 3.14420104e-01
-8.25348496e-01 -1.93720385e-01 4.77098018e-01 -7.73540318e-01
1.02428591e+00 -2.82765359e-01 5.62993467e-01 1.09948349e+00
4.52490032e-01 8.57312500e-01 1.09768939e+00 -3.73125017e-01
3.30925018e-01 -3.02922070e-01 2.70988625e-02 7.17407286e-01
1.47238687e-01 3.37302327e-01 -2.18998313e-01 6.02971017e-03
1.14371550e+00 2.70227939e-01 -2.57181823e-01 -8.97779822e-01
-1.52526414e+00 4.91481692e-01 3.72251421e-01 3.56495947e-01
-1.03363812e+00 4.54984605e-01 4.65975136e-01 5.21469235e-01
-4.46582794e-01 5.45963705e-01 -4.68541026e-01 -2.89905578e-01
-3.47735018e-01 3.89229476e-01 8.15195024e-01 1.25201249e+00
6.77292645e-01 -5.10912165e-02 -3.76234770e-01 2.74935126e-01
1.89781040e-02 3.22363198e-01 2.71790653e-01 -1.34141827e+00
6.27810299e-01 4.38818991e-01 8.12377930e-01 -4.84644741e-01
-3.80245984e-01 3.25896651e-01 -3.41954142e-01 1.08595371e+00
5.76071084e-01 -2.38809541e-01 -9.91125405e-01 1.51944554e+00
3.83246034e-01 -1.65530711e-01 6.05396554e-02 1.26639664e+00
2.39716128e-01 6.19279444e-01 1.99663535e-01 -2.16587737e-01
1.26247501e+00 -1.18886960e+00 -7.29376376e-01 -3.00793827e-01
3.03092390e-01 -4.65170950e-01 1.02281976e+00 5.90409100e-01
-1.08126926e+00 -8.08093369e-01 -9.45310056e-01 3.44262272e-02
-1.98938221e-01 4.51720327e-01 2.74940759e-01 -3.43009472e-01
-7.95216262e-01 8.35983455e-01 -1.14269161e+00 -4.54608321e-01
1.90058842e-01 5.16479552e-01 -5.01020730e-01 4.83349375e-02
-6.23389304e-01 1.32288158e+00 8.86706948e-01 2.34107330e-01
-1.89522111e+00 -9.74991098e-02 -7.88637340e-01 9.75532681e-02
1.00698745e+00 -4.81730849e-01 1.64492691e+00 -9.86724973e-01
-1.81628311e+00 7.15045989e-01 2.43248850e-01 -3.74229670e-01
6.41422033e-01 -5.86019933e-01 3.84171188e-01 2.48459533e-01
1.57652900e-01 8.65353644e-01 1.24102604e+00 -1.33092427e+00
-9.72686529e-01 -4.03578699e-01 4.92809951e-01 5.49128950e-01
4.83545363e-02 -5.71605563e-02 -3.76895159e-01 -1.80941612e-01
-2.60320663e-01 -1.13407195e+00 -1.98431537e-01 2.90240467e-01
-4.27848428e-01 -7.32141674e-01 1.11660075e+00 -3.61046433e-01
2.49296844e-01 -1.99831712e+00 8.11430335e-01 -1.92710176e-01
-2.46057138e-02 2.55574107e-01 -2.62304038e-01 5.69519281e-01
1.89334363e-01 -5.71641624e-01 -7.97501206e-02 -4.06859582e-03
5.86972870e-02 3.96149039e-01 -1.23269737e-01 4.55185801e-01
7.31198415e-02 7.31708586e-01 -1.23605239e+00 -2.73313314e-01
4.81288850e-01 9.93915349e-02 -1.38861805e-01 7.45665371e-01
-4.92016405e-01 8.20226252e-01 -6.43436134e-01 2.66531706e-01
3.10341623e-02 1.58501398e-02 2.37217501e-01 6.28334507e-02
-2.03721657e-01 2.40009278e-01 -1.01761496e+00 1.60722506e+00
-4.40993458e-01 5.10845900e-01 6.27949476e-01 -1.23816121e+00
5.45362949e-01 2.52244771e-01 4.03745443e-01 -2.15079561e-01
2.45951548e-01 1.61265641e-01 3.75431031e-01 -9.97461796e-01
1.35249436e-01 2.52322853e-01 -2.20749155e-03 4.03064370e-01
1.83659181e-01 -5.98812342e-01 4.81817991e-01 -9.37317684e-02
1.16598439e+00 7.60286868e-01 6.00229561e-01 -3.26374546e-03
4.77476716e-01 4.80743140e-01 2.29682148e-01 1.01297045e+00
-3.67426693e-01 -1.14116922e-01 3.82080585e-01 -2.22366169e-01
-1.10459447e+00 -8.76161456e-01 5.77114463e-01 1.17123389e+00
3.99179935e-01 2.53104120e-01 -7.31489539e-01 -6.60656095e-01
8.35175812e-02 6.54175043e-01 -5.38459301e-01 -9.53747854e-02
-1.03831923e+00 2.20448926e-01 -1.47961732e-02 5.77167034e-01
4.94111359e-01 -1.86878192e+00 -1.64983821e+00 1.35652944e-01
1.32592544e-01 -1.19689918e+00 -3.70668113e-01 5.58431745e-01
-9.19077337e-01 -1.47278321e+00 -6.65767789e-01 -1.21712542e+00
1.05379188e+00 2.29401574e-01 5.70211351e-01 -1.36114776e-01
-4.20532584e-01 5.56165576e-01 -3.33540201e-01 -3.84531677e-01
-5.17796576e-01 -2.86294192e-01 1.72127381e-01 -2.83075124e-01
-4.31794859e-03 -3.27567339e-01 -3.83978903e-01 3.15629333e-01
-5.65349817e-01 1.49622858e-01 1.03369331e+00 1.02580130e+00
3.21099162e-01 1.88884094e-01 1.55241951e-01 -1.98735178e-01
7.95214593e-01 -1.04439426e-02 -7.04760492e-01 6.29240796e-02
5.62225059e-02 1.67986318e-01 8.22305620e-01 -9.53869998e-01
-9.02895272e-01 4.53686625e-01 5.05434692e-01 -5.93310297e-01
-5.26960194e-01 2.35251799e-01 -5.05792424e-02 2.79226184e-01
3.44703674e-01 2.92334706e-01 2.98194826e-01 -2.42261738e-01
2.82414347e-01 5.58197916e-01 7.31788099e-01 -7.36108065e-01
6.77317441e-01 2.37894490e-01 1.27030700e-01 -6.42753124e-01
-4.19217318e-01 -5.34986377e-01 -1.08711386e+00 -4.02098238e-01
9.82929826e-01 -6.10803306e-01 -1.39927971e+00 5.98833382e-01
-1.13427579e+00 -1.07252347e+00 -1.54557720e-01 6.99001670e-01
-1.17142344e+00 1.66675627e-01 -6.36696041e-01 -7.84245968e-01
4.96186502e-02 -1.37024093e+00 1.08812761e+00 1.95558637e-01
-9.60466862e-02 -4.25934076e-01 -2.92616218e-01 -3.95386294e-02
1.22745201e-01 2.65417337e-01 9.95080233e-01 -5.97352445e-01
-6.28613412e-01 -9.94465954e-05 -4.06549778e-03 3.17261696e-01
4.19642001e-01 -4.64401156e-01 -2.94525027e-01 -7.11496055e-01
3.28977033e-02 -8.67917299e-01 4.67394799e-01 2.44294032e-01
1.08553433e+00 -2.82646179e-01 -6.16308391e-01 9.14588571e-02
1.04724407e+00 7.46823311e-01 2.25777000e-01 2.99159527e-01
4.78207111e-01 6.53794646e-01 1.27548039e+00 2.26976112e-01
-1.77988574e-01 8.41764808e-01 1.03612757e+00 3.73399377e-01
1.04826041e-01 -9.29430053e-02 6.34637773e-01 1.72404006e-01
-2.49048799e-01 -1.22240260e-01 -6.00466549e-01 4.66776937e-01
-2.26764321e+00 -8.54823828e-01 1.41688079e-01 2.02396297e+00
5.89562058e-01 2.12017640e-01 2.12948710e-01 7.87694603e-02
6.16111040e-01 -4.73426618e-02 -1.12299633e+00 -1.25833288e-01
6.51769102e-01 -2.75261879e-01 4.34156477e-01 4.20188487e-01
-1.12247252e+00 1.11272967e+00 5.46398163e+00 3.98773909e-01
-1.11352503e+00 -1.37273759e-01 -1.97644994e-01 -9.90765020e-02
7.83376098e-01 -9.44301859e-02 -4.92092997e-01 2.14232877e-01
3.03256333e-01 1.82218120e-01 7.75152743e-01 1.10714030e+00
3.87097776e-01 -5.25672495e-01 -1.77228653e+00 8.06522667e-01
-2.79392558e-03 -6.64716601e-01 -1.82586670e-01 -1.87934384e-01
2.87355959e-01 -9.89915803e-02 -2.18676597e-01 6.05665445e-01
6.47015989e-01 -8.80514026e-01 7.76430011e-01 4.16014522e-01
6.04050279e-01 -3.99766952e-01 3.53721440e-01 1.06603253e+00
-8.51726592e-01 -6.05025470e-01 -3.87189120e-01 -3.75072777e-01
1.79815024e-01 -2.23756313e-01 -1.32214153e+00 2.87363827e-01
6.38090909e-01 5.00018299e-01 2.13378653e-01 9.21694636e-01
-7.52558291e-01 2.33796872e-02 -1.68982744e-01 -5.02079308e-01
5.46591997e-01 -2.28047594e-01 6.57924056e-01 8.45691204e-01
1.75017670e-01 1.64186329e-01 7.67831862e-01 7.99772143e-01
3.17400008e-01 -3.74684036e-01 -8.42704058e-01 2.18532514e-02
2.48798698e-01 1.17114210e+00 -7.01695025e-01 -6.08184278e-01
-4.27137613e-02 1.21671200e+00 5.85057914e-01 5.48120499e-01
-6.69949591e-01 -5.50192416e-01 5.50394833e-01 -2.25052327e-01
4.63158906e-01 -6.92755580e-01 4.04947877e-01 -7.63116896e-01
2.84469455e-01 -9.94512618e-01 2.26169229e-02 -1.17645526e+00
-7.65555322e-01 2.37904936e-01 3.72049183e-01 -1.25441098e+00
-5.45363605e-01 -1.07922649e+00 -5.59894085e-01 7.82134235e-01
-1.26556063e+00 -8.09080243e-01 -5.34060717e-01 5.83001614e-01
1.08632910e+00 -2.21776754e-01 5.54651141e-01 -3.40184450e-01
-8.92784670e-02 -2.70283997e-01 -1.34801015e-01 1.45220324e-01
4.91372913e-01 -1.19657755e+00 -2.64789313e-01 5.13910413e-01
-2.77705103e-01 4.96611804e-01 7.54430056e-01 -6.94023311e-01
-1.57142472e+00 -6.24653637e-01 3.23506922e-01 -1.36803612e-01
5.63018441e-01 -4.40260947e-01 -6.76665485e-01 9.59206939e-01
2.64309913e-01 1.75566152e-02 -4.55102801e-01 -4.89597887e-01
2.71833450e-01 1.29457653e-01 -1.07510734e+00 7.37212539e-01
1.06348681e+00 -2.29907125e-01 -1.11567926e+00 5.97328305e-01
4.73222941e-01 -6.34312093e-01 -4.55783039e-01 4.56443608e-01
5.02408624e-01 -5.70927620e-01 8.89082193e-01 -9.96626914e-01
4.70203102e-01 -5.34717917e-01 2.83905417e-01 -1.55096197e+00
-3.32751364e-01 -6.59789026e-01 -1.85721219e-01 4.11009282e-01
-1.86837792e-01 -3.20062846e-01 5.83147645e-01 8.17729533e-02
-1.98488727e-01 -5.82348347e-01 -8.78278852e-01 -9.84535396e-01
-3.16427141e-01 1.14726953e-01 3.46378833e-02 5.03537476e-01
1.84145018e-01 2.05441818e-01 -4.36764002e-01 3.88033301e-01
5.13630211e-01 2.52383441e-01 1.22384107e+00 -1.09131837e+00
-7.52283931e-02 -3.02664727e-01 -2.71283090e-01 -1.48570776e+00
5.36076128e-01 -6.99822068e-01 8.69738460e-01 -1.68864250e+00
3.76948938e-02 -2.65487373e-01 -9.97796282e-02 7.79649734e-01
-2.84337392e-03 -5.20500124e-01 4.42273438e-01 3.65754783e-01
-5.76463401e-01 4.83234942e-01 1.57815063e+00 -2.59293497e-01
-3.18971127e-01 1.89644784e-01 2.62998909e-01 7.78565466e-01
7.35429764e-01 -4.09284294e-01 -4.50910836e-01 -2.72371501e-01
-3.81530702e-01 6.06820226e-01 5.66194296e-01 -1.03669381e+00
3.82351011e-01 -5.84046483e-01 3.75493795e-01 -5.04989803e-01
5.46434462e-01 -1.25913727e+00 -2.13119149e-01 1.24234474e+00
-6.04659140e-01 1.25054702e-01 6.01150505e-02 5.99426806e-01
3.63183469e-02 -6.26847446e-01 6.00405693e-01 -3.71576697e-01
-1.05200970e+00 3.52385454e-02 -7.15608239e-01 -3.67865562e-01
1.75632465e+00 -2.43027210e-02 -6.32231310e-02 -2.77917057e-01
-1.14854896e+00 6.11561894e-01 3.43668163e-01 6.83779776e-01
5.00170290e-01 -9.09633398e-01 -4.39239055e-01 -1.05123214e-01
-3.55169959e-02 1.36289299e-01 -3.33180606e-01 7.20747054e-01
-4.47129756e-01 5.09263277e-01 -7.88263083e-01 -8.51075053e-01
-1.34696794e+00 9.23216224e-01 2.95444191e-01 -2.26351708e-01
-9.62346852e-01 4.88072783e-01 2.38733947e-01 -4.49101299e-01
6.60558820e-01 -6.45861566e-01 -3.27044010e-01 -4.08687472e-01
1.67319834e-01 3.45597506e-01 -4.96789008e-01 -2.66941398e-01
-1.36120304e-01 3.40106070e-01 1.48319438e-01 -1.57155335e-01
1.26597762e+00 1.31118074e-01 -1.64375931e-01 6.85853243e-01
7.30281651e-01 -7.79581845e-01 -1.95607829e+00 -1.64250985e-01
4.13498282e-02 -3.57797354e-01 -4.89085436e-01 -9.54235435e-01
-4.63485003e-01 9.44640219e-01 3.18133771e-01 -4.42187376e-02
8.38178635e-01 -9.49083045e-02 2.49004081e-01 1.17111397e+00
9.87172544e-01 -1.40834224e+00 6.95749402e-01 7.19045877e-01
1.30986679e+00 -1.22916639e+00 -1.06434738e-02 -1.54033065e-01
-6.31540775e-01 1.13974845e+00 1.01325130e+00 -4.34558332e-01
2.50235200e-01 1.57468364e-01 -5.26595823e-02 -7.70133287e-02
-7.23767221e-01 -1.49441853e-01 4.28848825e-02 6.55611455e-01
-2.52642602e-01 -1.99590512e-02 -1.47867978e-01 -2.74178773e-01
9.85467285e-02 1.63666055e-01 4.05634016e-01 1.46259427e+00
-8.52878451e-01 -6.44692302e-01 -3.06493670e-01 4.05911088e-01
-8.98277983e-02 4.73351926e-01 -3.48911256e-01 1.05089235e+00
-6.10337919e-03 6.98488891e-01 8.88404530e-03 1.31710500e-01
4.74230200e-01 1.26607016e-01 9.06727612e-01 -8.53475392e-01
-5.32272577e-01 -9.09625217e-02 -9.26230624e-02 -7.97663987e-01
-4.98157769e-01 -7.13954508e-01 -1.44583929e+00 5.22517741e-01
-2.06074677e-02 6.33833185e-02 7.49632299e-01 1.04568243e+00
4.36527841e-02 6.48490846e-01 3.35578948e-01 -1.63837111e+00
-1.18316960e+00 -1.25865972e+00 -2.06701577e-01 5.90366662e-01
6.25235736e-01 -1.01445377e+00 -1.46488860e-01 9.06621367e-02] | [4.629035949707031, 0.8072786331176758] |
0214456f-d20d-4f72-953b-658462e5ffa4 | covid-19-literature-mining-and-retrieval | 2205.14781 | null | https://arxiv.org/abs/2205.14781v1 | https://arxiv.org/pdf/2205.14781v1.pdf | COVID-19 Literature Mining and Retrieval using Text Mining Approaches | The novel coronavirus disease (COVID-19) began in Wuhan, China, in late 2019 and to date has infected over 148M people worldwide, resulting in 3.12M deaths. On March 10, 2020, the World Health Organisation (WHO) declared it as a global pandemic. Many academicians and researchers started to publish papers describing the latest discoveries on covid-19. The large influx of publications made it hard for other researchers to go through a large amount of data and find the appropriate one that helps their research. So, the proposed model attempts to extract relavent titles from the large corpus of research publications which makes the job easy for the researchers. Allen Institute for AI released the CORD-19 dataset, which consists of 2,00,000 journal articles related to coronavirus-related research publications from PubMed's PMC, WHO (World Health Organization), bioRxiv, and medRxiv pre-prints. Along with this document corpus, they have also provided a topics dataset named topics-rnd3 consisting of a list of topics. Each topic has three types of representations like query, question, and narrative. These Datasets are made open for research, and also they released a TREC-COVID competition on Kaggle. Using these topics like queries, our goal is to find out the relevant documents in the CORD-19 dataset. In this research, relevant documents should be recognized for the posed topics in topics-rnd3 data set. The proposed model uses Natural Language Processing(NLP) techniques like Bag-of-Words, Average Word-2-Vec, Average BERT Base model and Tf-Idf weighted Word2Vec model to fabricate vectors for query, question, narrative, and combinations of them. Similarly, fabricate vectors for titles in the CORD-19 dataset. After fabricating vectors, cosine similarity is used for finding similarities between every two vectors. Cosine similarity helps us to find relevant documents for the given topic. | ['Rohit Chivukula', 'T. Jaya Lakshmi', 'Satti Thanuja Pavani', 'Sanku Satya Uday'] | 2022-05-29 | null | null | null | null | ['literature-mining'] | ['natural-language-processing'] | [-1.96763143e-01 -4.31715906e-01 -3.85618895e-01 -1.10943712e-01
-6.39163315e-01 -6.25980794e-01 7.58208632e-01 6.31125808e-01
-5.65505922e-01 8.20305943e-01 6.92532241e-01 -3.38380575e-01
-2.01005742e-01 -7.24038064e-01 -2.00505167e-01 -4.30843055e-01
-1.29452124e-01 7.09561169e-01 1.23791121e-01 -1.83194116e-01
5.14973640e-01 3.28756630e-01 -1.31234860e+00 1.86963916e-01
8.67739975e-01 6.81955040e-01 6.81804717e-01 3.99835110e-01
-6.58188343e-01 9.33788121e-02 -7.19914496e-01 -1.25430569e-01
-1.09656021e-01 -2.20646426e-01 -7.81134427e-01 -5.88195264e-01
5.34749068e-02 6.50596991e-02 -1.98786288e-01 7.62572348e-01
5.26015401e-01 2.14942824e-02 1.04307461e+00 -1.32328701e+00
-5.17613351e-01 1.99928701e-01 -5.68459272e-01 5.84733367e-01
2.73182005e-01 -3.19149435e-01 1.02554893e+00 -9.61796582e-01
1.29997635e+00 1.24098468e+00 3.69279325e-01 4.13370520e-01
-5.02714455e-01 -7.58573830e-01 -9.50825363e-02 2.95352191e-01
-1.21015656e+00 3.53797942e-01 3.56867135e-01 -7.26302624e-01
1.17744863e+00 4.95762914e-01 6.40598416e-01 1.36055803e+00
8.06681991e-01 4.59526598e-01 8.09129834e-01 -1.01945922e-02
9.83086079e-02 2.97145277e-01 4.64644402e-01 2.05005735e-01
5.08797586e-01 -1.63818970e-01 -1.30630553e-01 -6.90969348e-01
2.32014477e-01 6.10400915e-01 -3.40154052e-01 2.37107500e-01
-1.28910828e+00 1.05728388e+00 3.49314674e-03 6.90845013e-01
-5.56269646e-01 -3.76086533e-01 7.83110201e-01 2.07938358e-01
3.71539414e-01 6.37008190e-01 -7.83337295e-01 3.72104496e-02
-7.96887100e-01 7.70147204e-01 6.14649355e-01 9.31193590e-01
3.44395816e-01 -5.24654448e-01 -5.83379865e-01 9.07739162e-01
5.55662334e-01 7.61234641e-01 7.58382976e-01 -2.30824038e-01
3.09907734e-01 7.34438658e-01 -1.01447757e-02 -1.33374059e+00
-6.97283745e-01 -2.04877689e-01 -9.80469227e-01 -7.06911147e-01
-3.12341809e-01 -2.95903862e-01 -1.04438102e+00 1.52310872e+00
2.85141408e-01 1.49666980e-01 4.90884893e-02 8.01873207e-01
1.41448426e+00 1.14732933e+00 3.13936621e-01 -5.02011538e-01
2.00200033e+00 -5.86986721e-01 -1.09706807e+00 2.49545678e-01
5.68218172e-01 -9.97234702e-01 4.50447887e-01 4.03252661e-01
-4.06174421e-01 -2.08798289e-01 -7.50058413e-01 6.62691519e-02
-1.08401680e+00 -1.93296984e-01 5.87001145e-01 2.98194915e-01
-7.86506116e-01 1.10633276e-01 -2.78869033e-01 -1.05613530e+00
3.09235275e-01 -8.79802406e-02 -2.70301372e-01 -1.96483165e-01
-1.58135486e+00 1.10201931e+00 3.74977350e-01 -3.23311329e-01
-9.43989515e-01 -9.38978136e-01 -4.88883495e-01 -2.26534590e-01
8.94282013e-02 -7.91716695e-01 4.32697713e-01 2.55529076e-01
-5.92185378e-01 7.91331172e-01 -2.76281178e-01 -1.82993650e-01
-1.39483064e-01 -1.53264776e-01 -9.59630668e-01 1.47312105e-01
5.03157258e-01 6.11161411e-01 4.65133131e-01 -7.81684935e-01
-7.44397223e-01 -5.29719830e-01 -4.53592062e-01 -9.02337581e-02
-2.98561692e-01 5.07846475e-01 -7.65034497e-01 -8.44245255e-01
-3.15485120e-01 -8.93543303e-01 -3.22820038e-01 -5.26378214e-01
-4.75201517e-01 -7.57714450e-01 9.93812323e-01 -6.51091039e-01
1.35681772e+00 -1.99718237e+00 -1.21481217e-01 1.88163534e-01
4.17343527e-01 3.59598905e-01 -3.45319271e-01 8.66671860e-01
-2.75380001e-03 5.39339304e-01 6.68273568e-02 3.05345923e-01
-4.62182134e-01 1.60399944e-01 -5.41613281e-01 4.00221735e-01
5.75952381e-02 8.11581373e-01 -8.40432167e-01 -5.74560106e-01
-1.85475007e-01 6.28505886e-01 -4.32979435e-01 3.44637454e-01
-3.57859701e-01 1.12446904e-01 -1.05036294e+00 5.23491859e-01
8.55074763e-01 -2.09590539e-01 1.05328657e-01 -1.81581706e-01
-2.73678541e-01 6.48127347e-02 -5.27714550e-01 1.40931809e+00
9.53498483e-03 6.19665504e-01 -3.95518541e-01 -8.69038522e-01
9.20121193e-01 7.09800720e-01 7.69879043e-01 -4.24324274e-01
3.96995544e-01 -8.39696229e-02 -3.56095046e-01 -8.66382241e-01
4.55633342e-01 -2.64112614e-02 -2.28283346e-01 2.73867965e-01
4.74367151e-03 1.63035542e-02 2.57344186e-01 4.42174554e-01
1.16188908e+00 -4.09600109e-01 2.00666696e-01 -3.78653497e-01
3.63732666e-01 4.11462456e-01 6.42675698e-01 5.88431180e-01
1.17532559e-01 4.52780128e-01 3.08196753e-01 -5.34216762e-01
-8.83761704e-01 -9.94666278e-01 -5.87153435e-01 7.57527709e-01
-7.94300213e-02 -8.06630790e-01 -2.87271738e-01 -5.19063294e-01
9.47587863e-02 5.09176433e-01 -8.12058449e-01 2.06786782e-01
-3.57816726e-01 -7.83718884e-01 7.51402736e-01 -2.62025911e-02
3.08098823e-01 -1.07670176e+00 -6.06328607e-01 2.00186834e-01
-3.34692329e-01 -8.11167657e-01 -5.39735258e-01 -1.14694253e-01
-8.89442623e-01 -1.24626911e+00 -1.17048454e+00 -8.49583685e-01
3.86051506e-01 3.88453841e-01 7.73717880e-01 -6.68546855e-02
-7.77818739e-01 5.19688018e-02 -5.57585776e-01 -8.26517224e-01
5.80903403e-02 -9.93302241e-02 1.42561615e-01 -6.14692748e-01
7.16390550e-01 -1.00451402e-01 -6.35779440e-01 5.14210090e-02
-1.00295413e+00 -3.69486392e-01 4.84238058e-01 7.76442885e-01
5.12673736e-01 -1.58932030e-01 7.71658242e-01 -9.08352852e-01
1.07355225e+00 -1.15752673e+00 -3.75839651e-01 3.67850155e-01
-9.42083776e-01 -1.24576643e-01 2.79179841e-01 -3.43583763e-01
-6.94510758e-01 -6.95441663e-01 -1.55413881e-01 -3.83925647e-01
-7.84430578e-02 9.51011062e-01 2.22599760e-01 7.45510161e-01
5.24697602e-01 1.70604274e-01 -1.92386210e-01 -6.33624017e-01
3.89576674e-01 1.12056088e+00 8.60842168e-02 -1.30064026e-01
4.75650460e-01 2.65399843e-01 -2.66791075e-01 -1.21330631e+00
-4.13308412e-01 -1.17655647e+00 2.50741839e-01 -5.19370176e-02
1.25627935e+00 -8.36181700e-01 -9.06005561e-01 1.79407611e-01
-1.52619433e+00 6.98696792e-01 3.13168257e-01 7.47910142e-01
5.18929586e-02 2.07859695e-01 -4.65947658e-01 -4.37299073e-01
-9.75120008e-01 -1.00800812e+00 8.70261192e-01 1.82386771e-01
-5.72537720e-01 -9.03906345e-01 5.73690593e-01 2.45380044e-01
6.27742469e-01 1.40235409e-01 1.40498173e+00 -1.15088081e+00
-1.02130219e-01 -1.38200924e-01 -3.54045063e-01 -5.14714271e-02
2.94804364e-01 -1.36713609e-01 -6.03408039e-01 -2.22183287e-01
2.69306917e-02 1.20489433e-01 1.01743114e+00 5.09649932e-01
1.04670823e+00 -4.22998637e-01 -1.02356851e+00 2.54995823e-01
1.36185837e+00 6.87232912e-01 6.22066379e-01 2.75672346e-01
4.19831187e-01 5.56356728e-01 6.41989887e-01 4.40279365e-01
3.54529619e-01 4.93352532e-01 -4.44415305e-03 1.59424484e-01
4.17113572e-01 -5.65917715e-02 -1.27264425e-01 1.14961052e+00
-5.05385548e-02 -4.95795369e-01 -1.00312769e+00 7.18753099e-01
-1.47237468e+00 -8.37062418e-01 -2.70063698e-01 1.81862628e+00
7.92809129e-01 -2.97455162e-01 -1.69209629e-01 -4.60136652e-01
4.86124486e-01 4.31860425e-02 -3.35803449e-01 -6.32421374e-01
-2.24160343e-01 1.39438406e-01 3.34548652e-01 5.31617999e-02
-9.61799383e-01 7.52118170e-01 5.87430286e+00 9.12505805e-01
-1.18545401e+00 1.48988277e-01 3.43617588e-01 2.02952355e-01
-6.62610292e-01 -1.50969669e-01 -9.50808108e-01 5.85621953e-01
1.09614205e+00 -6.78694904e-01 -1.23492934e-01 7.29851604e-01
2.70397842e-01 -1.96851436e-02 -5.05445421e-01 9.80631232e-01
3.05252165e-01 -1.71512413e+00 3.73564124e-01 9.57029909e-02
6.97414577e-01 4.51422274e-01 -1.22083284e-01 4.78172988e-01
2.70944051e-02 -1.08934236e+00 -1.67068243e-01 7.26954460e-01
7.76423395e-01 -5.75456023e-01 1.06720030e+00 2.44603112e-01
-9.05739605e-01 2.75300860e-01 -5.77699423e-01 5.44629514e-01
6.42270073e-02 7.72375047e-01 -1.12476051e+00 6.64171338e-01
8.22143555e-01 6.01798773e-01 -1.77678928e-01 1.24711049e+00
3.35425049e-01 6.42570019e-01 -8.78687501e-02 -5.19060016e-01
3.53518307e-01 -1.24806911e-01 7.28946209e-01 1.58459663e+00
4.23535794e-01 1.77515507e-01 4.57329415e-02 6.88531637e-01
-1.63841665e-01 6.46915674e-01 -7.62476802e-01 -4.93047416e-01
5.01820207e-01 1.21047425e+00 -6.40208185e-01 -3.85677010e-01
-3.07569891e-01 6.06329441e-01 -2.59636402e-01 4.66152370e-01
-5.70494354e-01 -6.72243178e-01 7.78409064e-01 -1.86216757e-02
1.57109872e-01 1.15106225e-01 1.43378124e-01 -7.52723038e-01
-4.38303113e-01 -8.66492629e-01 6.13032401e-01 -6.57813191e-01
-1.61377871e+00 7.84150839e-01 3.00638169e-01 -1.16117370e+00
2.28083543e-02 -3.21020633e-01 -5.53510249e-01 1.09818292e+00
-1.09974790e+00 -7.31172442e-01 9.18172300e-02 3.63970459e-01
6.68680489e-01 -4.48248535e-01 1.08655977e+00 3.81084710e-01
-2.86256462e-01 1.80070236e-01 3.93363714e-01 4.56779078e-02
8.62370431e-01 -6.98646247e-01 2.51438051e-01 -9.32542980e-02
-2.76625007e-01 1.33647609e+00 7.85671115e-01 -1.18551731e+00
-1.53437901e+00 -1.33492529e+00 1.58665431e+00 -3.70840102e-01
5.18848002e-01 -1.56314000e-01 -8.23537529e-01 3.02831441e-01
4.71282959e-01 -5.98393977e-01 8.87383044e-01 5.31311631e-02
-3.07926059e-01 1.81041226e-01 -1.16760576e+00 5.34964323e-01
4.42631483e-01 -2.34609842e-01 -1.03847122e+00 7.78227329e-01
1.10034442e+00 3.54524776e-02 -8.13817382e-01 3.74745876e-01
5.05578697e-01 -4.36055586e-02 8.62228870e-01 -1.04692841e+00
4.41378117e-01 -2.49094009e-01 -1.16177566e-01 -1.31928062e+00
-1.18914790e-01 -2.76988685e-01 3.67239535e-01 1.06217957e+00
5.85754216e-01 -6.09360695e-01 2.90524185e-01 -5.82119562e-02
-5.58890067e-02 -1.18501759e+00 -6.83429658e-01 -6.76022947e-01
2.07463935e-01 -4.69023854e-01 5.38096726e-01 1.23896301e+00
1.31808892e-01 5.17500103e-01 -3.15381646e-01 -1.82259262e-01
2.79860079e-01 9.04335156e-02 4.22419101e-01 -1.47093821e+00
3.82986307e-01 -2.62875587e-01 -4.26829815e-01 -7.80126810e-01
-2.77697861e-01 -1.07226312e+00 -2.25097030e-01 -1.94135725e+00
4.97658163e-01 -4.44051713e-01 -2.91472286e-01 2.93674290e-01
-1.62777096e-01 -2.11769834e-01 -1.09123290e-02 2.78100073e-01
-3.40970069e-01 3.25365365e-01 1.15219450e+00 -4.30781931e-01
-1.45111769e-01 -3.00174326e-01 -8.01998258e-01 3.40713769e-01
7.25470543e-01 -6.51566744e-01 -3.28043938e-01 1.48379141e-02
4.39556390e-01 1.09545961e-01 -1.18756115e-01 -5.25733709e-01
2.09833130e-01 -3.63279134e-01 3.75120759e-01 -1.30264127e+00
2.19546214e-01 -5.75935543e-01 2.40806088e-01 3.68838161e-01
-3.54023099e-01 2.81505674e-01 3.33464175e-01 5.22111893e-01
-1.59246162e-01 -8.00707042e-02 1.68817848e-01 -8.54705051e-02
-4.13832664e-01 5.05771875e-01 -7.98579097e-01 4.97081429e-01
1.00041366e+00 2.01348901e-01 -6.27971888e-01 -2.19530687e-01
-2.74563134e-01 5.81845343e-01 -1.67970866e-01 8.77020478e-01
9.64905262e-01 -1.12885201e+00 -9.70543146e-01 -1.43753877e-02
3.89101803e-01 -3.58853549e-01 5.93493879e-01 7.21955180e-01
-3.87113601e-01 1.08499300e+00 -1.71183422e-01 -3.73625129e-01
-1.31787336e+00 8.76441360e-01 -4.79440242e-01 -2.91610390e-01
-5.23455024e-01 6.74121439e-01 2.64637142e-01 -3.47985357e-01
1.27114311e-01 -2.21456319e-01 -8.95572603e-01 5.23763776e-01
8.08777630e-01 1.70779541e-01 -3.29392105e-02 -5.62160254e-01
-7.12932944e-01 6.07205749e-01 -4.75065768e-01 -1.21614954e-03
1.43533838e+00 1.81584403e-01 -4.26517636e-01 2.99551249e-01
1.74042213e+00 -2.10546330e-02 2.93934256e-01 -2.25399248e-02
6.99638724e-02 -7.42954984e-02 -1.20264702e-01 -8.84684920e-01
-8.69612813e-01 6.54634058e-01 7.16887236e-01 -1.67388264e-02
8.41304302e-01 2.09459990e-01 1.07237375e+00 3.60052079e-01
1.60170719e-01 -8.53648126e-01 -3.81577373e-01 7.47633517e-01
1.05077004e+00 -9.86665905e-01 3.73931080e-02 -1.41270831e-01
-6.20650411e-01 8.79342854e-01 1.15373857e-01 2.17264757e-01
1.16617286e+00 -6.37689698e-03 1.98929593e-01 -7.55372226e-01
-1.02227485e+00 3.18720080e-02 4.72254664e-01 6.09658122e-01
5.55865586e-01 1.24581844e-01 -1.07665038e+00 8.64261866e-01
-1.12842180e-01 -1.05170356e-02 2.32289061e-01 8.66028666e-01
-5.03131032e-01 -1.11503112e+00 -2.05208972e-01 1.07644117e+00
-7.39239812e-01 -3.16844821e-01 -2.02933654e-01 6.14289701e-01
2.56322742e-01 1.12487614e+00 6.93106204e-02 -4.57029462e-01
1.43953085e-01 3.65573280e-02 -3.04245979e-01 -5.93609452e-01
-6.03805125e-01 4.34076861e-02 7.07855150e-02 -2.96728492e-01
-2.70342231e-01 -4.53834921e-01 -1.42587042e+00 -1.37543157e-01
-2.52370209e-01 8.23783517e-01 8.91787589e-01 8.61164510e-01
5.71173966e-01 3.40112597e-01 5.12643099e-01 8.02251920e-02
3.42457294e-02 -1.18116200e+00 -5.09757817e-01 3.26994181e-01
5.15962467e-02 -5.97588539e-01 -2.27739215e-01 -1.07153825e-01] | [8.770913124084473, 8.604622840881348] |
dd033814-dc0c-4561-a050-4bac814116de | sllen-semantic-aware-low-light-image | 2211.11571 | null | https://arxiv.org/abs/2211.11571v2 | https://arxiv.org/pdf/2211.11571v2.pdf | SLLEN: Semantic-aware Low-light Image Enhancement Network | How to effectively explore semantic feature is vital for low-light image enhancement (LLE). Existing methods usually utilize the semantic feature that is only drawn from the output produced by high-level semantic segmentation (SS) network. However, if the output is not accurately estimated, it would affect the high-level semantic feature (HSF) extraction, which accordingly interferes with LLE. To this end, we develop a simple and effective semantic-aware LLE network (SSLEN) composed of a LLE main-network (LLEmN) and a SS auxiliary-network (SSaN). In SLLEN, LLEmN integrates the random intermediate embedding feature (IEF), i.e., the information extracted from the intermediate layer of SSaN, together with the HSF into a unified framework for better LLE. SSaN is designed to act as a SS role to provide HSF and IEF. Moreover, thanks to a shared encoder between LLEmN and SSaN, we further propose an alternating training mechanism to facilitate the collaboration between them. Unlike currently available approaches, the proposed SLLEN is able to fully lever the semantic information, e.g., IEF, HSF, and SS dataset, to assist LLE, thereby leading to a more promising enhancement performance. Comparisons between the proposed SLLEN and other state-of-the-art techniques demonstrate the superiority of SLLEN with respect to LLE quality over all the comparable alternatives. | ['Chuheng Chen', 'DaCheng Tao', 'Jinhui Tang', 'Jinshan Pan', 'Charles A. Guo', 'Mingye Ju'] | 2022-11-21 | null | null | null | null | ['low-light-image-enhancement'] | ['computer-vision'] | [ 3.07201684e-01 6.15759343e-02 -1.01882100e-01 -4.01018977e-01
-4.63114470e-01 -8.15598667e-02 2.73283631e-01 -2.08335057e-01
-4.35312212e-01 6.97822332e-01 1.66415334e-01 9.24662650e-02
1.15068695e-02 -1.02457213e+00 -6.77790940e-01 -8.57029617e-01
4.89950091e-01 -3.70480299e-01 3.13489288e-01 -2.04919592e-01
4.39864509e-02 4.58651721e-01 -1.50971937e+00 4.07100283e-02
1.18209076e+00 1.42104447e+00 6.94095910e-01 4.56434824e-02
-5.68209171e-01 5.91256022e-01 -3.38973433e-01 -2.06727162e-01
1.08368844e-01 -4.89081860e-01 -4.77985620e-01 1.31123187e-02
-8.03364739e-02 -4.73841608e-01 -2.85674542e-01 1.13769996e+00
4.89866912e-01 1.94639713e-01 3.69496614e-01 -1.27220333e+00
-2.26484731e-01 2.70008862e-01 -6.22550607e-01 -1.81891635e-01
-2.48714257e-02 2.89101422e-01 9.35436606e-01 -9.39821839e-01
6.78279042e-01 1.19593084e+00 4.81401652e-01 3.17665786e-01
-9.25274253e-01 -7.97815442e-01 2.53450066e-01 1.44173130e-01
-1.12450504e+00 -3.61615151e-01 1.30118108e+00 4.05671224e-02
3.39489102e-01 1.06912747e-01 6.96504235e-01 8.28031898e-01
-7.06716720e-03 1.13796866e+00 1.37632155e+00 -3.24932277e-01
4.69799191e-02 4.16893393e-01 1.41112891e-03 8.44229400e-01
-8.43203738e-02 1.59260947e-02 -4.58121717e-01 2.74433970e-01
8.72721314e-01 -9.25384387e-02 -5.64663231e-01 -2.78927028e-01
-8.27835917e-01 5.30482590e-01 9.63021755e-01 3.69589597e-01
-5.35502911e-01 6.75105527e-02 2.35084921e-01 -9.05073509e-02
3.38961303e-01 2.99794585e-01 -4.30853575e-01 1.79879472e-01
-1.08135855e+00 -3.19449604e-01 4.42776710e-01 6.99769437e-01
1.19593322e+00 -1.61629915e-02 -3.99731755e-01 8.41075003e-01
5.22073388e-01 3.80717665e-01 2.15060711e-01 -8.91956210e-01
4.14155781e-01 9.31151927e-01 3.54902111e-02 -1.07505453e+00
-3.84300798e-01 -8.89827132e-01 -8.69403183e-01 1.90771490e-01
6.88579166e-03 9.99423489e-03 -7.74886847e-01 2.01130700e+00
3.52277309e-01 4.03011650e-01 1.18377097e-01 9.92640197e-01
1.04331708e+00 6.71022296e-01 1.32513270e-01 -2.22331628e-01
1.08406830e+00 -1.28143811e+00 -7.31203020e-01 -2.25331694e-01
3.34005892e-01 -6.97963595e-01 1.21615601e+00 2.55522013e-01
-8.88235688e-01 -7.77720869e-01 -1.15321743e+00 -1.66781843e-01
-2.94743747e-01 7.44601250e-01 6.67083740e-01 3.14146221e-01
-9.03375745e-01 4.83382225e-01 -5.66104352e-01 -1.12730727e-01
6.86109662e-01 3.77215296e-01 -2.87509948e-01 -1.99201539e-01
-1.43725562e+00 5.01749158e-01 4.58188593e-01 5.88582337e-01
-5.74187934e-01 -4.82432932e-01 -8.63535225e-01 7.07476661e-02
4.67741132e-01 -6.79591298e-01 7.23621607e-01 -9.45236325e-01
-1.65230775e+00 3.97427261e-01 -2.65710145e-01 -5.37847541e-02
4.41068619e-01 -5.13336323e-02 -3.28780770e-01 5.65390527e-01
2.68867183e-02 9.07675505e-01 8.58671963e-01 -1.66748047e+00
-4.97397006e-01 -1.65994078e-01 2.84438103e-01 4.14583355e-01
-6.79485917e-01 -2.78316766e-01 -6.19328141e-01 -6.16396248e-01
2.00480834e-01 -6.09520018e-01 -5.74834831e-02 2.94869035e-01
-5.05740523e-01 -1.26477972e-01 9.42352533e-01 -8.36377621e-01
1.35644603e+00 -2.25334835e+00 6.78818952e-03 8.67532641e-02
2.08651543e-01 5.42121708e-01 -2.07483694e-01 2.29738325e-01
6.71748295e-02 5.74637856e-03 -4.40299153e-01 -4.01663780e-01
-3.02097518e-02 1.66599005e-01 -8.30902383e-02 2.86639184e-01
3.84957135e-01 1.16894090e+00 -9.84024346e-01 -7.45714247e-01
6.33996665e-01 5.93414187e-01 -3.46830845e-01 3.49050671e-01
-2.89609551e-01 6.18806064e-01 -7.48492718e-01 5.97855687e-01
1.02745986e+00 -2.75220662e-01 -1.01688504e-01 -8.44878495e-01
-2.34759778e-01 1.76129758e-03 -1.07838309e+00 1.89022088e+00
-9.11822081e-01 3.09557915e-01 4.05023873e-01 -7.65848398e-01
1.17752790e+00 8.50502774e-02 3.63980323e-01 -7.85216272e-01
2.65230745e-01 4.63165075e-01 -3.76017541e-01 -3.38071108e-01
8.37757364e-02 -8.08019564e-02 1.51083022e-01 2.66808838e-01
1.43711314e-01 6.16076104e-02 -2.31630895e-02 2.37424552e-01
7.97871351e-01 3.16787988e-01 2.01060444e-01 1.87854059e-02
1.10655463e+00 -4.51650679e-01 6.46336198e-01 4.81788069e-01
-1.91332027e-01 2.29878500e-01 3.30906034e-01 4.58460599e-02
-7.65072346e-01 -9.91877377e-01 -2.43216470e-01 6.75617933e-01
7.88661182e-01 -3.03493023e-01 -8.44647765e-01 -9.97749150e-01
-3.39802772e-01 6.53354526e-01 -2.73486376e-01 -3.94672364e-01
-4.93653238e-01 -6.36666179e-01 2.46797860e-01 5.38217485e-01
1.35614669e+00 -1.21531987e+00 -4.47570056e-01 2.97541529e-01
-2.28007272e-01 -1.18762898e+00 -3.74969333e-01 1.02751747e-01
-7.37120032e-01 -8.73760223e-01 -6.23754740e-01 -4.57811803e-01
5.97550869e-01 2.84626871e-01 7.90404618e-01 1.16699494e-01
-2.61562355e-02 1.07144739e-03 -4.00076658e-01 -3.57755348e-02
-1.45341814e-01 7.44349286e-02 -5.27230084e-01 2.59375334e-01
2.63356566e-01 -5.91080844e-01 -9.17310834e-01 2.40900978e-01
-1.07780218e+00 4.77012694e-01 1.13493812e+00 1.05168688e+00
4.59120333e-01 3.17844301e-01 7.25965083e-01 -7.97344208e-01
2.46481538e-01 -2.78125823e-01 -2.77507007e-01 3.07966471e-01
-6.56947374e-01 4.17878106e-02 9.31199908e-01 9.16789472e-03
-1.47777879e+00 -2.65983641e-02 -5.12083471e-01 -4.71744776e-01
-2.31730729e-01 2.48719051e-01 -7.32536495e-01 -2.82550424e-01
-2.21450347e-02 4.48594600e-01 1.84907317e-02 -5.73070109e-01
4.31481421e-01 8.89333069e-01 4.07642692e-01 -3.08958948e-01
7.12355554e-01 5.70928931e-01 1.12088196e-01 -6.71197712e-01
-1.16920149e+00 -4.60588753e-01 -5.87057769e-01 -3.47089946e-01
7.73859680e-01 -9.93588984e-01 -5.94845116e-01 7.70317614e-01
-1.16088533e+00 -1.42794162e-01 -3.98146957e-01 3.95147651e-01
-4.10751611e-01 3.16210300e-01 -6.18766665e-01 -7.16791391e-01
-3.69543254e-01 -1.40748167e+00 1.29060590e+00 5.41005194e-01
4.12739038e-01 -1.06183267e+00 -6.73736811e-01 5.18877745e-01
4.55830961e-01 2.38462493e-01 8.61754537e-01 -2.16136649e-01
-7.47655213e-01 1.58742711e-01 -9.44769621e-01 8.21530938e-01
2.07657605e-01 -4.04415786e-01 -1.18091393e+00 -2.44457170e-01
2.03981563e-01 -4.12315667e-01 1.16599643e+00 2.88250506e-01
1.27534604e+00 -1.19962439e-01 -2.89325207e-01 8.30968916e-01
1.67566526e+00 5.02944067e-02 7.39633203e-01 3.92664611e-01
8.76349509e-01 4.93386894e-01 8.26497495e-01 3.17510545e-01
4.55307961e-01 6.24731123e-01 5.69017947e-01 -6.88520730e-01
-5.89026749e-01 -3.20299089e-01 4.35142875e-01 9.39200759e-01
2.13596269e-01 -3.14870954e-01 -2.42332593e-01 1.55435950e-01
-1.72138131e+00 -4.55064505e-01 -3.38573568e-02 1.84252322e+00
8.39023650e-01 1.50277391e-01 -4.06839043e-01 2.11332276e-01
6.81881428e-01 6.56188309e-01 -8.25234771e-01 -1.10928990e-01
-2.31877670e-01 3.44626844e-01 3.87177140e-01 2.51970053e-01
-1.05405796e+00 9.57801580e-01 4.36818743e+00 1.39542639e+00
-1.17573535e+00 2.63236582e-01 5.41501701e-01 3.02541167e-01
-6.48091912e-01 3.88491154e-02 -6.87961996e-01 6.40969396e-01
3.85887235e-01 3.77312630e-01 3.29670995e-01 5.82686424e-01
3.87529522e-01 -2.84115225e-01 -3.98491800e-01 7.94020057e-01
-1.06755253e-02 -9.92299676e-01 4.00293246e-02 -1.76933438e-01
5.83875537e-01 -3.65869075e-01 -8.00570920e-02 1.92573547e-01
-3.33456725e-01 -7.20185995e-01 5.21923482e-01 6.43719673e-01
1.20462370e+00 -7.97649920e-01 9.83181775e-01 4.22926426e-01
-1.50310385e+00 -1.90562442e-01 -2.76510984e-01 3.26375782e-01
2.94493079e-01 1.06859744e+00 -3.13183963e-01 1.00065160e+00
4.05611604e-01 1.10050678e+00 -4.96590018e-01 8.28069687e-01
-8.91254485e-01 5.80879807e-01 -2.42549106e-01 3.63482863e-01
5.22201657e-01 -3.46076608e-01 5.99321365e-01 9.96652246e-01
6.13370501e-02 9.52187926e-04 2.63474077e-01 1.15888047e+00
-2.74484843e-01 2.16198757e-01 -1.52065992e-01 -1.82497185e-02
4.07506138e-01 1.57381725e+00 -7.64054596e-01 -2.33498141e-01
-3.99829686e-01 1.13959336e+00 2.33199239e-01 4.14228022e-01
-7.87525177e-01 -7.15739608e-01 4.21437025e-01 5.38187698e-02
1.87625304e-01 5.46624288e-02 -5.29613346e-02 -1.09978747e+00
1.73410773e-01 -5.07495284e-01 4.50591482e-02 -9.97057498e-01
-1.13097620e+00 7.01021731e-01 -3.33821476e-01 -1.25860083e+00
2.71193445e-01 -4.06838894e-01 -5.46713769e-01 1.20526206e+00
-2.23413539e+00 -1.57896936e+00 -5.88299692e-01 4.28098798e-01
5.65887213e-01 1.25541642e-01 3.65754843e-01 4.46058750e-01
-7.66924262e-01 5.63655257e-01 -1.29270047e-01 -3.29428315e-02
6.36864603e-01 -1.04321706e+00 -1.52185500e-01 8.69443774e-01
-1.58301130e-01 4.59046781e-01 1.29231796e-01 -6.55716658e-01
-1.07328522e+00 -1.40502632e+00 4.82549638e-01 2.61992455e-01
3.80934775e-01 -2.22445101e-01 -8.33719552e-01 1.31976798e-01
-4.09292290e-03 1.06342502e-01 2.00557217e-01 -4.31390136e-01
6.88287839e-02 -3.51663619e-01 -1.18896103e+00 3.75602990e-01
1.19770408e+00 -6.30916595e-01 -3.03994954e-01 -7.05902949e-02
1.00443959e+00 -1.99015781e-01 -8.02375674e-01 8.18614483e-01
5.01562476e-01 -1.21489513e+00 1.03743243e+00 1.13665402e-01
6.90972507e-01 -5.13488472e-01 2.37318184e-02 -1.15655196e+00
-7.54223093e-02 -2.74783045e-01 -9.63775069e-02 1.53222108e+00
-3.70922014e-02 -6.63898647e-01 7.08915055e-01 9.06624869e-02
-2.94923186e-01 -1.16145349e+00 -6.00811958e-01 -7.33608603e-01
-4.07165408e-01 -4.71371770e-01 7.44398594e-01 5.80749452e-01
-7.85339117e-01 3.06293190e-01 -2.90780574e-01 1.47409007e-01
7.96912670e-01 3.68088722e-01 3.82603824e-01 -1.12060499e+00
-1.21060312e-01 -2.82045484e-01 -2.14699760e-01 -1.38658404e+00
4.38538998e-01 -9.67881322e-01 6.69875965e-02 -1.79218817e+00
2.18465060e-01 -9.40885961e-01 -6.08142257e-01 3.64376336e-01
-3.76442134e-01 4.00772780e-01 3.50496233e-01 1.50422901e-01
-6.06260061e-01 9.86736238e-01 1.86835325e+00 4.84273285e-02
-2.59976715e-01 -1.20958619e-01 -7.01958418e-01 9.54986155e-01
6.38583004e-01 -2.92421222e-01 -5.69629848e-01 -1.70545861e-01
-6.27097040e-02 2.83346176e-02 6.57713771e-01 -9.82045650e-01
2.50092775e-01 4.93137240e-02 5.26852131e-01 -6.63474858e-01
3.47129315e-01 -8.64816248e-01 -1.93709329e-01 1.92441091e-01
-4.84388284e-02 -7.59433627e-01 -9.13186669e-02 4.84826386e-01
-5.61727762e-01 -2.30718657e-01 8.81209970e-01 -1.59493327e-01
-9.82177198e-01 4.24121201e-01 3.30535471e-01 -1.55994847e-01
8.74424100e-01 -2.70921707e-01 -3.82075608e-01 -1.20489150e-01
-5.36961973e-01 3.34961653e-01 5.55898964e-01 1.97239682e-01
6.90328360e-01 -1.23586893e+00 -1.48796380e-01 5.30810952e-01
1.75096486e-02 1.72113612e-01 4.98223990e-01 1.20670795e+00
-2.00298667e-01 2.48454779e-01 -5.19270413e-02 -3.52430135e-01
-7.52275348e-01 3.96413326e-01 9.03263092e-02 -5.38825035e-01
-5.06104589e-01 8.29981744e-01 4.01692837e-01 -4.35376614e-01
2.21909434e-02 -2.47117400e-01 -3.99576217e-01 6.14703484e-02
3.07414323e-01 3.88279736e-01 -8.81278366e-02 -6.34238780e-01
-2.28388399e-01 7.20546722e-01 2.81118840e-01 1.56690836e-01
1.34728730e+00 -5.02451062e-01 -3.22207183e-01 2.73338437e-01
1.41773856e+00 1.57830894e-01 -1.44934058e+00 -3.13887656e-01
-3.64116222e-01 -4.29446608e-01 5.33967614e-01 -8.06545734e-01
-1.38504612e+00 1.11899745e+00 5.26066720e-01 -3.11342239e-01
1.76638353e+00 -9.70427413e-03 1.39377952e+00 -1.37313783e-01
4.93232459e-01 -9.70173180e-01 1.37708098e-01 1.82981696e-02
6.46911025e-01 -1.31317568e+00 -1.26008376e-01 -7.66573966e-01
-5.50812066e-01 1.11829817e+00 6.36955500e-01 8.23733658e-02
5.63303530e-01 1.57214031e-01 -2.01974154e-01 -1.52868286e-01
-2.92201668e-01 -3.99188876e-01 4.38825607e-01 2.93904424e-01
9.19372663e-02 -1.88913465e-01 -3.91896605e-01 8.60576332e-01
2.16877371e-01 2.23466262e-01 1.27835944e-01 5.53012908e-01
-4.60949063e-01 -1.22534120e+00 -1.28894933e-02 5.68586588e-01
-6.89972416e-02 -2.73792863e-01 6.27434850e-02 6.70358777e-01
7.89313912e-01 8.46127331e-01 -2.60927886e-01 -4.02372926e-01
1.99346974e-01 -2.30816782e-01 2.32300252e-01 -5.01316965e-01
-4.92923021e-01 1.69370711e-01 -7.36782178e-02 -8.66629243e-01
-7.58655012e-01 -2.66112655e-01 -1.40106511e+00 1.72798529e-01
-6.25767648e-01 -4.09331620e-02 6.11581028e-01 1.02754176e+00
1.48051247e-01 7.54436433e-01 1.00129628e+00 -7.24453449e-01
-2.80612618e-01 -7.21549690e-01 -7.97281384e-01 3.59604508e-01
2.56472826e-01 -8.56040180e-01 -4.71028715e-01 -2.63207376e-01] | [9.660361289978027, -0.9963712692260742] |
b8e79b6a-86b4-4c47-aa60-4329dc5ca4c5 | a-computation-and-communication-efficient | 2205.15580 | null | https://arxiv.org/abs/2205.15580v2 | https://arxiv.org/pdf/2205.15580v2.pdf | A Computation and Communication Efficient Method for Distributed Nonconvex Problems in the Partial Participation Setting | We present a new method that includes three key components of distributed optimization and federated learning: variance reduction of stochastic gradients, compressed communication, and partial participation. We prove that the new method has optimal oracle complexity and state-of-the-art communication complexity in the partial participation setting. Moreover, we observe that "1 + 1 + 1 is not 3": by mixing variance reduction of stochastic gradients with compressed communication and partial participation, we do not obtain a fully synergetic effect. We explain the nature of this phenomenon, argue that this is to be expected, and propose possible workarounds. | ['Peter Richtárik', 'Alexander Tyurin'] | 2022-05-31 | null | null | null | null | ['distributed-optimization'] | ['methodology'] | [-1.61561936e-01 1.57203898e-01 -3.35457832e-01 -1.61788315e-01
-1.27423751e+00 -9.60423231e-01 3.32225233e-01 2.83184111e-01
-8.19869459e-01 8.01073432e-01 4.85420048e-01 -6.44702733e-01
-5.25654733e-01 -6.84404492e-01 -1.06308341e+00 -8.81386280e-01
-4.08674657e-01 4.18467075e-01 -1.72282368e-01 1.77482888e-02
1.23784607e-02 4.33823884e-01 -1.04206693e+00 1.52531296e-01
5.94801426e-01 8.31443787e-01 -3.67537588e-01 1.09603965e+00
-2.00416803e-01 9.59340036e-01 -6.14639342e-01 -5.88327229e-01
8.94905210e-01 -5.18999696e-01 -9.52568829e-01 -3.25761735e-02
1.78099230e-01 -5.89262426e-01 -4.74273860e-01 1.07870257e+00
6.43063903e-01 1.17875926e-01 1.16359547e-01 -1.06538808e+00
-2.14664966e-01 1.35572195e+00 -3.52339059e-01 3.15892100e-01
1.50264159e-01 8.55876952e-02 1.06028140e+00 -4.23042029e-01
7.95772731e-01 1.11615038e+00 6.08236551e-01 3.48557264e-01
-1.23499310e+00 -2.75633633e-01 1.05033405e-01 -2.52692997e-01
-1.10773408e+00 -4.44031268e-01 4.69381362e-01 -1.19432390e-01
6.71423554e-01 6.27133310e-01 3.61447304e-01 5.77108145e-01
-1.82807654e-01 9.20155466e-01 1.30679357e+00 -4.68211055e-01
4.22871232e-01 4.30641681e-01 -8.76507312e-02 7.93760180e-01
6.20735407e-01 2.08760187e-01 -7.46099353e-01 -6.91706419e-01
2.08353072e-01 1.90810069e-01 -5.92015982e-01 -5.00833333e-01
-9.79774117e-01 9.29110050e-01 9.06714797e-02 1.87407032e-01
-4.52419698e-01 4.55634862e-01 4.48453188e-01 7.94085085e-01
4.30199593e-01 2.37641826e-01 -5.30229390e-01 -4.07293737e-01
-8.38250220e-01 1.57058492e-01 1.49050176e+00 9.07772958e-01
8.56871068e-01 -2.25076482e-01 -6.54379427e-02 3.14737350e-01
-1.06739745e-01 7.13222384e-01 1.43877715e-01 -1.38626039e+00
6.61535203e-01 2.56506145e-01 3.22988123e-01 -6.64565146e-01
-2.99048305e-01 -7.79387414e-01 -4.57304507e-01 -6.72202334e-02
6.49923623e-01 -7.58983195e-01 1.21854983e-01 1.90728855e+00
5.04474998e-01 -3.32298517e-01 1.41365200e-01 9.54379320e-01
-7.71482056e-03 2.98005283e-01 -7.30116665e-01 -3.98763448e-01
6.85182393e-01 -1.16617548e+00 -5.19144535e-01 4.89728972e-02
9.71181273e-01 -3.39974046e-01 7.22799003e-01 1.38852969e-01
-1.36227274e+00 2.92741239e-01 -8.87550056e-01 2.09402159e-01
-8.41764063e-02 -3.88615102e-01 1.05136323e+00 1.10979402e+00
-1.10080302e+00 9.82889295e-01 -9.34907615e-01 -7.93919265e-02
3.11863571e-01 4.22471106e-01 -2.56368935e-01 -2.32816890e-01
-4.20860529e-01 4.27356273e-01 -1.15553826e-01 -2.66339600e-01
-9.68927979e-01 -4.93974745e-01 -2.63353854e-01 3.16782147e-01
6.48868144e-01 -8.29624116e-01 1.11682642e+00 -5.68760753e-01
-1.30810761e+00 5.23058176e-01 -2.54538685e-01 -7.27256656e-01
9.83404100e-01 -1.34952351e-01 3.26150030e-01 1.84976473e-01
-2.42292419e-01 -3.50332916e-01 3.05476278e-01 -1.00723720e+00
-7.87550747e-01 -7.17106879e-01 2.15976715e-01 3.06267232e-01
-7.44956434e-01 -6.86499029e-02 -1.85945913e-01 -1.17732249e-01
4.49691899e-02 -8.47906828e-01 -5.11718512e-01 -7.67447576e-02
-3.16171050e-01 1.06482893e-01 2.48371661e-01 -3.00556242e-01
9.09290314e-01 -2.18449616e+00 -3.25473659e-02 4.11408544e-01
6.53996587e-01 -3.15581024e-01 -2.44791538e-01 7.58954585e-01
6.02824092e-01 4.91064042e-01 6.88675717e-02 -7.10691094e-01
4.04148698e-01 -2.88320659e-03 -1.77835003e-01 9.91796792e-01
-6.96476221e-01 7.10143626e-01 -9.10578609e-01 -7.98975751e-02
-2.42046893e-01 1.83097556e-01 -7.71923125e-01 1.07503429e-01
-1.03436783e-01 3.51060659e-01 -5.12037158e-01 5.74484408e-01
6.46538794e-01 -3.83933038e-01 6.59138918e-01 2.74893284e-01
-1.20110303e-01 4.06041205e-01 -1.52678633e+00 1.50070715e+00
-4.45314437e-01 3.22638065e-01 9.30659950e-01 -8.57694149e-01
2.99229085e-01 3.36382121e-01 6.60242796e-01 -2.48750165e-01
2.31407553e-01 6.02639496e-01 -3.65099847e-01 -3.14807832e-01
3.13954473e-01 1.18377529e-01 6.45673573e-02 1.25298345e+00
-6.30507544e-02 1.26391396e-01 -9.94734019e-02 4.72653121e-01
1.58975279e+00 -5.57288706e-01 -1.44132122e-01 -2.55806208e-01
1.18172675e-01 -3.23028713e-01 4.21430528e-01 1.44428158e+00
-3.49901140e-01 9.89885852e-02 8.37846816e-01 -2.22948015e-01
-9.19289231e-01 -7.30987728e-01 5.74731946e-01 1.38317037e+00
1.02893598e-01 -7.60414898e-01 -6.27499104e-01 -9.08080935e-01
1.65712044e-01 3.02574426e-01 -5.32495856e-01 1.67736158e-01
-1.36049390e-01 -7.04970241e-01 5.60244203e-01 1.39288306e-01
2.69784331e-01 -1.53304905e-01 -5.17351508e-01 -2.43323520e-02
-2.00717172e-04 -8.66269886e-01 -8.04276943e-01 4.25659806e-01
-1.03495383e+00 -1.21084762e+00 -2.89863646e-01 -2.06595749e-01
6.49538875e-01 5.79960585e-01 8.67271602e-01 2.07513377e-01
1.84762061e-01 7.83839166e-01 -1.21839240e-01 -1.32732540e-01
-3.70701253e-01 1.21487305e-01 -3.35392579e-02 -1.02610700e-01
-1.60929352e-01 -7.27954566e-01 -8.69945347e-01 -7.97749013e-02
-8.69888663e-01 -5.71797967e-01 5.96608877e-01 1.08515835e+00
3.27282876e-01 -8.53644162e-02 6.86582476e-02 -1.14482653e+00
8.79445493e-01 -5.26138067e-01 -7.78204620e-01 2.08945125e-01
-9.96786535e-01 2.63255715e-01 6.95911229e-01 -3.04298520e-01
-7.64716208e-01 -1.29633043e-02 2.53669880e-02 -2.32427537e-01
3.39371413e-01 3.67825955e-01 1.54931486e-01 -5.71749210e-01
6.65731192e-01 3.03811908e-01 1.77118957e-01 -6.36486113e-01
7.27305174e-01 6.74614549e-01 2.03456596e-01 -8.48852336e-01
5.48070967e-01 9.50342119e-01 -4.21824306e-02 -1.92678735e-01
-6.16760612e-01 -2.90130675e-01 1.76543683e-01 2.37188041e-01
-3.17688137e-02 -8.15173388e-01 -1.22960341e+00 9.72208083e-02
-1.08051002e+00 -3.35632950e-01 -9.54128683e-01 6.57441497e-01
-4.50211197e-01 4.52966899e-01 -8.83555949e-01 -1.15990460e+00
-7.54094779e-01 -1.04440773e+00 4.75000083e-01 4.75820638e-02
4.12679732e-01 -9.77836013e-01 1.10013083e-01 4.54916924e-01
1.06655657e+00 -8.22677370e-03 4.88545388e-01 -1.07855117e+00
-8.25486839e-01 -3.59549403e-01 -9.75022689e-02 2.74915129e-01
-2.78078496e-01 -4.65944707e-01 -6.23515904e-01 -8.46812248e-01
2.80453712e-01 -4.89099860e-01 6.88331246e-01 2.36206241e-02
1.02430081e+00 -9.70259309e-01 -7.10389391e-03 9.83533740e-01
1.73297060e+00 -3.23667258e-01 -1.98563077e-02 2.36311048e-01
2.93299824e-01 4.40996289e-01 1.24609523e-01 1.22334039e+00
5.63247144e-01 2.25154206e-01 5.43354750e-01 -2.87081264e-02
2.02622548e-01 -2.08199546e-01 3.30663770e-01 8.59665155e-01
1.31032854e-01 -7.88343474e-02 -5.18588662e-01 6.90701246e-01
-1.85979998e+00 -7.63975263e-01 9.28154588e-02 2.44611955e+00
8.11972022e-01 -1.17158383e-01 2.89973348e-01 -6.06633089e-02
5.02807677e-01 1.21238552e-01 -4.55299497e-01 -4.87375855e-01
-3.52941990e-01 2.14693695e-01 1.42878497e+00 6.54796004e-01
-4.39530939e-01 5.77604055e-01 7.62786150e+00 6.19878411e-01
-1.02640307e+00 7.58859873e-01 6.18091166e-01 -6.28603101e-01
-7.37790644e-01 3.19470197e-01 -5.78367174e-01 4.20215458e-01
8.40075374e-01 -4.07998323e-01 1.07916570e+00 1.06527030e+00
-2.24051341e-01 5.09131402e-02 -1.08263302e+00 8.96816194e-01
-1.78655028e-01 -1.49813163e+00 -2.80426592e-01 4.56190944e-01
1.04852688e+00 5.15349030e-01 -2.13864684e-01 -7.21413316e-03
6.20293498e-01 -7.38367915e-01 4.86688107e-01 1.25377163e-01
4.21528161e-01 -8.25393558e-01 8.03035438e-01 5.67069113e-01
-6.30258203e-01 -6.05003834e-01 -2.03703195e-01 -1.93334639e-01
8.82617608e-02 7.84388304e-01 -4.64696199e-01 5.78954101e-01
5.76243758e-01 -3.24534625e-01 -3.48929375e-01 1.06061268e+00
-4.15185578e-02 6.66665792e-01 -8.57871234e-01 -2.77296513e-01
3.78259979e-02 -1.85335070e-01 7.22733140e-01 9.98333275e-01
3.07531744e-01 -8.17329213e-02 2.49002323e-01 5.72167218e-01
-5.88604569e-01 2.22050279e-01 -5.24202287e-01 -2.92522222e-01
8.93037021e-01 1.02447724e+00 -3.73977065e-01 -3.92602235e-01
-4.01124984e-01 9.34933782e-01 4.42336172e-01 2.22142503e-01
-5.24867475e-01 -4.49116915e-01 5.03917873e-01 -1.63891718e-01
5.48682332e-01 -1.95750505e-01 -4.19349849e-01 -1.25070405e+00
2.76874304e-01 -9.79413152e-01 5.03313184e-01 2.86965609e-01
-1.12494302e+00 3.98929387e-01 -4.44139600e-01 -3.30393881e-01
-9.68446583e-03 -5.56143895e-02 -4.51400012e-01 5.38560152e-01
-1.34699106e+00 -7.05526710e-01 -5.30047752e-02 8.74310315e-01
-1.29547462e-01 2.23286692e-02 7.28717506e-01 4.67958897e-01
-3.20763230e-01 1.00639296e+00 7.25959778e-01 -1.44357875e-01
4.79311973e-01 -1.27017212e+00 1.07834497e-02 1.06085718e+00
1.22912750e-01 6.55099094e-01 8.09473634e-01 -1.77023247e-01
-2.36624384e+00 -6.60652637e-01 7.77943254e-01 -2.72685409e-01
6.31332099e-01 -5.18155873e-01 -1.56397924e-01 6.45074010e-01
9.51504782e-02 3.49448740e-01 8.15200448e-01 3.69506031e-01
-4.80648100e-01 -4.83237356e-01 -1.39268339e+00 3.96755397e-01
1.28829229e+00 -6.99387074e-01 9.11310837e-02 9.01307642e-01
7.49814630e-01 -4.98331547e-01 -8.21281731e-01 4.75520603e-02
4.19370174e-01 -1.14992213e+00 5.64956665e-01 -4.15542573e-01
-2.31213495e-01 2.28794679e-01 -5.22410154e-01 -1.13888323e+00
8.72153491e-02 -1.36859572e+00 -6.01855814e-01 8.98374856e-01
4.54226524e-01 -1.22239137e+00 1.19481182e+00 7.70846426e-01
2.26868719e-01 -8.82499754e-01 -1.06146049e+00 -9.66640651e-01
1.91134274e-01 -3.12478185e-01 8.34574103e-01 7.60756075e-01
2.92514622e-01 -1.50530953e-02 -5.01083970e-01 1.34391999e-02
8.08430552e-01 3.50641251e-01 1.03356099e+00 -5.02607703e-01
-1.04940569e+00 -3.07349801e-01 1.04620963e-01 -1.39270902e+00
-2.67791390e-01 -7.13395417e-01 -3.45897555e-01 -1.07202625e+00
5.59948504e-01 -5.03300071e-01 -4.16288286e-01 3.20479602e-01
1.52415767e-01 -1.27983376e-01 4.65349823e-01 3.10839146e-01
-1.10443115e+00 3.75861257e-01 9.73639727e-01 -1.53451711e-02
-2.64736354e-01 7.23642632e-02 -1.20390165e+00 1.31113514e-01
6.33687019e-01 -7.10421264e-01 -4.31376398e-02 -6.14835262e-01
4.01248217e-01 3.58200163e-01 5.21217426e-03 -5.55362105e-01
7.78456032e-01 -1.22576758e-01 -2.62516826e-01 -1.60361752e-01
1.19879380e-01 -5.96042275e-01 6.38705343e-02 6.44189477e-01
-4.34736639e-01 1.13181680e-01 -3.29603374e-01 7.73667753e-01
1.51997238e-01 -5.57049736e-02 3.49497825e-01 -2.73636401e-01
3.16920429e-01 3.51712763e-01 -3.60376209e-01 2.50713468e-01
6.33945942e-01 3.12455982e-01 -4.98909891e-01 -7.87756979e-01
-3.91372770e-01 4.65811700e-01 6.69204593e-01 -2.77306944e-01
2.10275948e-01 -8.94636333e-01 -8.76624584e-01 -6.32104129e-02
-4.87584591e-01 -2.28103042e-01 1.61937684e-01 1.26517189e+00
-2.74113446e-01 2.48089895e-01 5.27884722e-01 -2.40939870e-01
-1.19539881e+00 4.85572636e-01 4.10976827e-01 -4.95094508e-01
-4.40934092e-01 1.00065839e+00 -4.63401198e-01 -6.58089042e-01
7.09476471e-01 2.97003686e-01 8.81772518e-01 -4.95019495e-01
4.43377972e-01 6.17466688e-01 5.00952154e-02 2.50354856e-01
-3.88854831e-01 -1.30781174e-01 6.64343759e-02 -4.39523965e-01
1.37634778e+00 -2.99171209e-01 -3.21457654e-01 1.48234606e-01
1.41596389e+00 6.51241481e-01 -1.17445707e+00 -5.19931078e-01
-1.92330599e-01 -7.11678982e-01 -1.96475722e-02 -7.41324127e-01
-1.48255491e+00 5.85529268e-01 5.69055080e-01 2.30839282e-01
1.03075218e+00 7.32042119e-02 8.99652362e-01 7.88428724e-01
7.02828705e-01 -1.10516536e+00 -3.41188341e-01 4.04854000e-01
4.02883172e-01 -1.00937510e+00 1.67359561e-01 -1.59602866e-01
-1.04424268e-01 6.64582074e-01 6.08470328e-02 -7.73781389e-02
6.60808265e-01 4.39917803e-01 -1.96455419e-01 -6.21581785e-02
-1.16467249e+00 -6.17483966e-02 -5.02288222e-01 2.48321652e-01
1.48614764e-01 4.50367071e-02 -7.22895861e-01 7.21050382e-01
-1.23797089e-01 -2.06057608e-01 4.66362447e-01 1.28447104e+00
-2.33232886e-01 -1.23185658e+00 -4.15822595e-01 4.50220317e-01
-1.04297268e+00 -1.14229873e-01 -4.67630833e-01 3.53791773e-01
-1.63304850e-01 1.09584248e+00 -4.62800950e-01 -3.97599578e-01
-7.16253370e-02 -1.70206293e-01 4.17220950e-01 -2.27653846e-01
-9.48784709e-01 -2.80047487e-02 2.74918050e-01 -7.28765130e-01
-2.65278202e-03 -2.67300665e-01 -8.59710932e-01 -9.79624152e-01
-6.65489078e-01 6.99471891e-01 1.15062809e+00 8.14114511e-01
1.03591287e+00 8.21369812e-02 1.15902174e+00 -3.40654552e-01
-1.56399488e+00 -5.60278475e-01 -6.93033755e-01 2.53431529e-01
3.88804317e-01 2.48252839e-01 -1.09927952e+00 -7.17197776e-01] | [6.224489212036133, 4.996263027191162] |
ad035f51-78f6-4700-b7ef-d65306c22ad6 | can-deep-network-balance-copy-move-forgery | 2305.10247 | null | https://arxiv.org/abs/2305.10247v1 | https://arxiv.org/pdf/2305.10247v1.pdf | Can Deep Network Balance Copy-Move Forgery Detection and Distinguishment? | Copy-move forgery detection is a crucial research area within digital image forensics, as it focuses on identifying instances where objects in an image are duplicated and placed in different locations. The detection of such forgeries is particularly important in contexts where they can be exploited for malicious purposes. Recent years have witnessed an increased interest in distinguishing between the original and duplicated objects in copy-move forgeries, accompanied by the development of larger-scale datasets to facilitate this task. However, existing approaches to copy-move forgery detection and source/target differentiation often involve two separate steps or the design of individual end-to-end networks for each task. In this paper, we propose an innovative method that employs the transformer architecture in an end-to-end deep neural network. Our method aims to detect instances of copy-move forgery while simultaneously localizing the source and target regions. By utilizing this approach, we address the challenges posed by multi-object copy-move scenarios and report if there is a balance between the detection and differentiation tasks. To evaluate the performance of our proposed network, we conducted experiments on two publicly available copy-move datasets. The results and analysis aims to show the potential significance of our focus in balancing detection and distinguishment result and transferring the trained model in different datasets in the field. | ['Shizhen Chang'] | 2023-05-17 | null | null | null | null | ['image-forensics'] | ['computer-vision'] | [ 3.79333556e-01 -5.27786016e-01 3.34049314e-02 -4.26654220e-02
-1.21163428e+00 -6.53603971e-01 7.48971522e-01 6.05574157e-03
-1.63071245e-01 1.52797222e-01 -8.48917216e-02 -3.40023756e-01
2.99475323e-02 -4.59847897e-01 -6.25401378e-01 -8.64932179e-01
-3.12668495e-02 7.62034357e-02 4.35374349e-01 2.52218898e-02
8.60717177e-01 7.95332730e-01 -1.30287397e+00 6.54940188e-01
3.90978992e-01 1.07240891e+00 2.95777768e-01 6.20625913e-01
1.42823786e-01 7.71492898e-01 -1.06989753e+00 -6.95849717e-01
6.66890383e-01 -5.00869572e-01 -6.82659626e-01 3.22208822e-01
6.06981099e-01 -7.21693695e-01 -5.07589340e-01 1.20928490e+00
6.60019100e-01 -3.11190933e-02 2.97463685e-01 -1.39668107e+00
-6.61914289e-01 3.10400188e-01 -1.02334595e+00 9.65072572e-01
5.79291657e-02 4.80732769e-01 6.64548218e-01 -8.01560283e-01
5.12236297e-01 1.22009921e+00 6.18764937e-01 2.12709740e-01
-8.30001652e-01 -8.34056079e-01 -3.25385600e-01 5.97489655e-01
-1.39188015e+00 -6.66027188e-01 9.80444849e-01 -3.19007516e-01
7.90541470e-01 7.73112178e-02 2.56468989e-02 1.00351131e+00
3.46711785e-01 9.34410632e-01 9.49059248e-01 -4.03247774e-01
-1.17116958e-01 -1.38461381e-01 -1.51361926e-02 2.38242865e-01
3.41472864e-01 1.70232922e-01 -4.07796651e-01 -1.51517749e-01
6.00306451e-01 3.90183240e-01 -3.55782717e-01 -4.36952524e-02
-8.53404343e-01 8.06891799e-01 3.22307289e-01 4.90618438e-01
-3.70965719e-01 3.02471906e-01 6.38688862e-01 2.76836038e-01
2.22894296e-01 4.28418547e-01 2.26907402e-01 -4.30050828e-02
-1.51922131e+00 3.26618552e-01 1.65419251e-01 5.70728302e-01
4.37420338e-01 -2.69053310e-01 -2.67779619e-01 5.10195851e-01
4.33887867e-03 9.96235907e-02 5.38692236e-01 -5.75614035e-01
7.54103661e-01 5.42077899e-01 5.41220568e-02 -1.29479313e+00
2.21102987e-03 -4.25052166e-01 -5.11171699e-01 3.18204373e-01
4.25623983e-01 2.66800225e-01 -8.33352923e-01 1.30991066e+00
3.72122854e-01 1.92499995e-01 -2.94768810e-01 1.08421898e+00
2.28882745e-01 4.58629459e-01 -2.40137009e-03 3.67598593e-01
1.52648747e+00 -7.79052019e-01 -6.43147528e-01 -2.04700097e-01
7.33471394e-01 -8.79117489e-01 7.32344925e-01 3.48269254e-01
-1.08411038e+00 -3.66007745e-01 -1.07203114e+00 -6.06668927e-02
-4.89475727e-01 3.17430526e-01 1.12904675e-01 8.70022535e-01
-7.29852855e-01 6.32100642e-01 -6.53016269e-01 -7.11150765e-02
7.73973823e-01 1.30893990e-01 -3.17396075e-01 -1.31991267e-01
-1.01100802e+00 7.91446090e-01 4.90124553e-01 1.34036154e-01
-9.87570107e-01 -5.10536313e-01 -3.48037571e-01 3.16491991e-01
3.84134293e-01 4.36900966e-02 1.00354326e+00 -1.10077226e+00
-7.33874261e-01 1.15942609e+00 2.00230703e-01 -5.32671988e-01
7.51906574e-01 8.29259083e-02 -5.68856418e-01 4.81262267e-01
3.19976926e-01 1.78710163e-01 1.21059096e+00 -1.15891218e+00
-9.92701471e-01 -6.72491133e-01 -6.07698187e-02 -3.35767984e-01
-2.99007535e-01 6.14377260e-01 -4.36526150e-01 -9.48481798e-01
-1.27702042e-01 -4.84812558e-01 4.18967515e-01 2.19722033e-01
-5.46703577e-01 -1.15353122e-01 1.53446090e+00 -1.00220585e+00
1.14627242e+00 -2.43050027e+00 -4.06322718e-01 -1.37222221e-03
3.47744286e-01 7.83697307e-01 -1.89059660e-01 5.87607980e-01
-1.96737096e-01 1.26227900e-01 -3.58031541e-01 -4.18523669e-01
-1.67982683e-01 -2.19616920e-01 -7.40055561e-01 1.05031443e+00
6.02897644e-01 7.31271327e-01 -7.11855233e-01 -3.62379491e-01
-6.14027120e-02 2.58545965e-01 -2.18762115e-01 1.91904530e-01
8.52614716e-02 1.34864509e-01 -2.95720935e-01 9.09333646e-01
1.17548859e+00 -1.36222377e-01 -1.49609735e-02 -1.41105205e-01
6.73561841e-02 1.41963318e-01 -1.21378493e+00 1.21055210e+00
-1.12702109e-01 9.87392545e-01 3.15677017e-01 -1.17584836e+00
6.45897150e-01 1.09155238e-01 2.43896887e-01 -1.09335744e+00
1.92733184e-01 2.90560186e-01 4.30968478e-02 -7.77860940e-01
8.92718554e-01 -3.07484061e-01 2.09213234e-02 8.32274556e-01
-1.38794124e-01 4.73395795e-01 6.64680749e-02 2.34553277e-01
1.30507731e+00 -8.54916126e-02 -3.00632659e-02 1.80277690e-01
4.48474228e-01 -1.61773503e-01 3.44908893e-01 6.79647624e-01
-5.84322751e-01 7.90746152e-01 5.23114741e-01 -2.60599762e-01
-1.01872885e+00 -8.77029955e-01 -3.19556557e-02 9.25917983e-01
5.03444552e-01 5.75472191e-02 -6.31062627e-01 -8.64975393e-01
3.52843516e-02 7.45036066e-01 -6.08113170e-01 -3.15662205e-01
-8.96308303e-01 -7.58013308e-01 1.24021387e+00 4.27869380e-01
6.87838495e-01 -8.86791229e-01 -1.00253737e+00 9.14652124e-02
-3.48507285e-01 -1.32073462e+00 -6.78522289e-01 -8.66358951e-02
-6.01387084e-01 -1.48349619e+00 -6.84057415e-01 -5.92475891e-01
4.21446323e-01 9.22007978e-01 6.90983534e-01 5.19860268e-01
-6.05702698e-01 3.15790445e-01 -4.88223612e-01 -1.35070547e-01
-5.53124845e-01 -1.14482753e-01 -4.18919742e-01 3.61649007e-01
2.51358718e-01 -3.59899879e-01 -8.53748083e-01 5.58500946e-01
-1.56260335e+00 -4.81148869e-01 7.21285164e-01 4.61199105e-01
7.16015026e-02 3.39410663e-01 5.22508979e-01 -5.07504106e-01
7.66926944e-01 -7.32812166e-01 -3.88899654e-01 3.58318388e-01
-3.17518115e-01 -2.26197004e-01 5.23244083e-01 -3.55589390e-01
-8.82498205e-01 -2.47206435e-01 1.57648131e-01 -8.03995848e-01
-1.62098440e-04 9.23781022e-02 -2.67573446e-01 -2.50577867e-01
4.70672965e-01 6.47011399e-01 1.04974313e-02 -4.67724413e-01
2.35672206e-01 1.01035702e+00 7.92073309e-01 -3.41319263e-01
9.16478932e-01 7.50221133e-01 -1.62850499e-01 -5.94434202e-01
-1.72151148e-01 -7.81140327e-01 -5.63029110e-01 -1.69033304e-01
4.57204193e-01 -6.50785029e-01 -4.53940064e-01 8.63541484e-01
-1.32798624e+00 1.66626111e-01 8.76722187e-02 -1.33969069e-01
-1.92548931e-01 1.19692802e+00 -4.53781605e-01 -9.68133152e-01
-4.78352278e-01 -1.32551408e+00 1.37633955e+00 2.42374069e-03
2.24708498e-01 -7.21559107e-01 -2.30152383e-01 5.60619175e-01
5.36475718e-01 3.99587423e-01 7.76640117e-01 -9.80199993e-01
-8.05597723e-01 -6.97764337e-01 -5.55983007e-01 2.61424422e-01
8.09903666e-02 -2.41122440e-01 -9.84406650e-01 -5.63821375e-01
2.73237646e-01 -2.36817673e-01 9.18313801e-01 -1.08121976e-01
1.14109755e+00 -2.10224897e-01 -4.55516636e-01 4.51685578e-01
1.43966365e+00 1.51723787e-01 9.41275358e-01 5.77832222e-01
5.29968262e-01 6.31829977e-01 5.31768620e-01 4.36657757e-01
-5.08291870e-02 8.51139009e-01 7.68761218e-01 1.00562699e-01
-3.44925106e-01 -7.32567385e-02 4.21443760e-01 -4.05074470e-02
3.64296526e-01 -6.53003097e-01 -7.87650287e-01 8.73963416e-01
-1.45523691e+00 -1.34492171e+00 -8.44245329e-02 2.26256537e+00
3.66665214e-01 -3.16103995e-02 3.35038543e-01 4.24044818e-01
1.31228662e+00 3.51451308e-01 -4.95066792e-01 -2.09318370e-01
-1.52461797e-01 9.15118456e-02 5.58556557e-01 -6.92279264e-02
-1.18986213e+00 7.32594550e-01 5.59177446e+00 1.34266591e+00
-1.42661428e+00 2.27175012e-01 6.39209509e-01 -1.92654863e-01
3.26059282e-01 2.04522908e-02 -7.57497847e-01 9.27644074e-01
8.05527806e-01 2.40231931e-01 1.70283720e-01 6.20853603e-01
3.87769669e-01 -1.10611744e-01 -8.72920692e-01 1.01427221e+00
3.79518658e-01 -1.43633962e+00 -6.23401739e-02 2.37911299e-01
2.29914561e-01 -2.94921756e-01 2.41202593e-01 -1.59372436e-03
-2.71871805e-01 -8.38732958e-01 9.91878390e-01 -1.46875978e-01
6.63349748e-01 -6.99808121e-01 5.89447021e-01 3.85800034e-01
-9.50697064e-01 -2.22853333e-01 -2.66650677e-01 2.30428010e-01
3.14702928e-01 5.25362790e-01 -8.83616447e-01 4.99397784e-01
5.87320626e-01 3.46896380e-01 -5.68425357e-01 1.26999962e+00
-2.83517987e-02 5.89801073e-01 -1.47861964e-03 5.03545105e-01
3.57651800e-01 1.55979350e-01 6.44669592e-01 1.40995216e+00
4.19906557e-01 -2.37223908e-01 -2.25737959e-01 1.26833808e+00
-2.67047644e-01 -2.37525582e-01 -2.74106115e-01 -1.96423471e-01
4.68382508e-01 1.18051267e+00 -1.04719746e+00 -1.01877347e-01
-2.08359838e-01 1.34169114e+00 1.93463847e-01 -4.52044159e-02
-1.01020265e+00 -6.45682037e-01 5.92535257e-01 5.53083122e-01
9.15569901e-01 -2.62636662e-01 4.60990369e-02 -9.96777177e-01
3.49917233e-01 -9.67330158e-01 5.22468388e-01 -4.94178206e-01
-1.35355425e+00 2.36424923e-01 -7.69051313e-02 -1.32077217e+00
1.70137677e-02 -7.07172871e-01 -1.13004899e+00 7.79155433e-01
-1.63648188e+00 -1.19898748e+00 -2.83519059e-01 4.64509338e-01
5.46436310e-01 3.71224992e-02 8.19004104e-02 6.42279446e-01
-7.37722218e-01 9.76078033e-01 3.22747707e-01 5.82859933e-01
7.59328306e-01 -8.11728299e-01 6.80395722e-01 1.48237062e+00
-3.17032402e-03 6.16317809e-01 4.22747225e-01 -7.78835297e-01
-1.36931372e+00 -1.04736984e+00 5.13350368e-01 -3.70057255e-01
5.40797174e-01 -3.09030712e-01 -1.06255591e+00 4.49017733e-01
1.49792492e-01 2.25845531e-01 4.38938260e-01 -6.82592332e-01
-5.45454621e-01 2.40821630e-01 -1.60926819e+00 9.71214101e-02
7.59490967e-01 -5.50563037e-01 -3.80334347e-01 1.55507177e-01
1.19439065e-01 -1.56540066e-01 -4.14328396e-01 -4.92709205e-02
3.33313495e-01 -1.37380850e+00 1.09224820e+00 -4.84750450e-01
6.85844541e-01 -3.50163519e-01 -1.11838140e-01 -8.48316193e-01
-1.57867923e-01 -6.76388085e-01 -2.08270609e-01 1.39860976e+00
-2.27057859e-01 -5.15664876e-01 7.53932893e-01 2.31862769e-01
-7.07391649e-02 -6.22037411e-01 -1.29379320e+00 -9.39363539e-01
5.60890846e-02 -1.63629159e-01 5.35749257e-01 1.04042017e+00
-3.95403534e-01 -2.59405464e-01 -5.72211802e-01 3.80207241e-01
6.60830259e-01 3.29365402e-01 7.54829526e-01 -6.52655363e-01
-4.04682845e-01 -6.18605077e-01 -8.73727143e-01 -1.03451049e+00
-2.07453221e-01 -8.04346442e-01 -6.97639585e-02 -1.03828216e+00
3.57834727e-01 -2.74218827e-01 -1.15244806e-01 3.23662668e-01
-4.45225090e-01 5.19349873e-01 4.69943225e-01 6.35727763e-01
-5.02221882e-01 3.43494028e-01 8.70457649e-01 -2.00070992e-01
2.35483766e-01 2.98000555e-02 -6.91356361e-01 2.94756740e-01
6.58492744e-01 -7.70326972e-01 3.47404671e-03 -4.35910612e-01
-1.88124746e-01 -1.38003632e-01 8.90448034e-01 -1.01289678e+00
1.77323028e-01 8.33420604e-02 3.73866618e-01 -7.06709623e-01
2.03795314e-01 -6.80399597e-01 -2.35700488e-01 5.39874077e-01
-4.95896488e-02 2.99223334e-01 1.84199065e-01 8.49810183e-01
-2.17327505e-01 -4.02649969e-01 1.02041280e+00 -7.27085695e-02
-9.00409937e-01 -2.63018534e-02 -4.13142681e-01 1.06960461e-02
1.36424410e+00 -7.40837812e-01 -4.91257071e-01 -2.08934844e-01
-1.14232309e-01 -1.49168968e-01 5.36828339e-01 4.66209322e-01
7.75696039e-01 -1.05822277e+00 -7.28524864e-01 1.25505418e-01
1.68999329e-01 -4.38498467e-01 3.56678277e-01 8.38860452e-01
-6.59835100e-01 1.70728400e-01 -3.09798241e-01 -3.32378447e-01
-1.40048885e+00 7.91524112e-01 4.65777785e-01 -3.68645489e-01
-7.11508989e-01 5.83338499e-01 -2.78038792e-02 2.94668204e-03
-3.38419806e-04 1.06307343e-01 2.25096166e-01 1.68689489e-02
7.99768209e-01 8.07727933e-01 1.37343243e-01 -9.30874050e-01
-4.05912280e-01 2.44576201e-01 -4.07519132e-01 2.39728808e-01
1.28731644e+00 -1.60927549e-01 -4.22832929e-02 -3.43120188e-01
1.50162590e+00 -1.00443549e-02 -1.15243387e+00 -2.26328611e-01
1.90603048e-01 -1.06565392e+00 1.65731981e-01 -5.59342325e-01
-1.45804989e+00 9.79833663e-01 8.08593035e-01 1.90882251e-01
1.17906368e+00 -4.57711630e-02 1.26165283e+00 -5.95846251e-02
2.37057015e-01 -8.79164934e-01 3.26079071e-01 -1.04995474e-01
5.98402441e-01 -1.20623016e+00 1.18148632e-01 -2.73834407e-01
-4.10209596e-01 1.19047713e+00 3.11496258e-01 -2.81486273e-01
1.32137552e-01 1.38691232e-01 -8.55023339e-02 -4.33139086e-01
-1.74304903e-01 1.64142206e-01 -5.60671836e-02 5.72775483e-01
1.08855821e-01 -3.28592122e-01 -1.59022838e-01 3.01136494e-01
2.27307841e-01 -3.04556191e-01 6.71163142e-01 1.38321888e+00
-4.00533259e-01 -1.20644224e+00 -9.92502868e-01 3.27549726e-01
-8.71927679e-01 1.33265167e-01 -6.49660647e-01 7.94362366e-01
1.75397396e-01 1.02915931e+00 8.00465271e-02 -2.63350606e-01
-1.90638360e-02 -1.78460836e-01 3.69983912e-01 -1.57283083e-01
-8.25568736e-01 -2.18968809e-01 -3.22672755e-01 -5.61207354e-01
-1.81076646e-01 -7.84010708e-01 -1.01655006e+00 -3.46476585e-01
-5.86492360e-01 -1.77055940e-01 6.25740111e-01 8.67308199e-01
6.85169339e-01 2.57629097e-01 7.89898157e-01 -9.86873746e-01
-9.43577349e-01 -8.02620351e-01 -6.80948555e-01 8.16795230e-01
4.70714837e-01 -6.25390947e-01 -5.50007343e-01 -3.64429206e-02] | [12.351665496826172, 0.9808393716812134] |
1b6352f1-d74e-4866-a735-7189b065a6ab | use-of-transfer-learning-and-wavelet | 2103.03602 | null | https://arxiv.org/abs/2103.03602v1 | https://arxiv.org/pdf/2103.03602v1.pdf | Use of Transfer Learning and Wavelet Transform for Breast Cancer Detection | Breast cancer is one of the most common cause of deaths among women. Mammography is a widely used imaging modality that can be used for cancer detection in its early stages. Deep learning is widely used for the detection of cancerous masses in the images obtained via mammography. The need to improve accuracy remains constant due to the sensitive nature of the datasets so we introduce segmentation and wavelet transform to enhance the important features in the image scans. Our proposed system aids the radiologist in the screening phase of cancer detection by using a combination of segmentation and wavelet transforms as pre-processing augmentation that leads to transfer learning in neural networks. The proposed system with these pre-processing techniques significantly increases the accuracy of detection on Mini-MIAS. | ['Muhammad Bilal', 'Junaid Qadir', 'Muhammad Shahzad Younis', 'Ahmed Rasheed'] | 2021-03-05 | null | null | null | null | ['breast-cancer-detection', 'breast-cancer-detection'] | ['knowledge-base', 'medical'] | [ 4.28284049e-01 4.20886278e-02 -3.43715698e-01 -4.70094144e-01
-6.22908890e-01 3.01310439e-02 1.19403034e-01 3.69048089e-01
-8.67748260e-01 2.59559840e-01 -9.28686038e-02 -6.47557139e-01
5.76719232e-02 -1.17187965e+00 -4.92496371e-01 -6.37482762e-01
-1.25031129e-01 3.99958104e-01 5.47271430e-01 -2.14798734e-01
-5.74154221e-02 6.83292925e-01 -9.59366083e-01 6.81295872e-01
6.47055387e-01 1.05810690e+00 4.07361805e-01 9.17518973e-01
-2.45042995e-01 5.37943244e-01 -4.60390234e-03 -2.09375501e-01
1.88343853e-01 -4.56819743e-01 -5.84810317e-01 -1.43048977e-02
-5.78754097e-02 -5.61167836e-01 -4.00270224e-01 1.21455860e+00
4.91979092e-01 -2.06855148e-01 7.01504350e-01 -4.98127282e-01
-2.32339561e-01 4.03216511e-01 -9.57994103e-01 7.51191020e-01
-1.16525002e-01 -3.47705841e-01 4.54421677e-02 -7.92275786e-01
1.53880149e-01 1.17459846e+00 9.80212390e-01 6.38920546e-01
-4.86735702e-01 -7.34136879e-01 -7.83760011e-01 5.38044989e-01
-8.94169867e-01 -7.95073956e-02 5.23037493e-01 -5.40568903e-02
4.72606063e-01 4.29772228e-01 7.96572685e-01 6.16543710e-01
6.77694261e-01 6.59332156e-01 9.62649226e-01 -6.50742650e-01
-4.64110300e-02 1.75416112e-01 3.62965614e-01 9.75188494e-01
4.13448900e-01 1.12404920e-01 -2.19454691e-02 -7.54473060e-02
1.02668893e+00 4.05612230e-01 2.18035519e-01 1.63386226e-01
-6.94305182e-01 1.04102886e+00 8.24809432e-01 7.65660465e-01
-4.37902004e-01 6.97160885e-02 5.63073099e-01 3.12122907e-02
3.04822713e-01 -6.69089779e-02 -2.09554791e-01 2.58807480e-01
-9.44043815e-01 -2.85601646e-01 1.92645624e-01 2.11221397e-01
1.52980641e-01 -2.70031810e-01 -2.32355952e-01 6.77470207e-01
2.04936817e-01 3.82128596e-01 1.01969504e+00 -5.67682028e-01
-5.46974093e-02 9.34130847e-01 -2.00762302e-01 -9.06300604e-01
-6.99211359e-01 -6.34094119e-01 -1.13713658e+00 9.56387538e-03
1.94620103e-01 -9.45514739e-02 -1.64490294e+00 9.93515313e-01
4.26530987e-01 1.85814634e-01 1.49045482e-01 6.55901551e-01
1.00387442e+00 6.02287710e-01 3.22099298e-01 1.36695569e-02
1.73873234e+00 -6.33060336e-01 -7.39458203e-01 -1.08189516e-01
5.59146345e-01 -6.92023337e-01 6.12838984e-01 1.74626186e-01
-8.19812655e-01 -6.05754077e-01 -1.05046487e+00 -8.49452615e-02
-2.86274761e-01 4.78441745e-01 9.26903129e-01 9.28948760e-01
-6.86986804e-01 3.86817604e-01 -1.66517019e+00 -6.84459269e-01
8.46861482e-01 6.07584536e-01 -3.74542981e-01 -3.22369635e-01
-8.58197749e-01 1.00738871e+00 6.40731394e-01 4.87394519e-02
-5.62092304e-01 -6.50547922e-01 -7.53180563e-01 -1.99140962e-02
8.98523629e-02 -3.53633314e-01 1.25913334e+00 -1.07994068e+00
-9.93861973e-01 9.73497152e-01 -1.20418049e-01 -6.60562336e-01
6.02724969e-01 -7.34951720e-02 -1.90166339e-01 4.67649579e-01
-1.76995937e-02 5.80793500e-01 5.50998390e-01 -5.96613705e-01
-9.31376576e-01 -4.17924345e-01 -6.50340676e-01 1.70438305e-01
-4.77543801e-01 2.93320656e-01 -5.45597494e-01 -4.04466033e-01
4.61201102e-01 -7.99101412e-01 -3.93203616e-01 1.93338379e-01
7.64123946e-02 2.70037279e-02 1.33027220e+00 -1.10941625e+00
5.99353254e-01 -2.23845720e+00 -4.20611352e-01 2.81568527e-01
-3.58238593e-02 3.72935921e-01 3.06327671e-01 -2.28390500e-01
-7.78793097e-02 3.31794620e-02 -9.48042423e-02 1.13995470e-01
-8.89237285e-01 1.81642622e-01 5.41660249e-01 4.91311789e-01
4.06384081e-01 8.95693243e-01 -7.39041448e-01 -8.18700433e-01
4.41938996e-01 6.05777621e-01 -2.20534414e-01 4.18162756e-02
2.81437099e-01 6.46019280e-01 -6.60405576e-01 7.02896357e-01
6.90663755e-01 -1.25429899e-01 -1.24095283e-01 -4.41247940e-01
2.62205094e-01 -3.04259777e-01 -7.52156436e-01 1.34340227e+00
-1.39621645e-01 7.03952432e-01 9.33732763e-02 -1.48282874e+00
6.18235826e-01 4.83100057e-01 7.24687636e-01 -7.23942697e-01
5.53667545e-01 3.18680465e-01 4.92785841e-01 -8.91917348e-01
5.39888861e-03 -2.72709221e-01 2.99870342e-01 1.35173619e-01
-8.66077170e-02 1.36292130e-01 1.53042570e-01 -9.06654000e-02
1.11722386e+00 -3.49296242e-01 5.02887607e-01 -2.69897878e-01
4.74473894e-01 3.36530089e-01 5.64598441e-01 5.02761602e-01
-2.24865541e-01 4.81777847e-01 8.79676789e-02 -6.40902281e-01
-1.01383615e+00 -6.13246500e-01 -5.34015536e-01 6.92282379e-01
-1.26387894e-01 6.34557307e-01 -6.16904974e-01 -5.17463207e-01
-5.38384588e-03 2.24968478e-01 -1.08413470e+00 -2.38045320e-01
-7.57724881e-01 -1.20417535e+00 4.38848495e-01 7.33396888e-01
1.13404429e+00 -9.46031988e-01 -9.56320286e-01 1.11914888e-01
-4.60594967e-02 -8.72119308e-01 6.26803711e-02 3.77519816e-01
-1.35842133e+00 -1.14767241e+00 -1.12395334e+00 -1.17148018e+00
9.44260776e-01 2.76267856e-01 5.41504681e-01 3.89201999e-01
-9.21174586e-01 -7.09925359e-03 -2.64874965e-01 -8.42894197e-01
-8.31436992e-01 3.42022139e-03 -4.95046407e-01 -2.11775392e-01
5.42567790e-01 -2.76707821e-02 -8.54240298e-01 -1.01023711e-01
-9.48641002e-01 1.87119588e-01 1.05547178e+00 8.23486388e-01
5.48051119e-01 5.55509150e-01 2.56280720e-01 -1.20146906e+00
4.58877891e-01 -5.08605063e-01 -3.24592739e-01 1.53623834e-01
-7.84871653e-02 -1.94680437e-01 -3.81546542e-02 -4.24138993e-01
-1.05905032e+00 1.38504758e-01 -2.17441276e-01 1.76993355e-01
-1.10238634e-01 7.59149075e-01 3.77437979e-01 -4.64250445e-01
6.28337681e-01 1.16723225e-01 1.41911283e-01 -3.89618665e-01
-4.35047925e-01 6.37184381e-01 7.16177285e-01 1.35594815e-01
3.39238465e-01 5.07548273e-01 4.43411201e-01 -1.11644983e+00
-6.07828796e-01 -5.69854081e-01 -5.34450531e-01 -2.02972531e-01
1.03923166e+00 -8.10955524e-01 -3.33315879e-01 4.69669223e-01
-8.55790317e-01 3.54630277e-02 8.91746283e-02 6.42209888e-01
2.45120943e-01 3.25117081e-01 -1.02937543e+00 -5.91593325e-01
-7.52298117e-01 -1.03665745e+00 6.94521070e-01 6.35864377e-01
3.56602855e-02 -8.52631569e-01 -4.33729887e-01 2.63724834e-01
5.66857696e-01 5.21394074e-01 1.16078579e+00 -5.89326978e-01
-3.17546912e-02 -7.86661029e-01 -4.94739413e-01 2.01854661e-01
3.85998338e-01 -1.96914285e-01 -5.97390950e-01 -2.33199343e-01
2.36682236e-01 -4.02031727e-02 1.07566285e+00 1.04865980e+00
1.40533221e+00 3.61151814e-01 -6.85807884e-01 5.15351176e-01
1.41662383e+00 6.58141077e-01 4.76758808e-01 4.73087341e-01
2.66587168e-01 1.94490299e-01 4.76400554e-01 6.52568713e-02
-1.36423096e-01 -5.54194599e-02 3.12593579e-01 -7.38248825e-01
-1.35724276e-01 1.19786598e-01 -3.03642482e-01 4.61530477e-01
-2.60437012e-01 2.99298614e-01 -1.18516576e+00 4.80279028e-01
-1.51778924e+00 -6.67937994e-01 -3.44075590e-01 1.78133667e+00
8.20229232e-01 2.32088834e-01 -1.34096190e-01 5.31850934e-01
6.71082795e-01 -7.53733337e-01 -3.52385223e-01 -2.18526229e-01
4.14751053e-01 7.26869583e-01 9.40435588e-01 2.60033190e-01
-1.41050017e+00 4.31987584e-01 6.39955807e+00 5.88383138e-01
-1.60813785e+00 3.14101130e-01 1.08781207e+00 3.91321212e-01
4.42742616e-01 -6.40674472e-01 -4.72032368e-01 3.39246809e-01
7.37075686e-01 1.80239394e-01 -3.91050518e-01 7.77784467e-01
3.22317719e-01 -8.18282247e-01 -9.92402256e-01 6.28486574e-01
-3.03602964e-01 -1.24800706e+00 -1.68781564e-01 -2.28297085e-01
6.52083993e-01 -2.00514853e-01 2.43911430e-01 1.96516842e-01
-1.86947107e-01 -1.17706192e+00 -1.43047236e-02 3.47117245e-01
7.18230665e-01 -8.78501654e-01 1.37204528e+00 4.64195788e-01
-7.35371590e-01 -7.59553909e-02 -3.60876739e-01 5.43711185e-02
-3.36964875e-01 4.60565329e-01 -1.27955580e+00 1.21331088e-01
9.19193566e-01 1.14835843e-01 -5.60918331e-01 1.31261420e+00
3.20100904e-01 8.86378825e-01 -4.80368048e-01 -2.31849141e-02
3.24518800e-01 6.57675266e-02 -4.45973650e-02 1.36815906e+00
7.12060750e-01 3.75928164e-01 1.70146123e-01 3.88387591e-01
-1.02766911e-02 8.32801983e-02 -3.01613390e-01 1.06943570e-01
8.37261751e-02 1.23451245e+00 -1.48373580e+00 -4.77761090e-01
-2.96665311e-01 7.10069776e-01 -3.02682489e-01 -1.24076016e-01
-5.58540940e-01 -2.93557227e-01 -3.72932762e-01 2.63123035e-01
6.83801696e-02 -1.18529089e-02 -4.08056885e-01 -3.74318928e-01
-2.49191046e-01 -6.85395837e-01 4.56417322e-01 -2.18665019e-01
-6.51850462e-01 3.15118134e-01 1.00720860e-01 -7.84697473e-01
1.10298961e-01 -7.46121228e-01 -8.80840480e-01 7.31384099e-01
-1.59804058e+00 -1.31356394e+00 -7.72175014e-01 5.44226468e-01
7.38479495e-01 -1.43919677e-01 9.10454690e-01 4.97990727e-01
-4.12136674e-01 3.73198956e-01 1.26486138e-01 4.12794977e-01
5.25759041e-01 -1.12769067e+00 -6.51242957e-02 8.00717592e-01
-5.16064167e-01 2.49007672e-01 5.75201750e-01 -8.66374433e-01
-1.23622763e+00 -1.17450714e+00 3.48775089e-01 2.61584371e-01
1.50555000e-01 1.20650008e-01 -7.88114130e-01 4.35408264e-01
1.20064035e-01 4.93351072e-01 6.93964064e-01 -5.05751193e-01
6.05983734e-01 -7.63346627e-02 -1.55091381e+00 9.33177918e-02
3.19721587e-02 1.11513294e-01 -4.33009207e-01 4.26764548e-01
2.08714321e-01 -7.24842370e-01 -5.55873990e-01 7.56685138e-01
4.78075743e-01 -7.50164032e-01 8.26534867e-01 -2.97680229e-01
6.28042579e-01 3.48077148e-01 2.04039440e-01 -8.61864984e-01
-3.25248301e-01 1.55964747e-01 2.00055748e-01 6.38495386e-01
2.33473316e-01 -2.76294440e-01 1.28831995e+00 3.68863612e-01
2.06993535e-01 -6.61720693e-01 -8.26251209e-01 -7.94437900e-02
3.36577557e-02 -1.77131236e-01 3.91431414e-02 6.85138166e-01
-4.48654503e-01 -1.97619855e-01 2.16725230e-01 1.55129611e-01
6.81790352e-01 -3.83045256e-01 2.23016515e-01 -1.18878114e+00
-2.10747153e-01 -1.48607060e-01 -4.91429090e-01 -2.97672838e-01
-7.24577129e-01 -8.23736668e-01 -1.59689412e-01 -1.63354111e+00
7.39466906e-01 -2.85084188e-01 -4.26012754e-01 3.96619499e-01
-3.88350964e-01 5.78027368e-01 -2.31281146e-01 -7.57870614e-04
-1.94099434e-02 -1.42464101e-01 1.45607913e+00 -1.77924007e-01
-1.52366444e-01 4.55677629e-01 -4.98879641e-01 1.12592208e+00
8.23858559e-01 -6.05703354e-01 -2.92082988e-02 -2.85085678e-01
-1.12650603e-01 3.71060967e-01 2.64323860e-01 -1.26609242e+00
1.27949655e-01 7.38294199e-02 1.21895647e+00 -8.59546185e-01
2.52516299e-01 -1.03154659e+00 -2.70225089e-02 1.36873484e+00
-2.40635097e-01 -2.87313402e-01 6.47282124e-01 3.65465194e-01
-2.85607070e-01 -4.37689275e-01 1.06615412e+00 -5.19203424e-01
-6.81333899e-01 3.06008440e-02 -6.65142238e-01 -7.14186728e-01
1.24833906e+00 -4.23229814e-01 2.42685989e-01 -1.51984036e-01
-6.64405584e-01 1.69747218e-01 -2.46501088e-01 -5.51744830e-03
6.66209877e-01 -1.15429902e+00 -8.14347446e-01 1.61067784e-01
-3.46271783e-01 1.01476334e-01 5.43419905e-02 1.03583860e+00
-1.12073576e+00 3.08912605e-01 -5.54659426e-01 -6.09090507e-01
-1.81710649e+00 1.07367806e-01 4.28166628e-01 -5.15323520e-01
-7.80474961e-01 1.08108032e+00 -1.72581464e-01 1.73717231e-01
3.73865694e-01 -7.10626960e-01 -4.86968786e-01 -4.01435375e-01
5.23683190e-01 3.37538093e-01 2.39210293e-01 -2.55139589e-01
-2.26076275e-01 4.09516037e-01 -5.20028293e-01 1.30974650e-01
1.43896401e+00 3.10055882e-01 -7.35769197e-02 1.08234935e-01
1.14686680e+00 -4.52341318e-01 -6.97550476e-01 -2.05928504e-01
8.74682218e-02 -1.82068616e-01 7.89078295e-01 -7.66009510e-01
-1.18002093e+00 1.04560983e+00 1.46091342e+00 -1.00788083e-02
1.26918447e+00 -2.51359671e-01 8.66716146e-01 4.33343530e-01
-9.09958556e-02 -8.63383114e-01 7.45948032e-02 -7.75170512e-03
5.49573183e-01 -1.81954050e+00 2.48837382e-01 -5.35717249e-01
-2.51328021e-01 1.48704863e+00 4.23623860e-01 -4.60123479e-01
7.47028768e-01 5.51898956e-01 8.73858333e-02 -1.33344427e-01
7.66231259e-03 1.75453871e-01 3.43945473e-01 5.27522266e-01
7.92505741e-01 1.14946775e-01 -6.41831696e-01 5.85731030e-01
2.21792385e-01 2.87811905e-01 2.56335825e-01 1.38420177e+00
-8.62206519e-01 -9.31748986e-01 -8.38670373e-01 1.05743957e+00
-1.20266700e+00 2.69045651e-01 -2.29796499e-01 8.55834067e-01
3.44096154e-01 5.82968473e-01 1.06490344e-01 1.80889428e-01
-1.04218118e-01 -1.40535250e-01 6.02195144e-01 -5.89631855e-01
-4.96257395e-01 2.77519852e-01 -2.49525353e-01 -1.45116434e-01
-3.78261656e-01 -4.61173952e-01 -1.56210005e+00 8.71127695e-02
-3.59307975e-01 -6.76045120e-02 9.80101585e-01 9.52390552e-01
-2.93184847e-01 1.00145245e+00 2.30968714e-01 -6.60392463e-01
-3.58613670e-01 -1.21913362e+00 -3.40327173e-01 2.78306663e-01
3.36343795e-01 -4.65405107e-01 3.38267684e-01 2.64011294e-01] | [15.193923950195312, -2.5699000358581543] |
fc49298a-e63e-4401-a2b9-bef744ef3f13 | pcc-paraphrasing-with-bottom-k-sampling-and | 2208.08110 | null | https://arxiv.org/abs/2208.08110v3 | https://arxiv.org/pdf/2208.08110v3.pdf | PCC: Paraphrasing with Bottom-k Sampling and Cyclic Learning for Curriculum Data Augmentation | Curriculum Data Augmentation (CDA) improves neural models by presenting synthetic data with increasing difficulties from easy to hard. However, traditional CDA simply treats the ratio of word perturbation as the difficulty measure and goes through the curriculums only once. This paper presents \textbf{PCC}: \textbf{P}araphrasing with Bottom-k Sampling and \textbf{C}yclic Learning for \textbf{C}urriculum Data Augmentation, a novel CDA framework via paraphrasing, which exploits the textual paraphrase similarity as the curriculum difficulty measure. We propose a curriculum-aware paraphrase generation module composed of three units: a paraphrase candidate generator with bottom-k sampling, a filtering mechanism and a difficulty measure. We also propose a cyclic learning strategy that passes through the curriculums multiple times. The bottom-k sampling is proposed to generate super-hard instances for the later curriculums. Experimental results on few-shot text classification as well as dialogue generation indicate that PCC surpasses competitive baselines. Human evaluation and extensive case studies indicate that bottom-k sampling effectively generates super-hard instances, and PCC significantly improves the baseline dialogue agent. | ['Wai Lam', 'Hongyuan Lu'] | 2022-08-17 | null | null | null | null | ['paraphrase-generation', 'few-shot-text-classification', 'paraphrase-generation'] | ['computer-code', 'natural-language-processing', 'natural-language-processing'] | [ 5.25871992e-01 1.80441812e-01 -1.31898418e-01 -3.78869802e-01
-1.01332462e+00 -6.45978987e-01 8.90467227e-01 2.21409708e-01
-4.36850399e-01 8.69860113e-01 4.99980420e-01 -3.67536277e-01
1.22480445e-01 -8.18357766e-01 -5.92846930e-01 -6.28779054e-01
5.93030035e-01 7.69243181e-01 2.09094405e-01 -7.83079803e-01
4.31998610e-01 -2.36454815e-01 -1.49681616e+00 7.77362645e-01
1.49258006e+00 6.30818427e-01 4.40311700e-01 7.76943445e-01
-4.32336867e-01 7.52186239e-01 -1.20534074e+00 -6.63270712e-01
1.49513081e-01 -7.68431246e-01 -9.61442351e-01 1.46832615e-02
6.31123602e-01 -3.77528250e-01 -4.91366476e-01 9.30213213e-01
5.78116059e-01 4.54033434e-01 7.83471167e-01 -1.18786371e+00
-8.53748441e-01 1.05573821e+00 -5.19038141e-01 3.68797660e-01
5.24547935e-01 1.30020082e-01 1.01136792e+00 -1.08151138e+00
3.77023488e-01 1.32665694e+00 5.21986485e-01 8.58869910e-01
-1.29573667e+00 -7.80405283e-01 2.05485493e-01 3.63543779e-01
-8.60800028e-01 5.31240925e-02 8.78849566e-01 -2.24593773e-01
1.05156958e+00 3.29085231e-01 6.52704835e-01 1.54798925e+00
-6.28869701e-03 1.26864922e+00 1.07663476e+00 -5.08256555e-01
3.67614508e-01 1.90621018e-01 5.93640745e-01 2.99998969e-01
2.67717429e-02 -5.59821911e-02 -5.35507739e-01 -2.09006652e-01
4.26154315e-01 -1.83881193e-01 -2.38541543e-01 7.40602671e-04
-8.37338030e-01 1.17986035e+00 9.51181427e-02 -1.49057016e-01
1.32591069e-01 -1.95168138e-01 7.91274905e-01 7.38244951e-01
4.54154819e-01 9.85013664e-01 -3.49814773e-01 -2.29916796e-01
-8.51083159e-01 7.83798695e-01 8.49899828e-01 1.33903825e+00
2.85476774e-01 2.60863304e-01 -8.35629165e-01 1.50439954e+00
-4.03402805e-01 4.20728177e-01 1.13230062e+00 -7.39017367e-01
9.68881726e-01 6.37015879e-01 -5.07570105e-04 -5.31980693e-01
-8.40883479e-02 -2.58595616e-01 -8.84819746e-01 -2.21931398e-01
2.28006020e-01 -3.54091734e-01 -9.14407253e-01 1.78578568e+00
-6.16828986e-02 8.88266135e-03 2.35418350e-01 6.55461967e-01
1.12374437e+00 9.65976775e-01 -1.16995163e-02 -2.48095244e-01
1.34308887e+00 -1.42058015e+00 -7.58112073e-01 -3.72638851e-01
6.55615032e-01 -6.80952132e-01 1.70680654e+00 5.83350122e-01
-1.37429714e+00 -7.93196201e-01 -1.20274198e+00 -8.80814567e-02
-3.96860778e-01 1.56800464e-01 3.53164822e-01 6.74793065e-01
-7.72229970e-01 5.17104030e-01 -1.46978777e-02 -9.52821299e-02
5.36603630e-02 3.86558399e-02 1.92372575e-01 -1.03075191e-01
-1.70014584e+00 9.97033834e-01 6.56932652e-01 -4.85914141e-01
-9.07077610e-01 -6.77834332e-01 -1.18037224e+00 1.60845175e-01
3.39986742e-01 -6.50537193e-01 1.63085198e+00 -7.33111382e-01
-1.72370386e+00 5.80597401e-01 8.44182670e-02 -6.16616011e-01
5.32601416e-01 -3.28257263e-01 -1.02637976e-01 8.57319385e-02
8.11965987e-02 6.89008415e-01 8.99993420e-01 -1.03565955e+00
-5.39198279e-01 2.08269194e-01 2.03801572e-01 7.28758276e-01
-6.46689355e-01 -2.70614386e-01 -1.46892145e-01 -1.26147866e+00
-2.17643857e-01 -7.06801355e-01 -3.89982313e-02 -7.97622919e-01
-4.60827947e-01 -6.59500003e-01 6.07065678e-01 -4.89340484e-01
1.57670188e+00 -1.63176870e+00 5.08348703e-01 -7.96968937e-02
9.17238519e-02 5.51588535e-01 -4.74306762e-01 5.44750929e-01
-2.44804204e-01 -9.18653011e-02 -2.49963880e-01 -4.25401717e-01
-9.72621236e-03 1.55212775e-01 -6.58492506e-01 -2.49348462e-01
3.77789319e-01 8.50794673e-01 -1.24573684e+00 -3.21887612e-01
1.90966383e-01 -2.91357458e-01 -6.90324426e-01 4.98001754e-01
-4.43879426e-01 -1.33516103e-01 -1.85791403e-01 2.46073693e-01
3.19929272e-01 -1.21313199e-01 -1.90563112e-01 1.63049892e-01
1.57625988e-01 4.04389203e-01 -8.07147622e-01 1.75120044e+00
-5.24632633e-01 3.12286317e-01 -3.87638777e-01 -1.08253765e+00
1.21661651e+00 1.85420856e-01 -3.62507254e-02 -7.40220249e-01
2.04236925e-01 -5.61868399e-03 -8.48976746e-02 -5.67855120e-01
1.05485511e+00 -1.79857835e-02 -4.83486533e-01 5.75540066e-01
2.07003698e-01 -7.57080972e-01 5.56158364e-01 5.40878832e-01
9.70708072e-01 8.50896984e-02 4.70519438e-02 -2.34800741e-01
6.38769686e-01 1.94729030e-01 2.60630637e-01 1.09405172e+00
-8.20881277e-02 7.61830151e-01 6.12669528e-01 -1.56306416e-01
-1.43262827e+00 -9.85870659e-01 1.03346244e-01 1.49352241e+00
2.63961822e-01 -5.56010962e-01 -1.07384729e+00 -7.08643198e-01
-1.04538184e-02 1.15725768e+00 -8.04253697e-01 -7.64538050e-01
-5.60528755e-01 -7.95317113e-01 6.46294355e-01 6.16578162e-01
5.59232652e-01 -1.16408920e+00 -4.09765355e-02 2.20161021e-01
-5.01759171e-01 -8.66064966e-01 -8.50912869e-01 2.65670896e-01
-5.26734710e-01 -7.31628776e-01 -9.70108390e-01 -1.03041995e+00
5.72631955e-01 4.74403024e-01 1.09654891e+00 -9.57401916e-02
-1.70980632e-01 -1.04768962e-01 -6.68415964e-01 -4.26685363e-01
-8.16193879e-01 2.47524023e-01 -7.26166591e-02 -4.77854013e-01
4.21648324e-01 -3.98588061e-01 -4.90551531e-01 2.34123170e-01
-7.50011325e-01 3.86188030e-01 5.09327531e-01 1.35674822e+00
2.27577865e-01 -2.86771804e-01 1.04818153e+00 -1.01854670e+00
1.56497669e+00 -5.91155648e-01 -6.53745681e-02 2.73736775e-01
-5.96838415e-01 -4.67033610e-02 1.12323296e+00 -8.60820651e-01
-1.30409253e+00 -4.71729606e-01 -3.47747132e-02 -4.89325553e-01
-3.13831563e-03 3.16338122e-01 1.15821667e-01 5.08954465e-01
1.13479316e+00 7.06408620e-01 6.23920374e-02 -2.28678361e-01
5.01235127e-01 7.81206608e-01 4.99530405e-01 -8.68306756e-01
6.82539225e-01 -1.86145172e-01 -7.62277603e-01 -9.45454419e-01
-1.06765413e+00 -4.43440706e-01 -2.88836896e-01 -1.69372931e-01
4.53888804e-01 -8.82301152e-01 -3.46416384e-01 5.31828344e-01
-1.10645413e+00 -4.40650672e-01 -5.73316932e-01 1.52426539e-02
-6.46864414e-01 5.76629460e-01 -8.58875275e-01 -4.84840810e-01
-7.63988376e-01 -1.16912997e+00 9.64207292e-01 4.86984581e-01
-6.33686841e-01 -6.09480143e-01 7.09535554e-02 7.81504631e-01
2.88338751e-01 -1.25364900e-01 1.13200474e+00 -1.16595638e+00
5.01112789e-02 -2.09564686e-01 -1.30032435e-01 5.27010798e-01
-1.53459057e-01 -3.35386753e-01 -8.54494512e-01 -3.16328526e-01
1.04742907e-01 -1.02928817e+00 8.52060676e-01 1.35037467e-01
1.54351783e+00 -5.86965561e-01 1.55347735e-01 3.93865585e-01
7.42229879e-01 2.87314028e-01 4.99230653e-01 3.10701907e-01
5.09399235e-01 5.01613259e-01 6.70854509e-01 4.55685467e-01
3.73000473e-01 5.91793418e-01 -7.39994645e-02 5.04196808e-02
-2.22249568e-01 -3.75100166e-01 4.54714239e-01 1.17957473e+00
3.94621581e-01 -1.66050076e-01 -6.74863398e-01 5.39818287e-01
-1.67905927e+00 -1.08722079e+00 4.04190831e-02 2.02760816e+00
1.67308307e+00 4.57822144e-01 2.09613547e-01 2.25673094e-01
6.34697914e-01 8.70324299e-03 -6.68609858e-01 -8.92092645e-01
-5.15678078e-02 4.06795055e-01 -6.35388568e-02 5.62237442e-01
-7.92980134e-01 1.12063384e+00 5.87987661e+00 1.33808422e+00
-5.66160023e-01 -4.09930423e-02 5.96336126e-01 -2.71742046e-01
-4.99927223e-01 -3.52869660e-01 -9.78570282e-01 7.39511013e-01
6.58636928e-01 -6.84693635e-01 3.68008584e-01 1.14203835e+00
1.20252930e-01 -9.09368787e-03 -1.14737880e+00 6.18692756e-01
4.48611915e-01 -1.31751907e+00 5.88400960e-01 -4.57060188e-01
9.99759376e-01 -4.27666217e-01 -4.70129959e-02 1.13530314e+00
7.58248210e-01 -7.99302459e-01 4.64374989e-01 4.84803207e-02
6.82936788e-01 -9.61716533e-01 6.43473685e-01 4.76170123e-01
-9.13619757e-01 -2.09422007e-01 -6.20438337e-01 4.84372750e-02
-1.79321989e-01 7.48367086e-02 -1.09860635e+00 4.37279731e-01
3.83034170e-01 4.21398789e-01 -6.69615090e-01 7.51897633e-01
-3.70556742e-01 5.89557111e-01 1.42214894e-01 -6.22260809e-01
3.57397228e-01 -1.66287929e-01 5.72107613e-01 1.59169912e+00
-7.60959610e-02 2.95364499e-01 4.96567696e-01 9.19723988e-01
-1.84200078e-01 1.55777514e-01 -4.11055893e-01 2.59128600e-01
9.42988157e-01 9.84748781e-01 -2.43113577e-01 -6.74725354e-01
-2.10031793e-01 1.04987657e+00 3.81734252e-01 2.76030868e-01
-8.82825673e-01 -9.54208910e-01 3.66447121e-01 -2.45205030e-01
-1.09413676e-01 2.44248077e-01 -4.92081016e-01 -1.17197359e+00
-1.61398530e-01 -1.19873393e+00 3.16486746e-01 -8.17611277e-01
-1.49881375e+00 5.03470719e-01 3.31325263e-01 -1.22629738e+00
-4.05776173e-01 -3.12619358e-01 -1.00579679e+00 1.03370810e+00
-1.35062432e+00 -8.20178866e-01 -4.64986414e-01 5.21010637e-01
1.65054154e+00 -5.20453751e-01 8.34185779e-01 -9.99824852e-02
-7.22627580e-01 1.08415997e+00 2.36736313e-01 1.61748096e-01
7.03199565e-01 -1.71051550e+00 7.65725613e-01 5.80537915e-01
-3.92343163e-01 7.34580636e-01 8.93042028e-01 -7.73586333e-01
-9.72181261e-01 -1.08964980e+00 6.19701326e-01 -1.93805531e-01
7.58539379e-01 -4.89534646e-01 -1.09667230e+00 1.73769802e-01
4.70686585e-01 -8.79934490e-01 7.06134081e-01 -3.48980390e-02
-3.74957800e-01 6.56285509e-02 -9.40710247e-01 1.00109637e+00
6.91191494e-01 -3.72675151e-01 -1.36600876e+00 5.80982447e-01
1.14681673e+00 -6.53366029e-01 -5.94554782e-01 3.02321374e-01
2.28622526e-01 -5.87190628e-01 9.71806645e-01 -6.54356420e-01
1.04115164e+00 2.81568736e-01 2.54995435e-01 -1.73493516e+00
-2.10843965e-01 -6.30766571e-01 -3.85120094e-01 1.05892193e+00
3.65954489e-01 -1.59008503e-01 9.51578915e-01 3.47061276e-01
-4.83108521e-01 -8.52059722e-01 -7.63416290e-01 -7.92347968e-01
5.62281907e-01 8.14045891e-02 3.17427248e-01 1.19587779e+00
5.71456730e-01 7.46410370e-01 -3.93955261e-01 -5.06752729e-01
5.77802181e-01 1.21671885e-01 8.06532741e-01 -9.23338473e-01
-4.58741546e-01 -6.97347641e-01 4.07854259e-01 -1.38296139e+00
7.24189058e-02 -9.78401363e-01 1.70362592e-01 -1.31957102e+00
3.51377249e-01 -1.79234296e-01 1.07706033e-01 3.34872216e-01
-6.80738747e-01 -2.09024981e-01 1.05099998e-01 1.84771679e-02
-2.94301212e-01 8.95308197e-01 1.51007557e+00 -3.06150079e-01
-3.73814911e-01 1.27474055e-01 -7.49483943e-01 5.71567118e-01
8.52094710e-01 -7.90139213e-02 -9.61165547e-01 -2.78948307e-01
-6.66784570e-02 2.20796749e-01 -7.18039200e-02 -8.05922389e-01
2.19807848e-01 -2.51992941e-01 3.64993721e-01 -7.51916468e-01
4.82634574e-01 -1.98218539e-01 -6.31556988e-01 5.43082595e-01
-1.13203239e+00 1.65913150e-01 3.95251542e-01 6.62242234e-01
-2.77182851e-02 -7.72087991e-01 8.57247055e-01 -3.51458043e-01
-5.01858711e-01 -1.01318620e-01 -5.97086966e-01 5.77588379e-01
9.61408257e-01 -4.34454441e-01 -6.94468439e-01 -4.13138151e-01
-6.58150434e-01 8.68399441e-01 4.94369790e-02 8.49112391e-01
8.41398001e-01 -1.29661334e+00 -9.70625162e-01 2.57061094e-01
3.23123157e-01 1.54229730e-01 2.66658753e-01 1.19737335e-01
-2.06906855e-01 2.96015233e-01 -1.13444269e-01 -4.48496997e-01
-1.41947198e+00 3.56220454e-01 2.53616482e-01 -4.09731001e-01
-5.32570302e-01 1.23664999e+00 1.65585265e-01 -5.89305639e-01
7.47761846e-01 -2.49752745e-01 -4.76097435e-01 4.02213842e-01
6.74942434e-01 3.19404691e-01 3.17927785e-02 2.65499093e-02
5.24652362e-01 9.67705250e-02 -9.37929809e-01 -7.54787177e-02
9.77970481e-01 -2.19824184e-02 3.28424186e-01 3.68140340e-01
8.22203696e-01 -5.42554259e-01 -1.08761489e+00 -4.37735140e-01
-1.38588443e-01 -3.22350591e-01 -4.39393640e-01 -1.15299034e+00
-4.92057592e-01 8.79840016e-01 3.34929526e-02 2.98509121e-01
9.30114210e-01 -3.24719667e-01 9.74227190e-01 7.08895981e-01
1.37282312e-01 -1.54764807e+00 9.02559280e-01 9.47965860e-01
1.26155674e+00 -9.48268235e-01 -6.77265897e-02 -2.31108427e-01
-1.16864085e+00 9.82433259e-01 1.31517100e+00 -1.18986167e-01
-2.22044680e-02 5.70880286e-02 -2.46632993e-01 1.68809161e-01
-1.07152545e+00 1.58894137e-01 2.13212725e-02 3.91832143e-01
3.29275608e-01 -1.25000089e-01 -6.53624654e-01 8.83572578e-01
-7.32609391e-01 -2.91140825e-01 9.23741758e-01 8.84208143e-01
-7.87131488e-01 -1.20003307e+00 -1.33455992e-01 6.92541003e-01
-5.45095578e-02 -4.05040622e-01 -6.12912297e-01 5.36254764e-01
-1.67929098e-01 8.73785675e-01 -1.01575375e-01 -6.85653150e-01
6.64501846e-01 4.43652838e-01 2.54841447e-01 -1.08071458e+00
-1.09917760e+00 -1.55660450e-01 2.52828807e-01 -1.30069152e-01
3.09550971e-01 -5.20901203e-01 -1.05533004e+00 -3.56315851e-01
-5.44957578e-01 4.08205986e-01 2.55547404e-01 7.69949555e-01
-3.53659354e-02 7.74351895e-01 9.59336460e-01 -6.08414412e-01
-1.24203908e+00 -1.39861631e+00 -2.28527188e-01 7.02493548e-01
1.05021961e-01 -4.70141619e-01 -3.72129589e-01 -4.57813917e-03] | [11.832868576049805, 9.1710205078125] |
44c04c98-9a72-4837-8206-46427f58599d | differentially-private-cross-camera-person-re | 2306.02765 | null | https://arxiv.org/abs/2306.02765v1 | https://arxiv.org/pdf/2306.02765v1.pdf | Differentially Private Cross-camera Person Re-identification | Camera-based person re-identification is a heavily privacy-invading task by design, benefiting from rich visual data to match together person representations across different cameras. This high-dimensional data can then easily be used for other, perhaps less desirable, applications. We here investigate the possibility of protecting such image data against uses outside of the intended re-identification task, and introduce a differential privacy mechanism leveraging both pixelisation and colour quantisation for this purpose. We show its ability to distort images in such a way that adverse task performances are significantly reduced, while retaining high re-identification performances. | ['Keiichi Yasumoto', 'Yuki Matsuda', 'Lucas Maris'] | 2023-06-05 | null | null | null | null | ['person-re-identification'] | ['computer-vision'] | [ 6.38629138e-01 6.17303886e-03 2.22981483e-01 -5.64256489e-01
-4.21673596e-01 -9.82278943e-01 8.69947433e-01 2.22911350e-02
-8.77984822e-01 5.81348181e-01 3.70047778e-01 -8.80232230e-02
9.47599784e-02 -5.13630569e-01 -5.17981768e-01 -7.62187541e-01
1.38303228e-02 -2.36050449e-02 -1.41297638e-01 2.33165234e-01
1.73907027e-01 7.03875005e-01 -1.51452947e+00 1.96323141e-01
2.74608284e-01 7.03623116e-01 -5.59201479e-01 8.65884662e-01
7.14981318e-01 1.10796936e-01 -5.23825765e-01 -1.16288173e+00
8.93984973e-01 -2.23116040e-01 -4.33429390e-01 1.87164426e-01
6.87873006e-01 -4.41970706e-01 -5.14041185e-01 1.13692868e+00
7.27121472e-01 1.72966123e-01 5.18703878e-01 -1.20049441e+00
-8.69241357e-01 -1.49348810e-01 -4.60318476e-01 1.67738378e-01
8.13653350e-01 2.00176358e-01 4.89189208e-01 -2.80177891e-01
4.09171283e-01 1.19294369e+00 7.95561373e-01 1.00237942e+00
-1.66422808e+00 -9.58819091e-01 -2.35320583e-01 -7.29491860e-02
-1.70771253e+00 -8.04473221e-01 5.44480264e-01 -3.45621079e-01
5.76590776e-01 8.12662542e-01 7.11202383e-01 1.35740495e+00
-1.73663765e-01 3.46363336e-01 1.34330881e+00 -2.84467250e-01
1.08015575e-02 4.55577493e-01 -7.47035220e-02 2.99642295e-01
5.57952940e-01 4.25271094e-01 -5.62600255e-01 -4.56386417e-01
6.28260911e-01 1.80031002e-01 -5.16710520e-01 -6.33529902e-01
-1.05998516e+00 6.23389006e-01 2.59176433e-01 -2.35861223e-02
-4.30498794e-02 4.85573001e-02 4.97827947e-01 2.64819860e-01
1.44872442e-01 6.13480926e-01 -1.78268440e-02 1.91799570e-02
-7.08854735e-01 4.13265646e-01 5.51281810e-01 8.27663720e-01
6.43968821e-01 -5.01304448e-01 -3.56833726e-01 4.99455899e-01
4.29847725e-02 5.67724764e-01 6.72944307e-01 -7.69841254e-01
3.58681291e-01 1.91947967e-01 1.78602457e-01 -1.20559025e+00
-1.93074495e-01 2.11994462e-02 -1.08714163e+00 2.72194833e-01
5.52787542e-01 -3.66115093e-01 -6.88356638e-01 1.82309592e+00
2.02897400e-01 4.92983647e-02 -4.11052629e-02 8.93211305e-01
2.52556264e-01 1.99067399e-01 3.85969311e-01 8.56746137e-02
1.63215804e+00 -3.41115519e-02 -4.48934406e-01 -1.32898226e-01
2.92319655e-01 -4.20756698e-01 5.91389835e-01 3.59368801e-01
-1.01154876e+00 -4.32282060e-01 -1.08196771e+00 -1.69404566e-01
-5.38819492e-01 -2.37877756e-01 8.05035114e-01 1.64452899e+00
-1.33017266e+00 3.07295710e-01 -2.55939990e-01 -5.41780174e-01
7.80802429e-01 7.43441224e-01 -1.05047011e+00 -7.23402053e-02
-1.13087225e+00 8.07354450e-01 2.78195202e-01 -4.77877520e-02
-3.44995290e-01 -5.78765988e-01 -9.11907673e-01 5.25928522e-03
-1.08418427e-01 -7.75624633e-01 7.75266826e-01 -1.11183417e+00
-1.33770514e+00 1.49924421e+00 6.27727509e-02 -6.95109308e-01
8.12253773e-01 1.62640274e-01 -6.71665251e-01 -2.19792873e-02
-1.11695036e-01 7.66714931e-01 1.26725280e+00 -1.12154114e+00
-6.11178637e-01 -9.05066788e-01 -1.37751356e-01 4.59835857e-01
-7.40371346e-01 7.65736625e-02 -5.65086067e-01 -8.22803557e-01
-5.55958331e-01 -1.17828119e+00 -1.48846731e-01 1.69262618e-01
-5.59937537e-01 3.13057303e-01 7.84251571e-01 -8.37288141e-01
8.87579322e-01 -2.30416393e+00 -7.61669725e-02 6.03137612e-01
3.34833592e-01 6.80762947e-01 8.87075141e-02 7.28752930e-03
-1.92885980e-01 4.51501787e-01 -1.01630002e-01 -5.69456756e-01
2.27065571e-02 -1.60890847e-01 -7.56257102e-02 9.42262411e-01
-6.46190569e-02 9.54680502e-01 -5.96967638e-01 -1.27982855e-01
4.08925563e-01 6.46909237e-01 -3.64704162e-01 2.76003152e-01
6.00147426e-01 3.66560847e-01 -6.58635423e-02 5.40513158e-01
8.21738124e-01 4.39284265e-01 2.02885389e-01 -9.88474786e-02
2.30419651e-01 -3.32595646e-01 -1.06304789e+00 1.35785580e+00
-1.61140934e-01 8.51710021e-01 1.86615825e-01 -6.62166715e-01
7.31966019e-01 1.99328721e-01 2.73706973e-01 -6.77339315e-01
1.07987158e-01 -3.06945592e-01 -3.04991215e-01 -1.37687802e-01
6.42479539e-01 -1.69281498e-01 -3.54810208e-01 3.53023320e-01
-3.68052721e-01 2.94326723e-01 -3.37668687e-01 -5.79916611e-02
8.64679217e-01 -3.81037951e-01 3.09646368e-01 -2.38482937e-01
4.57796961e-01 -4.56094831e-01 3.43515486e-01 8.82624149e-01
-6.87799096e-01 8.91623557e-01 -1.30256563e-02 -2.87067413e-01
-1.35728121e+00 -9.20414269e-01 -3.91728282e-01 6.78779602e-01
2.73869395e-01 -2.19649747e-01 -8.88210893e-01 -7.61211991e-01
5.73445320e-01 4.01287109e-01 -8.76713037e-01 -5.02296507e-01
-2.90861249e-01 -7.95399129e-01 1.18104446e+00 2.73644924e-01
7.55465984e-01 -5.61990440e-01 -4.99214113e-01 -1.88798428e-01
2.53660351e-01 -9.17566955e-01 -7.59362102e-01 -1.00376576e-01
-4.06341493e-01 -1.02797794e+00 -1.31872976e+00 -5.37539959e-01
8.01388323e-01 5.47328532e-01 7.51924217e-01 8.31749514e-02
-3.81290704e-01 9.62502599e-01 -6.58234507e-02 -2.61623204e-01
-1.93101272e-01 -2.38764286e-01 3.06434482e-01 6.93738997e-01
8.06441963e-01 -3.78914505e-01 -7.50178039e-01 2.83692092e-01
-9.76713836e-01 -2.08298787e-01 1.60201311e-01 5.97380877e-01
1.95437476e-01 2.34096751e-01 1.55839220e-01 -9.11394656e-01
8.59435320e-01 -1.09417893e-01 -4.08379406e-01 3.21048200e-01
-5.69643497e-01 -1.23141393e-01 2.58843750e-01 -7.11685717e-01
-1.02626562e+00 4.30952430e-01 1.10532679e-01 -1.93527445e-01
-5.10219157e-01 -4.48853582e-01 -5.54774582e-01 -6.58678353e-01
5.51199555e-01 2.54598290e-01 4.85822931e-02 -3.88750017e-01
5.83764195e-01 1.04483283e+00 8.89400601e-01 -1.47522673e-01
1.01721239e+00 7.21682370e-01 -3.26862261e-02 -7.71930218e-01
-3.92336398e-02 -5.63630521e-01 -6.37184978e-01 -1.57254577e-01
9.80000913e-01 -1.11308587e+00 -1.01552391e+00 6.82921112e-01
-9.18497503e-01 9.14893597e-02 -3.38867992e-01 2.10571006e-01
-3.12184960e-01 7.11307049e-01 -2.46206105e-01 -8.92656267e-01
-9.62621719e-02 -7.94794738e-01 1.08581400e+00 3.52925330e-01
-3.61611277e-01 -9.63690042e-01 6.88549727e-02 5.20540953e-01
1.76953793e-01 3.58856648e-01 1.57217085e-01 -6.77460968e-01
-1.25631869e-01 -8.37887049e-01 -3.93832445e-01 3.31808150e-01
1.96172982e-01 -5.15932500e-01 -1.26812708e+00 -7.29320824e-01
-3.36047797e-03 -1.32573918e-01 8.61494958e-01 8.60334262e-02
1.17071140e+00 -5.08638620e-01 -4.54869896e-01 1.06203055e+00
1.40790141e+00 -1.74484141e-02 1.06213272e+00 5.91070235e-01
8.59604239e-01 8.19012523e-01 -1.14724927e-01 4.85923499e-01
3.78701299e-01 8.97812009e-01 6.94724247e-02 -3.28287870e-01
2.40828618e-02 -3.06216300e-01 1.35374680e-01 -3.26645941e-01
-1.46356248e-03 -2.76721925e-01 -6.35427058e-01 5.01391172e-01
-1.72412717e+00 -1.11140561e+00 7.95689523e-02 2.64497972e+00
7.79958069e-01 -3.40773612e-01 5.17989993e-01 3.48438844e-02
1.10537326e+00 -2.03757547e-02 -4.83269930e-01 -5.07903993e-01
-3.60009462e-01 -1.68353274e-01 1.32310843e+00 1.89110547e-01
-1.39035702e+00 5.55373549e-01 7.17603254e+00 5.04545212e-01
-6.71736181e-01 -1.65686026e-01 9.43765879e-01 -1.52716681e-01
-1.14032581e-01 -2.16293827e-01 -7.44296432e-01 7.14751601e-01
8.59890342e-01 -2.76985973e-01 6.11911595e-01 5.53276122e-01
-8.30521062e-02 -4.72228378e-02 -1.21046317e+00 1.57665372e+00
3.95223379e-01 -1.12852073e+00 4.88123186e-02 5.25425732e-01
4.57909226e-01 -6.97247267e-01 5.15724838e-01 -2.03759655e-01
4.50214177e-01 -1.19775438e+00 5.03435194e-01 4.01910931e-01
1.13035583e+00 -8.70573878e-01 5.25148571e-01 -4.68835123e-02
-8.75650048e-01 -1.92396060e-01 -4.04344201e-01 -4.59716376e-03
6.58526495e-02 3.07470739e-01 -5.14233649e-01 2.67645359e-01
7.32480764e-01 4.46541518e-01 -9.97463882e-01 1.06978059e+00
1.50193691e-01 1.10165849e-01 -3.98417145e-01 2.90793091e-01
-2.93359786e-01 4.07517962e-02 4.79148626e-01 1.48897994e+00
3.23650956e-01 3.75039689e-02 -2.65827090e-01 5.04219294e-01
-1.81169495e-01 -1.49779588e-01 -1.03974891e+00 2.68955976e-01
4.10284102e-01 9.78863299e-01 -2.36474916e-01 -2.35696509e-01
-3.36328208e-01 1.82220864e+00 5.76570146e-02 3.09479058e-01
-5.33247769e-01 -2.95964956e-01 1.07587707e+00 1.03375040e-01
1.99338377e-01 -1.22949272e-01 -3.34231585e-01 -1.37805772e+00
9.41840652e-03 -8.64350975e-01 6.33825660e-01 -5.57573438e-01
-1.57021534e+00 1.98880509e-01 -2.27996930e-01 -1.13603044e+00
-1.95413649e-01 -4.75507796e-01 -1.56477362e-01 1.12663901e+00
-1.32082236e+00 -1.34985495e+00 -8.17900300e-02 1.08801913e+00
-3.31124723e-01 -2.99182355e-01 9.22152519e-01 1.89624384e-01
-6.28037512e-01 1.23592722e+00 2.37306535e-01 4.12058115e-01
8.35441411e-01 -1.25395334e+00 5.47057927e-01 1.11191392e+00
1.98177159e-01 6.80427730e-01 4.36077774e-01 -5.41155040e-01
-1.43688083e+00 -1.18273664e+00 7.68127799e-01 -8.15487325e-01
8.21883380e-02 -8.55608642e-01 -6.24765337e-01 6.58226073e-01
3.52806388e-03 -2.23204438e-02 1.01040912e+00 1.67482924e-02
-7.23306954e-01 -1.97453171e-01 -1.79470503e+00 6.47727728e-01
9.68215108e-01 -1.01311684e+00 -4.62028801e-01 1.27584711e-01
1.89682290e-01 4.00508717e-02 -7.93582737e-01 -1.46911338e-01
7.78134882e-01 -1.12336147e+00 1.46697080e+00 -5.04303277e-01
-4.14947808e-01 -1.09357201e-01 -1.11568712e-01 -9.93038833e-01
-6.13533318e-01 -1.05972016e+00 1.52311489e-01 1.72125077e+00
-7.42886886e-02 -7.38905966e-01 1.12548721e+00 1.57465529e+00
9.91978288e-01 4.01194006e-01 -1.11403954e+00 -8.74581099e-01
-6.75848350e-02 -2.94886112e-01 7.61440456e-01 8.54676902e-01
9.27017853e-02 -7.08444715e-02 -9.11995113e-01 3.72435182e-01
9.75447714e-01 -4.96875674e-01 9.27326977e-01 -1.03282654e+00
-8.63347501e-02 -1.97240904e-01 -9.65686858e-01 -8.45435262e-01
8.62034708e-02 -7.67740130e-01 -1.80926338e-01 -5.36717832e-01
3.75392586e-01 -3.70320976e-01 -2.45797694e-01 2.54037857e-01
-2.35845476e-01 9.10988510e-01 4.16210175e-01 2.62488753e-01
-2.82580584e-01 9.15694982e-02 6.63649678e-01 -2.17743501e-01
-3.94850112e-02 9.86982360e-02 -1.15461457e+00 2.48178557e-01
7.85079777e-01 -2.67827421e-01 -3.01276058e-01 -3.26503634e-01
-1.22751042e-01 -4.10696685e-01 8.29855084e-01 -1.20617199e+00
2.03184158e-01 1.10086314e-01 9.59810555e-01 1.67992324e-01
4.31740910e-01 -1.16960537e+00 4.01319653e-01 1.71124712e-01
-5.62788546e-01 5.96950110e-03 2.18146011e-01 8.88966084e-01
1.25056669e-01 -6.39542704e-03 9.21469629e-01 -2.64327496e-01
-6.65036559e-01 4.64522034e-01 -3.74116719e-01 -2.95289934e-01
1.30502200e+00 -8.45039725e-01 -9.88686923e-03 -3.47317010e-01
-5.14901400e-01 -1.54808849e-01 1.17522323e+00 2.66410530e-01
4.58313495e-01 -1.39070570e+00 -6.67165220e-01 7.19674289e-01
2.73101926e-01 -6.20128930e-01 1.54466495e-01 1.20921962e-01
-2.14788780e-01 2.80098826e-01 -5.23161471e-01 -3.23060155e-01
-1.67401528e+00 9.88728166e-01 2.95589507e-01 5.42288758e-02
-8.33299875e-01 8.85298193e-01 2.82887608e-01 -1.47209421e-01
2.45120987e-01 3.86446953e-01 -9.88552943e-02 -1.54746184e-02
9.85565245e-01 2.69204557e-01 -5.78257553e-02 -1.01197624e+00
-5.30411780e-01 4.48923320e-01 -4.63207848e-02 -2.88905501e-01
8.21487069e-01 -7.87953675e-01 2.78716356e-01 -3.88313234e-01
1.50123703e+00 -1.68522894e-02 -1.55656958e+00 -7.80062303e-02
-5.08111641e-02 -1.10076129e+00 -2.25076750e-02 -7.29888320e-01
-8.32587898e-01 5.98988473e-01 1.11733675e+00 4.48705852e-01
1.23098099e+00 -4.25860643e-01 8.10931921e-01 6.53567761e-02
4.19926196e-01 -1.03921711e+00 -5.11631727e-01 -2.65106142e-01
5.48832834e-01 -1.30899978e+00 2.62104690e-01 -1.93193570e-01
-7.80164182e-01 7.76657224e-01 4.37170416e-02 -4.55011241e-02
5.43978035e-01 1.73430532e-01 -5.65865193e-04 2.01284811e-01
-9.86358225e-02 -3.21336300e-03 3.91880572e-01 1.58124304e+00
-1.96352854e-01 2.37642005e-01 1.69760734e-01 4.90626842e-01
-8.11414793e-02 -1.38974354e-01 5.19672155e-01 7.31053650e-01
1.82912841e-01 -1.16383719e+00 -7.78800130e-01 4.05637056e-01
-4.69331205e-01 3.57784554e-02 -7.65778959e-01 5.15306413e-01
2.28651121e-01 9.29467440e-01 3.17920893e-01 -4.97520566e-01
3.30459535e-01 -7.58182108e-02 3.93685490e-01 -2.22175568e-01
-8.14342439e-01 -4.76252854e-01 1.32575259e-01 -6.16663456e-01
-4.63677824e-01 -9.56136167e-01 -2.22815543e-01 -8.97066832e-01
-1.03360834e-02 -3.17006379e-01 5.76961100e-01 4.17813629e-01
6.22197688e-01 -2.24154159e-01 6.97061121e-01 -9.79083300e-01
-3.86027455e-01 -2.03847602e-01 -7.98183560e-01 1.06693053e+00
6.80816650e-01 -2.97972471e-01 -3.60817462e-01 3.34686100e-01] | [12.796225547790527, 0.8035993576049805] |
999adc9f-92ae-4af3-8d05-64688b8d9296 | do-you-follow-me-a-survey-of-recent | 2207.14627 | null | https://arxiv.org/abs/2207.14627v1 | https://arxiv.org/pdf/2207.14627v1.pdf | "Do you follow me?": A Survey of Recent Approaches in Dialogue State Tracking | While communicating with a user, a task-oriented dialogue system has to track the user's needs at each turn according to the conversation history. This process called dialogue state tracking (DST) is crucial because it directly informs the downstream dialogue policy. DST has received a lot of interest in recent years with the text-to-text paradigm emerging as the favored approach. In this review paper, we first present the task and its associated datasets. Then, considering a large number of recent publications, we identify highlights and advances of research in 2021-2022. Although neural approaches have enabled significant progress, we argue that some critical aspects of dialogue systems such as generalizability are still underexplored. To motivate future studies, we propose several research avenues. | ['Benoit Favre', 'Lina M. Rojas-Barahona', 'Léo Jacqmin'] | 2022-07-29 | null | null | null | null | ['dialogue-state-tracking'] | ['natural-language-processing'] | [ 2.72402614e-01 5.60671866e-01 -3.61871839e-01 -6.88061297e-01
-3.17641318e-01 -7.16648936e-01 9.47601974e-01 5.79272173e-02
-4.55468595e-01 8.84524047e-01 7.91511059e-01 -3.88596743e-01
4.62194271e-02 -3.28309685e-01 2.53810018e-01 -2.05188438e-01
1.31594792e-01 5.01011550e-01 1.76025387e-02 -7.93514073e-01
6.08823895e-01 2.49800533e-01 -1.23176908e+00 4.86324698e-01
7.93371558e-01 7.27043450e-01 1.95227697e-01 7.42594600e-01
-3.38961035e-01 9.00459468e-01 -7.93403208e-01 -3.78834397e-01
-1.08084798e-01 -9.14672732e-01 -1.46385086e+00 -6.34850934e-02
-2.83498392e-02 -6.49806082e-01 -4.15335178e-01 7.38139033e-01
7.45476365e-01 4.83620584e-01 4.32906210e-01 -1.07598734e+00
-1.24103665e-01 6.26347721e-01 1.42282471e-01 2.31169477e-01
8.12072933e-01 2.97887832e-01 1.11061311e+00 -5.24810970e-01
6.13966048e-01 1.39259064e+00 3.83388430e-01 1.14692378e+00
-1.01492727e+00 -2.32938915e-01 4.76883680e-01 1.78282663e-01
-5.62264442e-01 -9.35214102e-01 6.81260645e-01 -2.83541471e-01
1.14099765e+00 3.90482605e-01 8.00305903e-01 1.33596921e+00
-4.97545823e-02 1.22040260e+00 1.30201578e+00 -3.87809396e-01
1.26357868e-01 3.15321803e-01 3.76657814e-01 4.06366080e-01
-5.37706614e-01 -1.85055315e-01 -8.34868312e-01 -2.73520291e-01
5.34086585e-01 -7.70389915e-01 -4.17798549e-01 5.76514229e-02
-1.01453459e+00 9.41899717e-01 -5.81573583e-02 5.12356460e-01
-3.25859964e-01 -4.65499789e-01 7.61655629e-01 5.34059167e-01
4.84691918e-01 7.48312652e-01 -4.56472337e-01 -9.46411312e-01
-6.25351071e-01 4.09686357e-01 1.43896472e+00 6.45352364e-01
2.10690573e-01 -6.25158921e-02 -5.92499971e-01 1.25644386e+00
2.81159878e-01 -1.02180004e-01 3.94971222e-01 -1.06246483e+00
4.97076571e-01 6.00307941e-01 2.34688148e-01 -8.64801824e-01
-5.96193969e-01 1.85097963e-01 -7.01821446e-01 -3.09623033e-01
7.91810870e-01 -6.09346509e-01 -1.36202276e-01 1.82073152e+00
2.40271986e-01 -4.57965642e-01 2.35253796e-01 8.45071375e-01
7.06872940e-01 6.53074086e-01 5.22636622e-02 -4.21975076e-01
1.36992311e+00 -8.53122234e-01 -1.13302946e+00 -4.48223203e-01
4.26435143e-01 -5.62317193e-01 9.70668137e-01 4.11708534e-01
-1.14009607e+00 -2.69132704e-01 -7.25373864e-01 -1.52895585e-01
-1.96133852e-01 5.98350260e-03 6.79503202e-01 7.40370929e-01
-1.15975153e+00 4.20025080e-01 -6.18987620e-01 -8.80878568e-01
-1.46046951e-01 2.36584827e-01 -5.89891262e-02 4.49539810e-01
-1.75327635e+00 1.46856892e+00 2.03529984e-01 4.27183628e-01
-5.71208715e-01 -1.21543325e-01 -6.81448102e-01 -5.89326955e-02
3.57353747e-01 -3.83964121e-01 1.92498147e+00 -5.56633115e-01
-2.45193195e+00 7.41831362e-01 -3.59520853e-01 -4.24401462e-01
5.89415252e-01 -2.26248965e-01 -2.08288297e-01 1.56825230e-01
-3.71560454e-01 4.43266898e-01 4.09203053e-01 -9.04553831e-01
-1.03768337e+00 -3.03288728e-01 4.19231355e-01 7.48313785e-01
-3.35555404e-01 3.85374099e-01 -4.49826181e-01 -1.52044803e-01
3.40886302e-02 -9.93320882e-01 -1.38282463e-01 -2.71402150e-01
-5.46486378e-01 -7.16825664e-01 6.09622598e-01 -7.08822668e-01
1.53221655e+00 -1.68152916e+00 1.46107197e-01 -2.76457250e-01
2.11563379e-01 3.95545185e-01 1.21438220e-01 1.05963063e+00
2.10986942e-01 -4.85054478e-02 -1.66840971e-01 -3.57682705e-01
1.75944418e-01 -1.62758857e-01 -2.92575002e-01 2.89994270e-01
5.20903505e-02 8.85689378e-01 -9.03196752e-01 -2.55664974e-01
3.80854815e-01 9.48162004e-02 -4.04027760e-01 6.61794901e-01
-3.38140160e-01 7.57847905e-01 -5.83483338e-01 1.54139355e-01
1.55918762e-01 -1.73562616e-01 5.70763528e-01 -1.41912512e-02
-4.27742153e-01 9.85118389e-01 -6.23567045e-01 1.57044220e+00
-2.75036365e-01 9.62374151e-01 5.62286794e-01 -1.05658782e+00
9.14879620e-01 5.84738433e-01 4.25529331e-01 -5.11300087e-01
1.28840938e-01 -1.23352073e-01 3.73650491e-01 -7.74067342e-01
8.15691113e-01 -1.10387631e-01 -2.63881356e-01 7.43471801e-01
-1.66641191e-01 -1.01346508e-01 9.79863405e-02 3.56244951e-01
9.47172284e-01 3.06744426e-02 5.99885523e-01 -2.06572697e-01
6.29094660e-01 1.66646376e-01 3.41059804e-01 6.32703781e-01
-7.25539684e-01 1.34313852e-01 9.42738473e-01 -2.17918664e-01
-7.28785515e-01 -4.26141590e-01 -7.13375807e-02 1.29765224e+00
-2.21465230e-01 -4.19805050e-01 -9.55643296e-01 -6.15434825e-01
-1.92155391e-01 8.68689835e-01 -5.28555930e-01 -1.73982233e-02
-6.78788126e-01 -4.89001691e-01 6.66919887e-01 1.72442794e-01
8.05688441e-01 -1.31253588e+00 -5.90304732e-01 3.94546360e-01
-6.18991733e-01 -1.03009248e+00 -3.14436346e-01 -4.70013693e-02
-7.07367599e-01 -8.94121587e-01 -7.70628333e-01 -6.40880704e-01
2.22780108e-01 2.21020862e-01 9.02905583e-01 -5.31283841e-02
3.68259341e-01 4.47782189e-01 -3.36376935e-01 -1.01347208e-01
-8.30930173e-01 3.92567009e-01 1.53251722e-01 -1.01748958e-01
3.87272358e-01 -2.18704611e-01 -5.71215689e-01 4.67471629e-01
-2.19290361e-01 3.70254457e-01 1.82391167e-01 8.19902062e-01
-4.82333839e-01 -4.52968121e-01 9.70378697e-01 -1.10486150e+00
1.52919436e+00 -4.62735981e-01 -2.71971196e-01 2.06334770e-01
-6.88184381e-01 -2.11185262e-01 5.21042228e-01 -3.06369122e-02
-1.55107260e+00 -1.93561226e-01 -3.66813302e-01 4.94196475e-01
-3.54565054e-01 6.56747520e-01 -7.08172321e-02 2.44619340e-01
4.83358860e-01 2.48322040e-01 4.58913743e-01 -4.41288739e-01
3.12822670e-01 1.19313788e+00 2.55213857e-01 -6.57762408e-01
1.21920437e-01 1.27122495e-02 -6.88091099e-01 -1.17549765e+00
-6.73935533e-01 -4.23839480e-01 -6.90968275e-01 -6.92993283e-01
7.52317369e-01 -5.84481180e-01 -1.18586922e+00 5.55428088e-01
-1.32383573e+00 -6.69762850e-01 7.99598023e-02 2.78135836e-01
-5.81817508e-01 4.98588562e-01 -8.80339086e-01 -1.36834717e+00
-3.61968011e-01 -1.20767486e+00 4.62219238e-01 4.31929410e-01
-9.57925916e-01 -1.07724607e+00 -4.72018272e-02 6.47763848e-01
6.19464755e-01 -2.87886143e-01 7.72302270e-01 -1.20507145e+00
-7.02203512e-02 -1.23976536e-01 6.28916919e-02 1.58799112e-01
2.99223661e-01 -1.72784105e-01 -1.10555577e+00 -1.93515405e-01
3.27940881e-01 -5.29374003e-01 3.20987940e-01 2.45899782e-01
6.74384415e-01 -3.40925753e-01 -2.54477859e-01 -1.56145781e-01
4.43997413e-01 4.86902475e-01 3.02356511e-01 1.76693112e-01
1.76733434e-01 1.44671333e+00 7.92800486e-01 5.48928320e-01
8.28021705e-01 7.08775938e-01 1.71278529e-02 2.73887515e-01
8.05890709e-02 -3.45537841e-01 5.55677176e-01 1.00511825e+00
2.24172547e-01 -4.34539884e-01 -9.06767428e-01 3.24737340e-01
-1.90462041e+00 -9.47951376e-01 -3.51551771e-02 1.97905326e+00
1.12324226e+00 1.71624288e-01 4.14080411e-01 -4.26735505e-02
9.11575735e-01 4.66939300e-01 -7.04389811e-01 -6.35827482e-01
2.80001108e-02 -3.75426173e-01 -1.59320861e-01 7.64698982e-01
-7.32200503e-01 1.10886967e+00 7.45528841e+00 3.32926124e-01
-1.08484209e+00 -2.67036915e-01 7.23274887e-01 1.36404499e-01
8.14553723e-03 -3.29122618e-02 -7.78164327e-01 4.16572958e-01
1.05423367e+00 -3.74598742e-01 6.52486563e-01 5.57737470e-01
7.23999977e-01 -6.44217193e-01 -1.17398477e+00 5.19427359e-01
-3.18662561e-02 -9.32810843e-01 -4.18778241e-01 -1.02571538e-02
2.47009039e-01 -1.67867959e-01 -1.84240460e-01 4.87045050e-01
5.09197235e-01 -7.14060783e-01 2.15234563e-01 3.82801890e-01
5.88510633e-01 -4.00321633e-01 4.51383442e-01 7.03108668e-01
-8.34275544e-01 -6.53599016e-03 -2.90443725e-03 -4.61481839e-01
4.11710888e-01 1.56017929e-01 -1.20208657e+00 2.74648219e-01
2.14503258e-01 5.72025001e-01 -8.36437121e-02 5.90027034e-01
-3.56846750e-01 8.41106117e-01 -1.26053184e-01 -4.90156323e-01
2.72778034e-01 -4.14112091e-01 5.63219309e-01 1.14295018e+00
-2.22474486e-01 4.84011173e-01 1.64803401e-01 7.05216825e-01
1.63929209e-01 1.58902049e-01 -5.98677635e-01 -3.68973941e-01
7.07521796e-01 1.24390197e+00 -5.56415439e-01 -2.67222911e-01
-4.25994933e-01 1.01330113e+00 2.95937061e-01 2.86464810e-01
-2.80497342e-01 -1.34910390e-01 8.98134947e-01 -4.16609943e-01
-4.59949285e-01 -3.32535535e-01 -2.40453422e-01 -9.76329505e-01
-2.52974302e-01 -1.16987991e+00 3.23610783e-01 -3.66082847e-01
-1.08054531e+00 4.78267193e-01 -3.17296237e-02 -8.08368862e-01
-6.11783028e-01 -4.52383727e-01 -5.49599946e-01 1.02060568e+00
-1.19194329e+00 -5.70659220e-01 -8.50183051e-03 1.72291026e-01
1.06519806e+00 -1.58459112e-01 1.05651832e+00 -1.74031258e-01
-9.17186260e-01 4.89260167e-01 -3.34116817e-02 2.29285896e-01
8.77617955e-01 -1.17537773e+00 6.55417740e-01 1.94979027e-01
-4.05566007e-01 8.67584348e-01 8.78399372e-01 -6.79950714e-01
-1.51403916e+00 -2.91345537e-01 1.13628471e+00 -4.11588043e-01
6.17303729e-01 -3.11594993e-01 -8.48767400e-01 5.33432484e-01
7.14331031e-01 -8.23611915e-01 7.54444242e-01 3.54466408e-01
2.89114445e-01 3.52325141e-01 -1.06309259e+00 8.86653781e-01
8.27936053e-01 -7.90575981e-01 -6.69948280e-01 2.00624436e-01
3.89725357e-01 -6.78904235e-01 -8.93767059e-01 -2.38808300e-02
6.34532452e-01 -1.14328098e+00 4.41607118e-01 -6.93008125e-01
2.09683985e-01 3.33475798e-01 1.25482440e-01 -1.48647988e+00
-1.00461401e-01 -1.26778591e+00 1.10539095e-02 1.22126555e+00
3.56439054e-01 -8.16791534e-01 8.77718687e-01 1.38711238e+00
-1.48992687e-01 -8.01115036e-01 -7.83372521e-01 -2.37956434e-01
1.69380903e-01 -2.54379660e-01 2.22608268e-01 1.04109466e+00
1.03024232e+00 8.75519097e-01 -6.66633189e-01 -4.46679443e-01
-1.51900440e-01 -7.71890357e-02 7.98035622e-01 -1.25928164e+00
9.34352353e-02 -6.87416255e-01 4.36572641e-01 -1.73414457e+00
2.47816309e-01 -5.93720078e-01 2.59262294e-01 -1.60505736e+00
-9.72299054e-02 -2.18968596e-02 1.59060135e-01 2.78379917e-01
-2.73641974e-01 -5.00678539e-01 2.27858841e-01 5.52388094e-02
-4.64461356e-01 7.87050247e-01 1.26824844e+00 7.05692619e-02
-3.03533912e-01 5.03555536e-01 -8.66088092e-01 5.96677840e-01
1.07264352e+00 5.62449433e-02 -4.38018948e-01 -4.32363860e-02
-9.20155197e-02 6.24623537e-01 -3.14693987e-01 -3.83669317e-01
4.69548732e-01 -3.35319042e-01 -9.04235467e-02 -4.72137779e-01
7.44732499e-01 -4.28215802e-01 -5.72061777e-01 3.95730823e-01
-1.00841713e+00 1.25546724e-01 1.52034804e-01 3.72762054e-01
-8.42148811e-02 -3.34754258e-01 6.01996422e-01 -1.91729516e-01
-7.02021301e-01 -1.49446458e-01 -1.15221167e+00 1.19153924e-01
7.71832466e-01 -8.01646560e-02 -2.49199569e-01 -9.55212116e-01
-6.07150078e-01 5.99382579e-01 9.12638903e-02 6.04149580e-01
4.03778046e-01 -8.31067920e-01 -6.12780750e-01 -1.65383607e-01
-7.46490434e-02 -3.81545782e-01 2.71133423e-01 7.74691343e-01
-3.04429233e-02 9.34548616e-01 -1.46040618e-01 -2.33470470e-01
-1.27220583e+00 -1.15152411e-01 4.27725077e-01 -1.67253077e-01
-6.38955772e-01 7.38635361e-01 -1.69004455e-01 -6.74544811e-01
5.00656128e-01 6.28671870e-02 -7.89249659e-01 5.02451301e-01
5.11529684e-01 4.98705238e-01 -1.38313726e-01 -5.39977252e-01
-8.10927153e-02 -2.32984915e-01 -2.34868765e-01 -5.47205150e-01
9.96024907e-01 -4.83253181e-01 -1.00587197e-01 8.36214483e-01
8.11405599e-01 -2.72942543e-01 -1.21407104e+00 -3.69683027e-01
3.36069316e-01 -3.01128387e-01 -3.35193388e-02 -1.06150854e+00
-6.71223581e-01 8.75766098e-01 9.30353031e-02 7.39617646e-01
7.31783569e-01 -3.12791348e-01 6.81760252e-01 5.48638642e-01
2.63613373e-01 -1.50349998e+00 3.63173075e-02 1.10491025e+00
1.04905570e+00 -1.30547667e+00 -1.80690825e-01 -2.54895002e-01
-1.08288729e+00 1.14292574e+00 9.06885207e-01 4.97750074e-01
3.58452350e-01 7.10284635e-02 2.71927953e-01 -1.66631013e-01
-1.13266289e+00 -1.62346333e-01 7.61075364e-03 4.93885040e-01
1.03587139e+00 -2.73598403e-01 -5.94985306e-01 4.67306763e-01
-3.92433256e-01 -1.69543147e-01 4.61132437e-01 8.95499766e-01
-4.50521231e-01 -1.38299203e+00 -7.48938993e-02 6.85498655e-01
-2.15556413e-01 2.56882291e-02 -9.57782269e-01 5.07718801e-01
-8.87688339e-01 1.49922061e+00 -2.19128951e-01 -4.67274100e-01
5.05168259e-01 4.65850621e-01 1.50467604e-01 -7.82160461e-01
-1.05018497e+00 -1.39375940e-01 8.50306869e-01 -4.64172751e-01
-2.42299691e-01 -9.37116385e-01 -1.06422830e+00 -3.89794558e-01
-2.98472464e-01 4.04025793e-01 6.17521703e-01 1.08182895e+00
2.99238443e-01 3.34017009e-01 6.90076888e-01 -6.63676679e-01
-8.33887160e-01 -1.38149595e+00 -2.42948636e-01 4.53146994e-02
2.25088462e-01 -4.07952696e-01 -1.15110129e-01 -3.38465184e-01] | [12.888623237609863, 7.985328674316406] |
75e60d44-2667-4d23-94e7-00f5ef8acdf1 | language-agnostic-semantic-consistent-text-to | null | null | https://aclanthology.org/2022.mml-1.1 | https://aclanthology.org/2022.mml-1.1.pdf | Language-agnostic Semantic Consistent Text-to-Image Generation | Recent GAN-based text-to-image generation models have advanced that they can generate photo-realistic images matching semantically with descriptions. However, research on multi-lingual text-to-image generation has not been carried out yet much. There are two problems when constructing a multilingual text-to-image generation model: 1) language imbalance issue in text-to-image paired datasets and 2) generating images that have the same meaning but are semantically inconsistent with each other in texts expressed in different languages. To this end, we propose a Language-agnostic Semantic Consistent Generative Adversarial Network (LaSC-GAN) for text-to-image generation, which can generate semantically consistent images via language-agnostic text encoder and Siamese mechanism. Experiments on relatively low-resource language text-image datasets show that the model has comparable generation quality as images generated by high-resource language text, and generates semantically consistent images for texts with the same meaning even in different languages. | ['Byoung-Tak Zhang', 'SeongHo Choi', 'Woo Suk Choi', 'SeongJun Jung'] | null | null | null | null | mml-acl-2022-5 | ['multilingual-text-to-image-generation', 'multi-lingual-text-to-image-generation'] | ['computer-vision', 'natural-language-processing'] | [ 4.40251082e-01 4.26246598e-02 9.91045311e-02 -4.15148348e-01
-1.21522307e+00 -7.36014187e-01 9.59948301e-01 -7.04330623e-01
-3.01263742e-02 9.69393075e-01 2.88367331e-01 -4.64263149e-02
6.50545537e-01 -9.40624595e-01 -9.56802905e-01 -5.67130804e-01
7.87421703e-01 8.77031505e-01 -1.87668353e-01 -2.24624172e-01
-3.12607169e-01 -1.84837251e-03 -1.33618760e+00 6.82835460e-01
9.67975199e-01 6.59431815e-01 2.90819466e-01 5.59121251e-01
-5.34672856e-01 9.84612226e-01 -8.22857916e-01 -9.73476768e-01
5.12658358e-01 -1.25559175e+00 -7.77774513e-01 2.33688399e-01
8.88011456e-01 -2.64050692e-01 -2.32043907e-01 1.20393968e+00
6.86105430e-01 -3.03549409e-01 8.29975545e-01 -1.81161249e+00
-1.42297697e+00 5.76624215e-01 -6.49327576e-01 -4.62137699e-01
4.04162675e-01 2.96399802e-01 4.64303762e-01 -6.70308769e-01
1.04278922e+00 1.64076495e+00 4.96612549e-01 8.82829010e-01
-1.04728723e+00 -8.77700210e-01 -2.34967977e-01 -4.99045737e-02
-1.46199572e+00 -2.98150420e-01 6.35510743e-01 -2.65575498e-01
6.10622823e-01 2.14483634e-01 6.54837191e-01 1.59384227e+00
3.65787774e-01 6.25153422e-01 1.48790574e+00 -5.26284099e-01
-1.43348023e-01 3.49413484e-01 -1.20160627e+00 6.62313879e-01
2.33657986e-01 5.43957874e-02 -5.43503165e-01 1.98798552e-01
7.65891314e-01 -3.78411442e-01 -8.77389833e-02 9.01304372e-03
-1.33904243e+00 1.01409090e+00 2.37745270e-01 3.28594774e-01
-1.03874870e-01 4.20108110e-01 4.43810970e-01 5.30830085e-01
6.29651070e-01 2.43690684e-01 1.20651079e-02 2.98147410e-01
-1.00786078e+00 4.89847422e-01 2.91277915e-01 1.60048079e+00
7.79454410e-01 6.61281228e-01 -4.19324785e-01 7.14090228e-01
1.52880654e-01 1.14413750e+00 7.02600777e-01 -6.48425221e-01
7.22674370e-01 3.07317168e-01 -1.25894457e-01 -7.57231414e-01
3.31457496e-01 1.41504765e-01 -1.22783601e+00 1.41874224e-01
1.02518119e-01 -3.71220648e-01 -1.07740104e+00 1.78740442e+00
1.18008949e-01 8.85610357e-02 4.77287203e-01 7.22314537e-01
1.16948128e+00 8.01503479e-01 3.26374710e-01 1.90099970e-01
1.52940726e+00 -1.11393440e+00 -8.33735704e-01 -3.71229678e-01
3.36661905e-01 -1.26470649e+00 1.23788786e+00 -2.47272149e-01
-1.18430722e+00 -7.89285302e-01 -8.26021612e-01 -2.82537699e-01
-5.25853634e-01 1.02928961e-02 4.38297153e-01 8.39094043e-01
-1.27127755e+00 -1.85390100e-01 -6.48448169e-02 -4.90982860e-01
3.22844267e-01 -3.06642443e-01 -4.60466236e-01 -2.89921373e-01
-1.41012311e+00 8.97988498e-01 7.45590210e-01 -4.04015154e-01
-1.02562129e+00 -5.35239100e-01 -1.15976799e+00 -4.60874617e-01
8.18284322e-03 -1.33440530e+00 1.07623363e+00 -1.97111773e+00
-1.31556261e+00 1.30896306e+00 -2.62124270e-01 -4.53243911e-01
9.21861768e-01 3.57048184e-01 -5.07957757e-01 2.52609029e-02
6.09219790e-01 1.51786816e+00 1.15580499e+00 -1.58293533e+00
-5.18724322e-01 -1.07909039e-01 -1.99626312e-01 4.61131245e-01
1.20478258e-01 6.17123805e-02 -5.71967900e-01 -1.27622283e+00
-3.08891684e-01 -1.04027069e+00 -4.52498198e-02 6.48122653e-02
-5.70641398e-01 6.44696876e-02 8.22863936e-01 -7.70744562e-01
3.41975331e-01 -1.67469835e+00 -1.80133283e-01 -2.77918875e-01
-3.15329820e-01 -3.51201557e-02 -6.74929142e-01 6.14774048e-01
-2.05675676e-01 4.33552295e-01 -2.29838356e-01 -5.51876128e-01
-1.49350697e-02 2.53527522e-01 -6.33015692e-01 4.01810780e-02
2.16377497e-01 1.52984667e+00 -8.79369497e-01 -9.80759621e-01
3.04530859e-01 4.47607338e-01 -2.57417828e-01 3.50154996e-01
-4.79105622e-01 7.25843012e-01 -4.67772663e-01 5.75508296e-01
7.29884923e-01 2.38313377e-02 -1.08526237e-01 -4.00243670e-01
3.97427559e-01 -4.52338427e-01 -5.79228580e-01 1.97937202e+00
-6.31784081e-01 9.29674149e-01 -4.96332973e-01 -6.99973822e-01
8.00908864e-01 6.67331636e-01 2.47561932e-01 -1.09684527e+00
-5.71269393e-02 1.85627371e-01 -4.55045074e-01 -4.19926077e-01
6.77622914e-01 -6.78096592e-01 -3.05144757e-01 4.35922325e-01
1.42217129e-01 -1.05506134e+00 2.06729054e-01 2.65289724e-01
2.37966359e-01 4.19611782e-01 -1.81043163e-01 -5.22450358e-02
5.07257283e-01 2.74814993e-01 2.55991548e-01 7.00946093e-01
1.72917441e-01 1.14151013e+00 7.67160580e-02 -3.25661719e-01
-1.73391318e+00 -1.11217475e+00 1.55744359e-01 5.39075613e-01
1.78353816e-01 -3.45561542e-02 -1.14254975e+00 -6.70751572e-01
-3.61450106e-01 8.95569324e-01 -5.16152382e-01 5.64764589e-02
-1.78110406e-01 -4.49575007e-01 1.11021411e+00 2.49132827e-01
1.11247945e+00 -1.29762959e+00 1.42128229e-01 -2.30621528e-02
-7.21130550e-01 -1.57757378e+00 -9.43041086e-01 -7.57700443e-01
-4.29460496e-01 -8.41631949e-01 -1.20050275e+00 -1.24340630e+00
1.11288893e+00 9.76232216e-02 1.43083942e+00 -2.07696825e-01
-3.59811336e-01 4.10885870e-01 -3.87839317e-01 -5.73412418e-01
-1.12640309e+00 -3.40079308e-01 -2.70128220e-01 6.48775920e-02
-5.64566143e-02 -3.59748602e-02 -4.97491241e-01 3.07321370e-01
-1.49182391e+00 8.36840689e-01 4.11735147e-01 8.18160415e-01
7.72911131e-01 2.70161182e-01 8.07536125e-01 -9.71137464e-01
6.15430415e-01 -2.68824965e-01 -4.92601335e-01 6.23115897e-01
-5.04106760e-01 9.90004689e-02 1.06634736e+00 -3.99732232e-01
-1.59274828e+00 1.02394661e-02 -2.67191529e-02 -5.25007010e-01
-2.16118500e-01 9.90099311e-02 -2.64670759e-01 5.55448160e-02
4.25850213e-01 8.00756574e-01 -1.74709573e-01 1.91093370e-01
8.53861630e-01 6.22094929e-01 8.09312463e-01 -7.40021527e-01
1.27658761e+00 6.83378637e-01 -1.87114760e-01 -5.28296947e-01
-5.73910356e-01 2.00047925e-01 -2.70576835e-01 -1.59254476e-01
1.39616168e+00 -1.50426197e+00 4.37335260e-02 9.10705745e-01
-1.24490464e+00 -4.25850719e-01 -4.21710193e-01 1.25061989e-01
-8.89509797e-01 3.42066258e-01 -4.13389772e-01 -2.98890948e-01
-8.53798628e-01 -1.26037025e+00 1.48915291e+00 2.18024209e-01
1.22528002e-01 -1.12897480e+00 -5.52450269e-02 7.84395993e-01
4.82538968e-01 5.00570714e-01 8.32633793e-01 -3.86172272e-02
-6.32778525e-01 -1.04436800e-01 -3.75262111e-01 3.86994570e-01
2.61106700e-01 -6.28541857e-02 -7.49884427e-01 -4.11484390e-01
-2.29562432e-01 -6.45097136e-01 4.44887877e-01 2.73626685e-01
1.01037288e+00 -4.72207993e-01 -7.35369232e-03 7.68451035e-01
1.65953159e+00 1.17786050e-01 1.10745490e+00 1.81729887e-02
9.32316184e-01 5.68675697e-01 4.50360298e-01 -5.72597682e-02
5.74632347e-01 6.70294106e-01 2.12444253e-02 -6.35562241e-01
-7.83392251e-01 -9.89952803e-01 4.70195293e-01 6.49950206e-01
3.36889714e-01 -9.33030307e-01 -3.56038302e-01 6.79621041e-01
-1.58535147e+00 -1.21479988e+00 -3.16076458e-01 1.89773893e+00
1.03581274e+00 -3.25616091e-01 -2.70068049e-01 -6.92876160e-01
9.83389735e-01 1.50142595e-01 -4.76617098e-01 -4.15138155e-01
-7.02314377e-01 1.60683185e-01 6.11549318e-01 2.00510040e-01
-8.46967936e-01 1.44677210e+00 6.11252069e+00 1.41783917e+00
-1.06742275e+00 2.55746633e-01 1.16760039e+00 2.12587938e-01
-8.51603985e-01 1.47085592e-01 -4.64943707e-01 6.30601346e-01
5.15960097e-01 -5.47792554e-01 3.56219471e-01 6.87411666e-01
2.18643114e-01 9.24583748e-02 -7.62488723e-01 1.30401313e+00
6.39070570e-01 -1.35842943e+00 9.72402096e-01 -2.37155437e-01
1.58368063e+00 -2.08280697e-01 1.57690883e-01 9.83994380e-02
6.91527843e-01 -1.30484641e+00 1.16428888e+00 3.11036855e-01
1.57227576e+00 -7.87115216e-01 6.26011014e-01 2.92571019e-02
-1.27668941e+00 6.73880041e-01 -5.28430760e-01 6.74755156e-01
4.24109578e-01 3.67487788e-01 -7.91075289e-01 9.10279989e-01
5.52806556e-01 5.13677776e-01 -7.20856488e-01 3.11129272e-01
-3.41731459e-01 2.38823771e-01 6.39021546e-02 3.38270634e-01
3.38967264e-01 -3.73931408e-01 1.61650255e-01 9.48951364e-01
7.22887278e-01 -3.39281321e-01 6.06044941e-02 1.32904255e+00
-3.17144901e-01 2.86667615e-01 -1.17059076e+00 -5.15713021e-02
4.00200576e-01 1.12531722e+00 -6.17191732e-01 -6.94655895e-01
-4.28252369e-01 1.67681766e+00 -2.18469024e-01 4.93662059e-01
-1.02348793e+00 -1.72798991e-01 1.98198140e-01 -5.74810663e-04
-1.27817288e-01 1.00878030e-01 -1.10291176e-01 -1.27516890e+00
4.30953465e-02 -1.24634707e+00 2.11515844e-01 -1.58657897e+00
-1.62182987e+00 8.15476835e-01 1.44428769e-02 -1.34109008e+00
-5.61498046e-01 -1.31869033e-01 -6.14311934e-01 1.19281197e+00
-1.45262003e+00 -2.01308656e+00 -4.27793264e-01 1.00631177e+00
9.18957174e-01 -5.45990825e-01 7.82223761e-01 1.62458181e-01
-1.01561554e-01 8.62837374e-01 -6.82127057e-03 3.05668324e-01
1.00035799e+00 -1.06208801e+00 9.67983425e-01 9.81253266e-01
2.90812224e-01 1.54260695e-01 4.32731003e-01 -9.08645451e-01
-1.31025255e+00 -1.67235053e+00 8.58280063e-01 -4.81405467e-01
2.41148919e-01 -1.77431479e-01 -3.63030016e-01 7.76234031e-01
1.00703442e+00 -3.80527645e-01 2.90905267e-01 -9.97036576e-01
-3.11905146e-01 6.10344335e-02 -1.42673528e+00 1.08743656e+00
1.01080942e+00 -6.39658689e-01 -1.29274875e-01 6.50294423e-01
7.19347656e-01 -4.50106144e-01 -6.30771339e-01 2.22945690e-01
2.87156850e-01 -8.13281596e-01 1.07181823e+00 -3.30629498e-01
7.90853024e-01 -4.80261832e-01 -1.10592628e-02 -1.35292315e+00
1.99321136e-01 -7.66434014e-01 1.05240440e+00 1.52527213e+00
2.30421603e-01 -6.70517266e-01 5.96915483e-01 3.27683598e-01
-5.26855439e-02 -5.62211350e-02 -8.44632566e-01 -8.91025424e-01
5.13140619e-01 -2.07765713e-01 9.74502742e-01 1.11820185e+00
-7.52179801e-01 3.78894389e-01 -7.49708056e-01 -2.01682076e-01
6.73129559e-01 1.84128523e-01 1.08448887e+00 -3.76037061e-01
2.96068098e-02 -3.29957455e-01 -2.29312748e-01 -6.84665382e-01
5.48532903e-01 -1.17011082e+00 6.19167350e-02 -1.56936705e+00
5.35223067e-01 -3.94271165e-01 5.81373990e-01 5.15210629e-01
-2.87366122e-01 8.11045766e-01 1.74573064e-01 1.26675278e-01
-2.02912480e-01 8.00339043e-01 1.80121744e+00 -5.24010956e-01
6.03403270e-01 -6.43637121e-01 -5.46828985e-01 4.82858181e-01
7.71194100e-01 -4.89941537e-01 -8.05478036e-01 -6.62656665e-01
1.35044426e-01 1.52288944e-01 4.93140340e-01 -8.35271060e-01
-1.97663590e-01 -3.82857740e-01 5.85845172e-01 -5.21074653e-01
1.65946618e-01 -7.46310890e-01 9.65571344e-01 2.76327878e-01
-4.01894450e-01 3.87320161e-01 1.56784967e-01 3.34570557e-01
-4.46454287e-01 -2.47779533e-01 8.72053206e-01 -4.88741934e-01
-6.97621047e-01 4.55699921e-01 -2.64927566e-01 6.18805885e-01
1.09117734e+00 -3.16249371e-01 -4.27548885e-01 -9.31972682e-01
3.90272252e-02 1.04866885e-02 9.31689203e-01 8.99021447e-01
6.71568871e-01 -1.90895903e+00 -1.46815336e+00 1.88498408e-01
4.25205410e-01 -1.80880979e-01 4.04664785e-01 -7.64365122e-02
-8.30063820e-01 4.21468228e-01 -2.45249256e-01 -5.04414320e-01
-1.22744966e+00 4.52931553e-01 5.05587578e-01 -2.08040714e-01
-4.57175165e-01 6.47752464e-01 6.50042236e-01 -6.51150346e-01
-3.99230868e-01 8.30540806e-02 3.78815532e-01 -2.94861615e-01
2.33690783e-01 -2.39364579e-01 -1.55808762e-01 -1.12174571e+00
7.49627054e-02 8.05410743e-01 3.12475383e-01 -4.82675493e-01
6.60817862e-01 -1.43964186e-01 -7.57449493e-02 -9.77319404e-02
1.12124741e+00 -1.27735049e-01 -9.45092916e-01 -8.78828540e-02
-6.61681950e-01 -6.94906354e-01 -3.34675074e-01 -8.94945383e-01
-1.56834686e+00 5.10285258e-01 7.32192814e-01 -2.13405848e-01
1.15658128e+00 -1.25337586e-01 1.22152162e+00 -8.90424922e-02
4.91263986e-01 -9.74682868e-01 2.98803806e-01 1.82123467e-01
1.26048267e+00 -1.38476861e+00 -2.09718123e-01 -2.29092672e-01
-1.18131375e+00 7.39173532e-01 7.80854046e-01 1.31386518e-01
-8.12731013e-02 7.28245378e-02 4.50853974e-01 7.56204128e-02
-5.48776209e-01 -2.14831427e-01 3.64889979e-01 1.12646639e+00
4.66116935e-01 1.33789971e-01 -4.17107865e-02 -1.55088246e-01
-4.80229855e-01 -7.82306716e-02 4.63775784e-01 3.50857198e-01
3.34357977e-01 -1.43046582e+00 -5.22496164e-01 1.37058958e-01
-5.27076542e-01 -5.90044439e-01 -5.11675119e-01 7.50366569e-01
3.12230080e-01 1.11056626e+00 1.46984935e-01 -7.33350888e-02
-4.06797566e-02 9.13505852e-02 7.09575474e-01 -4.17547047e-01
-4.92128789e-01 -5.04910387e-02 1.20951772e-01 -2.10269526e-01
-4.84823376e-01 -2.42383167e-01 -1.04780245e+00 -4.38644707e-01
-4.83443849e-02 -2.97610983e-02 7.54466414e-01 6.56491816e-01
4.34816122e-01 3.72657955e-01 6.64734840e-01 -5.99072874e-01
-8.50890875e-02 -7.79530227e-01 -4.91458386e-01 1.13545394e+00
-2.71064341e-01 -6.58932477e-02 -2.35466883e-02 7.67631590e-01] | [12.068793296813965, -0.1793341040611267] |
417bbcfd-836f-4476-a77e-d13b71249745 | encore-ensemble-learning-using-convolution | 1906.08691 | null | https://arxiv.org/abs/1906.08691v1 | https://arxiv.org/pdf/1906.08691v1.pdf | ENCORE: Ensemble Learning using Convolution Neural Machine Translation for Automatic Program Repair | Automated generate-and-validate (G&V) program repair techniques typically rely on hard-coded rules, only fix bugs following specific patterns, and are hard to adapt to different programming languages. We propose ENCORE, a new G&V technique, which uses ensemble learning on convolutional neural machine translation (NMT) models to automatically fix bugs in multiple programming languages. We take advantage of the randomness in hyper-parameter tuning to build multiple models that fix different bugs and combine them using ensemble learning. This new convolutional NMT approach outperforms the standard long short-term memory (LSTM) approach used in previous work, as it better captures both local and long-distance connections between tokens. Our evaluation on two popular benchmarks, Defects4J and QuixBugs, shows that ENCORE fixed 42 bugs, including 16 that have not been fixed by existing techniques. In addition, ENCORE is the first G&V repair technique to be applied to four popular programming languages (Java, C++, Python, and JavaScript), fixing a total of 67 bugs across five benchmarks. | ['Moshi Wei', 'Thibaud Lutellier', 'Lawrence Pang', 'Viet Hung Pham', 'Lin Tan'] | 2019-06-20 | null | null | null | null | ['program-repair', 'program-repair'] | ['computer-code', 'reasoning'] | [-3.35617512e-01 -3.67081732e-01 -3.32292736e-01 -1.64777145e-01
-1.04724610e+00 -8.12717974e-01 3.16190459e-02 2.70758539e-01
-1.54974476e-01 5.80391645e-01 -1.96844831e-01 -9.34301019e-01
2.01033533e-01 -8.72678459e-01 -1.45558560e+00 6.40037730e-02
-1.97704002e-01 -2.08866656e-01 2.99226135e-01 -2.40837306e-01
5.19884408e-01 -3.16117734e-01 -1.33471251e+00 7.74760246e-01
1.04799664e+00 1.23088025e-01 1.23229891e-01 1.02204132e+00
-3.29190642e-01 8.81403089e-01 -8.29863787e-01 -6.42075539e-01
-1.68034792e-01 5.24326898e-02 -8.72053325e-01 -1.05726731e+00
7.00621068e-01 -8.82925689e-02 -6.35485724e-03 1.15933061e+00
4.21938032e-01 -5.15714824e-01 -9.94548574e-02 -1.10378492e+00
-1.36418843e+00 1.41778564e+00 -5.65855861e-01 1.94649398e-01
2.95797527e-01 3.16365689e-01 8.39828491e-01 -7.51006544e-01
5.45170605e-01 1.16977632e+00 1.39472950e+00 8.34105909e-01
-1.29353809e+00 -7.45943069e-01 -2.00753495e-01 -3.82963642e-02
-1.31115639e+00 8.81760716e-02 3.22776198e-01 -6.71232879e-01
1.98640573e+00 1.05703741e-01 8.94505456e-02 1.36737061e+00
1.02746904e+00 2.86713332e-01 8.12951744e-01 -7.09011734e-01
9.74064469e-02 -2.06545934e-01 6.61154211e-01 1.04266202e+00
2.42160216e-01 1.69152901e-01 -1.46786541e-01 -7.24433243e-01
1.29186437e-01 4.82110940e-02 -7.22475275e-02 3.08011949e-01
-1.18785906e+00 7.82565236e-01 4.04017717e-01 5.10772645e-01
1.98571011e-01 9.72617805e-01 8.75561178e-01 6.91964328e-01
1.45288870e-01 8.70314062e-01 -1.00883365e+00 -4.48135167e-01
-6.57848179e-01 2.46977150e-01 7.52879381e-01 8.80897999e-01
9.40880299e-01 2.00962931e-01 -1.98587626e-01 7.13110745e-01
1.67543426e-01 4.83407080e-01 6.79976940e-01 -6.27538741e-01
8.28173578e-01 8.32915545e-01 -1.39392704e-01 -8.33901167e-01
-9.22200456e-02 -2.80732960e-01 -2.61071980e-01 2.74915516e-01
-9.93467271e-02 -2.33747080e-01 -9.01935160e-01 1.73535883e+00
-2.93601185e-01 1.64878666e-01 -2.68345568e-02 2.29079261e-01
7.74290621e-01 3.95023644e-01 -1.61205739e-01 4.92300361e-01
6.99803054e-01 -9.18525636e-01 -2.18461126e-01 -4.51770216e-01
1.23718548e+00 -9.22680616e-01 1.30969357e+00 4.30549115e-01
-8.79182041e-01 -5.64491093e-01 -1.23135936e+00 -3.83058302e-02
-5.61313570e-01 2.45560393e-01 6.42838240e-01 6.45129025e-01
-1.40612769e+00 7.44426608e-01 -9.97927368e-01 -2.22211719e-01
-5.55582158e-03 4.58263040e-01 -3.00570399e-01 3.93982008e-02
-7.72130489e-01 7.20902383e-01 5.10034502e-01 -1.43483400e-01
-9.18901384e-01 -9.16553974e-01 -9.29967105e-01 6.76601306e-02
2.58651644e-01 -4.75679666e-01 1.52293730e+00 -7.02322423e-01
-1.17324054e+00 5.77845156e-01 5.28990589e-02 -4.11358029e-01
-9.07427669e-02 -4.16482568e-01 -3.85677576e-01 -7.80654490e-01
1.61606014e-01 4.48352993e-01 4.97507751e-01 -8.39113414e-01
-3.15941483e-01 1.22270472e-01 3.10858101e-01 -1.07529664e+00
-2.51945883e-01 4.79296297e-01 -1.56600952e-01 -6.33250237e-01
-5.08866072e-01 -1.00676441e+00 6.73424080e-02 -4.30712521e-01
-5.96162975e-01 -1.78809077e-01 6.03563249e-01 -8.92910480e-01
1.50226068e+00 -2.05699086e+00 4.44845706e-01 -1.25852050e-02
2.32579127e-01 4.73998606e-01 -6.40067756e-01 4.67330068e-01
-3.60528648e-01 7.42473483e-01 -3.67671132e-01 -1.78416565e-01
1.46254182e-01 2.75666088e-01 -4.11975980e-01 -2.61560064e-02
3.00942212e-01 1.14114547e+00 -8.74941945e-01 -1.35956615e-01
-2.60997713e-01 2.38148481e-01 -1.05064940e+00 -3.19718719e-02
-9.62114275e-01 -1.55349270e-01 -6.45726919e-02 9.16321576e-01
6.14659071e-01 -2.06486523e-01 -6.35191053e-02 4.50418383e-01
-4.27470565e-01 3.36599261e-01 -6.20012403e-01 1.82958126e+00
-8.19613934e-01 6.24546528e-01 -4.05933350e-01 -3.55178267e-01
9.68460381e-01 1.94472790e-01 -1.89872950e-01 -5.36787271e-01
-2.91417629e-01 7.92700112e-01 6.26772270e-02 -7.94428051e-01
5.34633577e-01 9.43771422e-01 -6.16403341e-01 7.33764946e-01
2.02386484e-01 2.16624774e-02 2.49096885e-01 4.03766632e-02
1.75136340e+00 2.78891593e-01 -5.83844297e-02 5.85470442e-03
1.39144182e-01 -2.35718906e-01 5.92402995e-01 1.12164640e+00
3.47180456e-01 5.67726433e-01 6.79208040e-01 -6.77259326e-01
-1.15474343e+00 -6.60507321e-01 2.72188038e-01 1.05608296e+00
-6.17180049e-01 -8.83643091e-01 -1.02372932e+00 -9.33439910e-01
-9.52349510e-03 8.28002930e-01 -7.58891940e-01 -4.69533682e-01
-8.75862837e-01 -6.07386529e-01 1.21323216e+00 6.39903665e-01
2.67293096e-01 -1.19636691e+00 -4.55254108e-01 5.27073264e-01
-2.02652052e-01 -5.49530506e-01 -6.98807120e-01 3.91994715e-01
-4.37023342e-01 -1.12740123e+00 2.34601926e-02 -7.02205539e-01
2.29220018e-01 -9.41792503e-02 1.45120203e+00 8.53529274e-01
-4.35119867e-01 2.08305880e-01 -3.60256702e-01 -1.76457509e-01
-9.73813295e-01 3.29522878e-01 -1.77685797e-01 -7.64066279e-01
5.58857024e-01 -5.37088811e-01 2.09786430e-01 -8.20800755e-03
-8.05053294e-01 -3.85316670e-01 4.72265333e-01 1.10498202e+00
1.47906527e-01 -2.27166817e-01 7.56841898e-02 -8.70922089e-01
6.70102060e-01 -8.62641215e-01 -9.80009198e-01 6.14601672e-01
-5.64477384e-01 1.93026066e-01 6.51301861e-01 -6.56889498e-01
-6.34702444e-01 -3.49522263e-01 -2.96378225e-01 -5.37193239e-01
2.26827353e-01 1.06500995e+00 3.01228315e-01 -5.40504098e-01
1.10920835e+00 1.41167194e-01 -4.35168505e-01 -5.69630742e-01
1.37741461e-01 5.91330707e-01 6.77829266e-01 -1.01138449e+00
6.94385350e-01 -3.77586663e-01 -7.09973216e-01 -1.67934701e-01
-4.01424289e-01 3.15172106e-01 -4.30897504e-01 2.19818294e-01
6.48431718e-01 -5.61767399e-01 -4.58854407e-01 8.21843088e-01
-1.80159426e+00 -6.38134360e-01 2.06258789e-01 5.91412783e-02
-1.31913781e-01 2.73335427e-01 -1.02329719e+00 -2.81435370e-01
-6.07140303e-01 -1.54793060e+00 1.07019246e+00 4.09173220e-02
-4.68370885e-01 -7.86728501e-01 5.95088601e-01 -1.36671677e-01
8.04942071e-01 2.14648947e-01 1.55313122e+00 -2.50520349e-01
-8.15959275e-01 -1.22875087e-01 -2.91660447e-02 4.53141540e-01
-9.80053395e-02 8.12016487e-01 -5.90031385e-01 -4.83858377e-01
-4.16318774e-01 -3.84059936e-01 8.63451779e-01 1.64474882e-02
1.39151156e+00 -5.32650828e-01 -2.81422794e-01 6.43703043e-01
1.74531388e+00 -5.28100878e-02 8.63576055e-01 5.86034119e-01
7.76635706e-01 -1.33965656e-01 1.03554569e-01 1.12140551e-01
4.91548449e-01 6.52013600e-01 7.53907979e-01 3.28957289e-01
-5.32298572e-02 -3.12341779e-01 8.85377765e-01 1.09542620e+00
3.38364005e-01 -1.52602717e-01 -1.36268365e+00 6.08842254e-01
-1.85049784e+00 -7.31833637e-01 -4.98357952e-01 2.09301233e+00
1.07670259e+00 1.54102728e-01 -3.37186366e-01 -2.29199044e-02
5.61276734e-01 -3.93200010e-01 -2.34372050e-01 -1.06647515e+00
1.78252339e-01 7.61536896e-01 3.79090637e-01 3.15263450e-01
-9.69780624e-01 1.00319147e+00 6.62661457e+00 5.20343840e-01
-1.49221182e+00 4.99465793e-01 1.71919148e-02 9.39214975e-02
-5.60327947e-01 2.83809245e-01 -8.24789703e-01 7.03021407e-01
1.27879655e+00 6.82934672e-02 6.15700841e-01 1.05920351e+00
-2.97906876e-01 1.95069224e-01 -1.28540099e+00 4.55605060e-01
2.56263558e-02 -1.68671250e+00 -3.76593322e-01 -2.77471095e-01
1.04017866e+00 6.17553651e-01 2.52798855e-01 8.32535982e-01
7.87313819e-01 -1.06939280e+00 8.64617646e-01 7.48475552e-01
7.42722452e-01 -6.91989124e-01 8.35559011e-01 5.46703674e-02
-7.94749796e-01 -3.15468490e-01 -5.45228302e-01 4.75717001e-02
-5.59768319e-01 5.59383452e-01 -6.48341715e-01 2.07372949e-01
1.31964576e+00 6.05559170e-01 -1.16880763e+00 1.11424983e+00
-3.34462374e-01 8.37759733e-01 5.12324795e-02 -2.10363790e-01
1.39809757e-01 5.47741950e-01 4.90731984e-01 1.43479419e+00
6.84832454e-01 -8.58608007e-01 1.22442082e-01 1.59736609e+00
-2.44126752e-01 -3.16616565e-01 -5.36612034e-01 -2.57463723e-01
5.42132497e-01 9.27557468e-01 -1.16260238e-01 -7.95879364e-02
-5.31244874e-01 6.18026376e-01 7.06726193e-01 1.53681576e-01
-1.12675118e+00 -6.44321620e-01 5.73826313e-01 -3.38466585e-01
5.11569917e-01 -3.39382410e-01 -3.38752747e-01 -1.17911577e+00
4.92798686e-01 -1.42993665e+00 1.99861020e-01 -8.43093097e-01
-1.05557168e+00 5.80920517e-01 -1.15794651e-01 -8.31746042e-01
-3.26644301e-01 -3.76360208e-01 -1.04741192e+00 8.09257805e-01
-1.20565867e+00 -1.36043096e+00 -3.42600048e-02 6.06861748e-02
3.00696015e-01 -3.27365786e-01 1.02282012e+00 5.76840281e-01
-8.28209639e-01 1.21463966e+00 2.13712618e-01 1.23989642e-01
6.76324487e-01 -1.41978300e+00 1.12548864e+00 1.22542632e+00
-1.28643930e-01 1.26631069e+00 4.12278980e-01 -8.73729110e-01
-1.69940650e+00 -1.68439353e+00 1.03301024e+00 -7.32043147e-01
8.87606323e-01 -3.99203628e-01 -1.50038648e+00 1.11373150e+00
3.56107950e-01 4.84271757e-02 3.86310488e-01 1.50843844e-01
-1.18080151e+00 1.38524801e-01 -1.02246988e+00 4.89943892e-01
5.78446031e-01 -7.55196273e-01 -3.00642371e-01 4.09175068e-01
1.24857736e+00 -6.25655532e-01 -1.05435288e+00 2.85143793e-01
3.97719294e-01 -7.98871875e-01 6.52738154e-01 -6.44081235e-01
9.88511920e-01 -2.96723157e-01 -3.64026129e-01 -1.35882545e+00
-3.85707796e-01 -6.64575517e-01 -2.12353647e-01 1.42014527e+00
7.86412597e-01 -8.08230460e-01 1.38198107e-01 2.58269787e-01
-5.52069128e-01 -5.08239388e-01 -7.44241297e-01 -7.50671864e-01
5.11959732e-01 -6.10728621e-01 1.06137669e+00 1.13040686e+00
-2.66754236e-02 -2.31561199e-01 -4.55766261e-01 2.78106183e-01
2.29878128e-01 -3.91681306e-02 8.90145719e-01 -9.00642097e-01
-9.42031503e-01 -3.95856440e-01 -2.11322829e-01 -2.88716048e-01
4.33377802e-01 -1.09826338e+00 1.32987902e-01 -8.44266176e-01
4.05001134e-01 -4.81066823e-01 -3.17594737e-01 1.36067760e+00
-3.07097942e-01 1.79419309e-01 -2.49411896e-01 3.91726904e-02
-4.10696864e-01 -2.30904315e-02 3.06637973e-01 -6.86791062e-01
-4.34418172e-02 -2.99342513e-01 -6.10227406e-01 5.13896048e-01
7.48102129e-01 -9.04328763e-01 1.85014680e-01 -1.25576818e+00
9.09834087e-01 -9.42394063e-02 5.92177033e-01 -1.12199640e+00
1.37711093e-01 -1.52985826e-02 -1.50923640e-01 -3.14615518e-01
-4.31410939e-01 -1.92452058e-01 4.17659223e-01 6.78901613e-01
-4.35119450e-01 7.43297219e-01 7.54606724e-01 1.81052282e-01
8.33269395e-03 -6.55520082e-01 7.18357563e-01 -3.26666713e-01
-8.34981263e-01 5.28187025e-03 -3.83609712e-01 3.96503229e-03
8.14726949e-01 2.74371862e-01 -9.18250740e-01 3.79766166e-01
-1.35754272e-01 8.08460191e-02 7.65682161e-01 1.07622159e+00
6.18788242e-01 -1.20645297e+00 -8.49940836e-01 2.26088747e-01
5.10936379e-01 -4.27580953e-01 1.07004428e-02 5.32539070e-01
-6.73796237e-01 2.86368400e-01 -1.41859353e-01 -7.44105339e-01
-1.32488370e+00 6.73546970e-01 5.83343089e-01 -2.40645885e-01
-5.55795014e-01 9.15949106e-01 -5.85938036e-01 -1.25606716e+00
-7.93371648e-02 -8.21343422e-01 3.11884612e-01 -8.28052104e-01
5.24680376e-01 1.15159303e-01 5.16309917e-01 -7.54471198e-02
-3.38849753e-01 5.23581982e-01 -3.59198183e-01 3.84683222e-01
1.53437924e+00 6.73139691e-01 -9.41657722e-01 4.72171187e-01
1.30462313e+00 2.47983202e-01 -4.79202479e-01 -1.73480213e-01
3.19395036e-01 -4.35178518e-01 -2.45280042e-01 -1.09550512e+00
-9.91333723e-01 8.34105909e-01 6.19039536e-01 -5.83752654e-02
6.05474412e-01 -3.02958488e-01 8.79410505e-01 6.23821437e-01
7.36735046e-01 -5.81634045e-01 2.44795904e-01 9.66643095e-01
7.84744322e-01 -8.59431088e-01 -6.36637747e-01 8.39278102e-02
3.28018159e-01 1.30025434e+00 1.13419330e+00 -1.66728735e-01
1.65752202e-01 7.85535812e-01 -1.40340373e-01 -6.43823063e-03
-1.10100996e+00 5.81148386e-01 2.44879741e-02 5.50293922e-01
6.64223731e-01 4.56238650e-02 -4.92074937e-02 4.09094423e-01
-1.24866657e-01 2.04002246e-01 1.05740595e+00 1.23929954e+00
-1.48292094e-01 -1.59131396e+00 -5.51250458e-01 5.57017803e-01
-6.26438856e-01 -4.26889926e-01 -3.36085886e-01 6.74309313e-01
4.99357045e-01 8.43141496e-01 -2.45221362e-01 -9.33534384e-01
1.40105739e-01 5.66857196e-02 5.15282154e-01 -9.47453260e-01
-1.29502797e+00 -5.14196217e-01 -6.62866607e-02 -7.11789072e-01
2.98518777e-01 -2.83041567e-01 -9.07765090e-01 -6.67810559e-01
-4.37416434e-01 -1.13538139e-01 7.29500771e-01 8.23376775e-01
8.00763667e-01 9.04047668e-01 9.56315771e-02 -4.34208751e-01
-6.89387560e-01 -8.86440694e-01 3.33910465e-01 -1.18616782e-02
5.05426049e-01 -4.07992780e-01 -2.16437459e-01 -2.00372025e-01] | [7.609827995300293, 7.734960079193115] |
90c06851-7bc7-416e-a98e-4ee1f6374c33 | video-pose-track-with-graph-guided-sparse | 2303.00138 | null | https://arxiv.org/abs/2303.00138v3 | https://arxiv.org/pdf/2303.00138v3.pdf | Texture-Based Input Feature Selection for Action Recognition | The performance of video action recognition has been significantly boosted by using motion representations within a two-stream Convolutional Neural Network (CNN) architecture. However, there are a few challenging problems in action recognition in real scenarios, e.g., the variations in viewpoints and poses, and the changes in backgrounds. The domain discrepancy between the training data and the test data causes the performance drop. To improve the model robustness, we propose a novel method to determine the task-irrelevant content in inputs which increases the domain discrepancy. The method is based on a human parsing model (HP model) which jointly conducts dense correspondence labelling and semantic part segmentation. The predictions from the HP model also function as re-rendering the human regions in each video using the same set of textures to make humans appearances in all classes be the same. A revised dataset is generated for training and testing and makes the action recognition model exhibit invariance to the irrelevant content in the inputs. Moreover, the predictions from the HP model are used to enrich the inputs to the AR model during both training and testing. Experimental results show that our proposed model is superior to existing models for action recognition on the HMDB-51 dataset and the Penn Action dataset. | ['Yalong Jiang'] | 2023-02-28 | null | null | null | null | ['human-parsing', 'multi-person-pose-estimation', 'multi-person-pose-estimation-and-tracking', 'graph-matching'] | ['computer-vision', 'computer-vision', 'computer-vision', 'graphs'] | [ 5.48392892e-01 6.41962364e-02 -1.30952463e-01 -4.76173431e-01
-2.99701840e-01 -2.05880091e-01 3.23290169e-01 -7.18417704e-01
-3.18765759e-01 4.79915917e-01 2.22033471e-01 2.51289219e-01
4.43324745e-01 -5.84138334e-01 -8.99676919e-01 -7.53876746e-01
4.18150783e-01 2.76538551e-01 9.28986967e-01 -1.05737112e-01
-2.58874856e-02 4.03964370e-01 -1.82100391e+00 9.11414206e-01
5.78853846e-01 1.24423838e+00 3.14941734e-01 5.76977253e-01
-1.88177943e-01 1.14554501e+00 -6.93709254e-01 -2.09971637e-01
6.43229246e-01 -6.72597229e-01 -7.48670101e-01 6.56761467e-01
4.17194694e-01 -7.85369396e-01 -6.05976105e-01 1.10126412e+00
2.87736773e-01 4.49633330e-01 4.38552171e-01 -1.43759012e+00
-4.50366497e-01 1.11559220e-01 -6.32802427e-01 1.88971847e-01
3.33489120e-01 2.14415506e-01 3.63219142e-01 -7.24831522e-01
6.18830562e-01 1.62028468e+00 2.55795270e-01 8.69263053e-01
-5.69239616e-01 -5.93895197e-01 5.21176159e-01 6.09890759e-01
-1.14661598e+00 -5.11727691e-01 8.59360039e-01 -4.04821992e-01
7.36065269e-01 1.03796937e-01 6.98092997e-01 1.29704380e+00
1.70299396e-01 1.11094463e+00 7.31687307e-01 -3.18267137e-01
2.55978376e-01 4.21624817e-02 -1.84250921e-01 6.57709241e-01
8.58888309e-03 -7.42331296e-02 -3.73929441e-01 2.70236403e-01
1.15043604e+00 1.42387986e-01 -3.82623732e-01 -6.00613773e-01
-1.09299302e+00 4.77931082e-01 1.74613088e-01 2.83987653e-02
-2.44802400e-01 -1.47654578e-01 4.98071164e-01 -4.01419960e-02
7.99777433e-02 -1.12661041e-01 -8.31986547e-01 -2.81405915e-02
-4.59993660e-01 4.00269143e-02 4.40074086e-01 1.08623183e+00
4.37242240e-01 1.11233965e-01 -3.38428706e-01 8.62123132e-01
4.40753490e-01 4.14308637e-01 7.15482712e-01 -1.12032449e+00
8.45900178e-01 8.57021987e-01 1.05954580e-01 -1.04282534e+00
-2.87350327e-01 -1.94067046e-01 -7.40488887e-01 2.08018169e-01
7.25866139e-01 9.06012729e-02 -1.20340073e+00 1.57640219e+00
6.06196523e-01 3.45787764e-01 2.43007764e-01 1.08066702e+00
9.34274912e-01 6.74261630e-01 2.28632510e-01 2.75532585e-02
1.22444260e+00 -1.46327221e+00 -7.34896660e-01 -3.59452963e-01
7.25281894e-01 -5.35085440e-01 8.98230791e-01 4.41418618e-01
-9.11088705e-01 -1.30287850e+00 -9.74317193e-01 -5.83403483e-02
-1.76663637e-01 4.02836919e-01 3.35572660e-01 4.78939027e-01
-6.70621455e-01 4.80672121e-01 -8.92317593e-01 -4.15686190e-01
5.95545590e-01 2.83909500e-01 -3.96646738e-01 -3.78227413e-01
-1.07406628e+00 6.62580788e-01 5.81472099e-01 4.00405198e-01
-7.79279351e-01 -1.65745974e-01 -9.95167434e-01 -1.08917318e-01
4.27339405e-01 -3.74518991e-01 1.26035464e+00 -1.85016191e+00
-1.59420562e+00 5.92472494e-01 -1.33406743e-02 -1.34165823e-01
7.61654675e-01 -1.73360690e-01 -5.91677666e-01 2.77588397e-01
1.17565393e-01 9.05411243e-01 7.81396508e-01 -9.86361146e-01
-9.81358409e-01 -4.46573198e-01 9.03902426e-02 3.97648960e-01
1.44300237e-01 1.46623899e-03 -9.37696695e-01 -7.16174245e-01
2.70094901e-01 -9.67830300e-01 -2.85407543e-01 1.69085294e-01
-1.15341172e-01 4.13709693e-02 9.77120996e-01 -9.35916126e-01
6.73809767e-01 -2.32650685e+00 1.24418005e-01 -8.83938223e-02
-3.48359555e-01 4.20651376e-01 -4.53782856e-01 -1.67761177e-01
-1.78406909e-01 -2.52006650e-01 -1.21223420e-01 2.24549025e-02
-3.41534793e-01 5.58789849e-01 -1.25192016e-01 3.71028453e-01
4.34061259e-01 7.89961576e-01 -5.97264946e-01 -6.34839714e-01
2.85508722e-01 2.39662290e-01 -4.71111238e-01 4.40228254e-01
-1.94383383e-01 9.40920651e-01 -6.20939553e-01 7.71422684e-01
7.54383624e-01 -9.58199054e-02 5.40557764e-02 -2.48744830e-01
3.16501409e-01 7.46717118e-03 -1.56871808e+00 1.66952455e+00
1.68598250e-01 2.19014809e-01 -2.24077255e-01 -1.10830986e+00
8.56138885e-01 2.22856209e-01 5.73569000e-01 -1.02191019e+00
2.57491559e-01 -1.65740088e-01 1.94811106e-01 -8.93268585e-01
1.89729586e-01 1.87666997e-01 1.44600570e-01 1.44936144e-01
-7.67474398e-02 3.32700253e-01 1.69706464e-01 -2.63112247e-01
7.87805498e-01 7.23112106e-01 9.79487151e-02 2.06788585e-01
8.15792561e-01 -5.35336928e-03 1.22078979e+00 4.79529619e-01
-6.69979513e-01 8.40251029e-01 4.37232316e-01 -8.47277224e-01
-9.29641545e-01 -7.43830800e-01 1.25907242e-01 1.05288672e+00
5.79586804e-01 8.44135061e-02 -8.24443042e-01 -1.04935205e+00
-3.14389646e-01 4.26352352e-01 -6.51389658e-01 -3.35590065e-01
-6.57665968e-01 -5.78386068e-01 3.19396615e-01 1.09846854e+00
1.17075849e+00 -1.40536606e+00 -7.11450160e-01 1.47413477e-01
-4.35011864e-01 -1.54721355e+00 -2.97254264e-01 -1.29312620e-01
-9.06053901e-01 -1.28496182e+00 -6.70215011e-01 -8.38552415e-01
8.09474707e-01 2.62952775e-01 6.70247078e-01 2.34732572e-02
-1.65918514e-01 2.94299841e-01 -6.53722823e-01 -2.06718385e-01
-4.21231180e-01 -4.48861718e-01 -7.10503086e-02 4.58925068e-01
4.68080133e-01 -6.63059205e-02 -6.72653019e-01 7.90660799e-01
-9.67210829e-01 2.44371340e-01 5.89297056e-01 6.18112504e-01
6.97056293e-01 2.49973685e-01 1.73872009e-01 -6.95769906e-01
-2.12094307e-01 -1.74129725e-01 -3.70877862e-01 3.05610389e-01
8.30432493e-03 -1.00339472e-01 5.03842950e-01 -6.92154646e-01
-1.39917672e+00 5.58861494e-01 -1.34383393e-02 -5.69455624e-01
-5.30137420e-01 -1.92590892e-01 -8.22263122e-01 1.07543394e-01
3.12280774e-01 1.86829925e-01 -2.23438274e-02 -3.92669678e-01
5.31732552e-02 5.16195536e-01 4.70279932e-01 -2.28601098e-01
5.27137399e-01 4.73598540e-01 -1.40968934e-01 -6.68133914e-01
-7.60032773e-01 -3.76157761e-01 -1.04953575e+00 -3.85501593e-01
1.29961348e+00 -9.64051545e-01 -8.12549144e-02 8.27143431e-01
-1.27800965e+00 -3.49915415e-01 -3.02369893e-01 6.13210380e-01
-6.88931584e-01 5.32112956e-01 -5.22634506e-01 -7.09527791e-01
1.41817033e-01 -1.39869809e+00 1.19411540e+00 4.41618502e-01
8.65388736e-02 -5.23936689e-01 -1.91708907e-01 6.75953448e-01
-1.38498381e-01 2.74372160e-01 7.99363256e-01 -6.38703585e-01
-6.76486850e-01 -7.63560534e-02 -1.51495442e-01 6.30324483e-01
1.02854475e-01 1.32285235e-02 -9.40867484e-01 -4.51761112e-03
7.79897422e-02 -1.97378963e-01 8.05205107e-01 4.12045538e-01
1.40794182e+00 -8.34897384e-02 -2.38390818e-01 6.50061727e-01
9.40622628e-01 7.89575160e-01 1.17961943e+00 3.97437215e-01
9.04480696e-01 8.10609818e-01 9.85814393e-01 3.21904451e-01
5.53343333e-02 7.09373295e-01 4.13543612e-01 -2.53310412e-01
-1.75398305e-01 -2.55867481e-01 5.29466927e-01 4.16168153e-01
-2.25136176e-01 -1.87446713e-01 -6.08898401e-01 1.25105187e-01
-2.05353928e+00 -9.43518102e-01 -1.91584617e-01 2.13535929e+00
4.07001406e-01 1.61421597e-01 8.35453048e-02 1.00750132e-02
8.11750233e-01 5.59036098e-02 -6.97607636e-01 -1.46006197e-01
-1.83446631e-01 -2.19636224e-02 4.15706426e-01 1.30118821e-02
-1.44476545e+00 9.79882956e-01 5.83419323e+00 5.48373520e-01
-8.63259315e-01 -1.54854879e-01 9.20752883e-01 1.66994348e-01
3.93937111e-01 -1.92233443e-01 -7.61802018e-01 4.47634876e-01
5.50824940e-01 6.31362438e-01 2.75135841e-02 1.08932960e+00
2.84370393e-01 -1.76810339e-01 -1.08232307e+00 1.01593101e+00
3.23861539e-01 -7.69914925e-01 3.10252815e-01 -2.28849843e-01
6.16326153e-01 -3.34168136e-01 -1.51255757e-01 5.45225143e-01
-1.55263886e-01 -7.79433131e-01 7.27718890e-01 6.15574479e-01
5.40653706e-01 -6.59042776e-01 9.27879155e-01 3.45902145e-01
-1.21377075e+00 -3.42506617e-01 -6.68202281e-01 -6.15074812e-03
-4.28028964e-02 -1.69253930e-01 -6.12344384e-01 4.30405408e-01
8.21027219e-01 7.66054273e-01 -7.76535749e-01 8.86499345e-01
-1.26427516e-01 3.12110245e-01 8.98189694e-02 2.88947523e-01
1.94319095e-02 -2.34417915e-01 2.85339683e-01 8.69084477e-01
5.50605021e-02 2.82192677e-01 3.94747585e-01 5.62775433e-01
2.23009780e-01 2.00283732e-02 -4.67454404e-01 2.10589580e-02
-1.04851812e-01 9.03941751e-01 -8.05056334e-01 -4.59913641e-01
-7.31561959e-01 1.39636040e+00 2.23625395e-02 6.58544958e-01
-1.04337740e+00 5.94246089e-02 5.81890106e-01 1.16392344e-01
5.83894432e-01 -2.68723406e-02 1.21479362e-01 -1.16464531e+00
1.59917638e-01 -1.10893893e+00 5.37948430e-01 -9.17638540e-01
-1.04083967e+00 5.86083531e-01 -4.33170237e-03 -1.64630258e+00
-1.84029058e-01 -9.52186704e-01 -3.73944938e-01 5.21807909e-01
-1.25253129e+00 -1.07891166e+00 -5.19381762e-01 7.96989799e-01
1.10539365e+00 -2.55578279e-01 5.29363275e-01 2.76145011e-01
-8.20531011e-01 3.95660311e-01 -2.50028282e-01 5.56242824e-01
5.36915600e-01 -7.53465831e-01 2.46647775e-01 1.00721920e+00
4.00685370e-02 -2.96210106e-02 3.50488871e-01 -8.51113677e-01
-1.09327471e+00 -1.34705448e+00 2.98486114e-01 -4.13154244e-01
1.18932448e-01 -1.67958811e-01 -1.01937604e+00 6.84252203e-01
-2.10508913e-01 2.16341972e-01 5.48291862e-01 -5.80269635e-01
-1.82017520e-01 1.95135840e-03 -1.06313205e+00 5.16836345e-01
1.21849322e+00 -8.02896544e-02 -6.36245251e-01 2.20047861e-01
6.30678952e-01 -5.86593986e-01 -4.24452454e-01 7.06951380e-01
6.44103289e-01 -8.72336686e-01 8.63770425e-01 -1.07660258e+00
5.41670322e-01 -5.29713631e-01 -4.25619870e-01 -7.13103890e-01
-4.39366907e-01 2.70017534e-02 -1.39819430e-02 1.02009320e+00
5.53786755e-02 -3.35569590e-01 8.58439028e-01 8.84718180e-01
5.39033562e-02 -6.51243806e-01 -9.16332960e-01 -7.12886155e-01
-3.22889864e-01 -2.52032936e-01 5.05298197e-01 7.13835418e-01
-4.99932051e-01 1.81638643e-01 -4.04857427e-01 2.63458341e-01
1.53262481e-01 1.75920501e-02 9.63321149e-01 -1.05331826e+00
-4.99461740e-01 -4.92555983e-02 -7.92117715e-01 -1.31596148e+00
7.35889822e-02 -4.26077753e-01 9.85046178e-02 -1.44436729e+00
1.93934470e-01 -3.20080705e-02 -3.22497457e-01 4.82583672e-01
-2.78571695e-01 1.61424458e-01 2.13425770e-01 2.73108393e-01
-7.78429329e-01 7.97553062e-01 1.55592775e+00 -1.88435122e-01
-3.39254886e-01 2.97243923e-01 -2.36805052e-01 1.03239429e+00
7.39723146e-01 -3.12785238e-01 -5.00522017e-01 -4.69087124e-01
-4.12656069e-01 1.08945876e-01 3.40465099e-01 -1.21997619e+00
-4.68895920e-02 -3.18594486e-01 1.11055076e+00 -6.60163701e-01
3.48819613e-01 -1.16138399e+00 1.05291486e-01 4.87533152e-01
-2.87741184e-01 -7.46408524e-03 2.86563456e-01 8.71979833e-01
-1.57951519e-01 -1.38451427e-01 9.00035203e-01 -3.26204270e-01
-1.25057304e+00 4.35898900e-01 -2.25702539e-01 1.83645878e-02
1.17097402e+00 -7.38108575e-01 -1.72458038e-01 -2.35434175e-01
-7.81533957e-01 1.18789844e-01 2.79450595e-01 8.26452553e-01
6.77789271e-01 -1.44152343e+00 -4.85749155e-01 5.17556310e-01
9.55021977e-02 2.31128007e-01 5.81017077e-01 7.94029593e-01
-4.33842927e-01 1.77106291e-01 -6.74191713e-01 -6.58313274e-01
-1.28174162e+00 6.09723508e-01 6.22856915e-01 -2.20615156e-02
-7.05911160e-01 7.63360620e-01 7.32890069e-01 -1.81757718e-01
5.48430264e-01 -4.33001488e-01 -4.65886444e-01 -1.23312943e-01
8.25750291e-01 2.93296963e-01 -1.64774761e-01 -9.37613904e-01
-3.69462639e-01 5.70670307e-01 -3.00842021e-02 6.64032064e-04
1.14718187e+00 3.64766689e-04 2.34488726e-01 3.47974181e-01
1.06722724e+00 -4.42757577e-01 -1.69287562e+00 -1.02503300e-01
-2.87549049e-01 -6.76441848e-01 -3.49012196e-01 -7.16071486e-01
-1.38631594e+00 7.75614023e-01 8.27852845e-01 -3.64861757e-01
1.21835399e+00 -1.11351341e-01 6.54991984e-01 2.73038089e-01
2.89023578e-01 -1.51924491e+00 3.05505127e-01 3.32894981e-01
9.35681820e-01 -1.25132918e+00 -2.26623759e-01 -5.67015469e-01
-1.14922261e+00 1.20963836e+00 1.21225762e+00 -1.93554461e-02
2.34986290e-01 -1.87317617e-02 1.96989834e-01 1.17760912e-01
-3.88962626e-01 -2.04156324e-01 4.20709223e-01 8.53406847e-01
4.58075060e-03 -2.78673768e-01 -5.42370714e-02 8.26856673e-01
2.84086704e-01 -1.45612331e-03 3.05473357e-01 9.84750390e-01
-4.74940240e-01 -1.02536750e+00 -4.31918651e-01 1.66777849e-01
-1.82314545e-01 4.28401172e-01 -5.76605976e-01 9.58579957e-01
4.89202321e-01 8.90846729e-01 1.62884369e-01 -5.50656438e-01
5.13415992e-01 1.94504097e-01 5.09271085e-01 -5.78475654e-01
-1.29640326e-01 9.26605836e-02 5.38592711e-02 -9.75454807e-01
-6.01442099e-01 -6.09641016e-01 -1.53898644e+00 4.15940642e-01
-1.35793716e-01 -2.50207722e-01 2.90614665e-01 1.03936327e+00
2.24171698e-01 7.75677443e-01 5.23749948e-01 -6.60587132e-01
-4.45174605e-01 -9.12442029e-01 -6.07124984e-01 5.64275920e-01
3.34387459e-03 -9.10006106e-01 -2.31032014e-01 3.99657637e-01] | [8.201454162597656, 0.644717276096344] |
64c2ff14-05a2-446f-a303-bcc2e8b06ced | openood-benchmarking-generalized-out-of | 2210.07242 | null | https://arxiv.org/abs/2210.07242v1 | https://arxiv.org/pdf/2210.07242v1.pdf | OpenOOD: Benchmarking Generalized Out-of-Distribution Detection | Out-of-distribution (OOD) detection is vital to safety-critical machine learning applications and has thus been extensively studied, with a plethora of methods developed in the literature. However, the field currently lacks a unified, strictly formulated, and comprehensive benchmark, which often results in unfair comparisons and inconclusive results. From the problem setting perspective, OOD detection is closely related to neighboring fields including anomaly detection (AD), open set recognition (OSR), and model uncertainty, since methods developed for one domain are often applicable to each other. To help the community to improve the evaluation and advance, we build a unified, well-structured codebase called OpenOOD, which implements over 30 methods developed in relevant fields and provides a comprehensive benchmark under the recently proposed generalized OOD detection framework. With a comprehensive comparison of these methods, we are gratified that the field has progressed significantly over the past few years, where both preprocessing methods and the orthogonal post-hoc methods show strong potential. | ['Ziwei Liu', 'Yixuan Li', 'Dan Hendrycks', 'Wayne Zhang', 'Kaiyang Zhou', 'Xuefeng Du', 'Yiyou Sun', 'Bo Li', 'Guangyao Chen', 'Haoqi Wang', 'Wenxuan Peng', 'Kunyuan Ding', 'Zitang Zhou', 'Dejian Zou', 'Pengyun Wang', 'Jingkang Yang'] | 2022-10-13 | null | null | null | null | ['open-set-learning'] | ['miscellaneous'] | [ 8.14220160e-02 -5.60872257e-03 -2.09857583e-01 -2.08898932e-01
-7.85869539e-01 -4.37517792e-01 7.60293663e-01 7.66070306e-01
-3.34647410e-02 5.59457958e-01 -7.03819767e-02 -3.09524089e-01
-3.14360589e-01 -6.70935035e-01 -3.31015378e-01 -5.30339539e-01
-3.34107190e-01 2.78517544e-01 3.97595525e-01 1.17919125e-01
4.40657198e-01 4.11237568e-01 -1.85349798e+00 -1.01076290e-01
1.12172925e+00 1.31669760e+00 -5.76434910e-01 -8.74485448e-02
-9.02031064e-02 6.02466702e-01 -4.47696745e-01 -4.86108959e-01
2.60780126e-01 2.41866652e-02 -7.18293607e-01 -2.85259455e-01
3.64551604e-01 4.22655493e-02 -1.23907164e-01 1.18617404e+00
5.10443866e-01 2.23076567e-02 8.20672870e-01 -1.73445654e+00
-3.34321171e-01 2.81727880e-01 -6.33247852e-01 1.56820089e-01
3.48784328e-01 1.43743679e-01 1.18474913e+00 -8.25668693e-01
4.57719654e-01 9.07546341e-01 6.88929677e-01 1.92848697e-01
-1.04225254e+00 -4.85622287e-01 1.71646267e-01 3.56077760e-01
-1.50082994e+00 -1.89092040e-01 6.43457651e-01 -6.91698909e-01
1.00523090e+00 2.42839143e-01 2.93555945e-01 9.96296585e-01
2.64166027e-01 8.06320965e-01 8.90617430e-01 -3.87335867e-01
6.39715016e-01 1.85002923e-01 1.60748243e-01 5.08019149e-01
6.75816655e-01 3.79548103e-01 -4.76415157e-01 -5.26762605e-01
2.05154177e-02 9.63071585e-02 -1.15873687e-01 -8.39217007e-01
-1.10389018e+00 7.72549748e-01 2.33226866e-02 2.45447144e-01
-1.52310893e-01 -3.08067262e-01 7.53944039e-01 1.28041297e-01
6.26013279e-01 5.04714429e-01 -3.03529739e-01 -2.76748925e-01
-1.04299831e+00 4.57454562e-01 9.44813013e-01 7.34073400e-01
5.86106479e-01 -1.15143573e-02 -2.16699362e-01 7.89290369e-01
4.87328172e-01 1.99483991e-01 4.39293571e-02 -6.63765669e-01
3.34218949e-01 7.32862115e-01 -4.99451086e-02 -1.21750975e+00
-2.77270138e-01 -4.26214725e-01 -8.38192761e-01 3.99708241e-01
3.48040611e-01 -2.84922160e-02 -4.98514950e-01 1.45703161e+00
4.88882720e-01 2.24723548e-01 4.42979485e-03 6.33651912e-01
6.39867783e-01 2.81789780e-01 1.46386353e-02 -8.68895873e-02
1.08860087e+00 -7.60916948e-01 -6.66777670e-01 -3.79911691e-01
7.94158757e-01 -7.33089685e-01 6.38602018e-01 7.25022197e-01
-6.00570977e-01 -4.28147167e-02 -1.16529465e+00 4.80884790e-01
-6.27277136e-01 -5.17732322e-01 7.34870493e-01 5.82667708e-01
-7.32175708e-01 3.52549493e-01 -7.09395468e-01 -3.88873875e-01
5.74068129e-01 -3.29590663e-02 -3.57882977e-01 -2.52506167e-01
-1.30372047e+00 1.13581502e+00 3.08258861e-01 -1.02145329e-01
-1.04632676e+00 -8.42532039e-01 -9.99483764e-01 -2.96262980e-01
7.32020438e-01 -5.30398607e-01 9.26699102e-01 -2.09689692e-01
-9.35368001e-01 8.97546291e-01 1.99553311e-01 -6.73455358e-01
5.69775045e-01 -4.97440457e-01 -7.79323280e-01 -1.99632287e-01
1.23363867e-01 1.16034590e-01 8.72655690e-01 -1.11791968e+00
-6.93120241e-01 -5.64286768e-01 -2.11441413e-01 -1.74524188e-01
-3.42521429e-01 2.59394735e-01 -2.21357182e-01 -7.97103882e-01
-8.41964781e-02 -6.20354176e-01 -3.89331490e-01 -1.19205534e-01
-5.27551234e-01 -3.15202355e-01 8.28868747e-01 -2.93981582e-01
1.97029316e+00 -2.20653534e+00 -1.07490204e-01 3.85382950e-01
3.33953768e-01 2.27224097e-01 2.43396625e-01 5.28339326e-01
-4.15225744e-01 -4.32464667e-02 -9.20225859e-01 -1.15644388e-01
2.42928684e-01 1.35031536e-01 -4.61891443e-01 7.03744054e-01
3.65163088e-01 4.09603894e-01 -9.71782923e-01 -4.26617533e-01
3.17186326e-01 9.43933725e-02 -5.13410330e-01 1.31186247e-01
-2.13033512e-01 9.10994560e-02 -7.16813743e-01 1.06841552e+00
6.87545240e-01 8.39347243e-02 -3.02697778e-01 2.88392067e-01
-2.10681170e-01 -1.47788487e-02 -1.53341448e+00 1.54105711e+00
-8.43542516e-02 3.72013122e-01 -1.23131208e-01 -1.09974468e+00
1.00487339e+00 2.42880985e-01 6.92156434e-01 -3.94964993e-01
7.04626217e-02 4.44450557e-01 -1.06310889e-01 -5.01762211e-01
4.51630890e-01 2.58274794e-01 -2.03571677e-01 4.39859599e-01
-1.87054917e-01 -1.26932025e-01 2.14612305e-01 1.70513406e-01
1.51102757e+00 -3.85413095e-02 6.56273961e-01 -3.02251518e-01
7.22918212e-01 1.18650079e-01 4.33683664e-01 7.10031390e-01
-5.34182549e-01 5.22202611e-01 6.97205722e-01 -3.74292254e-01
-6.15055740e-01 -1.12888312e+00 -8.22593629e-01 5.35690069e-01
2.93621905e-02 -7.16547549e-01 -7.24547744e-01 -9.47392762e-01
3.86934310e-01 8.37125480e-01 -3.79911661e-01 -2.21845031e-01
-1.66030452e-01 -9.04899955e-01 6.65977657e-01 3.91634107e-01
2.74293959e-01 -8.60924721e-01 -4.01616305e-01 1.42481312e-01
-1.25282928e-02 -1.05730653e+00 3.72660980e-02 1.76563799e-01
-8.47713232e-01 -1.27635682e+00 -4.46210802e-01 -2.85753131e-01
4.69706476e-01 3.05245295e-02 1.39511395e+00 -2.00166423e-02
-4.05712157e-01 5.07250607e-01 -4.64760840e-01 -7.72156298e-01
-5.04842162e-01 -7.04986155e-02 3.99795145e-01 -3.57515900e-03
6.93451524e-01 -3.51629198e-01 -2.47063369e-01 5.28680503e-01
-1.17976797e+00 -7.60710537e-01 4.35611784e-01 6.07955754e-01
5.71926534e-01 1.96149841e-01 7.19358087e-01 -8.68167400e-01
6.01719201e-01 -8.01370502e-01 -8.91195476e-01 1.88715793e-02
-1.19595850e+00 1.42240524e-01 3.95020962e-01 2.59476811e-01
-7.94948459e-01 -2.50132293e-01 -2.71841705e-01 -3.75206560e-01
-7.52623081e-01 7.31207550e-01 -2.42344365e-01 -9.27628279e-02
7.27069139e-01 -4.52408046e-02 3.93194892e-02 -4.57545102e-01
1.70511842e-01 8.01659405e-01 4.78130043e-01 -6.85828507e-01
1.01639724e+00 3.24805617e-01 -8.03136174e-03 -7.98508048e-01
-8.44136834e-01 -5.94755173e-01 -4.58411902e-01 -1.17919013e-01
6.29685521e-01 -7.65264273e-01 -1.72755569e-01 6.06629491e-01
-9.07189190e-01 1.04099214e-01 -1.25267282e-01 3.46596181e-01
-1.17419742e-01 8.11867654e-01 3.01050637e-02 -8.34379673e-01
-6.16925098e-02 -1.27550173e+00 9.86203730e-01 -2.04387665e-01
-4.93478417e-01 -1.08259046e+00 3.47643048e-01 1.86668828e-01
3.35561424e-01 4.46464449e-01 1.07946098e+00 -1.10859239e+00
-5.10424316e-01 -7.69887149e-01 -5.06792702e-02 5.49961627e-01
-2.62410194e-03 2.02529952e-01 -1.20469403e+00 -1.11679435e-01
2.09708363e-02 -4.68229223e-03 7.84514010e-01 1.87514424e-01
1.52495301e+00 8.67531970e-02 -3.88607055e-01 1.36092350e-01
1.19483161e+00 -3.10786683e-02 7.29518831e-01 4.84737933e-01
2.76298672e-01 5.87191045e-01 1.00630152e+00 7.23534882e-01
2.34841168e-01 5.97884655e-01 9.08363581e-01 -2.34433934e-02
3.40011597e-01 1.42546788e-01 4.12653029e-01 3.67206633e-01
2.55553350e-02 -1.56634808e-01 -1.25047159e+00 5.47648609e-01
-1.73060572e+00 -1.04401958e+00 -2.50166178e-01 2.61147237e+00
5.34733653e-01 3.44168663e-01 1.32467672e-01 4.62777525e-01
7.20223486e-01 3.21786553e-01 -4.17093754e-01 -2.81514496e-01
5.20242043e-02 4.62229103e-02 1.32258371e-01 3.82474139e-02
-1.24120033e+00 3.90733927e-01 7.10045338e+00 1.11081851e+00
-6.69904470e-01 -2.48811066e-01 5.68291485e-01 3.46704900e-01
-3.21664572e-01 -7.11478814e-02 -7.85236180e-01 4.36605722e-01
8.04146469e-01 -2.45934442e-01 6.13067597e-02 1.21453559e+00
-2.55588219e-02 -3.64098698e-01 -1.53077054e+00 8.72005463e-01
1.51770279e-01 -1.13700819e+00 -2.06878603e-01 2.72223234e-01
7.10681140e-01 2.12699592e-01 1.30319640e-01 4.20792460e-01
5.34450002e-02 -9.62573588e-01 5.00517726e-01 3.46376181e-01
5.09178460e-01 -7.17706919e-01 1.03097165e+00 1.79211095e-01
-9.79489148e-01 -1.39521211e-01 -1.25069380e-01 -7.49700367e-02
-4.40268479e-02 1.16884112e+00 -5.62118709e-01 1.04322302e+00
8.48983467e-01 9.22212660e-01 -5.13156235e-01 1.43798864e+00
-1.17222652e-01 7.99176097e-01 -3.01219016e-01 3.66657913e-01
1.51333407e-01 -3.03235978e-01 1.03210175e+00 9.19246137e-01
2.26567551e-01 -4.77605492e-01 2.44165808e-01 9.51315343e-01
2.34394282e-01 1.41651139e-01 -8.65337670e-01 -1.94514900e-01
3.90715033e-01 1.28273201e+00 -5.11023760e-01 2.71354448e-02
-7.89834499e-01 3.76709968e-01 -9.78789851e-02 1.93588585e-01
-1.01059914e+00 -5.34963191e-01 1.05055451e+00 2.48635001e-02
-7.73184597e-02 1.06702276e-01 -3.02240044e-01 -1.09680319e+00
1.60297319e-01 -1.09180498e+00 9.07515109e-01 -1.96180880e-01
-1.76568973e+00 5.70920408e-01 2.51276076e-01 -1.72306406e+00
-2.17769787e-01 -7.16361284e-01 -6.65502727e-01 5.46895206e-01
-1.53599334e+00 -4.25575703e-01 -2.89570183e-01 2.06006825e-01
2.79856890e-01 -3.77592921e-01 9.97046351e-01 6.35387599e-01
-9.98604417e-01 5.42017341e-01 1.59225121e-01 -6.21294379e-02
1.01387227e+00 -1.17651999e+00 2.14302704e-01 1.09903193e+00
-5.60845435e-02 5.78984022e-01 7.79508233e-01 -5.85108995e-01
-1.17678773e+00 -1.22814119e+00 6.44611955e-01 -8.36537361e-01
1.04436529e+00 -8.55150670e-02 -1.08803451e+00 4.39176440e-01
-6.40209243e-02 4.18225676e-01 7.73803413e-01 3.40835869e-01
-4.48264956e-01 -1.87252104e-01 -1.11417246e+00 3.21404606e-01
8.27091336e-01 -4.48844522e-01 -7.55348444e-01 1.27287552e-01
2.65825152e-01 -4.18910116e-01 -8.93616021e-01 6.78464472e-01
1.46057084e-01 -1.18273735e+00 9.43219721e-01 -5.65084338e-01
3.45783353e-01 -4.67312515e-01 -2.98065335e-01 -1.12282670e+00
1.35290250e-01 -5.16782165e-01 -4.12212998e-01 1.38886440e+00
3.48525912e-01 -1.01609325e+00 3.37251544e-01 6.63586497e-01
-6.63640499e-01 -1.07560742e+00 -8.82580042e-01 -1.01888037e+00
9.82421935e-02 -9.78457987e-01 8.23859930e-01 8.66300642e-01
1.74896717e-01 -3.62520739e-02 -2.44623244e-01 4.50879842e-01
6.77524269e-01 4.92960140e-02 7.96926141e-01 -1.65957868e+00
1.97125077e-02 -6.09342158e-01 -7.62357116e-01 -5.69739282e-01
7.52536580e-02 -1.10608745e+00 -1.84236858e-02 -1.28110254e+00
6.58871084e-02 -5.73319614e-01 -5.89375257e-01 4.09671843e-01
-2.09527239e-01 2.10528169e-02 -4.12295580e-01 6.61659166e-02
-8.55781078e-01 5.53676367e-01 6.70999467e-01 -6.91221282e-02
1.35578692e-01 1.79934412e-01 -7.22608745e-01 1.12943745e+00
4.25256252e-01 -6.13203228e-01 -1.26550436e-01 3.93738213e-04
3.12145203e-01 -3.27452689e-01 5.65055251e-01 -1.22303474e+00
-6.21698722e-02 -7.17026135e-03 8.01364053e-03 -6.26563847e-01
-3.27308565e-01 -9.89052474e-01 -9.77918729e-02 3.95665348e-01
-1.34384215e-01 2.05767620e-02 2.30172440e-01 8.43596220e-01
-6.28162920e-01 -3.34428161e-01 7.02077091e-01 1.58521265e-01
-1.02455235e+00 6.29588783e-01 -2.55124927e-01 5.23916841e-01
1.30864275e+00 -1.79875895e-01 -1.34696111e-01 -9.85632911e-02
-3.35582197e-01 4.24329758e-01 4.50482637e-01 6.44781232e-01
4.07796562e-01 -1.13713193e+00 -5.57514846e-01 2.83405125e-01
7.68822193e-01 1.66562781e-01 1.07274309e-01 1.02134931e+00
-2.36649320e-01 4.87291873e-01 7.02131763e-02 -7.29210377e-01
-8.22836637e-01 6.11798644e-01 2.61203945e-01 -2.95658290e-01
-6.05205417e-01 3.60667378e-01 8.73403698e-02 -4.34453368e-01
7.14782119e-01 -2.36683175e-01 -2.01429933e-01 6.76678419e-02
4.45738912e-01 7.65585363e-01 4.70374763e-01 -4.71621990e-01
-5.91255486e-01 2.72349685e-01 3.47737521e-02 5.14827669e-01
1.28378201e+00 4.68766689e-02 -3.56921345e-01 4.98019844e-01
7.78934360e-01 1.04450859e-01 -8.17759335e-01 -1.39797732e-01
4.35042620e-01 -5.74754953e-01 1.35668457e-01 -7.05327332e-01
-8.07599962e-01 7.93622375e-01 5.34926713e-01 5.01221061e-01
1.04409564e+00 -6.41302392e-02 5.88604808e-01 2.78880000e-01
5.19688368e-01 -1.07780540e+00 -8.09263624e-03 6.47573769e-01
7.66977966e-01 -1.38700056e+00 8.52243155e-02 -5.11083722e-01
-5.94433844e-01 1.04799366e+00 6.96739674e-01 2.22618863e-01
7.77299047e-01 3.59536797e-01 -1.65757075e-01 -2.94202805e-01
-4.58265454e-01 -1.27516359e-01 4.18057829e-01 5.47729552e-01
3.63519967e-01 -3.86633039e-01 -1.19930781e-01 6.65571690e-01
3.25996011e-01 -9.16493908e-02 2.71962821e-01 7.33811498e-01
-4.49277192e-01 -1.33440280e+00 -5.05972922e-01 7.28125870e-01
-5.21013260e-01 -8.65259767e-02 -9.28199068e-02 7.88010120e-01
6.26384765e-02 8.06646466e-01 -6.51154891e-02 -1.74943551e-01
4.62450862e-01 2.18555361e-01 -3.62553610e-03 -5.59968472e-01
-2.05916479e-01 -2.77242690e-01 9.71714407e-02 -7.58290470e-01
-2.69973814e-01 -7.58346856e-01 -1.09254777e+00 -4.78032865e-02
-3.91124904e-01 2.00732931e-01 3.74984711e-01 1.02242494e+00
5.18707275e-01 4.78799492e-01 6.11007094e-01 -3.84504557e-01
-7.95913458e-01 -6.06434107e-01 -7.85119832e-01 1.97233438e-01
1.32994175e-01 -1.22421312e+00 -6.69482470e-01 -4.01430398e-01] | [7.613095760345459, 2.6619298458099365] |
e5e180f4-fa71-45be-8eb7-5e8d661dfcc1 | incomplete-multi-view-clustering-via-1 | 2305.11489 | null | https://arxiv.org/abs/2305.11489v1 | https://arxiv.org/pdf/2305.11489v1.pdf | Incomplete Multi-view Clustering via Diffusion Completion | Incomplete multi-view clustering is a challenging and non-trivial task to provide effective data analysis for large amounts of unlabeled data in the real world. All incomplete multi-view clustering methods need to address the problem of how to reduce the impact of missing views. To address this issue, we propose diffusion completion to recover the missing views integrated into an incomplete multi-view clustering framework. Based on the observable views information, the diffusion model is used to recover the missing views, and then the consistency information of the multi-view data is learned by contrastive learning to improve the performance of multi-view clustering. To the best of our knowledge, this may be the first work to incorporate diffusion models into an incomplete multi-view clustering framework. Experimental results show that the proposed method performs well in recovering the missing views while achieving superior clustering performance compared to state-of-the-art methods. | ['Sifan Fang'] | 2023-05-19 | null | null | null | null | ['incomplete-multi-view-clustering'] | ['computer-vision'] | [-2.49792978e-01 -1.61074132e-01 -3.74471098e-01 -2.09909841e-01
-8.01583290e-01 -6.32767618e-01 3.90499145e-01 -2.08589256e-01
5.86438663e-02 2.16355383e-01 3.28209430e-01 2.36264735e-01
-1.42021522e-01 -5.28148532e-01 -3.18363786e-01 -9.15185630e-01
2.49026820e-01 7.09125042e-01 1.39754564e-01 2.50127077e-01
6.08569160e-02 4.73447330e-02 -1.24612308e+00 4.55997854e-01
8.69932115e-01 4.25759077e-01 4.93272930e-01 3.61399531e-01
2.00317074e-02 7.84572363e-01 -4.33533601e-02 -2.12457031e-01
2.45028958e-01 -3.85822326e-01 -6.93346918e-01 7.23539531e-01
6.85246987e-03 -3.98564965e-01 -1.71922252e-01 1.20333207e+00
3.37632626e-01 -2.02258248e-02 4.22586679e-01 -1.28424728e+00
-6.57417297e-01 5.73315799e-01 -1.06355679e+00 5.10513633e-02
3.05400640e-01 -4.02188212e-01 9.18357253e-01 -1.04367423e+00
8.72883320e-01 1.18615711e+00 2.58625925e-01 4.02616352e-01
-1.09702778e+00 -4.48978901e-01 5.44859707e-01 3.93382430e-01
-1.38330305e+00 -3.80470008e-01 1.05148542e+00 -5.27120292e-01
1.52824834e-01 2.00968403e-02 5.23383260e-01 9.62508738e-01
-2.29907393e-01 1.05385661e+00 1.31696951e+00 -1.43306345e-01
3.98469657e-01 9.95628908e-02 -8.65527708e-03 6.61024332e-01
3.29417348e-01 -3.41105193e-01 -4.57302272e-01 -2.95479685e-01
4.84124422e-01 7.32331038e-01 -2.95040995e-01 -8.99486423e-01
-1.33329713e+00 7.86710918e-01 2.37884387e-01 3.42598468e-01
-5.43950438e-01 -3.26837093e-01 1.70397982e-01 8.87651667e-02
7.05529094e-01 -2.61563241e-01 -3.95840347e-01 9.80429277e-02
-9.70826209e-01 -2.18583882e-01 7.24223077e-01 9.96336877e-01
8.63704562e-01 -1.49421051e-01 4.79264408e-01 8.74475360e-01
4.67064708e-01 7.42553174e-01 1.07129447e-01 -1.40401959e+00
6.77873433e-01 9.48648870e-01 -2.97365878e-02 -9.04265642e-01
-2.16191173e-01 -2.95424104e-01 -1.23207402e+00 9.66106877e-02
3.26346904e-01 -1.07562289e-01 -6.93489432e-01 1.51237643e+00
6.45049810e-01 1.89568639e-01 1.87436923e-01 9.21593845e-01
6.48582697e-01 5.19688964e-01 -6.03917062e-01 -7.15395451e-01
1.07742500e+00 -1.01744890e+00 -8.37261260e-01 -5.45058446e-03
3.48860502e-01 -7.24535823e-01 6.42536104e-01 5.61976552e-01
-1.06904459e+00 -3.54312241e-01 -9.68511164e-01 2.15014085e-01
-2.71195564e-02 1.31333187e-01 3.14884692e-01 4.02756512e-01
-8.23916674e-01 1.48952559e-01 -1.28373528e+00 -4.09538299e-01
3.96543086e-01 1.00175537e-01 -5.88695228e-01 -9.78600979e-01
-5.62088072e-01 4.37251478e-01 1.55322000e-01 -8.28795880e-02
-1.01178324e+00 -4.96115178e-01 -6.62597716e-01 -1.20843269e-01
9.31640089e-01 -7.55268037e-01 7.85915613e-01 -6.59272790e-01
-8.03529918e-01 6.44369781e-01 -5.90175748e-01 2.32312065e-02
4.06060994e-01 -1.49672151e-01 -2.42458433e-01 5.08299172e-01
5.70762277e-01 2.64750957e-01 7.70786047e-01 -2.07160115e+00
-4.80563730e-01 -1.02888989e+00 -1.64643839e-01 5.42978346e-01
-3.02125484e-01 -3.87303203e-01 -1.11382973e+00 -4.30099577e-01
7.59429097e-01 -1.30443347e+00 -4.17514920e-01 -2.82073408e-01
-4.09833968e-01 1.20902732e-02 1.23725343e+00 -8.27020884e-01
1.04171443e+00 -2.18416810e+00 7.55272150e-01 1.67073667e-01
5.82713842e-01 -1.73542023e-01 4.40732501e-02 6.34874821e-01
1.79971501e-01 -2.49784952e-03 -4.06090468e-01 -6.09657347e-01
-3.35884452e-01 4.44377899e-01 1.24448732e-01 7.92858601e-01
-6.21780217e-01 5.34938097e-01 -8.76080096e-01 -5.56625485e-01
2.67484188e-01 4.80154663e-01 -3.65506381e-01 2.67713815e-01
1.30612239e-01 8.48112285e-01 -6.58919573e-01 7.27371812e-01
9.25036311e-01 -7.42830157e-01 5.54331243e-01 -1.93608701e-01
1.87489286e-01 -5.10367811e-01 -1.58582103e+00 2.14906549e+00
-1.85269266e-01 -9.74326115e-03 5.46875715e-01 -9.51553345e-01
3.84195954e-01 4.85078692e-01 1.11908615e+00 -2.37403676e-01
-3.07497174e-01 5.85707240e-02 -1.28998667e-01 -4.49722081e-01
1.72331482e-01 -2.23028541e-01 2.56339349e-02 9.39298391e-01
-1.41808867e-01 1.63743451e-01 9.09015983e-02 7.59589612e-01
9.04555261e-01 -8.07665139e-02 3.16180170e-01 5.83991632e-02
5.53030372e-01 -8.67730901e-02 8.64542782e-01 4.14471924e-01
-1.92642510e-01 8.19631934e-01 1.63392767e-01 -2.93247491e-01
-1.03310215e+00 -1.06348801e+00 4.72554356e-01 6.00912571e-01
3.14403564e-01 -5.37332594e-01 -8.32597613e-01 -8.76058698e-01
-2.94472367e-01 2.99677998e-01 -4.38738018e-01 2.46522799e-01
-1.62533194e-01 -7.73566246e-01 -7.77487829e-02 4.84613568e-01
5.31751573e-01 -4.52252775e-01 -6.36396632e-02 5.54957353e-02
-7.77361989e-01 -1.33891070e+00 -4.93169814e-01 -1.68368623e-01
-1.05889153e+00 -1.42604148e+00 -7.99208641e-01 -5.98376572e-01
9.67691660e-01 1.11819100e+00 8.66160095e-01 2.19953522e-01
2.95254141e-02 8.74492407e-01 -5.52925110e-01 1.49857044e-01
-3.78084600e-01 -6.27744645e-02 1.63706347e-01 3.09033930e-01
2.47983694e-01 -6.95463359e-01 -7.04454064e-01 5.21073222e-01
-1.20929933e+00 8.96025449e-02 4.57294673e-01 7.15639353e-01
9.41683114e-01 5.00644922e-01 3.44105512e-01 -1.19592059e+00
2.73044676e-01 -5.45663059e-01 -2.79579163e-01 4.42882061e-01
-8.56267691e-01 -1.32798061e-01 6.52504563e-01 -1.10797636e-01
-1.31842005e+00 4.47073847e-01 3.35825533e-01 -8.39032292e-01
-2.47987151e-01 5.65448880e-01 -5.02034247e-01 2.82610297e-01
3.67134474e-02 2.55829155e-01 5.99520933e-03 -6.66047871e-01
7.92582333e-01 5.25544643e-01 1.57852352e-01 -3.46379787e-01
7.54076004e-01 1.35791361e+00 -5.29020131e-02 -6.81342661e-01
-1.04178202e+00 -1.09324634e+00 -1.21355855e+00 -3.01872045e-01
9.93347168e-01 -1.38923562e+00 -2.04319179e-01 3.97055507e-01
-7.70304203e-01 1.49035200e-01 -4.25321162e-02 5.86900651e-01
-5.13291776e-01 1.03071094e+00 -7.40875244e-01 -7.70509183e-01
-1.44350564e-03 -1.20726991e+00 9.68345582e-01 -2.04832554e-01
2.96440423e-01 -1.09734464e+00 1.53562218e-01 1.01608050e+00
-2.27988809e-01 8.84093344e-02 6.25120759e-01 -4.01813000e-01
-8.52171302e-01 -2.71086767e-02 1.43406335e-02 2.06541896e-01
3.15007985e-01 -3.06267709e-01 -8.13532770e-01 -6.79121494e-01
6.25240088e-01 -1.01211034e-01 1.02412689e+00 5.76208830e-01
7.89929748e-01 -1.65169626e-01 -5.44972718e-01 4.67013150e-01
1.56097615e+00 6.89880624e-02 2.31786683e-01 -8.89569148e-03
1.14559925e+00 6.99831247e-01 6.11613333e-01 7.56536484e-01
9.88944650e-01 3.89189720e-01 6.42862976e-01 -7.38030449e-02
-7.20805675e-02 -2.42847651e-01 1.41062990e-01 1.52855790e+00
-2.35945024e-02 -2.10645869e-01 -8.57996762e-01 6.57856703e-01
-2.19610667e+00 -1.34567463e+00 -4.58054245e-01 2.09674716e+00
1.25023469e-01 -2.63908505e-01 1.52049303e-01 9.55557451e-02
1.00537765e+00 4.08127874e-01 -7.56371975e-01 5.52214563e-01
-1.04830198e-01 -8.50848615e-01 2.39312395e-01 3.65593135e-01
-1.00518048e+00 5.84539890e-01 5.56927633e+00 3.92665476e-01
-4.83743340e-01 4.95425880e-01 5.28368175e-01 -1.03978992e-01
-4.59904313e-01 2.98112154e-01 -5.63335061e-01 2.94655710e-01
3.72615516e-01 1.61592141e-01 7.34047711e-01 6.54928029e-01
3.64761621e-01 -2.30868086e-01 -1.04000509e+00 1.11493504e+00
4.94088799e-01 -9.46161509e-01 -7.21642748e-02 5.79534829e-01
1.26012158e+00 2.55183786e-01 -1.15419880e-01 -1.78089142e-01
6.22839391e-01 -4.13026184e-01 2.70096242e-01 6.30210698e-01
3.88934821e-01 -8.65706027e-01 5.45773864e-01 9.06295061e-01
-1.40359390e+00 -2.51131982e-01 -3.14549088e-01 1.14533983e-01
4.84752864e-01 9.79371250e-01 -3.90830100e-01 1.03110766e+00
8.19460094e-01 1.31523252e+00 -5.30973375e-01 6.69170320e-01
2.09640097e-02 4.37455475e-01 -3.41838896e-02 7.57735193e-01
-4.72852625e-02 -6.07581377e-01 5.51460028e-01 5.52910149e-01
3.93464237e-01 2.28869751e-01 7.44833708e-01 6.19625568e-01
-6.55602738e-02 -1.65687636e-01 -9.56177950e-01 2.74242908e-01
4.43215102e-01 1.41008937e+00 -9.29493189e-01 -4.90292937e-01
-9.14123476e-01 1.23770726e+00 3.67884785e-01 5.35759449e-01
-4.50890929e-01 5.84679604e-01 1.27117157e-01 -3.31008360e-02
4.52428460e-01 -3.69375646e-01 -2.67246068e-01 -1.75823700e+00
2.58065939e-01 -9.14431036e-01 8.54587674e-01 -7.47799635e-01
-1.77003598e+00 2.33510032e-01 -8.16891789e-02 -1.40124190e+00
-1.51806474e-01 8.60337019e-02 -2.64222383e-01 4.85275775e-01
-1.36897349e+00 -1.38941634e+00 -2.87215799e-01 9.31465745e-01
6.92043662e-01 -1.87415391e-01 5.30197322e-01 1.92175359e-01
-4.74651694e-01 -1.29110351e-01 4.92995352e-01 -6.69599324e-02
6.89838409e-01 -1.24448109e+00 -1.19871438e-01 9.43893611e-01
3.62999767e-01 5.34640253e-01 2.60895789e-01 -8.56148005e-01
-1.67160344e+00 -9.43941295e-01 3.31426084e-01 -4.44178164e-01
4.11364526e-01 -2.00580731e-01 -9.51361179e-01 6.81138456e-01
4.14055139e-01 1.23706371e-01 1.13272834e+00 1.42266974e-01
-3.65988344e-01 -1.12983413e-01 -8.93942773e-01 2.84216434e-01
9.38218355e-01 -6.27488494e-01 -4.54835624e-01 2.58142203e-01
5.12446821e-01 -9.42105949e-02 -1.12703669e+00 2.63208419e-01
2.19424129e-01 -1.22690010e+00 1.08512056e+00 -1.43125519e-01
3.64059150e-01 -5.49143672e-01 -4.80536044e-01 -1.47079504e+00
-3.79998207e-01 -1.22412585e-01 -3.60171705e-01 1.46880078e+00
1.16697110e-01 -2.75125176e-01 9.26410198e-01 4.68638569e-01
1.19497620e-01 -5.03886700e-01 -8.01949322e-01 -4.65055496e-01
-1.35125339e-01 -1.57558993e-01 2.63119280e-01 1.33091509e+00
-1.12189025e-01 4.81094062e-01 -6.78206384e-01 5.39605021e-01
1.30127430e+00 7.15356529e-01 7.39576876e-01 -1.42611420e+00
-3.81518155e-01 1.59100413e-01 7.96881765e-02 -9.57509995e-01
2.81399578e-01 -8.06120694e-01 -3.03295910e-01 -1.99715328e+00
9.95601177e-01 -2.42172912e-01 -1.97229967e-01 3.87550215e-03
-3.90871793e-01 -1.98964123e-02 4.60537344e-01 8.19148004e-01
-1.14794338e+00 6.48383319e-01 1.39620924e+00 -2.44394317e-01
-6.41218647e-02 1.44022405e-01 -8.32288444e-01 9.22891676e-01
6.19351149e-01 -6.50506973e-01 -8.29845488e-01 -4.14503276e-01
2.68457420e-02 6.80116177e-01 1.09034635e-01 -6.55425966e-01
7.10664570e-01 -2.65147656e-01 3.08312565e-01 -1.14239323e+00
4.21832621e-01 -1.20760691e+00 2.75072187e-01 2.75304377e-01
5.06291650e-02 2.70472854e-01 -4.61086541e-01 1.32868373e+00
-4.53493357e-01 3.92364077e-02 5.36820471e-01 -7.32936382e-01
-6.48949027e-01 5.12879729e-01 -2.67135352e-01 2.07691103e-01
1.24215007e+00 -2.92659029e-02 -1.90304354e-01 -7.09721327e-01
-1.14353538e+00 5.49417913e-01 9.95003164e-01 3.51167917e-01
7.18038738e-01 -1.39172089e+00 -6.45090759e-01 7.12836757e-02
1.91721886e-01 2.00874150e-01 7.24819422e-01 9.59734619e-01
4.97933924e-02 1.66631758e-01 1.03855282e-01 -8.63547504e-01
-1.42771292e+00 1.06586933e+00 -9.74216536e-02 -5.36786556e-01
-6.40539646e-01 1.58015952e-01 3.99375498e-01 -6.97714269e-01
4.34395447e-02 3.87323886e-01 -2.49355271e-01 1.99624658e-01
3.55901897e-01 7.08489299e-01 -1.39367804e-01 -8.33891809e-01
-3.08865011e-01 7.54339516e-01 -1.10171691e-01 -3.52711469e-01
1.56160295e+00 -9.83431637e-01 -3.29352111e-01 7.31137395e-01
1.22900319e+00 1.10371083e-01 -1.36469114e+00 -4.33970451e-01
-1.68323725e-01 -4.29353863e-01 -7.98367616e-03 -5.34960628e-01
-1.46703076e+00 9.79070246e-01 3.24977994e-01 1.13028400e-01
1.09774303e+00 2.67692298e-01 6.23572886e-01 2.26514816e-01
4.60739255e-01 -1.14546621e+00 3.61903787e-01 8.39840993e-02
4.72730219e-01 -1.51167285e+00 4.00355935e-01 -6.70571327e-01
-1.07020307e+00 7.16442049e-01 6.75795913e-01 1.96529523e-01
8.77245545e-01 5.31007610e-02 1.51027665e-01 -5.45094550e-01
-8.51093948e-01 -1.65298805e-01 7.61031359e-02 6.16416872e-01
2.04237904e-02 1.38287798e-01 5.85291907e-02 2.53707409e-01
5.65081477e-01 -2.55648524e-01 6.39085174e-01 9.33074534e-01
-2.82200247e-01 -1.26410985e+00 -6.41323149e-01 3.72403145e-01
-3.74537736e-01 2.19857588e-01 -5.94010711e-01 4.88097370e-01
-9.93172526e-02 1.36340892e+00 -4.52590346e-01 -2.73915708e-01
7.97268376e-02 -7.23853800e-03 1.74987882e-01 -6.31672561e-01
-1.08327843e-01 6.36371255e-01 -3.75362784e-01 -4.56394285e-01
-1.16024768e+00 -9.04428899e-01 -1.14014018e+00 -2.42767289e-01
-4.08651501e-01 1.95346966e-01 4.14260775e-01 9.30988073e-01
4.87080514e-01 2.85714597e-01 9.31918859e-01 -6.67784214e-01
-4.65624183e-01 -5.26945829e-01 -9.69168246e-01 8.47563267e-01
1.97220147e-01 -4.79604393e-01 -3.60515565e-01 2.63910025e-01] | [8.302175521850586, 4.598137855529785] |
96dc2989-65d3-46c3-a1c3-d5cde855fff3 | reinforcement-learning-for-adaptive-video | 2105.08205 | null | https://arxiv.org/abs/2105.08205v1 | https://arxiv.org/pdf/2105.08205v1.pdf | Reinforcement Learning for Adaptive Video Compressive Sensing | We apply reinforcement learning to video compressive sensing to adapt the compression ratio. Specifically, video snapshot compressive imaging (SCI), which captures high-speed video using a low-speed camera is considered in this work, in which multiple (B) video frames can be reconstructed from a snapshot measurement. One research gap in previous studies is how to adapt B in the video SCI system for different scenes. In this paper, we fill this gap utilizing reinforcement learning (RL). An RL model, as well as various convolutional neural networks for reconstruction, are learned to achieve adaptive sensing of video SCI systems. Furthermore, the performance of an object detection network using directly the video SCI measurements without reconstruction is also used to perform RL-based adaptive video compressive sensing. Our proposed adaptive SCI method can thus be implemented in low cost and real time. Our work takes the technology one step further towards real applications of video SCI. | ['Weisong Shi', 'Aggelos K Katsaggelos', 'Xin Yuan', 'Sidi Lu'] | 2021-05-18 | null | null | null | null | ['video-compressive-sensing'] | ['computer-vision'] | [ 5.26182294e-01 -4.26529437e-01 -2.65647680e-01 -1.02405362e-01
-6.84432626e-01 -2.53547430e-01 1.84627444e-01 -6.45562172e-01
-3.85076791e-01 5.72095633e-01 1.37528524e-01 -2.49131724e-01
-4.12564687e-02 -5.21131158e-01 -1.12606847e+00 -5.81678629e-01
-1.50160074e-01 -2.16431513e-01 2.31063724e-01 4.92045954e-02
4.18385804e-01 4.87962008e-01 -1.10328674e+00 5.41086376e-01
2.08852962e-01 1.15487945e+00 7.58071244e-01 1.14183879e+00
4.16106075e-01 1.40414500e+00 -4.66494828e-01 1.28305241e-01
6.78339899e-01 -6.77822173e-01 -2.48137370e-01 4.09590423e-01
5.44817567e-01 -1.09491396e+00 -9.97010648e-01 1.07616019e+00
5.14362514e-01 -2.93599553e-02 2.23451331e-01 -9.47635710e-01
-6.82126820e-01 9.30562556e-01 -6.00387454e-01 3.70976955e-01
6.53142750e-01 3.74647081e-01 3.43572706e-01 -8.21934819e-01
3.59663427e-01 1.06461227e+00 4.59658384e-01 5.13038218e-01
-6.45009041e-01 -8.97194624e-01 -1.42954662e-01 9.54938710e-01
-1.28185022e+00 -9.00099218e-01 8.78760159e-01 -1.88327670e-01
7.51825809e-01 4.08537425e-02 9.72291410e-01 1.08690953e+00
3.38940144e-01 9.42688644e-01 9.18714583e-01 -7.00089633e-01
4.65394258e-01 -5.27170241e-01 -6.36079729e-01 3.98909032e-01
2.82109022e-01 3.33787978e-01 -4.99757290e-01 3.74164313e-01
1.18388271e+00 3.00143570e-01 -5.20630062e-01 3.22820507e-02
-1.30577791e+00 5.59464574e-01 2.93385386e-01 2.38018394e-01
-4.91704226e-01 7.13400126e-01 2.92265654e-01 6.87495887e-01
-2.26231843e-01 5.02124786e-01 3.71818454e-03 1.50312774e-03
-1.44653964e+00 -1.47037700e-01 4.56336230e-01 1.05539834e+00
6.38778329e-01 8.50586593e-01 -1.78840086e-01 5.90090632e-01
2.91187257e-01 7.31321871e-01 7.15498924e-01 -1.79832935e+00
3.56470108e-01 -2.98361033e-01 -6.21880367e-02 -8.63913834e-01
-7.64070153e-02 -2.54349083e-01 -8.85336876e-01 2.86721706e-01
-2.04732090e-01 -5.10345221e-01 -6.85153961e-01 1.41437376e+00
-5.67718223e-02 1.25499105e+00 2.69641072e-01 1.05252087e+00
2.37453476e-01 8.23816776e-01 -3.68426532e-01 -6.01633668e-01
7.45443583e-01 -1.02493787e+00 -8.10061455e-01 -2.49277487e-01
7.92907178e-02 -8.56032491e-01 4.47688729e-01 6.25402808e-01
-1.42816615e+00 -8.84588242e-01 -1.56988168e+00 3.42284560e-01
3.50280643e-01 2.09971637e-01 1.58738241e-01 4.58436579e-01
-1.31963158e+00 4.48061615e-01 -1.00760460e+00 -2.55630881e-01
2.97733545e-01 4.35654908e-01 -6.82739243e-02 -5.76161027e-01
-1.27721286e+00 4.17607844e-01 3.74789923e-01 -5.29886037e-03
-1.29095995e+00 -4.95588303e-01 -6.83541000e-01 2.13987291e-01
7.83146143e-01 -8.63121092e-01 1.13346112e+00 -1.40282547e+00
-1.96594977e+00 1.98870122e-01 3.72382075e-01 -8.21245372e-01
2.88564473e-01 -3.55805457e-01 -4.81724530e-01 8.92461121e-01
-1.38051137e-01 4.71437603e-01 1.60334718e+00 -1.21843433e+00
-4.48693812e-01 -8.06121752e-02 2.35440582e-01 1.99867621e-01
-2.56809741e-01 -1.03490569e-01 -6.19529426e-01 -7.42571354e-01
-1.39461726e-01 -9.36293304e-01 -2.61437505e-01 7.11781830e-02
9.92238596e-02 5.64536989e-01 1.17944419e+00 -7.50928998e-01
1.07917261e+00 -2.15359855e+00 2.08461002e-01 -2.29624853e-01
1.23651065e-01 6.54146492e-01 -3.41087550e-01 4.30398792e-01
1.61520481e-01 -2.69079506e-01 -1.30774245e-01 -3.40095200e-02
-6.01918936e-01 1.57846510e-01 -2.06020549e-01 4.25277919e-01
-5.56700900e-02 7.59210050e-01 -8.58748436e-01 -3.83901030e-01
5.89561045e-01 2.57513463e-01 -8.81136298e-01 4.57518280e-01
-8.57005194e-02 5.46247363e-01 -4.05572474e-01 6.57029092e-01
8.39670777e-01 -3.21506530e-01 5.56700587e-01 -4.46241349e-01
3.36407013e-02 -4.68187958e-01 -1.11942577e+00 1.79973960e+00
-3.21235657e-01 7.86079466e-01 5.18853188e-01 -1.39341879e+00
5.87587178e-01 3.88524562e-01 9.17031229e-01 -6.48596764e-01
4.68635969e-02 -7.10843271e-03 -5.91383465e-02 -1.12641227e+00
5.58588326e-01 -2.91986857e-02 3.36506516e-01 2.61564642e-01
-3.36066596e-02 -1.22713029e-01 2.08895747e-02 1.70299038e-01
1.45884514e+00 1.71842635e-01 4.86524522e-01 2.54902959e-01
6.47649586e-01 -3.03179115e-01 5.82646072e-01 1.02916706e+00
-1.83777198e-01 7.61848152e-01 -8.40192884e-02 -1.77791327e-01
-1.28344619e+00 -9.38540459e-01 3.77041399e-01 6.84988081e-01
3.95235717e-01 -1.55803666e-01 -6.69078112e-01 -9.47031155e-02
-2.53994375e-01 4.27695930e-01 -3.61077003e-02 -2.80250221e-01
-9.28718865e-01 -2.19505057e-01 5.25667727e-01 5.20927489e-01
8.72221947e-01 -9.91313279e-01 -7.98819423e-01 3.75851870e-01
-1.13060914e-01 -1.56225741e+00 -5.61216772e-01 -2.87095308e-01
-7.83825040e-01 -1.11851776e+00 -7.21749902e-01 -7.11866915e-01
2.54600823e-01 8.40084374e-01 6.40040934e-01 1.92549065e-01
6.36553168e-02 8.81149232e-01 -7.16562927e-01 -2.52117869e-02
-5.22982955e-01 -4.06931669e-01 2.60394871e-01 2.56452620e-01
1.67153031e-02 -6.31222069e-01 -1.01109374e+00 3.48758548e-02
-1.41468298e+00 -5.01933433e-02 9.51192021e-01 7.82305062e-01
4.16257352e-01 -1.34068310e-01 6.23075426e-01 -5.79102516e-01
3.29554886e-01 -7.46376514e-01 -6.94605350e-01 2.53042608e-01
-4.71591175e-01 -4.07930821e-01 1.01992261e+00 -4.65368211e-01
-8.61455321e-01 7.12683946e-02 -2.05903068e-01 -1.40263212e+00
1.68288440e-01 6.11352980e-01 1.60200626e-01 -3.77674043e-01
4.41393644e-01 4.77357656e-01 2.83840835e-01 6.72072098e-02
6.32825568e-02 9.39495206e-01 6.19171500e-01 -2.91295767e-01
6.44672692e-01 5.04912615e-01 3.99967097e-02 -1.06060815e+00
-4.65583295e-01 -3.67050290e-01 -4.79138792e-01 -5.52909911e-01
5.89062929e-01 -1.42647696e+00 -5.88348448e-01 5.38944542e-01
-8.43323946e-01 -5.65394640e-01 -1.70104101e-01 1.01983607e+00
-8.25721979e-01 9.15128767e-01 -1.00042772e+00 -5.54961681e-01
-2.47056738e-01 -1.28159046e+00 8.35132062e-01 1.81522578e-01
3.85504276e-01 -6.39398456e-01 -1.23728313e-01 5.48395157e-01
6.71192288e-01 -1.82757005e-01 3.70656639e-01 3.96813825e-02
-1.24298787e+00 2.57743634e-02 -9.70997065e-02 7.68966436e-01
2.55765170e-02 -1.46142945e-01 -6.34445906e-01 -9.74060297e-01
4.43736881e-01 -3.67521882e-01 7.74229288e-01 8.11452627e-01
1.61531901e+00 -3.57906163e-01 4.13700356e-04 1.03008306e+00
1.75806940e+00 5.11106908e-01 1.02917981e+00 2.91738510e-01
6.39466405e-01 -2.86402285e-01 6.53037608e-01 7.10141003e-01
5.38833588e-02 5.54523528e-01 5.11033058e-01 3.35543901e-01
-5.06539822e-01 -1.32360443e-01 9.46320057e-01 1.19374478e+00
-1.05104961e-01 -4.67605412e-01 -3.74076217e-01 1.81206197e-01
-1.68696868e+00 -1.43642020e+00 4.99849528e-01 2.09038997e+00
3.51865411e-01 -2.02136129e-01 -3.50910217e-01 8.48617628e-02
7.57922173e-01 3.88244301e-01 -7.86124051e-01 1.72218204e-01
-7.72529095e-02 1.80051669e-01 6.08344138e-01 4.62599605e-01
-8.06617856e-01 7.25185812e-01 6.70029354e+00 6.66587114e-01
-1.64784062e+00 1.05533540e-01 2.62481838e-01 -2.38293931e-01
5.94103783e-02 7.42686838e-02 -2.54173964e-01 6.82473004e-01
1.03511322e+00 -4.85100411e-03 1.07144392e+00 6.37033463e-01
3.99400234e-01 -2.84493691e-03 -9.68011856e-01 1.38513398e+00
4.78928417e-01 -1.71694994e+00 1.30095005e-01 -2.73507923e-01
8.31948817e-01 2.39024416e-01 1.68326706e-01 3.18978667e-01
-3.91845517e-02 -6.28746867e-01 6.31015062e-01 6.26932979e-01
1.21578622e+00 -3.06653261e-01 5.88746130e-01 2.73168027e-01
-9.67669845e-01 -5.70111454e-01 -5.18247306e-01 -3.00871599e-02
3.31374556e-01 2.99641043e-01 -7.37964451e-01 4.22251999e-01
6.77091897e-01 1.07041419e+00 -2.62682021e-01 1.11139703e+00
1.33134529e-01 8.77806246e-01 -4.22324613e-02 3.00578415e-01
2.03105107e-01 -1.75951689e-01 6.72676325e-01 1.15494192e+00
9.93877769e-01 3.39074016e-01 4.82681394e-01 3.39758843e-01
-3.77642721e-01 -4.33814704e-01 -7.24516451e-01 2.73317415e-02
4.63172674e-01 9.53983784e-01 -3.43725979e-01 -6.52011216e-01
-6.62992835e-01 1.06298792e+00 -8.32606703e-02 6.06636822e-01
-7.60898292e-01 7.49737844e-02 3.53671372e-01 2.91028619e-01
6.91860914e-01 -4.37846601e-01 3.35831314e-01 -1.53576577e+00
-2.43715510e-01 -1.19531894e+00 1.97539911e-01 -1.24209893e+00
-1.09584057e+00 2.15149283e-01 -3.77097726e-02 -1.59608817e+00
-4.11618620e-01 -5.23303151e-01 -4.09865737e-01 1.95216253e-01
-1.78815639e+00 -9.09582078e-01 -4.95551348e-01 8.69497418e-01
1.14788663e+00 -6.88779652e-01 3.32617283e-01 4.40753818e-01
-3.90117019e-01 4.39389437e-01 4.86403704e-01 1.96777917e-02
8.03969741e-01 -6.37228906e-01 -1.26323983e-01 1.19172359e+00
1.96524896e-02 3.67021918e-01 5.46633303e-01 -7.16025412e-01
-2.10032845e+00 -1.18294060e+00 -2.82394826e-01 3.78647149e-01
5.19007981e-01 3.26992393e-01 -7.59708643e-01 8.64334047e-01
7.57192194e-01 3.79286200e-01 5.27618587e-01 -8.07761371e-01
-4.05465104e-02 -4.08076882e-01 -1.09563136e+00 5.60208917e-01
8.22028756e-01 -5.37413716e-01 -2.31230229e-01 1.97291821e-01
6.97791576e-01 -5.53386271e-01 -8.47192824e-01 3.63317460e-01
5.47486424e-01 -1.16519547e+00 1.08179641e+00 -1.86949626e-01
7.71790743e-01 -3.71309906e-01 -5.72718441e-01 -1.28019691e+00
-4.75558132e-01 -6.96239829e-01 -6.10968292e-01 4.28340018e-01
-3.42360348e-01 -2.50725120e-01 8.39784086e-01 -3.09124112e-01
-2.91437000e-01 -4.90535617e-01 -8.54908109e-01 -4.74888951e-01
-3.99002999e-01 -4.92773563e-01 4.72479552e-01 7.40606308e-01
-2.37411588e-01 6.02340177e-02 -9.86031711e-01 5.36684692e-01
8.10431302e-01 5.74432686e-03 6.86779082e-01 -2.61432111e-01
-9.16665018e-01 8.10938552e-02 -5.35457492e-01 -1.60950649e+00
8.04229230e-02 -6.67704821e-01 -1.33969769e-01 -1.06735218e+00
7.37701207e-02 -2.37307996e-01 -6.53953791e-01 -1.90701738e-01
5.78321628e-02 2.49630317e-01 6.82075560e-01 5.99153638e-01
-8.32488239e-01 5.90210617e-01 1.38791251e+00 -4.15057600e-01
1.80232808e-01 8.83791409e-03 -3.68663013e-01 4.43927377e-01
7.62056112e-01 -2.76716083e-01 -5.10420561e-01 -8.84351671e-01
-2.53271759e-02 1.00279009e+00 2.47982785e-01 -1.52561736e+00
6.56536758e-01 -2.68646240e-01 7.28813827e-01 -2.75897950e-01
5.56654513e-01 -1.07133734e+00 1.73887163e-01 6.79475844e-01
-2.58116752e-01 1.07747734e-01 -8.59129131e-02 8.93758714e-01
-2.84917176e-01 -3.92500311e-01 9.72881675e-01 -4.55268323e-01
-9.43191171e-01 3.95135075e-01 -6.66035414e-01 -3.39490682e-01
1.14379156e+00 -2.48826355e-01 -1.96191259e-02 -8.47206473e-01
-4.53154713e-01 2.07777182e-03 4.40074146e-01 1.90628961e-01
1.26600170e+00 -1.26099515e+00 -6.63420439e-01 4.13008392e-01
-3.60922664e-01 -3.04620177e-01 5.42585075e-01 7.07668066e-01
-1.09826219e+00 7.54361004e-02 -6.30411565e-01 -6.61840975e-01
-9.64532673e-01 8.77699196e-01 3.34811270e-01 1.48997128e-01
-6.29243910e-01 3.38911712e-01 -3.14631581e-01 2.23366022e-01
1.66658405e-02 1.32357270e-01 -4.22833748e-02 -5.81494629e-01
8.35010648e-01 4.34127986e-01 -4.18047994e-01 -3.67060035e-01
2.32701153e-01 6.12526655e-01 -3.29210386e-02 -1.08484209e-01
1.34164417e+00 -5.64669073e-01 1.12913899e-01 2.08400026e-01
1.02059114e+00 8.04648325e-02 -1.66182375e+00 -2.34326035e-01
-3.99478883e-01 -7.66163170e-01 2.16998681e-01 -2.85788864e-01
-1.53458083e+00 5.24060369e-01 7.18749821e-01 -1.07658163e-01
1.48518276e+00 -2.94291973e-01 1.05239999e+00 5.03816307e-01
5.11516154e-01 -1.15837753e+00 4.50632930e-01 1.77558228e-01
7.98250973e-01 -1.30088413e+00 3.09826076e-01 2.44342443e-02
-5.01717627e-01 1.44099081e+00 5.74503183e-01 -4.87793654e-01
5.06814539e-01 4.72949296e-01 1.55528951e-02 1.06198661e-01
-6.67546928e-01 2.41261899e-01 -3.45601827e-01 5.37297308e-01
-3.63184251e-02 -7.82485306e-02 -4.35311049e-02 -2.54834712e-01
3.41697335e-01 5.62993586e-01 1.07195365e+00 9.27484334e-01
-4.80738491e-01 -8.75874996e-01 -6.97715521e-01 5.49019098e-01
-4.94674563e-01 9.78106707e-02 2.87348807e-01 3.51766288e-01
8.87092128e-02 9.74220037e-01 -3.44535969e-02 -6.80704176e-01
-1.57137051e-01 -4.16030794e-01 7.82981932e-01 -3.85311246e-01
-5.02562672e-02 9.70199630e-02 -5.43117285e-01 -6.99408114e-01
-7.65024185e-01 -6.51823878e-01 -9.89925325e-01 -3.36949855e-01
-4.46203351e-02 -1.82481498e-01 5.47764659e-01 8.38049114e-01
3.58723015e-01 5.48958123e-01 9.56677079e-01 -1.04177558e+00
-6.61751211e-01 -8.67542744e-01 -6.24132693e-01 1.11771367e-01
7.32744515e-01 -1.92492932e-01 -3.34019333e-01 3.89012843e-01] | [11.069012641906738, -2.0969159603118896] |
d076e7c0-7129-4f7b-a411-e8bfdaae744e | geometric-approach-for-non-pharmaceutical | 2301.08698 | null | https://arxiv.org/abs/2301.08698v1 | https://arxiv.org/pdf/2301.08698v1.pdf | Geometric approach for non pharmaceutical interventions in epidemiology | Various non pharmaceutical interventions have been settled to minimise the burden of the COVID-19 outbreak. We build a framework to analyse the dynamics of non pharmaceutical interventions, to distinguish between mitigations measures leading to objective scientific improvements and mitigations based on both political and scientific considerations. We analyse two possible strategies within this framework. Namely, we consider mitigations driven by the limited resources of the health system and mitigations where a constant set of measures is applied at different moments. We describe the optimal interventions for these scenarios. Our approach is mathematical and involves sir differential systems, it is qualitative and geometrical rather than computational. Along with the analysis of these scenarios, we collect several results that may be useful on their own, in particular on the ground when the variables are not known in real time. | ['Jean-Jacques Loeb', 'Laurent Evain'] | 2023-01-20 | null | null | null | null | ['epidemiology'] | ['medical'] | [ 4.97607619e-01 4.21949029e-01 1.23903766e-01 2.20494062e-01
-1.32995769e-02 -6.12412572e-01 8.74877334e-01 5.76936483e-01
-6.01359963e-01 1.04234338e+00 1.66469425e-01 -5.15386581e-01
-8.85700285e-01 -8.02627683e-01 -3.42765033e-01 -9.68388259e-01
-2.48124152e-01 4.87839907e-01 -1.76411751e-03 -4.12723213e-01
3.25185299e-01 7.39856601e-01 -1.20419931e+00 -1.56400263e-01
6.93726957e-01 4.02850002e-01 1.93455850e-03 7.85176456e-01
-5.59604242e-02 4.55390155e-01 -6.03966415e-01 2.09054761e-02
3.68026823e-01 -4.23114628e-01 -6.03416324e-01 -2.92924047e-01
-9.33327317e-01 7.23979399e-02 4.60391551e-01 8.25057507e-01
5.30366242e-01 -1.46135256e-01 1.00647342e+00 -9.57769573e-01
-1.33320078e-01 4.50099856e-01 -4.07776743e-01 3.09718430e-01
3.90807569e-01 3.32872659e-01 9.31634754e-02 -2.74592668e-01
6.29730642e-01 1.13059664e+00 4.57894146e-01 4.53198195e-01
-1.18594241e+00 -2.40355149e-01 -1.17282495e-01 1.41138006e-02
-1.11192799e+00 -4.06085908e-01 4.18584079e-01 -8.11557412e-01
6.01135075e-01 4.17402327e-01 6.63561583e-01 8.56678426e-01
5.78272581e-01 -9.10198018e-02 1.38973725e+00 -4.87169206e-01
4.50363994e-01 2.03987703e-01 -7.16656297e-02 -3.94926518e-02
8.48604023e-01 2.71256596e-01 8.08626339e-02 -4.28968161e-01
6.47971153e-01 8.94365460e-02 -4.11451936e-01 -2.04012349e-01
-9.05525804e-01 8.37064505e-01 -2.54933745e-01 6.81567669e-01
-9.98081625e-01 -1.94678232e-01 2.57480949e-01 3.14721614e-01
5.65039515e-01 5.12697816e-01 -4.83790100e-01 2.43124470e-01
-6.15378380e-01 3.63904983e-01 8.43585253e-01 1.86507583e-01
2.56161511e-01 -8.85558426e-02 -2.04389915e-02 -7.15244785e-02
3.20991606e-01 7.45720565e-01 -2.46949375e-01 -5.65836787e-01
1.62388533e-01 2.55128354e-01 4.47173208e-01 -8.61723304e-01
-6.14308536e-01 -3.53948563e-01 -8.26911926e-01 1.64324358e-01
5.05807340e-01 -7.22480655e-01 -4.64758992e-01 1.48002422e+00
4.81615096e-01 1.42821878e-01 2.21344084e-01 4.29177046e-01
2.27993488e-01 7.35934556e-01 2.02101752e-01 -1.09948182e+00
1.13352346e+00 5.36374450e-02 -8.75046372e-01 5.28994262e-01
5.51851034e-01 -7.75537193e-01 5.55761382e-02 3.64858449e-01
-1.04870296e+00 1.49375856e-01 -5.25037646e-01 9.85402524e-01
-7.88028598e-01 -2.92785496e-01 -6.34949729e-02 5.38973451e-01
-1.08803034e+00 7.40814924e-01 -6.19628608e-01 -7.02567339e-01
-2.84835905e-01 2.82993615e-01 2.34088421e-01 4.84713167e-01
-1.28953075e+00 1.17284656e+00 1.42052561e-01 4.19775844e-01
-5.91945708e-01 -6.45992935e-01 -4.62059408e-01 -1.57309055e-01
3.31915110e-01 -7.95280516e-01 7.39188433e-01 -5.13030529e-01
-1.34154487e+00 6.24671578e-01 1.22238241e-01 -6.29689991e-01
9.91312146e-01 1.31620020e-01 -1.23155326e-01 2.48118654e-01
-9.22688395e-02 -8.89968243e-04 3.87394398e-01 -1.16247404e+00
-4.29587841e-01 -2.52460927e-01 2.06891045e-01 1.38537884e-01
2.81717867e-01 5.58127105e-01 5.80683768e-01 -4.91614133e-01
-5.16103208e-01 -8.71549666e-01 -7.66885221e-01 -5.71705997e-01
-4.76145208e-01 4.00031917e-02 4.21929896e-01 -5.68596005e-01
8.27361107e-01 -1.60713124e+00 3.94258559e-01 3.06743592e-01
-6.77630156e-02 5.20257950e-01 2.10284069e-01 1.06911349e+00
-3.80940922e-02 3.64534676e-01 -3.56197476e-01 2.57072896e-01
-2.21285135e-01 5.84287941e-02 -3.52213591e-01 7.78093874e-01
2.90348291e-01 4.36435372e-01 -8.69727135e-01 -1.59897640e-01
4.41338658e-01 6.17655337e-01 -2.45804697e-01 1.11649029e-01
-1.37933999e-01 9.76342499e-01 -9.14225876e-01 2.91811943e-01
7.81696439e-01 3.18467408e-01 4.85506386e-01 2.05223039e-01
-7.75869727e-01 -1.12682968e-01 -1.22804511e+00 6.00912452e-01
-1.15001097e-01 1.85785145e-02 4.22751129e-01 -1.40837848e+00
5.98129451e-01 7.07497954e-01 9.47921872e-01 -3.14861238e-01
4.44736093e-01 2.21566260e-01 1.81703269e-02 -6.00483418e-01
2.75115483e-02 -4.15094018e-01 1.39599353e-01 4.27060962e-01
-3.55298638e-01 4.35275352e-03 1.89043134e-01 -2.41178975e-01
8.63056242e-01 -1.49488792e-01 6.84151709e-01 -1.01305032e+00
7.38645434e-01 6.67211935e-02 3.64740461e-01 4.49933022e-01
-2.27398306e-01 -5.34265116e-02 8.58925343e-01 -3.17327619e-01
-9.25837636e-01 -7.86543846e-01 -2.75016248e-01 2.72251695e-01
1.21524796e-01 1.96905565e-02 -9.33650911e-01 -1.88938171e-01
-2.83746153e-01 5.07852793e-01 -7.45106459e-01 -1.11361347e-01
-7.03742087e-01 -1.32052255e+00 1.33836910e-01 -2.91545361e-01
6.72780424e-02 -9.01052654e-01 -1.10205841e+00 4.46820527e-01
3.22560906e-01 -6.00330174e-01 3.12794894e-01 2.32530236e-01
-9.16781664e-01 -1.57547617e+00 -1.01123106e+00 -1.29391864e-01
5.31303883e-01 2.21653841e-02 7.85154402e-01 4.42026854e-02
-1.81093544e-01 4.80675131e-01 -2.63342112e-01 -1.00831950e+00
-6.29039228e-01 -3.08671564e-01 4.36995506e-01 -1.58970043e-01
1.68906853e-01 -4.33840215e-01 -7.40592539e-01 1.61041677e-01
-1.05361354e+00 -4.09010381e-01 3.75834286e-01 3.48065853e-01
3.77040476e-01 2.57246345e-01 8.13807786e-01 -8.49431455e-01
8.46371591e-01 -8.62273216e-01 -5.95120013e-01 3.04528683e-01
-4.76159096e-01 3.61727998e-02 4.71141249e-01 -2.17828676e-01
-9.90660846e-01 -2.15061739e-01 -8.07535946e-02 1.24592006e-01
-2.96149164e-01 3.72362584e-01 9.05956328e-02 -1.36899471e-01
4.16353166e-01 -7.05100819e-02 4.86862436e-02 -7.67851293e-01
2.01465920e-01 5.26388228e-01 -1.52395487e-01 -5.06676435e-01
7.80693710e-01 6.29785001e-01 3.21067065e-01 -1.25899732e+00
-1.77236140e-01 -3.44146609e-01 -5.28033316e-01 -4.58199412e-01
9.55612183e-01 -2.89770424e-01 -7.68927515e-01 6.10630095e-01
-1.27796102e+00 -3.36229742e-01 -4.38586861e-01 5.36800981e-01
-6.01697743e-01 -5.00602908e-02 -3.56649548e-01 -1.39283407e+00
-2.55405605e-01 -1.04247022e+00 5.99530160e-01 2.23036423e-01
-1.37176380e-01 -1.23242056e+00 5.07399619e-01 -1.29030824e-01
6.26587391e-01 1.00489342e+00 7.59980917e-01 -5.14202178e-01
-4.48073626e-01 3.29466581e-01 1.95574790e-01 -2.02832252e-01
2.61238545e-01 2.75703788e-01 -5.70014596e-01 -3.25192720e-01
5.67661345e-01 4.26530331e-01 4.87625271e-01 7.31149971e-01
3.33048433e-01 -7.20934987e-01 -4.54095632e-01 1.64021835e-01
1.68450308e+00 6.41068578e-01 4.76314694e-01 3.87868732e-01
-3.40345651e-02 1.26106775e+00 6.99009359e-01 5.92820704e-01
9.91867762e-03 6.50748432e-01 4.22970384e-01 -3.06149989e-01
5.23715496e-01 3.51538390e-01 1.08529836e-01 5.86105168e-01
-5.67602575e-01 -2.45070353e-01 -1.09269762e+00 6.29416764e-01
-1.73869729e+00 -9.58787739e-01 -5.32649755e-01 2.30744147e+00
5.46742380e-01 9.91355926e-02 3.91841769e-01 3.62941511e-02
8.66538882e-01 -1.58534888e-02 -1.40193971e-02 -9.43292558e-01
-1.11949474e-01 2.42152110e-01 8.40513289e-01 6.19551837e-01
-7.55563378e-01 1.70802593e-01 7.27926540e+00 3.60977501e-01
-1.06790757e+00 1.57352254e-01 4.56067234e-01 -7.16771046e-03
-2.86413401e-01 9.99643747e-03 -4.01628941e-01 5.59167862e-01
1.26988077e+00 -5.16964793e-01 1.67232484e-01 6.87523261e-02
9.32683527e-01 -3.53196114e-01 -4.67188567e-01 -1.73873790e-02
-4.24855143e-01 -1.31216812e+00 -2.05151230e-01 4.56238866e-01
5.21748424e-01 -1.39801383e-01 -9.07056928e-02 -1.85853288e-01
1.79707646e-01 -8.18793476e-01 4.47816581e-01 1.02408266e+00
3.94027799e-01 -9.10299182e-01 9.43168461e-01 6.61375582e-01
-9.80727196e-01 2.21512571e-01 -2.04048585e-02 -4.78509009e-01
6.23524368e-01 5.92271984e-01 -5.55610955e-01 7.90961564e-01
2.15937674e-01 8.72412845e-02 -7.52118602e-02 1.03554177e+00
-5.29613197e-02 4.29776579e-01 -4.28630531e-01 -1.96296886e-01
2.38835633e-01 -6.49263799e-01 9.08337295e-01 1.06914485e+00
2.71166891e-01 3.51899207e-01 -2.76455075e-01 8.50224674e-01
9.02734578e-01 1.85464382e-01 -1.02644289e+00 7.47396424e-02
1.10561736e-01 7.38437712e-01 -8.88009369e-01 -7.80326352e-02
5.65445833e-02 2.32882723e-01 -3.77892375e-01 3.93049628e-01
-7.94161618e-01 -2.64259249e-01 7.93832958e-01 4.37058836e-01
-2.47740708e-02 -1.43460304e-01 -9.40911099e-02 -8.39375734e-01
-2.09240928e-01 -4.57088917e-01 3.55883390e-01 -2.13301793e-01
-7.45426238e-01 3.80969912e-01 8.21116507e-01 -7.10113883e-01
-5.25316656e-01 -4.77157056e-01 -8.84229720e-01 8.33537638e-01
-1.57571340e+00 -6.20699525e-01 6.42079532e-01 1.12320155e-01
2.12787241e-02 2.73780674e-01 6.99448228e-01 1.60355881e-01
-6.06172800e-01 -2.44805366e-01 5.57462752e-01 -5.55594802e-01
1.82262450e-01 -9.28715289e-01 -9.26869512e-02 9.46626604e-01
-6.36412442e-01 7.10015595e-01 1.33433759e+00 -8.42231870e-01
-1.13310754e+00 -8.26373339e-01 1.17886937e+00 -1.76940501e-01
7.63871491e-01 -2.11235315e-01 -6.33858025e-01 1.74597904e-01
5.43007731e-01 -6.83500051e-01 4.03586209e-01 -2.63138562e-01
5.73294699e-01 2.24143863e-01 -1.53592598e+00 5.78051746e-01
6.73307180e-01 -7.52030537e-02 -6.24319553e-01 6.25883579e-01
5.04297078e-01 7.04476535e-02 -8.89014900e-01 5.80936611e-01
1.99586660e-01 -8.38876247e-01 9.98584032e-01 -7.53534198e-01
-5.33784293e-02 -2.10853770e-01 3.44092458e-01 -1.41403139e+00
-5.56102693e-02 -9.31217074e-01 3.57106030e-01 1.16670656e+00
2.28366897e-01 -1.07436395e+00 1.60901397e-01 4.21196282e-01
5.76234460e-01 -9.53976333e-01 -1.01755178e+00 -8.53430808e-01
3.58475238e-01 4.78483774e-02 4.57513869e-01 9.73314285e-01
-8.58239010e-02 -6.97658285e-02 -3.83876473e-01 1.18861981e-01
7.98709214e-01 -8.56831968e-02 3.02838713e-01 -1.09151042e+00
-8.31547454e-02 -4.25949574e-01 -3.21908087e-01 1.93511188e-01
-3.03373754e-01 -7.50331059e-02 -2.59390891e-01 -1.40790987e+00
-5.80551364e-02 -2.53636330e-01 -1.57075897e-01 1.60047021e-02
1.35082871e-01 -1.78916901e-01 2.21626043e-01 7.61595592e-02
8.35369620e-03 1.97449446e-01 8.97789478e-01 1.38309389e-01
-4.05761898e-01 1.89475968e-01 -5.78212380e-01 8.22104156e-01
1.15407634e+00 -6.11266315e-01 -3.02185506e-01 1.51888013e-01
4.91163820e-01 9.53319669e-02 2.74494022e-01 -6.91964090e-01
2.44677346e-02 -8.54018569e-01 -3.35130513e-01 -4.44558442e-01
-1.14088662e-01 -8.93728793e-01 9.14369822e-01 1.15253973e+00
-4.67511415e-02 -7.99460113e-02 9.84439850e-02 6.42694890e-01
2.46837765e-01 -4.57982898e-01 1.00376821e+00 -1.87803820e-01
1.04182810e-01 -7.85196424e-02 -6.91376209e-01 7.95296766e-03
1.55417073e+00 -1.53016508e-01 -4.44402248e-01 -1.84341460e-01
-8.27800035e-01 2.64798284e-01 5.24128139e-01 -8.30774754e-02
1.71709887e-03 -7.75866807e-01 -8.86691689e-01 -2.93516040e-01
-5.13657391e-01 -2.53624231e-01 5.19110918e-01 1.26906931e+00
-7.98228920e-01 7.71946073e-01 -3.17153484e-01 -2.05176175e-01
-1.09426200e+00 1.04949880e+00 6.22308016e-01 -4.36562359e-01
-2.37693310e-01 7.47672245e-02 3.50855321e-01 -9.44105685e-02
-4.52619679e-02 -1.25671715e-01 -7.97695041e-01 4.84952420e-01
5.98651350e-01 6.85787380e-01 -1.23614177e-01 -8.07756126e-01
-5.98865390e-01 7.69177139e-01 5.38329840e-01 -7.11090863e-02
1.58422160e+00 -3.63634795e-01 -2.77977467e-01 2.97433078e-01
7.64323950e-01 9.85572338e-02 -8.92731309e-01 8.89374465e-02
2.32985020e-01 -7.30813667e-02 -1.97414085e-01 -4.91241276e-01
-7.31867433e-01 5.41752279e-01 4.62774456e-01 8.25780153e-01
1.12014365e+00 -1.54468626e-01 2.13480115e-01 -3.53214070e-02
2.36942381e-01 -9.29115295e-01 -6.84345901e-01 2.21668288e-01
1.06216705e+00 -6.18521035e-01 7.18017370e-02 -5.47451854e-01
-2.21269220e-01 8.95619452e-01 -2.11569294e-01 -2.87081838e-01
7.73888946e-01 4.46408868e-01 -1.59035712e-01 -4.20279205e-01
-6.80360317e-01 -4.21930522e-01 -8.58895108e-02 8.74448359e-01
2.02365726e-01 2.24838212e-01 -1.39345574e+00 2.92917818e-01
4.23649758e-01 6.24647811e-02 7.33314872e-01 9.45967495e-01
-3.27908725e-01 -1.16115475e+00 -6.21732175e-01 2.79969484e-01
-7.45911419e-01 6.15277514e-02 -5.76371372e-01 1.04652846e+00
4.45284873e-01 1.00813687e+00 -2.69833118e-01 1.52807772e-01
4.43116546e-01 -1.66666269e-01 1.97407171e-01 -3.38937819e-01
-6.94976449e-01 1.38477415e-01 2.13084638e-01 -2.49411613e-01
-9.24601495e-01 -9.27046776e-01 -9.53011036e-01 -5.16565084e-01
-1.89873949e-01 5.18384457e-01 7.50618279e-01 1.05786765e+00
1.84832901e-01 6.01066887e-01 7.45267451e-01 -7.31212735e-01
-5.35197675e-01 -6.47150040e-01 -6.67417824e-01 -1.76563025e-01
3.16590756e-01 -6.18422091e-01 -5.20572007e-01 -1.40135735e-01] | [5.922079086303711, 4.32187032699585] |
ed552948-8068-42f3-a3e9-7556c15187aa | attention-based-joint-detection-of-object-and | 2007.02419 | null | https://arxiv.org/abs/2007.02419v1 | https://arxiv.org/pdf/2007.02419v1.pdf | Attention-based Joint Detection of Object and Semantic Part | In this paper, we address the problem of joint detection of objects like dog and its semantic parts like face, leg, etc. Our model is created on top of two Faster-RCNN models that share their features to perform a novel Attention-based feature fusion of related Object and Part features to get enhanced representations of both. These representations are used for final classification and bounding box regression separately for both models. Our experiments on the PASCAL-Part 2010 dataset show that joint detection can simultaneously improve both object detection and part detection in terms of mean Average Precision (mAP) at IoU=0.5. | ['Tara Vijaykumar', 'Keval Morabia', 'Jatin Arora'] | 2020-07-05 | null | null | null | null | ['semantic-part-detection'] | ['computer-vision'] | [ 5.56406863e-02 1.70855924e-01 -8.06412324e-02 -5.66854596e-01
-7.98532248e-01 -1.29508734e-01 5.11889756e-01 1.30766779e-01
-5.45325756e-01 3.91846746e-01 -1.05165750e-01 1.60141110e-01
3.12718391e-01 -5.47462225e-01 -1.00764525e+00 -3.66563827e-01
-2.00155839e-01 2.81927139e-01 9.43076789e-01 -8.76279697e-02
9.98307485e-04 8.71809483e-01 -2.07354832e+00 5.83005250e-01
3.25717837e-01 1.32414806e+00 1.84589416e-01 8.33222926e-01
-7.62694255e-02 8.93470526e-01 -6.40875697e-01 -5.73367953e-01
3.50363463e-01 3.53269845e-01 -7.71684349e-01 -6.87697157e-02
9.25190151e-01 -5.73169470e-01 -3.15060735e-01 9.49316382e-01
3.59960169e-01 1.03658676e-01 7.18288302e-01 -1.34014499e+00
-4.71163005e-01 6.03445113e-01 -1.02167368e+00 4.26438570e-01
1.90919206e-01 7.13124052e-02 8.34512293e-01 -1.03939795e+00
4.37502354e-01 1.70603240e+00 6.06738389e-01 5.93536615e-01
-1.11058629e+00 -7.34517932e-01 2.83636481e-01 4.62338328e-01
-1.51415813e+00 -3.65873933e-01 3.85643572e-01 -2.68054940e-02
1.44403124e+00 1.74368829e-01 4.07161236e-01 8.40828896e-01
7.85330161e-02 1.27204216e+00 6.48144543e-01 -3.52546483e-01
-1.67708635e-01 2.45124504e-01 5.38077831e-01 5.84884167e-01
5.84691107e-01 8.93160030e-02 -1.38411552e-01 -5.12659177e-02
5.22191346e-01 2.91995168e-01 -3.81048098e-02 -3.32110912e-01
-7.25257695e-01 9.31096315e-01 9.64663148e-01 3.82990167e-02
-5.28958678e-01 5.09752035e-01 3.13510239e-01 -2.08621055e-01
2.97311246e-01 1.77909493e-01 -7.21201718e-01 3.14301372e-01
-7.06003726e-01 3.93652707e-01 6.54571116e-01 1.19494998e+00
6.60688996e-01 -4.11497355e-01 -4.44765925e-01 8.99519503e-01
4.56002533e-01 4.52236444e-01 3.35749567e-01 -8.12380910e-01
1.89898893e-01 8.24724376e-01 3.93803194e-02 -6.27624154e-01
-5.18118858e-01 -3.50486130e-01 -3.24482858e-01 1.47908449e-01
1.20924987e-01 1.11831322e-01 -1.55337965e+00 1.50937629e+00
5.91868460e-01 1.90014914e-01 -2.43321106e-01 8.23668420e-01
1.29821956e+00 3.25931877e-01 5.55947900e-01 4.26223636e-01
1.93522775e+00 -1.39623177e+00 -5.58038831e-01 -3.33519846e-01
5.84012985e-01 -7.54893005e-01 2.03088537e-01 1.27630845e-01
-1.03728688e+00 -9.41467583e-01 -1.14734983e+00 -4.36890692e-01
-7.18951702e-01 3.87263358e-01 7.03141332e-01 3.35111111e-01
-8.80225301e-01 6.23619020e-01 -7.73330331e-01 -4.73477036e-01
9.22407806e-01 5.30686557e-01 -4.36735272e-01 -2.55368471e-01
-6.89128935e-01 1.23365128e+00 6.40616417e-01 6.76305965e-02
-8.06966126e-01 -4.95730400e-01 -8.91004503e-01 6.89117834e-02
4.32623118e-01 -6.05867565e-01 1.28391850e+00 -6.64050698e-01
-1.06252384e+00 9.64910388e-01 7.36080036e-02 -7.81566978e-01
4.38707858e-01 -6.84733510e-01 -3.47743273e-01 2.93229911e-02
7.31332675e-02 1.47951734e+00 9.04231191e-01 -8.98542702e-01
-1.16676128e+00 -5.83392262e-01 3.12484652e-02 -5.01188710e-02
2.10958838e-01 3.09175789e-01 -7.70861149e-01 -3.57824713e-01
-2.95548905e-02 -7.11484253e-01 -2.04048783e-01 3.52509290e-01
-1.14523754e-01 -5.74286103e-01 1.15727222e+00 -5.70319951e-01
6.60814404e-01 -2.15172935e+00 -2.81091064e-01 -1.51830405e-01
2.29985714e-01 4.73275393e-01 -4.57811654e-01 -1.75501257e-01
-2.17582345e-01 -1.94247395e-01 1.28817126e-01 -4.23504621e-01
-3.55323344e-01 1.68142304e-01 5.46863452e-02 5.71034372e-01
7.74766862e-01 1.18335044e+00 -4.91136372e-01 -6.31779850e-01
3.24330658e-01 6.40051365e-01 -4.04240876e-01 1.87213242e-01
-1.37584582e-01 -2.83755660e-01 -3.91185701e-01 8.57756197e-01
1.14537168e+00 -1.44121617e-01 -1.99842736e-01 -3.68344575e-01
2.90983289e-01 8.24575499e-02 -1.16068518e+00 1.41466641e+00
-2.78804153e-01 4.03852761e-01 1.87094837e-01 -6.27251983e-01
8.47375154e-01 -1.43527770e-02 8.96188840e-02 -6.08978033e-01
4.31182146e-01 2.59488970e-02 2.37276420e-01 -1.04727879e-01
5.77127576e-01 4.07824397e-01 2.57562816e-01 -3.98930274e-02
3.33517462e-01 1.78681016e-01 3.08240861e-01 9.65694338e-02
1.11537385e+00 2.58245915e-01 5.50428689e-01 -2.84140974e-01
6.36192858e-01 -3.70704383e-02 2.16470987e-01 7.72681832e-01
-4.93439585e-01 4.91005301e-01 1.49559513e-01 -5.92297256e-01
-8.71197462e-01 -9.49708998e-01 -4.89479274e-01 1.51433694e+00
2.68940121e-01 -3.26753646e-01 -5.62961340e-01 -1.20602024e+00
5.41734815e-01 6.12481236e-01 -9.88032699e-01 -2.57022828e-01
-7.46962070e-01 -6.45787179e-01 3.57856631e-01 1.17740881e+00
3.88311476e-01 -1.17222965e+00 -9.79373813e-01 1.62540942e-01
2.37165228e-01 -1.41419947e+00 -1.87767688e-02 7.12337434e-01
-6.84723079e-01 -1.37414408e+00 -7.39494205e-01 -7.71940649e-01
5.56116819e-01 4.14538413e-01 1.06764436e+00 2.12618813e-01
-8.99019003e-01 2.40520492e-01 -3.78109962e-01 -8.31482768e-01
-2.16941461e-01 -1.83694109e-01 -2.66988426e-01 -2.67264068e-01
7.33694315e-01 -6.94056973e-02 -5.69877982e-01 3.62534821e-01
-6.07291639e-01 -2.44822741e-01 8.35755110e-01 6.38615310e-01
4.70240146e-01 -3.28989178e-01 2.67093509e-01 -3.83707434e-01
2.74994913e-02 -3.05845976e-01 -7.25633442e-01 3.13541055e-01
-9.78059545e-02 -1.16976416e-02 5.43467812e-02 -5.12408316e-01
-7.82197595e-01 3.67448509e-01 -7.51956031e-02 -6.98853910e-01
-4.12641287e-01 -4.08717573e-01 -7.57916598e-03 -9.63110775e-02
6.02708519e-01 2.16779634e-02 -1.08702436e-01 -8.39426696e-01
5.20244956e-01 7.95519292e-01 6.58444524e-01 -1.73909098e-01
4.11890417e-01 4.05178905e-01 -4.03485447e-03 -5.23434460e-01
-9.20648694e-01 -9.47765470e-01 -7.31336534e-01 7.73422420e-02
9.67894912e-01 -1.31413925e+00 -6.95705771e-01 3.24416012e-01
-1.32396400e+00 1.75552860e-01 -2.70073026e-01 4.35015798e-01
-3.40956479e-01 1.32870361e-01 -5.93412220e-01 -8.06581259e-01
-6.53415024e-01 -1.02253366e+00 1.63313878e+00 4.13757384e-01
6.33940175e-02 -2.79583395e-01 -3.27399731e-01 2.59314686e-01
2.84477115e-01 -2.64797453e-03 4.23134834e-01 -1.03394783e+00
-5.75069249e-01 -6.76012278e-01 -8.89101207e-01 3.85279715e-01
-1.52831376e-01 7.22897723e-02 -1.19930851e+00 -2.83650219e-01
-3.34675521e-01 -3.32642019e-01 1.44445872e+00 3.29233468e-01
1.38443196e+00 1.60894275e-01 -8.12225401e-01 3.93494666e-01
1.40371013e+00 5.05976602e-02 6.45424426e-01 7.37204105e-02
6.06969655e-01 3.76993626e-01 7.69994617e-01 3.37660074e-01
-3.67997144e-03 8.77407789e-01 8.03056896e-01 1.30850852e-01
-6.04196966e-01 6.38769045e-02 9.87475440e-02 -1.57701820e-01
7.19960853e-02 -4.26416397e-02 -6.45020902e-01 6.74434364e-01
-1.86978674e+00 -6.51657522e-01 -2.12393433e-01 1.84016550e+00
2.66927600e-01 1.99505478e-01 3.89005452e-01 -1.80489570e-01
1.00115752e+00 -2.99870282e-01 -3.97358298e-01 -3.86125207e-01
2.15972081e-01 2.15685278e-01 8.19449067e-01 -1.39274284e-01
-1.57965481e+00 1.01029301e+00 7.06428432e+00 9.57895458e-01
-6.20380044e-01 1.54061362e-01 4.91817206e-01 -7.46556446e-02
5.60429096e-01 -3.95981759e-01 -1.50050616e+00 2.29428530e-01
8.90160799e-01 3.67735773e-01 -1.42445024e-02 1.50792015e+00
-6.19303763e-01 -3.98071706e-01 -1.33609474e+00 7.96101511e-01
1.68923363e-01 -9.26285148e-01 1.77535608e-01 -1.37353286e-01
5.38814664e-01 2.77524322e-01 -3.06123465e-01 8.28932285e-01
4.81646270e-01 -9.10941243e-01 7.00393319e-01 2.06403092e-01
4.57883477e-01 -7.97867954e-01 1.07570791e+00 2.33921587e-01
-1.31073356e+00 -2.31807485e-01 -7.10022211e-01 1.53500319e-01
-1.49296731e-01 1.99614450e-01 -9.26139295e-01 3.96612465e-01
9.73309994e-01 5.40631413e-01 -8.79524052e-01 1.34536326e+00
-2.11522698e-01 4.51310948e-02 -6.25282764e-01 -6.93547577e-02
2.01704264e-01 7.18739986e-01 3.45840096e-01 1.30832267e+00
-8.49772543e-02 1.39700882e-02 2.67736286e-01 9.64336753e-01
-2.63349086e-01 -7.67103806e-02 -2.33057410e-01 3.87088746e-01
3.29626352e-01 1.64854836e+00 -9.22358572e-01 -7.13263988e-01
-4.61630970e-01 8.76204193e-01 6.12470031e-01 -1.87468156e-01
-1.31620455e+00 -3.48616004e-01 7.79211402e-01 -4.08473127e-02
1.16795576e+00 4.26079035e-01 1.25314444e-01 -8.22261095e-01
-8.87406021e-02 -4.16717410e-01 4.47082996e-01 -6.31152213e-01
-1.12278092e+00 4.79918838e-01 1.02546394e-01 -7.99538076e-01
7.25236684e-02 -1.09679699e+00 -6.28174424e-01 5.69423497e-01
-1.41655898e+00 -1.51441646e+00 -3.56599540e-01 5.57859957e-01
6.21704459e-01 1.05324194e-01 7.09689617e-01 2.94456542e-01
-7.02853382e-01 6.90570772e-01 -2.21453682e-01 2.99773097e-01
4.53938365e-01 -1.13658404e+00 6.52572393e-01 6.35252059e-01
1.19763337e-01 3.18109483e-01 4.53862876e-01 -6.84502959e-01
-1.09244120e+00 -1.43849456e+00 3.88326079e-01 -6.20216548e-01
3.42492759e-01 -3.33610713e-01 -6.55821145e-01 7.45966256e-01
-2.92499393e-01 6.53304815e-01 1.21946894e-01 1.03766665e-01
-6.88303113e-01 1.04823753e-01 -1.45445216e+00 2.76525438e-01
1.06944561e+00 -5.19833826e-02 -7.26215124e-01 3.23201030e-01
8.57239783e-01 -2.68206030e-01 -6.88599765e-01 9.96887565e-01
6.37717009e-01 -7.87217796e-01 1.15174782e+00 -7.04156935e-01
2.02478930e-01 -3.04619014e-01 -4.69573855e-01 -7.52003133e-01
-5.37961364e-01 2.19584540e-01 -5.56829989e-01 9.04886901e-01
1.11199699e-01 -2.44296491e-01 5.76527715e-01 1.51439995e-01
1.36247827e-02 -7.33953893e-01 -9.31041837e-01 -9.79625165e-01
-3.21072042e-01 -4.60498601e-01 5.79309225e-01 1.43955693e-01
-4.83015358e-01 3.25152338e-01 1.77753028e-02 1.15524247e-01
5.63167274e-01 2.23125577e-01 8.44103694e-01 -1.20323861e+00
-1.16005443e-01 -5.51757336e-01 -1.04793870e+00 -9.29613590e-01
-1.17973387e-01 -6.54064119e-01 2.43488342e-01 -1.41668701e+00
7.01704860e-01 1.46033466e-01 -5.56920946e-01 7.26899624e-01
-3.53071511e-01 5.97040832e-01 4.14023787e-01 -1.15394101e-01
-1.00631166e+00 3.64401042e-01 1.02991831e+00 -1.56033590e-01
1.12235032e-01 -5.59493154e-03 -4.74393517e-01 8.93999338e-01
4.27692831e-01 -6.28429770e-01 3.17490935e-01 4.78716753e-02
-5.50483882e-01 -3.11422676e-01 6.51664853e-01 -1.40772820e+00
-1.80303037e-01 3.78497928e-01 9.48232234e-01 -1.16652095e+00
5.94628453e-01 -1.01071405e+00 -3.70034724e-01 7.19465494e-01
-1.37338772e-01 -3.12654644e-01 6.39177918e-01 7.76206553e-01
1.34958094e-02 -8.35209936e-02 1.13855159e+00 -1.04396641e-01
-1.01629353e+00 9.69577432e-02 -3.94535623e-02 -3.60407144e-01
1.51217377e+00 -2.02434197e-01 -3.35689008e-01 5.96267246e-02
-8.21373880e-01 4.41164345e-01 5.59324771e-02 7.85083294e-01
5.80108047e-01 -1.20344019e+00 -8.52178097e-01 3.19105983e-01
4.22392011e-01 -5.78780770e-02 2.22351134e-01 7.68033385e-01
-2.56625801e-01 5.78364372e-01 -3.78317177e-01 -7.52682209e-01
-1.73044765e+00 9.45117116e-01 3.42321575e-01 -2.95163929e-01
-5.32704055e-01 1.24015403e+00 3.96636724e-01 -2.54478484e-01
5.01791298e-01 -4.88460869e-01 -3.36609125e-01 -3.24640423e-03
8.20547283e-01 5.34400344e-01 5.27344793e-02 -7.32121885e-01
-7.27676988e-01 5.07453263e-01 -6.73110962e-01 5.75129449e-01
1.27076185e+00 1.51308268e-01 4.16863598e-02 -1.30533621e-01
1.44194210e+00 -4.48862642e-01 -1.11758101e+00 -2.73834258e-01
-6.15671054e-02 -2.83780724e-01 9.05991718e-02 -1.04334772e+00
-1.10428774e+00 6.89149439e-01 1.02485061e+00 -2.44884610e-01
7.47947156e-01 5.24261594e-01 5.47672451e-01 4.18691874e-01
3.04147035e-01 -7.95428872e-01 -1.09081566e-01 3.92660350e-01
9.74263072e-01 -1.43513906e+00 2.97282517e-01 -6.07302606e-01
-3.16026628e-01 1.05689991e+00 1.13361919e+00 -4.64919120e-01
7.12613046e-01 2.98742115e-01 -3.47802311e-01 -3.64443094e-01
-8.76527488e-01 -8.20436478e-01 7.70381570e-01 5.23841381e-01
2.51026899e-01 1.89187065e-01 -1.48854613e-01 5.25815308e-01
1.73451841e-01 -1.09943077e-01 1.89650394e-02 1.05507064e+00
-8.31421256e-01 -6.10185087e-01 -5.36126018e-01 6.50850952e-01
-7.00257719e-01 1.74818560e-01 -3.51669610e-01 1.13532794e+00
3.27274382e-01 7.28292346e-01 4.29233968e-01 -3.55217040e-01
2.55879939e-01 7.98490569e-02 6.28114641e-01 -6.85090840e-01
-6.12695158e-01 2.26568375e-02 6.60343096e-02 -1.00365293e+00
-3.23685914e-01 -3.88482749e-01 -1.15952718e+00 7.47557729e-02
-9.36182678e-01 -2.60696411e-01 8.08009744e-01 8.12312543e-01
2.27833226e-01 7.21655250e-01 1.87778354e-01 -1.11144376e+00
-7.13453948e-01 -1.38390899e+00 -5.47128320e-01 3.95510733e-01
2.33459368e-01 -9.49846447e-01 -1.29512653e-01 -4.81730878e-01] | [9.251012802124023, 0.8955840468406677] |
0d1be7fe-65e7-4fcf-a902-2e01b7184e95 | keyword-localisation-in-untranscribed-speech | 2202.01107 | null | https://arxiv.org/abs/2202.01107v1 | https://arxiv.org/pdf/2202.01107v1.pdf | Keyword localisation in untranscribed speech using visually grounded speech models | Keyword localisation is the task of finding where in a speech utterance a given query keyword occurs. We investigate to what extent keyword localisation is possible using a visually grounded speech (VGS) model. VGS models are trained on unlabelled images paired with spoken captions. These models are therefore self-supervised -- trained without any explicit textual label or location information. To obtain training targets, we first tag training images with soft text labels using a pretrained visual classifier with a fixed vocabulary. This enables a VGS model to predict the presence of a written keyword in an utterance, but not its location. We consider four ways to equip VGS models with localisations capabilities. Two of these -- a saliency approach and input masking -- can be applied to an arbitrary prediction model after training, while the other two -- attention and a score aggregation approach -- are incorporated directly into the structure of the model. Masked-based localisation gives some of the best reported localisation scores from a VGS model, with an accuracy of 57% when the system knows that a keyword occurs in an utterance and need to predict its location. In a setting where localisation is performed after detection, an $F_1$ of 25% is achieved, and in a setting where a keyword spotting ranking pass is first performed, we get a localisation P@10 of 32%. While these scores are modest compared to the idealised setting with unordered bag-of-word-supervision (from transcriptions), these models do not receive any textual or location supervision. Further analyses show that these models are limited by the first detection or ranking pass. Moreover, individual keyword localisation performance is correlated with the tagging performance from the visual classifier. We also show qualitatively how and where semantic mistakes occur, e.g. that the model locates surfer when queried with ocean. | ['Herman Kamper', 'Dan Oneata', 'Kayode Olaleye'] | 2022-02-02 | null | null | null | null | ['keyword-spotting'] | ['speech'] | [ 5.55539310e-01 6.31982207e-01 3.79014872e-02 -4.91206914e-01
-1.41333842e+00 -7.53119528e-01 8.45518112e-01 2.54946947e-01
-6.53979421e-01 3.08305830e-01 2.30847865e-01 -3.52299631e-01
2.84677058e-01 -1.81479916e-01 -1.02775729e+00 -8.70922327e-01
-3.23387906e-02 3.89802814e-01 5.60464501e-01 1.20846145e-02
2.82392323e-01 1.95896342e-01 -1.79813790e+00 5.73804379e-01
2.97784567e-01 1.12887955e+00 7.74812102e-01 9.83587444e-01
-1.12378724e-01 8.64315987e-01 -7.95697987e-01 -1.70593470e-01
9.08013955e-02 -3.58274966e-01 -8.48776817e-01 2.64571190e-01
4.41803515e-01 -7.23944828e-02 -2.20423788e-02 8.68180096e-01
4.44071263e-01 -1.04491636e-01 7.80042171e-01 -1.00410604e+00
-4.45371896e-01 3.11229825e-01 -2.58126646e-01 2.23501369e-01
6.80719137e-01 2.03459799e-01 1.27087283e+00 -1.21162474e+00
6.06649101e-01 1.03596187e+00 5.97677708e-01 3.56686264e-01
-1.14716423e+00 -1.58157021e-01 2.25757182e-01 -3.05027906e-02
-1.55295956e+00 -8.42023015e-01 3.59976113e-01 -5.36011338e-01
1.18636274e+00 4.29648399e-01 3.04028600e-01 9.70645607e-01
-1.40321791e-01 8.60242248e-01 1.03684926e+00 -7.52576828e-01
2.74805933e-01 6.43314302e-01 -1.91186368e-01 7.80750096e-01
-5.40406883e-01 6.51596040e-02 -8.54457438e-01 6.16094768e-02
4.44219351e-01 -4.12344933e-01 -4.38267916e-01 -3.91231000e-01
-1.36374438e+00 6.88461363e-01 5.47049344e-01 3.46269369e-01
-3.25363904e-01 2.31692687e-01 2.29737878e-01 2.15737000e-01
5.52454710e-01 4.90275770e-01 -4.71921712e-01 6.49618804e-02
-1.35899019e+00 -8.05361569e-02 4.94292766e-01 9.76594090e-01
9.29939508e-01 -9.42774490e-02 -3.27326030e-01 6.74954712e-01
3.09578687e-01 5.61671972e-01 5.68661630e-01 -6.93926036e-01
3.80123794e-01 2.60780901e-01 3.04138750e-01 -8.22333395e-01
-3.09009492e-01 -4.16208386e-01 -8.33026469e-02 2.05483243e-01
5.01672268e-01 1.71765983e-01 -1.31465614e+00 1.68015635e+00
1.46596491e-01 8.31280723e-02 2.43121520e-01 1.05535746e+00
7.80468583e-01 7.31879771e-01 9.11010876e-02 -1.31588966e-01
1.58908153e+00 -8.66738021e-01 -4.86187816e-01 -6.39984071e-01
7.68531561e-01 -6.91752553e-01 1.28017509e+00 2.27937803e-01
-9.19093788e-01 -4.92115080e-01 -1.02320421e+00 1.14506088e-01
-5.50834119e-01 2.34061629e-01 1.46655098e-01 4.69483227e-01
-1.69884408e+00 7.58679658e-02 -4.38498169e-01 -6.32130504e-01
-4.01071385e-02 3.48609865e-01 -4.03034747e-01 1.51925012e-01
-1.25285244e+00 1.14173639e+00 1.56714618e-01 -4.63234223e-02
-1.19601476e+00 -3.85973267e-02 -9.52892959e-01 1.75787061e-02
4.72591460e-01 -1.93322673e-01 1.57473087e+00 -1.36665511e+00
-1.04684019e+00 1.13763762e+00 -5.94840527e-01 -5.43560445e-01
2.62928039e-01 3.12256485e-01 -3.41515422e-01 5.03122032e-01
3.63082588e-01 9.78753984e-01 1.21136022e+00 -1.50917006e+00
-7.44565070e-01 -8.63930136e-02 2.52732839e-02 5.42194903e-01
1.57411337e-01 3.35348666e-01 -7.44966984e-01 -3.83997709e-01
1.26344875e-01 -6.77916229e-01 5.78531884e-02 -1.40676767e-01
-3.63305211e-01 -1.70552328e-01 6.33930624e-01 -8.30445170e-01
8.97690654e-01 -2.41785622e+00 -1.35504618e-01 2.07227126e-01
-2.20040022e-03 1.02940820e-01 -1.41952500e-01 5.01230597e-01
-8.69747922e-02 2.24358320e-01 -1.53463170e-01 -5.79048574e-01
-2.40848307e-02 1.65051028e-01 -4.86284107e-01 5.25529444e-01
3.19794148e-01 9.52778816e-01 -7.64864087e-01 -5.30716658e-01
1.89253896e-01 4.26446170e-01 -2.97153056e-01 2.43975431e-01
-2.83120036e-01 1.64413303e-01 -8.56639594e-02 5.45590818e-01
7.98095837e-02 -2.79315978e-01 -2.44969111e-02 1.22859046e-01
-1.70117095e-01 5.28814733e-01 -1.01751184e+00 1.52549469e+00
-4.80784148e-01 7.70318627e-01 3.32858741e-01 -9.74576652e-01
7.16664314e-01 6.16756797e-01 -1.45705968e-01 -7.70895064e-01
-2.42591426e-01 3.40922117e-01 -2.27502987e-01 -4.27460134e-01
4.17589068e-01 -9.53610316e-02 -1.88125551e-01 2.06386968e-01
3.27507555e-01 -2.91879978e-02 -2.93545276e-01 4.81067151e-01
1.20499992e+00 1.04236200e-01 1.25016421e-01 -1.60601988e-01
3.54757756e-01 1.93846062e-01 5.52928932e-02 1.08803904e+00
-1.51299253e-01 8.52721334e-01 5.21534979e-01 -7.38325343e-03
-9.51990187e-01 -8.36019576e-01 -1.10273005e-03 1.41695940e+00
1.60894260e-01 -2.97281176e-01 -9.44690049e-01 -7.11200476e-01
-2.53641605e-01 8.13456297e-01 -7.22793519e-01 -5.83665557e-02
-5.25630303e-02 -3.04017097e-01 5.40525138e-01 3.62938493e-01
5.48447520e-02 -1.40231693e+00 -7.86202192e-01 -2.11714879e-02
-2.10105523e-01 -1.00436962e+00 -3.90376449e-01 7.64880896e-01
-1.60737962e-01 -7.23690391e-01 -8.70837986e-01 -1.02784216e+00
7.95689702e-01 1.78469852e-01 1.06519938e+00 2.09517449e-01
-4.99304309e-02 6.83919728e-01 -5.73606730e-01 -2.74803400e-01
-4.22272921e-01 -1.44411638e-01 -4.72539850e-02 1.66949809e-01
2.73969531e-01 8.60903934e-02 -6.53221309e-01 4.11366433e-01
-6.05176210e-01 -3.34210368e-03 8.28716218e-01 8.87616456e-01
4.46872413e-01 -2.06986994e-01 4.19930071e-01 -4.50817049e-01
2.24901140e-01 -3.09150159e-01 -4.63014364e-01 2.91604012e-01
-4.17740166e-01 -4.73776646e-02 1.84007809e-01 -4.74485248e-01
-5.89929998e-01 5.10434330e-01 -1.01519242e-01 -3.61411422e-01
-6.14582300e-01 4.40934002e-01 -1.42822668e-01 -9.57009941e-02
6.74606442e-01 5.02383411e-01 2.92618368e-02 -2.65523583e-01
2.93811798e-01 9.42442477e-01 6.25618815e-01 -1.09466106e-01
5.93711615e-01 2.63737172e-01 -6.48267150e-01 -9.78756011e-01
-7.97578394e-01 -8.88210893e-01 -4.38901842e-01 -2.82121480e-01
9.57063258e-01 -1.12950015e+00 -3.96699309e-01 1.49151608e-01
-1.15172899e+00 -4.88090336e-01 -1.76339731e-01 1.88397363e-01
-6.55618310e-01 1.13905922e-01 -2.15134338e-01 -1.23066282e+00
2.33309746e-01 -1.30796409e+00 1.58420336e+00 -1.42318323e-01
-2.07103699e-01 -7.77400136e-01 -5.19015729e-01 3.26314837e-01
2.76730329e-01 -3.29424798e-01 4.88895953e-01 -1.22651303e+00
-4.36833113e-01 -2.88294435e-01 -2.68344939e-01 1.69533968e-01
-2.37347767e-01 -4.43079412e-01 -1.48078287e+00 -2.09841356e-01
2.70296168e-03 -3.41943532e-01 8.82132292e-01 1.61444649e-01
7.74483919e-01 -4.78710830e-01 -4.11133379e-01 1.13209948e-01
1.26590300e+00 2.84738868e-01 3.77859026e-01 3.62034291e-01
5.22971809e-01 9.04614568e-01 6.34226501e-01 -1.27294157e-02
1.94985762e-01 9.69489396e-01 6.52022302e-01 -4.21578407e-01
-1.02757223e-01 -3.58893961e-01 4.04905587e-01 3.08014363e-01
4.38658923e-01 -5.83628953e-01 -1.12169397e+00 9.31639314e-01
-1.71991682e+00 -8.60785007e-01 3.80692966e-02 2.32965589e+00
7.28051543e-01 2.86852539e-01 1.33119330e-01 2.81574845e-01
8.13604355e-01 2.22995803e-01 -2.07133129e-01 -3.13696623e-01
-1.38888806e-01 -9.40050036e-02 6.82821155e-01 9.20028389e-01
-1.24784517e+00 1.21371877e+00 6.44792128e+00 8.63231182e-01
-1.20068669e+00 2.81912923e-01 7.10434794e-01 7.84718525e-03
-1.61066055e-01 6.04501888e-02 -7.71779895e-01 5.27248800e-01
1.21105111e+00 4.94869113e-01 3.46969962e-01 7.47909904e-01
2.82775640e-01 -5.59410751e-01 -9.84051585e-01 7.94665933e-01
4.09715742e-01 -9.89306271e-01 -2.46971205e-01 1.06272720e-01
1.35469556e-01 1.06874011e-01 1.92949519e-01 2.18185827e-01
6.31978661e-02 -1.23817432e+00 1.30075419e+00 4.35384214e-01
9.94236290e-01 -5.44633389e-01 7.09795773e-01 5.93213797e-01
-9.66479897e-01 9.62935835e-02 -1.22316986e-01 -3.22433398e-03
1.57497928e-01 1.24215320e-01 -1.43124139e+00 6.82898685e-02
7.03883111e-01 1.01204611e-01 -6.73833728e-01 1.00267804e+00
-4.95584399e-01 9.36646879e-01 -3.27201158e-01 -1.06959559e-01
5.77727675e-01 4.50003415e-01 5.22321820e-01 1.48645902e+00
1.26227334e-01 -7.77088329e-02 1.33528456e-01 5.76192439e-01
3.20564359e-01 1.74586415e-01 -7.01958299e-01 7.47446045e-02
3.70017529e-01 1.06566012e+00 -8.54594648e-01 -3.56742799e-01
-5.19704700e-01 1.36362481e+00 1.93937987e-01 5.85717320e-01
-7.62933969e-01 -4.30508852e-01 2.73219377e-01 3.52014959e-01
7.21578062e-01 5.59966788e-02 1.42636057e-02 -6.73812151e-01
6.86864778e-02 -6.76034391e-01 -2.51507238e-02 -1.37302661e+00
-6.37228727e-01 6.03665173e-01 -1.09235741e-01 -9.19576585e-01
-5.40424883e-01 -6.33551896e-01 -2.30984047e-01 1.07642543e+00
-1.41893697e+00 -1.12587631e+00 9.44867581e-02 3.84884000e-01
6.33174598e-01 -8.35848227e-02 8.96637559e-01 -8.37899819e-02
-3.49782258e-02 4.70437467e-01 -1.96975783e-01 2.40371168e-01
6.04075432e-01 -1.44308352e+00 4.13703352e-01 9.21002209e-01
7.54631937e-01 5.24412096e-01 9.77171779e-01 -5.11299551e-01
-9.69670594e-01 -9.61003602e-01 1.40147126e+00 -7.47046888e-01
6.05068028e-01 -8.83131504e-01 -1.02396488e+00 5.81749499e-01
2.58508682e-01 1.27099633e-01 3.69957238e-01 -1.23929851e-01
-2.51083016e-01 2.69157082e-01 -9.48248923e-01 3.43851417e-01
7.43128777e-01 -1.08052993e+00 -7.64274716e-01 6.29050016e-01
9.30599749e-01 -4.00009364e-01 -2.15493947e-01 1.48362845e-01
2.52626002e-01 -9.62308228e-01 8.85126948e-01 -2.87162304e-01
-5.48443869e-02 -5.23457527e-01 -3.86154383e-01 -1.04970062e+00
-3.40210216e-04 -4.91326123e-01 3.51026833e-01 1.22063279e+00
9.47837174e-01 -3.96786451e-01 5.09898007e-01 3.86243403e-01
-2.01785073e-01 -5.88825643e-01 -1.12186050e+00 -5.60460806e-01
-5.65355897e-01 -6.89760089e-01 2.07035184e-01 6.59260094e-01
1.22500482e-04 4.29081440e-01 -2.77163357e-01 5.72282076e-01
1.46793291e-01 -4.63797629e-01 4.38239545e-01 -8.50983739e-01
-2.17107385e-01 -1.88893929e-01 -6.83264077e-01 -1.23159182e+00
1.73264116e-01 -6.63715720e-01 6.14290476e-01 -1.55140650e+00
-2.20512077e-02 -1.87632769e-01 -2.79514521e-01 7.76254296e-01
6.82971105e-02 5.17789960e-01 -7.11526722e-02 4.23922569e-01
-9.19364274e-01 1.15687035e-01 4.67130274e-01 -1.48632517e-02
-5.49317943e-03 -7.50787854e-02 -5.80496013e-01 6.95222199e-01
4.38566923e-01 -3.61391693e-01 -3.76766920e-01 -3.03329259e-01
2.00033277e-01 2.10915834e-01 8.33183885e-01 -7.57994950e-01
4.70806956e-01 2.23294541e-01 3.22846919e-01 -3.54985386e-01
5.76033831e-01 -8.25766146e-01 -1.87933162e-01 -2.88621872e-03
-4.68781829e-01 -1.43038377e-01 1.08861849e-01 6.88689709e-01
-3.23906779e-01 -3.92985940e-01 6.06523812e-01 -2.85220057e-01
-1.10980213e+00 -3.80378217e-01 -6.66982532e-01 -2.15534776e-01
8.43788624e-01 -4.20310259e-01 -1.08091585e-01 -7.34041035e-01
-1.06622136e+00 -1.91966584e-03 6.06280863e-01 2.43374169e-01
6.75259113e-01 -9.21029985e-01 -2.65706271e-01 3.05220008e-01
4.39060599e-01 -4.28820178e-02 -2.98740387e-01 8.66248906e-01
-1.79887876e-01 6.53818309e-01 5.88217378e-01 -8.86263788e-01
-1.27587032e+00 6.52254581e-01 4.72851664e-01 2.44275704e-01
-4.14395988e-01 1.21766925e+00 4.97219801e-01 -3.18129808e-01
5.96220613e-01 -2.65024573e-01 -1.79018900e-01 1.44269645e-01
4.67667073e-01 -4.60090488e-01 3.45680952e-01 -1.26150632e+00
-7.14436710e-01 4.41155404e-01 2.12181106e-01 -6.78981125e-01
8.14479470e-01 -2.88803250e-01 2.88885087e-01 7.09266782e-01
1.06990993e+00 1.23895183e-01 -1.28544569e+00 -1.39841482e-01
2.23615021e-01 -1.69879973e-01 2.94806272e-01 -1.18728220e+00
-6.10356808e-01 8.60872269e-01 4.95254219e-01 3.80945235e-01
8.28011751e-01 6.37754560e-01 4.44811434e-02 1.83550566e-01
3.07282239e-01 -8.69531035e-01 5.04615530e-02 3.65329206e-01
9.34431732e-01 -1.22370708e+00 -4.21328306e-01 -2.17502460e-01
-1.07620227e+00 5.01012921e-01 2.19535589e-01 1.87215731e-01
2.13541701e-01 2.27287993e-01 3.77934992e-01 -3.76306087e-01
-7.49682069e-01 -7.00229704e-01 5.17783821e-01 5.74302554e-01
2.67172694e-01 -2.39805579e-01 2.68609136e-01 3.06416482e-01
-1.74707636e-01 -5.76750815e-01 1.72318831e-01 7.90122271e-01
-8.26508462e-01 -4.58659470e-01 -4.23020095e-01 2.08950177e-01
-4.24666077e-01 -4.75873441e-01 -6.52942419e-01 5.88147640e-01
6.65341690e-02 1.32053900e+00 2.12090760e-01 -4.00017709e-01
2.65719980e-01 6.35802388e-01 -9.48475301e-02 -9.55629587e-01
-5.67925692e-01 3.91425818e-01 1.85594335e-01 -5.85020423e-01
-5.23129702e-01 -7.31111109e-01 -1.23897731e+00 5.17271817e-01
-6.81486607e-01 4.72067177e-01 9.49414432e-01 1.07068408e+00
2.20303208e-01 3.22761804e-01 3.68280649e-01 -1.00898385e+00
-2.40518481e-01 -8.75247478e-01 -5.61378777e-01 7.98678398e-02
7.45846808e-01 -6.23713315e-01 -9.80357707e-01 1.50237799e-01] | [10.490619659423828, 1.384629249572754] |
46f2a5d3-1fdb-4046-9905-658481ed2e72 | relation-regularized-scene-graph-generation | 2202.10826 | null | https://arxiv.org/abs/2202.10826v1 | https://arxiv.org/pdf/2202.10826v1.pdf | Relation Regularized Scene Graph Generation | Scene graph generation (SGG) is built on top of detected objects to predict object pairwise visual relations for describing the image content abstraction. Existing works have revealed that if the links between objects are given as prior knowledge, the performance of SGG is significantly improved. Inspired by this observation, in this article, we propose a relation regularized network (R2-Net), which can predict whether there is a relationship between two objects and encode this relation into object feature refinement and better SGG. Specifically, we first construct an affinity matrix among detected objects to represent the probability of a relationship between two objects. Graph convolution networks (GCNs) over this relation affinity matrix are then used as object encoders, producing relation-regularized representations of objects. With these relation-regularized features, our R2-Net can effectively refine object labels and generate scene graphs. Extensive experiments are conducted on the visual genome dataset for three SGG tasks (i.e., predicate classification, scene graph classification, and scene graph detection), demonstrating the effectiveness of our proposed method. Ablation studies also verify the key roles of our proposed components in performance improvement. | ['Xuelong Li', 'Heng Tao Shen', 'Nicu Sebe', 'Peng Wang', 'Jingkuan Song', 'Lianli Gao', 'Yuyu Guo'] | 2022-02-22 | null | null | null | null | ['scene-graph-generation'] | ['computer-vision'] | [ 5.90565085e-01 4.34642106e-01 -1.24409646e-01 -6.58674896e-01
4.18093391e-02 -3.05019230e-01 5.56397319e-01 3.90622497e-01
4.43653651e-02 1.74630731e-01 2.33711123e-01 -1.49446532e-01
-1.83058500e-01 -1.18946767e+00 -1.13025987e+00 -4.12877440e-01
-9.01548639e-02 2.12063909e-01 4.16879296e-01 1.79030150e-01
1.32850140e-01 5.38806021e-01 -1.50725138e+00 4.16102916e-01
7.65997887e-01 1.03563964e+00 5.49860895e-01 6.06995463e-01
-1.56571388e-01 1.31548357e+00 -3.81211758e-01 -4.52050239e-01
7.30974078e-02 -3.32539648e-01 -9.65662956e-01 6.42035067e-01
3.16472679e-01 -2.25448728e-01 -7.02148616e-01 1.18052351e+00
-2.97494698e-02 3.27572912e-01 5.76586306e-01 -1.47230065e+00
-1.24364877e+00 8.95445168e-01 -5.15039146e-01 7.67975077e-02
1.99138463e-01 9.14669931e-02 1.41987693e+00 -7.27711439e-01
4.66235459e-01 1.62239373e+00 4.51015308e-02 2.93862611e-01
-1.18725348e+00 -5.78061342e-01 5.51231682e-01 4.54464853e-01
-1.48300898e+00 -1.04161941e-01 9.15715516e-01 -4.49165076e-01
7.86797702e-01 2.17505425e-01 7.72311628e-01 4.94761527e-01
-8.20302367e-02 9.95368659e-01 6.10299706e-01 -3.83699477e-01
-1.35219619e-01 1.29320085e-01 2.92598456e-01 1.08035803e+00
5.18680096e-01 -1.63125828e-01 -6.05872333e-01 2.64716059e-01
8.46173465e-01 1.45080462e-01 -4.01502699e-01 -4.69988078e-01
-1.07566512e+00 6.82803929e-01 1.15534270e+00 -1.64698232e-02
-2.68099725e-01 4.10505801e-01 9.21285748e-02 -1.62096277e-01
3.15222383e-01 3.94266278e-01 1.76974833e-02 5.24713576e-01
-7.78840706e-02 2.94724405e-02 4.20627862e-01 1.35771012e+00
1.02411044e+00 -2.55891562e-01 -6.61154449e-01 7.11849332e-01
6.58484519e-01 2.17089027e-01 -4.50491868e-02 -6.46037579e-01
3.46172154e-01 1.30161572e+00 -3.30586880e-01 -1.50859225e+00
-2.83959001e-01 -4.05574113e-01 -8.15324187e-01 -2.71016985e-01
-5.53174466e-02 3.88484597e-01 -1.00936913e+00 1.78146648e+00
3.76658857e-01 4.72964913e-01 5.63786440e-02 9.48397040e-01
1.34425342e+00 5.79911232e-01 2.13970363e-01 1.29006609e-01
1.52111733e+00 -1.06199133e+00 -6.12686932e-01 -3.37965369e-01
5.23616731e-01 -4.03219700e-01 9.74108338e-01 -7.27041587e-02
-8.18132401e-01 -7.93501854e-01 -9.37570572e-01 -2.27683455e-01
-2.75956184e-01 1.23607211e-01 1.06464696e+00 1.65363461e-01
-8.94449890e-01 3.69553208e-01 -6.45903409e-01 -3.64189237e-01
8.73581946e-01 3.24911267e-01 -2.71063566e-01 -2.97173768e-01
-9.89429116e-01 5.15971541e-01 8.19384933e-01 3.03975165e-01
-9.75583017e-01 -3.69667888e-01 -1.11216164e+00 4.44931149e-01
5.11126876e-01 -6.36228561e-01 8.52568924e-01 -7.01566339e-01
-9.08469498e-01 9.59245682e-01 -7.27801472e-02 -4.46055263e-01
-1.20461516e-01 -1.27862077e-02 -3.88198882e-01 1.89578772e-01
1.37744322e-01 9.17534590e-01 6.57016218e-01 -1.39020514e+00
-7.85766065e-01 -1.91505775e-01 4.35535908e-01 4.29146081e-01
-1.54395536e-01 -1.31276026e-01 -9.31751847e-01 -3.49247456e-01
4.05492336e-01 -6.73859894e-01 -2.44931594e-01 -8.11753422e-03
-8.72698784e-01 -5.76320946e-01 7.58613110e-01 -4.08113539e-01
9.68979299e-01 -2.25325632e+00 -1.33750653e-02 4.59841311e-01
5.69657981e-01 6.05501942e-02 -2.82234222e-01 9.78571475e-02
-1.49604768e-01 -3.58629748e-02 -2.02023610e-01 -2.84530632e-02
-9.83205363e-02 3.43055189e-01 -2.70836890e-01 3.65753055e-01
7.31808066e-01 1.29593801e+00 -9.71890688e-01 -4.66718704e-01
1.97579548e-01 2.73250371e-01 -5.76792598e-01 5.68232298e-01
-4.64663893e-01 3.47087175e-01 -5.81026137e-01 4.51100349e-01
6.60667121e-01 -8.83084714e-01 2.81183094e-01 -7.08417594e-01
4.52603102e-01 2.76205808e-01 -9.90840793e-01 1.37593091e+00
-1.05128832e-01 5.95012784e-01 -5.11387646e-01 -1.12313628e+00
1.22367764e+00 -2.73668140e-01 1.37620062e-01 -7.10831523e-01
1.95968896e-01 -3.38728905e-01 1.58621028e-01 -5.27291775e-01
2.90297806e-01 3.83423835e-01 2.45031387e-01 1.72217831e-01
1.58204675e-01 3.07076443e-02 2.28725761e-01 6.76604211e-01
1.07012069e+00 6.94746673e-02 1.92666158e-01 -1.73557043e-01
6.10728920e-01 -2.36594632e-01 5.32601833e-01 6.78460598e-01
-8.94704089e-03 3.92986119e-01 7.11962283e-01 -3.16490471e-01
-6.21889591e-01 -9.36016738e-01 1.91227540e-01 9.85656679e-01
9.16121721e-01 -6.47003055e-01 -4.83422458e-01 -8.15116227e-01
1.06408522e-01 6.59382463e-01 -7.55919278e-01 -5.92990637e-01
-4.76212770e-01 -5.95657289e-01 2.03314185e-01 7.13245630e-01
6.09792590e-01 -1.20715630e+00 -3.87476295e-01 -9.73053947e-02
-9.65649858e-02 -1.49998438e+00 -4.88985956e-01 -1.02563798e-02
-5.27889431e-01 -1.38300753e+00 3.93246897e-02 -1.15214527e+00
1.16552496e+00 5.37107527e-01 1.16788304e+00 4.24570709e-01
-2.80614883e-01 4.40762579e-01 -3.93507600e-01 -1.56640634e-01
-2.58760899e-01 -3.15242082e-01 -3.22992831e-01 5.04531801e-01
2.48102903e-01 -4.33991671e-01 -5.51054001e-01 1.96892485e-01
-8.46601725e-01 6.92350924e-01 7.37796128e-01 6.60229981e-01
8.15047383e-01 2.92533308e-01 2.69690186e-01 -1.25709140e+00
4.07848686e-01 -2.74132818e-01 -7.84171045e-01 6.64397657e-01
-3.82736325e-01 3.03686738e-01 3.54761153e-01 -2.80810475e-01
-1.15395248e+00 1.59304962e-01 3.04631531e-01 -5.63022494e-01
-9.06278417e-02 5.09498835e-01 -5.31498611e-01 -4.67651635e-02
4.04845148e-01 2.69633055e-01 -2.86950320e-01 -1.13656513e-01
6.93138182e-01 2.78767645e-01 5.86942732e-01 -5.35796463e-01
8.43911171e-01 3.64409566e-01 3.00844997e-01 -4.01033431e-01
-1.21247268e+00 -4.72795427e-01 -4.78082210e-01 -2.47207314e-01
1.12479603e+00 -9.12717104e-01 -9.91620898e-01 1.95350587e-01
-1.13405430e+00 -2.51820117e-01 -1.83947906e-01 4.16559845e-01
-3.88825476e-01 1.91761449e-01 -3.60625267e-01 -7.14786947e-01
-2.71503776e-02 -1.01076925e+00 1.14150536e+00 4.81264651e-01
3.15252632e-01 -8.74370992e-01 -4.71253276e-01 2.55160302e-01
-9.40662697e-02 1.40670285e-01 1.17029178e+00 -5.59280992e-01
-1.10350955e+00 5.67770042e-02 -1.00517118e+00 1.84391007e-01
3.39960963e-01 -4.24980409e-02 -8.36576104e-01 -6.12143055e-02
-4.09289598e-01 -1.60856053e-01 7.90546060e-01 3.16944063e-01
1.72746825e+00 -2.72227228e-01 -6.39709532e-01 7.44452536e-01
1.21923435e+00 3.08691651e-01 5.28238058e-01 -1.34923905e-01
1.31820190e+00 6.54206157e-01 7.17607796e-01 1.88975915e-01
6.42697215e-01 5.57909548e-01 6.94103181e-01 -2.53553271e-01
-5.20481646e-01 -5.61514318e-01 -1.87074635e-02 5.88036597e-01
2.64015030e-02 -3.73803347e-01 -7.58712828e-01 3.55668247e-01
-1.99490976e+00 -5.19472182e-01 -3.83523792e-01 1.63778520e+00
5.51695704e-01 1.04519270e-01 -2.33745694e-01 -2.40710855e-01
1.03581166e+00 7.88628384e-02 -6.14662588e-01 1.25133544e-01
-2.30292231e-01 -7.27096796e-02 3.29987794e-01 3.60038489e-01
-1.09405529e+00 1.20997965e+00 5.00182343e+00 5.49216509e-01
-8.36071670e-01 -3.00459504e-01 9.07732964e-01 4.21580195e-01
-4.35674697e-01 1.76366970e-01 -6.97830379e-01 1.46897301e-01
1.20892957e-01 -2.90069997e-01 5.82438588e-01 9.01212573e-01
-2.70665795e-01 8.91968310e-02 -1.22581923e+00 1.08999872e+00
2.86670357e-01 -1.38875687e+00 5.54334641e-01 -1.66604772e-01
6.00745678e-01 -2.87528485e-01 -1.01457871e-01 2.20068216e-01
6.25601113e-01 -9.20192003e-01 7.07291186e-01 4.72820848e-01
6.20607316e-01 -5.72701871e-01 5.02131402e-01 5.85117452e-02
-1.63964093e+00 4.98013059e-03 -5.32185793e-01 7.89721534e-02
1.16286024e-01 4.92784142e-01 -1.25559914e+00 8.18625629e-01
6.76318288e-01 1.05568063e+00 -9.55720305e-01 9.56431985e-01
-6.60712481e-01 6.85255110e-01 3.09604090e-02 -9.44227129e-02
6.63665980e-02 -1.73241064e-01 3.70870292e-01 9.88917887e-01
-1.48266152e-01 3.67419809e-01 3.25201809e-01 1.41345370e+00
-2.79290676e-01 -7.92366490e-02 -5.29107094e-01 -2.54311711e-01
3.20487857e-01 1.51656449e+00 -1.11288595e+00 -4.09837991e-01
-4.67159629e-01 9.39325213e-01 8.06571424e-01 5.11611223e-01
-9.37334120e-01 -3.71955544e-01 5.48813581e-01 -5.76625504e-02
4.06929225e-01 -7.15995282e-02 -7.55476132e-02 -1.04801250e+00
-1.16345011e-01 -4.70554024e-01 4.98086780e-01 -1.01785588e+00
-1.28414428e+00 5.41133642e-01 -5.40158339e-03 -8.38065982e-01
3.91960979e-01 -7.02596784e-01 -3.77545446e-01 6.76676810e-01
-1.32722807e+00 -1.41758370e+00 -7.52033651e-01 7.01910496e-01
2.75950223e-01 6.56059757e-02 4.25909013e-01 1.98859930e-01
-4.45754766e-01 3.94369543e-01 -7.38457620e-01 3.59307885e-01
1.24698691e-01 -1.17764187e+00 5.21368802e-01 1.13248682e+00
5.96666217e-01 6.67970598e-01 1.76121607e-01 -8.79124522e-01
-1.40503752e+00 -1.70991921e+00 5.60493767e-01 -2.71905392e-01
4.26426739e-01 -7.41874635e-01 -1.03552496e+00 8.30706596e-01
-1.67564496e-01 4.43716496e-01 4.00007963e-01 1.03989497e-01
-5.39828598e-01 -1.03600889e-01 -7.01732695e-01 6.25816166e-01
1.67058122e+00 -6.06920421e-01 -3.24830741e-01 5.25638819e-01
1.27466607e+00 -4.67914194e-01 -5.90761602e-01 7.13828087e-01
1.45017952e-01 -6.06245697e-01 1.01984024e+00 -8.13147485e-01
4.52373683e-01 -6.16154909e-01 -2.49315903e-01 -1.00672066e+00
-7.16516376e-01 -1.03026636e-01 -2.55681545e-01 1.31587350e+00
2.70819157e-01 -5.35370946e-01 5.77001929e-01 4.55806643e-01
-2.73305476e-01 -6.45370662e-01 -3.22263509e-01 -5.77246726e-01
-7.75645614e-01 -4.45608139e-01 9.18507040e-01 9.60296094e-01
-3.61199588e-01 7.26015270e-01 -1.49655908e-01 7.14477420e-01
5.56786537e-01 4.90923047e-01 7.75892735e-01 -1.16659272e+00
-4.01466250e-01 -2.92736232e-01 -8.04015875e-01 -1.33797038e+00
3.04389387e-01 -1.20102608e+00 2.58491606e-01 -1.91020894e+00
6.63876951e-01 -6.18786633e-01 -3.57941717e-01 6.81192815e-01
-7.64656544e-01 1.51776433e-01 1.67757764e-01 -7.75981769e-02
-9.82731164e-01 6.49838269e-01 1.69192994e+00 -4.05636400e-01
-4.04274464e-02 -2.62538493e-01 -9.42391872e-01 6.06218517e-01
4.41105306e-01 -2.73979366e-01 -8.26650143e-01 -4.76187557e-01
2.65419304e-01 -1.73588812e-01 6.42558694e-01 -7.86174715e-01
2.19259813e-01 -2.59579390e-01 3.88138086e-01 -5.56626618e-01
5.73383197e-02 -8.17977250e-01 1.31131306e-01 3.37664992e-01
-5.25943875e-01 -3.70543301e-01 1.85250212e-02 7.95187533e-01
-1.97495624e-01 1.38526246e-01 5.08296549e-01 8.11948627e-02
-1.04365826e+00 6.92460358e-01 1.83126107e-01 -1.48216546e-01
1.08479786e+00 -1.16525225e-01 -4.75549817e-01 -3.33625108e-01
-6.95054591e-01 3.31600368e-01 9.99424681e-02 5.75196266e-01
8.81560326e-01 -1.50270998e+00 -5.83998263e-01 4.02725518e-01
5.41386247e-01 5.14401376e-01 1.29656702e-01 4.71872389e-01
-3.32062423e-01 2.57986844e-01 -5.51222675e-02 -8.53500545e-01
-1.38785267e+00 7.33921885e-01 1.69581354e-01 -5.32974210e-03
-7.18560636e-01 1.31777442e+00 1.10570467e+00 -4.11592275e-01
1.50082737e-01 -5.74592888e-01 -4.87770915e-01 -4.59018648e-01
2.99265951e-01 -6.70298114e-02 -2.20078155e-01 -7.55299568e-01
-3.99540007e-01 3.74073714e-01 -1.86283916e-01 4.02170539e-01
1.13944006e+00 -1.25331432e-01 -3.88669431e-01 1.77062199e-01
1.03194463e+00 -3.28411192e-01 -1.32376301e+00 -4.96568352e-01
-5.99563494e-03 -7.06600964e-01 -1.70542136e-01 -4.39728975e-01
-1.36000073e+00 7.36966193e-01 3.70152056e-01 3.78725320e-01
1.36296654e+00 6.10205591e-01 3.03164482e-01 3.50756645e-01
1.95917919e-01 -3.26919377e-01 4.59367931e-01 3.12999368e-01
8.61098111e-01 -1.19816566e+00 -7.34350383e-02 -1.23632514e+00
-6.75679445e-01 8.24741364e-01 1.16914451e+00 -2.03343228e-01
4.92486537e-01 -1.36373997e-01 -4.81200367e-01 -6.90564692e-01
-6.08557582e-01 -5.99806905e-01 7.79042006e-01 6.06917322e-01
1.53046906e-01 2.50909358e-01 9.59790871e-02 5.70166111e-01
-3.80316265e-02 -3.17729175e-01 1.00035779e-01 5.20110965e-01
-3.74706924e-01 -6.97470784e-01 5.66698126e-02 6.69767380e-01
4.00523469e-02 -1.63316309e-01 -4.50024694e-01 4.42515194e-01
2.35616848e-01 9.07038748e-01 2.78669626e-01 -5.70297301e-01
3.16968769e-01 -5.12098134e-01 6.20937645e-01 -9.61453140e-01
-6.72603399e-02 -2.36034364e-01 1.08560205e-01 -7.78167486e-01
-5.21316707e-01 -2.25086764e-01 -1.59038830e+00 1.18235528e-01
-6.30384386e-01 5.79358079e-02 1.96652696e-01 8.35286736e-01
4.08788443e-01 1.10557163e+00 5.98772585e-01 -3.31458181e-01
2.10150450e-01 -6.39922023e-01 -6.47925198e-01 8.40903878e-01
5.58370873e-02 -8.22417080e-01 8.39209259e-02 3.43764991e-01] | [10.284773826599121, 1.6479332447052002] |
629bb3c0-af70-4657-a1a0-30fe6ea3932b | federated-chinese-word-segmentation-with | null | null | https://aclanthology.org/2021.findings-acl.376 | https://aclanthology.org/2021.findings-acl.376.pdf | Federated Chinese Word Segmentation with Global Character Associations | null | ['Yan Song', 'Han Qin', 'Guimin Chen', 'Yuanhe Tian'] | null | null | null | null | findings-acl-2021-8 | ['chinese-word-segmentation'] | ['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] | [-7.474621772766113, 3.589365243911743] |
f6b87e0c-fdb9-4bea-a066-9088c8a02b88 | linear-bandit-algorithms-with-sublinear-time | 2103.02729 | null | https://arxiv.org/abs/2103.02729v2 | https://arxiv.org/pdf/2103.02729v2.pdf | Linear Bandit Algorithms with Sublinear Time Complexity | We propose two linear bandits algorithms with per-step complexity sublinear in the number of arms $K$. The algorithms are designed for applications where the arm set is extremely large and slowly changing. Our key realization is that choosing an arm reduces to a maximum inner product search (MIPS) problem, which can be solved approximately without breaking regret guarantees. Existing approximate MIPS solvers run in sublinear time. We extend those solvers and present theoretical guarantees for online learning problems, where adaptivity (i.e., a later step depends on the feedback in previous steps) becomes a unique challenge. We then explicitly characterize the tradeoff between the per-step complexity and regret. For sufficiently large $K$, our algorithms have sublinear per-step complexity and $\tilde O(\sqrt{T})$ regret. Empirically, we evaluate our proposed algorithms in a synthetic environment and a real-world online movie recommendation problem. Our proposed algorithms can deliver a more than 72 times speedup compared to the linear time baselines while retaining similar regret. | ['Sujay Sanghavi', 'Inderjit S. Dhillon', 'Eric Price', 'Sanjay Shakkottai', 'Tongzheng Ren', 'Shuo Yang'] | 2021-03-03 | null | null | null | null | ['movie-recommendation'] | ['miscellaneous'] | [-2.44626272e-02 2.02452078e-01 -8.06927979e-01 -3.84055793e-01
-1.33035541e+00 -1.09998477e+00 -1.45551041e-01 1.63789943e-01
-5.98277032e-01 7.68390238e-01 9.82701313e-03 -8.33204329e-01
-4.94923770e-01 -7.08163679e-01 -1.24582422e+00 -6.73584282e-01
-3.42300534e-01 7.40565956e-01 7.76643753e-02 -9.29493383e-02
2.76976079e-01 2.45551065e-01 -8.49866927e-01 2.94463158e-01
6.50167882e-01 1.44583642e+00 -1.23789638e-01 8.18294764e-01
-1.45210564e-01 5.38402438e-01 -3.97849903e-02 -6.19872749e-01
1.07552528e+00 -2.93152004e-01 -9.22987700e-01 1.87826529e-01
4.99451905e-01 -5.16528249e-01 -5.23017466e-01 9.90683556e-01
2.65956610e-01 2.55222350e-01 9.54812765e-02 -7.94881463e-01
-2.73605198e-01 1.19203472e+00 -9.24239397e-01 2.21511021e-01
8.74489769e-02 -6.44154474e-02 1.53670597e+00 -2.45496646e-01
2.24975556e-01 1.03201103e+00 3.60830396e-01 2.72152632e-01
-1.32848012e+00 -6.43977225e-01 7.84862936e-01 1.04038961e-01
-9.53075588e-01 -4.03618246e-01 4.59152937e-01 -1.51588932e-01
7.06775725e-01 7.33144403e-01 6.01497352e-01 3.23260188e-01
-5.04717171e-01 1.05007160e+00 9.54754829e-01 -3.58012229e-01
5.32749116e-01 2.24020943e-01 4.47657585e-01 5.62206209e-01
3.25848162e-01 2.17306055e-03 -5.12198269e-01 -2.68521696e-01
6.57602072e-01 2.00713649e-01 -2.99197644e-01 -2.08927080e-01
-7.86047399e-01 9.38982964e-01 5.32125711e-01 -2.54142016e-01
-3.01032037e-01 5.24965346e-01 2.58609354e-01 8.25930357e-01
2.51732528e-01 2.87110418e-01 -7.99123168e-01 -2.93289602e-01
-9.25858259e-01 3.85573655e-01 1.03778720e+00 1.12779427e+00
6.64655864e-01 -3.33181292e-01 -2.27355570e-01 6.19322538e-01
-8.46570805e-02 4.54814583e-01 1.45087287e-01 -1.03362989e+00
1.11072707e+00 2.83725947e-01 8.49082470e-01 -8.68939400e-01
-2.38108695e-01 -7.65632391e-01 -5.65884829e-01 -1.39681503e-01
7.40882635e-01 -3.79901648e-01 -5.95286608e-01 1.66533935e+00
6.99409127e-01 -2.56409496e-01 -3.46829206e-01 9.97078538e-01
7.35082291e-03 7.24194109e-01 -6.55251741e-01 -8.03169012e-01
1.06490123e+00 -1.47389150e+00 -4.92770612e-01 -5.74789405e-01
7.48628020e-01 -5.31880677e-01 8.37215483e-01 5.21674991e-01
-1.50503528e+00 1.98910400e-01 -1.03919303e+00 2.72869587e-01
4.17028219e-02 -1.95705488e-01 1.18019438e+00 9.59476888e-01
-6.59953654e-01 5.90454161e-01 -8.69308770e-01 1.81699276e-01
3.70429933e-01 7.24520624e-01 -8.54657143e-02 -4.31556115e-03
-5.21787286e-01 1.96427509e-01 2.90710717e-01 2.10077479e-01
-6.42823994e-01 -8.31742764e-01 -2.40989968e-01 1.29107311e-01
9.04390991e-01 -5.61248660e-01 1.53268456e+00 -1.15049303e+00
-1.67391264e+00 4.55170006e-01 -2.19170228e-01 -6.56128705e-01
7.31300712e-01 -3.60885054e-01 2.18209103e-01 -1.74315020e-01
-4.25111860e-01 -2.91771472e-01 6.75789237e-01 -7.48246968e-01
-1.02428520e+00 -5.47618091e-01 8.87548029e-01 2.19345942e-01
-6.54643059e-01 -1.94801524e-01 -5.56246698e-01 -3.62786055e-01
3.33658814e-01 -1.12050128e+00 -6.38166189e-01 6.47518225e-03
-1.24153189e-01 4.02999222e-02 -2.83511598e-02 -3.74775261e-01
1.48145640e+00 -1.96207917e+00 1.30767196e-01 4.09633815e-01
-1.06419012e-01 -8.63871537e-03 -1.64612308e-01 6.45763338e-01
1.52205974e-01 5.35348989e-02 1.91941872e-01 -1.84681304e-02
1.50727808e-01 1.52588785e-01 -4.92902100e-01 5.89683592e-01
-8.77290666e-01 6.80765808e-01 -5.93991816e-01 3.74130532e-02
-3.32021981e-01 -5.19580722e-01 -9.62452233e-01 1.76078439e-01
-4.38555151e-01 -3.15843970e-02 -7.04666734e-01 3.99808288e-01
8.27529490e-01 -6.91115618e-01 3.82075429e-01 -2.19452865e-02
-3.13241908e-04 3.47238064e-01 -1.61620891e+00 1.63744724e+00
-7.56618798e-01 1.42279372e-01 4.21771944e-01 -1.32543862e+00
3.78573328e-01 1.12228328e-02 3.52051705e-01 -6.35562301e-01
1.90348953e-01 2.72456467e-01 -2.21979052e-01 -9.51039121e-02
1.84303924e-01 -2.34937053e-02 5.25361896e-02 8.76691163e-01
-3.60105783e-01 3.94967467e-01 3.41480434e-01 1.90660596e-01
1.38176906e+00 -3.66816580e-01 1.83819488e-01 -1.45614460e-01
2.27470413e-01 -1.89357830e-04 6.99057519e-01 1.32547927e+00
1.52801022e-01 1.28798246e-01 7.40218997e-01 -8.56377065e-01
-9.21730697e-01 -7.80595601e-01 2.48067483e-01 1.67595577e+00
2.27136314e-01 -2.91348368e-01 -5.49561501e-01 -5.66289663e-01
2.87014902e-01 3.83333951e-01 -8.48216414e-01 3.20164889e-01
-5.53139567e-01 -8.71503055e-01 -2.09928945e-01 4.37369227e-01
2.94400722e-01 -2.23893493e-01 -4.05765653e-01 5.01768172e-01
9.47822928e-02 -9.99273002e-01 -8.15407634e-01 2.31014177e-01
-1.20381773e+00 -8.79651070e-01 -4.77838218e-01 -5.53949773e-01
7.71485925e-01 5.70148766e-01 9.02276635e-01 -9.78873521e-02
-6.42578825e-02 1.52021676e-01 -4.41148698e-01 -2.92717218e-01
2.64637351e-01 1.47426233e-01 -6.99497983e-02 7.50581920e-02
1.67762171e-02 -7.18221426e-01 -1.17965686e+00 2.41033003e-01
-7.10077047e-01 1.90184623e-01 6.95221007e-01 9.42193985e-01
7.62676120e-01 -1.02000237e-01 3.36292684e-01 -1.41798043e+00
4.27920133e-01 -3.53816748e-01 -1.16691816e+00 4.60862428e-01
-8.19493353e-01 2.97849625e-01 8.32131088e-01 -5.19709647e-01
-9.17438984e-01 2.61326998e-01 1.91068232e-01 -2.05219686e-02
6.07971787e-01 5.11590898e-01 4.15311575e-01 -1.91111028e-01
6.71352684e-01 3.01282495e-01 -3.01451206e-01 -4.43635881e-01
5.38697422e-01 6.01133525e-01 7.49091282e-02 -8.82885396e-01
8.30882967e-01 6.13879621e-01 4.03423347e-02 -1.07262783e-01
-1.38797593e+00 -5.57948112e-01 2.45530810e-02 2.63224155e-01
-2.14221507e-01 -8.61351132e-01 -1.46052790e+00 -1.05281726e-01
-7.37763464e-01 -6.05389774e-01 -2.07159460e-01 3.42717171e-01
-7.46880352e-01 2.65956223e-01 -5.93232334e-01 -1.17649305e+00
-7.40783393e-01 -8.15429688e-01 4.62618351e-01 2.96590298e-01
3.39012176e-01 -5.85514486e-01 9.91491675e-02 6.55168056e-01
4.01227504e-01 -4.68945131e-02 6.97407961e-01 -4.44362402e-01
-8.12795460e-01 -4.21239346e-01 -2.23647282e-01 -7.33599365e-02
-1.82512831e-02 -7.50806272e-01 -4.35541749e-01 -8.29294086e-01
-2.25879878e-01 -3.94872397e-01 7.25492835e-01 3.67998868e-01
1.44368327e+00 -1.17989910e+00 -1.44080937e-01 8.00234497e-01
1.59603727e+00 3.02413762e-01 1.93761796e-01 4.77968097e-01
1.15375355e-01 9.50476378e-02 9.56160367e-01 1.04860055e+00
-6.24749362e-02 6.02278471e-01 5.19055784e-01 2.31172532e-01
5.73644876e-01 -1.21840507e-01 1.72157094e-01 5.89209080e-01
-1.75905526e-01 -1.10936284e-01 -4.53450471e-01 6.97128654e-01
-2.48931623e+00 -7.12766826e-01 2.98102915e-01 2.68608022e+00
1.22907865e+00 3.46151501e-01 4.68691379e-01 -1.78351283e-01
3.68596822e-01 9.97836590e-02 -9.89346504e-01 -6.62819386e-01
2.47100040e-01 3.24562937e-01 1.27950263e+00 5.90742350e-01
-9.21601295e-01 8.64657938e-01 6.09676027e+00 1.00498939e+00
-8.64401281e-01 2.04112574e-01 8.33933890e-01 -1.01864815e+00
-2.19557643e-01 1.84860900e-01 -6.63503230e-01 4.22639638e-01
8.62579226e-01 -4.19008940e-01 1.17082107e+00 1.18440711e+00
4.98110540e-02 5.82901575e-02 -1.34177530e+00 1.17179286e+00
-4.00388241e-01 -1.61522126e+00 -2.98718065e-01 8.23939294e-02
1.04914856e+00 -2.49712355e-02 2.52773494e-01 1.58793375e-01
7.29738057e-01 -8.25342894e-01 4.99619603e-01 -9.84844342e-02
5.49491405e-01 -9.80749726e-01 5.28369486e-01 6.26714766e-01
-1.04655075e+00 -5.97361684e-01 -6.23690724e-01 -2.45844871e-01
1.38042524e-01 5.38002849e-01 -5.05617201e-01 4.23859626e-01
8.30266356e-01 1.49444729e-01 1.68795720e-01 1.17337430e+00
4.79015661e-03 6.70917928e-01 -8.78029287e-01 -3.16833526e-01
4.94934082e-01 -4.68440056e-01 2.37272024e-01 1.05960584e+00
2.09479272e-01 5.41253805e-01 1.96399406e-01 2.73754358e-01
-2.36358166e-01 3.32304627e-01 8.85290578e-02 -2.33206779e-01
5.72856367e-01 1.12227583e+00 -4.28476006e-01 -9.09578651e-02
-4.62019384e-01 1.01480949e+00 7.96483576e-01 3.56114388e-01
-9.31727171e-01 -3.77963424e-01 7.30361819e-01 4.86015156e-02
8.12340021e-01 -1.25365764e-01 -3.38437438e-01 -9.41938579e-01
4.57578540e-01 -7.78240383e-01 7.24175274e-01 1.25749215e-01
-1.06672859e+00 2.10342750e-01 -4.20300633e-01 -8.75823677e-01
-3.93911526e-02 -5.56449175e-01 -2.87213445e-01 4.51357633e-01
-1.27127731e+00 -9.03219640e-01 2.26489324e-02 4.94815946e-01
5.67953110e-01 1.35678574e-01 6.11549616e-01 2.87396640e-01
-6.57667637e-01 1.34523869e+00 8.45095575e-01 -2.36481458e-01
3.11395884e-01 -1.15032744e+00 1.17789410e-01 6.78350866e-01
6.69827610e-02 6.43761098e-01 8.35491717e-01 -4.36416976e-02
-2.13690639e+00 -7.75913358e-01 1.42691284e-01 1.13577627e-01
1.03705668e+00 -3.72693211e-01 -1.76521450e-01 9.93488371e-01
-2.12760210e-01 1.27664730e-01 7.79726446e-01 6.41940773e-01
-6.18821204e-01 -6.84723496e-01 -1.11752796e+00 5.38445592e-01
1.44885099e+00 -1.83646590e-01 1.10390089e-01 8.93042684e-01
8.49635184e-01 -8.16331685e-01 -8.53416562e-01 1.01784900e-01
8.99220705e-01 -8.30981672e-01 8.50542247e-01 -1.14262295e+00
9.73175466e-02 1.18913300e-01 -1.58314213e-01 -1.00059915e+00
-4.95868325e-01 -1.30837202e+00 -6.33578300e-01 6.32602155e-01
8.34948123e-01 -7.03465104e-01 1.24696767e+00 1.06747460e+00
3.33693534e-01 -1.05172670e+00 -1.07581365e+00 -9.25146043e-01
5.40167801e-02 -4.00012732e-01 6.30151033e-01 5.17605186e-01
4.13553089e-01 3.28930281e-02 -7.29870617e-01 2.48872682e-01
5.85753024e-01 8.99628401e-01 9.73241627e-01 -6.10884488e-01
-1.31696439e+00 -3.52115512e-01 6.52106330e-02 -1.88769901e+00
-3.80866021e-01 -6.20885551e-01 -2.00707003e-01 -1.16721141e+00
6.09725535e-01 -7.54006088e-01 -6.32297993e-01 5.07053792e-01
7.07264990e-02 -1.07836574e-01 2.05387741e-01 -5.31243086e-02
-9.75006402e-01 1.03469186e-01 9.93448496e-01 -2.34172434e-01
-4.81421173e-01 4.49499398e-01 -9.23529625e-01 4.72768933e-01
8.06707323e-01 -5.09902000e-01 -3.52939487e-01 -7.20930278e-01
8.39584768e-01 5.22100866e-01 -4.66211826e-01 -4.57947701e-01
4.86764669e-01 -7.75624990e-01 -2.62142360e-01 -3.70905876e-01
2.55230308e-01 -1.01735234e+00 9.12958011e-02 5.71389318e-01
-5.90004265e-01 -3.37983757e-01 8.27226639e-02 9.60429251e-01
3.18972200e-01 -3.96254927e-01 6.60020888e-01 -7.33838230e-02
8.27226937e-02 7.44396508e-01 9.64389816e-02 -9.98031534e-03
1.21265984e+00 -5.40162101e-02 -3.43494654e-01 -6.03149831e-01
-6.36740327e-01 5.06317496e-01 2.81305984e-02 -1.22820579e-01
1.59604535e-01 -1.11093700e+00 -4.60239798e-01 -2.79939026e-01
-4.95405942e-02 -7.68448636e-02 4.00294125e-01 7.93978095e-01
-5.28685093e-01 7.15319991e-01 3.42102259e-01 -1.15965441e-01
-1.19226313e+00 7.83388555e-01 1.36341035e-01 -5.95361233e-01
-3.15866053e-01 1.23511767e+00 1.61995925e-02 -1.11583173e-01
7.23247945e-01 -9.14026722e-02 4.84869570e-01 -2.72450626e-01
8.45740974e-01 5.01280725e-01 9.95323621e-03 3.20038766e-01
-7.18792900e-02 2.72707045e-01 -7.81528592e-01 -9.82128233e-02
1.62860310e+00 -1.47463739e-01 -1.00046054e-01 1.28994644e-01
1.08231580e+00 4.38602231e-02 -1.15778303e+00 -7.54553437e-01
-2.12564841e-01 -8.81325722e-01 2.11416587e-01 -7.37382710e-01
-1.25480938e+00 2.89799154e-01 6.69735909e-01 2.75985479e-01
1.25753903e+00 -8.75168666e-03 1.01390076e+00 9.80644464e-01
9.13783967e-01 -1.41398954e+00 -2.30216339e-01 1.87668815e-01
6.54176235e-01 -1.25249684e+00 3.19082946e-01 -5.03824830e-01
-2.26730466e-01 1.03150952e+00 3.85583967e-01 -2.88027018e-01
5.95053673e-01 9.94133279e-02 -5.14577985e-01 2.02268228e-01
-1.01818824e+00 8.68075639e-02 -4.08853255e-02 -1.74162358e-01
1.33745119e-01 4.07822162e-01 -6.96001232e-01 8.81391108e-01
-2.80786127e-01 -1.99193810e-03 2.95062333e-01 8.26615572e-01
-7.21215069e-01 -1.19586229e+00 -1.92334637e-01 7.63808727e-01
-7.93670058e-01 9.11289752e-02 -6.07472435e-02 2.63437301e-01
-6.30435109e-01 9.47471380e-01 2.90987734e-02 -1.95555896e-01
1.71667308e-01 -3.80398750e-01 7.73971558e-01 -2.28129536e-01
-5.01437306e-01 1.97490938e-02 1.52690634e-01 -9.32360947e-01
-6.17616959e-02 -3.18622202e-01 -9.73173738e-01 -6.78937137e-01
-4.39467162e-01 4.11717623e-01 5.93826652e-01 8.12016785e-01
2.91021526e-01 5.15045645e-03 1.20069361e+00 -2.28663906e-01
-1.29079187e+00 -5.78850806e-01 -7.04957843e-01 1.07017942e-01
3.32094580e-01 -2.05569044e-01 -4.40163642e-01 -3.99303764e-01] | [4.7486467361450195, 3.5106918811798096] |
a39d3ddc-6a38-4f5c-9f3d-965c6d1f87d3 | deep-reinforcement-learning-for-multi-class | 2205.12070 | null | https://arxiv.org/abs/2205.12070v1 | https://arxiv.org/pdf/2205.12070v1.pdf | Deep Reinforcement Learning for Multi-class Imbalanced Training | With the rapid growth of memory and computing power, datasets are becoming increasingly complex and imbalanced. This is especially severe in the context of clinical data, where there may be one rare event for many cases in the majority class. We introduce an imbalanced classification framework, based on reinforcement learning, for training extremely imbalanced data sets, and extend it for use in multi-class settings. We combine dueling and double deep Q-learning architectures, and formulate a custom reward function and episode-training procedure, specifically with the added capability of handling multi-class imbalanced training. Using real-world clinical case studies, we demonstrate that our proposed framework outperforms current state-of-the-art imbalanced learning methods, achieving more fair and balanced classification, while also significantly improving the prediction of minority classes. | ['David A. Clifton', 'Andrew A. S. Soltan', 'Alexander S. Lachapelle', "Odhran O'Donoghue", 'Rasheed el-Bouri', 'Jenny Yang'] | 2022-05-24 | null | null | null | null | ['imbalanced-classification'] | ['miscellaneous'] | [ 8.46969783e-02 2.52711158e-02 -5.24158001e-01 -4.73458081e-01
-8.41375351e-01 -9.44815949e-02 -1.17962919e-01 7.97508717e-01
-4.76822555e-01 1.07079232e+00 -9.49860588e-02 -5.14501333e-01
-3.19938034e-01 -8.71098816e-01 -3.28167021e-01 -4.77046341e-01
-3.19691360e-01 1.00958848e+00 -2.18733966e-01 -5.00932150e-02
7.31216818e-02 2.98910707e-01 -1.30805159e+00 5.27536690e-01
1.11051810e+00 1.01087570e+00 -4.88406152e-01 5.19318104e-01
1.41159698e-01 9.40766811e-01 -8.36263478e-01 -5.35451114e-01
1.32366329e-01 -2.59300023e-01 -6.34417832e-01 -2.22304780e-02
-1.31921014e-02 -3.38023871e-01 -7.03311637e-02 4.72259343e-01
1.04958117e+00 -1.07650742e-01 3.18395317e-01 -1.25949037e+00
-4.40323874e-02 2.34048977e-01 -6.79860413e-01 5.57057858e-01
8.18012729e-02 2.78912395e-01 9.85534132e-01 -3.79071116e-01
4.75441784e-01 8.65600407e-01 7.76753783e-01 5.81279695e-01
-8.51491392e-01 -7.86720097e-01 4.45459001e-02 3.59080762e-01
-8.09181988e-01 -1.73288897e-01 6.18273973e-01 -3.05475682e-01
1.13032103e+00 1.20087035e-01 8.88674140e-01 9.71586168e-01
4.74044621e-01 9.63752091e-01 8.72851908e-01 -1.27886221e-01
5.71245730e-01 -3.85097027e-01 -2.63796505e-02 2.79299676e-01
4.88729864e-01 2.17340216e-01 -3.14584911e-01 -3.71232867e-01
3.53741497e-01 3.75783920e-01 1.83444366e-01 -3.98743629e-01
-9.94833589e-01 9.47632968e-01 3.44412655e-01 3.06152739e-02
-6.00098431e-01 -5.64780645e-02 8.61640155e-01 2.92807579e-01
7.63982952e-01 5.76931536e-01 -6.02208734e-01 -5.28355896e-01
-9.55340743e-01 5.83302081e-01 7.49333680e-01 3.26517880e-01
2.03262419e-01 -1.63118318e-01 -2.03111783e-01 8.01453292e-01
-2.06462905e-01 1.93716571e-01 8.64809573e-01 -6.46523178e-01
6.31537259e-01 9.32407916e-01 -6.97016437e-03 -8.09632123e-01
-9.18061316e-01 -1.02683187e+00 -1.07267284e+00 1.31563321e-01
1.41161159e-01 -1.15499638e-01 -8.85710895e-01 1.55980802e+00
5.17502010e-01 1.45371348e-01 2.26509288e-01 7.33027995e-01
5.33808291e-01 1.88686058e-01 2.83376634e-01 -4.35162634e-01
1.24374902e+00 -6.91363156e-01 -8.56190979e-01 -3.24555427e-01
8.27688515e-01 -1.91537961e-01 8.41712832e-01 6.41889095e-01
-1.13077879e+00 -1.45686775e-01 -1.00160575e+00 2.74492353e-01
-2.73710191e-01 -2.19066903e-01 8.99012089e-01 7.16040790e-01
-4.41120476e-01 6.18532062e-01 -9.42937791e-01 1.80817172e-01
1.07530725e+00 5.00566721e-01 -2.07762048e-01 -2.81694680e-01
-1.35680246e+00 9.02095675e-01 2.80482173e-01 5.07628955e-02
-7.37693787e-01 -1.11011147e+00 -6.89672410e-01 1.92620084e-02
2.21934870e-01 -8.34070683e-01 1.18900049e+00 -8.91989827e-01
-1.13957953e+00 9.33919191e-01 4.19161141e-01 -5.41279852e-01
8.21474552e-01 -1.92709893e-01 -4.98568416e-01 -2.77911760e-02
-3.63800973e-02 2.53931552e-01 5.97999871e-01 -6.66911006e-01
-4.61346000e-01 -8.86604786e-01 -4.38526161e-02 4.89060819e-01
-2.42065296e-01 -2.43940040e-01 1.91201463e-01 -5.88112354e-01
-2.78980583e-01 -5.65312207e-01 -5.94355404e-01 -1.04799658e-01
-3.42205107e-01 -1.21368445e-01 5.59644520e-01 -5.50332546e-01
1.28169250e+00 -2.02676821e+00 -5.90256155e-02 6.32627159e-02
5.85304379e-01 5.90087414e-01 2.94894408e-02 2.19635338e-01
-3.16186070e-01 1.33655192e-02 -3.35834712e-01 -2.90739328e-01
-2.69287199e-01 3.21493089e-01 1.05218723e-01 3.44670087e-01
6.92846239e-01 8.21889222e-01 -1.18318820e+00 -3.93433154e-01
2.40786634e-02 1.67290360e-01 -1.00873637e+00 5.08001387e-01
-9.35754925e-02 5.37501037e-01 -3.29396516e-01 7.43054807e-01
6.34639382e-01 -6.21239364e-01 4.22374964e-01 1.87676206e-01
4.44456130e-01 1.81076825e-01 -1.15597475e+00 1.30381751e+00
-5.56298614e-01 1.03957608e-01 -3.97486001e-01 -1.43497622e+00
8.45398605e-01 1.26920059e-01 7.07977116e-01 -1.25101483e+00
4.83506825e-03 3.93753201e-01 4.02600616e-01 -6.65642381e-01
2.53385097e-01 -5.80631614e-01 -1.22069702e-01 3.32639843e-01
-1.36145577e-01 -7.98737109e-02 3.66069794e-01 -8.92716721e-02
1.35477734e+00 -2.81033397e-01 5.76912880e-01 7.68331364e-02
1.81513786e-01 -1.44570366e-01 1.04647291e+00 4.54750389e-01
-5.00923693e-01 4.51785535e-01 1.01210523e+00 -7.82740772e-01
-9.54852700e-01 -7.93293893e-01 -3.34041178e-01 8.42615128e-01
-8.55737180e-02 -6.79295138e-02 -4.69675124e-01 -7.72270679e-01
2.82971472e-01 3.00255239e-01 -6.91765666e-01 -5.35222292e-01
-5.90660453e-01 -1.61268485e+00 4.61925417e-01 5.14143646e-01
2.00889632e-01 -1.21596384e+00 -1.00735164e+00 6.34844005e-01
-5.51131852e-02 -8.70016754e-01 3.40754800e-02 5.32805145e-01
-1.06662798e+00 -1.35706508e+00 -6.35401428e-01 -5.21066487e-01
3.64742011e-01 -3.38665485e-01 1.71857798e+00 3.55530828e-01
-7.58939266e-01 -1.79777727e-01 -3.97815198e-01 -6.99519396e-01
-3.75940740e-01 2.43249074e-01 -7.92655274e-02 -1.34042829e-01
4.30743098e-01 -4.70799685e-01 -1.08163524e+00 8.54065716e-02
-1.07494462e+00 -3.86166945e-02 5.32864630e-01 1.26738262e+00
6.59276605e-01 2.66467147e-02 1.38517022e+00 -1.27444780e+00
4.57096338e-01 -8.65733564e-01 -2.03882411e-01 7.20384996e-03
-7.98322916e-01 -3.34698945e-01 7.71035612e-01 -1.28687263e-01
-6.25330210e-01 -3.62917095e-01 -4.42248583e-01 -1.21370539e-01
6.18903264e-02 4.99181777e-01 2.73484234e-02 2.85828859e-01
6.74536824e-01 -1.64284766e-01 3.36352348e-01 -1.48500085e-01
-1.06403716e-01 7.80126154e-01 -2.27578450e-02 -2.37074018e-01
1.25218138e-01 3.69169176e-01 -1.09452657e-01 -2.00722441e-01
-1.04610062e+00 -4.64927286e-01 -2.77072012e-01 6.13406580e-03
3.41184527e-01 -1.07838917e+00 -8.53870928e-01 7.04936862e-01
-5.40030479e-01 -4.87569749e-01 -6.34410441e-01 3.01675975e-01
-7.13317335e-01 -5.41318348e-03 -6.64303839e-01 -4.54000175e-01
-5.99040627e-01 -1.24135160e+00 1.06144881e+00 1.72367513e-01
-1.72843575e-01 -9.33038056e-01 6.20501935e-01 5.06449878e-01
4.40418184e-01 5.36395907e-01 1.27598059e+00 -9.92753088e-01
-1.00613274e-01 -2.77231485e-01 -8.90560746e-02 1.91383392e-01
1.42514035e-01 -2.99250275e-01 -7.28957474e-01 -5.68689406e-01
-3.43795359e-01 -7.33401120e-01 6.24673843e-01 4.17288929e-01
1.51481652e+00 -1.06630117e-01 -2.90756017e-01 5.81347883e-01
1.46786332e+00 2.72521794e-01 6.31544411e-01 3.11319292e-01
3.45674306e-01 3.53898138e-01 7.31992245e-01 9.16966379e-01
4.99851733e-01 6.14457309e-01 7.72554815e-01 -4.32722807e-01
2.16310456e-01 1.50549352e-01 -3.16685438e-01 7.06987023e-01
2.45615572e-01 -2.26842627e-01 -1.09347200e+00 5.72121978e-01
-1.79144812e+00 -8.14876437e-01 4.64425355e-01 2.12273026e+00
1.18560994e+00 2.83817917e-01 3.49823385e-01 5.32575130e-01
4.65148747e-01 2.10246220e-02 -1.11853909e+00 -5.51287174e-01
4.20418419e-02 4.84329879e-01 3.57031897e-02 8.89552087e-02
-1.08539796e+00 4.51720536e-01 6.83024216e+00 6.22100353e-01
-1.26707757e+00 3.54297087e-02 1.44408989e+00 -3.87420148e-01
-5.65045550e-02 -6.28547549e-01 -4.63801146e-01 5.95362425e-01
1.08599424e+00 -5.40651754e-02 2.26582717e-02 5.44927418e-01
2.19319880e-01 -2.67634485e-02 -1.02657068e+00 9.28029180e-01
-2.18704700e-01 -1.22038066e+00 -1.03460923e-01 -2.28079110e-02
7.92230606e-01 6.60335273e-03 1.37035564e-01 5.86781144e-01
2.71673292e-01 -1.23607969e+00 4.29312699e-02 1.35253698e-01
7.99662292e-01 -1.03266931e+00 1.11548376e+00 3.18197489e-01
-2.61185318e-01 -4.73923385e-01 -1.88862294e-01 -2.37699762e-01
-1.51092365e-01 1.37387812e+00 -8.59872699e-01 6.38573050e-01
6.61953211e-01 9.08876419e-01 -3.76983553e-01 1.26785386e+00
3.02907258e-01 5.14781296e-01 -1.13251165e-01 2.34837495e-02
-3.90333533e-02 1.23368524e-01 4.85026352e-02 9.18814957e-01
3.87532115e-02 5.99481612e-02 2.05959857e-01 1.97114065e-01
-3.87098700e-01 3.21263343e-01 -3.18266690e-01 1.69762015e-01
3.12505066e-01 1.19969451e+00 -5.97312450e-01 -4.39819723e-01
-1.48923844e-01 6.23981595e-01 5.17434418e-01 -8.49614739e-02
-8.09946775e-01 -3.26377243e-01 7.23564029e-01 -1.21258982e-01
1.36170924e-01 3.82503241e-01 -5.38447976e-01 -1.08237672e+00
1.26537651e-01 -1.44952333e+00 8.31865847e-01 -1.19005360e-01
-1.50003314e+00 5.52178741e-01 -6.61999047e-01 -1.46525311e+00
-4.57604438e-01 -4.95682418e-01 -3.12853783e-01 4.10977244e-01
-2.02627373e+00 -7.59955704e-01 -2.53985435e-01 3.00721198e-01
2.36337543e-01 -1.12802461e-01 9.70475495e-01 7.63736844e-01
-7.66046405e-01 7.26288795e-01 3.31338614e-01 5.05018532e-02
6.62588716e-01 -1.34361649e+00 8.83514360e-02 1.03560321e-01
-4.25871074e-01 -1.52767310e-02 5.35068631e-01 -4.41497892e-01
-1.10683858e+00 -1.15002155e+00 6.10843718e-01 -2.79056937e-01
3.02894443e-01 -2.78804928e-01 -8.79687667e-01 3.27714533e-01
-3.33228260e-01 3.72105688e-01 1.23514426e+00 2.60646403e-01
-2.42117181e-01 -3.61682475e-01 -1.47910774e+00 1.45992965e-01
7.47031212e-01 -1.38062164e-01 -4.92501050e-01 7.07540035e-01
4.09553707e-01 -9.24804151e-01 -1.12795496e+00 8.21736574e-01
4.99834836e-01 -1.03752422e+00 9.09757078e-01 -1.02210915e+00
8.40178192e-01 1.01214707e-01 2.46122882e-01 -1.65160751e+00
-1.07372075e-01 -2.66895026e-01 -3.14835250e-01 4.80155766e-01
2.73913622e-01 -5.59428036e-01 1.19632471e+00 2.29209930e-01
-1.55796986e-02 -1.58499503e+00 -1.23403656e+00 -3.81550580e-01
3.08095157e-01 -1.15738750e-01 7.35925376e-01 9.33649540e-01
1.70350671e-01 7.02288002e-02 -3.87883067e-01 -2.43181020e-01
4.05969590e-01 3.91923666e-01 3.45810443e-01 -1.19954705e+00
-5.43569982e-01 -2.90300578e-01 -7.14469552e-01 -2.90312082e-01
-8.21838751e-02 -6.87934101e-01 -1.68911010e-01 -1.25085294e+00
3.99259627e-01 -7.60884345e-01 -6.74043715e-01 6.34362936e-01
-7.20569193e-01 5.74994564e-01 -8.01181570e-02 -5.75881964e-03
-9.00097668e-01 7.02309132e-01 1.30997097e+00 -1.78452224e-01
-1.71179548e-01 3.77398640e-01 -7.63087213e-01 3.12594563e-01
7.46485949e-01 -5.14545321e-01 -2.48176858e-01 -1.44531116e-01
5.56837201e-01 3.83141518e-01 -1.12817017e-03 -1.14452124e+00
-1.46791816e-01 4.40146960e-03 6.44370735e-01 -4.12829161e-01
9.21560600e-02 -7.09009290e-01 -1.48665756e-01 1.02826691e+00
-4.34153140e-01 2.32786208e-01 3.17448825e-01 5.06358027e-01
-4.43087727e-01 1.71572328e-01 9.06512976e-01 2.90786917e-03
-3.14463198e-01 5.94791472e-01 -1.24748580e-01 5.63860774e-01
1.28092206e+00 -4.80369180e-02 -4.48612928e-01 -1.88493475e-01
-8.48308682e-01 6.26661599e-01 1.17262438e-01 4.75660056e-01
5.90258956e-01 -1.18346179e+00 -8.39905918e-01 3.37094843e-01
2.87887424e-01 3.20011556e-01 6.93952024e-01 6.91042423e-01
-5.57756662e-01 3.07285525e-02 -5.82194567e-01 -5.31753719e-01
-9.52331722e-01 4.83528525e-01 5.50854445e-01 -1.14670813e+00
-3.66621941e-01 3.83143544e-01 -1.26994494e-02 -7.72345185e-01
9.62323099e-02 -1.00437284e-01 -1.38819218e-01 3.02503526e-01
7.05665231e-01 4.23458487e-01 8.19864333e-01 6.66567236e-02
-5.03532887e-01 2.04570398e-01 -3.44406337e-01 6.29108369e-01
1.53823435e+00 3.88948798e-01 1.83354784e-02 3.83354843e-01
1.08523881e+00 -4.83750045e-01 -1.13335216e+00 3.57735604e-02
-6.13245480e-02 -2.50531316e-01 -1.28143355e-01 -1.22321820e+00
-1.24745655e+00 1.02866840e+00 9.54825222e-01 1.02550201e-01
1.38217616e+00 -4.51199323e-01 7.77310193e-01 1.74480125e-01
2.57644832e-01 -1.03395855e+00 3.32033217e-01 3.34489137e-01
3.86495709e-01 -1.50366855e+00 1.27890125e-01 -7.09263012e-02
-5.46386838e-01 8.74068379e-01 7.30232120e-01 -2.12315425e-01
6.34997249e-01 3.27710062e-01 2.41414130e-01 -2.61745840e-01
-1.02990222e+00 3.00286680e-01 -3.02135557e-01 5.06674945e-01
3.93237114e-01 2.69597054e-01 -2.93455660e-01 5.38132071e-01
9.51291397e-02 3.95970583e-01 4.58987385e-01 9.41585600e-01
-1.93722434e-02 -1.16289449e+00 -8.98583326e-03 1.12069011e+00
-1.02401853e+00 -8.42440277e-02 2.78274775e-01 6.20600879e-01
1.42001405e-01 5.56272089e-01 4.97444898e-01 1.01675941e-02
4.51785803e-01 -1.14491403e-01 2.52736300e-01 -5.01444876e-01
-8.25779498e-01 -3.84869367e-01 -2.29105297e-02 -7.18648374e-01
-3.62934828e-01 -3.74589205e-01 -1.17431831e+00 -1.87754929e-01
-2.33853936e-01 9.20569822e-02 5.23865879e-01 8.22349429e-01
7.85058379e-01 1.02762222e+00 9.08258498e-01 -4.23645258e-01
-9.16172922e-01 -7.72882819e-01 -6.28507257e-01 7.21277237e-01
6.49096429e-01 -5.96289337e-01 -7.84500614e-02 -4.84072596e-01] | [15.068684577941895, -2.414060354232788] |
3ccf2841-36fc-4387-8a66-9740a77471ea | expression-conditional-gan-for-facial | 1905.05416 | null | https://arxiv.org/abs/1905.05416v1 | https://arxiv.org/pdf/1905.05416v1.pdf | Expression Conditional GAN for Facial Expression-to-Expression Translation | In this paper, we focus on the facial expression translation task and propose a novel Expression Conditional GAN (ECGAN) which can learn the mapping from one image domain to another one based on an additional expression attribute. The proposed ECGAN is a generic framework and is applicable to different expression generation tasks where specific facial expression can be easily controlled by the conditional attribute label. Besides, we introduce a novel face mask loss to reduce the influence of background changing. Moreover, we propose an entire framework for facial expression generation and recognition in the wild, which consists of two modules, i.e., generation and recognition. Finally, we evaluate our framework on several public face datasets in which the subjects have different races, illumination, occlusion, pose, color, content and background conditions. Even though these datasets are very diverse, both the qualitative and quantitative results demonstrate that our approach is able to generate facial expressions accurately and robustly. | ['Wei Wang', 'Xinya Chen', 'Songsong Wu', 'Yan Yan', 'Hao Tang', 'Nicu Sebe', 'Dan Xu'] | 2019-05-14 | null | null | null | null | ['facial-expression-generation'] | ['computer-vision'] | [ 4.87570435e-01 -2.18471542e-01 -3.03838570e-02 -8.37355733e-01
-3.99514377e-01 -4.76219922e-01 7.26607800e-01 -7.67789781e-01
-1.10795952e-01 8.48546088e-01 -1.75092742e-01 4.44627017e-01
3.87024641e-01 -7.23999023e-01 -5.73861301e-01 -1.01314104e+00
3.61677885e-01 -3.78953107e-02 -3.72973859e-01 -2.32755288e-01
-2.84314185e-01 6.67193472e-01 -1.75089335e+00 1.16615780e-01
6.27687812e-01 1.24654675e+00 -3.65182102e-01 2.59832561e-01
5.28176352e-02 8.69913042e-01 -7.10144639e-01 -6.12666190e-01
1.99070871e-01 -8.26359928e-01 -5.45133054e-01 5.53644419e-01
4.18881029e-01 -2.45452151e-01 1.51227787e-01 1.12193120e+00
6.30227149e-01 -1.08506680e-01 6.38233781e-01 -1.62597871e+00
-5.79422832e-01 -6.08016402e-02 -7.06849635e-01 -5.99667072e-01
3.87956709e-01 1.24309085e-01 5.71462631e-01 -8.66800487e-01
8.11309636e-01 1.35273135e+00 3.58309537e-01 9.88890588e-01
-1.22557342e+00 -1.03757346e+00 2.26734072e-01 2.57675676e-03
-1.42998707e+00 -6.58075809e-01 1.02665758e+00 -4.10704404e-01
8.97475332e-03 1.76463887e-01 4.45679039e-01 1.43234479e+00
-1.78558499e-01 6.48582637e-01 1.76255429e+00 -4.99608904e-01
2.14075774e-01 2.70847976e-01 -4.41790968e-01 6.53402090e-01
-3.23570132e-01 -7.14951977e-02 -6.17249906e-01 -2.17382625e-01
6.63899243e-01 -1.61979005e-01 -3.82065326e-01 -2.46547610e-01
-7.22368896e-01 7.27437913e-01 7.45395273e-02 7.85589293e-02
-4.31291282e-01 1.47203460e-01 3.25084865e-01 3.80444437e-01
5.49738526e-01 -5.34366742e-02 -2.28324875e-01 -2.91151386e-02
-6.86459720e-01 3.12599570e-01 6.70559824e-01 1.04509985e+00
9.84833360e-01 3.16204607e-01 -5.61703205e-01 1.14303243e+00
2.79529870e-01 6.89971864e-01 1.83269709e-01 -1.00986493e+00
-7.36116767e-02 3.25052053e-01 1.69685513e-01 -1.12536824e+00
-3.24661545e-02 1.23251742e-02 -8.51331234e-01 4.09259379e-01
-5.68103082e-02 -5.43845236e-01 -9.16846097e-01 2.42825460e+00
4.21502650e-01 2.70881683e-01 1.05625816e-01 9.60928619e-01
9.09206390e-01 5.54597914e-01 2.66689271e-01 -4.99550045e-01
1.43537939e+00 -8.72045457e-01 -1.20315075e+00 -2.71611921e-02
2.61048287e-01 -7.67864227e-01 1.04888833e+00 2.75724083e-01
-9.92054343e-01 -4.92798567e-01 -9.75861609e-01 1.48492843e-01
-9.67820436e-02 5.81836462e-01 7.58095860e-01 8.18544149e-01
-1.01905215e+00 8.70026797e-02 -4.99449760e-01 -3.07569951e-01
5.30769825e-01 3.52703750e-01 -6.81754827e-01 -9.03708711e-02
-1.03753197e+00 5.84152997e-01 -6.78956434e-02 2.43430018e-01
-8.65023911e-01 -3.27389151e-01 -8.67363811e-01 -9.79581103e-02
9.48813260e-02 -5.34398735e-01 1.12255323e+00 -1.84252095e+00
-2.19918895e+00 1.14052248e+00 -3.65369290e-01 1.17483929e-01
5.39726555e-01 -3.94456387e-02 -5.07628798e-01 -1.28365094e-02
-1.09767482e-01 7.93943167e-01 1.00183654e+00 -1.30686760e+00
-1.18773542e-01 -5.09496152e-01 -6.86929524e-02 9.34818313e-02
-3.69040102e-01 5.80121636e-01 -6.28911912e-01 -7.32214570e-01
-2.95324892e-01 -1.01397359e+00 3.10490802e-02 4.25565898e-01
-1.62713856e-01 -7.50376731e-02 1.07085764e+00 -5.07312179e-01
8.44376266e-01 -2.29431653e+00 2.02413514e-01 3.51729363e-01
-2.14514986e-01 7.99173042e-02 -2.37474322e-01 5.25049344e-02
-1.11455411e-01 3.94456498e-02 -4.58824813e-01 -5.09972274e-01
1.65272847e-01 3.45359713e-01 -1.21989146e-01 3.43608528e-01
5.55508673e-01 7.71180272e-01 -4.68814433e-01 -4.81238693e-01
-1.30152628e-01 8.55348527e-01 -3.96288991e-01 5.20799279e-01
-1.40847459e-01 8.31915677e-01 -4.99779314e-01 9.06427443e-01
1.06891596e+00 1.86954916e-01 2.42043555e-01 -9.60200801e-02
1.98787570e-01 -4.83837634e-01 -1.03106546e+00 1.62070262e+00
-5.17399848e-01 4.94873136e-01 5.25090933e-01 -7.41918325e-01
1.36550450e+00 5.04258454e-01 4.75705773e-01 -5.98345160e-01
4.71769959e-01 1.63760945e-01 -3.32036942e-01 -4.71944571e-01
1.47155404e-01 -4.66378659e-01 6.28838614e-02 3.89249146e-01
3.15690905e-01 -6.76343665e-02 1.26311839e-01 -3.09023201e-01
6.21798158e-01 3.07006925e-01 1.44579694e-01 -1.59379616e-01
9.13341284e-01 -6.24719560e-01 9.16176677e-01 1.68198377e-01
-2.70374686e-01 5.23146927e-01 7.05052435e-01 -2.64284194e-01
-6.64104819e-01 -8.57756913e-01 -2.32473373e-01 1.15970218e+00
2.42164936e-02 -1.85969353e-01 -9.78211224e-01 -7.31714904e-01
-2.97799140e-01 2.32870802e-01 -8.06384861e-01 -7.47143626e-02
-2.72061586e-01 -8.87637258e-01 7.19830394e-01 4.62550700e-01
8.13747346e-01 -1.22225702e+00 -2.18140855e-01 -1.66781396e-01
-2.94985950e-01 -1.47221971e+00 -3.92755061e-01 -4.07844216e-01
-3.49515170e-01 -7.30537474e-01 -8.28350902e-01 -8.00798237e-01
8.73045027e-01 -3.67694616e-01 1.03238964e+00 2.28586458e-02
-2.11820692e-01 4.92765844e-01 -4.24576968e-01 -5.24922729e-01
-3.68590385e-01 -2.45533153e-01 -2.07594007e-01 8.67253184e-01
2.74311125e-01 -7.01213419e-01 -6.16681516e-01 4.23116863e-01
-9.67096269e-01 4.98923548e-02 3.67584974e-01 7.79554725e-01
5.95552444e-01 -2.26330593e-01 5.85884571e-01 -9.06866312e-01
6.29218400e-01 -2.11230233e-01 -5.07555902e-01 2.52638370e-01
-3.36582184e-01 -2.87820846e-01 4.11512166e-01 -3.89465421e-01
-1.38698161e+00 3.16164464e-01 -2.78319538e-01 -3.30044925e-01
-3.50162089e-01 1.32777900e-01 -7.50466406e-01 -3.93570036e-01
3.77145529e-01 1.35378450e-01 2.05656633e-01 -3.13072830e-01
3.36291343e-01 5.81397772e-01 5.46492755e-01 -8.49481583e-01
6.64516151e-01 6.10414445e-01 2.10719362e-01 -5.80047846e-01
-5.51681459e-01 1.53216004e-01 -5.68013847e-01 -3.47062796e-01
8.67695749e-01 -1.07525253e+00 -7.02331066e-01 8.73168766e-01
-1.20663464e+00 -2.40893170e-01 4.40740623e-02 2.16104761e-01
-6.59195185e-01 1.91324372e-02 -5.92714906e-01 -8.38920772e-01
-3.67850065e-01 -1.18623722e+00 1.27630329e+00 2.89367229e-01
-2.22907271e-02 -7.45716929e-01 -8.11135173e-02 1.61220938e-01
5.08396208e-01 8.63856792e-01 6.88390374e-01 3.17623839e-02
-3.57615501e-01 -9.40058157e-02 -2.68889040e-01 7.17381001e-01
4.66263324e-01 3.07664156e-01 -1.15781987e+00 -2.78335571e-01
-1.18819714e-01 -6.54901266e-01 4.88329023e-01 -2.26917285e-02
1.26165605e+00 -4.11755174e-01 1.74194798e-02 8.25384498e-01
1.30878770e+00 1.75676569e-01 8.98803234e-01 1.08066080e-02
5.51635146e-01 9.25696075e-01 6.21884406e-01 5.65807402e-01
1.05946243e-01 1.06519437e+00 2.43279129e-01 -5.94802678e-01
-1.13698266e-01 -6.77115321e-02 6.14383280e-01 3.36704284e-01
-3.14229101e-01 -1.86438814e-01 -3.57755035e-01 1.46768451e-01
-1.66145837e+00 -9.04371679e-01 2.06738830e-01 1.75877607e+00
9.99002695e-01 -5.82038224e-01 6.89548403e-02 -1.97689056e-01
7.23618567e-01 9.83623639e-02 -4.45751518e-01 -5.54808080e-01
-5.16847253e-01 6.72504365e-01 -2.64052879e-02 2.49903575e-01
-1.04112506e+00 1.06731939e+00 6.62046146e+00 7.42468774e-01
-1.67321908e+00 1.00973770e-01 9.78105128e-01 -4.85936888e-02
-1.86998457e-01 -3.33978444e-01 -5.12699306e-01 4.62348700e-01
5.18413544e-01 -1.27133802e-01 2.76271820e-01 8.73898566e-01
2.11557001e-01 3.10882598e-01 -8.20939422e-01 1.16897869e+00
4.32398021e-01 -8.56825292e-01 1.84785038e-01 -2.22655818e-01
8.28837037e-01 -6.46585107e-01 3.16177279e-01 1.55533895e-01
7.72385150e-02 -1.36868739e+00 5.05840063e-01 4.70367849e-01
1.11557305e+00 -8.64155054e-01 7.81480014e-01 -9.39641818e-02
-8.39134216e-01 1.64959520e-01 7.17827007e-02 5.32271229e-02
4.97418009e-02 2.77315587e-01 -2.75934428e-01 5.57584107e-01
6.43238842e-01 5.73130429e-01 -3.98673981e-01 3.26699346e-01
-6.59426689e-01 4.73926574e-01 -7.25514069e-02 2.20239684e-01
-9.55678374e-02 -4.56981003e-01 1.57573894e-01 1.10119987e+00
5.14367580e-01 6.36739656e-02 3.89172845e-02 1.17418110e+00
-3.76255214e-01 4.52703565e-01 -5.34723103e-01 1.75458491e-01
1.30275622e-01 1.57770574e+00 -2.59743363e-01 -4.46693599e-02
-4.16368484e-01 1.30139327e+00 1.47340313e-01 6.60009325e-01
-1.03837121e+00 -2.12783322e-01 9.26342189e-01 -7.37191662e-02
2.05432892e-01 1.74981236e-01 1.31119147e-01 -1.12386596e+00
2.75400370e-01 -1.08761501e+00 -9.51504428e-03 -8.45208228e-01
-1.22800148e+00 8.09727728e-01 -1.46568969e-01 -9.46865678e-01
-2.59738684e-01 -7.65516281e-01 -6.37607396e-01 8.15873682e-01
-1.59948230e+00 -1.54095685e+00 -7.52335072e-01 8.98785293e-01
-1.14733521e-02 -2.96150595e-01 1.06651938e+00 5.87457418e-01
-7.14356482e-01 9.49928463e-01 -3.29959631e-01 3.00191164e-01
9.99142349e-01 -7.11828649e-01 -2.77923763e-01 6.22399330e-01
3.37744579e-02 4.43107933e-01 5.02694547e-01 -4.93516773e-02
-1.13135684e+00 -1.20422256e+00 4.93516266e-01 -7.26514608e-02
1.16635635e-01 -7.25213587e-01 -6.77620530e-01 6.96988821e-01
3.76037061e-01 2.45011538e-01 9.22528446e-01 -2.36615147e-02
-4.58202600e-01 -4.16473806e-01 -1.45711589e+00 5.60538113e-01
9.71045017e-01 -3.33536208e-01 2.70884126e-01 1.03600442e-01
2.35553384e-01 -3.33911866e-01 -8.82693172e-01 9.42053318e-01
6.53106511e-01 -1.06365788e+00 3.98905218e-01 -5.14860630e-01
4.87799674e-01 -4.32470649e-01 -3.15267116e-01 -1.07804251e+00
-6.54523671e-02 -8.93669486e-01 2.68480539e-01 1.80590451e+00
2.06278130e-01 -6.57990277e-01 8.19469869e-01 5.22214115e-01
3.14057797e-01 -7.54804969e-01 -8.02823186e-01 -6.17924869e-01
-1.01179972e-01 -1.49577662e-01 1.05157900e+00 1.03432727e+00
-4.96053666e-01 3.47034365e-01 -8.45740676e-01 -1.02011792e-01
2.72914499e-01 3.81388634e-01 1.16534543e+00 -9.25345361e-01
-1.82949558e-01 -2.06256777e-01 -6.12203956e-01 -7.21644759e-01
7.05205679e-01 -6.32279038e-01 9.57571715e-02 -8.69126856e-01
3.29081863e-01 -3.36159021e-01 -3.32113296e-01 7.22234130e-01
-2.32651699e-02 6.39496446e-01 1.82807431e-01 -1.30368173e-01
-2.65376389e-01 9.47207749e-01 1.32509243e+00 -1.56229362e-01
6.36112243e-02 -2.40490716e-02 -7.75536001e-01 7.11234272e-01
7.49290884e-01 -2.60759890e-01 -5.00139832e-01 -2.92830974e-01
-1.76776513e-01 -1.35323107e-01 4.03686285e-01 -7.17271745e-01
-2.04551101e-01 -2.88695753e-01 4.14673507e-01 3.17343175e-02
6.12674117e-01 -7.86410511e-01 2.38707647e-01 -9.08923510e-04
-2.30912492e-01 -4.41233963e-02 2.89653510e-01 1.80527851e-01
-5.53107619e-01 1.78296000e-01 1.12428069e+00 1.31476998e-01
-6.73172593e-01 5.89935064e-01 -1.61713824e-01 -1.69579983e-01
1.21667635e+00 -9.97900963e-02 -8.86580423e-02 -7.23846674e-01
-6.79830670e-01 -8.53492245e-02 5.37168562e-01 4.83183026e-01
4.58351910e-01 -1.88569713e+00 -1.05210447e+00 4.61485744e-01
4.56260860e-01 -3.60990286e-01 2.70019561e-01 7.83096313e-01
-2.18469530e-01 -1.86597496e-01 -5.64965367e-01 -5.50381541e-01
-1.70157230e+00 2.63225973e-01 4.30117577e-01 9.40957386e-03
2.18706392e-02 7.94929028e-01 4.65310812e-01 -6.24954283e-01
5.40717132e-02 3.38501483e-01 -1.26853451e-01 -1.84236422e-01
5.25716841e-01 -1.01820096e-01 -1.57788396e-01 -1.18823099e+00
-3.20150346e-01 8.38162124e-01 1.83183998e-01 -9.95549038e-02
1.04013669e+00 3.84446718e-02 -4.76563126e-01 2.03399226e-01
1.24397242e+00 9.18047652e-02 -1.19048321e+00 -5.57261556e-02
-3.44853908e-01 -6.15924239e-01 -3.77773613e-01 -7.73358464e-01
-1.55735219e+00 7.46621013e-01 8.60512853e-01 -4.38121229e-01
1.72895718e+00 -8.43598396e-02 4.53512728e-01 -3.72598432e-02
2.36777410e-01 -9.63215351e-01 8.75354633e-02 3.38855982e-01
1.12390792e+00 -1.06564295e+00 -1.58899486e-01 -7.48026073e-01
-6.67673528e-01 8.79643977e-01 8.40939403e-01 1.57721221e-01
5.26258469e-01 3.70718509e-01 4.84056771e-01 -3.07477396e-02
-6.33638263e-01 -4.09796871e-02 1.25372186e-01 8.15146387e-01
7.22451270e-01 1.77794900e-02 -3.91154408e-01 4.92201507e-01
-3.43506724e-01 2.78145403e-01 1.90049529e-01 5.93116045e-01
1.23594306e-01 -1.45972228e+00 -3.03644925e-01 -1.23188235e-01
-7.28327215e-01 3.32522810e-01 -6.93823755e-01 7.57060826e-01
4.07368630e-01 8.54717433e-01 -3.83893363e-02 -2.81243026e-01
3.18964243e-01 1.73787877e-01 7.49921739e-01 -3.55046988e-01
-2.50825852e-01 3.47919539e-02 1.01729795e-01 -5.90760827e-01
-8.72705519e-01 -4.96738523e-01 -9.49449599e-01 -2.49107525e-01
-5.16426861e-02 6.80050924e-02 5.17471552e-01 7.00707614e-01
5.57635486e-01 2.67049819e-01 9.88491893e-01 -5.79043925e-01
-2.36839980e-01 -9.95839655e-01 -8.01771760e-01 8.48710775e-01
1.78017363e-01 -8.42329383e-01 -1.35006025e-01 4.00759399e-01] | [13.102945327758789, 0.4330434799194336] |
a13bc368-193f-4110-aa4a-36b9d1524d2d | bigissue-a-realistic-bug-localization | 2207.10739 | null | https://arxiv.org/abs/2207.10739v2 | https://arxiv.org/pdf/2207.10739v2.pdf | BigIssue: A Realistic Bug Localization Benchmark | As machine learning tools progress, the inevitable question arises: How can machine learning help us write better code? With significant progress being achieved in natural language processing with models like GPT-3 and Bert, the applications of natural language processing techniques to code are starting to be explored. Most of the research has been focused on automatic program repair (APR), and while the results on synthetic or highly filtered datasets are promising, such models are hard to apply in real-world scenarios because of inadequate bug localization. We propose BigIssue: a benchmark for realistic bug localization. The goal of the benchmark is two-fold. We provide (1) a general benchmark with a diversity of real and synthetic Java bugs and (2) a motivation to improve bug localization capabilities of models through attention to the full repository context. With the introduction of BigIssue, we hope to advance the state of the art in bug localization, in turn improving APR performance and increasing its applicability to the modern development cycle. | ['Caiming Xiong', 'Yingbo Zhou', 'Bo Pang', 'Erik Nijkamp', 'Paul Kassianik'] | 2022-07-21 | null | null | null | null | ['program-repair', 'program-repair'] | ['computer-code', 'reasoning'] | [-2.06421509e-01 1.48610696e-01 -4.07524824e-01 -1.60827905e-01
-9.42082703e-01 -2.77545184e-01 1.67882279e-01 7.20837712e-01
9.66967419e-02 4.44408089e-01 4.42175120e-02 -5.60561299e-01
6.15015216e-02 -6.81487858e-01 -6.60481334e-01 2.44980082e-01
-5.17576039e-01 6.55079773e-03 4.52163547e-01 -4.38817590e-01
6.16346836e-01 -1.65411443e-01 -1.63242507e+00 4.66330230e-01
1.01097572e+00 2.08502725e-01 1.60254225e-01 7.77852893e-01
-6.08983412e-02 1.09482074e+00 -8.32112074e-01 -3.66713285e-01
-5.47599941e-02 -2.57726848e-01 -1.13886356e+00 -1.78112775e-01
2.86995500e-01 1.32506132e-01 3.54103714e-01 1.21005869e+00
1.30094573e-01 -4.73182112e-01 -1.63467154e-01 -1.27578795e+00
-5.13855040e-01 7.65664279e-01 -6.36302173e-01 1.34316951e-01
8.03234875e-01 3.15057784e-01 1.16800463e+00 -5.59196174e-01
6.71405435e-01 1.11326027e+00 1.10597062e+00 4.88456845e-01
-1.16330981e+00 -2.98171323e-02 -2.81750560e-02 1.00417033e-01
-1.05601633e+00 -1.54384464e-01 4.39621419e-01 -7.89817691e-01
1.70337021e+00 1.72737435e-01 4.27749872e-01 8.21124613e-01
7.07026243e-01 6.14266455e-01 7.04333305e-01 -9.82753038e-01
1.55754671e-01 1.51790157e-01 5.97885430e-01 1.18255687e+00
4.57961112e-01 -1.03624351e-01 -4.41911310e-01 -6.76951587e-01
1.84706643e-01 -1.67346686e-01 -1.59132227e-01 -1.55236006e-01
-1.16732657e+00 8.93428266e-01 1.77201316e-01 6.59707069e-01
-3.53849195e-02 4.64902937e-01 7.56538570e-01 6.78650498e-01
5.15740633e-01 1.10094845e+00 -7.69300759e-01 -7.85586238e-01
-1.10448098e+00 7.16871738e-01 1.11175227e+00 9.86644268e-01
8.30711663e-01 -4.67930809e-02 1.32381469e-01 5.83271444e-01
5.41905284e-01 -7.89024830e-02 5.08584201e-01 -7.69458175e-01
6.42945707e-01 1.11380851e+00 1.11777112e-01 -1.04321492e+00
-2.86207229e-01 -1.74252957e-01 1.77456245e-01 3.60515028e-01
4.08475369e-01 -8.90147015e-02 -4.01700109e-01 1.39701068e+00
-1.75304696e-01 4.78643505e-03 -2.05930203e-01 3.86820734e-01
4.12382275e-01 4.11732346e-01 -1.87486574e-01 5.33790886e-02
1.27208734e+00 -8.44791412e-01 -2.62895197e-01 -8.03421855e-01
1.51657701e+00 -8.37016761e-01 1.48815608e+00 6.76533163e-01
-8.02364051e-01 -2.23300740e-01 -1.21264219e+00 1.80959120e-01
-2.76925892e-01 8.68810564e-02 1.14164686e+00 9.18906212e-01
-1.19736588e+00 6.21320963e-01 -1.27626765e+00 -6.07926250e-01
1.27333611e-01 2.30533808e-01 -4.55090851e-01 -4.09190774e-01
-7.11935282e-01 8.61783803e-01 2.40266651e-01 -2.05712855e-01
-8.32296312e-01 -8.24108422e-01 -1.15808165e+00 -1.46863498e-02
6.59680784e-01 -5.76634288e-01 1.54458368e+00 -8.18295419e-01
-6.83492124e-01 6.63405299e-01 -3.86310695e-03 -2.88179725e-01
-2.72896537e-03 -3.27966779e-01 -2.05960721e-01 -5.57998598e-01
3.70478094e-01 3.00724655e-01 5.12869716e-01 -8.47399473e-01
-6.68525755e-01 -1.89803347e-01 4.35162485e-01 -7.77548015e-01
-3.59052688e-01 7.13311076e-01 -1.04429074e-01 -4.76581603e-01
-2.94209033e-01 -6.62773430e-01 -3.11545849e-01 -3.88027668e-01
-1.49864942e-01 -4.01892781e-01 4.84223783e-01 -7.93267131e-01
1.72672188e+00 -1.98769879e+00 2.03919038e-01 -1.92615643e-01
9.52125639e-02 6.56285584e-02 -4.72978681e-01 7.06134260e-01
-3.35863113e-01 5.22448003e-01 -2.75725812e-01 -1.86801359e-01
1.28426597e-01 -5.58431679e-03 -3.41387957e-01 3.14068377e-01
8.16185653e-01 7.55595386e-01 -1.20799994e+00 -3.07221442e-01
-3.45224977e-01 -1.12616614e-01 -1.18875074e+00 -8.42618011e-03
-8.11414540e-01 -2.27317318e-01 -3.51362824e-01 1.02711570e+00
1.51518017e-01 -9.75308046e-02 -2.20916614e-01 7.19794512e-01
-2.61914968e-01 6.05681121e-01 -1.01914692e+00 1.83194411e+00
-3.63981992e-01 6.86509192e-01 -4.92454357e-02 -5.90906322e-01
8.28897297e-01 1.20843843e-01 1.18556507e-01 -5.26842117e-01
-4.23823953e-01 2.59868592e-01 1.68363824e-01 -1.03621697e+00
7.17530608e-01 1.48262516e-01 -4.93407279e-01 6.29278004e-01
7.72258639e-02 -1.90987900e-01 5.87099791e-01 1.80213884e-01
1.92104566e+00 3.03999633e-01 2.91532576e-01 -3.24590564e-01
2.54496574e-01 5.48792839e-01 5.70352852e-01 6.77876055e-01
5.79294413e-02 3.19682091e-01 1.08813834e+00 -6.28817976e-01
-7.95974910e-01 -3.95679533e-01 1.30263358e-01 9.64624465e-01
-3.64355654e-01 -1.33521330e+00 -1.03349638e+00 -9.54486847e-01
-8.99313912e-02 8.87048006e-01 -4.56598550e-01 -2.64075249e-01
-6.46080852e-01 -8.82748425e-01 7.88396478e-01 3.61316442e-01
-1.80824354e-01 -1.09643054e+00 -7.51740217e-01 4.80019599e-01
-9.72730964e-02 -3.97000104e-01 -1.46496713e-01 1.45992726e-01
-9.56243455e-01 -1.33758044e+00 8.34395885e-02 -7.12860405e-01
4.95208085e-01 6.40253499e-02 1.75851154e+00 8.87936532e-01
-6.69737518e-01 3.69105160e-01 -6.41844690e-01 -4.14687186e-01
-8.16723704e-01 2.36166969e-01 -3.74505281e-01 -9.29647803e-01
6.52483463e-01 -3.46162945e-01 3.63476388e-02 2.13749915e-01
-9.47429001e-01 -4.17422891e-01 3.18249643e-01 9.55770850e-01
-5.65098748e-02 3.51437837e-01 5.54089785e-01 -8.82137597e-01
9.34156239e-01 -7.36973584e-01 -8.80208254e-01 2.14196935e-01
-6.54141843e-01 1.28082871e-01 3.31427366e-01 -1.62421584e-01
-7.81743884e-01 -3.73284161e-01 -3.58228713e-01 4.18264270e-01
-1.70233652e-01 1.26639986e+00 6.38600364e-02 -2.71686733e-01
1.16224170e+00 -2.80651063e-01 -2.02864930e-01 -4.78348285e-01
8.99097696e-02 4.75128800e-01 2.06521496e-01 -1.00462091e+00
5.18380105e-01 -2.15050861e-01 -4.89797890e-01 -6.84644997e-01
-4.11381185e-01 -3.06238919e-01 -2.23402277e-01 2.27830842e-01
3.09007615e-01 -6.66947007e-01 -3.11452866e-01 2.14079410e-01
-1.31203723e+00 -4.65887517e-01 -3.93381268e-02 -2.23830808e-02
-4.46583748e-01 6.41470492e-01 -7.06981778e-01 -7.05325305e-01
3.77534181e-02 -1.61126602e+00 1.09530807e+00 -4.79315743e-02
-7.48605609e-01 -1.04178894e+00 5.94579935e-01 3.25401187e-01
3.93121809e-01 2.56948382e-01 1.56182706e+00 -3.16033751e-01
-7.98998833e-01 -5.41252851e-01 9.48892906e-02 2.83624768e-01
1.64425019e-02 3.58567178e-01 -6.90275729e-01 -2.43571088e-01
2.08103850e-01 -3.96269172e-01 4.85108346e-01 -1.54815957e-01
6.88649714e-01 -1.71182081e-01 -3.46838742e-01 -2.95753330e-02
1.56372285e+00 -1.96701333e-01 7.13865638e-01 7.09754527e-01
4.54778522e-01 6.43061578e-01 1.06976485e+00 3.65599126e-01
5.73964953e-01 5.54157913e-01 8.71563375e-01 3.86780083e-01
5.74212782e-02 -7.63006136e-02 7.13556647e-01 7.37109184e-01
1.14635982e-01 5.27243726e-02 -1.63490891e+00 9.53423917e-01
-1.96712172e+00 -8.07454228e-01 -4.13553327e-01 2.21130610e+00
8.99885476e-01 2.92577058e-01 3.59682739e-02 1.31309256e-01
2.03058898e-01 -2.47674525e-01 2.88497545e-02 -5.89862585e-01
4.95834649e-01 1.80888519e-01 -1.85775027e-01 3.45987678e-01
-9.27029610e-01 7.63848901e-01 6.38582087e+00 5.39742351e-01
-1.10715818e+00 1.94762155e-01 2.13442355e-01 2.34833702e-01
-3.02478462e-01 4.56227362e-01 -8.39564025e-01 3.06127667e-01
1.16800725e+00 -7.48203918e-02 4.90999371e-01 1.28467405e+00
7.28215799e-02 -4.39644128e-01 -1.56328821e+00 4.61602032e-01
1.50297150e-01 -1.29278708e+00 -5.34167290e-01 -1.43843979e-01
6.69079900e-01 2.84989446e-01 -3.98083091e-01 7.32285440e-01
1.66230366e-01 -1.07487571e+00 7.31371880e-01 4.42558408e-01
1.11699529e-01 -6.19216025e-01 9.57395017e-01 6.71635747e-01
-8.09559107e-01 -3.68464947e-01 -4.86861825e-01 -4.89088982e-01
-2.58259177e-01 6.24866605e-01 -1.10464358e+00 5.29304445e-01
7.75236785e-01 7.61623502e-01 -1.27861261e+00 1.60128677e+00
-3.13793600e-01 8.89917552e-01 -9.41223279e-02 -1.47390842e-01
1.35495797e-01 3.07954758e-01 4.53646392e-01 1.32713926e+00
2.76378542e-01 -7.15660691e-01 2.39118874e-01 1.24860859e+00
1.90011367e-01 -2.81751715e-02 -6.50874674e-01 -5.16724169e-01
2.01182693e-01 1.09046400e+00 -6.42909765e-01 5.98166557e-03
-9.03780222e-01 2.97024369e-01 5.39305747e-01 -3.49039398e-02
-6.23864055e-01 -5.04581571e-01 5.82841873e-01 4.23384190e-01
-4.02577743e-02 -2.11750120e-01 -4.58159626e-01 -1.09793127e+00
4.70303565e-01 -1.56239510e+00 2.14831784e-01 -6.59790337e-01
-1.02215528e+00 4.45653856e-01 -2.30485685e-02 -9.01659310e-01
-3.67551863e-01 -5.85876942e-01 -9.01631594e-01 7.03726768e-01
-1.30772734e+00 -9.13946509e-01 1.85939875e-02 -1.52150646e-01
6.34094536e-01 -1.39194697e-01 1.01377165e+00 4.22485739e-01
-4.05442357e-01 4.86273110e-01 -3.58297527e-01 -2.98662573e-01
8.87462080e-01 -1.42942035e+00 9.52181578e-01 1.18667161e+00
1.83685496e-01 1.19559026e+00 8.54764104e-01 -9.16563392e-01
-1.72111213e+00 -1.10454834e+00 1.06022370e+00 -1.00453436e+00
1.20128322e+00 -5.25677562e-01 -1.07112670e+00 8.26403737e-01
2.17978895e-01 -5.38368523e-02 3.94313782e-01 5.54500818e-01
-4.64383066e-01 2.39802867e-01 -9.68008518e-01 3.70123714e-01
5.68930328e-01 -5.24720132e-01 -6.87712371e-01 5.94552875e-01
7.68234313e-01 -2.40906417e-01 -9.87219572e-01 2.23557368e-01
8.74251425e-02 -1.00572085e+00 5.43176711e-01 -6.82565331e-01
7.57415831e-01 -4.74813879e-01 6.07535690e-02 -1.24305880e+00
-4.36067060e-02 -7.62578666e-01 3.86877619e-02 1.29156768e+00
7.18490958e-01 -5.10376513e-01 7.49465346e-01 5.93801558e-01
-4.12396848e-01 -8.29805553e-01 -6.36170149e-01 -5.78357637e-01
8.08967352e-02 -8.41355205e-01 4.63486403e-01 9.99876857e-01
8.15300703e-01 4.36052196e-02 -1.28768042e-01 1.68596476e-01
1.21003039e-01 -2.36250088e-02 9.16667700e-01 -1.41131687e+00
-8.29166293e-01 -1.73162803e-01 -7.10510850e-01 -5.01743853e-01
5.49737997e-02 -7.25438178e-01 3.95740986e-01 -1.22231352e+00
3.51011962e-01 -3.08176696e-01 1.87500760e-01 9.86973643e-01
-4.41923976e-01 -9.72233247e-03 -8.72596577e-02 -1.26396328e-01
-8.49571109e-01 1.49493171e-02 5.11134028e-01 -2.23193064e-01
-6.86876848e-02 5.55979200e-02 -9.24051702e-01 8.24519455e-01
4.80265349e-01 -8.66239607e-01 -1.64666161e-01 -6.67849898e-01
1.09075296e+00 1.12127274e-01 3.24381769e-01 -1.07874918e+00
9.63112339e-02 7.69733265e-02 -4.26160127e-01 -4.56250235e-02
-3.89681071e-01 -4.01569694e-01 -3.90575156e-02 5.21694064e-01
-1.09911434e-01 4.85415667e-01 6.33612573e-01 2.87057340e-01
-4.28914160e-01 -9.66650367e-01 3.73625487e-01 -3.14759731e-01
-8.44051242e-01 -2.29007512e-01 -4.95346606e-01 1.00365914e-01
8.93965364e-01 -8.85966495e-02 -7.12282062e-01 1.47340804e-01
-2.91165084e-01 2.32332319e-01 1.11453485e+00 7.88759530e-01
3.44719112e-01 -8.51036012e-01 -5.23943841e-01 2.30319649e-01
7.63552845e-01 -2.51814187e-01 -3.61956656e-01 7.89473355e-01
-7.44048178e-01 4.18297499e-01 5.36869839e-02 -5.84914625e-01
-1.41785276e+00 5.76584041e-01 1.19430646e-01 -4.33693647e-01
-4.49451298e-01 8.36227775e-01 -2.75860161e-01 -5.87835073e-01
1.89511538e-01 -6.59420013e-01 1.34866059e-01 -4.93110090e-01
7.55420327e-01 2.17970923e-01 6.35390699e-01 -3.80268767e-02
-4.56104457e-01 6.50462359e-02 -3.55537444e-01 4.25241709e-01
1.70241714e+00 5.03717780e-01 -8.04213941e-01 3.88969839e-01
5.59735477e-01 2.39943579e-01 -6.69995546e-01 2.45788857e-01
1.02267122e+00 -5.42997897e-01 -1.35306358e-01 -8.68727505e-01
-5.84265947e-01 7.93067873e-01 3.72238457e-01 5.21189809e-01
7.49280989e-01 1.33304834e-01 2.84407586e-01 5.57645679e-01
9.32545364e-01 -6.21199965e-01 2.74749458e-01 6.83088422e-01
7.81109631e-01 -1.22062826e+00 -3.63435075e-02 -3.77709329e-01
5.29569387e-02 1.18736029e+00 7.20953405e-01 -2.33875200e-01
7.43853375e-02 7.26414502e-01 -2.15268403e-01 -5.35763919e-01
-1.11106622e+00 3.61401916e-01 -1.70996547e-01 6.13342464e-01
1.05521393e+00 -3.03534478e-01 -3.88388544e-01 7.96788514e-01
-1.39709450e-02 2.44386017e-01 1.05604148e+00 1.46223307e+00
-5.67929029e-01 -1.77148581e+00 -6.51603341e-01 5.59796751e-01
-6.86832130e-01 -2.64208019e-01 -4.86734062e-01 9.40979600e-01
2.61490762e-01 1.04317999e+00 -5.05998790e-01 -1.82161450e-01
3.99033427e-01 9.80283916e-02 7.22725868e-01 -1.55067849e+00
-9.08460617e-01 -3.29706967e-01 4.35630083e-01 -8.11093330e-01
1.32896379e-01 -8.06624532e-01 -9.45362508e-01 -1.03168719e-01
-7.64732540e-01 3.07496339e-01 6.68229520e-01 8.80014241e-01
4.40935671e-01 6.84831619e-01 -2.08839297e-01 -6.97337866e-01
-6.55013144e-01 -9.43455815e-01 -1.07007682e-01 -1.62301525e-01
2.50191987e-01 -4.60043430e-01 -1.69233888e-01 1.61762550e-01] | [7.589453220367432, 7.704395771026611] |
f175fd59-7e95-4343-ba48-41767cddba17 | a-large-dataset-of-historical-japanese | 2004.08686 | null | https://arxiv.org/abs/2004.08686v1 | https://arxiv.org/pdf/2004.08686v1.pdf | A Large Dataset of Historical Japanese Documents with Complex Layouts | Deep learning-based approaches for automatic document layout analysis and content extraction have the potential to unlock rich information trapped in historical documents on a large scale. One major hurdle is the lack of large datasets for training robust models. In particular, little training data exist for Asian languages. To this end, we present HJDataset, a Large Dataset of Historical Japanese Documents with Complex Layouts. It contains over 250,000 layout element annotations of seven types. In addition to bounding boxes and masks of the content regions, it also includes the hierarchical structures and reading orders for layout elements. The dataset is constructed using a combination of human and machine efforts. A semi-rule based method is developed to extract the layout elements, and the results are checked by human inspectors. The resulting large-scale dataset is used to provide baseline performance analyses for text region detection using state-of-the-art deep learning models. And we demonstrate the usefulness of the dataset on real-world document digitization tasks. The dataset is available at https://dell-research-harvard.github.io/HJDataset/. | ['Melissa Dell', 'Kaixuan Zhang', 'Zejiang Shen'] | 2020-04-18 | null | null | null | null | ['document-layout-analysis'] | ['computer-vision'] | [ 1.29585853e-03 -2.45461076e-01 -4.96414565e-02 -2.09610865e-01
-1.10504186e+00 -1.02103996e+00 5.77550292e-01 3.09635282e-01
-2.23277658e-01 3.06419641e-01 6.00111902e-01 -5.43023825e-01
-7.60202259e-02 -6.51943684e-01 -6.60899580e-01 -5.38980722e-01
-3.74672301e-02 4.89258617e-01 -1.23732917e-01 -1.63590014e-01
6.13186181e-01 5.40993154e-01 -1.08228707e+00 8.26916516e-01
8.57869029e-01 8.53794456e-01 3.48264188e-01 9.01774108e-01
-3.25568169e-01 6.21393800e-01 -1.13033509e+00 -4.37370032e-01
1.03692077e-01 8.86610616e-03 -8.81679595e-01 1.74909621e-01
1.07750773e+00 -6.85052395e-01 -4.69584525e-01 7.85025418e-01
7.24970877e-01 -2.00603068e-01 7.48624384e-01 -7.69090295e-01
-9.61360216e-01 9.19822514e-01 -7.95772016e-01 2.62994796e-01
1.22097567e-01 -1.10948220e-01 1.29779768e+00 -1.08648658e+00
9.95357096e-01 1.04560149e+00 7.25608051e-01 9.12089497e-02
-1.03171408e+00 -3.18329573e-01 1.01188026e-01 2.01216564e-02
-1.30439353e+00 -4.97800916e-01 8.97039890e-01 -6.76575661e-01
9.61474717e-01 2.67286569e-01 4.12227839e-01 1.28322840e+00
6.80262595e-02 1.34910548e+00 6.95767164e-01 -8.22634816e-01
1.29172400e-01 -1.91904649e-01 3.90638679e-01 8.07264686e-01
2.15947285e-01 -5.74994624e-01 -2.47147515e-01 2.83584952e-01
5.79828799e-01 -8.02238211e-02 -1.17570400e-01 -2.13956118e-01
-1.01319194e+00 5.80105424e-01 5.35403609e-01 4.27822351e-01
1.29967406e-01 -1.86246142e-01 6.51712000e-01 -5.59860794e-03
5.12037277e-01 6.32867157e-01 -4.45226580e-01 -1.10857002e-01
-1.23872316e+00 4.37376857e-01 5.06416321e-01 1.19847810e+00
2.76165396e-01 2.01739967e-02 -1.75174430e-01 1.15649712e+00
4.71224077e-02 5.24076581e-01 2.34761499e-02 -5.92951775e-01
1.15041101e+00 8.04483175e-01 -5.78122623e-02 -1.18612683e+00
-6.38033807e-01 -3.78097594e-01 -8.06991577e-01 3.63921262e-02
5.99658489e-01 2.01152246e-02 -1.04725206e+00 9.23710167e-01
2.72680949e-02 -7.94701517e-01 -5.13691545e-01 5.46458602e-01
9.37276363e-01 6.55426502e-01 -4.83864009e-01 6.32356286e-01
1.50492775e+00 -1.04909313e+00 -7.54570901e-01 -2.30720520e-01
8.98427188e-01 -9.85528708e-01 1.41281867e+00 7.86272228e-01
-8.69965255e-01 -5.77964485e-01 -1.29900157e+00 -7.33276486e-01
-8.61281455e-01 7.77939737e-01 4.87820923e-01 6.46352112e-01
-8.62231672e-01 2.35864699e-01 -5.81210077e-01 -4.25195634e-01
9.46489096e-01 -9.94431153e-02 -2.56520748e-01 -2.42167205e-01
-7.67204881e-01 4.90489215e-01 4.70038533e-01 3.91563118e-01
-7.51121759e-01 -5.63592494e-01 -8.47731531e-01 2.55898476e-01
4.21888143e-01 6.63271397e-02 1.36878431e+00 -5.36649048e-01
-8.26249242e-01 8.60821903e-01 1.41594052e-01 -1.06807813e-01
6.36268377e-01 -6.94730759e-01 -4.70530093e-01 1.27094045e-01
8.46167207e-02 5.19321859e-01 7.29124308e-01 -1.34811902e+00
-6.20907009e-01 -5.05119681e-01 -2.08255321e-01 -1.67149588e-01
-6.72659338e-01 9.01574939e-02 -9.80592787e-01 -1.10848129e+00
-2.53211915e-01 -6.92959428e-01 2.75932997e-01 -5.25634065e-02
-1.04646170e+00 2.87237689e-02 8.88867795e-01 -1.29029596e+00
1.66153443e+00 -2.07103848e+00 -1.32526383e-01 2.96977371e-01
3.34332317e-01 1.25909761e-01 -3.54544014e-01 6.13417625e-01
1.90116271e-01 4.85171348e-01 -2.02941984e-01 -3.97746861e-01
1.83096632e-01 -2.47507036e-01 -6.25063837e-01 4.06637788e-01
1.72904208e-01 1.02328587e+00 -6.23101950e-01 -6.92300200e-01
1.42203644e-01 3.18288684e-01 -4.31780428e-01 -1.01670489e-01
-4.06916976e-01 -1.52106330e-01 -1.70645192e-01 1.15029061e+00
7.66477823e-01 -2.02001810e-01 4.43047523e-01 -2.32376531e-01
-1.39942586e-01 3.92847300e-01 -1.03198230e+00 1.83627415e+00
-8.31000283e-02 1.47333515e+00 7.00172335e-02 -4.88777310e-01
7.77470648e-01 -1.81066841e-01 1.71964884e-01 -1.00469017e+00
2.70875126e-01 -7.02689290e-02 -7.94445798e-02 -4.13820714e-01
1.19614291e+00 8.19233358e-01 -3.98445696e-01 5.06648004e-01
-7.76261985e-02 1.17100246e-01 5.76035142e-01 5.64890444e-01
1.22911429e+00 1.22456901e-01 -8.31735581e-02 -3.25601190e-01
1.33419171e-01 2.62077242e-01 2.11587355e-01 8.54687691e-01
9.05813947e-02 7.95931578e-01 7.32369721e-01 -5.70581019e-01
-1.63861382e+00 -9.41050768e-01 -2.61574209e-01 1.23544633e+00
-3.54669780e-01 -7.68213093e-01 -1.02607787e+00 -6.68626010e-01
7.86778405e-02 4.99289304e-01 -8.84468436e-01 3.61800134e-01
-8.27172101e-01 -5.91093481e-01 9.30349648e-01 9.17564511e-01
2.26671904e-01 -1.19557953e+00 -3.07078302e-01 -2.05287114e-01
-1.86362654e-01 -8.76885116e-01 -4.45384175e-01 2.40399495e-01
-5.38420796e-01 -1.11474586e+00 -7.24368989e-01 -8.25453043e-01
8.40078771e-01 2.06055835e-01 1.23200834e+00 1.37989014e-01
-6.72951341e-01 -1.36329725e-01 -4.24695522e-01 -5.25736630e-01
-2.32714206e-01 6.05454504e-01 -4.00767416e-01 -5.56054175e-01
2.35615268e-01 1.88677445e-01 -4.58555311e-01 1.13027960e-01
-9.90769982e-01 1.68897882e-02 6.85694516e-01 6.58147871e-01
3.71211320e-01 4.47934985e-01 1.54821023e-01 -1.12034750e+00
7.58736789e-01 -1.78536400e-01 -7.43287623e-01 4.08821195e-01
-3.17324132e-01 -1.50213674e-01 6.43005788e-01 -2.16203541e-01
-1.20229936e+00 -2.32671618e-01 1.84307955e-02 9.08410698e-02
-2.59735644e-01 6.32950783e-01 -5.04611790e-01 5.41505873e-01
6.62415564e-01 -1.54778451e-01 -6.39588356e-01 -1.09209144e+00
6.77171886e-01 1.03863764e+00 9.66425419e-01 -7.07356870e-01
7.38587022e-01 4.18411821e-01 -4.82857466e-01 -1.08301902e+00
-8.97130907e-01 -2.49256447e-01 -1.33110595e+00 -1.52042642e-01
7.12417006e-01 -7.02332377e-01 -1.03048570e-01 8.20971370e-01
-1.03326976e+00 -6.70674741e-01 1.13296159e-01 -3.07344258e-01
2.30304487e-02 3.66597176e-01 -8.92025054e-01 -6.58061743e-01
-4.85235214e-01 -8.67357314e-01 1.27652717e+00 7.23515972e-02
-3.18257749e-01 -7.82142580e-01 -3.24323960e-02 5.18991053e-01
-6.72995150e-02 3.21873724e-01 1.33925378e+00 -5.03503680e-01
-5.29080927e-01 -4.38312590e-01 -4.19874161e-01 1.40753109e-02
4.74783443e-02 7.80442834e-01 -1.19195223e+00 -2.68857121e-01
-7.56241024e-01 -2.78062016e-01 1.04630315e+00 4.59834725e-01
1.57708204e+00 -2.51913667e-01 -4.21278894e-01 5.95877230e-01
1.27805650e+00 2.35358924e-01 7.78302372e-01 9.21102405e-01
1.23790026e+00 5.57545483e-01 4.15419042e-01 5.97744107e-01
1.79003179e-01 3.92021418e-01 1.49687827e-01 -4.15196151e-01
-1.92989528e-01 -3.62671733e-01 -2.42006201e-02 7.44341493e-01
3.72737080e-01 -8.65339637e-01 -1.49738002e+00 7.36736059e-01
-1.59470165e+00 -9.13948357e-01 -1.88731655e-01 1.76754487e+00
7.70462930e-01 5.71063519e-01 1.44984633e-01 4.27391291e-01
6.13182843e-01 3.48830104e-01 -3.21464032e-01 -3.59616697e-01
-4.05864298e-01 -1.24059491e-01 5.00981808e-01 1.11023523e-01
-1.55375302e+00 1.02380252e+00 6.25155210e+00 8.90558839e-01
-7.45507956e-01 -3.47087771e-01 7.58591413e-01 -2.46763587e-01
-3.20420355e-01 -2.87831306e-01 -1.03354633e+00 5.08149385e-01
6.61221147e-01 5.69809616e-01 2.01423481e-01 8.89932334e-01
8.49754736e-02 -1.26884297e-01 -9.94982600e-01 6.94440842e-01
2.28564039e-01 -1.62838960e+00 -9.38384831e-02 3.87121379e-01
8.20723116e-01 -7.44913071e-02 2.17185065e-01 1.62149668e-01
3.13961655e-01 -9.56103981e-01 1.19220257e+00 4.33319271e-01
9.29206550e-01 -9.29488301e-01 6.12755835e-01 -7.24219084e-02
-1.03839517e+00 -2.68221617e-01 -3.27121943e-01 2.21455112e-01
-1.45459389e-02 7.18773782e-01 -7.81800210e-01 2.91460127e-01
8.52396846e-01 7.35142112e-01 -1.32117105e+00 1.01287401e+00
-3.53715390e-01 5.65125108e-01 -8.26301128e-02 -3.65208951e-03
2.76435435e-01 -5.85692264e-02 1.13536678e-01 1.68274570e+00
5.46651073e-02 -5.62464178e-01 -1.86258465e-01 8.99803102e-01
-4.77654725e-01 3.67846601e-02 -7.37400830e-01 -4.78214860e-01
7.11852014e-01 1.47478008e+00 -1.11130309e+00 -3.15044791e-01
-3.64321381e-01 8.02569985e-01 6.16598785e-01 4.80255723e-01
-7.43513584e-01 -9.08212006e-01 2.59632975e-01 1.48025647e-01
5.27966261e-01 -5.36287844e-01 -8.26902926e-01 -9.91933525e-01
2.19314218e-01 -1.14408600e+00 5.36594987e-01 -9.50931907e-01
-1.19691694e+00 3.05363089e-01 -1.98169753e-01 -9.51015115e-01
2.28691846e-01 -1.00442922e+00 -5.55635095e-01 5.29722095e-01
-9.12232697e-01 -1.43556774e+00 -4.11868989e-01 8.51738527e-02
8.80903900e-01 -2.61604339e-01 4.62453365e-01 4.38512594e-01
-1.23043513e+00 8.13482046e-01 7.36808181e-01 1.04004776e+00
8.71102870e-01 -1.63054323e+00 1.15944445e+00 1.09872556e+00
4.18570966e-01 8.25248659e-01 3.12453538e-01 -8.66355717e-01
-1.41655648e+00 -1.04529440e+00 7.26791441e-01 -9.97735083e-01
6.96497619e-01 -1.26655185e+00 -9.89649177e-01 8.11097801e-01
3.94583493e-01 -5.19143403e-01 6.54775083e-01 4.68608171e-01
-5.31848729e-01 2.71985605e-02 -6.54286265e-01 7.71229804e-01
8.13330412e-01 -7.48548746e-01 -5.22612631e-01 3.43921542e-01
3.59416515e-01 -5.77055693e-01 -8.00666571e-01 -1.25125080e-01
8.12547505e-01 -6.23083055e-01 8.78409207e-01 -4.39882815e-01
9.79624093e-01 -1.65258363e-01 -3.84939052e-02 -1.01342964e+00
-3.89499307e-01 -3.57599139e-01 -4.26296145e-01 1.69884944e+00
6.90643191e-01 7.74385482e-02 7.17086077e-01 4.81725007e-01
-3.56203824e-01 -5.44275641e-01 -3.20290685e-01 -6.40947104e-01
2.44984806e-01 -4.35632974e-01 6.79738224e-01 8.53579462e-01
-2.38177311e-02 3.53201717e-01 -2.34445408e-01 -3.77308838e-02
3.92908841e-01 1.74417704e-01 9.68122721e-01 -1.13751090e+00
1.12415798e-01 -8.33455443e-01 1.80202961e-01 -9.19115126e-01
-2.38130875e-02 -7.13615656e-01 4.87896353e-02 -1.84764802e+00
1.42810956e-01 -3.29256356e-01 -9.92435515e-02 6.52629614e-01
-2.42248718e-02 -3.31772529e-02 1.24770544e-01 3.55272233e-01
-6.80318534e-01 2.92336017e-01 9.32286322e-01 -6.44627512e-01
-1.36676162e-01 -5.66466153e-01 -7.15938151e-01 6.16829574e-01
1.08783770e+00 -1.32012084e-01 -2.13112921e-01 -9.45360482e-01
4.28377062e-01 -7.42022395e-01 6.28547594e-02 -8.12650084e-01
2.73169335e-02 1.39360681e-01 1.22453010e+00 -1.44704449e+00
-3.60173732e-02 -4.36300009e-01 -5.70455968e-01 -1.30818477e-02
-5.26225328e-01 2.70353258e-01 5.21126151e-01 3.83166879e-01
1.31318450e-01 -1.69035792e-01 3.76288146e-01 -5.32330237e-02
-9.17899609e-01 -1.69765100e-01 -5.42531133e-01 1.27746850e-01
4.55824822e-01 -1.14348382e-01 -1.17893267e+00 1.35921175e-02
-8.36432874e-02 2.48927966e-01 6.45327747e-01 7.58403063e-01
5.23413420e-01 -1.18272269e+00 -3.81897271e-01 1.49362907e-01
4.04826105e-01 1.63972110e-01 1.74736425e-01 3.63839567e-01
-1.04843867e+00 6.79266989e-01 -3.44992548e-01 -3.40205252e-01
-1.20054650e+00 7.82181084e-01 -1.52624384e-01 -1.39228001e-01
-9.43891883e-01 6.92620277e-01 3.22646387e-02 -4.26765919e-01
5.40188551e-01 -5.59822321e-01 -2.01722980e-01 3.48629802e-01
8.60582769e-01 6.10495806e-01 3.15871388e-01 -3.56807083e-01
-1.99999020e-01 2.34959364e-01 -5.22119045e-01 -7.46423975e-02
1.45170617e+00 -1.23062238e-01 -8.57412294e-02 2.44110212e-01
1.23528588e+00 2.62671620e-01 -1.18746424e+00 -1.74489170e-02
4.42846775e-01 -5.73704362e-01 1.01588078e-01 -1.03248262e+00
-8.82969677e-01 1.01128030e+00 4.15835112e-01 3.18429559e-01
8.83948803e-01 -1.09147891e-01 8.02324295e-01 6.14835739e-01
-1.52004138e-01 -1.73025596e+00 1.27596974e-01 6.14693642e-01
1.04682410e+00 -1.24042094e+00 4.04843092e-01 -2.11958829e-02
-4.75166410e-01 1.26600361e+00 7.47205913e-01 1.65844515e-01
2.47230798e-01 6.60263538e-01 1.76988140e-01 -4.90577012e-01
-3.85764271e-01 1.63908117e-02 6.08532429e-01 5.37365437e-01
6.36703670e-01 5.94765507e-02 2.74120033e-01 4.30759609e-01
-4.76347625e-01 -5.42506695e-01 4.98187155e-01 1.16362751e+00
-3.83612156e-01 -8.70650887e-01 -5.76426506e-01 6.65410936e-01
-3.48713785e-01 -2.87196487e-01 -1.08043027e+00 9.60598707e-01
-1.36919454e-01 7.40327001e-01 3.02799076e-01 -2.08829105e-01
1.84611663e-01 -1.44598901e-01 3.80559653e-01 -3.31211299e-01
-3.91750038e-01 4.38973308e-01 3.31860244e-01 -2.01575622e-01
3.25569421e-01 -6.54081881e-01 -1.02907896e+00 -3.50145370e-01
-1.20176040e-01 -4.27967012e-01 7.24117339e-01 5.53307950e-01
3.76326293e-01 7.32331812e-01 1.60937086e-01 -8.86495054e-01
-1.25965655e-01 -1.13393092e+00 -5.33336401e-01 4.19494182e-01
3.59998584e-01 -2.87078261e-01 8.96082297e-02 2.81872720e-01] | [11.68596363067627, 2.6655690670013428] |
bb5aaccf-a37c-4848-b89b-9b7258549968 | dynamic-curriculum-learning-for-low-resource | 2011.14608 | null | https://arxiv.org/abs/2011.14608v1 | https://arxiv.org/pdf/2011.14608v1.pdf | Dynamic Curriculum Learning for Low-Resource Neural Machine Translation | Large amounts of data has made neural machine translation (NMT) a big success in recent years. But it is still a challenge if we train these models on small-scale corpora. In this case, the way of using data appears to be more important. Here, we investigate the effective use of training data for low-resource NMT. In particular, we propose a dynamic curriculum learning (DCL) method to reorder training samples in training. Unlike previous work, we do not use a static scoring function for reordering. Instead, the order of training samples is dynamically determined in two ways - loss decline and model competence. This eases training by highlighting easy samples that the current model has enough competence to learn. We test our DCL method in a Transformer-based system. Experimental results show that DCL outperforms several strong baselines on three low-resource machine translation benchmarks and different sized data of WMT' 16 En-De. | ['Jingbo Zhu', 'Tong Xiao', 'Qi Ju', 'Shen Huang', 'Zeyang Wang', 'Kai Feng', 'Yufan Jiang', 'Bojie Hu', 'Chen Xu'] | 2020-11-30 | null | https://aclanthology.org/2020.coling-main.352 | https://aclanthology.org/2020.coling-main.352.pdf | coling-2020-8 | ['low-resource-neural-machine-translation'] | ['natural-language-processing'] | [ 3.80788803e-01 -2.83623803e-02 -5.34372866e-01 -3.67338508e-01
-1.11740398e+00 -7.50178635e-01 7.58889496e-01 -1.18102349e-01
-8.37336779e-01 9.82651889e-01 1.45588741e-01 -9.18815136e-01
3.70231956e-01 -4.91492659e-01 -1.01048982e+00 -3.53400856e-01
4.34592336e-01 1.18552673e+00 2.73472816e-01 -5.46503484e-01
6.92576990e-02 1.89916417e-01 -9.27773774e-01 6.84425294e-01
1.03044438e+00 4.15930480e-01 4.25236851e-01 3.98161173e-01
-1.46980226e-01 3.65501970e-01 -7.30656803e-01 -6.95014715e-01
4.32101041e-01 -5.22689521e-01 -1.06072342e+00 -3.31144184e-01
5.27035296e-01 -2.98036456e-01 2.22256258e-01 8.38082731e-01
6.89867795e-01 -8.51649418e-02 4.75811869e-01 -9.12647426e-01
-5.46278179e-01 9.18453515e-01 -3.90485376e-01 3.11105400e-01
-2.53162980e-02 1.05994850e-01 9.78663623e-01 -1.01415992e+00
9.01210129e-01 1.34441233e+00 4.28810447e-01 9.41049755e-01
-1.44170296e+00 -7.42635250e-01 1.69026449e-01 3.51203710e-01
-8.77248228e-01 -5.02109706e-01 5.04279137e-01 -2.39029184e-01
1.29179454e+00 3.00021112e-01 4.15614754e-01 1.39504659e+00
1.02740169e-01 9.10452664e-01 1.48323667e+00 -8.24886262e-01
5.09787947e-02 2.18298674e-01 -1.72738358e-01 3.54183704e-01
-7.77831525e-02 1.26600638e-01 -4.95676070e-01 7.94671476e-02
4.95213628e-01 -4.84682828e-01 -6.81215450e-02 -2.52512276e-01
-1.46434677e+00 6.90014064e-01 1.34146065e-01 4.43191677e-01
2.56470516e-02 -5.09151667e-02 4.22522932e-01 8.48491669e-01
6.40012205e-01 8.92742813e-01 -9.02067184e-01 -5.26071131e-01
-9.07860339e-01 1.08612873e-01 7.13020802e-01 8.60176802e-01
5.62051535e-01 -1.25906572e-01 -2.95804352e-01 1.22283041e+00
-1.88896999e-01 3.80613893e-01 7.10331202e-01 -6.73121631e-01
1.14006889e+00 4.72152710e-01 -1.08261846e-01 -2.61050463e-01
7.84780905e-02 -3.50176543e-01 -5.91212034e-01 4.80264574e-02
6.05130255e-01 -2.17563272e-01 -9.91871417e-01 1.80453432e+00
9.43841338e-02 4.92122360e-02 1.03319444e-01 8.78266811e-01
3.32789570e-01 8.94325554e-01 -1.66469991e-01 -2.45298162e-01
9.18196738e-01 -1.33472431e+00 -5.67066431e-01 -4.34989840e-01
9.39170599e-01 -1.12522483e+00 1.46374464e+00 5.00863850e-01
-1.15722501e+00 -5.75424373e-01 -9.80904520e-01 -2.53493756e-01
-4.17531699e-01 2.42987156e-01 4.37988609e-01 4.85405952e-01
-1.10174263e+00 8.65979671e-01 -8.74851227e-01 -4.87554550e-01
1.82944521e-01 6.36609912e-01 -2.88843066e-01 -2.49504134e-01
-1.40116394e+00 1.32071483e+00 3.28074843e-01 4.69657183e-02
-7.61224270e-01 -6.47608697e-01 -3.95551234e-01 -5.86337335e-02
1.87390506e-01 -5.60320199e-01 1.73711681e+00 -1.18068373e+00
-1.84072649e+00 8.65834177e-01 -1.20891176e-01 -5.23492873e-01
7.54580796e-01 -4.17753339e-01 -8.30877721e-02 -2.99926728e-01
-1.33845106e-01 9.53018367e-01 5.40243685e-01 -1.18899298e+00
-6.33988202e-01 -2.40998790e-02 1.44653276e-01 3.66788715e-01
-5.90289354e-01 2.84202665e-01 -4.80536222e-01 -5.70886850e-01
-2.30033830e-01 -1.15212405e+00 -2.40387127e-01 -4.18658882e-01
-4.58990782e-02 -4.62489337e-01 5.28592527e-01 -6.17055237e-01
1.35024369e+00 -1.82628536e+00 4.16824669e-01 -9.73852500e-02
-1.65762708e-01 6.32017493e-01 -4.48305488e-01 4.44074869e-01
1.90444306e-01 2.78532863e-01 -1.04781270e-01 -4.75988358e-01
1.99240521e-02 4.62589175e-01 -3.22070390e-01 -1.38350964e-01
4.45769340e-01 9.44488823e-01 -9.71723080e-01 -6.17203832e-01
-7.63781145e-02 2.39776090e-01 -8.00303698e-01 1.77783445e-01
-4.90576744e-01 4.92803961e-01 -3.16954762e-01 4.67394441e-01
3.57546806e-01 -1.32686988e-01 4.37202305e-01 2.98062321e-02
-8.19474235e-02 8.41972828e-01 -5.83749533e-01 1.76971781e+00
-6.65737689e-01 6.82127714e-01 -3.47832441e-01 -8.83208215e-01
6.98649108e-01 3.43097448e-01 1.38996854e-01 -1.15499151e+00
-3.72587815e-02 6.03939831e-01 4.59512144e-01 -3.42447162e-01
6.52161777e-01 -6.69092759e-02 7.43564889e-02 4.89059597e-01
1.34366341e-02 -4.20891345e-02 3.63013417e-01 -5.62942140e-02
1.01935673e+00 4.48892593e-01 -7.48881474e-02 -2.12308303e-01
2.06594869e-01 2.97505558e-01 7.81848013e-01 4.66757178e-01
-1.06305918e-02 5.26997387e-01 3.20629865e-01 -3.65069777e-01
-1.31290758e+00 -7.96999633e-01 -2.43602376e-02 1.39457798e+00
-2.32682779e-01 -5.44921398e-01 -9.93283272e-01 -9.52624977e-01
-2.45076284e-01 7.47837067e-01 -4.07471329e-01 -2.31771529e-01
-1.35182476e+00 -8.14057827e-01 3.78366709e-01 3.90624017e-01
2.95606762e-01 -1.10907900e+00 -2.33592451e-01 4.23334122e-01
-4.89159256e-01 -1.11587238e+00 -6.94072068e-01 4.23862696e-01
-1.24858606e+00 -4.61124450e-01 -6.12303197e-01 -9.98753726e-01
7.05108464e-01 1.62135765e-01 1.55770683e+00 -2.55713705e-02
2.74296165e-01 -3.29664916e-01 -4.57522720e-01 -3.27384204e-01
-8.23519528e-01 8.74324262e-01 4.42806333e-02 -4.53172177e-01
5.39891958e-01 -6.14520073e-01 -3.49404484e-01 5.77466309e-01
-5.35595834e-01 4.68874961e-01 1.03206789e+00 1.06298327e+00
4.87652421e-01 -5.07873118e-01 5.01112938e-01 -1.07420588e+00
8.01788867e-01 -2.79416054e-01 -5.52935660e-01 5.54829299e-01
-8.58813584e-01 1.98495224e-01 9.47074115e-01 -7.66793549e-01
-8.17658246e-01 -2.39853665e-01 -7.05998912e-02 -4.52114791e-01
2.11316347e-01 4.17961389e-01 -7.14438930e-02 2.71370947e-01
6.62653208e-01 1.16571933e-01 -2.03055874e-01 -6.82936966e-01
1.66305825e-01 7.90965378e-01 9.00855064e-02 -1.07803178e+00
7.51003087e-01 -2.42871955e-01 -4.12783384e-01 -2.69978374e-01
-6.16944730e-01 -1.17156900e-01 -6.82205260e-01 1.20574154e-01
5.95318496e-01 -7.07100570e-01 -1.66634381e-01 1.72371328e-01
-1.20786631e+00 -9.05976057e-01 -7.82608986e-02 4.38081115e-01
-3.84780914e-01 -1.80992614e-02 -8.29796135e-01 -1.67405501e-01
-3.87191772e-01 -1.63358879e+00 9.25668538e-01 -2.53299505e-01
-2.81426281e-01 -8.02120864e-01 2.29688600e-01 5.92510402e-01
6.12118065e-01 -2.48598292e-01 1.12856114e+00 -6.67351902e-01
-6.86587811e-01 2.49868229e-01 -3.70679260e-03 4.93647963e-01
1.25700280e-01 -1.72253773e-01 -7.52686203e-01 -5.86143017e-01
-2.45409951e-01 -6.18229032e-01 6.42858207e-01 -6.64140433e-02
9.88692462e-01 -4.55094904e-01 -3.33659977e-01 5.96161842e-01
1.23308945e+00 1.65647104e-01 5.54942906e-01 6.82637691e-01
5.85309088e-01 4.67562884e-01 7.64858544e-01 -3.05699944e-01
3.14035684e-01 1.06199586e+00 1.30395889e-01 -1.28764316e-01
-2.98530430e-01 -3.27025443e-01 7.49236524e-01 1.46527135e+00
-2.82523096e-01 -2.18580559e-01 -1.08880937e+00 4.38817650e-01
-1.74349964e+00 -6.91126645e-01 2.48242959e-01 2.17179441e+00
1.52558041e+00 4.19016153e-01 -3.97779560e-03 -1.57269821e-01
5.60918212e-01 -2.09833905e-01 -3.56088728e-01 -8.14761460e-01
2.76168678e-02 4.53789145e-01 4.60481405e-01 5.95595181e-01
-7.81441450e-01 1.44086063e+00 6.04503059e+00 9.81712162e-01
-1.31602645e+00 4.33111817e-01 7.16326892e-01 -2.73601443e-01
-2.66220003e-01 9.98619571e-02 -1.04606223e+00 5.66362321e-01
1.22280538e+00 1.25773875e-02 5.98156154e-01 6.75885618e-01
1.51238993e-01 2.71438926e-01 -1.37612379e+00 5.22234499e-01
-1.68512166e-01 -1.21323574e+00 3.04080814e-01 1.58515528e-01
8.39530766e-01 4.43333626e-01 4.21700478e-02 7.41787255e-01
5.64804792e-01 -9.06885564e-01 5.94497263e-01 -3.59537676e-02
9.18466389e-01 -5.63056171e-01 6.04467154e-01 4.76283073e-01
-6.39402688e-01 8.92637148e-02 -4.91031528e-01 -8.53057206e-03
-7.19898716e-02 2.95434147e-01 -1.29178274e+00 4.18372840e-01
4.98158693e-01 4.05324072e-01 -7.28303909e-01 7.72886932e-01
-3.40608746e-01 1.07922840e+00 -1.95452616e-01 -8.95733610e-02
3.50656122e-01 -2.07177415e-01 3.16326886e-01 1.29326165e+00
2.62492090e-01 -2.43909642e-01 2.26600930e-01 3.69121343e-01
-4.35734361e-01 2.76333958e-01 -4.06102866e-01 9.91446152e-03
5.50573349e-01 1.07658136e+00 -3.76674563e-01 -3.39272588e-01
-3.87158990e-01 1.13990879e+00 8.07748556e-01 2.05662295e-01
-6.75561607e-01 1.66096285e-01 5.81230879e-01 1.26095369e-01
1.46099955e-01 -3.52890253e-01 5.09818038e-03 -1.30878484e+00
2.25946471e-01 -1.52718794e+00 1.23381466e-01 -5.39169729e-01
-1.10737824e+00 7.81879127e-01 -1.81864034e-02 -1.40131283e+00
-4.94926542e-01 -6.36214495e-01 -3.53648663e-01 8.34764421e-01
-1.66016138e+00 -1.20276046e+00 3.94257545e-01 2.52879679e-01
9.83839035e-01 -2.25025341e-01 9.09727573e-01 6.30529165e-01
-5.49485683e-01 1.00344741e+00 3.80887508e-01 1.98697552e-01
1.24019992e+00 -1.37231636e+00 8.31494629e-01 7.79407322e-01
2.82522172e-01 7.63470232e-01 6.61793768e-01 -5.11780083e-01
-1.39125538e+00 -9.43921566e-01 1.34381831e+00 -6.53935194e-01
6.28843546e-01 -7.60660827e-01 -9.13389683e-01 7.46988773e-01
5.38806438e-01 -2.00597286e-01 5.61083496e-01 3.34871918e-01
-4.13730651e-01 -2.18885377e-01 -7.29389131e-01 7.57736206e-01
1.12935960e+00 -3.12745064e-01 -7.19778061e-01 4.40565199e-01
9.06052291e-01 -5.59303105e-01 -8.15241814e-01 5.17340958e-01
5.27325809e-01 -4.44191307e-01 7.64869213e-01 -9.13086355e-01
6.52622163e-01 -3.59653793e-02 -9.07075629e-02 -1.79714239e+00
-1.76275089e-01 -7.92187512e-01 3.44227366e-02 1.25600493e+00
9.32762861e-01 -5.78730166e-01 8.42000484e-01 2.54845321e-01
-3.28258991e-01 -1.11793137e+00 -9.24228489e-01 -9.41490233e-01
7.53833234e-01 4.23291922e-02 6.73274815e-01 1.21912956e+00
-3.54650877e-02 6.99793398e-01 -3.67943287e-01 -2.62816668e-01
2.15549186e-01 1.17863595e-01 8.78790081e-01 -1.08542562e+00
-5.98090529e-01 -5.84103167e-01 1.99452102e-01 -1.20220530e+00
1.04156829e-01 -1.17133820e+00 1.19333649e-02 -1.46004021e+00
1.58233717e-01 -7.00948477e-01 -3.37530285e-01 6.83812916e-01
-3.25978369e-01 4.24397849e-02 2.61409134e-01 3.82698268e-01
-4.41779047e-01 4.00922418e-01 1.47292995e+00 -2.05711603e-01
-1.23486958e-01 -1.24664247e-01 -3.58916163e-01 1.76212594e-01
1.11899018e+00 -5.52249432e-01 -4.93481308e-01 -1.16190994e+00
4.36029881e-01 -6.22472279e-02 -8.68408084e-02 -7.70402730e-01
1.06394202e-01 -2.95775384e-01 1.60672367e-01 -3.96448135e-01
1.84896663e-01 -6.84371650e-01 -1.41456887e-01 2.72271127e-01
-5.40864646e-01 5.44713080e-01 2.47631967e-01 -8.82835984e-02
-1.80524662e-01 -1.45678878e-01 6.52337313e-01 -1.46539673e-01
-3.84680420e-01 1.24172397e-01 -2.11449876e-01 1.46176234e-01
3.79828006e-01 -9.34345126e-02 -3.93518060e-01 1.37340175e-02
-4.10316318e-01 3.80624205e-01 5.34885406e-01 8.00162017e-01
2.65511543e-01 -1.35507071e+00 -8.33701789e-01 8.05397257e-02
1.78795338e-01 -2.54629347e-02 -4.28666502e-01 8.43857467e-01
-2.65297979e-01 4.24992651e-01 -3.37088495e-01 -5.08883953e-01
-1.27836049e+00 3.74757767e-01 3.31894487e-01 -8.16739738e-01
-3.72545570e-01 7.50327110e-01 -1.83167890e-01 -8.89203548e-01
1.08969003e-01 -3.09458017e-01 -1.16039276e-01 -2.97888909e-02
2.19885468e-01 2.94319186e-02 4.16416615e-01 -3.91016573e-01
-3.82783562e-02 3.28461558e-01 -7.45185554e-01 -3.72983843e-01
1.40820217e+00 9.08956900e-02 -5.27153127e-02 3.86549503e-01
1.14333606e+00 -9.06697735e-02 -1.07931769e+00 -4.36659127e-01
2.59212285e-01 -3.65929395e-01 -1.97550997e-01 -1.08006561e+00
-8.11749279e-01 1.09270298e+00 5.56189239e-01 -1.89629629e-01
1.06199110e+00 -2.91584402e-01 9.69671249e-01 6.99783385e-01
5.93141377e-01 -1.37510586e+00 5.09003103e-02 9.06702459e-01
8.22195292e-01 -1.30663836e+00 -2.78844833e-01 -1.72337458e-01
-3.69912177e-01 9.91312683e-01 8.78315210e-01 1.94472030e-01
8.58495012e-02 1.93015739e-01 3.36414695e-01 3.42475653e-01
-1.17137456e+00 2.64783707e-02 1.48819700e-01 2.47356057e-01
7.67178953e-01 4.57297266e-03 -5.43106496e-01 1.23249017e-01
-5.90674818e-01 6.90013543e-03 2.75189877e-01 9.39086616e-01
-1.96884155e-01 -1.93365896e+00 -1.15556218e-01 2.79648274e-01
-5.99258423e-01 -4.18716103e-01 -5.56302726e-01 8.57075036e-01
8.98993239e-02 7.58352041e-01 -1.91840172e-01 -4.92593855e-01
3.57723236e-01 3.53916317e-01 7.52231300e-01 -8.64561319e-01
-8.95414412e-01 1.95116010e-02 3.62856865e-01 -3.88703972e-01
-3.21329325e-01 -6.98072493e-01 -8.34808350e-01 -2.04761028e-01
-3.95937949e-01 3.25832039e-01 8.00210655e-01 8.75987768e-01
2.58330673e-01 3.29193056e-01 5.87865174e-01 -6.43930435e-01
-9.01149511e-01 -1.23351812e+00 4.06448126e-01 2.66236275e-01
1.65827468e-01 -4.48161930e-01 -1.30666316e-01 4.65003736e-02] | [11.599347114562988, 10.18521499633789] |
ea9b2db7-7a7a-45ea-af36-debd5575c6b7 | 190910305 | 1909.10305 | null | https://arxiv.org/abs/1909.10305v2 | https://arxiv.org/pdf/1909.10305v2.pdf | Deep Multi-Facial patches Aggregation Network for Expression Classification from Face Images | Emotional Intelligence in Human-Computer Interaction has attracted increasing attention from researchers in multidisciplinary research fields including psychology, computer vision, neuroscience, artificial intelligence, and related disciplines. Human prone to naturally interact with computers face-to-face. Human Expressions is an important key to better link human and computers. Thus, designing interfaces able to understand human expressions and emotions can improve Human-Computer Interaction (HCI) for better communication. In this paper, we investigate HCI via a deep multi-facial patches aggregation network for Face Expression Recognition (FER). Deep features are extracted from facial parts and aggregated for expression classification. Several problems may affect the performance of the proposed framework like the small size of FER datasets and the high number of parameters to learn. For That, two data augmentation techniques are proposed for facial expression generation to expand the labeled training. The proposed framework is evaluated on the extended Cohn-Konade dataset (CK+) and promising results are achieved. | ['Alice Othmani', 'Amine Djerghri', 'Ahmed Rachid Hazourli'] | 2019-09-23 | null | null | null | null | ['facial-expression-generation', 'emotional-intelligence'] | ['computer-vision', 'natural-language-processing'] | [ 2.40297645e-01 4.92069274e-02 1.06844634e-01 -7.04278409e-01
-1.22178495e-01 -4.81164828e-02 3.25282037e-01 -4.32592511e-01
-3.32474530e-01 6.54728949e-01 -1.27512828e-01 3.02767128e-01
1.21679775e-01 -4.14961159e-01 -2.13453278e-01 -7.64069974e-01
-1.13081440e-01 -1.14766225e-01 -5.06859481e-01 -3.27643275e-01
-1.49375156e-01 7.56682456e-01 -2.02185941e+00 3.84296507e-01
5.87120116e-01 1.63489246e+00 -3.11072797e-01 3.63193601e-01
4.56494801e-02 9.02246177e-01 -5.08894920e-01 -5.88850200e-01
-1.66725412e-01 -5.92481792e-01 -6.41227603e-01 2.35464990e-01
-8.42873529e-02 -7.47326761e-02 2.61934161e-01 1.03921533e+00
5.65442204e-01 3.02672654e-01 6.90056980e-01 -1.61200619e+00
-4.06306833e-01 2.72457022e-02 -8.87738109e-01 -2.85599619e-01
3.90285999e-01 -1.64957181e-01 3.88449460e-01 -1.03924572e+00
4.79498416e-01 1.43322325e+00 5.13898194e-01 8.85710537e-01
-7.23405838e-01 -9.46388662e-01 -1.65012553e-01 4.29171205e-01
-1.62189555e+00 -5.81756055e-01 9.66852665e-01 -3.54333848e-01
5.96049070e-01 2.10480958e-01 6.76840961e-01 1.17579246e+00
-2.52010673e-01 7.10429728e-01 1.27389336e+00 -5.89535117e-01
3.50699248e-03 5.86406469e-01 1.37914509e-01 7.11180210e-01
-2.92019516e-01 -2.50255316e-01 -5.28680682e-01 -4.72681932e-02
6.10826194e-01 -2.33448774e-01 -1.42120942e-01 1.66783601e-01
-4.42408860e-01 7.11082816e-01 4.95583981e-01 6.27904296e-01
-7.62643039e-01 3.91565263e-02 6.21059656e-01 2.36946583e-01
4.66727167e-01 2.54058152e-01 -1.63895503e-01 -2.17877090e-01
-4.29016113e-01 1.07460424e-01 6.29471719e-01 7.16666520e-01
1.01669085e+00 1.45366326e-01 -2.47943066e-02 1.28527260e+00
2.99507469e-01 4.46485758e-01 3.41699779e-01 -8.45305443e-01
-1.33376792e-01 8.01863730e-01 -2.38979042e-01 -1.38511646e+00
-2.72659987e-01 4.17180620e-02 -1.26816595e+00 2.06889123e-01
-1.33091465e-01 -3.94553185e-01 -3.76624465e-01 1.72638774e+00
3.48601103e-01 1.46566123e-01 -4.25000079e-02 1.08007252e+00
9.64619637e-01 6.57950580e-01 4.21961844e-01 -3.96171689e-01
1.50854266e+00 -8.21583211e-01 -1.09597623e+00 -1.18127584e-01
5.37086666e-01 -5.20933032e-01 8.32550347e-01 5.39271355e-01
-8.22883010e-01 -8.60394478e-01 -8.04604411e-01 1.24832712e-01
-3.36617768e-01 5.24457216e-01 8.55329752e-01 8.38690221e-01
-9.70452130e-01 2.59964317e-01 -3.66838157e-01 -3.82882923e-01
7.41407454e-01 8.67765665e-01 -6.74194515e-01 1.81241065e-01
-1.16641772e+00 6.61405981e-01 4.35414687e-02 4.42672282e-01
-2.10611746e-01 -1.99054815e-02 -8.00402343e-01 1.35113925e-01
-1.20006263e-01 -3.47036391e-01 9.25135612e-01 -2.00357080e+00
-1.58023822e+00 1.08653224e+00 -2.87635177e-01 -6.77888421e-03
-8.65085050e-02 -2.04685345e-01 -6.08255744e-01 9.28611532e-02
-4.30000335e-01 7.44579196e-01 8.45520675e-01 -1.16422510e+00
-1.86633170e-01 -8.73621285e-01 -3.36803347e-01 1.10566646e-01
-7.85473585e-01 5.37435412e-01 -4.39738303e-01 -3.72051775e-01
-3.52076292e-01 -9.77320194e-01 5.60749248e-02 -3.58198211e-03
-8.84950757e-02 -5.11719108e-01 1.04938972e+00 -5.31588793e-01
9.86468315e-01 -2.24061179e+00 2.54366428e-01 4.44378197e-01
-1.93205904e-02 4.65183944e-01 -9.64955017e-02 -1.80005133e-01
-3.25220287e-01 4.00919951e-02 2.16487065e-01 -4.72405165e-01
-1.16739243e-01 2.22480278e-02 2.69785523e-01 2.73991078e-01
4.95989382e-01 7.59738207e-01 -2.33399138e-01 -5.57523310e-01
5.21445349e-02 8.80775809e-01 -4.94836211e-01 7.08882511e-01
5.22574596e-02 6.63658202e-01 -5.42906463e-01 7.56719053e-01
7.27359295e-01 -2.17112992e-02 5.73213547e-02 -4.42414492e-01
2.07520768e-01 -6.81409478e-01 -7.88011670e-01 1.34460270e+00
-5.99049449e-01 6.54968262e-01 3.55935067e-01 -1.10300398e+00
1.25990498e+00 6.20588303e-01 5.76453030e-01 -8.27035606e-01
9.30061698e-01 2.88669858e-02 -4.44836952e-02 -8.90747845e-01
1.80391982e-01 -2.52794206e-01 1.47467867e-01 2.07675189e-01
1.19975261e-01 1.73559219e-01 -1.40597224e-01 -2.88974255e-01
5.08558810e-01 3.84752527e-02 2.43162900e-01 3.03535294e-02
9.53864753e-01 -4.44688082e-01 4.64720488e-01 -1.84981167e-01
-2.51618832e-01 -2.75905766e-02 4.20695841e-01 -5.12069702e-01
-6.01955235e-01 -3.03049743e-01 -6.72684684e-02 1.15772951e+00
-1.74094248e-03 -8.09585378e-02 -1.16081274e+00 -4.12361890e-01
-5.06846011e-01 2.62119919e-01 -7.65031278e-01 -2.73745984e-01
-1.78661883e-01 -6.26931965e-01 5.88680685e-01 4.27380860e-01
7.13264644e-01 -1.55604053e+00 -4.88520443e-01 -1.57141805e-01
-2.49449044e-01 -1.31087518e+00 1.44537687e-01 -2.02394091e-02
-4.57224607e-01 -8.20007801e-01 -7.97232568e-01 -9.20527756e-01
7.00634062e-01 -2.69374371e-01 8.30302000e-01 3.19544166e-01
-4.89489079e-01 3.12952757e-01 -4.60796654e-01 -7.92338431e-01
-2.71556735e-01 -2.05973953e-01 1.14672266e-01 4.99238431e-01
7.75583625e-01 -4.81138110e-01 -5.04132569e-01 3.76994103e-01
-7.80584514e-01 -1.74692079e-01 6.25557303e-01 7.70665109e-01
1.87065542e-01 -1.39138922e-01 6.50567055e-01 -5.97781897e-01
7.59768009e-01 -3.80792677e-01 -8.96361694e-02 2.76993513e-01
-5.41534014e-02 -2.52232283e-01 4.29144740e-01 -5.09550512e-01
-1.32351232e+00 7.79033005e-02 -3.59908193e-01 -4.92403597e-01
-5.24825752e-01 2.07225546e-01 -5.69902122e-01 -4.19285685e-01
5.75280547e-01 -3.64296705e-01 3.41845274e-01 -1.16491951e-02
4.04978776e-03 1.15483105e+00 3.52974892e-01 -5.74609160e-01
6.15106337e-02 4.98107523e-02 8.09499398e-02 -1.06108022e+00
-5.37914157e-01 -3.08938920e-01 -5.32626987e-01 -7.19491184e-01
1.04704368e+00 -9.49998319e-01 -1.32841337e+00 6.93626285e-01
-1.39045453e+00 5.55781983e-02 3.80787790e-01 3.75736296e-01
-4.03335571e-01 6.58629090e-02 -5.73954821e-01 -1.23735774e+00
-5.22834241e-01 -1.11221075e+00 1.19442189e+00 5.60616136e-01
-4.70237523e-01 -7.69077182e-01 -8.40237439e-02 4.96333659e-01
4.61455017e-01 4.25464571e-01 7.12627769e-01 -2.75809616e-01
-1.20593021e-02 -1.51413798e-01 -3.61841142e-01 7.03967810e-01
6.05647005e-02 1.29360065e-01 -1.36713648e+00 1.37103096e-01
7.78077096e-02 -9.96879458e-01 3.57433528e-01 7.36697540e-02
1.49364960e+00 -2.76572317e-01 -1.78631604e-01 4.47861046e-01
1.06107187e+00 4.69502628e-01 8.39148223e-01 5.27757518e-02
5.57726264e-01 1.16188693e+00 5.56515932e-01 6.97580695e-01
1.61361042e-02 7.21588433e-01 2.02675939e-01 -4.86980975e-01
3.31615150e-01 3.06996197e-01 4.11937654e-01 5.87574482e-01
-4.93923008e-01 -1.25857636e-01 -6.59700990e-01 9.39753950e-02
-1.60310054e+00 -9.51470673e-01 8.31879228e-02 1.62039900e+00
5.98409295e-01 -5.73674858e-01 8.38004276e-02 3.05152506e-01
6.97447240e-01 -3.05542797e-01 -3.86913121e-01 -7.64032483e-01
-1.00323252e-01 6.47702038e-01 -4.40145433e-01 1.37094095e-01
-1.02677095e+00 9.39569056e-01 4.68340206e+00 6.78758323e-01
-1.46271300e+00 -8.20476748e-03 1.24798524e+00 9.23048109e-02
2.37431705e-01 -7.42978334e-01 -5.11881351e-01 1.04245387e-01
9.07855034e-01 6.19693585e-02 5.11628568e-01 1.08973908e+00
1.71225995e-01 -8.63418207e-02 -8.39432776e-01 1.58431506e+00
2.98568934e-01 -6.88917398e-01 -8.94611031e-02 -8.21288973e-02
4.39539492e-01 -6.56070530e-01 2.23496169e-01 3.62331450e-01
-4.78449076e-01 -1.32967663e+00 1.73367485e-01 6.03754282e-01
8.07906866e-01 -1.01827013e+00 1.07652724e+00 7.93968886e-02
-8.78286958e-01 -3.23707685e-02 -9.87933725e-02 -7.72741660e-02
-3.45345028e-02 2.11634532e-01 -5.18734515e-01 1.54370263e-01
9.41103399e-01 3.65998983e-01 -4.77081686e-01 3.01714659e-01
2.12812826e-01 2.71733761e-01 -1.63932800e-01 -3.97707820e-01
-1.40827999e-01 -4.34585065e-01 -1.72099695e-01 1.03948414e+00
4.82145220e-01 4.82495040e-01 -1.94787428e-01 8.82115722e-01
-2.20718503e-01 5.29304326e-01 -6.60970867e-01 -3.13180566e-01
8.34403746e-03 1.65955007e+00 -3.98189723e-01 -1.15127251e-01
-4.38460350e-01 1.23462963e+00 3.31905335e-01 3.40846032e-01
-9.26944435e-01 -3.38200361e-01 7.12067664e-01 8.50840996e-04
-2.80768216e-01 1.87865570e-01 -1.99643187e-02 -7.31066942e-01
-1.89644061e-02 -1.04306793e+00 2.64647547e-02 -8.93293917e-01
-1.06250691e+00 1.21793556e+00 -1.81969851e-01 -8.62601399e-01
-3.77124012e-01 -8.33057165e-01 -4.71998841e-01 8.36532652e-01
-1.06653500e+00 -1.28236961e+00 -8.78106475e-01 9.10728335e-01
1.44522086e-01 -2.89962709e-01 1.11170816e+00 5.22995472e-01
-8.32744956e-01 7.61712372e-01 -6.06843531e-01 3.12209815e-01
7.35110223e-01 -5.74274600e-01 -4.11982119e-01 3.50349456e-01
-8.45054239e-02 5.98241508e-01 3.34074736e-01 -4.42305207e-02
-1.36935508e+00 -8.35347056e-01 6.49322569e-01 1.10312380e-01
5.96356280e-02 -2.83596963e-01 -7.50444233e-01 3.23069602e-01
4.86176610e-01 1.17276490e-01 1.11222947e+00 1.41886786e-01
-1.15440488e-01 -2.86406785e-01 -1.26194322e+00 4.99153227e-01
6.58540964e-01 -5.27067065e-01 1.25365555e-01 -4.24650088e-02
8.83421581e-03 9.68659967e-02 -9.52555239e-01 5.97439051e-01
7.53871679e-01 -1.26001573e+00 5.38904130e-01 -4.77838993e-01
4.02882189e-01 6.53470606e-02 -1.71062514e-01 -1.15833664e+00
-1.10201836e-02 -5.49463093e-01 1.40180528e-01 1.27828264e+00
1.38477281e-01 -2.73606956e-01 8.46215665e-01 1.00103772e+00
3.69825631e-01 -9.14398491e-01 -5.51231444e-01 -1.97281286e-01
-3.86046767e-01 -3.03929418e-01 6.00948632e-01 9.28301096e-01
1.77931175e-01 7.48595536e-01 -5.17949700e-01 -2.22934768e-01
2.30268463e-01 -2.61344999e-01 9.55068469e-01 -1.40318286e+00
-7.20504671e-02 -3.83147866e-01 -6.68527901e-01 -5.68964958e-01
5.60569704e-01 -4.38833386e-01 -1.18130438e-01 -7.59201109e-01
2.72273093e-01 -3.77392679e-01 -1.16425931e-01 5.87208569e-01
-1.08709082e-01 4.29727525e-01 6.91192150e-02 -3.20076734e-01
-5.91538370e-01 9.88463223e-01 1.14726388e+00 -5.25473384e-03
-1.42502770e-01 -1.77154228e-01 -5.75511754e-01 7.44312167e-01
7.06390560e-01 2.17793677e-02 -3.40945095e-01 -3.34367715e-02
-2.25873519e-04 1.49363503e-01 3.31355989e-01 -1.11283422e+00
9.63084251e-02 1.10091925e-01 6.85091972e-01 9.73096937e-02
6.97048306e-01 -1.06291544e+00 1.98752925e-01 -4.63668555e-02
-1.93375811e-01 1.34224743e-02 3.11662555e-01 5.04417270e-02
-6.84520304e-01 -6.01550862e-02 1.02010775e+00 5.56520075e-02
-8.65950823e-01 4.59196210e-01 -3.52046311e-01 -4.79997963e-01
1.31880713e+00 -1.84831843e-01 1.02252483e-01 -6.31099343e-01
-7.79314637e-01 -1.28822148e-01 1.31455332e-01 5.38381457e-01
7.03938603e-01 -1.54682553e+00 -2.79480934e-01 5.29645205e-01
2.66660333e-01 -2.86904573e-01 4.20810074e-01 9.02891159e-01
-3.01927507e-01 1.05375104e-01 -6.83439434e-01 -4.29034680e-01
-1.79168701e+00 1.81111783e-01 3.33046883e-01 1.19947225e-01
2.98595458e-01 9.34970379e-01 1.56398773e-01 -2.89809167e-01
3.95431489e-01 2.33425513e-01 -5.41599452e-01 1.72541156e-01
7.09976256e-01 2.23918572e-01 -7.81271607e-02 -1.16678667e+00
-3.02824646e-01 6.35244668e-01 5.24275117e-02 5.85306324e-02
1.27952206e+00 5.26813902e-02 -6.19155109e-01 2.07908228e-01
1.55579209e+00 -4.21432823e-01 -4.82875258e-01 4.85993475e-02
5.73530747e-03 -2.67691970e-01 1.12414993e-01 -5.72106063e-01
-1.23415768e+00 1.27199459e+00 1.06618333e+00 -1.19178988e-01
1.69129121e+00 -1.23695873e-01 6.89238250e-01 5.68346143e-01
3.27542216e-01 -1.12741160e+00 4.49662179e-01 2.02954471e-01
1.04115164e+00 -1.32512176e+00 -3.74329478e-01 -4.66671288e-01
-1.07796717e+00 1.31748044e+00 1.05394161e+00 2.89562404e-01
9.07964408e-01 4.56386954e-01 3.17025989e-01 -3.47000122e-01
-5.72617829e-01 -1.17137522e-01 3.71490449e-01 5.72284222e-01
1.00637865e+00 -1.78948089e-01 -1.34292422e-02 1.06244195e+00
3.47229429e-02 3.60968798e-01 -1.54246196e-01 4.63776350e-01
-2.17822671e-01 -1.12130725e+00 -3.85738194e-01 2.44910911e-01
-6.65645838e-01 3.33086878e-01 -8.01653147e-01 5.26856422e-01
2.71868736e-01 9.93101895e-01 2.07791030e-01 -5.04197776e-01
3.59677821e-01 1.98469788e-01 5.25798142e-01 -1.74318805e-01
-6.99064016e-01 6.06123693e-02 4.11854424e-02 -6.26643896e-01
-1.08686280e+00 -3.30256999e-01 -1.13214207e+00 -2.11382583e-01
-2.42540702e-01 1.71978340e-01 7.01931775e-01 8.22668254e-01
5.29633939e-01 2.14372560e-01 9.21816289e-01 -9.51692760e-01
4.91209887e-02 -1.15532732e+00 -6.45874977e-01 7.67329335e-01
-6.65246323e-02 -7.81859577e-01 -2.09199175e-01 2.62200147e-01] | [13.562982559204102, 1.8475781679153442] |
795b47dc-4643-4152-bf45-0d8feccd21e1 | read-listen-and-see-leveraging-multimodal | 2105.12306 | null | https://arxiv.org/abs/2105.12306v1 | https://arxiv.org/pdf/2105.12306v1.pdf | Read, Listen, and See: Leveraging Multimodal Information Helps Chinese Spell Checking | Chinese Spell Checking (CSC) aims to detect and correct erroneous characters for user-generated text in the Chinese language. Most of the Chinese spelling errors are misused semantically, phonetically or graphically similar characters. Previous attempts noticed this phenomenon and try to use the similarity for this task. However, these methods use either heuristics or handcrafted confusion sets to predict the correct character. In this paper, we propose a Chinese spell checker called ReaLiSe, by directly leveraging the multimodal information of the Chinese characters. The ReaLiSe model tackles the CSC task by (1) capturing the semantic, phonetic and graphic information of the input characters, and (2) selectively mixing the information in these modalities to predict the correct output. Experiments on the SIGHAN benchmarks show that the proposed model outperforms strong baselines by a large margin. | ['Xian-Ling Mao', 'Heyan Huang', 'Yunbo Cao', 'Zizhen Wang', 'Chao Li', 'Qingyu Zhou', 'Zhongli Li', 'Heng-Da Xu'] | 2021-05-26 | null | https://aclanthology.org/2021.findings-acl.64 | https://aclanthology.org/2021.findings-acl.64.pdf | findings-acl-2021-8 | ['chinese-spell-checking'] | ['natural-language-processing'] | [ 6.20556235e-01 -4.55984950e-01 1.98837504e-01 -1.02791011e-01
-1.12532449e+00 -8.84150445e-01 8.23642671e-01 1.70062721e-01
-4.09968823e-01 4.78265464e-01 1.97846755e-01 -4.77729797e-01
5.01108766e-01 -2.69669980e-01 -4.15831178e-01 -7.65600383e-01
5.41289032e-01 3.26724201e-01 3.53984177e-01 -4.09361795e-02
8.73667181e-01 1.79900631e-01 -1.54288542e+00 9.60655570e-01
1.47055304e+00 8.87831211e-01 5.55236340e-01 1.19399524e+00
-5.28004289e-01 9.94683862e-01 -7.99219012e-01 -6.49057090e-01
-9.17846039e-02 -9.19885695e-01 -8.10604155e-01 7.68305548e-03
4.34749842e-01 1.92464307e-01 2.47967377e-01 1.46333742e+00
4.52937484e-01 -5.03944010e-02 5.96302330e-01 -9.82250154e-01
-5.02215505e-01 7.11034238e-01 -2.50529736e-01 -5.13015985e-02
7.14273632e-01 -2.80059222e-03 7.93671727e-01 -1.11311972e+00
3.00427556e-01 1.04140484e+00 6.61887050e-01 8.08406889e-01
-6.11788571e-01 -6.86756551e-01 9.44286361e-02 3.78284186e-01
-1.59382164e+00 -5.04672587e-01 4.74166006e-01 -2.60154307e-01
8.70674670e-01 7.61210561e-01 2.46324971e-01 1.18838477e+00
-7.87893087e-02 1.27316988e+00 1.17950416e+00 -8.92926574e-01
1.71188146e-01 2.67769873e-01 1.56799674e-01 3.27953517e-01
-1.64157987e-01 -2.44352430e-01 -3.59694660e-01 4.26965766e-02
5.90367541e-02 -1.69836387e-01 -4.35480624e-01 2.82454401e-01
-1.22806859e+00 3.33034188e-01 -1.88292518e-01 4.11721349e-01
6.32985309e-02 -1.85250655e-01 4.07128811e-01 2.45620146e-01
1.65980935e-01 7.52657533e-01 -4.66552883e-01 -6.46251559e-01
-1.03292811e+00 1.67735890e-01 7.54941523e-01 1.48435271e+00
3.57488692e-01 -3.09731126e-01 -3.57053339e-01 8.36693287e-01
3.08548391e-01 7.91232467e-01 9.25595820e-01 -5.59916973e-01
8.62295628e-01 5.21362484e-01 2.49607354e-01 -9.80474770e-01
-1.37442470e-01 -7.36401901e-02 -5.91983497e-01 -7.82643929e-02
6.25668764e-01 -1.97146401e-01 -9.99486208e-01 1.09210336e+00
-1.02091551e-01 1.72923699e-01 1.42743036e-01 8.17000985e-01
5.56129098e-01 6.25025153e-01 4.03121524e-02 2.19512403e-01
1.19218242e+00 -1.18901050e+00 -7.94110179e-01 -3.22708309e-01
8.45731020e-01 -1.31827760e+00 1.11571944e+00 6.73576653e-01
-9.48957443e-01 -5.47469676e-01 -1.01213753e+00 -2.36578658e-02
-5.71935475e-01 6.72304153e-01 -8.89761224e-02 1.12044036e+00
-7.59575188e-01 6.57301843e-01 -6.60668492e-01 -3.04169711e-02
1.53937200e-02 -3.91513444e-02 -9.78734866e-02 -1.70328096e-01
-1.02417076e+00 6.84165418e-01 5.89104652e-01 2.23751783e-01
-2.84743667e-01 -3.34329486e-01 -9.67519939e-01 2.03661043e-02
3.98397744e-01 3.56547534e-02 1.38088059e+00 -1.44842815e+00
-1.66509974e+00 6.28208399e-01 -4.07828480e-01 -1.90739796e-01
8.27484310e-01 -5.66402495e-01 -8.50842834e-01 1.08199857e-01
-2.99579501e-02 3.81885439e-01 7.76580632e-01 -1.16566348e+00
-1.15449488e+00 -1.29292384e-01 -5.90384305e-01 2.95485795e-01
-7.08543360e-02 2.87558526e-01 -1.05728364e+00 -8.53393435e-01
7.08226413e-02 -1.10107255e+00 -3.89275476e-02 -5.72176993e-01
-9.09412444e-01 -1.39001012e-01 8.02990794e-01 -9.47000206e-01
1.52350640e+00 -2.39458394e+00 -6.57622516e-02 4.45767850e-01
-1.76588848e-01 8.07648063e-01 -2.24033743e-01 4.70332921e-01
2.20919356e-01 1.39466912e-01 -3.48786920e-01 -7.37437904e-01
-7.84472004e-02 -2.02465683e-01 -3.02239686e-01 3.55470151e-01
2.54153758e-01 7.27052331e-01 -1.11729610e+00 -6.30143821e-01
1.77525476e-01 1.57505974e-01 -3.22540551e-01 1.61098897e-01
-1.98734358e-01 1.55794308e-01 -2.52349913e-01 6.08608305e-01
9.65436757e-01 5.86005040e-02 3.83903652e-01 2.58566856e-01
-3.09768945e-01 4.92380977e-01 -1.38296711e+00 1.39237809e+00
-2.49019757e-01 8.07137012e-01 -3.48124206e-01 -4.11278278e-01
7.08166540e-01 2.59485930e-01 -2.46030837e-01 -1.01582825e+00
2.76506878e-02 5.12838721e-01 -1.37590066e-01 -4.92157131e-01
8.61604929e-01 3.99467826e-01 -2.28353709e-01 2.10691482e-01
-4.65996027e-01 -1.03418037e-01 1.62559777e-01 2.21666813e-01
7.92012811e-01 3.05125087e-01 3.05853009e-01 4.37634625e-02
9.92765069e-01 8.56725648e-02 4.19453532e-01 1.08366966e+00
-1.98441401e-01 1.12674439e+00 3.99974674e-01 6.39340952e-02
-9.10211623e-01 -7.12756097e-01 2.61963785e-01 8.41286182e-01
2.74966598e-01 -7.49725699e-01 -1.13122022e+00 -8.93200576e-01
-2.52314627e-01 9.03433323e-01 -4.09297764e-01 -5.74017875e-02
-5.53834736e-01 -2.37517774e-01 1.10289228e+00 7.05849648e-01
4.90669429e-01 -1.00409186e+00 -4.32010710e-01 1.55902892e-01
-3.59787405e-01 -1.26171780e+00 -7.51786113e-01 2.76780799e-02
-2.69229859e-01 -1.02787113e+00 -8.28027010e-01 -9.47790802e-01
5.10999680e-01 2.91020900e-01 7.20771670e-01 5.57363451e-01
8.82599056e-02 -2.44447291e-02 -8.82290483e-01 -4.77197886e-01
-7.71781564e-01 1.70173943e-01 -2.42726296e-01 2.81229526e-01
5.33797026e-01 4.11779732e-01 -1.71597853e-01 4.55773383e-01
-7.68499196e-01 3.73340279e-01 5.75787246e-01 6.05949104e-01
3.81405681e-01 5.57497032e-02 5.56481518e-02 -1.15913713e+00
4.40728098e-01 -2.16980740e-01 -4.51065421e-01 5.29644668e-01
-5.27853131e-01 9.45736170e-02 8.65791798e-01 -4.17489976e-01
-1.30351388e+00 3.26441586e-01 -3.22320431e-01 1.02113478e-01
-4.99548346e-01 4.80965823e-01 -4.83629465e-01 1.22312829e-01
4.33962435e-01 6.23408675e-01 -4.40305471e-01 -7.41259217e-01
1.26334816e-01 1.26182759e+00 8.02045941e-01 -4.10762727e-01
4.97674465e-01 2.01350629e-01 -4.59363252e-01 -9.11538243e-01
-5.75607121e-01 -7.62139559e-01 -6.54400885e-01 -1.21198550e-01
7.97148645e-01 -8.02357733e-01 -3.79536986e-01 1.12876046e+00
-1.17412484e+00 -2.03585833e-01 3.88242036e-01 4.27409351e-01
-2.90741742e-01 9.98112321e-01 -6.17284775e-01 -9.70732749e-01
-2.63383295e-02 -1.20602620e+00 1.31732523e+00 2.63712406e-01
-3.61605644e-01 -6.22357011e-01 -9.30365026e-02 1.54850140e-01
2.20133126e-01 -2.49506757e-01 7.04876006e-01 -1.02055657e+00
-3.37733269e-01 -5.15614212e-01 -3.31586957e-01 3.90889347e-01
7.07094744e-02 2.40131870e-01 -9.89498913e-01 1.00829609e-01
-3.29310775e-01 6.43201098e-02 8.00257504e-01 -1.79148749e-01
1.11554074e+00 -1.70956373e-01 -3.52250487e-01 4.86971557e-01
1.16374278e+00 4.89806503e-01 9.83340442e-01 2.56104141e-01
9.40568805e-01 4.65162665e-01 6.03039861e-01 4.93124336e-01
2.78939068e-01 6.88941658e-01 1.20815106e-01 2.68243581e-01
-4.08784077e-02 -4.62313890e-01 4.76431489e-01 7.32391536e-01
3.52866165e-02 -5.13192654e-01 -1.12134576e+00 4.10987586e-01
-1.78257120e+00 -8.56895089e-01 -6.37260973e-01 2.15527844e+00
9.84670818e-01 1.99114457e-01 -2.40837887e-01 4.33616906e-01
8.88808846e-01 -2.69273520e-01 1.01897508e-01 -4.50131863e-01
-5.58239460e-01 2.32086144e-02 5.66409647e-01 3.80631208e-01
-1.28258073e+00 1.28899240e+00 5.56674290e+00 1.16554952e+00
-1.13387084e+00 -3.64176989e-01 4.60596591e-01 2.93724030e-01
-2.81500995e-01 4.11291197e-02 -1.02005279e+00 8.63760889e-01
8.75732303e-01 3.67955476e-01 3.00485730e-01 5.50417960e-01
6.24308586e-02 -2.32901096e-01 -9.21938121e-01 9.71131027e-01
4.29530770e-01 -1.12939882e+00 4.81949858e-02 -3.94079447e-01
9.90484238e-01 -2.62667060e-01 -4.91599403e-02 2.21044660e-01
2.38943592e-01 -1.03723323e+00 1.24240065e+00 7.24712312e-01
6.32440984e-01 -7.25433409e-01 8.02878857e-01 4.03768241e-01
-9.20102835e-01 6.80853426e-02 1.64401103e-02 -4.38645110e-02
-9.67300981e-02 1.07233874e-01 -6.33042395e-01 4.09179688e-01
6.54296696e-01 7.34489202e-01 -1.08322918e+00 1.47852099e+00
-5.94058037e-01 8.51141512e-01 3.21106166e-02 -4.12303507e-01
4.60166872e-01 -1.79274455e-02 3.91169995e-01 1.82631838e+00
4.23686832e-01 -1.45611882e-01 -8.55847895e-02 5.60137212e-01
2.11260632e-01 2.67744184e-01 -1.28930077e-01 -2.84291267e-01
3.60961854e-01 8.82300138e-01 -7.18384624e-01 -5.57180583e-01
-4.63859200e-01 1.70962024e+00 9.01181921e-02 3.22985023e-01
-7.32052982e-01 -8.12554061e-01 5.08692086e-01 -5.86530626e-01
5.78959107e-01 -1.28827810e-01 -5.96170783e-01 -1.18524945e+00
4.35188077e-02 -1.23659909e+00 2.20430985e-01 -7.69690394e-01
-9.68771636e-01 5.76085627e-01 -6.25577390e-01 -1.46993148e+00
-2.32571498e-01 -7.66820312e-01 -5.80525219e-01 1.06365967e+00
-1.57653487e+00 -9.14856076e-01 -2.96879500e-01 4.54140395e-01
8.62291753e-01 -1.59462318e-01 5.98267853e-01 2.81782627e-01
-6.67656958e-01 9.81772006e-01 4.24749136e-01 5.94330966e-01
9.76006925e-01 -1.59647036e+00 7.37911284e-01 1.23190498e+00
1.84517294e-01 6.31949842e-01 7.52716184e-01 -8.95177782e-01
-1.12293589e+00 -9.27944601e-01 1.60796762e+00 -5.35480499e-01
2.75985181e-01 -4.75615561e-01 -8.99279237e-01 2.86818027e-01
1.75197691e-01 -4.04116958e-01 5.90084672e-01 -2.91441888e-01
-4.54683512e-01 4.51709330e-01 -7.38538802e-01 8.26809704e-01
5.76231480e-01 -6.66565597e-01 -4.54529434e-01 -5.51298149e-02
4.27350521e-01 -6.49308145e-01 -9.45080221e-02 -1.66396648e-01
5.84883988e-01 -9.04036045e-01 5.12269139e-01 -6.62784934e-01
6.38773024e-01 -5.64893425e-01 -1.09246612e-01 -1.45785785e+00
-1.50748938e-01 -6.63639665e-01 2.38722473e-01 1.27033103e+00
7.41965353e-01 -3.13790143e-02 5.11793256e-01 6.36408389e-01
-4.86906081e-01 -5.34759425e-02 -6.31409764e-01 -6.89009428e-01
7.90395290e-02 -8.61079335e-01 8.18467379e-01 8.70573342e-01
3.59532446e-01 -9.39056277e-02 -6.27291620e-01 3.16751242e-01
1.23887554e-01 -2.10178152e-01 5.94015002e-01 -8.44598114e-01
-1.65298074e-01 -7.51700521e-01 -2.52118617e-01 -1.21339715e+00
3.03356554e-02 -5.93017638e-01 5.41134715e-01 -1.01004398e+00
8.62700194e-02 -4.28658336e-01 4.46740584e-03 1.43289536e-01
-7.66604781e-01 1.75237119e-01 4.68306690e-01 8.33127350e-02
-9.34399247e-01 4.78043444e-02 8.47634017e-01 -8.65264535e-02
-1.33530676e-01 3.16301942e-01 -6.18787766e-01 7.97274947e-01
6.91338360e-01 -2.08096042e-01 2.04392895e-01 -6.18650079e-01
4.82761830e-01 -2.42762759e-01 1.69618562e-01 -1.21054208e+00
5.62275887e-01 -1.54439330e-01 4.76177156e-01 -7.40675807e-01
-4.68510203e-02 -9.28422630e-01 -2.82206744e-01 2.12939680e-01
-6.17767334e-01 2.80967861e-01 1.15789860e-01 5.74808359e-01
-2.90630728e-01 -6.54056847e-01 7.04378843e-01 -9.93637666e-02
-9.85575974e-01 -3.25177640e-01 -8.67948830e-01 1.81684196e-01
8.46451283e-01 -3.67503166e-01 -2.52786040e-01 -3.28881830e-01
-1.82663083e-01 8.84431321e-03 5.62045932e-01 5.77940345e-01
6.64963067e-01 -1.10485566e+00 -6.42144978e-01 4.05553341e-01
4.66771930e-01 -4.18606937e-01 1.45550221e-01 7.41489708e-01
-9.49978709e-01 4.66380537e-01 1.30665541e-01 -3.29611301e-01
-1.53211296e+00 4.01857406e-01 4.65158790e-01 -2.74664373e-03
-2.74510771e-01 9.14339125e-01 -3.10868740e-01 -3.24486107e-01
4.05266583e-01 -4.02816057e-01 -4.33722198e-01 -1.23176128e-02
1.08803773e+00 4.64117527e-01 3.44289541e-01 -8.23419392e-01
-4.35244381e-01 4.03286397e-01 -2.42217213e-01 -2.99176335e-01
5.83545387e-01 -3.75791788e-01 1.01165920e-01 2.14563712e-01
1.17307115e+00 5.28026760e-01 -1.11792779e+00 -4.03365284e-01
5.21154583e-01 -6.65364206e-01 -2.73862809e-01 -1.05418861e+00
-6.72404408e-01 9.43816960e-01 2.97326773e-01 -3.24453339e-02
9.26825583e-01 -3.49068552e-01 8.40925217e-01 3.22921246e-01
5.19329682e-04 -1.54516745e+00 -2.10955948e-01 9.77745891e-01
3.45027834e-01 -1.33892405e+00 -4.99859720e-01 -4.46496040e-01
-1.26903474e+00 1.33406174e+00 6.19925976e-01 2.91620135e-01
2.85248756e-01 3.28793138e-01 3.64582211e-01 3.86032194e-01
-4.74015772e-01 -2.86576331e-01 6.01186037e-01 7.00236619e-01
5.42909324e-01 1.33483529e-01 -3.27296078e-01 7.84386337e-01
-1.83948189e-01 -2.94909537e-01 6.91793323e-01 1.02655983e+00
-4.63293910e-01 -1.19776559e+00 -6.18619502e-01 2.11836293e-01
-5.71584046e-01 -6.48922920e-01 -8.75635684e-01 3.20606172e-01
2.44858980e-01 1.27376127e+00 7.80792460e-02 -4.37424600e-01
2.57092208e-01 1.56593725e-01 6.11185059e-02 -4.42205459e-01
-9.19997633e-01 2.23026857e-01 3.48681509e-02 -4.47339147e-01
-5.29085333e-03 -1.06591904e+00 -1.21921849e+00 -7.89127201e-02
-3.09197247e-01 1.54762864e-01 6.71499074e-01 1.26333308e+00
2.59315223e-01 6.86889663e-02 5.50301254e-01 -4.36345637e-01
-4.22298938e-01 -8.53896558e-01 -3.43294293e-01 7.44056582e-01
4.57246840e-01 1.70989148e-02 -3.11256677e-01 5.71945459e-02] | [10.929628372192383, 10.82861614227295] |
6b5a0642-45ad-4be5-a4ed-532fdfcf145b | temporal-collaborative-filtering-with | null | null | https://www.cs.cmu.edu/~jgc/publication/PublicationPDF/Temporal_Collaborative_Filtering_With_Bayesian_Probabilidtic_Tensor_Factorization.pdf | https://www.cs.cmu.edu/~jgc/publication/PublicationPDF/Temporal_Collaborative_Filtering_With_Bayesian_Probabilidtic_Tensor_Factorization.pdf | Temporal Collaborative Filtering with Bayesian Probabilistic Tensor Factorization | Real-world relational data are seldom stationary, yet
traditional collaborative filtering algorithms generally
rely on this assumption. Motivated by our sales prediction problem, we propose a factor-based algorithm that
is able to take time into account. By introducing additional factors for time, we formalize this problem as
a tensor factorization with a special constraint on the
time dimension. Further, we provide a fully Bayesian
treatment to avoid tuning parameters and achieve automatic model complexity control. To learn the model
we develop an efficient sampling procedure that is capable of analyzing large-scale data sets. This new algorithm, called Bayesian Probabilistic Tensor Factorization (BPTF), is evaluated on several real-world problems including sales prediction and movie recommendation. Empirical results demonstrate the superiority of
our temporal model. | ['Tzu-Kuo Huang', 'Liang Xiong', 'Jaime G. Carbonell', 'Jeff Schneider', 'Xi Chen'] | 2019-05-04 | null | null | null | null | ['movie-recommendation'] | ['miscellaneous'] | [-2.24392071e-01 -3.98001969e-01 -5.83974540e-01 -4.57742065e-01
-2.95095265e-01 -3.30683410e-01 5.47421217e-01 -8.79174396e-02
-4.05290484e-01 2.34885603e-01 1.80309683e-01 -3.91450435e-01
-7.68948734e-01 -9.21878695e-01 -4.77995902e-01 -5.06496847e-01
-2.47589931e-01 6.55623853e-01 3.53267133e-01 -1.25562504e-01
3.17027509e-01 3.55925590e-01 -1.27599990e+00 2.48011082e-01
9.22360122e-01 8.56524944e-01 1.45376161e-01 3.93512398e-01
-5.63453175e-02 6.55228972e-01 -9.89581868e-02 -7.04115093e-01
3.91441643e-01 1.09150715e-01 -7.18551695e-01 3.70334536e-01
-4.57083173e-02 -2.49442592e-01 -3.98273200e-01 7.76813865e-01
-9.83507410e-02 6.36702180e-01 5.25864601e-01 -1.33456898e+00
-4.49049354e-01 1.04567516e+00 -5.39368451e-01 3.36601228e-01
1.64741054e-01 -3.70988399e-01 1.31714559e+00 -7.05244720e-01
4.73178118e-01 1.25333118e+00 6.25866115e-01 9.99634787e-02
-1.58296609e+00 -3.04979324e-01 5.84039390e-01 3.14319670e-01
-1.38710308e+00 -7.87499994e-02 8.36143374e-01 -6.29927635e-01
6.87649250e-01 1.98964939e-01 5.62442422e-01 1.02727306e+00
1.33834139e-01 1.23746479e+00 1.15652955e+00 -2.41153568e-01
2.68824309e-01 -3.26887369e-02 4.70919818e-01 3.11762869e-01
1.58699527e-01 1.46123454e-01 -6.88844800e-01 -6.19085193e-01
6.39449477e-01 6.52425170e-01 -6.52065575e-02 -6.47794247e-01
-1.10908091e+00 9.79052365e-01 -9.12507996e-02 2.78676212e-01
-5.08180737e-01 2.95421518e-02 2.83350170e-01 3.31632376e-01
8.81557167e-01 1.37263108e-02 -7.01556802e-01 -1.54058188e-01
-8.17340851e-01 4.20739889e-01 1.01330483e+00 7.79382944e-01
6.24750972e-01 -3.03125262e-01 -1.84072509e-01 8.46509814e-01
4.67890382e-01 4.20934886e-01 2.67301887e-01 -9.91040111e-01
1.65999234e-01 4.25873250e-01 4.88608241e-01 -1.09467518e+00
-3.60500246e-01 -5.45969486e-01 -6.13832772e-01 -4.68806326e-01
5.80930889e-01 2.84020394e-01 -5.61447084e-01 1.56893802e+00
5.05510926e-01 4.71124321e-01 -2.90283769e-01 7.13095903e-01
-1.13652788e-01 4.87583429e-01 -7.35108331e-02 -6.63437724e-01
1.22105277e+00 -8.20143998e-01 -9.10453022e-01 2.37102181e-01
5.68674147e-01 -5.35365999e-01 8.76848042e-01 1.08402252e+00
-7.98079908e-01 -4.09874052e-01 -6.61215365e-01 2.03274697e-01
-1.10590145e-01 2.79940572e-02 1.47784114e+00 9.60957766e-01
-6.00400448e-01 9.97832716e-01 -1.43694448e+00 -2.17390075e-01
-2.93794870e-02 2.97204286e-01 -8.14026408e-03 -2.89495111e-01
-1.05664384e+00 4.84553903e-01 2.41510227e-01 1.58466667e-01
-6.60836875e-01 -6.99778497e-01 -3.96537155e-01 9.26787872e-03
7.43329346e-01 -8.85732353e-01 1.34368575e+00 -6.74822569e-01
-1.51105738e+00 1.18101820e-01 -3.58794034e-01 -5.95389128e-01
3.65317017e-01 -3.38279516e-01 -6.13848150e-01 7.15795532e-03
-1.22815177e-01 -2.61704773e-01 1.18895435e+00 -1.05334044e+00
-5.84999919e-01 -5.31764269e-01 3.95685524e-01 -2.33503401e-01
-6.61871195e-01 2.45505080e-01 -5.37884891e-01 -8.20370078e-01
3.08894277e-01 -1.08309340e+00 -7.08886862e-01 -4.75949496e-01
-9.79225039e-02 -5.03152549e-01 3.30990314e-01 -2.86118180e-01
1.63322747e+00 -2.10782528e+00 3.45613509e-01 4.64288294e-01
3.88273537e-01 -1.98216438e-01 -4.44011129e-02 6.38024330e-01
1.51293650e-01 -1.18139274e-02 2.06906870e-01 -6.65874481e-01
8.92224237e-02 6.43405974e-01 -5.75733900e-01 4.53468174e-01
-4.04265016e-01 5.36185861e-01 -9.31751430e-01 -1.98081613e-01
1.28438964e-01 3.37463945e-01 -8.89625609e-01 -5.84029173e-03
-3.89523357e-01 1.80743322e-01 -6.36790037e-01 3.99757802e-01
7.06877410e-01 -4.44330961e-01 5.49687803e-01 -3.60628366e-01
-1.50703982e-01 1.59811988e-01 -1.54781687e+00 1.69654155e+00
-5.00955582e-01 -2.67824948e-01 -4.04312648e-02 -7.82753587e-01
4.86497611e-01 1.07139446e-01 9.63447928e-01 -7.30905458e-02
2.20643312e-01 -9.95866060e-02 -1.46392062e-02 -3.52483273e-01
8.62309217e-01 -2.16715366e-01 2.05553696e-02 6.19857371e-01
1.81861192e-01 3.20941240e-01 4.84877110e-01 5.91748297e-01
1.04157400e+00 1.26351252e-01 -5.42401895e-02 -2.41970748e-01
8.38870555e-02 -2.31750086e-01 9.00382817e-01 7.93663144e-01
2.49679782e-03 -6.30088225e-02 3.77286404e-01 -5.07135510e-01
-7.72465944e-01 -8.35008681e-01 -1.70815974e-01 1.38675869e+00
-1.35497004e-01 -1.21242011e+00 -3.29666227e-01 -7.34248757e-01
3.95435929e-01 5.57831764e-01 -8.42772365e-01 9.44815949e-02
-6.56476691e-02 -1.04526305e+00 -1.29504904e-01 6.48906887e-01
-2.89571613e-01 -3.29919681e-02 7.87080377e-02 4.12718534e-01
-1.04517184e-01 -1.05204189e+00 -4.75331247e-01 9.94895324e-02
-1.15830493e+00 -8.15033555e-01 -1.09423898e-01 1.32186115e-01
2.26538718e-01 7.48463750e-01 9.31936145e-01 -2.37201154e-01
2.38444284e-01 7.56650388e-01 -6.00667119e-01 -1.76349923e-01
-8.78292918e-02 4.81561832e-02 5.74303746e-01 4.96616840e-01
5.70783257e-01 -9.11940634e-01 -4.44794506e-01 6.13802969e-01
-9.09559429e-01 -1.02769919e-01 3.24731648e-01 7.78468311e-01
6.13711238e-01 5.69157004e-01 2.62423664e-01 -1.09641361e+00
5.64506531e-01 -6.89808488e-01 -7.17004120e-01 2.58384615e-01
-1.12121105e+00 1.68352112e-01 5.03364742e-01 -8.36001575e-01
-1.14494312e+00 9.38000083e-02 3.11513752e-01 -7.49233961e-01
2.04042852e-01 9.24345493e-01 -1.02930404e-01 1.67231739e-01
3.62499595e-01 1.05995890e-02 -2.39244372e-01 -8.61906946e-01
6.75798893e-01 4.23863351e-01 1.39265925e-01 -1.04905796e+00
7.77629495e-01 7.67122805e-01 -2.84857918e-02 -5.34570873e-01
-1.14394975e+00 -7.92652249e-01 -7.80749619e-01 -2.50496000e-01
3.56119514e-01 -8.55516970e-01 -1.04840922e+00 8.69416893e-02
-8.04012299e-01 -1.39882416e-01 -8.16450045e-02 9.30964351e-01
-6.01989150e-01 6.61774397e-01 -8.36357474e-01 -1.27440727e+00
2.61783421e-01 -6.84062541e-01 6.21248782e-01 -2.07920149e-01
3.21685486e-02 -1.08883798e+00 3.59936297e-01 5.89185357e-01
1.44414634e-01 -3.31888258e-01 6.50772095e-01 -7.44789600e-01
-7.58471906e-01 -1.89985082e-01 7.92333297e-03 3.58382046e-01
-1.07539684e-01 1.83033139e-01 -5.73272824e-01 -2.80321121e-01
9.28489715e-02 1.54081851e-01 1.05051374e+00 3.55501175e-01
1.17341745e+00 -1.70391858e-01 -4.92125839e-01 4.66315240e-01
1.19281149e+00 -7.33378902e-02 2.07069099e-01 4.71719913e-02
7.69082189e-01 5.37896037e-01 9.64660823e-01 9.83949780e-01
6.99105263e-01 7.85815835e-01 1.94345072e-01 6.51094794e-01
4.66228098e-01 -5.48349380e-01 2.91838527e-01 1.15153801e+00
-5.99308729e-01 -1.83591053e-01 -6.70435429e-01 4.33197021e-01
-2.36074972e+00 -1.07229352e+00 -2.80141085e-01 2.29619813e+00
5.57503998e-01 1.94094494e-01 5.92182636e-01 3.18854660e-01
3.98587644e-01 -1.69118308e-02 -4.01232302e-01 -3.07729580e-02
1.60277873e-01 -2.84197964e-02 7.39265859e-01 4.46077019e-01
-1.16127086e+00 8.33912194e-01 7.06976700e+00 9.74196494e-01
-7.34312832e-01 2.76586562e-01 7.50592425e-02 -5.32617643e-02
-4.83791679e-01 1.85228929e-01 -9.15654957e-01 3.42868328e-01
1.33466399e+00 -3.15917313e-01 7.40907669e-01 9.88754153e-01
4.38997865e-01 3.45310122e-02 -1.15319455e+00 8.94446790e-01
-1.63280591e-01 -9.80327249e-01 -7.31429383e-02 3.18286180e-01
7.16498554e-01 2.51664594e-02 -1.40065392e-02 3.11274230e-01
7.73519695e-01 -4.11653489e-01 5.57224274e-01 8.75732005e-01
3.84696275e-02 -8.02746534e-01 2.63823837e-01 5.10247350e-01
-1.09477055e+00 -4.08002287e-01 -3.83726090e-01 -3.74833822e-01
3.38802189e-01 1.08911455e+00 -4.61557031e-01 8.89353931e-01
6.12861454e-01 1.13329136e+00 -1.63752392e-01 1.00679922e+00
-4.94270884e-02 8.27584565e-01 -4.90619749e-01 2.06667349e-01
-1.19048409e-01 -5.07407904e-01 3.63171875e-01 7.82445371e-01
2.57991850e-01 1.47295043e-01 5.34513116e-01 4.42424089e-01
2.13424698e-01 2.74828255e-01 -2.89304912e-01 -3.31503808e-01
3.71944904e-01 1.11778152e+00 -8.75126362e-01 -1.37373477e-01
-6.49156272e-01 7.32277930e-01 2.10616916e-01 3.57134879e-01
-7.64444470e-01 2.69518226e-01 7.31818199e-01 9.59868357e-02
6.31704330e-01 -7.34012604e-01 7.34600499e-02 -1.73084617e+00
-7.13891769e-03 -7.08877444e-01 5.80577254e-01 -2.15439051e-01
-1.75771010e+00 3.15180540e-01 1.88089058e-01 -1.13629961e+00
-1.61498711e-01 -6.66705728e-01 6.12030551e-03 5.52042902e-01
-1.14928508e+00 -1.19892502e+00 3.41735035e-01 9.15935040e-01
2.37244219e-01 1.42746195e-01 6.22784317e-01 5.30410409e-01
-6.91488564e-01 2.82216698e-01 4.34630156e-01 -3.57076198e-01
5.82868099e-01 -1.25797141e+00 -1.45176398e-02 8.19909334e-01
2.92776316e-01 1.20596480e+00 8.68180633e-01 -6.80140853e-01
-1.96770048e+00 -7.64708459e-01 9.40502524e-01 -8.08227241e-01
1.34458911e+00 -4.26358670e-01 -7.00736523e-01 9.50881243e-01
-3.06768745e-01 2.63899053e-03 1.25510442e+00 1.11989665e+00
-8.72280657e-01 -2.62335420e-01 -7.61672676e-01 1.50806263e-01
1.19583583e+00 -5.83580434e-01 -5.73958874e-01 6.80605710e-01
8.03598762e-01 9.23253596e-02 -1.53173006e+00 3.18955749e-01
6.88968539e-01 -7.02243090e-01 1.00665426e+00 -9.93872821e-01
1.25059769e-01 -3.77249599e-01 -4.46539968e-01 -1.09942746e+00
-7.84679174e-01 -7.55167425e-01 -6.42044842e-01 1.08346188e+00
2.23867476e-01 -4.63762224e-01 7.73188770e-01 9.36507583e-01
1.25903353e-01 -4.49468613e-01 -7.05609620e-01 -1.17345786e+00
-2.12321132e-01 -1.09422195e+00 5.07721841e-01 1.24102592e+00
3.57716054e-01 4.11662966e-01 -9.09159184e-01 2.87937641e-01
6.94198489e-01 4.09711182e-01 7.84544826e-01 -1.60870099e+00
-9.83847618e-01 -2.19976932e-01 -1.68697044e-01 -1.35880792e+00
1.17410854e-01 -6.45225883e-01 -4.28293526e-01 -9.82570171e-01
3.22933376e-01 -6.00913107e-01 -4.75921839e-01 1.99401170e-01
-2.94671766e-02 -1.30983487e-01 1.20298371e-01 4.32097256e-01
-8.21882725e-01 5.55500388e-01 1.07291353e+00 1.42491162e-01
-2.33054504e-01 5.33373356e-01 -5.78276753e-01 6.29457295e-01
3.50255787e-01 -4.92876649e-01 -6.61805451e-01 -2.17330009e-01
8.04646969e-01 3.73967171e-01 -1.28354698e-01 -3.34374487e-01
2.74918526e-01 -6.57180369e-01 -1.16657525e-01 -7.91488469e-01
3.64185363e-01 -1.04292655e+00 4.66471016e-01 1.71765968e-01
-2.42196769e-01 -1.48050502e-01 -2.88987011e-01 1.10943639e+00
-9.16768610e-02 -9.30583626e-02 2.31869429e-01 2.16678023e-01
-2.94391870e-01 6.82762444e-01 -4.05635357e-01 -6.68836951e-01
7.42810726e-01 2.78357565e-01 5.79214916e-02 -2.98899293e-01
-1.28440678e+00 1.43770203e-01 2.43687004e-01 4.28065419e-01
1.58891156e-01 -1.35934401e+00 -4.00580317e-01 -1.06161818e-01
2.69261718e-01 -5.63364625e-01 5.66781402e-01 1.14173925e+00
2.27801740e-01 3.97892803e-01 3.49629909e-01 -4.20459330e-01
-9.17121232e-01 1.42966425e+00 -1.60876438e-01 -4.89431590e-01
-3.89917016e-01 5.17435193e-01 -1.27086416e-01 -3.19640309e-01
6.92497790e-02 -5.21382093e-01 -2.43645176e-01 2.14649469e-01
7.13160574e-01 4.87159312e-01 5.98051995e-02 -3.35417479e-01
-1.42313808e-01 7.47131854e-02 -3.61242473e-01 -3.37600708e-02
1.48578846e+00 -4.34283674e-01 -2.63523191e-01 7.85048425e-01
6.82204306e-01 1.69558406e-01 -1.08737755e+00 -7.95863092e-01
1.70694962e-01 -9.64042127e-01 3.28249991e-01 -4.39848721e-01
-1.02450168e+00 4.01167452e-01 1.94676578e-01 6.86500728e-01
9.69358325e-01 -1.26931086e-01 6.25769496e-01 5.14753282e-01
7.44941652e-01 -9.63749647e-01 -2.81003803e-01 5.15286565e-01
4.50315952e-01 -1.02154386e+00 6.05655462e-02 -6.56511843e-01
-3.30171913e-01 1.05542576e+00 1.46927372e-01 -1.38446555e-01
1.30335164e+00 1.39669016e-01 -4.93832797e-01 -1.58821754e-02
-1.23657382e+00 -2.05977350e-01 2.49653220e-01 2.58169025e-01
3.44079345e-01 4.19652075e-01 -7.02612460e-01 1.14747787e+00
3.71946283e-02 3.32904220e-01 4.34244961e-01 8.37884665e-01
-2.47284964e-01 -1.59784114e+00 -2.95063287e-01 4.86561507e-01
-5.80595732e-01 7.81750828e-02 6.95051476e-02 2.25569248e-01
-1.82476133e-01 1.08167017e+00 -1.79843903e-01 -8.51083159e-01
1.62693575e-01 5.69901802e-02 6.52228534e-01 -5.96433520e-01
-4.01735902e-01 5.60885847e-01 1.38892904e-01 -8.09013903e-01
-8.14707458e-01 -1.00896287e+00 -5.18209577e-01 -7.17719376e-01
-5.79550564e-01 5.39676428e-01 7.89443195e-01 9.88766372e-01
3.62845182e-01 3.42472017e-01 9.36914206e-01 -6.55411780e-01
-7.90890455e-01 -9.51529384e-01 -1.08865869e+00 2.04458401e-01
-1.85461923e-01 -1.05070078e+00 -4.15453970e-01 6.14193603e-02] | [9.6392822265625, 5.500072002410889] |
08089f43-90ee-4de5-849b-3ae9ee1f5f91 | learning-free-form-deformation-for-3d-face | 2105.14857 | null | https://arxiv.org/abs/2105.14857v2 | https://arxiv.org/pdf/2105.14857v2.pdf | Learning Free-Form Deformation for 3D Face Reconstruction from In-The-Wild Images | The 3D Morphable Model (3DMM), which is a Principal Component Analysis (PCA) based statistical model that represents a 3D face using linear basis functions, has shown promising results for reconstructing 3D faces from single-view in-the-wild images. However, 3DMM has restricted representation power due to the limited number of 3D scans and the global linear basis. To address the limitations of 3DMM, we propose a straightforward learning-based method that reconstructs a 3D face mesh through Free-Form Deformation (FFD) for the first time. FFD is a geometric modeling method that embeds a reference mesh within a parallelepiped grid and deforms the mesh by moving the sparse control points of the grid. As FFD is based on mathematically defined basis functions, it has no limitation in representation power. Thus, we can recover accurate 3D face meshes by estimating appropriate deviation of control points as deformation parameters. Although both 3DMM and FFD are parametric models, it is difficult to predict the effect of the 3DMM parameters on the face shape, while the deformation parameters of FFD are interpretable in terms of their effect on the final shape of the mesh. This practical advantage of FFD allows the resulting mesh and control points to serve as a good starting point for 3D face modeling, in that ordinary users can fine-tune the mesh by using widely available 3D software tools. Experiments on multiple datasets demonstrate how our method successfully estimates the 3D face geometry and facial expressions from 2D face images, achieving comparable performance to the state-of-the-art methods. | ['Seong-Whan Lee', 'Myeong-Seok Oh', 'Harim Jung'] | 2021-05-31 | null | null | null | null | ['3d-face-reconstruction', '3d-face-modeling', 'face-reconstruction'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-2.15591609e-01 2.08997279e-01 9.29341651e-03 -3.03153366e-01
-4.93090034e-01 -4.06574547e-01 4.43816543e-01 -5.66761553e-01
3.21627975e-01 1.92542091e-01 5.42392731e-02 1.06080644e-01
5.79087250e-02 -8.08827460e-01 -6.75271988e-01 -7.14384496e-01
-1.14904260e-02 7.70028949e-01 -1.08677901e-01 -3.10587753e-02
-6.45109592e-03 1.28070807e+00 -1.67321396e+00 1.39062822e-01
3.29573452e-01 8.54783833e-01 -1.49654686e-01 1.45831570e-01
-2.71895558e-01 -2.18875155e-01 -1.76807806e-01 -3.56380194e-01
4.54307616e-01 -4.24719483e-01 -4.46408004e-01 4.07746196e-01
5.94984889e-01 -3.63753766e-01 -1.69466285e-03 7.61752427e-01
3.95493060e-01 -1.89910978e-01 9.96290326e-01 -1.19078588e+00
-3.90335590e-01 -3.33196253e-01 -6.34966135e-01 -6.70249641e-01
5.98004580e-01 -1.65043056e-01 4.11800504e-01 -1.38358879e+00
7.94254243e-01 1.71794450e+00 9.39163268e-01 9.19228911e-01
-1.65593970e+00 -6.26965106e-01 -2.02618226e-01 -5.07798553e-01
-1.62514842e+00 -5.84287822e-01 1.21735168e+00 -8.78993452e-01
4.78685856e-01 4.34436351e-01 8.24373841e-01 6.19908452e-01
3.81251663e-01 2.25382015e-01 1.26084530e+00 -5.29570997e-01
2.19861701e-01 -1.47379667e-01 -3.88722956e-01 1.04964256e+00
-8.44990183e-03 1.57189056e-01 -4.23419595e-01 -7.04449177e-01
1.44325268e+00 -1.59550652e-01 -2.57784307e-01 -6.97085798e-01
-6.83399677e-01 7.32810855e-01 5.90011030e-02 6.53334185e-02
-4.29181457e-01 -1.72832794e-02 -1.43945858e-01 -6.57622591e-02
8.40980709e-01 1.51781887e-01 -4.86094028e-01 4.18023057e-02
-1.14053345e+00 3.99428993e-01 7.85957813e-01 6.69236183e-01
1.07287502e+00 1.73654184e-01 2.51355171e-01 9.80210602e-01
7.49002039e-01 8.15082073e-01 9.98200625e-02 -1.39944565e+00
-2.54801273e-01 7.64209092e-01 -8.83167014e-02 -1.29498053e+00
-2.54697412e-01 2.32303128e-01 -6.55979037e-01 6.71277046e-01
3.41732711e-01 1.62567735e-01 -9.53747153e-01 1.61453629e+00
7.04060555e-01 2.44010344e-01 -3.91523063e-01 8.76491606e-01
8.69607449e-01 4.58724022e-01 -2.31245697e-01 -4.01403874e-01
1.00097513e+00 -1.07420214e-01 -5.51298916e-01 8.29323828e-02
2.52803653e-01 -7.92247415e-01 9.51605678e-01 1.58031598e-01
-1.37215996e+00 -4.98193830e-01 -6.59416020e-01 1.70943081e-01
7.09877983e-02 -2.66525950e-02 4.41803575e-01 5.64447761e-01
-1.21128988e+00 6.48716033e-01 -1.10386491e+00 -2.81171918e-01
5.37565291e-01 5.89652836e-01 -8.19268107e-01 -4.10024077e-02
-5.94315112e-01 9.12766635e-01 -3.42404306e-01 1.99829981e-01
-5.71266830e-01 -9.63366389e-01 -9.53798234e-01 -1.74314335e-01
-1.53548405e-01 -7.28288591e-01 7.35813916e-01 -6.75759673e-01
-1.76919675e+00 1.37132931e+00 -5.79558492e-01 4.97552961e-01
3.25381726e-01 2.38084316e-01 -1.35675475e-01 1.61518589e-01
-1.47828802e-01 6.30427122e-01 1.12970436e+00 -1.64429665e+00
3.87097329e-01 -6.72975183e-01 -2.91519940e-01 -6.20506853e-02
-3.50871012e-02 -3.92615907e-02 -6.51647031e-01 -5.60079694e-01
6.09825969e-01 -1.04676247e+00 -9.11769569e-02 5.91810882e-01
-8.37630555e-02 5.29791368e-03 1.10961175e+00 -7.46220291e-01
6.95023835e-01 -2.15904403e+00 3.95973057e-01 4.81957704e-01
5.30186780e-02 2.16114283e-01 -1.18883140e-01 4.19402421e-01
-7.11522177e-02 2.96175629e-01 -3.87401789e-01 -4.95838642e-01
-1.53407499e-01 3.49018455e-01 3.93383689e-02 6.48756742e-01
3.14236581e-01 7.91637778e-01 -5.80276787e-01 -5.09820044e-01
3.62482309e-01 1.08748937e+00 -7.92438447e-01 2.93554068e-01
-1.11499570e-01 7.71891177e-01 -5.30808508e-01 7.82530010e-01
1.02980614e+00 1.19099669e-01 1.95019513e-01 -4.73429322e-01
-2.01622024e-03 -3.30646932e-01 -1.21680796e+00 1.60309076e+00
-4.24644083e-01 1.61089525e-01 5.83870769e-01 -6.83727264e-01
1.30119312e+00 4.93948013e-01 7.44161606e-01 -1.88789114e-01
2.11069718e-01 3.11121762e-01 -3.80518079e-01 -4.03680265e-01
-1.91612527e-01 -4.07107383e-01 2.62001961e-01 4.05843496e-01
7.94326514e-02 -7.30028570e-01 -4.82510656e-01 -3.61779213e-01
4.70208853e-01 4.47729379e-01 9.80861932e-02 -5.20979226e-01
5.74867249e-01 -3.31488669e-01 6.07142031e-01 -1.35345489e-01
3.27569544e-01 8.82881165e-01 3.99599195e-01 -6.34537518e-01
-9.63902593e-01 -1.06321216e+00 -5.34837663e-01 2.11741969e-01
-1.24845900e-01 -3.39361727e-01 -9.62985396e-01 -4.17084336e-01
3.81981760e-01 3.00545633e-01 -6.98805034e-01 7.11769983e-03
-6.55514538e-01 -6.44615829e-01 2.43396103e-01 2.16280743e-01
1.67852983e-01 -7.87145674e-01 -3.05498004e-01 -8.97962227e-02
8.41000751e-02 -9.19567108e-01 -5.26980698e-01 -4.90509182e-01
-1.12101293e+00 -1.25190651e+00 -4.18009043e-01 -7.70217538e-01
1.26173115e+00 -5.53517826e-02 9.45628941e-01 3.51749241e-01
-2.13749841e-01 7.06902087e-01 7.49679506e-02 -1.86088324e-01
-7.15374172e-01 -5.26072264e-01 2.86259770e-01 3.09128314e-01
1.82719141e-01 -9.05145645e-01 -3.52451414e-01 5.56914270e-01
-7.24891484e-01 4.01923433e-02 3.85591090e-02 6.14929199e-01
1.07847142e+00 1.36593347e-02 1.89876989e-01 -7.43874431e-01
2.94611305e-01 -7.24959299e-02 -6.55412555e-01 -3.79254967e-02
-4.62284148e-01 -8.08063075e-02 2.47112408e-01 -3.87654573e-01
-8.66081953e-01 3.57359767e-01 -2.41452351e-01 -9.36776638e-01
-1.84453204e-01 3.68023515e-01 -4.15282100e-01 -5.11236250e-01
4.76601779e-01 2.07985025e-02 8.04479480e-01 -9.05922532e-01
1.08025312e-01 5.57339311e-01 2.46078074e-01 -7.70974338e-01
9.63407099e-01 7.16432214e-01 5.21672189e-01 -1.08333397e+00
-3.49514186e-01 3.34829241e-02 -1.08130836e+00 -4.66478884e-01
6.79843962e-01 -7.49736786e-01 -8.59913230e-01 5.00451326e-01
-1.09661996e+00 -2.47907013e-01 -1.15121454e-01 2.77879000e-01
-6.39826775e-01 2.83394545e-01 -3.91215533e-01 -8.06752384e-01
-1.93414643e-01 -1.23376572e+00 1.42746532e+00 -2.08250269e-01
-5.49920142e-01 -1.13121140e+00 1.16616841e-02 3.21394861e-01
2.42569044e-01 8.47417474e-01 1.27740312e+00 2.23284125e-01
-3.26211095e-01 -3.52450222e-01 5.05897045e-01 4.68890637e-01
3.88916671e-01 3.66171628e-01 -9.87086356e-01 -3.76722515e-01
2.41840422e-01 4.73464094e-03 2.60435641e-01 5.71744084e-01
1.06827104e+00 -1.99047193e-01 -2.99204469e-01 8.35907340e-01
1.32334578e+00 2.89025274e-03 5.66879213e-01 -2.34402597e-01
8.00343513e-01 6.62850499e-01 2.45647788e-01 3.47882956e-01
2.28615060e-01 9.60613310e-01 4.22855467e-01 -7.50690475e-02
-3.68253827e-01 -5.09281158e-01 3.38704139e-01 8.42315495e-01
-4.15025890e-01 5.41457176e-01 -8.43672276e-01 8.21046159e-02
-1.42276728e+00 -7.50588119e-01 -8.03195685e-02 2.48366070e+00
8.34932446e-01 -3.67266774e-01 8.10665488e-02 6.58358410e-02
6.13920629e-01 -1.39582857e-01 -4.35024291e-01 -4.34274733e-01
1.49832085e-01 4.21917647e-01 -7.54561126e-02 7.48439848e-01
-6.45120978e-01 6.89748049e-01 6.84659576e+00 5.89630008e-01
-1.40846980e+00 -1.31968305e-01 4.43738312e-01 4.39309366e-02
-4.97900873e-01 -4.99805771e-02 -8.42512250e-01 3.80973786e-01
5.68392932e-01 3.96311209e-02 5.94941318e-01 8.67381513e-01
5.80982983e-01 1.79941192e-01 -1.11551964e+00 1.06639874e+00
1.20058939e-01 -1.36366868e+00 3.18584651e-01 4.10565615e-01
7.38366961e-01 -5.31143248e-01 -1.06027380e-01 -2.48729423e-01
-2.82766372e-01 -1.20133042e+00 6.89589143e-01 6.20699942e-01
1.32186377e+00 -6.22277141e-01 3.60159844e-01 4.82820392e-01
-9.70603287e-01 5.40355384e-01 -3.67668241e-01 2.13508327e-02
1.57603726e-01 5.64190447e-01 -7.16672301e-01 2.52472997e-01
6.15522861e-01 4.79070753e-01 -2.53899157e-01 5.16372085e-01
-1.16231863e-03 5.32667994e-01 -3.81747484e-01 5.74163079e-01
-4.55543786e-01 -6.57049716e-01 5.72617650e-01 7.31224120e-01
5.64546347e-01 3.17705244e-01 1.64581388e-01 1.15613329e+00
-5.45070320e-02 1.00358360e-01 -6.91180766e-01 6.33013472e-02
5.59313178e-01 1.27295387e+00 -5.66685140e-01 3.09515953e-01
-3.24373513e-01 5.32681108e-01 1.40172720e-01 2.91768789e-01
-3.75033557e-01 3.83538604e-01 9.46714163e-01 8.92986298e-01
2.98061460e-01 -5.34463882e-01 -1.65031657e-01 -1.00426733e+00
-2.69024614e-02 -8.15438628e-01 -1.79308772e-01 -7.75877237e-01
-1.33462441e+00 6.01148903e-01 1.04143843e-01 -1.13272846e+00
-3.00802350e-01 -7.20280647e-01 -4.59511250e-01 1.05954242e+00
-1.17382061e+00 -1.49074972e+00 -2.45921940e-01 6.10174656e-01
6.20654225e-02 -5.41255772e-02 1.37413347e+00 -3.19044106e-02
-2.56431341e-01 4.82599169e-01 -1.86880276e-01 4.79117618e-04
4.68428701e-01 -7.32668400e-01 2.96141088e-01 3.47287834e-01
9.80293751e-02 7.74361312e-01 4.49726492e-01 -7.75769353e-01
-1.84828877e+00 -8.71575594e-01 6.48526669e-01 -7.17594028e-01
-4.93857563e-02 -5.05362868e-01 -1.08190000e+00 6.12116754e-01
-4.13841218e-01 3.89889985e-01 8.20133746e-01 -7.96652287e-02
-2.41297841e-01 -2.99248588e-03 -1.58565211e+00 3.42379600e-01
9.71320271e-01 -5.07126987e-01 -3.87394369e-01 6.86446652e-02
9.59028527e-02 -5.91642737e-01 -1.28236628e+00 5.38821399e-01
8.36645305e-01 -9.31117296e-01 1.01209986e+00 -3.94598037e-01
1.11212380e-01 -4.01152790e-01 -2.11782500e-01 -1.34811223e+00
-3.51753920e-01 -7.22454965e-01 -2.80841827e-01 1.29676652e+00
3.41688655e-02 -4.93113726e-01 9.72410977e-01 1.03511405e+00
2.69137379e-02 -1.01252496e+00 -1.15465832e+00 -6.84331715e-01
2.10501790e-01 -3.38750452e-01 8.54589522e-01 1.03797507e+00
-3.75637621e-01 -1.81222931e-01 -1.59115732e-01 2.54889250e-01
6.28570318e-01 2.64264345e-01 8.69218826e-01 -1.66919625e+00
2.23716442e-02 -9.98515934e-02 -4.71826017e-01 -7.97246218e-01
6.67621195e-01 -1.04832327e+00 -2.90343791e-01 -1.16261399e+00
-7.79391378e-02 -6.44332588e-01 4.19917732e-01 4.98963058e-01
2.38717020e-01 3.43902409e-01 2.98220851e-02 2.89106458e-01
5.36206841e-01 5.63369453e-01 1.61112416e+00 1.96551740e-01
-2.79612988e-01 -1.05801657e-01 -3.83873999e-01 1.06947780e+00
3.45990717e-01 -2.54121751e-01 -2.47550488e-01 -3.56513321e-01
-2.53527164e-01 1.49236903e-01 3.70379210e-01 -4.77961272e-01
-1.82408646e-01 -2.44597793e-01 6.64178789e-01 -3.97811294e-01
7.35989869e-01 -1.04558468e+00 8.79027605e-01 1.74448222e-01
1.64035499e-01 -1.03440054e-01 2.69174367e-01 1.58265933e-01
-8.16298127e-02 -1.21429481e-01 1.04940617e+00 -2.46463418e-01
-2.03706950e-01 6.60859406e-01 -1.59486502e-01 -1.65309623e-01
8.05056214e-01 -5.70309341e-01 3.88279051e-01 -2.87833631e-01
-6.69672012e-01 -3.64731938e-01 1.13447392e+00 2.22722501e-01
8.48901093e-01 -1.72967470e+00 -7.17741013e-01 1.00042820e+00
-2.07542762e-01 2.29964867e-01 1.04031198e-01 4.70976502e-01
-5.97270191e-01 6.89413846e-02 -2.44019940e-01 -6.66778207e-01
-1.39779174e+00 2.91240960e-01 4.97705847e-01 3.73467743e-01
-5.42117238e-01 6.80953085e-01 1.06929570e-01 -7.31880784e-01
-1.70503095e-01 -1.62181333e-02 1.58734340e-02 -6.66763708e-02
3.13429505e-01 2.48114869e-01 1.24389179e-01 -1.18433201e+00
-4.55341727e-01 1.39596999e+00 4.70891565e-01 3.45590003e-02
1.48381078e+00 1.66793287e-01 -5.41510344e-01 2.95206368e-01
1.34763491e+00 2.59078145e-01 -1.36771250e+00 1.09858133e-01
-4.73325789e-01 -8.19160402e-01 2.57609248e-01 -3.70050162e-01
-1.33915007e+00 7.73185611e-01 2.42015183e-01 -1.74336731e-01
1.04192770e+00 7.87032396e-02 6.07319117e-01 -2.78540760e-01
6.02360189e-01 -6.16221666e-01 -1.10170461e-01 2.01983452e-01
1.33630085e+00 -8.16171169e-01 1.43439978e-01 -9.58471179e-01
-5.05145900e-02 1.36041498e+00 3.78071547e-01 -9.23595279e-02
1.16916502e+00 3.81371796e-01 1.63485125e-01 -4.55621272e-01
-2.60954946e-01 4.37002093e-01 6.00022972e-01 6.85341835e-01
3.95357847e-01 -5.30842282e-02 -1.24884568e-01 3.85156006e-01
-2.78076798e-01 5.30128032e-02 8.48141611e-02 7.40683138e-01
-2.61960216e-02 -1.48028469e+00 -7.02252328e-01 2.35686004e-01
-2.29664877e-01 3.93422753e-01 -3.21433127e-01 8.09560001e-01
1.77744687e-01 5.72227716e-01 1.32064894e-01 -3.52816671e-01
5.17276347e-01 4.13230926e-01 9.26735342e-01 -6.99814141e-01
-9.81074423e-02 3.20577741e-01 -2.40678132e-01 -7.04582274e-01
-5.19576132e-01 -8.42563987e-01 -1.21273196e+00 -3.96227598e-01
-1.76498100e-01 6.44332496e-03 7.36366749e-01 6.97502553e-01
7.04792976e-01 -2.96342582e-01 7.82590926e-01 -1.43086541e+00
-2.74358183e-01 -6.60997689e-01 -8.71955991e-01 4.98971313e-01
1.28443226e-01 -1.05877364e+00 -5.25620401e-01 3.50396842e-01] | [13.101070404052734, 0.006868502125144005] |
09d9fa24-527c-4347-a04f-00b418bb54e2 | quaternion-neural-networks-for-multi-channel | 2005.08566 | null | https://arxiv.org/abs/2005.08566v2 | https://arxiv.org/pdf/2005.08566v2.pdf | Quaternion Neural Networks for Multi-channel Distant Speech Recognition | Despite the significant progress in automatic speech recognition (ASR), distant ASR remains challenging due to noise and reverberation. A common approach to mitigate this issue consists of equipping the recording devices with multiple microphones that capture the acoustic scene from different perspectives. These multi-channel audio recordings contain specific internal relations between each signal. In this paper, we propose to capture these inter- and intra- structural dependencies with quaternion neural networks, which can jointly process multiple signals as whole quaternion entities. The quaternion algebra replaces the standard dot product with the Hamilton one, thus offering a simple and elegant way to model dependencies between elements. The quaternion layers are then coupled with a recurrent neural network, which can learn long-term dependencies in the time domain. We show that a quaternion long-short term memory neural network (QLSTM), trained on the concatenated multi-channel speech signals, outperforms equivalent real-valued LSTM on two different tasks of multi-channel distant speech recognition. | ['Titouan Parcollet', 'Mirco Ravanelli', 'Nicholas Lane', 'Mohamed Morchid', 'Xinchi Qiu'] | 2020-05-18 | null | null | null | null | ['distant-speech-recognition'] | ['speech'] | [-4.60805409e-02 -2.40175929e-02 3.75001281e-01 -1.34582043e-01
-9.10331190e-01 -2.94077903e-01 4.78312343e-01 -1.53768688e-01
-7.35641956e-01 4.02405888e-01 3.69558096e-01 -3.11828732e-01
3.48784715e-01 -5.39160252e-01 -9.68371928e-01 -6.47888541e-01
-1.44409612e-01 2.61519309e-02 -1.86827764e-01 -3.69533122e-01
-1.80562600e-01 4.24149573e-01 -1.12341571e+00 1.67473644e-01
2.82283902e-01 9.89640653e-01 3.19566488e-01 1.16865659e+00
2.21000552e-01 1.08933163e+00 -1.02164996e+00 -5.58772087e-01
-1.65189758e-01 -3.47546577e-01 -6.05079532e-01 -1.85896046e-02
6.80408850e-02 -2.37416208e-01 -6.21329904e-01 1.03397548e+00
6.79700315e-01 1.36320814e-01 3.54289562e-01 -8.61590087e-01
-7.55455554e-01 8.01171720e-01 -4.28025611e-02 2.80026793e-01
3.55735689e-01 -4.42764968e-01 1.33845735e+00 -1.02442741e+00
1.67309150e-01 1.33409309e+00 6.25843406e-01 1.93615824e-01
-1.06119323e+00 -3.26306969e-01 -2.05188952e-02 5.55837095e-01
-1.51568079e+00 -4.49802428e-01 6.59708440e-01 -9.03832018e-02
1.47016048e+00 3.47498477e-01 7.58729696e-01 1.18333709e+00
3.06493163e-01 8.34513545e-01 7.50698984e-01 -3.44276398e-01
-5.92548735e-02 -6.95320070e-02 -9.74908397e-02 5.27723193e-01
-4.86080289e-01 -1.49277300e-01 -8.92958105e-01 1.22291416e-01
8.76308680e-01 3.83027866e-02 -4.31468129e-01 2.10226536e-01
-1.59100914e+00 6.14502788e-01 5.74495852e-01 6.37384415e-01
-5.49978554e-01 4.01779741e-01 4.22486097e-01 7.07627833e-01
3.87241453e-01 3.87066752e-01 -3.00386518e-01 -2.73775488e-01
-7.16978192e-01 -4.79853600e-01 1.02671087e+00 7.57521808e-01
5.73396146e-01 4.98487979e-01 3.67629051e-01 8.62743855e-01
5.37667394e-01 8.28278363e-01 8.41413796e-01 -4.57592428e-01
5.26391447e-01 -2.71663755e-01 -1.01056404e-01 -1.03233445e+00
-3.95585090e-01 -8.81837130e-01 -1.28207850e+00 -3.54531229e-01
-2.05006134e-02 -2.81993955e-01 -6.95258558e-01 1.67267430e+00
-8.22485387e-02 4.15643096e-01 2.64980763e-01 8.63388836e-01
5.72005272e-01 1.16520274e+00 -8.12752917e-02 -2.77882844e-01
1.16506863e+00 -9.24494982e-01 -1.18049943e+00 -7.08680749e-02
5.11507690e-01 -9.14654791e-01 6.29824281e-01 4.44549799e-01
-1.26091075e+00 -7.06198335e-01 -1.42451751e+00 -2.84504086e-01
-5.35615921e-01 1.14209279e-01 9.11078788e-03 3.17187697e-01
-1.14789283e+00 6.53674722e-01 -1.00849080e+00 1.53801173e-01
-5.60933650e-01 4.39072549e-01 -4.44235444e-01 5.83655834e-01
-1.78983951e+00 1.15249515e+00 1.71888292e-01 8.28667045e-01
-6.09021544e-01 -3.92931610e-01 -8.71262431e-01 1.71811506e-02
-6.62006950e-03 -6.52722955e-01 1.49145746e+00 -8.89164209e-01
-2.11143661e+00 4.16048557e-01 -1.21492177e-01 -6.80809796e-01
1.55882537e-01 -5.26651144e-01 -6.75415635e-01 2.16669831e-02
-4.23958212e-01 2.01556817e-01 1.32446754e+00 -8.39534998e-01
-5.05662560e-01 -2.28189811e-01 -1.26052871e-01 4.48904127e-01
-3.77013624e-01 1.23063192e-01 -3.72798800e-01 -9.37748969e-01
4.15572315e-01 -9.93229568e-01 -9.93890688e-02 -7.04109669e-01
-5.18710077e-01 -9.42550004e-02 3.50138575e-01 -9.58470702e-01
1.28018606e+00 -2.13394356e+00 7.17327476e-01 8.00428838e-02
-1.54288888e-01 3.66041839e-01 -2.14094803e-01 6.89320326e-01
-3.19094598e-01 -1.13173246e-01 1.40479892e-01 -8.00947368e-01
1.99570060e-01 3.16308379e-01 -8.48851740e-01 6.88922405e-01
3.08931053e-01 8.31064939e-01 -7.67156303e-01 4.78053577e-02
2.53833532e-01 1.03377497e+00 -3.25671621e-02 4.80591506e-01
1.29987627e-01 4.54262048e-01 -2.81903669e-02 1.16942205e-01
4.28090602e-01 -6.85346872e-02 1.33168891e-01 -4.21881258e-01
-1.98074117e-01 8.45056832e-01 -1.34644556e+00 1.65480566e+00
-1.07329369e+00 7.23453760e-01 2.28630498e-01 -1.02099454e+00
8.02188694e-01 8.12693536e-01 1.73176229e-01 -8.35866332e-01
5.03105596e-02 3.82569194e-01 -9.14663151e-02 -2.79177189e-01
4.58885610e-01 -4.98260200e-01 -2.30675898e-02 2.10364878e-01
3.16368282e-01 -2.61655688e-01 -2.76970148e-01 -1.53362542e-01
7.77492642e-01 -2.39050388e-01 3.13370287e-01 3.96328568e-01
6.27650559e-01 -9.50041831e-01 3.39770615e-01 2.77750909e-01
1.60942271e-01 7.94992328e-01 1.56976819e-01 -1.87420040e-01
-8.72113645e-01 -1.18005300e+00 1.49412662e-01 1.01623988e+00
-1.25347048e-01 -4.99665290e-01 -3.31534594e-01 2.26591438e-01
-2.85498679e-01 5.82697511e-01 -2.38697946e-01 -1.43573135e-01
-9.08914864e-01 -5.96818328e-01 7.76849866e-01 5.33470809e-01
3.47497404e-01 -7.85777807e-01 -3.02678853e-01 7.56818891e-01
-3.26986402e-01 -1.29682875e+00 -5.07429957e-01 6.37105346e-01
-6.90217257e-01 -2.79617071e-01 -1.15021563e+00 -7.10802913e-01
7.45120633e-04 8.55154172e-02 1.00715363e+00 -4.50610846e-01
1.62256479e-01 4.30279642e-01 -3.59105766e-01 -3.13536644e-01
-4.90367621e-01 1.19422683e-02 2.39243269e-01 3.99841964e-01
-8.37791190e-02 -7.43114293e-01 -2.85571069e-01 1.16374120e-01
-9.60852921e-01 -8.38983804e-02 6.52728438e-01 1.02664316e+00
5.92409134e-01 -2.81511635e-01 5.40856063e-01 -1.17881984e-01
6.57849014e-01 -3.01222920e-01 -3.53208452e-01 3.30136150e-01
1.79758653e-01 2.39484012e-01 8.78495216e-01 -4.29197639e-01
-7.82458544e-01 -2.70122379e-01 -4.38535154e-01 -4.54488784e-01
2.55459011e-01 7.82316148e-01 -1.90669402e-01 1.70216337e-01
1.92733645e-01 2.35698357e-01 -3.48454088e-01 -5.56324363e-01
6.24844551e-01 9.34165359e-01 4.60035950e-01 6.40914962e-02
6.97804689e-01 1.62250444e-01 -8.33272189e-03 -1.42068410e+00
-8.24894965e-01 -4.53730136e-01 -7.49528885e-01 7.74377445e-03
9.37020481e-01 -1.24946976e+00 -7.53564358e-01 8.19259584e-01
-1.66758347e+00 -1.29601285e-01 -7.25000128e-02 1.03510404e+00
-4.86008406e-01 4.02638286e-01 -9.74246860e-01 -8.62247825e-01
-2.64544040e-01 -1.07373822e+00 1.22345388e+00 -4.55792397e-02
-9.57546458e-02 -1.22757757e+00 9.85216126e-02 4.97058295e-02
2.11993888e-01 -3.64421129e-01 4.48181599e-01 -3.94398034e-01
-7.09949374e-01 -2.22220093e-01 2.58416682e-01 9.71305609e-01
-4.66553215e-03 -6.01172745e-02 -9.86307144e-01 -3.32428962e-01
4.94541466e-01 -2.01135308e-01 7.21073925e-01 1.19476482e-01
6.68096423e-01 -2.17198282e-01 1.94144771e-01 6.81709588e-01
1.05508304e+00 3.27892214e-01 5.10422945e-01 8.91769603e-02
8.92531157e-01 3.17698628e-01 1.61106065e-01 3.60685289e-01
5.87665319e-01 6.88762069e-01 2.86249340e-01 -2.88390338e-01
-4.79593165e-02 -2.37966865e-01 7.92586505e-01 2.15103817e+00
-4.53005768e-02 -3.12986970e-01 -7.37754881e-01 2.96390414e-01
-1.53016531e+00 -9.41189170e-01 9.09019113e-02 2.13841534e+00
6.91480458e-01 -6.02934510e-03 -2.36087188e-01 4.12962288e-01
4.92770344e-01 4.33880150e-01 -2.66132265e-01 -5.40874302e-01
-4.73008245e-01 4.15798873e-01 4.65664655e-01 8.04677367e-01
-1.07786155e+00 7.17966676e-01 5.93726110e+00 6.35993302e-01
-1.74397862e+00 1.10356189e-01 2.61778414e-01 2.31233731e-01
-1.63767353e-01 -3.16354752e-01 -5.17361939e-01 -4.99932133e-02
1.51305473e+00 -3.51296999e-02 5.12779176e-01 3.16260517e-01
1.46688491e-01 2.78148502e-01 -1.29679191e+00 1.30846930e+00
2.34430835e-01 -8.85107815e-01 6.97603673e-02 -1.98391438e-01
2.95753211e-01 3.76499087e-01 4.71228033e-01 1.77289903e-01
-7.39238337e-02 -1.29079521e+00 1.04598582e+00 8.22592318e-01
6.39387250e-01 -5.35006523e-01 5.52926898e-01 3.99180114e-01
-1.33582234e+00 1.85916051e-02 -2.91070491e-01 -2.20422581e-01
4.74253386e-01 5.54350197e-01 -1.00115955e+00 6.82685494e-01
4.98167962e-01 8.57103467e-01 -2.18219534e-01 7.60919631e-01
-3.52769881e-01 6.44094646e-01 -4.37828600e-01 -1.97709411e-01
2.72895694e-01 -4.80417162e-01 5.64591348e-01 1.33983517e+00
5.26061356e-01 -1.31912991e-01 -4.34177518e-01 4.29975718e-01
1.21479109e-02 -1.08534936e-02 -3.84745508e-01 -3.14900219e-01
2.05061659e-01 8.71552706e-01 -2.23426029e-01 -2.98877954e-01
-6.09609187e-01 1.38044333e+00 3.30635399e-01 7.63020813e-01
-7.84469128e-01 -4.69079107e-01 7.11375535e-01 -5.66272676e-01
3.31572324e-01 -8.67544770e-01 8.43051299e-02 -1.54687178e+00
1.47188514e-01 -9.05395329e-01 8.45218915e-03 -8.79294872e-01
-1.17337394e+00 9.29911375e-01 -4.65725660e-01 -1.19643593e+00
-4.70669895e-01 -7.20647097e-01 -1.35532081e-01 1.19124949e+00
-1.64089608e+00 -1.12498438e+00 3.85333121e-01 6.04892790e-01
2.86870986e-01 1.40938044e-01 1.06579459e+00 5.06019831e-01
-5.12098312e-01 4.07993883e-01 3.09971541e-01 1.73309341e-01
5.80672145e-01 -1.47471881e+00 5.47899127e-01 6.62125289e-01
8.13014090e-01 9.15221989e-01 7.01987982e-01 -1.93546742e-01
-1.75184762e+00 -8.82676065e-01 1.23707199e+00 -1.46178573e-01
1.12932289e+00 -6.34769082e-01 -9.31872249e-01 9.83452022e-01
4.13348138e-01 -6.88143000e-02 7.23756552e-01 -1.24459974e-01
-4.17883813e-01 -4.06797856e-01 -4.06467952e-02 7.33280241e-01
6.91344082e-01 -1.42626071e+00 -8.66568863e-01 3.68381798e-01
7.85940647e-01 -5.11063755e-01 -1.14609301e+00 1.59672007e-01
3.99088413e-01 -9.26650226e-01 1.09234047e+00 -4.04943109e-01
-1.13373503e-01 -3.19555312e-01 -6.02804780e-01 -1.68751240e+00
-6.74229860e-02 -1.10052836e+00 -5.51521266e-03 8.44661057e-01
4.94136423e-01 -8.33940566e-01 2.01461941e-01 -1.79341640e-02
-2.50463575e-01 -4.93912637e-01 -1.38258553e+00 -5.91758251e-01
6.61646798e-02 -6.77278697e-01 3.61012757e-01 4.17645812e-01
5.47718145e-02 8.73569012e-01 -8.63983989e-01 5.58037996e-01
2.31590047e-01 -1.51760951e-01 4.17743236e-01 -7.95203924e-01
-5.76983452e-01 -1.85857654e-01 -5.49374938e-01 -1.77723861e+00
3.41595978e-01 -7.10221112e-01 1.80297308e-02 -1.29539216e+00
-5.09058952e-01 9.60292220e-02 -3.06653559e-01 -7.11095259e-02
-1.67553984e-02 1.64067745e-01 3.51013035e-01 8.35499242e-02
-4.34401304e-01 1.07477105e+00 1.16488135e+00 -7.51488209e-02
7.15549514e-02 1.65558711e-01 1.77520663e-01 6.97174728e-01
3.92957568e-01 -1.17493317e-01 -2.33107582e-01 -8.22119057e-01
5.71726263e-01 5.60105979e-01 2.68606722e-01 -1.14999616e+00
5.82192242e-01 4.70853299e-01 2.51034379e-01 -6.95131838e-01
1.06427038e+00 -7.81674981e-01 2.86505103e-01 3.56814802e-01
-3.35957825e-01 4.05824691e-01 -1.59739956e-01 5.45862198e-01
-1.05838466e+00 1.38582706e-01 4.85619456e-01 3.15456614e-02
-3.59471768e-01 2.15336472e-01 -6.56810880e-01 -3.31897259e-01
4.31677878e-01 1.84821248e-01 1.39598414e-01 -8.65668356e-01
-1.11982000e+00 -1.62372366e-01 -2.96089292e-01 4.59056199e-01
5.97195446e-01 -1.34483492e+00 -4.92723644e-01 4.33001071e-01
-4.32203799e-01 -1.79867551e-01 1.79346249e-01 8.06779683e-01
-2.35561073e-01 8.67826581e-01 8.84846151e-02 -5.53032696e-01
-1.16586208e+00 2.51502246e-01 7.47755587e-01 -1.80720001e-01
-2.96884298e-01 1.13501203e+00 1.81814060e-02 -4.54833865e-01
4.77547735e-01 -7.87478387e-01 -2.81931520e-01 3.85689229e-01
2.14997888e-01 3.18436772e-01 4.04479116e-01 -1.17856097e+00
-2.57102877e-01 6.04347765e-01 3.72549534e-01 -7.74698675e-01
1.30500507e+00 -4.22273457e-01 -3.77464205e-01 1.31685984e+00
1.78629386e+00 -8.96789059e-02 -7.69283772e-01 -3.72998863e-01
1.46920504e-02 2.20902890e-01 3.23186107e-02 -3.05682421e-01
-7.70086765e-01 1.46827412e+00 1.84999377e-01 6.52029455e-01
1.04499710e+00 -3.29819322e-01 9.38731790e-01 7.82089829e-01
1.87156796e-01 -1.07917023e+00 3.86418611e-01 1.05004239e+00
1.32444978e+00 -8.79160941e-01 -4.90648091e-01 -1.00174770e-01
-4.60450530e-01 1.52414000e+00 -1.40141873e-02 -1.11167401e-01
9.93895411e-01 2.21153677e-01 4.16241795e-01 2.61783719e-01
-7.80037582e-01 -4.35794778e-02 3.95110041e-01 3.50878775e-01
6.79070950e-01 2.76599467e-01 1.06800370e-01 5.36226571e-01
-6.98444963e-01 -4.62444484e-01 5.59911132e-01 5.13377786e-01
-1.00553624e-01 -1.13763309e+00 -6.38623595e-01 -3.06041956e-01
-4.64154094e-01 -1.77694887e-01 -2.39799306e-01 4.36514676e-01
-1.37547493e-01 1.09735012e+00 1.59648389e-01 -4.44813222e-01
5.96127331e-01 2.03858744e-02 4.89915460e-01 -4.39285547e-01
-5.17717957e-01 4.47339952e-01 1.08684070e-01 -4.70909804e-01
-4.90013540e-01 -4.30778682e-01 -1.30588806e+00 1.69569224e-01
-3.45352143e-01 2.25187153e-01 1.07108688e+00 8.58455718e-01
1.24442808e-01 8.31668913e-01 8.43966067e-01 -9.42329705e-01
-1.17714095e+00 -1.07874346e+00 -9.80424047e-01 2.81661272e-01
1.19571924e+00 -2.01369375e-01 -4.39281493e-01 9.23966244e-02] | [14.865700721740723, 6.187085151672363] |
813180e6-2582-4ff7-b01f-f596259a2b6d | rsvqa-visual-question-answering-for-remote | 2003.07333 | null | https://arxiv.org/abs/2003.07333v2 | https://arxiv.org/pdf/2003.07333v2.pdf | RSVQA: Visual Question Answering for Remote Sensing Data | This paper introduces the task of visual question answering for remote sensing data (RSVQA). Remote sensing images contain a wealth of information which can be useful for a wide range of tasks including land cover classification, object counting or detection. However, most of the available methodologies are task-specific, thus inhibiting generic and easy access to the information contained in remote sensing data. As a consequence, accurate remote sensing product generation still requires expert knowledge. With RSVQA, we propose a system to extract information from remote sensing data that is accessible to every user: we use questions formulated in natural language and use them to interact with the images. With the system, images can be queried to obtain high level information specific to the image content or relational dependencies between objects visible in the images. Using an automatic method introduced in this article, we built two datasets (using low and high resolution data) of image/question/answer triplets. The information required to build the questions and answers is queried from OpenStreetMap (OSM). The datasets can be used to train (when using supervised methods) and evaluate models to solve the RSVQA task. We report the results obtained by applying a model based on Convolutional Neural Networks (CNNs) for the visual part and on a Recurrent Neural Network (RNN) for the natural language part to this task. The model is trained on the two datasets, yielding promising results in both cases. | ['Sylvain Lobry', 'Devis Tuia', 'Jesse Murray', 'Diego Marcos'] | 2020-03-16 | null | null | null | null | ['object-counting'] | ['computer-vision'] | [ 3.36215436e-01 -6.64133579e-02 1.75739303e-01 -4.73060846e-01
-7.10407972e-01 -7.01894224e-01 5.91152310e-01 5.05307615e-01
-4.39792067e-01 4.72558409e-01 -1.22211151e-01 -8.47482264e-01
-2.34584108e-01 -1.49569070e+00 -5.29008806e-01 -4.52768683e-01
-1.43725321e-01 2.94993311e-01 1.89545289e-01 -5.93415976e-01
2.98409481e-02 9.35356140e-01 -2.36250496e+00 6.10256791e-01
7.39116788e-01 1.12931025e+00 7.92026222e-01 7.13824570e-01
-5.81850052e-01 1.09543812e+00 -3.55690569e-01 9.67785642e-02
2.70834267e-01 -1.28845587e-01 -1.15239835e+00 6.52469173e-02
5.83609521e-01 -4.21071768e-01 2.74346352e-01 9.56094444e-01
2.54964828e-01 1.64428636e-01 6.30704701e-01 -9.28258479e-01
-6.95409298e-01 2.19550848e-01 -4.97995093e-02 3.38181674e-01
3.65432382e-01 1.16367429e-01 1.16680443e+00 -9.71100390e-01
6.91015482e-01 1.25741434e+00 1.35665953e-01 2.85948575e-01
-1.08805799e+00 -1.36396289e-01 1.30172327e-01 4.58350956e-01
-1.47578537e+00 -2.35886097e-01 5.13880670e-01 -7.21367538e-01
8.79994869e-01 6.88544452e-01 5.82977533e-01 5.59080601e-01
-3.51368934e-01 6.29362047e-01 1.15035272e+00 -4.64904904e-01
3.01393241e-01 6.19733155e-01 1.82029992e-01 6.24051869e-01
-1.06273532e-01 6.75982982e-02 2.16883216e-02 1.16921268e-01
7.03559577e-01 6.95741251e-02 -2.22821876e-01 -7.19987750e-02
-9.51441109e-01 1.08648467e+00 1.14031136e+00 6.32166088e-01
-6.58056915e-01 -7.92164654e-02 2.03910451e-02 5.10173857e-01
5.72217226e-01 2.70257443e-01 -3.73350054e-01 6.75177157e-01
-8.38089228e-01 1.67860493e-01 7.29036689e-01 5.27280629e-01
1.34317279e+00 -1.95505977e-01 -2.29961067e-01 6.46992981e-01
5.63550174e-01 7.51281798e-01 9.85572562e-02 -6.15006626e-01
5.30226111e-01 9.79469240e-01 1.52671203e-01 -1.34821653e+00
-4.68354344e-01 -9.37155187e-02 -8.62025499e-01 6.73993170e-01
2.66500026e-01 7.04068020e-02 -1.09425318e+00 1.14384067e+00
4.45561379e-01 -6.20738268e-01 1.42418340e-01 9.91386592e-01
1.49677455e+00 1.16615665e+00 3.99292111e-01 2.22787485e-01
1.62866259e+00 -5.56164861e-01 -4.61131781e-01 -3.98270190e-01
3.02716225e-01 -4.46283162e-01 1.26025224e+00 7.72863403e-02
-4.98605490e-01 -8.30899417e-01 -8.13253045e-01 -3.12565893e-01
-1.24698222e+00 5.97065926e-01 9.10572857e-02 2.88764983e-01
-1.33461440e+00 2.44504601e-01 -3.53854448e-01 -6.95061564e-01
3.24194252e-01 8.87946114e-02 -3.43015254e-01 -1.67787783e-02
-1.28849697e+00 1.13541758e+00 5.70434034e-01 5.00040591e-01
-8.75870645e-01 -3.35847944e-01 -9.83804822e-01 3.25758517e-01
2.13335365e-01 -4.82497275e-01 7.63807952e-01 -1.25287974e+00
-9.30508792e-01 1.20181942e+00 1.29839957e-01 -4.16777223e-01
3.77316505e-01 -3.17975767e-02 -2.09921747e-01 4.68697786e-01
3.15610468e-01 9.72438693e-01 8.52531314e-01 -1.39546847e+00
-6.34918869e-01 -6.00276947e-01 4.36766773e-01 1.78737059e-01
-7.11228475e-02 1.61164626e-01 -7.00092316e-02 -3.08254987e-01
3.81748453e-02 -3.99350256e-01 -2.42025986e-01 4.08032984e-01
-1.90895900e-01 -7.48711973e-02 9.67782378e-01 -1.08944929e+00
7.54325509e-01 -2.08432245e+00 -1.30660385e-01 4.53388274e-01
1.14666425e-01 5.75815022e-01 -5.07880270e-01 5.08095264e-01
-1.54408783e-01 2.99606919e-01 -3.34432751e-01 2.11634114e-01
-1.87430784e-01 4.66178775e-01 -4.42579687e-01 3.69395614e-02
5.56219757e-01 9.79369879e-01 -7.42957056e-01 -6.11040294e-01
5.36040604e-01 4.69859421e-01 5.69638722e-02 6.58605814e-01
-6.95147514e-01 5.95986605e-01 -6.04534388e-01 5.27841270e-01
6.92304313e-01 -3.25614840e-01 -7.89092556e-02 -1.66792914e-01
-3.55376333e-01 2.42365636e-02 -9.99163568e-01 1.01073945e+00
-7.08632827e-01 6.25211000e-01 2.15454102e-02 -1.18071663e+00
1.28792024e+00 2.19217241e-01 1.11766525e-01 -9.07952368e-01
-1.29032448e-01 1.74763530e-01 -6.66208744e-01 -1.14731634e+00
5.77188730e-01 -4.32745032e-02 3.68394583e-01 3.46671969e-01
-1.42324477e-01 -2.80503869e-01 3.04589003e-01 -1.09849878e-01
4.97304767e-01 2.11260721e-01 3.98883402e-01 -7.49191269e-02
9.99082744e-01 4.57120687e-01 -2.90622234e-01 7.99429357e-01
2.97177613e-01 4.05804396e-01 1.31160647e-01 -9.55089629e-01
-9.87820923e-01 -8.41526806e-01 -2.74063013e-02 1.31845427e+00
-1.19014464e-01 1.09276213e-01 -4.96234655e-01 -4.37296748e-01
-1.62064075e-01 5.32257318e-01 -6.96198165e-01 5.26068032e-01
-2.40450770e-01 -4.44367945e-01 4.81117591e-02 1.52266383e-01
7.99343169e-01 -1.62318933e+00 -1.14905143e+00 7.53427595e-02
-4.01329130e-01 -1.13160682e+00 5.38569033e-01 -2.58408058e-02
-8.73991191e-01 -1.07686937e+00 -6.24619544e-01 -5.42079806e-01
6.38677061e-01 4.69599754e-01 1.34703481e+00 2.88856536e-01
-4.50572044e-01 4.13601339e-01 -6.21682644e-01 -4.65557456e-01
-5.56626022e-01 3.28316987e-01 -9.84435737e-01 2.15900645e-01
2.79176712e-01 -3.07862192e-01 -4.20905620e-01 1.41908199e-01
-1.41802537e+00 -1.81474146e-02 8.28525484e-01 3.09838921e-01
4.98210669e-01 -2.72196531e-01 2.85305709e-01 -6.51992261e-01
3.30173492e-01 -6.26937926e-01 -9.62972105e-01 5.30890405e-01
-1.16140336e-01 2.49061853e-01 4.29968923e-01 -1.13493279e-02
-9.81917024e-01 3.79437059e-01 -4.65808302e-01 -8.30465034e-02
-7.24330366e-01 1.06478703e+00 -1.47431985e-01 9.22815502e-03
1.09795165e+00 1.36836514e-01 -1.21560536e-01 -6.28965259e-01
5.65565825e-01 9.65793490e-01 2.14379981e-01 -7.04712421e-02
7.66661763e-01 6.13329232e-01 -8.68196189e-02 -1.25948000e+00
-9.83636737e-01 -6.73942387e-01 -5.67228138e-01 -4.21362579e-01
1.11171985e+00 -8.58763993e-01 -5.13894618e-01 5.52941151e-02
-1.16824186e+00 -2.10725203e-01 -4.56565350e-01 1.85993716e-01
-3.34400892e-01 1.10369073e-02 1.09741889e-01 -1.18334341e+00
-6.33262277e-01 -7.92513788e-01 1.25396979e+00 8.01375136e-02
3.36875409e-01 -8.48501921e-01 -1.51117006e-02 3.68990868e-01
8.87678802e-01 4.09932375e-01 1.13286543e+00 -4.10335422e-01
-9.21974003e-01 1.16317108e-01 -8.09232712e-01 4.76241589e-01
1.41612455e-01 -3.24498303e-02 -1.08708680e+00 1.76543742e-01
-1.95796251e-01 -3.59110475e-01 1.01383901e+00 2.53649414e-01
1.04633212e+00 -5.46384394e-01 -5.78088798e-02 4.35875244e-02
1.81382024e+00 -2.79924572e-01 9.17674422e-01 3.95662367e-01
5.74869275e-01 1.15016437e+00 5.64617157e-01 2.58713096e-01
6.69020832e-01 6.74502075e-01 1.04397774e+00 -5.69830894e-01
1.47196680e-01 -2.54544057e-02 1.49464846e-01 -6.15025684e-02
-2.33089492e-01 -1.56823103e-03 -1.25524366e+00 8.52585375e-01
-1.87982905e+00 -1.05641115e+00 -5.50430238e-01 1.92924857e+00
5.54029644e-01 -6.14935338e-01 5.08431606e-02 2.20212936e-01
5.32253027e-01 3.43674660e-01 -2.29590818e-01 -2.88108468e-01
-7.11112991e-02 3.58224660e-01 3.07755142e-01 6.14772558e-01
-1.33487105e+00 9.64157164e-01 5.50365925e+00 5.01066566e-01
-1.27707922e+00 -3.80571596e-02 4.79801714e-01 5.96465409e-01
-3.39644045e-01 -3.58262658e-02 -5.92223585e-01 -2.26214319e-01
9.61696267e-01 2.42465407e-01 3.25871557e-01 7.65648305e-01
5.21717489e-01 -5.36204040e-01 -7.26751745e-01 8.27178001e-01
1.80472657e-02 -1.30116808e+00 5.03633142e-01 -2.29891092e-01
3.46064031e-01 2.94248819e-01 -1.43297473e-02 6.17474541e-02
2.14044735e-01 -1.24328613e+00 6.25103116e-01 9.69445407e-01
6.02735579e-01 -3.85031641e-01 8.08437824e-01 6.59960389e-01
-1.38630772e+00 -2.55555838e-01 -4.11887497e-01 -1.71923265e-02
-1.92045972e-01 4.43831325e-01 -8.83986950e-01 7.61101961e-01
9.44146872e-01 6.31493509e-01 -1.10607374e+00 8.49359095e-01
-3.58274400e-01 2.47147381e-01 -3.86697710e-01 -1.65254474e-01
3.19282353e-01 -3.24596554e-01 3.01804364e-01 9.39931512e-01
3.56400400e-01 7.42557645e-02 7.28551522e-02 1.23978710e+00
2.75528908e-01 5.50234258e-01 -1.01170039e+00 3.15845422e-02
-1.74215764e-01 1.57910752e+00 -5.10013461e-01 -2.85500407e-01
-2.19613552e-01 5.73934734e-01 2.56688207e-01 5.60574651e-01
-3.03800732e-01 -3.52005780e-01 1.63648039e-01 3.63348007e-01
3.66936594e-01 -7.58238435e-02 1.63331643e-01 -1.06552112e+00
1.48774013e-01 -8.80053341e-01 5.42818904e-01 -1.41390204e+00
-1.16963661e+00 8.92886698e-01 1.30365461e-01 -1.20920157e+00
-3.95132124e-01 -7.85800219e-01 -1.86037660e-01 1.08591878e+00
-2.12512755e+00 -1.49021649e+00 -9.24914539e-01 6.42121434e-01
1.96984991e-01 2.23732844e-01 1.02195120e+00 2.83972561e-01
1.40200257e-01 -4.83436882e-01 -2.80024350e-01 3.65250468e-01
4.36242186e-02 -1.11824918e+00 1.89295888e-01 7.14170456e-01
4.09894198e-01 1.00967064e-01 4.37846750e-01 -2.41590649e-01
-9.77100670e-01 -1.53432167e+00 1.20365214e+00 -2.86770165e-01
4.52117711e-01 -5.70785590e-02 -1.10046482e+00 5.74883461e-01
7.72121474e-02 1.51500210e-01 5.28355718e-01 -4.68042880e-01
-3.52813095e-01 -2.61938274e-01 -1.23616993e+00 2.76151061e-01
4.58410710e-01 -7.42829621e-01 -6.38924718e-01 3.80271941e-01
3.23231459e-01 -2.31120884e-02 -5.80347896e-01 3.17540556e-01
2.27168918e-01 -1.05492938e+00 9.64748144e-01 -5.63411891e-01
6.12117648e-01 -5.73451579e-01 -3.41245055e-01 -9.53722060e-01
1.35568380e-01 2.77723998e-01 4.70421940e-01 7.47676849e-01
6.72240555e-01 -4.23120409e-01 2.94138581e-01 2.83971548e-01
4.15077180e-01 -2.51544099e-02 -7.41603494e-01 -3.56493056e-01
-7.91533366e-02 -3.82294178e-01 6.91051722e-01 7.05532372e-01
-5.57643354e-01 5.38620174e-01 -1.15744226e-01 5.26828051e-01
1.45588279e-01 5.48042059e-01 8.42776120e-01 -1.65768278e+00
1.02275290e-01 -2.18736321e-01 -3.94870132e-01 -8.03326309e-01
-7.37671107e-02 -8.24850798e-01 1.21794781e-02 -2.24340129e+00
-1.05081528e-01 -4.27592725e-01 3.94370258e-01 8.84649873e-01
-7.23747462e-02 2.70002067e-01 3.65972638e-01 1.42297968e-01
-2.48593748e-01 4.02773082e-01 9.82115626e-01 -4.69868422e-01
-2.48508140e-01 -2.01572068e-02 -3.55597138e-01 5.22350013e-01
9.38417017e-01 -3.78984153e-01 -1.19136967e-01 -5.11263192e-01
8.67084503e-01 1.32145688e-01 1.01976180e+00 -9.52441573e-01
-3.09179686e-02 -1.65994868e-01 1.00639798e-01 -8.80650401e-01
1.45835191e-01 -1.25561452e+00 2.04312459e-01 3.80056053e-01
-2.98624277e-01 -1.09807365e-01 2.47588053e-01 1.70858353e-01
-4.90332305e-01 -4.25929040e-01 5.10135233e-01 -4.41502601e-01
-8.76030684e-01 2.52260149e-01 -5.57619929e-01 -3.03368449e-01
8.25869620e-01 -8.06606282e-03 -3.60135525e-01 -6.74825013e-01
-1.05929101e+00 3.15121084e-01 2.65925191e-02 4.75507259e-01
7.47411370e-01 -1.07917321e+00 -1.13580489e+00 2.44172305e-01
6.55479789e-01 3.18070836e-02 2.44689256e-01 3.62108052e-01
-8.52861583e-01 3.60086679e-01 -3.37127745e-01 -6.97590351e-01
-1.19072950e+00 5.56887329e-01 6.00487828e-01 -3.23839664e-01
-2.89327562e-01 2.52151400e-01 -1.69814914e-01 -8.15300584e-01
-2.91316479e-01 -4.85358208e-01 -1.11282909e+00 5.40965974e-01
7.33213127e-01 3.39884534e-02 2.30671793e-01 -9.57051158e-01
-2.05794916e-01 6.85521007e-01 5.51559210e-01 -3.24529707e-01
1.70752692e+00 -7.07972869e-02 -3.66653800e-01 4.51662183e-01
1.04962170e+00 -3.36724490e-01 -6.79722905e-01 -4.42753673e-01
1.06986187e-01 -3.71231586e-01 1.20139875e-01 -7.61058867e-01
-1.11782873e+00 1.38758111e+00 9.69698489e-01 8.75241578e-01
1.16236973e+00 3.05386841e-01 7.97909126e-03 7.98293054e-01
2.07036585e-01 -8.17318082e-01 -2.46967405e-01 5.53703487e-01
1.34549940e+00 -1.58618402e+00 -1.76576003e-01 -3.25222045e-01
-4.97323930e-01 1.31858528e+00 1.15163438e-01 1.91622704e-01
7.22415805e-01 -3.04649800e-01 5.37099779e-01 -6.72192156e-01
-4.09420252e-01 -1.17298782e+00 6.11846030e-01 7.94814885e-01
7.32683018e-02 2.01911349e-02 -9.39148739e-02 -2.73438066e-01
4.30785976e-02 1.01303183e-01 3.31239402e-01 7.93883979e-01
-8.89901280e-01 -9.25286710e-01 -6.26103878e-01 3.62825990e-01
-1.83453083e-01 -3.11781406e-01 -5.17408550e-01 6.08791113e-01
1.46456435e-01 1.18359756e+00 1.16237469e-01 4.36225981e-02
3.72537553e-01 -1.31808311e-01 -6.36458695e-02 -7.55835116e-01
-7.76176333e-01 -5.20021439e-01 1.51054680e-01 -4.32706863e-01
-9.96083736e-01 -2.98553467e-01 -8.97678912e-01 9.43178460e-02
-9.90604237e-02 1.37632474e-01 1.01323497e+00 1.05726123e+00
2.87322730e-01 1.83326885e-01 6.39715493e-01 -7.46215224e-01
-2.90485710e-01 -9.31766510e-01 -4.01123166e-01 2.47676626e-01
5.66063762e-01 -7.72853419e-02 -2.80045569e-01 2.47484699e-01] | [9.723637580871582, -1.2558733224868774] |
45599fd9-85d1-45f9-85a8-f148198b6fa1 | multi-view-optimization-of-local-feature | 2003.08348 | null | https://arxiv.org/abs/2003.08348v2 | https://arxiv.org/pdf/2003.08348v2.pdf | Multi-View Optimization of Local Feature Geometry | In this work, we address the problem of refining the geometry of local image features from multiple views without known scene or camera geometry. Current approaches to local feature detection are inherently limited in their keypoint localization accuracy because they only operate on a single view. This limitation has a negative impact on downstream tasks such as Structure-from-Motion, where inaccurate keypoints lead to large errors in triangulation and camera localization. Our proposed method naturally complements the traditional feature extraction and matching paradigm. We first estimate local geometric transformations between tentative matches and then optimize the keypoint locations over multiple views jointly according to a non-linear least squares formulation. Throughout a variety of experiments, we show that our method consistently improves the triangulation and camera localization performance for both hand-crafted and learned local features. | ['Johannes L. Schönberger', 'Marc Pollefeys', 'Mihai Dusmanu'] | 2020-03-18 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/2556_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123460647.pdf | eccv-2020-8 | ['camera-localization'] | ['computer-vision'] | [ 8.95986101e-04 -4.96977806e-01 -1.43794760e-01 -3.77801478e-01
-1.24174654e+00 -1.03193355e+00 5.85381925e-01 1.25087738e-01
-3.93577605e-01 3.68391782e-01 -4.02899571e-02 5.40441871e-02
-1.95585966e-01 -3.59535903e-01 -8.01121771e-01 -4.18296814e-01
2.68508017e-01 2.33657569e-01 3.46316427e-01 1.42071560e-01
7.42393672e-01 9.23490703e-01 -1.27158737e+00 -1.39060125e-01
3.05709511e-01 9.40446496e-01 2.34769076e-01 8.98150682e-01
4.54755276e-01 3.07238519e-01 -1.45747527e-01 -1.29749447e-01
6.42612159e-01 -1.51062995e-01 -6.75268173e-01 3.66831392e-01
1.18662405e+00 -4.21714097e-01 -1.86704293e-01 1.02756429e+00
2.67573595e-01 1.91645950e-01 4.29123551e-01 -1.13205349e+00
-1.67957589e-01 -3.13139200e-01 -8.68729711e-01 1.36366650e-01
7.73513079e-01 -2.20630646e-01 1.05342662e+00 -1.43977523e+00
7.42687166e-01 9.71271396e-01 1.00328434e+00 -1.74290463e-02
-1.21208656e+00 -3.58468950e-01 2.69860089e-01 2.07535606e-02
-1.99726307e+00 -7.57597327e-01 7.40331769e-01 -4.56215769e-01
8.94500613e-01 1.50787264e-01 4.33889747e-01 3.98442417e-01
2.24433467e-01 2.10420296e-01 7.50700533e-01 -5.49530685e-01
-6.11266382e-02 -5.58969267e-02 -3.39891464e-01 1.08473539e+00
7.29298517e-02 1.29695728e-01 -7.81297445e-01 -3.99533838e-01
1.36743259e+00 3.45955998e-01 -2.32399911e-01 -9.18647110e-01
-1.71850896e+00 7.91933417e-01 4.56945866e-01 6.21576533e-02
-2.51020432e-01 4.03537422e-01 -1.44145429e-01 2.21279263e-01
2.86097556e-01 4.86715645e-01 -5.41736960e-01 -6.32984862e-02
-7.56049752e-01 1.47528037e-01 4.24060374e-01 1.32694113e+00
1.28804302e+00 -4.04876530e-01 5.71047843e-01 5.12599409e-01
2.68948317e-01 7.81602502e-01 4.08159830e-02 -1.29967356e+00
5.69124997e-01 5.55069387e-01 3.59461904e-01 -1.49203777e+00
-3.73611301e-01 -2.80576721e-02 -3.72814715e-01 2.49067619e-01
4.06286985e-01 -6.86628819e-02 -7.56682754e-01 1.16474855e+00
5.72664320e-01 2.11578533e-01 -2.59583294e-01 8.27612162e-01
4.10165578e-01 3.37665409e-01 -6.39701664e-01 -6.87979683e-02
8.88118982e-01 -9.81699705e-01 -2.85811454e-01 -3.14061522e-01
4.37575608e-01 -1.30211461e+00 4.93393719e-01 2.77457237e-01
-1.01322734e+00 -4.31485921e-01 -7.72437155e-01 -2.99801320e-01
-1.53760552e-01 1.54233456e-01 3.51642460e-01 1.54781371e-01
-1.18558931e+00 2.89242357e-01 -8.29867423e-01 -4.57517684e-01
-2.94558462e-02 6.28939092e-01 -7.24890769e-01 -1.58295676e-01
-2.73244113e-01 8.34484041e-01 -2.42481586e-02 2.67970383e-01
-5.55141151e-01 -6.44500315e-01 -1.02958047e+00 -1.80704147e-01
5.65626323e-01 -9.16049242e-01 1.24820256e+00 -6.28502846e-01
-1.31441116e+00 8.47041547e-01 -7.09824741e-01 8.25921372e-02
3.67948711e-01 -5.69327712e-01 7.33758584e-02 2.41989568e-01
2.73995072e-01 5.64413249e-01 9.77396607e-01 -1.39329398e+00
-1.03849256e+00 -4.39090729e-01 8.25578421e-02 5.31863928e-01
2.42073253e-01 -7.28520900e-02 -7.53471911e-01 -4.12774235e-01
7.67484248e-01 -9.80390310e-01 -4.55485463e-01 3.46612424e-01
-2.36405984e-01 2.53335685e-01 1.03996050e+00 -2.84590751e-01
8.52019131e-01 -1.92391658e+00 2.89127499e-01 5.88790536e-01
2.62730181e-01 -3.94708604e-01 4.75670546e-02 5.42719960e-01
1.29758567e-01 -1.81911185e-01 2.69549012e-01 -3.25426549e-01
-4.87891674e-01 -9.46218893e-02 -1.66204110e-01 9.79154348e-01
-1.22776367e-01 7.76565075e-01 -9.43406105e-01 -4.65778202e-01
6.86653137e-01 4.64007944e-01 -4.54697341e-01 9.58249643e-02
2.20140040e-01 3.79838824e-01 -5.46174347e-01 8.56702864e-01
6.15801454e-01 -1.87369898e-01 -5.90175129e-02 -4.80229199e-01
-4.31987524e-01 -1.54829249e-01 -1.56736350e+00 2.00182796e+00
-5.55344105e-01 6.01284742e-01 1.07118279e-01 -5.64890325e-01
7.22805500e-01 1.56152695e-01 8.35369766e-01 -2.83819467e-01
-1.03533059e-01 2.30151355e-01 -6.47120893e-01 -1.70248061e-01
6.48096323e-01 -6.13613687e-02 -8.41961056e-02 2.78783381e-01
-6.86860010e-02 -3.24518800e-01 -2.32096151e-01 -1.18460869e-02
1.01470733e+00 1.10395841e-01 8.82339478e-01 -1.08870566e-01
4.85817492e-01 1.67893931e-01 4.50548768e-01 5.85506082e-01
1.30609319e-01 1.05692267e+00 -8.05119611e-03 -5.02471745e-01
-1.19255042e+00 -9.99071896e-01 -1.95422128e-01 7.14607894e-01
5.92785537e-01 -6.56036854e-01 -3.53708595e-01 -7.59848654e-01
1.10660188e-01 -1.01326734e-01 -2.66591936e-01 2.28150025e-01
-7.49268174e-01 1.97644196e-02 2.46300753e-02 5.95824420e-01
9.67817307e-02 -3.64052683e-01 -7.03695476e-01 2.38343887e-02
-6.20675087e-02 -1.37234831e+00 -1.02650583e+00 -1.51726110e-02
-1.04688609e+00 -1.35974956e+00 -5.30416250e-01 -7.96881557e-01
1.22858381e+00 1.05727208e+00 7.52544820e-01 1.25109196e-01
-2.35421240e-01 9.03969765e-01 -2.56531477e-01 -3.30485255e-02
1.72751531e-01 1.47906691e-02 2.08257407e-01 3.78329237e-03
2.05903187e-01 -3.55254292e-01 -6.15346909e-01 6.83013141e-01
-4.87609297e-01 -2.93664355e-02 4.64610666e-01 7.74799585e-01
9.03513670e-01 -7.42513686e-02 -2.13665828e-01 -5.45166492e-01
1.25831917e-01 -9.30757374e-02 -1.02405930e+00 2.91351765e-01
-4.50588465e-01 -9.63211618e-03 3.66043478e-01 -2.53444433e-01
-8.38069260e-01 7.65355289e-01 3.73366266e-01 -6.93509758e-01
-1.08389169e-01 4.06842202e-01 4.94614094e-02 -8.54723573e-01
4.94743884e-01 5.21808714e-02 -2.43070096e-01 -3.82120788e-01
4.41296071e-01 3.12300920e-01 5.34590602e-01 -4.18917030e-01
1.19326556e+00 7.90672421e-01 2.55179673e-01 -9.01862800e-01
-7.04671562e-01 -1.07757437e+00 -1.43779743e+00 -1.42692685e-01
6.11221254e-01 -9.77988601e-01 -6.10441267e-01 2.57769078e-01
-1.17033839e+00 1.25096798e-01 -1.09904101e-02 6.39781415e-01
-7.75361121e-01 5.14427304e-01 -9.97906104e-02 -5.94936609e-01
1.52958920e-02 -1.14188313e+00 1.66526270e+00 1.40994951e-01
-2.65644312e-01 -1.23834729e+00 2.58097351e-01 -1.31761003e-02
-4.13286649e-02 3.38351876e-01 2.95156211e-01 1.63169578e-01
-1.07230842e+00 -4.72875178e-01 -1.94347411e-01 -1.58548787e-01
3.80634338e-01 1.16988257e-01 -8.57877910e-01 -5.44512331e-01
-1.04345731e-01 2.62923981e-03 4.35107559e-01 4.53841209e-01
6.95387542e-01 -4.81922291e-02 -5.86192250e-01 9.55272257e-01
1.74143612e+00 1.32207379e-01 4.25948650e-02 4.29893583e-01
9.65305984e-01 3.14406604e-01 9.07920957e-01 4.97368962e-01
3.89553785e-01 9.09607053e-01 3.10170293e-01 -2.17762545e-01
2.14528650e-01 -5.37732780e-01 2.29898095e-01 6.42515540e-01
-1.80442125e-01 1.49792224e-01 -1.03307927e+00 6.46146774e-01
-2.05235076e+00 -7.96072662e-01 -8.09967965e-02 2.43872976e+00
3.82462174e-01 -3.59097093e-01 -7.43545368e-02 -3.54228646e-01
7.50656009e-01 5.96502535e-02 -4.18457896e-01 7.87848420e-03
1.55261785e-01 -1.08759433e-01 9.34092283e-01 9.65699136e-01
-1.23903978e+00 9.76755559e-01 7.24940109e+00 3.08708489e-01
-1.01976824e+00 -7.05156475e-02 8.12154859e-02 -4.96353023e-02
-1.79311633e-01 3.39800179e-01 -1.03753972e+00 -6.90857098e-02
3.43586743e-01 1.98591836e-02 4.67748910e-01 8.47179830e-01
1.90010458e-01 -4.11794871e-01 -1.26894546e+00 1.40344262e+00
3.11716914e-01 -1.38868558e+00 -1.14889480e-01 3.59109551e-01
9.54926491e-01 1.06975384e-01 -7.16121271e-02 -4.63477463e-01
1.45190567e-01 -8.35830331e-01 7.26850271e-01 5.37199020e-01
9.21847165e-01 -7.54133940e-01 4.62425053e-01 3.45094115e-01
-1.50094295e+00 1.11073991e-02 -3.73168498e-01 -4.18405943e-02
3.05535376e-01 3.44929129e-01 -9.11820412e-01 4.18774426e-01
5.93832076e-01 8.03849161e-01 -5.85733414e-01 1.18044794e+00
-6.23487048e-02 -1.07059874e-01 -6.89045608e-01 5.20657301e-01
2.07988501e-01 -2.48576373e-01 5.59961915e-01 8.52813542e-01
5.18369138e-01 9.53552574e-02 4.38635767e-01 4.94248897e-01
1.44154966e-01 -6.48041964e-02 -1.02931452e+00 5.14551997e-01
6.24707401e-01 1.36519134e+00 -8.36529911e-01 -4.26272377e-02
-7.82280803e-01 1.10585046e+00 3.32733095e-01 4.29642558e-01
-5.68539560e-01 -3.93380463e-01 6.08709872e-01 2.66062856e-01
5.23854733e-01 -8.37421834e-01 -1.29578590e-01 -1.47051454e+00
2.46301264e-01 -5.47391474e-01 2.63777971e-01 -7.42819607e-01
-9.50751781e-01 2.64676690e-01 2.45188847e-01 -1.45706391e+00
-4.58076000e-01 -4.92506295e-01 -4.59199071e-01 7.30947912e-01
-1.30126643e+00 -1.45813584e+00 -4.09931868e-01 8.94769967e-01
6.35173798e-01 1.46331787e-01 6.67777061e-01 9.62791517e-02
-8.29729661e-02 4.48732734e-01 3.29299837e-01 3.62613075e-03
9.42154706e-01 -1.16634572e+00 5.15456975e-01 1.05073774e+00
5.56358635e-01 8.16743374e-01 4.46023047e-01 -6.01622641e-01
-1.76361513e+00 -8.17233026e-01 7.84401178e-01 -8.51199985e-01
5.11085212e-01 -3.46560568e-01 -4.07556355e-01 1.02229619e+00
-1.88086122e-01 4.14508581e-01 2.47175753e-01 1.84334978e-01
-1.80658877e-01 -1.71941072e-01 -1.04059875e+00 4.25203443e-01
1.05871165e+00 -8.87472808e-01 -3.32773060e-01 2.62259603e-01
2.48044923e-01 -7.56892741e-01 -9.01718318e-01 2.29174241e-01
7.78591037e-01 -8.65464628e-01 1.26196432e+00 -1.94891319e-01
-1.88121513e-01 -6.26298368e-01 -3.22137147e-01 -1.14588892e+00
-3.99892002e-01 -7.23126769e-01 6.91001415e-02 9.75094080e-01
2.07440898e-01 -4.18777943e-01 8.80731046e-01 6.79100513e-01
9.08862278e-02 -5.15289783e-01 -9.64902461e-01 -5.76785266e-01
-5.19972503e-01 -1.94339722e-01 2.37693399e-01 8.25021088e-01
-2.06086025e-01 3.31861109e-01 -2.97316551e-01 6.29333794e-01
7.38076210e-01 4.46678102e-01 1.16410005e+00 -1.01897097e+00
-5.18382117e-02 -7.36045986e-02 -7.22016990e-01 -1.48211396e+00
-5.76846339e-02 -4.28983420e-01 2.02968642e-01 -1.10724759e+00
2.38801301e-01 -3.98253471e-01 1.26063049e-01 1.84459105e-01
-1.13165520e-01 3.95879030e-01 7.02663139e-02 3.32173884e-01
-7.12946892e-01 9.80961472e-02 9.43599641e-01 1.98619753e-01
-2.08329722e-01 5.84430695e-02 -4.35619354e-01 1.05940962e+00
5.28639019e-01 -5.67783713e-01 -3.73715013e-01 -7.85342097e-01
3.19144130e-01 1.22992180e-01 5.67627847e-01 -8.22698295e-01
5.96185386e-01 -5.00668287e-01 5.39833128e-01 -7.00127423e-01
5.20775735e-01 -1.18312025e+00 1.14356391e-01 -5.90764545e-02
1.19504578e-01 6.97232842e-01 -8.55682120e-02 7.33274877e-01
-2.87291646e-01 -2.61069238e-01 5.24264932e-01 -2.74737358e-01
-8.74066532e-01 4.45734710e-01 -7.22755194e-02 -2.65547752e-01
1.26098609e+00 -6.60837770e-01 4.54622135e-02 -5.14379561e-01
-5.46607792e-01 5.40417656e-02 1.07790446e+00 2.64253944e-01
9.78781760e-01 -1.45431268e+00 -3.38229716e-01 4.50977474e-01
3.34501773e-01 3.66455555e-01 -1.24498062e-01 1.00957406e+00
-7.81861722e-01 6.47728503e-01 1.82470143e-01 -1.00655746e+00
-1.63012564e+00 5.13216615e-01 3.86491299e-01 1.93744183e-01
-6.77480876e-01 7.10115194e-01 3.18205893e-01 -3.78103524e-01
1.77041531e-01 -4.84105706e-01 2.66170949e-01 -1.75563231e-01
3.05311143e-01 6.23521328e-01 7.75183514e-02 -1.06247211e+00
-5.04489005e-01 1.57781994e+00 -9.22081843e-02 -2.28007257e-01
1.14007032e+00 -6.15359724e-01 -2.00908706e-02 3.22491556e-01
1.44548023e+00 4.30320024e-01 -1.40944695e+00 -3.64242345e-01
-8.23966861e-02 -1.07717359e+00 2.73891687e-01 -1.50579616e-01
-7.79704094e-01 6.77412570e-01 3.96213859e-01 -2.83488512e-01
7.91716397e-01 1.18721128e-01 6.84155047e-01 7.54799485e-01
7.03381240e-01 -8.03835988e-01 1.75985768e-02 4.48715895e-01
7.67382979e-01 -1.47311246e+00 4.36072290e-01 -5.76365530e-01
-2.84514546e-01 1.38760650e+00 5.51154435e-01 -2.21582845e-01
6.27350032e-01 2.37354755e-01 1.88837454e-01 -4.18758720e-01
-4.26424235e-01 1.93955805e-02 5.46410501e-01 5.85485101e-01
2.39277363e-01 -3.47863406e-01 3.59258443e-01 -3.71382445e-01
2.69644745e-02 -3.43302101e-01 3.86928648e-01 1.13360226e+00
-5.08643925e-01 -1.08777332e+00 -7.51981437e-01 6.46686628e-02
-2.48311207e-01 2.53821872e-02 -4.75638002e-01 8.32931399e-01
-3.40333395e-02 7.50395298e-01 4.72970791e-02 -2.77991533e-01
4.77085680e-01 -3.26954454e-01 7.33311474e-01 -7.01065958e-01
-2.40631714e-01 4.31068987e-01 -1.62441000e-01 -9.39551175e-01
-6.19141817e-01 -9.06911731e-01 -1.08075428e+00 -3.03902924e-01
-7.21692801e-01 -2.59363689e-02 5.94762564e-01 7.96095848e-01
4.02843803e-01 -2.21004531e-01 7.97340274e-01 -1.19776452e+00
-3.94173622e-01 -3.40341449e-01 -4.79305118e-01 1.54703796e-01
8.32352281e-01 -6.60374224e-01 -2.74690688e-01 3.45155954e-01] | [7.770287990570068, -2.305175304412842] |
53cb034d-a78a-4047-9755-519acce60731 | super-efficient-super-resolution-for-fast | 2112.14340 | null | https://arxiv.org/abs/2112.14340v1 | https://arxiv.org/pdf/2112.14340v1.pdf | Super-Efficient Super Resolution for Fast Adversarial Defense at the Edge | Autonomous systems are highly vulnerable to a variety of adversarial attacks on Deep Neural Networks (DNNs). Training-free model-agnostic defenses have recently gained popularity due to their speed, ease of deployment, and ability to work across many DNNs. To this end, a new technique has emerged for mitigating attacks on image classification DNNs, namely, preprocessing adversarial images using super resolution -- upscaling low-quality inputs into high-resolution images. This defense requires running both image classifiers and super resolution models on constrained autonomous systems. However, super resolution incurs a heavy computational cost. Therefore, in this paper, we investigate the following question: Does the robustness of image classifiers suffer if we use tiny super resolution models? To answer this, we first review a recent work called Super-Efficient Super Resolution (SESR) that achieves similar or better image quality than prior art while requiring 2x to 330x fewer Multiply-Accumulate (MAC) operations. We demonstrate that despite being orders of magnitude smaller than existing models, SESR achieves the same level of robustness as significantly larger networks. Finally, we estimate end-to-end performance of super resolution-based defenses on a commercial Arm Ethos-U55 micro-NPU. Our findings show that SESR achieves nearly 3x higher FPS than a baseline while achieving similar robustness. | ['Danny Loh', 'Paul Whatmough', 'James Ward', 'Dibakar Gope', 'Kartikeya Bhardwaj'] | 2021-12-29 | null | null | null | null | ['adversarial-defense'] | ['adversarial'] | [ 2.37506360e-01 1.00059360e-02 8.53851363e-02 -4.54693705e-01
-9.39256012e-01 -7.24850953e-01 4.86588538e-01 -6.66197240e-01
-7.95115650e-01 7.87303388e-01 -1.40193209e-01 -4.32362527e-01
2.67860204e-01 -7.46046960e-01 -1.09010375e+00 -5.71382880e-01
-1.11725681e-01 -1.99435875e-01 5.11724770e-01 -5.43495774e-01
-5.01501374e-03 8.40912521e-01 -1.26924503e+00 2.75742531e-01
6.11795604e-01 9.64520097e-01 -3.03981304e-01 9.33193207e-01
4.38881725e-01 8.66898119e-01 -1.04868805e+00 -6.36962414e-01
7.16535628e-01 2.46482133e-03 -6.21360421e-01 -4.57946986e-01
1.17830968e+00 -9.80238557e-01 -8.52074087e-01 1.51653099e+00
5.53085089e-01 -2.07992941e-01 1.27004161e-01 -1.16279662e+00
-9.77515340e-01 8.90461385e-01 -7.73793042e-01 7.90679276e-01
-2.32638091e-01 2.11410940e-01 5.18924892e-01 -3.98903877e-01
3.95005941e-01 1.53184426e+00 6.07578933e-01 1.21801245e+00
-1.25494742e+00 -1.42367220e+00 8.23920295e-02 -3.54291797e-02
-1.23613870e+00 -5.72214901e-01 1.48853272e-01 1.36541292e-01
8.39062631e-01 4.28455137e-03 -2.19636574e-01 1.60551047e+00
4.51820195e-01 1.44107312e-01 1.22974551e+00 2.72399336e-01
3.33996743e-01 -1.38748080e-01 2.51988649e-01 5.03213108e-01
6.51273847e-01 3.48031580e-01 -4.42555845e-01 -1.64574832e-01
1.19864202e+00 -3.09624851e-01 -1.60933703e-01 3.46085161e-01
-6.61822677e-01 8.50714803e-01 6.59206629e-01 -1.20908082e-01
-2.02436298e-01 4.89069372e-01 4.31652814e-01 3.84336174e-01
2.64108986e-01 3.05017889e-01 -3.81013453e-01 1.29925936e-01
-6.53693914e-01 1.63635358e-01 4.99822199e-01 8.20424438e-01
3.05088937e-01 7.31193125e-01 3.30459088e-01 3.68609726e-01
-8.74425992e-02 4.92945582e-01 2.43240580e-01 -1.32709253e+00
5.12534261e-01 -1.88765526e-02 -4.17355038e-02 -6.51142657e-01
-1.23123541e-01 -5.13129234e-01 -1.16510963e+00 9.82362866e-01
3.00747216e-01 -4.72045273e-01 -1.19136441e+00 2.11095214e+00
1.58387631e-01 4.62854534e-01 6.38713419e-01 9.69395101e-01
6.70152843e-01 4.37059432e-01 2.07977489e-01 1.30133435e-01
1.38058555e+00 -6.06305957e-01 -3.47505391e-01 -3.04028928e-01
-1.75837576e-01 -4.12220627e-01 6.20171845e-01 4.03332800e-01
-1.16107798e+00 -6.33214235e-01 -1.54581463e+00 -7.17103109e-02
-2.18184054e-01 -4.39474016e-01 4.75686610e-01 9.89435375e-01
-1.21901333e+00 7.81057179e-01 -9.84828889e-01 7.05478564e-02
9.30806220e-01 5.10086298e-01 -5.78682482e-01 7.93652236e-02
-1.35385036e+00 8.43492270e-01 2.74383754e-01 -4.40118849e-01
-1.38414931e+00 -9.06158209e-01 -6.84384286e-01 5.08453790e-03
8.64444450e-02 -4.37203407e-01 1.10764730e+00 -8.13348949e-01
-1.32176602e+00 6.24951303e-01 2.92264611e-01 -1.27905023e+00
4.29132372e-01 -6.22061968e-01 -6.08046174e-01 5.10257244e-01
-1.67787999e-01 6.63130283e-01 1.25718558e+00 -1.21401680e+00
-6.97475672e-01 -3.68823975e-01 6.01691604e-01 -2.19150707e-01
-4.52520221e-01 3.81908834e-01 3.18369754e-02 -7.48416960e-01
-1.77207068e-01 -8.17589641e-01 -4.58506048e-01 1.57398313e-01
-2.27947682e-01 3.36101890e-01 1.13463295e+00 -4.99490231e-01
5.72874725e-01 -2.27623701e+00 -1.47027671e-01 -2.81319469e-01
6.82157815e-01 7.37522423e-01 -3.13441098e-01 -4.32450801e-01
-1.99256510e-01 5.56769073e-01 -8.32921639e-02 -1.00373112e-01
-3.88987124e-01 1.67027131e-01 -8.60873342e-01 4.57241416e-01
3.59365404e-01 5.82757115e-01 -4.86846745e-01 -2.19048992e-01
-2.73807496e-02 9.27398264e-01 -5.64685404e-01 -7.67646879e-02
3.44487913e-02 2.68958688e-01 -2.84839869e-01 6.13273859e-01
1.15750968e+00 -5.40813617e-02 -1.19981110e-01 -2.32137546e-01
1.81803433e-03 -9.02307034e-03 -9.24021602e-01 1.18361378e+00
-2.58925229e-01 7.45379925e-01 6.00816965e-01 -5.58265030e-01
6.90335155e-01 2.03250527e-01 -1.61096305e-02 -8.01838100e-01
3.33680540e-01 2.81058699e-02 -4.02228348e-02 2.10041374e-01
4.61957067e-01 5.90671636e-02 -1.23951405e-01 1.88981265e-01
-1.64778084e-02 1.76959574e-01 -2.42592409e-01 4.16012883e-01
1.54902732e+00 -1.98227867e-01 -1.75041277e-02 -1.45379305e-01
2.30057955e-01 -9.71024036e-02 6.72936201e-01 1.00534368e+00
-6.08986676e-01 4.31478709e-01 3.54705483e-01 -5.01560092e-01
-1.23971117e+00 -1.39311302e+00 -1.82815030e-01 9.96423006e-01
3.53295356e-01 2.65714005e-02 -1.36328006e+00 -5.35641193e-01
-1.43467337e-01 4.37508702e-01 -6.44259274e-01 -3.00367147e-01
-8.08809400e-01 -9.35585439e-01 1.50922549e+00 8.60980213e-01
1.19863534e+00 -6.18117869e-01 -8.03070188e-01 -7.63870180e-02
1.65884808e-01 -1.67734838e+00 -1.42433196e-01 1.36674821e-01
-9.10896361e-01 -7.36702919e-01 -5.16049087e-01 -4.88234252e-01
4.52898324e-01 5.60689747e-01 9.81229246e-01 -5.10962792e-02
-3.37362617e-01 3.53620686e-02 -1.12429097e-01 -3.91554177e-01
-5.68415463e-01 -1.51923046e-01 5.95399618e-01 -4.19446468e-01
1.14920288e-01 -8.89477253e-01 -5.72258115e-01 1.94994017e-01
-1.13538027e+00 -4.29571003e-01 8.47913086e-01 6.48615122e-01
6.34907544e-01 3.02040368e-01 6.23682141e-01 -7.15912580e-01
4.74557459e-01 -3.72956604e-01 -9.33412611e-01 -6.57125339e-02
-2.81202644e-01 1.86552584e-01 1.09962738e+00 -7.99098790e-01
-9.87462163e-01 -2.70153023e-02 -2.84585595e-01 -8.30510259e-01
-2.56633461e-01 -2.37982705e-01 -1.78856999e-01 -6.00717843e-01
1.17760646e+00 -1.21297371e-02 -8.45667254e-03 -2.22935766e-01
4.36998904e-01 3.82887989e-01 1.17454708e+00 -3.15420717e-01
1.58182228e+00 8.22331786e-01 1.53095499e-01 -7.22450852e-01
-9.14404869e-01 3.64863753e-01 -3.17574203e-01 2.38689512e-01
1.04560173e+00 -1.35155571e+00 -7.60554850e-01 5.05833864e-01
-9.19647455e-01 -3.49827856e-01 8.60343874e-02 1.81780174e-01
-9.50634927e-02 4.13457006e-01 -1.14103913e+00 -5.67918777e-01
-7.22945988e-01 -1.07772958e+00 6.56803131e-01 5.22931516e-01
2.83985615e-01 -4.02086407e-01 -3.63657594e-01 4.86300051e-01
7.89702117e-01 5.85609913e-01 5.33922851e-01 -5.83418548e-01
-7.94745207e-01 -9.53389481e-02 -6.85312450e-01 6.30621672e-01
-2.11528242e-01 -4.19046879e-02 -1.18270791e+00 -6.28283739e-01
1.87398911e-01 -4.89930272e-01 1.09851277e+00 1.79647207e-01
1.11885571e+00 -4.66086179e-01 1.68629035e-01 1.13882649e+00
1.71788383e+00 1.86246224e-02 8.14466000e-01 6.16393268e-01
6.53413653e-01 6.37292787e-02 4.23476666e-01 1.68674693e-01
-7.50692859e-02 2.73935467e-01 1.03086734e+00 2.50294842e-02
-2.00703129e-01 1.26556084e-01 6.74996436e-01 -9.83957574e-02
-4.74355109e-02 -1.26541317e-01 -4.47831273e-01 1.77163199e-01
-1.31367040e+00 -1.17608726e+00 1.77902341e-01 2.02545738e+00
7.32603431e-01 8.14833939e-01 4.58703563e-02 -8.43004957e-02
7.60261953e-01 3.09508950e-01 -1.00344181e+00 -5.05841017e-01
-2.38967896e-01 5.10709584e-01 1.08301449e+00 3.84485841e-01
-1.29127705e+00 1.21977818e+00 6.17306519e+00 8.66050899e-01
-1.33580601e+00 2.48398945e-01 7.52241790e-01 -2.79903620e-01
2.69757420e-01 -4.25821304e-01 -1.11880922e+00 1.57767624e-01
1.30008829e+00 -1.15625009e-01 6.74181759e-01 1.16126001e+00
-1.14065684e-01 3.29327703e-01 -6.72645986e-01 7.06084013e-01
-5.25553375e-02 -1.44307959e+00 2.71512836e-01 1.93823203e-01
6.35993183e-01 4.31864172e-01 5.77706635e-01 3.82641375e-01
8.65473628e-01 -1.31348383e+00 5.77522516e-01 -1.06860690e-01
1.16951680e+00 -1.15013456e+00 6.97163284e-01 2.19961762e-01
-1.09082854e+00 -2.58454680e-01 -9.06967103e-01 9.65097770e-02
-1.56863183e-01 1.44420072e-01 -2.27762759e-01 1.83902651e-01
1.06759024e+00 -5.26647270e-02 -6.04737997e-01 3.11370760e-01
-1.79303229e-01 6.07659340e-01 -3.28066409e-01 5.67406118e-01
2.18724146e-01 5.18923700e-01 5.94281137e-01 1.02859855e+00
2.10275166e-02 5.00481963e-01 -2.39258558e-01 9.79029834e-01
-5.13679266e-01 -7.90681958e-01 -5.21855414e-01 2.18964517e-01
8.82938385e-01 1.41613054e+00 -5.40627003e-01 -2.22509652e-01
-2.74777085e-01 1.09469569e+00 3.28808516e-01 1.90493837e-01
-1.36136413e+00 -2.65311569e-01 1.37756586e+00 -2.19819352e-01
1.96671054e-01 -2.88556039e-01 -5.68002939e-01 -1.09431159e+00
-1.19984366e-01 -1.21619666e+00 2.59877443e-01 -3.69504273e-01
-1.15676248e+00 1.22634518e+00 -2.74769366e-01 -1.00240326e+00
5.46555258e-02 -7.02410519e-01 -5.22688448e-01 6.71259522e-01
-1.63312507e+00 -9.38514411e-01 -4.55991596e-01 8.08771372e-01
5.77228963e-01 -3.36288422e-01 7.45217443e-01 2.25754872e-01
-8.25333714e-01 1.12880743e+00 -4.78157364e-02 5.72959960e-01
6.78305745e-01 -1.16389096e+00 1.11996758e+00 1.49088216e+00
-1.55218363e-01 7.13170588e-01 8.36157501e-01 -5.36185563e-01
-1.35189390e+00 -1.40136230e+00 -1.43151164e-01 -6.35220826e-01
7.41860449e-01 -2.51714557e-01 -1.06463540e+00 6.58681154e-01
1.39112398e-01 3.94402742e-01 2.64918983e-01 -5.02254546e-01
-1.16587555e+00 -1.54943287e-01 -1.66666913e+00 9.30674911e-01
9.85054851e-01 -5.39339364e-01 -3.67336690e-01 -1.39417976e-01
1.22227716e+00 -6.39009118e-01 -9.45769608e-01 4.88346726e-01
4.52475727e-01 -1.04879427e+00 1.42910576e+00 -5.52824616e-01
5.52908182e-01 -2.06305981e-01 -4.45556521e-01 -9.19393301e-01
-4.04925466e-01 -6.90361440e-01 -2.43604362e-01 1.04106438e+00
2.45502189e-01 -7.56703794e-01 9.54503179e-01 6.90942466e-01
2.93028373e-02 -4.22721028e-01 -1.09740019e+00 -1.02449918e+00
4.49552119e-01 -1.61910400e-01 4.56055760e-01 7.72126973e-01
-5.66278696e-01 6.67245165e-02 -3.58647734e-01 1.19963157e+00
1.35744679e+00 -5.09909093e-01 6.72986448e-01 -9.60859001e-01
-1.67583212e-01 -2.98926145e-01 -5.19610405e-01 -6.15239382e-01
2.33187601e-01 -3.31254572e-01 -1.82910234e-01 -9.19206202e-01
9.54066515e-02 -5.96874058e-02 -4.00838524e-01 5.84815919e-01
-1.19891718e-01 8.06475759e-01 2.80077577e-01 1.10533699e-01
-5.66843748e-01 -6.08620606e-02 7.76144683e-01 4.27469276e-02
1.02427542e-01 -2.46008769e-01 -1.00774705e+00 9.84781086e-01
1.16597795e+00 -4.41698998e-01 -3.21967065e-01 -7.13466406e-01
-2.83674210e-01 1.67345483e-04 6.36211872e-01 -1.44347298e+00
2.23704427e-01 -1.04790658e-01 4.82120931e-01 -1.25734836e-01
4.95433956e-01 -6.84296072e-01 -1.53185904e-01 5.72918475e-01
-2.76602477e-01 1.77214131e-01 6.03975534e-01 5.19147992e-01
1.39695525e-01 1.48400947e-01 1.53555548e+00 -1.94471464e-01
-7.51485229e-01 4.76450801e-01 -4.36363846e-01 1.53441623e-01
9.06075656e-01 -8.95400047e-02 -1.07494223e+00 -2.21714497e-01
-2.80965090e-01 -1.48778260e-01 5.48143029e-01 5.06744385e-01
8.00943494e-01 -8.87837291e-01 -8.64675164e-01 4.18716148e-02
-4.33774948e-01 1.18452944e-02 4.29858178e-01 3.22268419e-02
-7.46859670e-01 5.15539087e-02 -6.23810351e-01 -1.27161875e-01
-1.59125423e+00 6.93795741e-01 4.42578107e-01 -4.15164381e-02
-8.14400554e-01 1.13076532e+00 3.23726803e-01 1.30112305e-01
3.91424417e-01 1.25239164e-01 -7.22182244e-02 -5.91245115e-01
1.08084130e+00 3.71537089e-01 -3.43366086e-01 -6.12476528e-01
-4.20936137e-01 2.59141624e-01 -6.22234285e-01 -7.88304359e-02
1.32785618e+00 7.90830478e-02 1.62405550e-01 -5.47389328e-01
9.52818155e-01 -3.37269418e-02 -1.75845504e+00 -1.30303070e-01
-3.36269557e-01 -3.34907949e-01 2.57404447e-01 -7.04421401e-01
-1.38011813e+00 6.90310419e-01 8.44725549e-01 4.17342409e-02
1.25055695e+00 -2.66461283e-01 1.11588860e+00 4.69167054e-01
5.82278311e-01 -7.67381966e-01 1.91038921e-01 5.32961011e-01
6.05559766e-01 -1.25443435e+00 8.08359534e-02 -3.89692754e-01
-5.52250087e-01 9.18096781e-01 1.04526484e+00 -6.66349173e-01
4.37646568e-01 9.75545168e-01 2.75138646e-01 1.89207822e-01
-8.43418181e-01 8.22319090e-02 -3.57451439e-01 9.24945831e-01
-2.07243532e-01 -2.32086658e-01 2.86007524e-01 5.41811168e-01
-3.84060323e-01 -3.57606709e-01 8.49391043e-01 7.46072352e-01
-5.14807820e-01 -6.49980128e-01 -5.51739633e-01 1.64869893e-02
-1.36818027e+00 -1.25629619e-01 -3.04722667e-01 8.09151590e-01
-8.44527036e-02 1.15262258e+00 -9.44858491e-02 -7.72770822e-01
9.29872468e-02 -6.27328277e-01 2.55797118e-01 -2.44002521e-01
-5.57316899e-01 -4.50522363e-01 -1.22924514e-01 -8.67095530e-01
-4.77951691e-02 -1.29110873e-01 -1.24992943e+00 -7.46447027e-01
-7.94366971e-02 -3.02972496e-01 7.41167128e-01 7.11970508e-01
4.01796073e-01 7.14630783e-01 6.30013764e-01 -9.71167564e-01
-1.16765428e+00 -5.21926522e-01 -2.84941941e-01 -2.82075051e-02
6.42107606e-01 -2.38324851e-01 -6.35617256e-01 -1.27918273e-01] | [5.5665459632873535, 7.915107250213623] |
767b77ef-8106-4631-9258-fc3834967246 | deep-forecast-deep-learning-based-spatio | 1707.08110 | null | http://arxiv.org/abs/1707.08110v1 | http://arxiv.org/pdf/1707.08110v1.pdf | Deep Forecast: Deep Learning-based Spatio-Temporal Forecasting | The paper presents a spatio-temporal wind speed forecasting algorithm using
Deep Learning (DL)and in particular, Recurrent Neural Networks(RNNs). Motivated
by recent advances in renewable energy integration and smart grids, we apply
our proposed algorithm for wind speed forecasting. Renewable energy resources
(wind and solar)are random in nature and, thus, their integration is
facilitated with accurate short-term forecasts. In our proposed framework, we
model the spatiotemporal information by a graph whose nodes are data generating
entities and its edges basically model how these nodes are interacting with
each other. One of the main contributions of our work is the fact that we
obtain forecasts of all nodes of the graph at the same time based on one
framework. Results of a case study on recorded time series data from a
collection of wind mills in the north-east of the U.S. show that the proposed
DL-based forecasting algorithm significantly improves the short-term forecasts
compared to a set of widely-used benchmarks models. | ['Borhan M. Sanandaji', 'Amir Ghaderi', 'Faezeh Ghaderi'] | 2017-07-24 | null | null | null | null | ['spatio-temporal-forecasting'] | ['time-series'] | [-4.47209746e-01 -3.98683816e-01 -1.46024689e-01 -1.23234853e-01
2.61779636e-01 -5.68928063e-01 9.34532821e-01 1.57942753e-02
1.38808459e-01 9.52206671e-01 3.80103797e-01 -6.08544886e-01
-2.42939949e-01 -1.31629074e+00 -4.93144393e-01 -8.97269487e-01
-6.37706876e-01 5.97355552e-02 -2.99961179e-01 -5.87650061e-01
-6.55960888e-02 6.49343550e-01 -1.49270368e+00 -2.87046820e-01
9.84088719e-01 1.00903273e+00 2.80142397e-01 6.85359716e-01
-2.02201530e-01 1.10698283e+00 -5.98616540e-01 3.09574306e-01
3.63312848e-02 -2.46702045e-01 -4.60885048e-01 -4.25396889e-01
-3.74339104e-01 -1.44446850e-01 -9.52175260e-02 7.33944654e-01
5.49775660e-01 4.43266273e-01 7.61340201e-01 -1.17779696e+00
-5.78346670e-01 8.39013875e-01 -3.79287094e-01 5.10868311e-01
-1.47478387e-01 -1.91115320e-01 9.16375935e-01 -7.43548095e-01
6.43995821e-01 7.40730643e-01 7.32179582e-01 1.23806395e-01
-8.76959920e-01 -6.53526366e-01 3.98373067e-01 1.78834096e-01
-1.19408238e+00 -2.11735293e-01 1.01022184e+00 -6.17218256e-01
1.40257120e+00 5.71863167e-02 9.86433148e-01 9.00985360e-01
8.19644511e-01 2.62172371e-01 6.69442832e-01 -2.05932379e-01
5.88766873e-01 -2.47551158e-01 1.22360870e-01 1.68657899e-01
2.81028926e-01 3.21386933e-01 -3.50372076e-01 1.21687897e-01
4.25720125e-01 1.86379701e-01 -2.31124371e-01 -2.63682492e-02
-9.65718865e-01 8.03468347e-01 7.30035841e-01 8.52691889e-01
-8.87158990e-01 2.44788423e-01 4.13032800e-01 1.77821085e-01
1.19338048e+00 -2.45857239e-02 -5.35752714e-01 -2.63858400e-02
-1.20974457e+00 1.74812779e-01 7.89826691e-01 4.73722667e-01
3.83600503e-01 1.03483367e+00 3.83556008e-01 3.87563646e-01
2.09417582e-01 8.92889202e-01 8.60633492e-01 -1.31242305e-01
4.22505885e-01 2.72296965e-01 5.70487142e-01 -1.39731205e+00
-9.96061504e-01 -9.95289207e-01 -1.51063466e+00 1.39289632e-01
-9.37663168e-02 -8.60570550e-01 -5.46674788e-01 1.49143958e+00
9.81876999e-02 7.13501155e-01 2.60231018e-01 6.69039488e-01
6.01355374e-01 1.31211293e+00 1.39671639e-01 -6.01591647e-01
8.69210124e-01 -9.02022243e-01 -1.09524012e+00 1.92711845e-01
6.89906538e-01 -3.23128432e-01 1.56163022e-01 1.88843280e-01
-8.14941108e-01 -8.11170399e-01 -1.03632581e+00 4.79490548e-01
-1.12318444e+00 2.97913849e-01 6.51693285e-01 2.15320915e-01
-1.33995688e+00 7.72073448e-01 -9.70175505e-01 -3.84411275e-01
-1.16371617e-01 -2.64190678e-02 5.33927381e-02 7.57188022e-01
-1.57972777e+00 9.70751524e-01 3.50617737e-01 6.79122388e-01
-6.58712804e-01 -6.77886009e-01 -7.77315855e-01 4.39543635e-01
-1.11266181e-01 -7.84994662e-01 9.24779415e-01 -6.03127122e-01
-1.50296748e+00 -2.34764233e-01 -4.16293681e-01 -9.11082327e-01
2.06533745e-01 -1.53406337e-01 -1.06663597e+00 -3.82077277e-01
-1.96047008e-01 -1.45461053e-01 6.73125148e-01 -1.00534534e+00
-7.98529923e-01 -1.83534071e-01 -1.89145222e-01 1.44329853e-02
-4.63207424e-01 -3.77819598e-01 6.17642879e-01 -8.93345535e-01
-3.10053140e-01 -6.12719476e-01 -5.04672110e-01 -8.90218675e-01
-3.19322109e-01 -4.86647487e-01 1.01200640e+00 -8.30761850e-01
1.62263215e+00 -1.78904283e+00 1.36038095e-01 2.91191608e-01
5.41858785e-02 2.58377105e-01 1.35658205e-01 1.08142292e+00
-4.03029859e-01 2.46582121e-01 6.34799302e-02 -4.19291615e-01
6.21811002e-02 4.36748296e-01 -1.08721042e+00 3.35180342e-01
-2.19767183e-01 9.20985818e-01 -1.07486498e+00 4.34943974e-01
6.41815364e-01 7.14086413e-01 3.66128325e-01 1.29282907e-01
-3.15590769e-01 3.30246270e-01 -3.70685548e-01 3.19091305e-02
7.08938539e-01 -4.46567059e-01 2.99337178e-01 1.48065491e-02
-6.94067240e-01 1.91631436e-01 -1.17032886e+00 1.32692540e+00
-9.91795659e-01 4.56430167e-01 -2.81451941e-01 -1.15422034e+00
1.06179965e+00 5.07379353e-01 5.21613657e-01 -6.32214189e-01
-1.33808210e-01 2.41197035e-01 -2.16697112e-01 -2.42187515e-01
5.73287368e-01 1.87199697e-01 4.25205678e-01 7.66390324e-01
3.02903224e-02 -3.51046994e-02 3.63368660e-01 -1.41600790e-02
5.35693109e-01 1.55029088e-01 3.92032593e-01 -4.84958827e-01
4.41020608e-01 -2.66083419e-01 5.69216073e-01 4.75415677e-01
3.47886860e-01 4.45277616e-02 1.84833840e-01 -1.12309480e+00
-8.72201502e-01 -6.17930055e-01 3.67774582e-03 8.23824108e-01
-2.71664262e-01 -4.01576698e-01 -2.83691227e-01 -4.09119189e-01
2.89718881e-02 1.07107234e+00 -7.37580180e-01 3.36504400e-01
-4.04703915e-01 -9.27952826e-01 -1.35062672e-02 6.59277499e-01
2.38910422e-01 -1.07341456e+00 -7.64094114e-01 5.35363615e-01
1.28396764e-01 -9.96463537e-01 1.10012658e-01 2.86724567e-01
-7.92637885e-01 -5.95008016e-01 -8.77820492e-01 -4.14028466e-01
4.72841293e-01 2.82348722e-01 1.31874943e+00 -1.57901943e-01
1.87877834e-01 -4.63741161e-02 -4.33884352e-01 -6.09769225e-01
-5.76520413e-02 3.48166496e-01 4.57656831e-01 1.05145156e-01
-2.00861972e-02 -1.06547940e+00 -6.97183132e-01 -1.27229139e-01
-7.69427896e-01 9.83794183e-02 3.26762162e-03 6.04891956e-01
2.28594288e-01 7.70547807e-01 1.21678102e+00 -7.44114578e-01
6.71340585e-01 -1.10595083e+00 -1.17862725e+00 2.41331384e-01
-9.94798005e-01 -1.50227129e-01 1.12058485e+00 4.68411818e-02
-1.01751661e+00 -1.14796869e-01 1.75596058e-01 -1.04242496e-01
-8.14853013e-02 1.02132773e+00 5.01221836e-01 2.88980275e-01
2.71525979e-01 4.78213072e-01 -5.48223138e-01 -5.78797519e-01
4.85697836e-01 4.61333692e-01 3.45340937e-01 1.23408012e-01
9.42467391e-01 5.07592976e-01 2.33843401e-01 -8.27235162e-01
-3.71730357e-01 -1.55568048e-01 -5.96141160e-01 -5.38745165e-01
6.12213671e-01 -1.11824989e+00 -7.56591797e-01 6.69748127e-01
-1.23708081e+00 -4.00251120e-01 -3.26371640e-01 6.58491135e-01
-4.14295420e-02 -1.34521037e-01 -4.52232599e-01 -1.37317824e+00
-7.53823221e-01 -6.33696914e-01 8.37418318e-01 3.99959207e-01
1.38217760e-02 -1.94429779e+00 4.37790006e-01 -5.96553683e-01
1.10975015e+00 5.22206247e-01 8.32663536e-01 -4.65756059e-01
-1.41503096e-01 -4.04962301e-02 7.73003250e-02 2.85004228e-01
5.41522324e-01 3.09084505e-01 -8.37210655e-01 -3.40517968e-01
3.30828466e-02 2.67771035e-01 6.45074010e-01 7.49286115e-01
6.50889277e-01 -2.87126034e-01 -3.27921599e-01 6.06349587e-01
1.66063368e+00 4.56163436e-01 9.72488970e-02 1.09381735e-01
6.99229121e-01 3.03452462e-01 1.63333699e-01 8.20007801e-01
7.59844601e-01 3.83749515e-01 5.25370896e-01 -3.45273018e-01
2.57786095e-01 -1.75565958e-01 4.17593330e-01 1.22339082e+00
-6.18411541e-01 -6.22606218e-01 -1.04046881e+00 8.52993429e-01
-2.09058976e+00 -1.14977646e+00 -2.31566563e-01 1.93827188e+00
1.11233383e-01 -1.77543730e-01 6.43494949e-02 1.17646448e-01
4.59167004e-01 8.42493534e-01 -5.14834344e-01 -4.92228270e-01
-4.52255994e-01 3.03970754e-01 6.01908565e-01 6.07569635e-01
-1.09309244e+00 6.33433938e-01 6.28995609e+00 4.92534846e-01
-1.48687589e+00 -1.20127738e-01 6.27223551e-01 3.50769609e-02
-3.94982725e-01 -3.78166378e-01 -7.29635775e-01 5.99725485e-01
1.45137417e+00 -7.48548090e-01 4.72572953e-01 7.60031581e-01
9.23484623e-01 8.78127366e-02 -6.17795289e-01 5.05138338e-01
-1.90136909e-01 -1.71488655e+00 4.50021867e-03 5.88143058e-02
1.16548765e+00 5.21508992e-01 9.60756466e-02 9.09858197e-02
7.26278245e-01 -1.13982189e+00 3.93142730e-01 1.02817225e+00
3.59904885e-01 -9.49935317e-01 1.22426915e+00 3.92754227e-01
-1.81530535e+00 -2.37891629e-01 -3.71054471e-01 -5.25510788e-01
4.10284370e-01 1.28201902e+00 -5.57485163e-01 1.37178802e+00
7.69383609e-01 1.50992870e+00 -1.42344665e-02 4.83967811e-01
-3.49585414e-01 8.98989439e-01 -3.27965140e-01 -4.35142294e-02
3.12837869e-01 -5.59646308e-01 2.15954363e-01 1.04744411e+00
8.23682010e-01 -4.78048511e-02 3.07425740e-03 5.91133118e-01
-1.13846781e-02 6.55647367e-02 -1.14419341e+00 9.74415317e-02
5.00804424e-01 1.39897406e+00 -5.40373623e-01 -4.61748540e-01
-3.57105196e-01 1.88826814e-01 2.78465301e-02 9.18423831e-01
-6.35906935e-01 -2.86647588e-01 6.46403551e-01 -1.97751924e-01
5.64458013e-01 -4.35122073e-01 -4.06499535e-01 -1.05983424e+00
-6.36242777e-02 -1.22910708e-01 2.32607186e-01 -9.96124327e-01
-1.49827433e+00 8.25902045e-01 -3.10000241e-01 -1.42013299e+00
-9.79326725e-01 -2.51858860e-01 -9.96288717e-01 1.35743439e+00
-2.03168583e+00 -1.11314249e+00 -9.57056060e-02 4.53823626e-01
4.33810115e-01 -2.93544084e-01 1.10836363e+00 -9.07275081e-02
-6.77703381e-01 -2.17708439e-01 7.20646977e-01 2.72801686e-02
-9.49791446e-02 -1.40796983e+00 1.12361383e+00 1.15515912e+00
2.24829718e-01 4.74438339e-01 5.81065714e-01 -7.40282714e-01
-1.29134393e+00 -1.34384382e+00 1.17167878e+00 7.40962625e-02
1.10386777e+00 -3.50128084e-01 -7.16285050e-01 7.74906754e-01
7.52849162e-01 3.67802650e-01 3.51170331e-01 1.06507391e-01
-8.73277662e-04 -3.41011405e-01 -7.75192499e-01 2.53389329e-01
6.40939713e-01 -4.25245881e-01 -2.80393869e-01 5.62813699e-01
3.38965356e-01 -4.02413085e-02 -9.73144352e-01 4.89637047e-01
4.04581100e-01 -7.98429668e-01 6.72004640e-01 -4.25399035e-01
3.08807880e-01 -4.31179196e-01 1.17306471e-01 -2.34682941e+00
-4.53112334e-01 -6.83947265e-01 -5.91247976e-01 1.07540023e+00
2.75383621e-01 -1.07811832e+00 3.06484282e-01 2.00980961e-01
1.15067497e-01 -5.45287430e-01 -9.82180834e-01 -7.74067700e-01
1.59837127e-01 -3.68429810e-01 1.15251696e+00 1.26939952e+00
1.03153400e-01 2.59481907e-01 -6.81932569e-01 4.16073948e-01
2.68945187e-01 5.64747930e-01 3.76172692e-01 -1.45903647e+00
1.70904472e-01 -3.78058195e-01 -1.35285094e-01 -6.53445184e-01
2.89526969e-01 -6.30800545e-01 -3.77994835e-01 -1.86503208e+00
-5.76061726e-01 -3.13496083e-01 -5.32982826e-01 3.62146497e-01
1.61329955e-01 -2.01601297e-01 3.90876345e-02 1.57378856e-02
2.07782928e-02 8.10309947e-01 6.79942489e-01 -4.36624438e-02
-1.52709767e-01 1.91874728e-01 3.15337442e-02 6.88847423e-01
1.05039418e+00 -4.57303114e-02 -6.90485895e-01 -6.26643836e-01
5.08962929e-01 3.28601211e-01 -6.66639432e-02 -1.08497703e+00
4.17537391e-01 -4.42510545e-02 4.15549845e-01 -8.79284680e-01
-8.34313035e-02 -9.02893245e-01 6.27051473e-01 5.48797905e-01
-4.81685922e-02 7.19297469e-01 2.00060189e-01 4.64897931e-01
-3.52109015e-01 2.92526811e-01 1.25834003e-01 1.17499299e-01
-6.22727692e-01 2.98551023e-01 -5.88115036e-01 -5.05243480e-01
8.30790818e-01 4.13229555e-01 -6.87286317e-01 -8.08213055e-01
-7.73796439e-01 3.01280588e-01 -3.41236927e-02 5.47457218e-01
2.66779482e-01 -1.26387715e+00 -8.24214280e-01 2.09834740e-01
-4.12012160e-01 -2.97385424e-01 3.58721733e-01 4.94724959e-01
-3.02330971e-01 8.01192880e-01 -9.31520090e-02 -3.79921913e-01
-5.03444910e-01 4.72271830e-01 5.65991998e-01 -6.08864069e-01
-6.71614826e-01 4.20188814e-01 -1.10037357e-01 -3.23021263e-01
-1.36424586e-01 -7.73267269e-01 -9.15194452e-01 4.48622912e-01
6.65305674e-01 5.28792322e-01 3.41359556e-01 -7.85803795e-01
-2.58021683e-01 7.15928912e-01 4.76149648e-01 8.52938592e-02
1.92807388e+00 -2.96621531e-01 -1.73381895e-01 1.11964428e+00
6.95563972e-01 -7.46277943e-02 -7.25013316e-01 -1.17614809e-02
-6.68843761e-02 1.34056658e-01 3.85453910e-01 -8.87615204e-01
-1.59254634e+00 8.21397781e-01 5.82925260e-01 1.06113923e+00
1.17750144e+00 -7.31476307e-01 8.31359267e-01 2.02053726e-01
4.98309344e-01 -1.00916910e+00 -1.00050724e+00 6.56236589e-01
7.03243732e-01 -7.48077095e-01 -7.10193664e-02 7.25069568e-02
-2.37928480e-01 1.27533567e+00 7.75595009e-03 -2.21466020e-01
1.08652258e+00 4.69375908e-01 1.72293559e-01 -7.57257827e-03
-1.21042347e+00 -1.71466291e-01 1.10402934e-01 4.99494433e-01
4.81426299e-01 3.91081989e-01 -2.12081134e-01 4.98991907e-01
-3.35694939e-01 3.13462496e-01 6.11768007e-01 5.83492398e-01
-1.14383893e-02 -6.37302518e-01 -3.72556150e-02 3.45326692e-01
-2.12489411e-01 -9.46768373e-02 3.45400035e-01 5.97740293e-01
-1.12779677e-01 1.21145666e+00 2.33188346e-01 -4.12815124e-01
1.13928817e-01 8.58339593e-02 -4.60215062e-01 -1.87917873e-01
-6.54004276e-01 -6.35062009e-02 -9.81791914e-02 -4.21971738e-01
-5.74832678e-01 -4.95818794e-01 -1.21269321e+00 -4.05799091e-01
-1.45510390e-01 6.23637378e-01 1.02311683e+00 1.06718683e+00
5.76227009e-01 9.69673574e-01 1.29422557e+00 -1.16510141e+00
-2.23931000e-01 -1.10920346e+00 -1.01350856e+00 4.17039618e-02
5.52895248e-01 -5.05147398e-01 -6.23470128e-01 -2.91907992e-02] | [6.329470157623291, 2.821597099304199] |
59ab8eae-13ba-468e-b120-2332cd27ca78 | a-mask-based-model-for-mandarin-chinese | null | null | http://www.interspeech2020.org/uploadfile/2020/1021/20201021034849937.pdf | http://www.interspeech2020.org/uploadfile/2020/1021/20201021034849937.pdf | A Mask-based Model for Mandarin Chinese Polyphone Disambiguation | Polyphone disambiguation serves as an essential part of Mandarin text-to-speech (TTS) system. However, conventional system modeling the entire Pinyin set causes the case that prediction belongs to the unrelated polyphonic character instead of the current input one, which has negative impacts on TTS performance. To address this issue, we introduce a mask-based model for polyphone disambiguation. The model takes a mask vector extracted from the context as an extra input. In our model, the mask vector not only acts as a weighting factor in Weightedsoftmax to prevent the case of mis-prediction but also eliminates the contribution of non-candidate set to the overall loss. Moreover, to mitigate the uneven distribution of pronunciation, we introduce a new loss called Modified Focal Loss. The experimental result shows the effectiveness of the proposed mask based model. We also empirically studied the impact of Weighted-softmax and Modified Focal Loss. It was found that Weighted-softmax can effectively prevent the model from predicting outside the candidate set. Besides, Modified Focal Loss can reduce the adverse impacts of the uneven distribution of
pronunciation. | ['Haiteng Zhang'] | 2020-10-21 | null | null | null | null | ['polyphone-disambiguation'] | ['natural-language-processing'] | [ 1.4964750e-01 -2.1888910e-02 -1.4091270e-01 -2.8943697e-01
-7.0542395e-01 -2.1897736e-01 1.7653441e-01 -1.7972720e-01
-4.9266687e-01 5.9230471e-01 3.7586066e-01 -3.4950125e-01
1.9154653e-01 -4.2769691e-01 -4.9836373e-01 -6.2153554e-01
4.4974405e-01 -2.4637088e-01 5.9336710e-01 -1.3184555e-01
1.9715129e-01 -1.6829042e-02 -1.2884102e+00 2.9419911e-01
9.4469273e-01 8.8735670e-01 8.4594584e-01 4.7919515e-01
-2.3016301e-01 5.5743569e-01 -1.1756430e+00 -1.7761779e-01
4.3171245e-01 -4.7443214e-01 -1.6494308e-01 -2.4237898e-01
2.4584943e-01 -4.0683526e-01 -3.8397345e-01 1.1774741e+00
8.4398240e-01 2.6136360e-01 4.4275308e-01 -1.0075639e+00
-2.4524982e-01 9.9896240e-01 -4.8544535e-01 5.1269889e-01
-1.1039575e-01 7.0066430e-02 8.9193296e-01 -7.6631504e-01
9.7844459e-02 1.3456563e+00 3.6475626e-01 4.7564110e-01
-7.5117731e-01 -9.6276063e-01 3.0785459e-01 2.2199626e-01
-1.3438810e+00 -6.5469545e-01 9.1112024e-01 -1.9924174e-01
8.3833367e-01 3.8986900e-01 2.6367837e-01 7.0436233e-01
8.8629127e-02 8.8314623e-01 8.8757867e-01 -7.8756338e-01
-1.6646051e-01 1.7925729e-01 2.8173271e-01 2.0269878e-01
-3.7394986e-01 3.5476336e-01 -5.9849542e-01 5.0323527e-02
4.3393430e-01 -3.1998292e-01 -3.5570577e-01 6.0488284e-01
-8.8116091e-01 4.6415442e-01 -1.2715617e-01 3.0570674e-01
-1.2386497e-01 1.9072245e-01 1.3361830e-01 2.0567584e-01
1.5965892e-01 4.5235977e-02 -6.5953857e-01 -4.4385505e-01
-9.9436241e-01 -5.9021831e-02 3.9066175e-01 8.1175178e-01
4.7644007e-01 6.1803138e-01 -4.6501222e-01 1.1752775e+00
4.4855198e-01 4.6125120e-01 8.7787914e-01 -4.3944356e-01
7.8145653e-01 1.8583359e-01 -8.6116269e-02 -7.0414078e-01
2.3113834e-02 -7.4050367e-01 -4.4896537e-01 1.6915411e-02
2.4739392e-01 -4.9954408e-01 -1.0354172e+00 2.0527511e+00
2.2879523e-01 2.7637717e-01 -1.2103129e-01 7.7936059e-01
4.6383998e-01 1.0119029e+00 -8.9309458e-04 -4.7357878e-01
1.1109167e+00 -8.3879155e-01 -1.1222750e+00 -3.1221431e-01
1.8998434e-01 -1.4485072e+00 1.0931044e+00 3.1800967e-01
-8.1301975e-01 -7.4830818e-01 -1.3020078e+00 2.7949736e-01
-6.9462739e-02 3.1800306e-01 -3.1023908e-02 7.9352808e-01
-5.2593535e-01 4.8885122e-01 -4.6098810e-01 1.9921905e-01
-1.8489626e-01 3.4083048e-01 1.7167921e-01 4.0902433e-01
-1.6890011e+00 8.5800213e-01 6.4415669e-01 7.3042125e-02
-4.2245227e-01 -5.9156549e-01 -5.7587886e-01 1.8405600e-01
4.3195379e-01 7.8710116e-02 1.3700083e+00 -9.8901427e-01
-1.4556881e+00 6.5095432e-02 -3.6099741e-01 -4.7525379e-01
4.4853842e-01 -3.1333315e-01 -9.5539057e-01 -3.1958959e-01
-2.1402600e-01 5.4563421e-01 1.0841829e+00 -9.6786034e-01
-9.7305042e-01 3.8611375e-02 -3.0244935e-01 3.4872660e-01
-3.6266822e-01 1.0406877e-01 -5.4101169e-01 -1.3115669e+00
-6.8934299e-02 -7.7876323e-01 1.2372384e-01 -5.9217274e-01
-6.2837118e-01 -1.6622280e-01 1.2111002e+00 -1.0420905e+00
2.0317404e+00 -2.5163031e+00 -4.5952511e-01 1.1656614e-01
-2.9441798e-01 7.2156334e-01 6.0681071e-02 2.6510349e-01
4.2901281e-02 1.6372998e-01 -1.1468286e-01 -1.6896413e-01
-5.4129265e-02 1.9240230e-01 -3.2919762e-01 -9.9388763e-02
1.5659280e-01 3.2183662e-01 -5.7920527e-01 -5.0679588e-01
4.2767844e-01 5.0053447e-01 -2.7095437e-01 8.6151220e-02
-2.0116846e-01 6.6679880e-02 -1.1580741e-01 3.6091039e-01
7.3150659e-01 5.5788988e-01 -2.2985433e-01 -1.7009488e-01
-2.3069355e-01 6.1140370e-01 -1.3835582e+00 1.0817674e+00
-5.2737278e-01 5.3146815e-01 3.0777853e-03 -5.2079064e-01
1.0066757e+00 5.6844312e-01 1.2523091e-01 -6.7784083e-01
1.6456063e-01 2.5434697e-01 5.9138453e-01 -2.4804123e-01
6.2539136e-01 -2.5524098e-01 1.6270816e-01 2.5468487e-01
-2.0159543e-01 2.3731016e-01 -2.0577368e-01 -1.6520612e-01
7.5259542e-01 -4.0532097e-02 1.3135642e-01 -9.1599859e-02
5.3785723e-01 -4.4994047e-01 1.0036705e+00 5.2563065e-01
-4.3593425e-01 4.8554239e-01 1.7299914e-01 1.9350482e-02
-6.9389969e-01 -9.6236402e-01 -8.5555270e-02 1.1105818e+00
1.7085651e-01 -4.2330489e-01 -8.5632241e-01 -5.7322615e-01
-9.9367224e-02 1.0234345e+00 3.9088178e-02 -3.9459637e-01
-7.3820615e-01 -6.3984394e-01 8.2242823e-01 3.6295560e-01
4.6469107e-01 -1.1833255e+00 -1.8917148e-01 4.1722441e-01
-3.3731383e-01 -9.9764222e-01 -1.1467906e+00 3.6140361e-01
-5.7897884e-01 -6.5160817e-01 -4.9340004e-01 -8.9053148e-01
1.8139972e-01 2.2207570e-01 4.7862396e-01 3.7267611e-03
2.2292724e-01 -3.6637211e-01 -3.8326162e-01 -5.8509851e-01
-7.2055280e-01 4.6856355e-02 1.6941904e-01 7.6258756e-02
5.2141351e-01 -3.8152546e-01 -5.7103115e-01 4.5644441e-01
-8.1176072e-01 -1.4313813e-01 4.5722771e-01 7.7746290e-01
5.6758875e-01 5.8445007e-01 1.0060564e+00 -5.9207481e-01
8.8502163e-01 -3.4251314e-01 -5.5558509e-01 1.3857366e-01
-6.7611814e-01 1.6867544e-01 9.2414385e-01 -8.0449253e-01
-1.2444030e+00 4.8818629e-02 -5.0446349e-01 -3.3904758e-01
1.5939283e-01 3.3696479e-01 -5.9892380e-01 2.0822589e-01
1.0702310e-01 3.0417535e-01 -1.6825835e-01 -7.5635034e-01
2.7588734e-03 1.2566068e+00 4.1930366e-01 -1.7377269e-01
6.6583294e-01 -2.9805779e-01 -4.4039577e-01 -1.0740391e+00
-3.1107596e-01 -3.9301240e-01 -1.1375861e-01 -1.4194769e-01
6.0443896e-01 -8.1235272e-01 -4.0402529e-01 4.9340406e-01
-1.3242502e+00 5.3902216e-02 -7.8775808e-03 7.5384945e-01
-6.6220671e-02 5.1022512e-01 -4.5920148e-01 -1.1689616e+00
-4.5939913e-01 -1.2107742e+00 4.4023708e-01 3.0862051e-01
-3.9464960e-01 -4.0800118e-01 -2.8575709e-01 2.9115942e-01
5.0179809e-01 -3.3548602e-01 9.6150595e-01 -6.9481856e-01
-3.8032749e-01 -2.1335496e-01 1.7925082e-01 9.5065504e-01
5.1580238e-01 6.1744988e-02 -1.2583959e+00 9.5352560e-02
3.4576708e-01 3.9261022e-01 7.4508929e-01 4.1596743e-01
9.1721922e-01 -4.2118242e-01 -5.2338976e-02 2.8348142e-01
1.1822164e+00 9.4293469e-01 6.3861477e-01 -3.2877911e-02
6.4248776e-01 3.6478862e-01 7.8466189e-01 3.4752408e-01
-2.6061818e-02 8.4710163e-01 2.0236377e-01 5.0738189e-02
-4.2108420e-01 -4.2234042e-01 5.1346684e-01 1.3420082e+00
3.0606619e-01 -4.2508149e-01 -4.7303811e-01 5.3845990e-01
-1.5072188e+00 -8.0798024e-01 2.7276492e-02 2.5893767e+00
1.0611216e+00 6.8997681e-01 -1.3303220e-01 5.1227897e-01
1.3305137e+00 1.7084731e-01 -3.3220553e-01 -6.5692425e-01
-3.4402406e-01 2.1693227e-01 5.3381479e-01 8.8216621e-01
-1.0167129e+00 1.0915756e+00 5.3728714e+00 1.5590173e+00
-1.2910939e+00 1.2866828e-01 4.7926965e-01 -1.7593034e-01
-2.9267925e-01 2.4802992e-02 -1.0783074e+00 1.0085727e+00
7.3144859e-01 -1.4348233e-01 3.6165458e-01 6.3410848e-01
5.7137942e-01 -1.3892289e-01 -7.5138193e-01 7.6452011e-01
-1.7657207e-01 -7.5564158e-01 1.0253670e-01 -7.9481646e-02
3.8041681e-01 -1.2589817e-01 -8.7971706e-04 2.9907840e-01
-5.9976380e-02 -6.4885563e-01 1.0062178e+00 2.2483175e-01
6.8936157e-01 -1.0467075e+00 6.4697939e-01 4.1523042e-01
-1.1959728e+00 -1.6478330e-02 -1.9662476e-01 -1.4413664e-02
3.0812156e-01 6.3333488e-01 -1.1020552e+00 1.8217947e-01
1.3727704e-01 2.3592056e-03 -1.0208893e-01 1.1130954e+00
-2.7871305e-01 1.0194459e+00 -4.5031103e-01 -1.1965666e-02
6.2852450e-02 1.4818263e-01 9.6476269e-01 1.4164451e+00
3.1241703e-01 -1.4820604e-01 2.3313791e-02 4.5757264e-01
2.4995362e-02 1.6966511e-01 -2.1610409e-01 -5.1299360e-02
1.0155643e+00 5.7987219e-01 -2.9950202e-01 -1.6555420e-01
-3.1687915e-01 8.5408789e-01 -1.5009964e-01 4.4243684e-01
-8.2428771e-01 -7.4028760e-01 5.6512922e-01 1.0695888e-01
4.8418096e-01 4.8952240e-02 -2.1941729e-01 -6.8860549e-01
1.0459976e-01 -8.7382394e-01 8.7938972e-02 -4.1690069e-01
-9.1883528e-01 6.0204303e-01 -2.0228525e-01 -1.2070006e+00
-1.6343866e-01 -1.1771208e-01 -7.7021730e-01 1.2472061e+00
-1.4681516e+00 -6.0544670e-01 4.0151027e-01 2.9384148e-01
8.9550781e-01 -1.6588725e-01 4.9817115e-01 7.2407186e-01
-6.9827443e-01 1.1189985e+00 5.1586065e-02 -4.5589181e-03
9.3729663e-01 -9.9054480e-01 2.4419568e-01 1.3167906e+00
9.5939433e-04 6.2951791e-01 7.1970832e-01 -8.5083658e-01
-7.3546439e-01 -1.1078105e+00 1.1556542e+00 6.5064602e-02
1.1058018e-01 -2.2185220e-01 -7.9362941e-01 3.4610096e-01
8.5361801e-02 -5.5023390e-01 5.5062068e-01 -3.0180916e-01
7.1741389e-03 -4.4669914e-01 -9.6920842e-01 7.8383934e-01
5.3919572e-01 -4.9293631e-01 -7.0370007e-01 -2.3535916e-01
1.2853488e+00 -2.6452023e-01 -4.4807830e-01 4.7360781e-01
6.6530901e-01 -7.2310728e-01 7.3607814e-01 -3.3130553e-02
1.6081782e-01 -5.1694536e-01 -3.5725880e-01 -1.2739921e+00
-1.5410663e-01 -9.4817531e-01 -2.9025263e-01 1.7070549e+00
6.6922480e-01 -4.1513497e-01 3.8626412e-01 3.3450216e-01
-4.5154753e-01 -6.2678057e-01 -1.0233818e+00 -9.1716009e-01
-2.7151066e-01 -6.2893772e-01 6.9625551e-01 5.9501243e-01
-1.2087445e-01 4.3416864e-01 -7.5698787e-01 1.6005029e-01
1.9779894e-01 -4.7062293e-01 2.6852864e-01 -6.6323984e-01
-5.8502764e-01 -4.3397442e-01 -1.1683915e-01 -1.2870816e+00
-2.1834151e-01 -5.3975803e-01 3.6765081e-01 -9.3102682e-01
-3.9584342e-01 -5.6551468e-01 -8.2792932e-01 3.5044199e-01
-6.0720950e-01 -1.1524009e-01 5.1955193e-01 2.2669362e-02
-3.0712737e-02 5.8404619e-01 1.1277920e+00 1.7935688e-02
-5.1352340e-01 4.6336681e-01 -4.8551136e-01 7.1389914e-01
9.8145497e-01 -6.8796849e-01 -4.4047099e-01 -4.4368011e-01
-2.6247966e-01 1.6423003e-01 -2.3202968e-01 -9.6820939e-01
1.6728967e-01 -1.5271826e-01 9.5269710e-02 -8.6741155e-01
4.6899951e-01 -9.3775022e-01 3.4378145e-02 5.9145772e-01
-1.7018396e-01 -8.1058718e-02 3.6924195e-01 5.3363049e-01
-3.5926569e-01 -4.4850501e-01 6.6377449e-01 1.0724863e-01
-5.1726007e-01 -9.1612481e-02 -5.9252697e-01 -2.2889238e-02
7.2576296e-01 -3.9919278e-01 -8.3843641e-02 -3.0080330e-01
-2.7251109e-01 1.6569522e-01 -1.4182309e-02 6.3778150e-01
4.2422813e-01 -1.1933659e+00 -4.6023715e-01 3.4902880e-01
-1.6804028e-01 -3.0129430e-01 1.6709979e-01 3.6645216e-01
-2.5644454e-01 4.2456174e-01 1.4084460e-01 -1.2231241e-01
-1.4673253e+00 2.3166731e-01 2.3338285e-01 7.6142915e-02
-2.0058429e-01 9.1666210e-01 1.4000151e-01 -1.5511926e-01
7.7156246e-01 -5.1358145e-01 -1.5745075e-01 -1.7827818e-01
3.4858796e-01 5.1967716e-01 3.5909817e-02 -6.5737915e-01
-3.8383862e-01 2.3962072e-01 -1.9712679e-01 -3.3615050e-01
6.7447865e-01 -3.6500522e-01 2.3175719e-01 6.2590760e-01
8.1529200e-01 6.7197889e-01 -1.1761311e+00 -2.7614710e-01
-7.1681507e-02 -3.5797548e-01 1.9498962e-01 -1.0401602e+00
-9.2028040e-01 9.2370600e-01 5.1074404e-01 2.1250809e-02
1.1415128e+00 -5.2303052e-01 1.4128898e+00 -1.0028795e-01
-6.4511500e-02 -1.4724282e+00 -2.4535650e-01 6.9389445e-01
5.3586948e-01 -1.0497814e+00 -4.9911255e-01 -4.0994757e-01
-7.2263443e-01 9.1378057e-01 8.0497509e-01 2.3259220e-01
7.5912082e-01 3.7711444e-01 3.2612294e-01 5.1016200e-01
-5.7146823e-01 2.2970047e-02 1.7966262e-01 3.9984614e-01
4.5965669e-01 2.5807998e-01 -7.0517665e-01 1.0063732e+00
-5.1667929e-01 -2.0866117e-01 4.1513571e-01 8.4792018e-01
-7.1450490e-01 -1.3430660e+00 -4.9482369e-01 4.0362373e-01
-8.5927474e-01 -6.2094700e-01 -2.7571476e-01 3.4155604e-01
5.3971153e-01 1.2897865e+00 -1.2604874e-02 -8.6825830e-01
2.6464134e-01 3.7237042e-01 -1.9044016e-02 -4.5806882e-01
-7.0768863e-01 6.8200147e-01 9.0163246e-02 9.1500349e-02
1.9751234e-01 -2.0899118e-01 -1.3443593e+00 -5.0458182e-02
-9.3597120e-01 2.6208714e-01 6.1583614e-01 1.0707512e+00
1.9210964e-01 7.9488450e-01 7.9177046e-01 -5.0448544e-02
-7.9468155e-01 -1.3457980e+00 -5.3326809e-01 2.0091961e-01
5.4096860e-01 -3.8095111e-01 -4.4772741e-01 -1.1118881e-02] | [14.703180313110352, 6.6198625564575195] |
5682e5fd-b794-462e-89dd-96f7248bfa3e | scideberta-learning-deberta-for-science | null | null | https://ieeexplore.ieee.org/abstract/document/9791256 | https://ieeexplore.ieee.org/abstract/document/9791256 | SciDeBERTa: Learning DeBERTa for Science Technology Documents and Fine-Tuning Information Extraction Tasks | Deep learning-based language models (LMs) have transcended the gold standard (human baseline) of SQuAD 1.1 and GLUE benchmarks in April and July 2019, respectively. As of 2022, the top five LMs on the SuperGLUE benchmark leaderboard have exceeded the gold standard. Even people with good general knowledge will struggle to solve problems in specialized fields such as medicine and artificial intelligence. Just as humans learn specialized knowledge through bachelor’s, master’s, and doctoral courses, LMs also require a process to develop the ability to understand domain specific knowledge. Thus, this study proposes SciDeBERTa and SciDeBERTa(CS) as a pre-trained LM (PLM) specialized in the science technology domain. We further pretrain the DeBERTa, which was trained with a general corpus, with the science technology domain corpus. Experiments verified that SciDeBERTa(CS) continually pre-trained in the computer science domain achieved 3.53% and 2.17% higher accuracies than SciBERT and S2ORC-SciBERT, respectively, which are science technology domain specialized PLMs, in the task of recognizing entity names in SciERC dataset. In the JRE task of the SciERC dataset, SciDeBERTa(CS) demonstrated a 6.7% higher performance than baseline SCIIE. In the Genia dataset, SciDeBERTa achieved the best performance compared to S2ORC-SciBERT, SciBERT, BERT, DeBERTa and SciDeBERTa(CS). Furthermore, re-initialization technology and optimizers after Adam were explored during fine-tuning to verify the language understanding of PLMs. | ['Eunhui Kim', 'Yuna Jeong'] | 2022-06-08 | null | null | null | ieee-access-2022-6 | ['general-knowledge', 'joint-entity-and-relation-extraction'] | ['miscellaneous', 'natural-language-processing'] | [-4.35214847e-01 -1.25207230e-01 -8.63627344e-02 -3.24632347e-01
-6.72310472e-01 -8.76361549e-01 6.15076602e-01 -1.51117012e-01
-7.31764376e-01 9.49687958e-01 -2.69510180e-01 -6.46170139e-01
-2.32848987e-01 -8.09648097e-01 -1.26093233e+00 -1.02169193e-01
1.66859061e-01 6.81443751e-01 -1.84263363e-01 -4.58951622e-01
7.64760235e-03 5.05234420e-01 -1.15488446e+00 4.98871207e-01
1.10038662e+00 9.17370856e-01 3.89902681e-01 6.00946069e-01
-3.36054683e-01 9.29027736e-01 -9.33858216e-01 -8.12688828e-01
2.32019842e-01 -1.54966757e-01 -1.20634317e+00 -7.42503941e-01
7.04872727e-01 2.73380905e-01 -4.79773283e-01 8.87249470e-01
3.59500200e-01 1.96062490e-01 4.86149848e-01 -1.20968366e+00
-9.57535803e-01 1.27089596e+00 -3.85251611e-01 1.31491646e-01
2.42081761e-01 9.49712992e-02 9.11120296e-01 -8.71287048e-01
5.49042642e-01 1.19956219e+00 6.05953276e-01 6.59629583e-01
-6.45360589e-01 -1.09000337e+00 2.50187814e-01 3.06274503e-01
-1.49855304e+00 9.71544906e-02 1.74378559e-01 -4.00404662e-01
1.19350147e+00 1.76200509e-01 2.84523576e-01 1.19992673e+00
3.24726850e-01 8.51788759e-01 1.15718555e+00 -2.71343857e-01
-1.10626735e-01 3.88840228e-01 2.28922471e-01 7.68957317e-01
3.47628146e-01 -8.21011811e-02 -7.10421979e-01 2.96614200e-01
6.48751616e-01 -2.99818724e-01 -1.68739155e-01 4.46839988e-01
-1.52346683e+00 2.77182430e-01 4.16496038e-01 5.36452651e-01
-3.80156130e-01 -4.52095605e-02 2.33356774e-01 6.42739117e-01
3.75694335e-02 1.25315464e+00 -1.12271059e+00 -2.85578847e-01
-7.98865438e-01 1.69413835e-01 1.20172751e+00 1.52603734e+00
4.50534761e-01 2.56979227e-01 -2.06452563e-01 6.09453857e-01
-6.11718744e-02 5.81944108e-01 6.88544571e-01 -7.16694593e-01
6.24447644e-01 9.66350496e-01 -1.86865479e-01 -5.42536914e-01
-2.85465062e-01 -1.07725585e+00 -9.82500792e-01 -2.09479362e-01
3.49158019e-01 -4.99386072e-01 -1.14354706e+00 1.52070153e+00
-3.67610693e-01 5.49114227e-01 4.85790253e-01 6.95574641e-01
1.34958839e+00 9.79569077e-01 3.21638077e-01 4.04688954e-01
1.22452462e+00 -1.15266252e+00 -2.18076289e-01 -4.25273150e-01
6.62417352e-01 -7.28634477e-01 1.01849389e+00 7.51215518e-01
-1.18924236e+00 -9.00707006e-01 -1.12159383e+00 -1.23750970e-01
-8.48435283e-01 4.25762087e-01 5.17683804e-01 3.40922385e-01
-9.79826510e-01 5.77170432e-01 -3.99315357e-01 -3.24891984e-01
-6.07310794e-03 4.82750982e-01 -4.77324396e-01 5.42609692e-02
-1.66005373e+00 1.25901937e+00 6.73828661e-01 -1.03880651e-01
-1.14954805e+00 -1.22163510e+00 -6.59327388e-01 2.33544722e-01
1.21383846e-01 -6.81565344e-01 1.26309741e+00 -4.71972167e-01
-1.62098563e+00 1.15516996e+00 1.66395232e-01 -7.75180519e-01
5.03294647e-01 -5.93644202e-01 -6.67369962e-01 -2.12321043e-01
-9.36677773e-03 6.07377052e-01 -2.94677094e-02 -9.60377991e-01
-8.19224238e-01 1.37332261e-01 3.82803917e-01 -5.21522854e-03
-2.83122301e-01 1.19494967e-01 -6.17328703e-01 -5.65979898e-01
-3.39423448e-01 -7.26235509e-01 5.02715223e-02 -8.27976644e-01
-3.84180963e-01 -4.87524748e-01 3.81730646e-01 -7.82982230e-01
1.41478074e+00 -1.76002765e+00 2.26061061e-01 1.47634983e-01
1.21922120e-01 7.33955204e-01 -2.82324821e-01 1.64758891e-01
-1.75113425e-01 4.36451674e-01 -2.63495352e-02 -6.52918220e-02
1.25918955e-01 1.39751449e-01 -2.44372457e-01 5.32745086e-02
8.62552077e-02 1.15185606e+00 -1.10786736e+00 -2.57684022e-01
-2.93672197e-02 2.27867201e-01 -2.26805344e-01 3.00888777e-01
-3.90695244e-01 4.66458976e-01 -5.54003239e-01 5.62579274e-01
4.36467499e-01 -5.76148987e-01 4.46714982e-02 -3.25589515e-02
-3.03921103e-01 4.44545835e-01 -8.75240982e-01 1.71999645e+00
-1.08580220e+00 6.44923568e-01 -6.51012287e-02 -1.13480532e+00
1.22657073e+00 4.22820449e-01 3.32740486e-01 -8.09039235e-01
2.08980218e-01 4.27545458e-01 9.17480886e-02 -3.96499127e-01
4.94250923e-01 3.51931825e-02 -1.19460061e-01 1.79347366e-01
2.20949769e-01 -2.08444417e-01 2.57724047e-01 2.93662965e-01
9.72300410e-01 2.13112265e-01 -1.72140345e-01 -5.65553963e-01
1.12003374e+00 2.42831185e-02 5.92040658e-01 8.21047783e-01
1.99773118e-01 5.05312383e-01 9.46416631e-02 -3.28914136e-01
-7.37607896e-01 -9.36033189e-01 -9.08418000e-02 1.26908541e+00
1.27066553e-01 -3.95236641e-01 -7.00973034e-01 -7.55410314e-01
-4.30140309e-02 1.25071728e+00 -1.58663690e-01 -2.33492538e-01
-9.57112491e-01 -5.34040689e-01 9.94119883e-01 5.97566009e-01
1.07336342e+00 -1.08383954e+00 -6.34907633e-02 1.95947781e-01
-4.25440595e-02 -1.42011893e+00 -2.64362454e-01 -6.33341121e-03
-4.08303618e-01 -8.77776623e-01 -9.87185419e-01 -1.25002253e+00
4.02773470e-01 -3.13911229e-01 1.57163739e+00 4.68346886e-02
-1.38905972e-01 5.45439243e-01 -2.95244992e-01 -5.17701805e-01
-4.60880011e-01 6.22362792e-01 1.41640335e-01 -3.80221874e-01
6.31548762e-01 -2.85780013e-01 -1.45134285e-01 3.30709934e-01
-5.28453052e-01 1.37548313e-01 7.80872345e-01 7.80155778e-01
3.69632006e-01 -2.79485472e-02 7.89198875e-01 -9.08228099e-01
4.47447211e-01 -5.94692886e-01 -7.13422596e-01 8.54014814e-01
-7.83948004e-01 1.13316223e-01 8.73287261e-01 -5.34348667e-01
-1.18013191e+00 -3.07083726e-01 -1.09182656e-01 -4.43171263e-01
-4.42286618e-02 9.47778761e-01 -2.51747906e-01 -1.38432950e-01
7.46386468e-01 2.43624151e-01 -6.71194315e-01 -7.73611844e-01
2.00831801e-01 7.97753930e-01 9.04843748e-01 -1.35408854e+00
6.68650448e-01 -3.11980575e-01 -2.89127737e-01 -6.70800328e-01
-1.08448339e+00 -4.02990252e-01 -4.92542624e-01 -3.17941024e-03
8.69713902e-01 -1.26760101e+00 -8.97975504e-01 6.44344270e-01
-1.32279277e+00 -4.52547669e-01 2.56541353e-02 7.81395495e-01
-3.99285592e-02 -2.94216070e-03 -7.71583378e-01 -3.52761865e-01
-5.35757899e-01 -1.07723081e+00 7.99891055e-01 7.02393413e-01
-1.51165754e-01 -1.11887801e+00 -3.04000586e-01 5.67709506e-01
4.49896812e-01 -1.11299031e-03 9.35356438e-01 -9.67140555e-01
-5.77732384e-01 -1.54309422e-01 -1.74802586e-01 8.18829775e-01
-2.28294462e-01 -2.06195906e-01 -6.97496176e-01 -2.14203998e-01
-2.29744345e-01 -3.71196389e-01 5.56864560e-01 -5.35914525e-02
1.29219675e+00 -1.56303540e-01 -1.83005124e-01 9.52439845e-01
1.06566441e+00 3.28407407e-01 5.27950943e-01 9.01022077e-01
6.87796295e-01 3.26647609e-01 4.03581560e-01 -5.34051359e-02
6.62144482e-01 4.36025947e-01 -8.82142112e-02 6.26864806e-02
-2.34241158e-01 -4.00058120e-01 6.10986590e-01 1.25433409e+00
-2.29649916e-01 -2.88147599e-01 -1.44693589e+00 6.39728963e-01
-1.53559375e+00 -4.55533653e-01 -2.93624938e-01 1.92575598e+00
1.22530794e+00 2.74780303e-01 -5.28480887e-01 -5.42888045e-01
6.19833529e-01 -3.97991985e-01 -6.55228376e-01 -4.99233216e-01
-5.47971249e-01 6.89369440e-01 7.58336782e-01 2.85128921e-01
-9.85863268e-01 1.49717283e+00 5.31360483e+00 9.47066963e-01
-1.19606733e+00 -9.97524559e-02 6.02092028e-01 1.92500278e-01
-8.38915333e-02 -1.34609923e-01 -1.27574337e+00 2.86703408e-01
1.44053364e+00 -6.69771373e-01 4.24223512e-01 8.99106622e-01
-4.54783946e-01 4.66925919e-01 -1.27483249e+00 9.75751340e-01
-7.47256652e-02 -1.42093468e+00 1.51752487e-01 -2.87927002e-01
1.28341067e+00 2.67071813e-01 1.25470057e-01 1.06008875e+00
7.55357563e-01 -1.51745462e+00 5.71686625e-01 6.51935279e-01
7.65636086e-01 -6.22804463e-01 9.82680619e-01 5.15271306e-01
-8.67462993e-01 1.90588698e-01 -3.82153690e-01 -5.93352281e-02
-1.31426141e-01 3.71031761e-01 -6.72425151e-01 1.05826831e+00
8.44318271e-01 7.34991968e-01 -5.09168983e-01 1.00169218e+00
-7.05745459e-01 9.67542112e-01 -1.42489374e-01 -5.33141978e-02
4.37020659e-01 2.65936926e-02 3.52138251e-01 1.49266458e+00
4.17420417e-01 2.93593973e-01 1.46858349e-01 9.65902746e-01
-6.68998718e-01 -1.30389929e-01 7.31999204e-02 -5.32818019e-01
6.26176476e-01 1.13590157e+00 -3.68444435e-02 -5.83276629e-01
-3.92409950e-01 9.12504494e-01 1.17869914e-01 5.00426948e-01
-8.31893086e-01 -6.54483020e-01 4.54715312e-01 -1.18824430e-01
-9.23402756e-02 -2.66104192e-01 -4.53925610e-01 -1.16391349e+00
-2.66444534e-01 -1.34219313e+00 4.30423468e-01 -7.04546630e-01
-1.52034426e+00 9.06434298e-01 -2.71090791e-02 -9.60819364e-01
-1.21864721e-01 -1.02027559e+00 -5.29743791e-01 1.36155748e+00
-1.42606854e+00 -1.31884885e+00 -1.77143440e-01 4.47201580e-01
4.01951224e-01 -9.23801839e-01 7.05386281e-01 5.94867229e-01
-7.02262342e-01 1.06524825e+00 1.91159338e-01 5.51482856e-01
8.25548112e-01 -1.16520941e+00 6.95032358e-01 6.45571172e-01
-2.31377352e-02 1.09333754e+00 3.90690327e-01 -7.20962584e-01
-1.60887253e+00 -1.20575237e+00 1.23348677e+00 -6.24150455e-01
8.01392972e-01 -1.55783087e-01 -1.00680482e+00 1.01969433e+00
3.04954112e-01 -4.92839068e-01 3.60221446e-01 7.09546879e-02
-3.29700917e-01 -1.88107103e-01 -9.35416400e-01 2.73589015e-01
9.16658223e-01 -5.84259152e-01 -8.52815270e-01 4.52681452e-01
8.51881266e-01 -6.83493435e-01 -1.34158921e+00 6.73933268e-01
4.26255673e-01 -2.26830363e-01 1.13211977e+00 -1.09105313e+00
6.17955029e-01 -1.75519139e-01 4.40087579e-02 -1.25596189e+00
-1.45647556e-01 -5.78747571e-01 -6.38752580e-02 1.32465696e+00
7.11344004e-01 -4.07205850e-01 7.48876333e-01 5.83274782e-01
-3.83384049e-01 -7.02537298e-01 -5.28839588e-01 -1.20066404e+00
7.74608076e-01 -2.58811980e-01 7.46621907e-01 1.32920146e+00
-2.63139218e-01 3.26145530e-01 -1.46019235e-01 2.39238366e-01
1.88384563e-01 7.36373886e-02 7.19924331e-01 -1.20155585e+00
-2.94328898e-01 -6.88410878e-01 -1.38748750e-01 -1.12802672e+00
6.54707551e-01 -1.32554305e+00 -2.36228108e-01 -1.68912089e+00
-1.81970656e-01 -4.84756321e-01 -7.63165534e-01 6.98390484e-01
-1.81092978e-01 -2.00969622e-01 2.58324355e-01 -1.00317687e-01
-4.82594430e-01 1.37667209e-01 1.23383427e+00 -4.36597615e-01
-2.95612551e-02 3.37595791e-02 -8.14270556e-01 7.32240319e-01
7.69872606e-01 -4.64845933e-02 -1.00825600e-01 -8.93685877e-01
4.39618647e-01 -2.49565095e-01 1.65056422e-01 -1.10032260e+00
7.17233360e-01 -2.18353271e-01 3.58579248e-01 -2.99289733e-01
-3.86141948e-02 -4.66480792e-01 -3.99225131e-02 3.01637858e-01
-4.31666762e-01 1.83753714e-01 6.10921204e-01 -3.07673573e-01
-4.09642905e-01 -3.99258643e-01 6.58202112e-01 -4.91886407e-01
-1.03671920e+00 1.43188819e-01 -1.62644714e-01 3.78627390e-01
7.28468657e-01 2.28463009e-01 -5.28774500e-01 -2.79156357e-01
-6.00363970e-01 7.77210951e-01 -6.90228418e-02 6.83890700e-01
4.93390888e-01 -9.18953180e-01 -1.11582005e+00 -2.85406616e-02
-1.48548663e-01 2.15387806e-01 6.27748221e-02 4.52240944e-01
-8.17719162e-01 8.49906206e-01 -1.75842687e-01 -9.99468043e-02
-6.95292652e-01 1.44875333e-01 4.24448818e-01 -5.83764911e-01
-2.30910182e-01 1.47561753e+00 2.33959034e-01 -9.70232904e-01
6.03055179e-01 -3.24775487e-01 -2.71727502e-01 -3.14199001e-01
4.89315361e-01 6.90109134e-01 2.62012184e-01 -4.36792940e-01
-4.47722703e-01 5.19156158e-01 -2.28009909e-01 1.30199730e-01
1.36687946e+00 3.70215595e-01 -1.26715004e-01 2.45961457e-01
1.15359509e+00 -6.84010983e-02 -4.31804478e-01 -2.40934953e-01
4.02313769e-01 2.70903230e-01 7.32192025e-02 -1.71587145e+00
-1.10545743e+00 9.18638885e-01 2.68969446e-01 -4.42858756e-01
8.92459273e-01 -1.51600122e-01 9.83438790e-01 7.69871175e-01
6.56287670e-01 -9.87602174e-01 -4.24966753e-01 1.13094413e+00
1.05031216e+00 -1.07164216e+00 -1.12955838e-01 -6.99982718e-02
-6.16787076e-01 1.10903704e+00 1.14725685e+00 1.63365975e-01
5.29329479e-01 1.77489311e-01 1.74646825e-01 -2.20820718e-02
-7.81218827e-01 1.17559303e-02 7.83793807e-01 4.88382280e-02
8.06465864e-01 5.70863299e-02 -3.30901295e-01 1.18269682e+00
-8.54289412e-01 1.05113678e-01 2.15316340e-01 7.20931232e-01
-1.06203943e-01 -1.20058191e+00 -1.16798595e-01 2.04937756e-01
-6.30377531e-01 -3.37957978e-01 -6.17909431e-01 9.61600184e-01
4.36515570e-01 8.38057995e-01 -1.57445431e-01 -6.25714779e-01
6.45776510e-01 1.79702505e-01 2.47642770e-01 -7.20303416e-01
-1.18999040e+00 -5.72368085e-01 1.07682705e-01 -1.15477532e-01
-7.87454844e-02 -1.00062080e-01 -1.61727989e+00 -4.26124007e-01
-4.70383838e-02 6.99415028e-01 7.87322521e-01 9.01419520e-01
4.40664083e-01 7.54588902e-01 -5.73399886e-02 -3.73770744e-02
-5.21847367e-01 -1.05868340e+00 -2.04211637e-01 2.66711146e-01
-1.33674502e-01 -2.87833273e-01 -2.10981175e-01 5.33846729e-02] | [10.134785652160645, 8.773910522460938] |
4836162e-df9a-4356-9ac3-374f1bc44330 | mlrmbo-a-modular-framework-for-model-based | 1703.03373 | null | http://arxiv.org/abs/1703.03373v3 | http://arxiv.org/pdf/1703.03373v3.pdf | mlrMBO: A Modular Framework for Model-Based Optimization of Expensive Black-Box Functions | We present mlrMBO, a flexible and comprehensive R toolbox for model-based
optimization (MBO), also known as Bayesian optimization, which addresses the
problem of expensive black-box optimization by approximating the given
objective function through a surrogate regression model. It is designed for
both single- and multi-objective optimization with mixed continuous,
categorical and conditional parameters. Additional features include multi-point
batch proposal, parallelization, visualization, logging and error-handling.
mlrMBO is implemented in a modular fashion, such that single components can be
easily replaced or adapted by the user for specific use cases, e.g., any
regression learner from the mlr toolbox for machine learning can be used, and
infill criteria and infill optimizers are easily exchangeable. We empirically
demonstrate that mlrMBO provides state-of-the-art performance by comparing it
on different benchmark scenarios against a wide range of other optimizers,
including DiceOptim, rBayesianOptimization, SPOT, SMAC, Spearmint, and
Hyperopt. | ['Michel Lang', 'Jakob Richter', 'Daniel Horn', 'Janek Thomas', 'Jakob Bossek', 'Bernd Bischl'] | 2017-03-09 | null | null | null | null | ['smac-1', 'smac'] | ['playing-games', 'playing-games'] | [-6.68802679e-01 -4.32630479e-01 -2.77967632e-01 -4.69581425e-01
-1.04243743e+00 -5.09341180e-01 2.39923209e-01 9.98672917e-02
-4.00698125e-01 8.17154706e-01 -1.14553213e-01 -3.55976909e-01
-5.06300926e-01 -2.11724266e-01 -5.04724085e-01 -8.59981179e-01
-7.68348798e-02 9.17567611e-01 -2.61544466e-01 -1.70487657e-01
2.10776955e-01 5.70379615e-01 -1.22659647e+00 -1.45234078e-01
1.01600730e+00 1.08724713e+00 2.10015789e-01 6.90881252e-01
2.92592674e-01 2.87903607e-01 -6.11273289e-01 -4.59642977e-01
2.91639626e-01 -2.20112950e-02 -3.11627865e-01 -4.02798623e-01
6.49031997e-02 1.21130832e-01 2.55258474e-02 7.45346963e-01
9.97925878e-01 4.38560665e-01 5.79962611e-01 -1.11360216e+00
-3.43804955e-01 5.29747784e-01 -5.29600799e-01 -3.80359590e-02
2.65476376e-01 7.50954211e-01 9.30146992e-01 -1.12969398e+00
8.11305270e-02 1.64024866e+00 7.34152675e-01 2.13677380e-02
-1.63877130e+00 -3.50985557e-01 1.56273499e-01 -3.99975991e-03
-1.59720933e+00 -2.63731778e-01 2.68147588e-01 -4.64829445e-01
1.05772173e+00 7.67699778e-01 3.83358061e-01 7.63751149e-01
4.44045454e-01 7.31634617e-01 1.17588210e+00 -7.19942376e-02
3.25478494e-01 1.32002145e-01 9.21321362e-02 7.51711965e-01
-2.97570750e-02 3.63741100e-01 -7.44402409e-01 -4.63975966e-01
3.63028079e-01 -2.73490012e-01 -2.20680743e-01 -3.08075607e-01
-1.36188698e+00 8.00447345e-01 2.52940923e-01 -3.18053633e-01
-3.09305161e-01 3.47559094e-01 2.13029936e-01 1.39677480e-01
7.25686073e-01 7.45858490e-01 -8.35284770e-01 -3.40307772e-01
-1.09508526e+00 4.28771406e-01 8.23569059e-01 8.90181601e-01
6.31702602e-01 2.71908522e-01 -6.99074507e-01 1.04919529e+00
7.78112471e-01 5.60281575e-01 5.25036514e-01 -1.06458032e+00
3.02562684e-01 4.24941361e-01 2.52645880e-01 -6.84943378e-01
-7.65431046e-01 -7.68883526e-01 -7.44342983e-01 3.03124011e-01
1.25386611e-01 -2.48455927e-01 -7.90574074e-01 1.30713439e+00
5.81215024e-01 -6.44380823e-02 -1.33018598e-01 1.04273391e+00
1.03024471e+00 7.23864794e-01 -1.84081048e-01 -3.09576392e-01
1.00962853e+00 -1.24754119e+00 -6.90325022e-01 -3.78212363e-01
3.60078752e-01 -8.52493942e-01 1.16201365e+00 7.02890277e-01
-1.20183337e+00 -3.62188518e-01 -1.02038920e+00 4.19841520e-02
-3.75449419e-01 3.04267555e-01 6.17107928e-01 7.02502549e-01
-8.24342191e-01 8.76319051e-01 -1.07078457e+00 1.42404497e-01
1.59833506e-01 6.24467432e-01 -1.34711236e-01 1.46937687e-02
-7.15634584e-01 1.13671625e+00 5.50514698e-01 3.88794750e-01
-1.16862559e+00 -1.26770198e+00 -8.45640063e-01 1.73495896e-02
5.08501530e-01 -8.91363561e-01 1.09502959e+00 -3.76738429e-01
-2.15058899e+00 5.34621298e-01 -3.73514175e-01 -1.50684893e-01
5.62275469e-01 -5.02228737e-01 -2.42088482e-01 -5.06576955e-01
-2.90273875e-01 2.91002035e-01 7.82371938e-01 -7.42251456e-01
2.08808836e-02 -4.06920075e-01 -3.98339212e-01 2.81287074e-01
1.81844771e-01 5.28815806e-01 -6.94507062e-01 -8.40679169e-01
-1.91037953e-01 -8.31130207e-01 -5.81109405e-01 -2.70654410e-01
-5.97279072e-01 -3.11338976e-02 4.15975988e-01 -9.53023732e-01
1.49677348e+00 -1.95382667e+00 5.82700133e-01 3.41549367e-01
-8.00218806e-02 4.83714789e-02 -9.70681459e-02 3.40913385e-01
-1.34310380e-01 2.81231582e-01 -5.70439696e-01 -6.38884902e-01
2.12660924e-01 6.40847906e-03 1.56013995e-01 7.63737082e-01
3.07120867e-02 1.10280359e+00 -8.15179050e-01 -3.91676366e-01
4.46990460e-01 3.41895878e-01 -7.13347316e-01 5.83169699e-01
-4.77420837e-01 7.36785591e-01 -2.51774490e-01 1.07882464e+00
5.43328941e-01 -2.85794675e-01 6.97498247e-02 -2.68406626e-02
-2.20614687e-01 1.32556483e-01 -1.63204873e+00 1.76968837e+00
-7.49224782e-01 4.17403281e-01 4.34062839e-01 -6.61501467e-01
1.04578567e+00 1.14025511e-01 3.42832386e-01 4.14383505e-03
-1.88050307e-02 2.72470623e-01 -1.53107807e-01 -2.34112963e-01
3.03104281e-01 2.42659584e-01 9.64638740e-02 3.17547321e-01
2.46416420e-01 -4.64936256e-01 3.94196779e-01 -4.29070324e-01
8.36271763e-01 6.15646482e-01 5.75234413e-01 -5.05167902e-01
3.92842144e-01 -9.06021148e-02 6.91648304e-01 7.97355354e-01
3.50127697e-01 6.53954983e-01 1.99363112e-01 -2.62832284e-01
-4.34207469e-01 -1.17402029e+00 -4.33587760e-01 1.34307528e+00
-1.68238997e-01 -6.86039209e-01 -1.85583651e-01 -2.92658657e-01
4.11666244e-01 9.35256720e-01 -2.08129868e-01 -1.46890715e-01
-4.14635956e-01 -1.53571141e+00 4.76510599e-02 1.80078462e-01
1.06679350e-01 -8.59519482e-01 -3.99765819e-01 3.74459475e-01
8.25140625e-02 -6.57839835e-01 -5.00454187e-01 2.82693416e-01
-1.12079477e+00 -9.23873246e-01 -4.80248004e-01 -7.36475587e-02
5.04175067e-01 -3.28150928e-01 1.48420811e+00 -1.93357423e-01
-5.75942874e-01 3.08165967e-01 2.35703401e-02 -2.93731570e-01
-2.45722041e-01 -5.17497621e-02 8.88428539e-02 1.03061728e-01
-1.95009053e-01 -2.97562003e-01 -6.00898981e-01 8.54634941e-01
-5.59414506e-01 -8.29675719e-02 1.97107136e-01 9.33506966e-01
8.98902178e-01 -3.07520241e-01 2.77735233e-01 -3.69125009e-01
8.01826894e-01 -6.06846988e-01 -1.30925477e+00 6.60425246e-01
-8.85568500e-01 2.52110869e-01 5.89465439e-01 -3.65343779e-01
-8.20156574e-01 -1.52314350e-01 -8.50873813e-02 -4.83225673e-01
2.76275754e-01 8.79377484e-01 -3.19389224e-01 -1.53159067e-01
5.45626938e-01 -1.07445322e-01 1.22822054e-01 -9.57496703e-01
5.25092244e-01 6.27828896e-01 3.63201112e-01 -8.80051315e-01
5.10533392e-01 6.98477626e-02 1.78883418e-01 -5.95243573e-01
-7.47922659e-01 -4.86878008e-01 -3.67755681e-01 -2.25758418e-01
5.39680421e-01 -8.42556059e-01 -1.04042840e+00 4.29313123e-01
-8.49519134e-01 -6.22319758e-01 -1.53737903e-01 5.63649237e-01
-6.05142415e-01 2.23035039e-03 -3.92629690e-02 -7.55837679e-01
-5.46252728e-01 -1.63445699e+00 9.51245785e-01 3.13450396e-01
-2.82491505e-01 -1.21536434e+00 1.67691350e-01 2.39033997e-01
6.45450711e-01 4.14908886e-01 7.27886081e-01 -7.60782659e-01
-3.93119872e-01 -2.53812432e-01 3.06266904e-01 4.43772078e-01
-1.64322659e-01 4.90658522e-01 -6.38556898e-01 -5.44158995e-01
-8.33004564e-02 -3.71514499e-01 6.37686908e-01 7.87436247e-01
1.54889369e+00 -4.21995938e-01 -2.37942412e-01 1.35518718e+00
1.20357323e+00 -1.02788500e-01 2.78701395e-01 4.04585570e-01
3.92885149e-01 3.64017099e-01 9.48636651e-01 7.48916566e-01
2.26675525e-01 9.84458148e-01 7.04229534e-01 -1.07558817e-01
4.40238953e-01 1.16244152e-01 4.36561406e-01 6.93859994e-01
1.97454952e-02 -2.49255031e-01 -9.78411615e-01 -7.58427829e-02
-2.13407779e+00 -4.39409524e-01 -1.32387742e-01 2.63916421e+00
9.57721889e-01 -2.66036928e-01 1.00398280e-01 -5.17555952e-01
5.39610565e-01 1.24826699e-01 -7.02550709e-01 -4.29558635e-01
-1.19298980e-01 2.79417366e-01 6.75580204e-01 6.57893896e-01
-8.42177212e-01 8.60246778e-01 7.22244358e+00 1.01712418e+00
-9.19447422e-01 2.86459506e-01 9.05906498e-01 -7.76580334e-01
1.16192803e-01 9.60832369e-03 -1.00419438e+00 5.56302130e-01
1.12424374e+00 -2.13020757e-01 1.15881765e+00 8.57826114e-01
6.49939775e-01 -1.86282083e-01 -1.12101829e+00 9.21108067e-01
-2.79976606e-01 -1.27975845e+00 -6.38429761e-01 -1.39853910e-01
7.05528498e-01 4.48785454e-01 5.77395745e-02 4.66971815e-01
6.88532650e-01 -1.45754313e+00 6.12340212e-01 7.80023217e-01
5.52214563e-01 -7.42978454e-01 7.34283447e-01 2.52573848e-01
-7.67315030e-01 -3.64089906e-02 -3.51424843e-01 4.63240862e-01
3.86155009e-01 8.89094591e-01 -4.46436942e-01 9.27162766e-01
9.74477768e-01 6.65481329e-01 -6.42568052e-01 1.53344715e+00
-2.86265254e-01 7.47929931e-01 -7.46805191e-01 2.78365165e-02
-2.01929644e-01 -6.44274116e-01 9.25250471e-01 1.19964457e+00
4.10206825e-01 -4.17414218e-01 1.71075970e-01 1.03307843e+00
2.27898553e-01 2.95099646e-01 8.33622888e-02 2.00946495e-01
3.12449813e-01 1.45819557e+00 -1.68646842e-01 -2.78030001e-02
3.91995385e-02 3.08699012e-01 3.22420180e-01 5.51461160e-01
-1.03622556e+00 -2.49466479e-01 8.19305718e-01 -2.35229418e-01
3.37962955e-01 -3.59896690e-01 -3.17773432e-01 -1.03134382e+00
-1.92279533e-01 -1.29199505e+00 6.55346811e-01 -8.86704445e-01
-1.10597539e+00 3.31099212e-01 4.27520305e-01 -7.84999609e-01
-2.56853610e-01 -9.64577198e-01 -6.69495642e-01 1.20190573e+00
-1.31077349e+00 -6.62937105e-01 -2.95681834e-01 2.28945807e-01
3.13749135e-01 -3.85289997e-01 6.23330891e-01 7.12020472e-02
-1.16474795e+00 5.14840782e-01 7.36107111e-01 -5.19030690e-01
7.24518418e-01 -1.38567162e+00 2.04074934e-01 6.39206827e-01
-1.18259080e-01 8.79848242e-01 8.72669816e-01 -5.56805015e-01
-1.78609550e+00 -9.52636838e-01 2.05763400e-01 -4.48985964e-01
6.45513475e-01 -3.18671197e-01 -6.86283231e-01 5.33608437e-01
1.36007351e-04 1.87249497e-01 7.24587560e-01 3.75631839e-01
1.17181353e-01 -5.38297772e-01 -1.08000576e+00 6.36378706e-01
4.20207888e-01 1.83443785e-01 -2.01233864e-01 6.21955931e-01
4.92637098e-01 -9.01235700e-01 -1.48446441e+00 7.47770250e-01
4.84631360e-01 -7.82043755e-01 1.30652928e+00 -7.60027707e-01
1.66604593e-01 -5.15181780e-01 -1.29094720e-01 -1.61354363e+00
3.95999923e-02 -1.37910295e+00 -7.17202961e-01 1.12309384e+00
5.65052032e-01 -8.29515457e-01 1.43260211e-01 7.75339723e-01
-3.40226114e-01 -1.12350047e+00 -1.26594198e+00 -9.72540557e-01
-2.07776412e-01 -5.46794713e-01 9.17847157e-01 3.56588274e-01
-3.02165836e-01 -4.56313416e-02 -5.03318191e-01 2.25289121e-01
5.99273384e-01 1.16858803e-01 1.14200842e+00 -8.87148917e-01
-9.43657160e-01 -6.83777094e-01 1.24172084e-01 -1.03445482e+00
-4.85470518e-02 -9.84651387e-01 3.52099948e-02 -1.38484824e+00
-2.16533750e-01 -4.61777985e-01 -4.34836149e-01 4.23868030e-01
-4.65827703e-01 -1.18089184e-01 1.01530105e-01 1.29533529e-01
-4.49896216e-01 9.54362094e-01 1.06389785e+00 -2.77284905e-02
-5.35082519e-01 4.32003081e-01 -4.19015974e-01 6.38026655e-01
6.48508430e-01 -7.85441518e-01 5.56576625e-02 -2.17545316e-01
2.79671848e-01 1.20198451e-01 3.11099410e-01 -5.95138967e-01
-1.39662817e-01 -4.67553228e-01 2.30798289e-01 -6.53685927e-01
3.52557778e-01 -5.54557204e-01 5.29184639e-01 2.37642407e-01
-1.91722423e-01 5.25885105e-01 4.78370190e-01 2.40645528e-01
-5.43723069e-02 -4.56909418e-01 9.42018032e-01 -7.97697157e-02
-1.21459544e-01 3.21733445e-01 -6.51286915e-02 2.93301731e-01
8.85508835e-01 1.78801417e-01 -2.56035417e-01 -1.96169183e-01
-6.89428270e-01 7.99302399e-01 2.18174502e-01 2.86303639e-01
4.82670426e-01 -1.06171775e+00 -1.01859868e+00 -1.00315802e-01
3.20637412e-02 1.40778959e-01 -1.17988169e-01 1.32737577e+00
-7.00018704e-01 3.70159000e-01 4.36616838e-01 -7.60320067e-01
-1.04560602e+00 3.82653981e-01 6.54910207e-01 -6.59080923e-01
-2.21455544e-01 7.39769161e-01 -2.25947008e-01 -9.68071818e-01
2.82757491e-01 -2.51672864e-01 2.22293288e-01 -2.46144384e-01
2.81598777e-01 9.13587868e-01 1.75121725e-01 -1.19215362e-01
-4.83252943e-01 3.73544276e-01 4.13925171e-01 -1.37782753e-01
1.53736436e+00 1.26587525e-01 -5.68141222e-01 6.01392567e-01
8.74887347e-01 -1.27522320e-01 -1.32202303e+00 8.82010460e-02
-6.12213537e-02 -5.94428897e-01 3.07533324e-01 -1.25305498e+00
-9.75056887e-01 5.17089307e-01 3.72059911e-01 -3.86714339e-01
9.55386221e-01 -2.73892790e-01 -6.82317838e-02 3.74653488e-01
1.38159290e-01 -9.79401052e-01 -1.49297193e-01 3.40177268e-01
1.43002975e+00 -1.31832731e+00 7.26680756e-01 1.44533634e-01
-5.93359172e-01 1.05699241e+00 5.20793676e-01 1.15554415e-01
7.83318460e-01 8.89627263e-02 -1.15277298e-01 3.11185215e-02
-7.90597796e-01 1.36774540e-01 8.77323091e-01 2.93444604e-01
4.07049030e-01 -2.34683380e-02 -2.28828102e-01 7.09273994e-01
-2.82633960e-01 -3.89430851e-01 -4.19703424e-02 5.97545445e-01
-3.08950036e-03 -1.10069251e+00 -7.84834146e-01 4.58268762e-01
-5.44642389e-01 -3.20983708e-01 5.70774972e-02 9.13674176e-01
-5.10984302e-01 9.73398507e-01 -2.33997583e-01 -1.78255402e-02
9.99238268e-02 -2.85631996e-02 2.10006192e-01 -5.10571897e-01
-1.17971194e+00 1.77114785e-01 1.48183078e-01 -8.28049481e-01
9.50013101e-02 -5.44616461e-01 -7.03441799e-01 -1.32039145e-01
-4.64642555e-01 2.20989913e-01 1.16394532e+00 8.71485710e-01
5.51275909e-01 5.05995214e-01 5.19370973e-01 -9.72619832e-01
-9.69920218e-01 -1.17005873e+00 -2.22822085e-01 -3.02920133e-01
2.51518279e-01 -7.86721230e-01 -5.83172441e-01 -5.11731505e-01] | [6.518116474151611, 3.9694504737854004] |
eee1da0e-06f5-48b8-a75d-2dad5bf45479 | end-to-end-simultaneous-learning-of-single | 2107.02958 | null | https://arxiv.org/abs/2107.02958v1 | https://arxiv.org/pdf/2107.02958v1.pdf | End-to-End Simultaneous Learning of Single-particle Orientation and 3D Map Reconstruction from Cryo-electron Microscopy Data | Cryogenic electron microscopy (cryo-EM) provides images from different copies of the same biomolecule in arbitrary orientations. Here, we present an end-to-end unsupervised approach that learns individual particle orientations from cryo-EM data while reconstructing the average 3D map of the biomolecule, starting from a random initialization. The approach relies on an auto-encoder architecture where the latent space is explicitly interpreted as orientations used by the decoder to form an image according to the linear projection model. We evaluate our method on simulated data and show that it is able to reconstruct 3D particle maps from noisy- and CTF-corrupted 2D projection images of unknown particle orientations. | ['Daniel Ratner', 'Chuck Yoon', 'Michael Kagan', 'Geoffrey Woollard', 'Harshit Gupta', 'Frederic Poitevin', 'Youssef S. G. Nashed'] | 2021-07-07 | null | null | null | null | ['cryogenic-electron-microscopy-cryo-em'] | ['computer-vision'] | [ 3.46297145e-01 1.74459472e-01 6.60915136e-01 -4.77407515e-01
-7.60190666e-01 -5.20060897e-01 8.03702474e-01 -3.35725754e-01
-7.01841950e-01 8.08639109e-01 2.55063355e-01 -1.47941053e-01
2.39016786e-01 -5.86569309e-01 -1.16583359e+00 -1.25496256e+00
1.93916082e-01 1.36488712e+00 -1.15826271e-01 3.23078126e-01
4.15211469e-01 6.22897625e-01 -9.25041616e-01 4.64029044e-01
8.29007253e-02 3.87005329e-01 7.53586233e-01 1.08383632e+00
1.86658874e-01 5.44685841e-01 -3.14338446e-01 -1.20085321e-01
1.37336060e-01 -5.20386994e-01 -7.43280888e-01 4.57890987e-01
2.70969540e-01 -2.26989299e-01 -2.21777707e-01 7.99487770e-01
3.61986995e-01 -2.46974468e-01 1.04251897e+00 -1.76588371e-01
-7.82889187e-01 -2.46050861e-02 1.39334247e-01 5.02292477e-02
2.74642318e-01 3.93612117e-01 6.56380892e-01 -9.44981694e-01
1.49738955e+00 1.16881526e+00 3.41770768e-01 6.14938438e-01
-2.09408140e+00 1.10190563e-01 -2.45657369e-01 4.69908975e-02
-1.00044239e+00 -4.96954709e-01 6.96507335e-01 -5.61074674e-01
1.15941119e+00 -2.92157859e-01 6.88382745e-01 1.23274839e+00
9.03933883e-01 1.68172657e-01 1.60670173e+00 -4.26322550e-01
4.19639975e-01 -2.19702780e-01 2.88931374e-02 5.72540402e-01
-3.80022712e-02 2.98915714e-01 -3.60401094e-01 -1.73433930e-01
8.50106239e-01 1.31644234e-01 -1.11526974e-01 -7.88223684e-01
-1.41733122e+00 4.40872103e-01 2.58721888e-01 -3.67851034e-02
-6.66795671e-01 2.06343904e-02 -6.29757121e-02 1.72493532e-01
3.89876485e-01 3.40044439e-01 -2.88688004e-01 -3.58880870e-02
-7.09252894e-01 2.64085144e-01 9.13939178e-01 5.68661749e-01
8.53865683e-01 -2.79101342e-01 5.60066998e-01 1.73523277e-01
5.55812001e-01 8.75103772e-01 3.59125793e-01 -1.10859537e+00
9.17668566e-02 8.66499990e-02 3.65981460e-01 -4.34727013e-01
-3.42019618e-01 3.10405254e-01 -6.15121424e-01 5.93249023e-01
3.99134338e-01 -3.99425179e-02 -1.23813021e+00 1.58422697e+00
1.94784313e-01 5.08506820e-02 1.73831254e-01 9.70945239e-01
4.26470488e-01 8.78738701e-01 -3.55938941e-01 -3.69807988e-01
1.02818310e+00 -5.47393739e-01 -7.79328167e-01 -9.78592262e-02
6.37891516e-02 -5.79561949e-01 4.24941033e-01 3.99832934e-01
-1.35078418e+00 -4.03117478e-01 -1.19154215e+00 -1.79971799e-01
-3.94251853e-01 5.35305031e-02 2.83106305e-02 2.58203685e-01
-1.11278701e+00 7.79783368e-01 -1.59028029e+00 -2.38467500e-01
1.27278969e-01 4.44450825e-01 -9.37020540e-01 1.11897096e-01
-3.20865393e-01 1.07392049e+00 2.52871841e-01 -1.28508518e-02
-1.48887384e+00 -2.38846630e-01 -5.42756855e-01 -3.59929979e-01
-2.32932478e-01 -1.06023586e+00 9.02445734e-01 -2.87875921e-01
-2.05501795e+00 1.14834595e+00 -7.68525422e-01 -5.43036163e-01
2.23945782e-01 1.42586082e-01 9.36647691e-03 5.16385913e-01
-1.00185812e-01 7.52619147e-01 9.24787641e-01 -1.65611565e+00
7.29484856e-02 -7.35963523e-01 -3.46418649e-01 1.34474263e-01
5.94040692e-01 -2.05161154e-01 2.15563807e-03 1.76247885e-03
8.22661817e-01 -1.05341351e+00 -3.08281153e-01 -2.61161685e-01
-5.76371551e-01 5.38632095e-01 8.96354258e-01 -6.84965193e-01
1.15354314e-01 -1.65663493e+00 9.77732539e-01 -5.49831055e-02
4.30928707e-01 -3.03858668e-01 1.61738381e-01 6.42824590e-01
-1.46369353e-01 -3.31943303e-01 -3.02854329e-01 -8.54497433e-01
4.14294414e-02 5.05190313e-01 -3.71996969e-01 9.63755131e-01
1.93370983e-01 9.70125437e-01 -6.51557684e-01 -2.85001069e-01
5.48917055e-01 8.71530235e-01 -4.35441285e-01 4.78390992e-01
-2.13381529e-01 1.16313982e+00 -1.22891270e-01 2.40977213e-01
8.20808887e-01 -5.77346265e-01 7.81787872e-01 -2.23005876e-01
-2.72284299e-01 5.93163788e-01 -8.27050209e-01 1.64901710e+00
-7.01220855e-02 3.18957657e-01 3.81061018e-01 -8.76320779e-01
7.03212202e-01 4.50530827e-01 3.25197190e-01 -5.31815290e-01
-1.35321423e-01 1.25764996e-01 -4.35101712e-04 -2.97009021e-01
7.82319605e-02 -8.29610884e-01 2.10150868e-01 8.29312086e-01
5.12575865e-01 -2.84413129e-01 -1.26077607e-02 2.10664481e-01
1.03537309e+00 5.82571208e-01 7.50817060e-02 -4.72998738e-01
4.92543638e-01 -2.41417259e-01 1.46186739e-01 6.28739178e-01
1.46315604e-01 9.85175788e-01 2.39416420e-01 -1.05830371e+00
-2.08000064e+00 -1.57873309e+00 -4.14331913e-01 3.12843323e-01
-5.09575270e-02 -2.56296545e-01 -8.47340465e-01 -2.70623624e-01
-3.27242166e-01 1.86130747e-01 -7.56116450e-01 2.84424037e-01
-7.79668212e-01 -9.76841807e-01 4.49129231e-02 8.38468447e-02
-1.94456279e-01 -1.25222576e+00 -7.11427510e-01 3.14234585e-01
-4.60817153e-03 -1.04181445e+00 8.39924067e-02 6.25479877e-01
-1.00391579e+00 -8.86451185e-01 -6.09340668e-01 -6.36865079e-01
1.41342723e+00 1.19719595e-01 1.04714203e+00 -3.31990540e-01
-1.66785479e-01 1.08982220e-01 2.38011591e-02 4.33719121e-02
-7.18943238e-01 -2.58427650e-01 4.21595216e-01 -1.50455788e-01
3.71030807e-01 -9.46034133e-01 -5.42066216e-01 2.85351723e-01
-8.55250776e-01 3.74171466e-01 2.54126996e-01 1.09697366e+00
1.10586751e+00 -4.07641917e-01 -1.62485838e-01 -1.01959229e+00
3.01685721e-01 5.92935365e-03 -7.56835282e-01 -1.14779130e-01
-3.25047731e-01 5.10096729e-01 9.68831837e-01 -6.89587295e-02
-1.02292669e+00 4.03875709e-01 -3.28711390e-01 -2.64826357e-01
-5.29711843e-01 2.87767649e-02 -2.79961735e-01 3.65355611e-02
4.29463983e-01 7.92791426e-01 3.67035151e-01 -6.57184780e-01
2.38890931e-01 3.22558582e-01 7.29826868e-01 -7.35227048e-01
5.50481558e-01 1.30541968e+00 2.09269673e-01 -7.28792489e-01
-3.56070966e-01 -5.16211949e-02 -1.24590755e+00 -8.39713588e-03
9.47520018e-01 -7.00842559e-01 -9.23987091e-01 5.13652265e-01
-1.24333346e+00 -3.60335648e-01 -1.48670614e-01 5.82015991e-01
-1.09438944e+00 5.79807699e-01 -9.15613770e-01 -5.31249344e-01
-1.36805430e-01 -1.64306748e+00 1.45725596e+00 -2.33162940e-01
8.50230828e-02 -9.63506103e-01 6.05689168e-01 2.15747971e-02
1.16386175e-01 -7.91938975e-02 9.21512783e-01 -1.13875009e-01
-9.37516510e-01 -9.43004563e-02 1.82588175e-01 1.48651406e-01
-1.52941551e-02 -1.01398163e-01 -8.73255074e-01 -5.58546185e-01
4.97040242e-01 -3.64700377e-01 9.22044396e-01 6.76232219e-01
5.85085452e-01 -2.82719254e-01 -4.70582724e-01 8.39671433e-01
1.49037826e+00 1.89086739e-02 8.59570026e-01 2.83662885e-01
6.14616096e-01 5.94647527e-01 -1.48132637e-01 2.10624605e-01
1.74801454e-01 6.62902534e-01 4.11214292e-01 3.97902429e-01
1.23248175e-01 -3.28697085e-01 5.32034099e-01 9.90942061e-01
-3.67317468e-01 -2.95734107e-01 -6.69992805e-01 2.34720498e-01
-1.45238209e+00 -1.01411128e+00 1.91413209e-01 2.21348572e+00
6.36969924e-01 9.27141830e-02 -1.55946344e-01 -2.12512732e-01
4.38351989e-01 1.72189713e-01 -6.59258306e-01 -7.13135600e-02
-3.82314265e-01 2.47726724e-01 4.19110298e-01 1.00634181e+00
-7.76968420e-01 7.98946857e-01 8.16653633e+00 -7.50315487e-02
-1.08129227e+00 1.31005824e-01 3.78356457e-01 1.41548170e-02
-3.48213643e-01 4.55220520e-01 -9.24043417e-01 7.39432096e-01
1.28630841e+00 2.22146392e-01 6.00462258e-01 2.76791811e-01
2.18324631e-01 2.51704063e-02 -1.18614268e+00 6.38903618e-01
4.43204902e-02 -1.66685975e+00 1.59435734e-01 5.73623478e-01
5.57059884e-01 3.50296646e-01 -1.38416756e-02 -4.29195046e-01
5.35230875e-01 -9.16686356e-01 5.69751143e-01 9.19285536e-01
5.56884348e-01 -3.78712684e-01 5.94864130e-01 6.44619465e-01
-5.57107508e-01 4.74067032e-01 -8.63135576e-01 -1.15087971e-01
8.67001712e-01 4.92048681e-01 -1.08032811e+00 1.63035512e-01
4.38858330e-01 6.53878808e-01 5.20489812e-02 1.76479027e-01
-2.19034091e-01 3.37497503e-01 -5.53202152e-01 3.56503010e-01
6.04566298e-02 -8.67312014e-01 6.71742618e-01 1.01175845e+00
1.99856818e-01 1.46916077e-01 -7.93362260e-02 1.07602262e+00
-8.41466784e-02 -5.16584396e-01 -7.07052529e-01 -1.13137327e-01
1.51987076e-01 1.13451719e+00 -8.18695188e-01 -2.95292675e-01
-1.37372874e-02 1.10550904e+00 5.67544281e-01 4.00032163e-01
-3.91776204e-01 1.71365306e-01 4.38186646e-01 2.85664618e-01
7.70253122e-01 -6.81361556e-01 5.36709242e-02 -1.13941669e+00
-4.30774279e-02 -4.47889775e-01 -2.30743662e-01 -1.08847535e+00
-1.24066091e+00 4.59163666e-01 -3.39986861e-01 -8.65129530e-01
-5.13072670e-01 -1.19716597e+00 -3.86853069e-01 9.76148963e-01
-1.09762156e+00 -9.02421117e-01 3.01200271e-01 -6.43471554e-02
-9.26338509e-02 -1.63677819e-02 1.15322649e+00 -1.77737072e-01
-1.13453001e-01 -2.21860066e-01 7.71650672e-01 -1.38075978e-01
4.90577191e-01 -1.60077977e+00 6.82500422e-01 6.60678029e-01
2.79478908e-01 9.71079230e-01 1.02282333e+00 -7.85370767e-01
-1.81419563e+00 -5.67604899e-01 8.76944542e-01 -1.08770287e+00
2.97759742e-01 -6.09823883e-01 -8.93651247e-01 1.15076399e+00
4.26402211e-01 2.63001233e-01 7.25044131e-01 -3.02186042e-01
-5.37152141e-02 4.77671653e-01 -1.21962404e+00 3.24007392e-01
7.15622187e-01 -9.55145597e-01 -8.72817755e-01 4.72116500e-01
2.63719559e-01 -4.84284014e-01 -1.05442202e+00 -4.52298932e-02
6.52666330e-01 -1.14400089e+00 1.17473090e+00 -8.15854371e-01
6.03434026e-01 -5.79891026e-01 -2.62644738e-01 -1.33919060e+00
-3.59622419e-01 -5.21375537e-01 -1.87107101e-01 2.03814760e-01
2.82322973e-01 -5.38810909e-01 1.07616961e+00 3.70569766e-01
-7.44412839e-02 -5.24994135e-01 -1.15592706e+00 -3.62922430e-01
2.43998513e-01 6.06325790e-02 2.93680340e-01 3.42843115e-01
6.64759427e-02 4.74255532e-01 -4.65692610e-01 3.65644366e-01
9.22728658e-01 3.96718979e-01 8.14304769e-01 -1.04142678e+00
-4.26384956e-01 3.66379052e-01 -6.79215014e-01 -1.60568666e+00
2.87226379e-01 -7.77889132e-01 2.32204631e-01 -1.16402709e+00
6.88174367e-01 3.07285547e-01 4.12559845e-02 1.88890161e-04
3.50986212e-01 3.71234477e-01 1.17362529e-01 6.09961569e-01
-7.88214862e-01 5.51036060e-01 1.28471160e+00 3.12191769e-02
3.25819194e-01 -4.19546634e-01 -1.38621226e-01 6.53253794e-01
2.92745322e-01 -7.30151653e-01 1.45542890e-01 -5.37241399e-01
2.85989702e-01 1.53536260e-01 5.53221881e-01 -9.08263922e-01
2.99279481e-01 2.80683309e-01 7.61778891e-01 -9.83738780e-01
7.92336702e-01 -7.55930066e-01 5.97036064e-01 3.69631797e-01
-1.67179659e-01 3.18074197e-01 -2.88358152e-01 6.56101167e-01
-4.01580967e-02 -2.56329864e-01 7.92061269e-01 -6.04375541e-01
3.22339609e-02 1.37608469e-01 -8.10843050e-01 -4.24509317e-01
5.49093783e-01 -1.99000433e-01 -2.96049923e-01 -1.86599851e-01
-1.28966701e+00 -2.97510624e-01 1.41830778e+00 -3.95136982e-01
8.95166337e-01 -1.03531837e+00 -6.13878489e-01 7.01364338e-01
-1.84674650e-01 7.53501952e-02 2.65365332e-01 5.84158719e-01
-1.00879729e+00 7.97729611e-01 -5.63271761e-01 -9.38312411e-01
-9.76199210e-01 6.50322378e-01 4.83811498e-01 -4.06619132e-01
-9.72027421e-01 4.55390304e-01 3.27870905e-01 -1.04091263e+00
-5.22339523e-01 -1.35574505e-01 1.98317379e-01 -7.30976105e-01
6.71140671e-01 -6.48521259e-02 2.76211407e-02 -1.07580376e+00
-1.12664551e-01 6.86926425e-01 -3.77162069e-01 -5.03471434e-01
1.67703629e+00 -2.76966721e-01 -2.31934711e-01 4.58492875e-01
1.30120230e+00 -1.28127739e-01 -1.81643116e+00 -1.13178521e-01
-3.43594074e-01 -2.78447270e-01 -1.72186121e-01 -4.87901837e-01
-5.79116940e-01 1.02777314e+00 4.79395777e-01 -1.44832641e-01
4.78240907e-01 2.25391284e-01 5.61801910e-01 8.61467779e-01
7.45816946e-01 -7.17416108e-01 -2.87457585e-01 5.51717162e-01
3.81308049e-01 -1.12383008e+00 2.08566099e-01 8.62215161e-02
-2.42754802e-01 1.11321425e+00 -1.56869784e-01 -4.43859309e-01
5.27250648e-01 6.55423343e-01 5.80099449e-02 -5.38178802e-01
-1.08359611e+00 2.81188279e-01 -3.38839710e-01 7.57068515e-01
2.68722266e-01 7.87823126e-02 1.32899463e-01 5.04572690e-02
-1.04841612e-01 -3.94802876e-02 5.24295092e-01 1.04305446e+00
-3.61081779e-01 -1.35213661e+00 -3.71903807e-01 1.20486297e-01
-4.25463557e-01 8.23258758e-02 -4.26292837e-01 2.26008549e-01
-1.83577433e-01 1.41045719e-01 2.40782127e-01 -1.45109922e-01
1.06751621e-02 3.92533332e-01 1.03023732e+00 -5.27030051e-01
4.66985852e-02 2.67902613e-01 -2.70157903e-01 -4.54107940e-01
-7.15547562e-01 -8.78208578e-01 -1.40103495e+00 -2.13652685e-01
-2.76342500e-02 4.04777974e-01 8.11195970e-01 9.55357909e-01
6.39641285e-01 1.43442169e-01 5.59980154e-01 -1.56453943e+00
-4.39500362e-01 -1.07102001e+00 -7.95548856e-01 6.78983688e-01
4.30943996e-01 -4.32978898e-01 -5.62212884e-01 6.89451694e-01] | [13.276959419250488, -3.073982000350952] |
6d0aa7ce-9bd7-4ebe-a58f-894d441da6cf | towards-fleet-wide-sharing-of-wind-turbine | 2212.03529 | null | https://arxiv.org/abs/2212.03529v2 | https://arxiv.org/pdf/2212.03529v2.pdf | Towards Fleet-wide Sharing of Wind Turbine Condition Information through Privacy-preserving Federated Learning | Terabytes of data are collected every day by wind turbine manufacturers from their fleets. The data contain valuable real-time information for turbine health diagnostics and performance monitoring, for predicting rare failures and the remaining service life of critical parts. And yet, this wealth of data from wind turbine fleets remains inaccessible to operators, utility companies, and researchers as manufacturing companies prefer the privacy of their fleets' turbine data for business strategic reasons. The lack of data access impedes the exploitation of opportunities, such as improving data-driven turbine operation and maintenance strategies and reducing downtimes. We present a distributed federated machine learning approach that leaves the data on the wind turbines to preserve the data privacy, as desired by manufacturers, while still enabling fleet-wide learning on those local data. We demonstrate in two case studies that wind turbines which are scarce in representative training data benefit from more accurate fault detection models with federated learning, while no turbine experiences a loss in model performance by participating in the federated learning process. When comparing conventional and federated training processes, the average model training time rises significantly by a factor of up to 14 in the federated training due to increased communication and overhead operations. Thus, model training times might constitute an impediment that needs to be further explored and alleviated in federated learning applications, especially for large wind turbine fleets. | ['Angela Meyer', 'Stefan Jonas', 'Lorin Jenkel'] | 2022-12-07 | null | null | null | null | ['fault-detection'] | ['miscellaneous'] | [-4.12245959e-01 2.15449244e-01 -5.40798753e-02 -1.89760268e-01
-2.04659984e-01 -6.59490108e-01 -6.27862960e-02 1.40040204e-01
-1.12459518e-01 7.86327958e-01 -4.70540076e-01 -2.64075994e-01
-7.48546839e-01 -9.80976701e-01 -1.71397507e-01 -9.36518669e-01
-6.71972871e-01 7.24296987e-01 -2.64362454e-01 1.71970017e-02
-2.72022814e-01 7.65490770e-01 -1.76893985e+00 2.58989036e-01
7.16642380e-01 1.11675072e+00 -1.38852179e-01 3.67127955e-01
7.86351860e-02 5.50194561e-01 -9.25454080e-01 -5.26881695e-01
9.26123142e-01 1.30559757e-01 -7.05842376e-01 -2.26047412e-02
-1.76233083e-01 -4.27089274e-01 -1.34602711e-01 8.94907176e-01
5.32992005e-01 1.58919737e-01 -7.93586578e-03 -1.82440317e+00
5.50499596e-02 5.99234223e-01 -1.40987113e-01 2.63823159e-02
-3.54695678e-01 3.05125535e-01 9.20854211e-01 -3.57631505e-01
1.95440993e-01 4.97025877e-01 6.98244393e-01 3.21268171e-01
-7.50497222e-01 -5.15896142e-01 -7.45020136e-02 -9.69893038e-02
-1.24777210e+00 -3.20781648e-01 5.12793839e-01 -2.58465856e-01
1.02908027e+00 8.10102105e-01 6.70896888e-01 4.02356774e-01
5.23970783e-01 3.73300731e-01 5.30000508e-01 -3.47459503e-02
6.59878731e-01 1.26625165e-01 -5.28674960e-01 3.78488660e-01
8.33314896e-01 5.52942276e-01 -6.90331042e-01 -3.19456011e-01
4.19605374e-01 3.87761801e-01 -3.57496679e-01 -6.19687736e-01
-6.95933938e-01 5.94761610e-01 2.17428952e-01 1.71403244e-01
-5.78683734e-01 -3.27803522e-01 1.04142690e+00 8.98961246e-01
6.86778426e-01 6.17335439e-01 -1.25041294e+00 -2.28805572e-01
-8.95280004e-01 -1.22058868e-01 1.08356404e+00 8.38417172e-01
7.58344293e-01 6.71380997e-01 5.69602966e-01 2.23516762e-01
-2.68180192e-01 2.92156011e-01 5.03981888e-01 -6.58796549e-01
4.47882950e-01 5.59899807e-01 2.87512302e-01 -5.85121393e-01
-3.09457302e-01 -7.55352199e-01 -1.10595810e+00 4.12556678e-01
-8.48426204e-03 -6.70001626e-01 -2.89434850e-01 8.19097221e-01
4.25587863e-01 -9.58179682e-02 -1.26295621e-02 8.72714221e-01
-1.97929710e-01 1.85367480e-01 -2.70272642e-01 -2.06462950e-01
1.03076530e+00 -8.03610623e-01 -7.68744528e-01 -5.33813536e-02
1.01612568e+00 -5.67360282e-01 3.89254749e-01 9.12641108e-01
-8.50573301e-01 -4.33625102e-01 -1.00993633e+00 7.12767422e-01
-6.64486289e-01 4.32867296e-02 9.34535384e-01 9.88106191e-01
-6.56679928e-01 1.04462278e+00 -1.08049119e+00 -1.84042856e-01
7.58257926e-01 7.38555849e-01 -6.48526251e-01 -2.24088028e-01
-9.05417264e-01 1.00721049e+00 3.28882307e-01 2.37860531e-01
-1.05366182e+00 -1.09033585e+00 -7.88521945e-01 3.56674135e-01
1.88030273e-01 -8.76689970e-01 1.28611839e+00 -2.93661684e-01
-8.67220283e-01 1.23460982e-02 5.38129508e-01 -8.85436594e-01
4.81280118e-01 -1.40596643e-01 -8.64714563e-01 -4.07720208e-02
-2.53994793e-01 -2.10071713e-01 1.10988331e+00 -1.01577914e+00
-1.17810392e+00 -4.44944173e-01 -1.37725040e-01 -9.41523686e-02
-6.00728810e-01 -3.20403874e-01 6.15342617e-01 -3.42779011e-01
-3.77601266e-01 -6.05009258e-01 -3.89853358e-01 -9.87116396e-02
1.43207297e-01 1.48076177e-01 1.65609515e+00 -6.67805552e-01
1.06187415e+00 -2.18492007e+00 -4.28692639e-01 1.92598343e-01
1.33824795e-01 4.38092887e-01 4.03175987e-02 1.01180887e+00
-3.03525358e-01 4.33711678e-01 5.73805496e-02 -2.39962399e-01
-1.22528039e-02 6.33096457e-01 -3.56628031e-01 5.86696982e-01
2.27031410e-01 3.59901965e-01 -8.35146725e-01 1.85667783e-01
6.92281365e-01 -4.18039262e-02 -1.34418771e-01 2.85841465e-01
-6.05703145e-02 4.59402919e-01 -7.11764276e-01 9.05061364e-01
8.47381592e-01 1.79519564e-01 1.70014173e-01 -1.87670082e-01
-1.09307364e-01 -1.63540527e-01 -9.33447599e-01 1.39826572e+00
-9.28541780e-01 2.66093075e-01 7.75547385e-01 -1.17905569e+00
1.06771648e+00 9.31099832e-01 1.07131541e+00 -1.46485239e-01
1.08148210e-01 1.26134992e-01 -1.71452850e-01 -4.27367270e-01
5.14692187e-01 -2.40049034e-01 -3.31686549e-02 6.58116162e-01
4.46505398e-02 -1.07445128e-01 -1.49149403e-01 -1.99911341e-01
1.36302137e+00 -4.21725392e-01 -1.13698968e-03 -9.30948630e-02
1.92852169e-01 1.54878065e-01 6.93932056e-01 1.12046078e-01
-1.45194352e-01 -6.01616800e-02 6.83602318e-02 -1.05409729e+00
-9.44927692e-01 -5.99362791e-01 -1.58463232e-02 5.45447648e-01
-4.07636166e-01 -5.81281245e-01 -2.97494143e-01 -9.29180980e-01
4.11979198e-01 6.80325627e-01 -1.51484579e-01 -4.93704647e-01
-1.97580054e-01 -4.46978927e-01 2.91024595e-01 4.47488308e-01
3.58915359e-01 -5.29895961e-01 -9.47431386e-01 1.50977805e-01
4.96201545e-01 -6.75082982e-01 -1.45888790e-01 7.62069762e-01
-1.06169748e+00 -1.12304533e+00 8.02713484e-02 -1.31579578e-01
7.71426618e-01 3.77766609e-01 9.59618032e-01 3.16399395e-01
-5.53806841e-01 1.92606032e-01 -5.54540992e-01 -6.31703973e-01
-2.02759460e-01 1.69571906e-01 3.77930671e-01 -1.42217174e-01
1.34424895e-01 -5.98866165e-01 -3.59943837e-01 5.96178353e-01
-6.71596885e-01 -6.72198951e-01 4.64845628e-01 1.25573039e+00
9.83349085e-02 1.43114853e+00 8.68334234e-01 -1.00349534e+00
2.87641436e-01 -7.35839784e-01 -8.07559788e-01 1.00437775e-01
-1.10435593e+00 -2.49285981e-01 1.21001089e+00 1.21156305e-01
-1.16239738e+00 1.25364408e-01 4.77889568e-01 -8.02238464e-01
-1.75077990e-01 7.09114254e-01 -1.98230818e-01 -2.08149537e-01
3.19603950e-01 -2.98029631e-01 4.55535561e-01 -6.98279619e-01
3.36890668e-02 9.59361732e-01 3.17377836e-01 -2.20751643e-01
1.13050425e+00 3.40056390e-01 1.24066643e-01 -7.09605455e-01
-2.32651427e-01 -5.39216995e-01 -3.37820947e-01 -4.27327417e-02
-5.97971084e-04 -1.15347493e+00 -8.29119682e-01 1.07810810e-01
-6.46575212e-01 1.87047482e-01 -1.00420237e+00 7.79365659e-01
-2.54587263e-01 8.03575739e-02 -2.97987312e-01 -1.03126943e+00
-6.31044924e-01 -6.08596325e-01 7.81917751e-01 -2.31329929e-02
-9.26835164e-02 -9.67167735e-01 -2.27044091e-01 4.65232968e-01
7.08135664e-01 3.94662291e-01 7.46264160e-01 -9.66330051e-01
-6.29166305e-01 -9.96859014e-01 5.71675479e-01 7.18137205e-01
8.68974984e-01 -1.04992621e-01 -1.08603323e+00 -1.19062734e+00
2.72388488e-01 -2.96183914e-01 1.30572408e-01 -5.30872382e-02
1.08194506e+00 -4.73956108e-01 -2.72433370e-01 4.30778831e-01
1.19595063e+00 2.00070515e-02 1.32273287e-01 -1.19542882e-01
2.84953356e-01 9.52033758e-01 1.24443412e+00 1.14012682e+00
-3.00096095e-01 2.15662662e-02 1.05113721e+00 -1.67783275e-01
5.03515720e-01 -1.79362118e-01 1.94087863e-01 9.00410116e-01
6.23104125e-02 -2.72842199e-01 -5.44208109e-01 8.48653555e-01
-1.76295006e+00 -1.03580117e+00 2.26264417e-01 2.64492035e+00
2.72760037e-02 -4.38067950e-02 -1.97087958e-01 4.69445050e-01
7.04841673e-01 -1.32722661e-01 -8.55977595e-01 -6.47144377e-01
1.42566651e-01 4.30044740e-01 9.78939533e-01 -2.07452267e-01
-7.76713192e-01 3.03659022e-01 5.65932989e+00 7.57934928e-01
-1.07323587e+00 1.21912360e-01 2.67017186e-01 -3.61235291e-01
-2.79051659e-04 3.19159508e-01 -8.28073025e-02 2.78679043e-01
1.44509292e+00 -6.71192527e-01 5.25947154e-01 1.29407799e+00
4.44363236e-01 1.77949995e-01 -9.79920089e-01 6.37733698e-01
-7.07169294e-01 -1.49093843e+00 -4.01100039e-01 5.83608925e-01
7.31713831e-01 3.44857872e-01 -3.42026830e-01 6.31713867e-02
5.27792335e-01 -1.01428306e+00 2.93342292e-01 3.44728053e-01
6.68438911e-01 -1.30185759e+00 1.22017860e+00 5.75694561e-01
-1.45679283e+00 -7.22665906e-01 -4.64585006e-01 -4.18629080e-01
1.18437767e-01 1.02768755e+00 -1.03726315e+00 1.43253946e+00
9.92344260e-01 5.90542197e-01 -1.14586860e-01 1.10021091e+00
3.63979518e-01 6.89947248e-01 -2.98288614e-01 4.43658739e-01
-2.61823952e-01 -1.77066237e-01 1.93551153e-01 1.02581844e-01
6.16038918e-01 -5.44157505e-01 2.59533763e-01 2.77776778e-01
-6.75804690e-02 -3.32309783e-01 -1.01693499e+00 -3.66420060e-01
1.16616094e+00 1.69170511e+00 -1.89976528e-01 3.84276599e-01
-3.84482473e-01 5.51636457e-01 -1.28657624e-01 -9.01925564e-02
-1.66601270e-01 -4.75031644e-01 1.13088989e+00 5.37344664e-02
5.73257387e-01 -2.62989879e-01 -1.45659700e-01 -7.24175155e-01
4.23778929e-02 -9.28370059e-01 6.76404774e-01 -3.76819074e-01
-1.75444102e+00 5.99839270e-01 -3.23391557e-01 -1.69066453e+00
-6.53017938e-01 -5.41057467e-01 -8.25169504e-01 8.77325416e-01
-1.28730690e+00 -1.26222336e+00 -3.54356617e-02 4.63889956e-01
4.49890703e-01 -5.63526869e-01 1.16405988e+00 2.31542140e-01
-6.42676175e-01 2.23773688e-01 2.58291990e-01 -1.93112254e-01
3.03036749e-01 -9.07278180e-01 4.75697160e-01 7.80414104e-01
9.48252752e-02 4.49135661e-01 5.54390490e-01 -5.77290833e-01
-1.91012621e+00 -1.46208680e+00 6.39986336e-01 -2.04748064e-01
5.83708465e-01 -8.85112584e-02 -6.39577091e-01 5.74318171e-01
4.37266856e-01 5.63084602e-01 9.26800966e-01 1.04866646e-01
3.11720580e-01 -5.56551278e-01 -1.72269297e+00 -2.50154793e-01
5.58248103e-01 -5.38592339e-01 -7.94994310e-02 5.83479881e-01
4.92121279e-01 -1.22040631e-02 -1.59704483e+00 1.81864709e-01
-6.37549013e-02 -8.56798470e-01 4.66330588e-01 -8.10787320e-01
-4.15531874e-01 -4.18603539e-01 1.30510628e-01 -1.82892942e+00
-5.71819656e-02 -1.24004734e+00 -2.69642055e-01 1.36363721e+00
8.69063847e-03 -9.81750906e-01 1.03782058e+00 9.17628527e-01
-3.38076264e-01 -6.83579028e-01 -1.56441522e+00 -1.12623858e+00
-2.00205240e-02 -2.64320195e-01 1.51451540e+00 1.14539993e+00
6.84420913e-02 -2.86639988e-01 -5.00954866e-01 6.51577175e-01
4.51889485e-01 1.03690669e-01 7.76530802e-01 -1.32812572e+00
-7.50989914e-02 4.11315501e-01 -5.10989308e-01 -1.07443631e-01
5.28484508e-02 -5.77066004e-01 -4.39848691e-01 -9.80200291e-01
-8.56856704e-01 -7.68518627e-01 -3.50802839e-01 8.16761613e-01
7.74646640e-01 -4.18940812e-01 1.20445557e-01 7.01114684e-02
9.96418372e-02 7.34508991e-01 9.95120406e-01 -2.13889271e-01
3.04689556e-01 5.12026310e-01 -4.18129534e-01 3.27713966e-01
1.19692063e+00 -6.06207907e-01 -7.71244764e-01 -5.59026539e-01
2.88628172e-02 2.97978252e-01 3.15765083e-01 -1.00739717e+00
4.77020860e-01 -7.32906759e-02 2.08580136e-01 -6.29776537e-01
1.70988739e-01 -1.71911633e+00 9.33450758e-01 8.41119230e-01
4.60601807e-01 2.64126927e-01 1.80499688e-01 5.85624814e-01
-5.51708102e-01 -3.14091891e-01 9.13960710e-02 -9.97863710e-02
-3.97227973e-01 5.56448698e-01 -3.07284713e-01 -4.35167760e-01
1.18837154e+00 -2.95846939e-01 -5.53248286e-01 -7.63703063e-02
-5.71442425e-01 6.31381392e-01 5.58792233e-01 3.49820971e-01
4.06385213e-01 -8.47176313e-01 -5.11114836e-01 7.38714993e-01
-8.91519040e-02 3.38341355e-01 6.59932315e-01 5.99490643e-01
-2.14855522e-01 4.53310639e-01 -2.55367845e-01 -2.14289889e-01
-8.92534435e-01 6.92160249e-01 3.06727737e-01 -2.51538008e-01
-4.97327179e-01 5.19993007e-01 -6.49145365e-01 -5.94177008e-01
-2.19110116e-01 -2.77446151e-01 5.99367917e-01 4.90856469e-02
1.84286892e-01 7.90417731e-01 1.13348258e+00 -1.11350462e-01
-3.24436188e-01 -1.67914391e-01 -1.10523347e-02 5.97272873e-01
1.56479692e+00 -1.10765323e-01 -1.33484051e-01 -7.73011073e-02
9.16077197e-01 -9.36310664e-02 -1.22650099e+00 2.73569405e-01
-1.26658231e-01 -1.01978528e+00 5.92524111e-01 -8.89833212e-01
-1.86970437e+00 5.77758014e-01 6.29012048e-01 5.46381652e-01
1.32673037e+00 -5.13645530e-01 8.98739219e-01 2.95331478e-01
1.17655694e+00 -9.33209181e-01 -5.97696722e-01 -1.16426416e-01
4.46614087e-01 -8.18258107e-01 1.24123059e-01 -1.43425718e-01
-6.83141828e-01 9.47727144e-01 4.63401884e-01 2.98581392e-01
8.68973434e-01 6.46321237e-01 3.38547416e-02 -3.31709057e-01
-1.32395661e+00 3.49623591e-01 -4.94970888e-01 1.13688338e+00
-2.61051685e-01 2.52922714e-01 1.75392374e-01 7.89981604e-01
-3.27077776e-01 6.39383122e-02 4.78415787e-01 1.49010873e+00
-8.56706128e-02 -1.53312361e+00 -2.57459283e-01 9.42699254e-01
-1.38758838e-01 3.46595198e-01 -1.29979160e-02 7.10264921e-01
5.19296348e-01 1.23933017e+00 2.38022536e-01 -6.63991392e-01
3.55907887e-01 -4.18207422e-02 -1.43833697e-01 -5.49779952e-01
-9.85896230e-01 -5.09530127e-01 3.70058537e-01 -4.99042720e-01
-1.35749906e-01 -7.54465878e-01 -1.16856444e+00 -5.70809364e-01
-8.90509903e-01 7.99925625e-01 9.39030290e-01 6.01336956e-01
8.04112494e-01 5.03393292e-01 1.53527296e+00 -5.76922953e-01
-1.33138287e+00 -6.13915980e-01 -1.24580669e+00 -5.14952540e-02
2.48364538e-01 -5.79305768e-01 -5.96925139e-01 -1.99955270e-01] | [5.8369140625, 6.43674898147583] |
c954bd39-af9b-41ef-acea-8a12b9bf7f41 | domain-adaptation-in-robot-fault-diagnostic | 1809.08626 | null | https://arxiv.org/abs/1809.08626v3 | https://arxiv.org/pdf/1809.08626v3.pdf | Domain Adaptation for Robot Predictive Maintenance Systems | Industrial robots play an increasingly important role in a growing number of fields. For example, robotics is used to increase productivity while reducing costs in various aspects of manufacturing. Since robots are often set up in production lines, the breakdown of a single robot has a negative impact on the entire process, in the worst case bringing the whole line to a halt until the issue is resolved, leading to substantial financial losses due to the unforeseen downtime. Therefore, predictive maintenance systems based on the internal signals of robots have gained attention as an essential component of robotics service offerings. The main shortcoming of existing predictive maintenance algorithms is that the extracted features typically differ significantly from the learnt model when the operation of the robot changes, incurring false alarms. In order to mitigate this problem, predictive maintenance algorithms require the model to be retrained with normal data of the new operation. In this paper, we propose a novel solution based on transfer learning to pass the knowledge of the trained model from one operation to another in order to prevent the need for retraining and to eliminate such false alarms. The deployment of the proposed unsupervised transfer learning algorithm on real-world datasets demonstrates that the algorithm can not only distinguish between operation and mechanical condition change, it further yields a sharper deviation from the trained model in case of a mechanical condition change and thus detects mechanical issues with higher confidence. | ['Arash Golibagh Mahyari', 'Thomas Locker'] | 2018-09-23 | null | null | null | null | ['industrial-robots'] | ['robots'] | [ 4.57695663e-01 1.47767857e-01 1.28484771e-01 -2.26336882e-01
2.98010409e-02 -1.98596597e-01 1.71287671e-01 5.92600584e-01
-1.84469998e-01 7.57001162e-01 -7.83452630e-01 -7.14859888e-02
-5.66530645e-01 -8.06050003e-01 -7.43968308e-01 -8.54430020e-01
8.06767941e-02 6.23518765e-01 2.63934165e-01 -1.23997293e-01
4.23636287e-01 8.00939620e-01 -1.73152065e+00 9.14293006e-02
8.13920081e-01 1.19470704e+00 5.66075861e-01 1.02313995e-01
7.12383166e-02 5.13296843e-01 -7.38559306e-01 2.42576092e-01
3.55516404e-01 -3.03911090e-01 -6.35424197e-01 4.26925868e-01
-3.02298218e-01 -1.51814237e-01 4.89039011e-02 1.07319605e+00
-4.05545123e-02 1.11588277e-01 4.57755148e-01 -1.28409529e+00
2.03950834e-02 1.79096296e-01 -3.06580842e-01 -8.41342881e-02
8.12533721e-02 1.13636330e-02 5.58229089e-01 -4.81182337e-01
2.12693632e-01 8.76038134e-01 4.91353393e-01 -2.09025443e-01
-1.28025365e+00 -4.24842566e-01 1.01670235e-01 4.15224612e-01
-9.76393640e-01 -4.44364287e-02 8.76023948e-01 -4.31911051e-01
6.60812497e-01 2.66441982e-03 5.45991540e-01 6.78865790e-01
7.43060827e-01 1.83970153e-01 8.66173506e-01 -4.91731286e-01
3.17899466e-01 3.51512164e-01 -1.85332358e-01 4.74309087e-01
5.30680478e-01 6.17796835e-03 1.22113023e-02 2.75371999e-01
5.76405168e-01 3.73564184e-01 -2.69111931e-01 -5.95149457e-01
-7.32307494e-01 5.85511684e-01 3.47417623e-01 6.82278872e-01
-7.62779653e-01 -2.24277809e-01 4.30104315e-01 6.02145672e-01
-6.98211640e-02 6.78424060e-01 -6.00736856e-01 -1.59834027e-01
-3.91154051e-01 -1.13847382e-01 6.40688241e-01 7.10659683e-01
8.05330396e-01 7.13600719e-04 4.10053045e-01 6.76570237e-01
-1.89217195e-01 1.69476960e-02 5.08771896e-01 -5.03816605e-01
6.65179491e-02 9.24485028e-01 2.83744097e-01 -1.28770697e+00
-4.56601083e-01 -5.81708908e-01 -7.83727884e-01 3.08554530e-01
1.60951927e-01 2.56408695e-02 -5.80993235e-01 1.27584243e+00
2.78744280e-01 -1.94178000e-01 -1.47116080e-01 6.94549680e-01
-2.80487686e-01 7.73434043e-01 -1.31947100e-01 -4.63457733e-01
9.07773435e-01 -4.80714381e-01 -8.46929729e-01 -2.92371303e-01
5.13768554e-01 -7.61392713e-01 6.52079225e-01 9.98950601e-01
-3.95820081e-01 -8.12704265e-01 -1.30471003e+00 6.11603796e-01
-3.49359810e-01 3.24810982e-01 3.63195807e-01 -4.67075221e-03
-4.13766563e-01 1.17978680e+00 -8.07403922e-01 -5.26296794e-01
-5.72666675e-02 4.52725053e-01 -4.51437056e-01 -1.21885620e-01
-9.43973303e-01 1.19418371e+00 7.02328622e-01 6.00788057e-01
-4.39059526e-01 -1.37708530e-01 -4.98466551e-01 1.32826760e-01
6.63026214e-01 1.99837044e-01 1.06901526e+00 -9.63353872e-01
-1.36049283e+00 1.81083649e-01 3.76947701e-01 -4.87179846e-01
5.65815508e-01 -4.34234679e-01 -6.44376636e-01 -2.07220949e-02
3.44321690e-03 -1.34647593e-01 1.14517617e+00 -1.18763363e+00
-8.73468637e-01 -2.33212888e-01 -1.85173646e-01 -1.47459298e-01
-3.18139344e-01 -5.77606797e-01 1.35184556e-01 -2.52714187e-01
3.06832522e-01 -1.08257103e+00 8.27098787e-02 -2.26365745e-01
-2.91862786e-01 -6.03029840e-02 1.16502368e+00 -6.05904341e-01
7.49629140e-01 -2.28253007e+00 9.61406454e-02 2.58257836e-01
-3.76617968e-01 3.40567231e-01 2.68865168e-01 6.36187673e-01
-3.07500362e-01 -5.68242848e-01 -2.12704763e-01 2.94488996e-01
-5.05015373e-01 3.38817805e-01 -7.87851810e-02 5.34747362e-01
5.51695883e-01 -1.20610492e-02 -7.50898242e-01 -1.07661322e-01
4.80812192e-01 1.19289081e-03 -2.57255346e-01 3.96177977e-01
5.44573478e-02 7.08858490e-01 -3.80454391e-01 2.94506997e-01
2.90260673e-01 9.72075462e-02 2.60487080e-01 -3.37242246e-01
-2.07674965e-01 -6.54908642e-02 -1.17117441e+00 1.21308482e+00
-6.70391083e-01 3.82881463e-01 1.06107585e-01 -1.78994012e+00
1.33271241e+00 4.72813338e-01 7.50713587e-01 -5.67642570e-01
4.45721835e-01 2.76541263e-01 1.80684119e-01 -5.53549945e-01
2.69015580e-01 -1.62371561e-01 1.36395963e-02 -1.84879992e-02
-3.51553247e-03 -2.68415183e-01 8.24436266e-03 -5.06645679e-01
1.06882358e+00 2.04292655e-01 2.75459528e-01 -7.80362561e-02
7.37301528e-01 4.15604375e-02 7.51502812e-01 2.34554932e-01
-4.97852713e-02 -7.10338354e-02 3.42262477e-01 -4.60043848e-01
-6.84171259e-01 -7.67645538e-01 -1.48610264e-01 5.33909142e-01
2.97860295e-01 2.88184613e-01 -3.52468252e-01 -5.80516815e-01
1.13550730e-01 9.26338553e-01 -3.21826696e-01 -9.94659007e-01
-5.37201107e-01 -4.40220416e-01 -2.27831066e-01 2.85083979e-01
4.32607889e-01 -1.10686338e+00 -1.17742467e+00 5.63169062e-01
-1.02190964e-01 -1.07263589e+00 2.22927183e-01 7.25537121e-01
-1.19237292e+00 -1.36662567e+00 -1.24485403e-01 -7.69895375e-01
8.26263666e-01 1.11914620e-01 4.58791822e-01 1.80998951e-01
-4.80356485e-01 4.86743934e-02 -4.56189454e-01 -5.05407870e-01
-8.99923146e-01 1.30927367e-02 2.67560750e-01 6.62686378e-02
1.70516625e-01 -4.62841809e-01 -1.32804126e-01 5.53172231e-01
-5.73448896e-01 -3.00695986e-01 1.01221883e+00 9.22341049e-01
3.23352098e-01 1.11925983e+00 1.09379256e+00 -7.83457875e-01
4.00479525e-01 -3.00118208e-01 -9.17264223e-01 1.41985893e-01
-1.00974739e+00 2.08580539e-01 9.20993745e-01 -4.41543698e-01
-1.09800220e+00 1.70103922e-01 3.38681102e-01 -4.51087475e-01
-3.13165218e-01 5.40671289e-01 -1.63527831e-01 3.35633010e-01
1.76461920e-01 8.65058005e-02 3.42072636e-01 -5.67932487e-01
-1.98646054e-01 6.07764065e-01 5.64709544e-01 -1.89414084e-01
9.29292679e-01 1.16687186e-01 3.29422444e-01 -8.95292222e-01
-4.90125567e-01 -5.49667716e-01 -7.49793768e-01 -4.76002723e-01
5.08195460e-01 -4.51538086e-01 -5.42262852e-01 4.94817019e-01
-9.89595294e-01 1.90305427e-01 -2.64505565e-01 5.79759061e-01
-5.06984055e-01 2.75579751e-01 -3.79342020e-01 -8.23307037e-01
-2.48721745e-02 -1.18288577e+00 4.43145543e-01 1.10200480e-01
-2.46660665e-01 -5.30906558e-01 -3.37428719e-01 9.30501148e-02
2.25391611e-01 3.19243401e-01 1.22283828e+00 -6.78333580e-01
-2.60235876e-01 -8.57874632e-01 1.95942938e-01 8.22657883e-01
9.34536874e-01 -2.20170561e-02 -6.95580661e-01 -4.15471077e-01
6.36935413e-01 4.60879132e-03 2.16468140e-01 8.43706727e-02
7.34097362e-01 1.92144439e-02 -3.82995784e-01 -2.28240967e-01
1.25670660e+00 7.66499817e-01 4.99069422e-01 4.36591059e-01
2.33472764e-01 9.82118130e-01 1.38934577e+00 3.92298192e-01
-4.98467118e-01 5.44192791e-01 7.89291203e-01 1.12222642e-01
5.36392152e-01 -6.09226488e-02 3.52347285e-01 6.03839934e-01
1.24149444e-02 2.01446205e-01 -3.97612989e-01 4.29435045e-01
-1.81674528e+00 -7.01610565e-01 -5.90030923e-02 2.64730000e+00
4.55011189e-01 6.75480068e-01 -2.12040558e-01 7.90824771e-01
1.03156555e+00 -5.33682048e-01 -5.26471853e-01 -5.39024651e-01
4.48066324e-01 -3.85233723e-02 4.55791116e-01 -1.08675717e-03
-9.91342366e-01 4.40436214e-01 4.37711239e+00 3.14643085e-01
-1.35042083e+00 -3.33440870e-01 2.26289928e-01 2.42253989e-01
7.59200096e-01 5.94567768e-02 -2.46413767e-01 4.15440410e-01
9.03021038e-01 1.69038162e-01 2.12899223e-01 1.16988635e+00
2.82766461e-01 -5.23978114e-01 -1.41052330e+00 7.50092387e-01
-1.65907666e-01 -5.88994503e-01 -2.95971125e-01 5.40473359e-03
1.60926193e-01 -6.12490356e-01 -3.93934250e-01 2.42154360e-01
-3.12586069e-01 -6.01707399e-01 6.90848887e-01 4.74493027e-01
2.43787721e-01 -1.00607586e+00 1.16521263e+00 4.56144571e-01
-8.72873306e-01 -4.82407719e-01 -2.66274899e-01 -2.61920720e-01
3.02816361e-01 7.40680099e-01 -1.29260576e+00 8.29496920e-01
7.08762586e-01 4.08712357e-01 -2.73967832e-01 9.08951819e-01
-1.95029184e-01 2.72961736e-01 -1.62050381e-01 6.79227859e-02
-1.47471875e-01 -4.30763364e-01 5.46280503e-01 5.00477791e-01
5.58750570e-01 -5.12772620e-01 2.00865388e-01 7.24210560e-01
4.04397339e-01 -1.06812149e-01 -7.63372123e-01 -1.13575749e-01
2.91486979e-01 1.11860132e+00 -9.75242019e-01 7.89477900e-02
-1.53325245e-01 1.20068061e+00 1.12175249e-01 -4.01642583e-02
-8.20370734e-01 -5.93280673e-01 2.96529740e-01 3.13157320e-01
4.56343502e-01 -1.95127502e-01 9.42122489e-02 -3.90566677e-01
4.41744216e-02 -5.39521277e-01 -3.22578289e-02 -4.18889523e-01
-1.13336587e+00 2.94774085e-01 -1.06506400e-01 -1.58707082e+00
-4.00277972e-01 -6.52185142e-01 -3.18271220e-01 5.96202612e-01
-1.18868434e+00 -3.40591133e-01 -3.40056807e-01 8.98974612e-02
6.71737254e-01 9.72815156e-02 6.82195246e-01 3.70355427e-01
-5.35567224e-01 4.84628938e-02 1.46111324e-02 -9.78335813e-02
9.29251909e-01 -9.94131744e-01 -4.07468319e-01 7.93148458e-01
-1.91792905e-01 2.77822554e-01 1.05256557e+00 -7.94482470e-01
-1.29308879e+00 -1.10291958e+00 4.33805168e-01 1.45250171e-01
4.95206565e-01 -8.91547278e-02 -1.14792454e+00 5.25122762e-01
-1.66946664e-01 -2.82366753e-01 2.19327614e-01 -1.48995062e-02
3.38031232e-01 -5.55620074e-01 -1.26691651e+00 2.29576051e-01
2.73989767e-01 -1.44763708e-01 -7.52246439e-01 2.42429629e-01
3.55069309e-01 -3.74716967e-02 -9.72984791e-01 6.12186074e-01
3.73372763e-01 -9.90013301e-01 4.08401608e-01 -2.16560140e-01
1.88736051e-01 -3.06787908e-01 3.00970793e-01 -1.57340038e+00
-3.61850649e-01 -1.97112992e-01 1.73243135e-01 1.21413648e+00
1.53888643e-01 -8.42616916e-01 4.64482546e-01 4.02902782e-01
-3.45371634e-01 -6.44056082e-01 -9.81090128e-01 -1.00478542e+00
-5.44141293e-01 -1.40850350e-01 2.72685707e-01 9.20197546e-01
1.48374632e-01 4.42011356e-01 -1.94742233e-01 4.88231122e-01
2.62033165e-01 9.05970857e-02 7.45672762e-01 -1.72899854e+00
-2.94548303e-01 9.20875091e-03 -7.42971599e-01 -4.30475652e-01
-8.26380774e-02 -3.63679677e-01 6.54958010e-01 -1.36777782e+00
-2.72136986e-01 -6.54139578e-01 -5.05804420e-01 4.23831046e-01
1.47812188e-01 -2.92718321e-01 -2.40910985e-02 2.35919416e-01
-1.73888244e-02 5.63971877e-01 9.46847379e-01 -1.16193220e-01
-4.03782487e-01 4.96731043e-01 -1.00171685e-01 7.83976912e-01
8.93062413e-01 -4.67687666e-01 -5.50990522e-01 -1.52078131e-02
-2.49550911e-03 1.37217805e-01 7.90484250e-02 -1.37677300e+00
3.13767493e-02 7.44328019e-04 3.49831849e-01 -4.73073691e-01
2.49698535e-01 -1.69250369e+00 4.33361411e-01 1.05178738e+00
1.31344616e-01 1.22619852e-01 2.46569142e-01 8.25681508e-01
-3.09100002e-01 -6.29404545e-01 8.26594174e-01 1.72560394e-01
-5.76539397e-01 -3.32820982e-01 -6.15278900e-01 -8.38859141e-01
1.42487288e+00 -2.88088650e-01 9.08086449e-02 -1.30437732e-01
-5.83697379e-01 1.93295807e-01 3.42281491e-01 7.43373215e-01
3.10626954e-01 -6.91993535e-01 -1.62742436e-01 2.63004601e-01
6.99695870e-02 2.38546774e-01 2.10753441e-01 9.42160726e-01
-3.54314804e-01 4.72244054e-01 -4.16436702e-01 -7.29044020e-01
-1.01032412e+00 7.79404342e-01 8.96903425e-02 -2.44940981e-01
-7.54823685e-01 2.66832769e-01 -1.71511859e-01 3.45344795e-03
1.04512192e-01 -7.70186603e-01 -1.80124655e-01 7.42350444e-02
5.65910563e-02 4.90276128e-01 7.32315183e-01 -3.96747380e-01
-8.27972963e-02 1.72268152e-01 -2.64538527e-01 4.88475174e-01
1.34990644e+00 -1.44881215e-02 -1.53883487e-01 9.03270900e-01
6.96565628e-01 -2.27247477e-01 -1.23677492e+00 1.01452209e-01
3.36003691e-01 -2.92161137e-01 -1.69190392e-02 -6.96200132e-01
-9.15421903e-01 4.85039860e-01 9.21137571e-01 4.61708218e-01
1.17633402e+00 -2.76518047e-01 5.57933092e-01 6.19866312e-01
8.83808911e-01 -1.43058181e+00 1.48472577e-01 1.77211776e-01
8.75904799e-01 -1.17124629e+00 7.38019943e-02 -4.59333241e-01
-4.82299238e-01 1.22321403e+00 5.29823661e-01 -2.07827374e-01
4.09458548e-01 -2.67529767e-02 -2.49409497e-01 -6.59202933e-02
-3.28719586e-01 3.91121745e-01 -2.21311644e-01 4.83850747e-01
3.66617851e-02 1.88900642e-02 -1.59476280e-01 4.18530881e-01
1.64512604e-01 1.89022988e-01 5.16721487e-01 1.40192628e+00
-6.48089528e-01 -1.12069368e+00 -4.82774943e-01 6.00651741e-01
-1.29615843e-01 6.64290965e-01 -6.59803376e-02 9.60173666e-01
2.55970240e-01 1.03795874e+00 1.28898084e-01 -3.80966395e-01
7.63987780e-01 2.94933736e-01 4.84766483e-01 -7.36781240e-01
-1.51697576e-01 1.03910714e-01 -7.57695641e-03 -4.52525496e-01
-2.88493395e-01 -8.06555450e-01 -1.53958821e+00 2.67906755e-01
-7.52786040e-01 3.02351743e-01 7.86290228e-01 1.05001998e+00
3.25940102e-01 1.01880383e+00 9.41832840e-01 -7.85405278e-01
-1.01326203e+00 -1.23744011e+00 -7.27520466e-01 4.90486771e-01
1.03181921e-01 -1.33099616e+00 -4.37532961e-01 6.11622036e-02] | [6.841690540313721, 2.352938413619995] |
e1c4f836-4dd5-4a23-9f95-a1266b862741 | predicting-lymph-node-metastasis-using | null | null | http://openaccess.thecvf.com/content_CVPR_2020/html/Zhao_Predicting_Lymph_Node_Metastasis_Using_Histopathological_Images_Based_on_Multiple_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Zhao_Predicting_Lymph_Node_Metastasis_Using_Histopathological_Images_Based_on_Multiple_CVPR_2020_paper.pdf | Predicting Lymph Node Metastasis Using Histopathological Images Based on Multiple Instance Learning With Deep Graph Convolution | Multiple instance learning (MIL) is a typical weakly-supervised learning method where the label is associated with a bag of instances instead of a single instance. Despite extensive research over past years, effectively deploying MIL remains an open and challenging problem, especially when the commonly assumed standard multiple instance (SMI) assumption is not satisfied. In this paper, we propose a multiple instance learning method based on deep graph convolutional network and feature selection (FS-GCN-MIL) for histopathological image classification. The proposed method consists of three components, including instance-level feature extraction, instance-level feature selection, and bag-level classification. We develop a self-supervised learning mechanism to train the feature extractor based on a combination model of variational autoencoder and generative adversarial network (VAE-GAN). Additionally, we propose a novel instance-level feature selection method to select the discriminative instance features. Furthermore, we employ a graph convolutional network (GCN) for learning the bag-level representation and then performing the classification. We apply the proposed method in the prediction of lymph node metastasis using histopathological images of colorectal cancer. Experimental results demonstrate that the proposed method achieves superior performance compared to the state-of-the-art methods.
| [' Jianhua Yao', ' Xinjuan Fan', ' Bjoern Menze', ' Sen Yang', ' Jiarui Sun', ' Jun Zhang', ' Niyun Zhou', ' Hailing Liu', ' Yuqi Fang', ' Fan Yang', 'Yu Zhao'] | 2020-06-01 | null | null | null | cvpr-2020-6 | ['histopathological-image-classification'] | ['medical'] | [ 4.94865716e-01 1.93110537e-02 -2.93299019e-01 -2.83929497e-01
-1.21117818e+00 -7.99939856e-02 4.68387872e-01 2.87470520e-01
-3.24893802e-01 6.45807981e-01 -1.19566955e-01 -2.93577820e-01
-4.10468012e-01 -1.08232892e+00 -6.60697162e-01 -1.20032668e+00
1.17643224e-02 3.16488802e-01 -1.50712952e-01 5.21950461e-02
6.70366064e-02 3.46399963e-01 -1.12589383e+00 3.11361760e-01
8.21463943e-01 1.12592137e+00 4.01631445e-02 5.39180160e-01
-6.45923615e-02 1.00499439e+00 -4.01505560e-01 -2.74044812e-01
7.31339157e-02 -5.87953031e-01 -8.34841430e-01 3.99191290e-01
1.94828495e-01 -1.45163517e-02 -3.54396969e-01 9.52463269e-01
4.78914499e-01 2.21926346e-01 8.54587913e-01 -1.43614900e+00
-5.30772507e-01 2.50302434e-01 -6.01415455e-01 1.18984781e-01
-3.40524256e-01 4.50601317e-02 9.71156716e-01 -8.13501298e-01
3.68417621e-01 8.40713739e-01 6.18639648e-01 5.19426763e-01
-1.20389354e+00 -5.94378293e-01 1.20490029e-01 1.23686686e-01
-1.39133298e+00 -1.01340123e-01 1.16439998e+00 -4.58345652e-01
4.59683657e-01 1.46512359e-01 7.61671782e-01 9.91459548e-01
5.38262308e-01 8.93568337e-01 1.14512706e+00 -3.73908639e-01
1.71264887e-01 1.02314122e-01 5.41635826e-02 1.14585888e+00
2.47948375e-02 -7.69706666e-02 -7.04485252e-02 -4.12228078e-01
7.73288548e-01 5.42559505e-01 -1.44080698e-01 -3.52797061e-01
-9.88969266e-01 1.11470091e+00 8.11884105e-01 3.30604732e-01
-4.70113635e-01 3.13387692e-01 4.99187171e-01 1.32564113e-01
7.24869788e-01 -1.55484602e-02 -1.99152127e-01 5.19631267e-01
-9.57854033e-01 8.20834488e-02 5.94128609e-01 6.82610512e-01
6.98144138e-01 -1.33339271e-01 -4.26139444e-01 8.56873512e-01
4.30881977e-01 2.01163322e-01 8.21666360e-01 -2.74875909e-01
1.16351917e-01 8.67732286e-01 -4.04679835e-01 -9.44068968e-01
-4.83291924e-01 -1.00729489e+00 -1.26147318e+00 1.14082256e-02
-5.72508201e-03 -4.65534180e-02 -8.51372540e-01 1.54076314e+00
4.78836477e-01 7.49513447e-01 2.15667248e-01 5.83440900e-01
1.03852141e+00 5.76547205e-01 3.33216310e-01 -2.46531799e-01
1.21211374e+00 -1.11067748e+00 -5.54281831e-01 -2.29949187e-02
7.65360594e-01 -3.10119390e-01 7.70089746e-01 2.10687146e-02
-6.00140870e-01 -2.99522251e-01 -9.18460667e-01 1.31382257e-01
-4.07269180e-01 2.40885615e-01 7.04505563e-01 3.14243883e-01
-7.56086111e-01 3.76928955e-01 -9.18531656e-01 -1.80684235e-02
6.24870479e-01 6.34340942e-01 -5.59853733e-01 -1.20966047e-01
-1.00894880e+00 4.18679565e-01 3.54739755e-01 1.29958287e-01
-1.03509212e+00 -7.06388056e-01 -1.15294850e+00 1.45596061e-02
1.95087522e-01 -9.05668855e-01 8.50688457e-01 -1.15084612e+00
-1.28284013e+00 9.21014190e-01 -3.59658450e-02 -3.46492201e-01
2.95877874e-01 3.84203345e-01 -2.58925170e-01 2.16823474e-01
1.30645066e-01 2.01897740e-01 9.65706766e-01 -1.24825466e+00
-4.97405857e-01 -4.97998059e-01 -9.89162084e-03 1.40659913e-01
-2.90896177e-01 -5.69998562e-01 -1.52294949e-01 -7.07652032e-01
1.30014867e-01 -8.25189531e-01 -4.01069820e-01 -2.48680383e-01
-6.05834663e-01 -4.26191092e-01 7.71445692e-01 -6.12932801e-01
1.09207547e+00 -2.11965227e+00 3.00351322e-01 3.16902697e-01
3.51212054e-01 2.07567811e-02 -6.37859851e-02 4.13408637e-01
1.25416380e-03 1.67063624e-01 -3.75274956e-01 -5.33867300e-01
-1.32818267e-01 2.04537779e-01 2.18120024e-01 7.46118128e-01
4.27372515e-01 9.96694684e-01 -1.10375988e+00 -8.57667387e-01
3.34745318e-01 5.71675658e-01 -4.95405704e-01 4.06178623e-01
-9.61183831e-02 8.05848122e-01 -6.80293858e-01 8.25517952e-01
6.01989329e-01 -5.55542111e-01 1.91664889e-01 -3.34976494e-01
3.84519786e-01 -2.88565487e-01 -7.64965951e-01 1.60604310e+00
-5.87748349e-01 7.72100762e-02 -2.99697459e-01 -1.38439262e+00
6.75854385e-01 2.51676321e-01 7.56172836e-01 -3.47285330e-01
2.58618593e-01 2.54850864e-01 -8.66798162e-02 -7.48858869e-01
-1.63904488e-01 -4.91068661e-01 -1.82426929e-01 9.70319286e-02
4.04319346e-01 1.92891836e-01 -1.50288641e-01 1.15488425e-01
1.26253426e+00 -4.52445634e-02 5.62219203e-01 -4.71968018e-02
8.68255258e-01 -1.26052588e-01 6.88732028e-01 5.13847649e-01
-2.84634858e-01 4.54057306e-01 5.08873999e-01 -3.54065299e-01
-6.14306748e-01 -7.28175223e-01 -3.27274650e-01 8.22546184e-01
1.81090742e-01 -1.18290514e-01 -6.55022740e-01 -1.21329248e+00
5.27194701e-02 4.46580470e-01 -9.92405415e-01 -3.71513367e-01
-4.19325203e-01 -1.05946791e+00 3.53823185e-01 5.09883285e-01
5.20935118e-01 -1.06945968e+00 -5.51598556e-02 2.64983773e-01
2.12678891e-02 -8.99255991e-01 -2.00893432e-01 2.12549925e-01
-8.06763053e-01 -1.27106965e+00 -6.47695482e-01 -1.04456222e+00
1.08020031e+00 1.17489181e-01 1.01588941e+00 5.69079518e-01
-3.72338265e-01 4.39340472e-01 -4.22162473e-01 -2.37706974e-01
-3.39572340e-01 1.75827608e-01 -3.92139852e-01 6.02868557e-01
2.31799468e-01 -4.57138926e-01 -9.95542109e-01 -7.26169571e-02
-1.09109962e+00 9.22531262e-02 1.04134381e+00 1.48458743e+00
1.16198146e+00 1.84865683e-01 9.03354228e-01 -1.39856052e+00
3.19233507e-01 -7.82182992e-01 -1.59346193e-01 2.84781575e-01
-3.66077930e-01 -1.02994218e-01 8.76883566e-01 -1.88709214e-01
-8.12663615e-01 9.09085199e-02 -3.06925297e-01 -4.53544110e-01
-9.85357240e-02 9.09947813e-01 -1.61142886e-01 -4.08667982e-01
3.57825935e-01 4.84811753e-01 2.63155073e-01 -9.99908149e-02
6.95979968e-02 6.96076870e-01 1.83413908e-01 -2.98392594e-01
6.76061273e-01 4.66859221e-01 2.80294359e-01 -5.57957172e-01
-1.13056099e+00 -5.50819457e-01 -5.87744772e-01 -2.08025157e-01
9.12218213e-01 -9.12322104e-01 -6.04638994e-01 4.80694711e-01
-6.06942892e-01 -4.10065800e-01 -7.51983076e-02 3.87497991e-01
-6.75492525e-01 1.12460135e-03 -7.12236106e-01 -5.26776254e-01
-5.41448116e-01 -1.35869861e+00 1.39552653e+00 2.35815689e-01
4.69544798e-01 -1.44762361e+00 1.04974158e-01 4.80397284e-01
2.88246691e-01 8.46769154e-01 1.00946844e+00 -8.61073732e-01
-4.83104557e-01 -6.37719750e-01 -9.54414159e-02 3.67694676e-01
3.00465733e-01 -2.30222747e-01 -8.81612301e-01 -6.67011499e-01
-1.55849546e-01 -4.43144411e-01 9.88714099e-01 4.66039062e-01
1.64330876e+00 -4.61912721e-01 -6.18986964e-01 8.40750635e-01
1.86619949e+00 -7.53963366e-02 4.58005399e-01 3.35012108e-01
9.85789537e-01 1.91498443e-01 4.72371459e-01 3.03998828e-01
3.17997456e-01 3.28629673e-01 7.52096474e-01 -2.93383628e-01
7.22607374e-02 -2.71854967e-01 -1.30866602e-01 9.13367927e-01
-8.69372301e-03 -2.55341798e-01 -7.24268198e-01 4.90296513e-01
-1.94708085e+00 -8.52425575e-01 2.62424827e-01 2.03026199e+00
6.69395864e-01 -1.07508227e-01 -1.47763491e-01 3.04271132e-01
7.06664562e-01 1.31381571e-01 -5.03843486e-01 2.87729055e-02
3.63959312e-01 3.01018298e-01 3.38286966e-01 2.53100097e-01
-1.44175732e+00 5.60805857e-01 4.50802088e+00 1.09510696e+00
-1.24756193e+00 2.91290700e-01 8.86507094e-01 2.88882703e-01
-1.70805618e-01 -1.91849545e-01 -5.48643172e-01 4.92550462e-01
7.00785518e-01 8.93334150e-02 2.18964671e-03 8.46024871e-01
-2.41760109e-02 2.99616843e-01 -9.15609837e-01 8.14365983e-01
2.70790815e-01 -1.30582786e+00 2.38450840e-01 1.90117911e-01
6.62505507e-01 -3.45165908e-01 1.45463303e-01 5.92845738e-01
-1.51674062e-01 -1.16324484e+00 1.85318783e-01 7.24171042e-01
7.64368892e-01 -1.05492878e+00 1.08314621e+00 5.18525958e-01
-1.16826046e+00 -1.13926634e-01 -4.16681528e-01 5.13599634e-01
-3.24880272e-01 5.71793973e-01 -7.82577872e-01 8.83515000e-01
2.01577127e-01 9.63537455e-01 -5.38824558e-01 1.01735258e+00
3.63422632e-02 8.16936791e-01 6.60182908e-02 -1.05757177e-01
4.37008500e-01 -1.26440758e-02 3.47002447e-01 1.16482270e+00
2.15436101e-01 1.43845394e-01 5.79056442e-01 5.63076735e-01
-2.40126327e-01 3.47331315e-01 -6.31443143e-01 1.11060455e-01
2.26254597e-01 1.68982971e+00 -7.03142107e-01 -1.68220028e-01
-4.78458762e-01 1.03159153e+00 6.03375494e-01 1.90840945e-01
-7.40854025e-01 -3.86258274e-01 4.04938012e-02 3.19546536e-02
2.40079895e-01 2.91170329e-01 1.90056011e-01 -1.24549222e+00
-2.06243068e-01 -7.57950783e-01 6.21581554e-01 -2.90406376e-01
-1.64996684e+00 5.49710274e-01 -3.13632131e-01 -1.61762464e+00
-9.10805240e-02 -6.01032495e-01 -9.13332522e-01 6.18676007e-01
-1.73701096e+00 -1.85363376e+00 -5.27802050e-01 7.69614160e-01
5.00175178e-01 -3.01519841e-01 1.05631733e+00 2.01367766e-01
-6.97422862e-01 8.15697253e-01 1.09776445e-01 4.80064988e-01
4.03711468e-01 -1.38784337e+00 -5.01052260e-01 4.12919790e-01
-6.02260605e-02 4.26348567e-01 3.08133304e-01 -5.53291976e-01
-1.45723879e+00 -1.78956962e+00 4.37174141e-01 -5.59367947e-02
5.39864779e-01 -2.26743490e-01 -7.47369409e-01 8.44772458e-01
1.23571306e-01 7.58949280e-01 1.21949661e+00 -2.27808326e-01
3.31230760e-02 -2.72555768e-01 -1.47146463e+00 2.13379756e-01
6.06354475e-01 -4.07892346e-01 -1.77029446e-01 6.85011208e-01
6.41732156e-01 -4.05997634e-01 -1.26498783e+00 6.61363661e-01
3.44263136e-01 -5.52750230e-01 9.01900649e-01 -7.01949656e-01
5.94014823e-01 -1.85017094e-01 -1.42840996e-01 -1.56542253e+00
-5.01330733e-01 -6.73389435e-02 -7.43271485e-02 1.10829651e+00
3.19728762e-01 -5.87800264e-01 7.93178976e-01 1.35590844e-02
-2.32681215e-01 -1.45519948e+00 -8.60559583e-01 -4.59967703e-01
1.02742612e-01 -8.10046028e-03 5.32595932e-01 1.05804539e+00
-1.54290408e-01 2.57175863e-01 -2.58544415e-01 3.08285147e-01
9.59876478e-01 3.34039092e-01 4.70178097e-01 -1.05663085e+00
-5.14500737e-01 -2.52273351e-01 -9.88218129e-01 -2.87478477e-01
5.53594708e-01 -1.43726933e+00 -8.37099329e-02 -1.62684739e+00
5.05981386e-01 -5.58922768e-01 -9.31123972e-01 4.91560072e-01
-4.21406716e-01 2.56558001e-01 -1.56910181e-01 1.15769617e-01
-7.21914172e-01 6.14300549e-01 1.34941113e+00 -5.52778482e-01
7.11586326e-02 2.13678733e-01 -6.74127996e-01 5.70640147e-01
6.76314712e-01 -5.35019994e-01 -4.05495554e-01 1.28428087e-01
-6.30514175e-02 3.16163838e-01 7.71069407e-01 -9.85097647e-01
1.16308458e-01 -1.16609119e-01 6.12055361e-01 -4.86526072e-01
1.15591951e-01 -9.00558770e-01 5.63061610e-02 5.41464269e-01
-4.95758742e-01 -3.56089056e-01 -7.48279318e-02 9.98170435e-01
-4.03537422e-01 -3.02442968e-01 7.63269067e-01 -2.89604545e-01
-6.43464208e-01 8.91529739e-01 -1.92881767e-02 -1.85555711e-01
1.25424075e+00 2.35564306e-01 -2.13493586e-01 7.78545514e-02
-9.30147886e-01 2.89562643e-01 7.97807872e-02 -3.23446095e-02
8.45597804e-01 -1.56361032e+00 -8.68295550e-01 2.54211336e-01
4.35434490e-01 1.89487606e-01 4.74301308e-01 1.11326957e+00
-4.51828003e-01 -1.35463206e-02 9.51051041e-02 -6.03840113e-01
-1.01473272e+00 4.37499434e-01 5.84505260e-01 -7.98662663e-01
-6.09924912e-01 9.73169208e-01 2.71837533e-01 -6.07135952e-01
4.25105281e-02 2.22485408e-01 -5.13886809e-01 -1.44784093e-01
2.52437294e-01 1.01837632e-03 2.21584082e-01 -5.94469011e-01
-2.84888566e-01 3.29764009e-01 -1.62056088e-01 5.06532252e-01
1.52555931e+00 2.47642368e-01 -3.84052694e-01 4.72588450e-01
1.59429419e+00 -2.69209534e-01 -9.39858556e-01 -2.42296323e-01
-4.05667841e-01 -4.89423126e-02 3.29960465e-01 -5.73285401e-01
-1.46777248e+00 8.23247075e-01 7.68099487e-01 1.51775092e-01
1.11891329e+00 -9.70905572e-02 8.38684201e-01 1.14830486e-01
3.83275390e-01 -6.39424920e-01 1.15989745e-01 3.06892190e-02
8.13741803e-01 -1.47627461e+00 -5.28981015e-02 -3.72727215e-01
-3.55750650e-01 1.10725880e+00 6.73486650e-01 -6.31355166e-01
1.08521295e+00 1.72040593e-02 -3.17719504e-02 -3.88484687e-01
-6.84532344e-01 -1.62564903e-01 4.41328526e-01 2.67012507e-01
4.97222215e-01 3.42383057e-01 -3.49337399e-01 9.35041130e-01
2.16837764e-01 4.07368615e-02 1.35256797e-01 9.04087424e-01
-3.16147953e-01 -9.96985078e-01 9.05712787e-03 1.03219771e+00
-8.01167250e-01 -1.14322498e-01 -2.85997894e-02 8.60504746e-01
2.17895076e-01 6.95704401e-01 -9.57384929e-02 -5.36502361e-01
4.02446352e-02 7.62738138e-02 4.83637154e-01 -7.36952782e-01
-6.97045445e-01 5.67446835e-02 -3.14327478e-01 -2.69168794e-01
-5.89762211e-01 -4.14704412e-01 -1.17674017e+00 1.72322735e-01
-5.87020397e-01 1.72182426e-01 4.61588502e-01 1.10302043e+00
1.26978844e-01 8.60390782e-01 9.56340194e-01 -8.46496224e-01
-5.49384296e-01 -9.03949857e-01 -8.25798452e-01 5.73515773e-01
5.26471913e-01 -8.58823121e-01 -5.34126461e-01 -2.14670375e-02] | [15.109024047851562, -2.7713935375213623] |
51de5024-67f4-4f3b-bf12-eeab59460de4 | iitp-ai-nlp-ml-cl-scisumm-2020-cl-laysumm | null | null | https://aclanthology.org/2020.sdp-1.30 | https://aclanthology.org/2020.sdp-1.30.pdf | IITP-AI-NLP-ML@ CL-SciSumm 2020, CL-LaySumm 2020, LongSumm 2020 | The publication rate of scientific literature increases rapidly, which poses a challenge for researchers to keep themselves updated with new state-of-the-art. Scientific document summarization solves this problem by summarizing the essential fact and findings of the document. In the current paper, we present the participation of IITP-AI-NLP-ML team in three shared tasks, namely, CL-SciSumm 2020, LaySumm 2020, LongSumm 2020, which aims to generate medium, lay, and long summaries of the scientific articles, respectively. To solve CL-SciSumm 2020 and LongSumm 2020 tasks, three well-known clustering techniques are used, and then various sentence scoring functions, including textual entailment, are used to extract the sentences from each cluster for a summary generation. For LaySumm 2020, an encoder-decoder based deep learning model has been utilized. Performances of our developed systems are evaluated in terms of ROUGE measures on the associated datasets with the shared task. | ['Pushpak Bhattacharyya', 'Sriparna Saha', 'Naveen Saini', 'Harshavardhan Kundarapu', 'Santosh Kumar Mishra'] | null | null | null | null | emnlp-sdp-2020-11 | ['scientific-article-summarization'] | ['natural-language-processing'] | [ 3.48479636e-02 1.11375395e-02 -2.35589538e-02 -7.85962641e-02
-1.64950991e+00 -5.72708368e-01 7.94352412e-01 4.39381182e-01
-2.63829261e-01 1.22870302e+00 7.83535957e-01 -6.40087575e-02
-2.36060247e-01 -3.74122351e-01 -7.52820790e-01 -3.78260106e-01
1.60203204e-01 6.00200057e-01 -2.57352740e-01 6.83675855e-02
1.04070711e+00 -2.04644967e-02 -1.37989938e+00 8.21108878e-01
1.34522629e+00 6.54836893e-01 6.72814727e-01 1.09910202e+00
-4.76337194e-01 9.26608145e-01 -1.17107654e+00 -6.10157251e-01
-4.85101610e-01 -7.58389831e-01 -9.15872097e-01 -1.78551421e-01
6.81368828e-01 -9.36589465e-02 -1.18537605e-01 1.06190622e+00
7.72997260e-01 6.74068630e-02 1.06545043e+00 -8.93912375e-01
-8.36895406e-01 1.24743903e+00 -6.98428094e-01 2.05887377e-01
6.02437913e-01 -2.23746523e-01 1.26018667e+00 -1.05653608e+00
8.42471182e-01 1.22045827e+00 3.86878699e-01 4.16441590e-01
-6.76228821e-01 -4.79703009e-01 -2.90214837e-01 5.74558377e-01
-1.17562199e+00 -5.80687761e-01 7.12222159e-01 -3.50119263e-01
1.20863104e+00 5.35465419e-01 1.84310466e-01 1.19997263e+00
5.99287570e-01 1.01737261e+00 5.58725476e-01 -3.50859910e-01
3.07562947e-01 -1.51886329e-01 3.52578551e-01 3.37326199e-01
6.07720733e-01 -8.14825952e-01 -7.68193483e-01 6.48297146e-02
-2.25982040e-01 -4.94078815e-01 -1.65085450e-01 6.43334270e-01
-1.36649430e+00 6.12048805e-01 -1.58179477e-02 4.32771355e-01
-6.61132932e-01 -1.41923362e-02 7.26693809e-01 1.33050963e-01
8.34628761e-01 8.33469391e-01 -6.72880039e-02 -1.65930748e-01
-1.71927702e+00 6.75160646e-01 9.85057473e-01 1.17380047e+00
1.27741143e-01 -1.36803344e-01 -8.96813571e-01 8.36476564e-01
2.12174699e-01 4.90519941e-01 8.00528944e-01 -1.28764188e+00
9.78571117e-01 3.84027153e-01 1.61026403e-01 -1.03466964e+00
-2.30289772e-01 -7.18963742e-01 -1.16013741e+00 -4.76727754e-01
-3.77069443e-01 -4.54222590e-01 -5.75803518e-01 1.29410541e+00
-2.12358594e-01 2.07305491e-01 5.06896913e-01 4.88168567e-01
1.62460339e+00 1.36848605e+00 -2.49705777e-01 -7.51703322e-01
1.08758652e+00 -1.31786287e+00 -1.12269115e+00 -1.42504141e-01
5.22150338e-01 -1.11979139e+00 5.61264932e-01 3.24831694e-01
-1.62274241e+00 -5.51709831e-01 -1.20160115e+00 -4.89221990e-01
-2.41863534e-01 6.52994692e-01 1.25653103e-01 -1.88447997e-01
-1.26574934e+00 8.00676346e-01 -3.50003809e-01 -4.28642407e-02
4.15183753e-01 -1.01586074e-01 -6.21015467e-02 -3.96710746e-02
-1.22430396e+00 9.05148208e-01 6.49436116e-01 1.89706370e-01
-7.70462990e-01 -8.80837142e-01 -6.73437715e-01 3.38067800e-01
-4.30458644e-03 -9.30242777e-01 1.36599183e+00 -1.80058435e-01
-1.42083764e+00 8.57204735e-01 -5.04820883e-01 -5.97701848e-01
4.23343569e-01 -3.34904253e-01 -1.94495708e-01 1.05458587e-01
5.52937448e-01 4.50272709e-01 2.68972009e-01 -1.22913325e+00
-7.23100424e-01 -1.24846414e-01 -6.12342179e-01 4.51867074e-01
-2.78673172e-01 2.68877178e-01 -5.25687099e-01 -5.40012002e-01
-4.19047266e-01 -3.75391185e-01 -8.75343755e-02 -9.02277529e-01
-1.10408592e+00 -6.46336079e-01 3.27765644e-01 -1.12808156e+00
1.56587112e+00 -1.76102626e+00 5.67042530e-01 -3.77010435e-01
1.87663466e-01 3.35052997e-01 -4.02286559e-01 7.95328617e-01
2.89231658e-01 1.63489118e-01 -3.64790529e-01 -8.57289910e-01
1.47433311e-01 -2.30685845e-01 -5.46779096e-01 -1.70712829e-01
6.47956952e-02 9.09575820e-01 -1.00383425e+00 -9.61035967e-01
-1.94897860e-01 -7.37826303e-02 -7.70096555e-02 4.14806604e-01
-4.95095402e-01 1.15021877e-01 -4.29823637e-01 1.66524410e-01
6.64412618e-01 -1.98476255e-01 8.48971531e-02 -1.96988112e-03
-4.84501779e-01 6.57967806e-01 -5.78413427e-01 1.95814216e+00
-4.12420958e-01 7.23500848e-01 -1.30021513e-01 -9.02655542e-01
7.46944129e-01 4.65446770e-01 2.91876704e-01 -5.07215977e-01
1.00381531e-01 5.20108581e-01 -2.04205737e-01 -7.47814834e-01
1.09870589e+00 3.04114163e-01 -2.74168223e-01 4.35006917e-01
1.68337241e-01 -2.93612421e-01 1.03342915e+00 7.48841703e-01
1.02080595e+00 -9.41320956e-02 3.12671393e-01 -2.84785002e-01
7.50255644e-01 1.29010707e-01 1.84811756e-01 8.49485397e-01
3.53909552e-01 8.03999901e-01 6.91902518e-01 7.98968077e-02
-1.13998890e+00 -8.88420224e-01 7.01922253e-02 5.78703046e-01
-1.77979887e-01 -6.13552034e-01 -1.00935972e+00 -4.23552483e-01
-2.69194514e-01 1.33402550e+00 -3.13467115e-01 -1.03569329e-01
-6.71461105e-01 -7.61726737e-01 6.16711259e-01 1.16067395e-01
4.39064801e-01 -1.38190138e+00 -2.42654398e-01 3.59888762e-01
-8.16892743e-01 -1.03921568e+00 -6.80069506e-01 -7.66974017e-02
-7.02859461e-01 -6.93069994e-01 -1.26003337e+00 -1.00319183e+00
3.40364307e-01 2.12187067e-01 1.19615531e+00 -4.05691057e-01
-1.03245132e-01 -2.62000978e-01 -3.57297599e-01 -5.50484657e-01
-7.09057212e-01 5.50262511e-01 -3.58187139e-01 -2.85074234e-01
1.18167438e-01 -3.13477188e-01 -2.36298621e-01 -6.70664370e-01
-7.39772499e-01 4.47380483e-01 9.40021753e-01 3.90491933e-01
4.87500370e-01 -1.06675662e-01 1.11114788e+00 -7.30919123e-01
1.39730465e+00 -6.41215146e-01 -4.20524403e-02 5.69893420e-01
-5.17455816e-01 1.29771575e-01 8.21704865e-01 -2.88361907e-02
-1.19368637e+00 -4.93441880e-01 -1.43369451e-01 -1.25295937e-01
1.84174806e-01 1.14258265e+00 -1.61056995e-01 7.75658488e-01
3.48220170e-01 6.69947147e-01 -2.91371912e-01 -6.30323946e-01
5.71801662e-01 1.24288070e+00 8.31364751e-01 -1.83525503e-01
1.22037306e-01 -2.14147523e-01 -1.37255117e-01 -8.51329982e-01
-1.49630547e+00 -3.65521938e-01 -3.03214550e-01 -2.46001706e-01
9.02263522e-01 -9.02455032e-01 -3.52210969e-01 5.01149714e-01
-1.94057190e+00 3.21212500e-01 -1.53449073e-01 3.06615233e-01
-4.49422836e-01 6.73222721e-01 -7.96559453e-01 -5.90710998e-01
-1.27372777e+00 -9.74738300e-01 1.20503628e+00 3.82334948e-01
-5.36367953e-01 -7.14423001e-01 1.74434498e-01 6.60769045e-01
2.17159986e-01 2.44213969e-01 8.72283518e-01 -8.02427888e-01
-1.56680852e-01 -2.13163003e-01 -1.74660057e-01 4.19877201e-01
-4.20920476e-02 1.42903835e-01 -5.71685374e-01 -5.07288538e-02
-7.28644505e-02 -2.53926039e-01 1.30455780e+00 7.00427353e-01
1.27860451e+00 -8.54745567e-01 -1.95958823e-01 2.84726918e-01
9.72920895e-01 2.36904383e-01 6.71055615e-01 1.34792730e-01
6.62715137e-01 5.37169039e-01 3.89582008e-01 4.70793307e-01
6.56050682e-01 2.37383083e-01 1.48071414e-02 4.07475054e-01
-4.41770405e-01 -2.60368865e-02 4.69755888e-01 1.74006796e+00
1.58824757e-01 -8.10511470e-01 -5.72755873e-01 6.10739768e-01
-1.95421982e+00 -1.20919859e+00 -5.65917015e-01 1.73464346e+00
1.29643357e+00 -2.74850875e-02 -3.70831579e-01 -5.56433089e-02
7.30837464e-01 5.32249093e-01 -2.90231705e-01 -6.21052563e-01
-2.93844372e-01 -2.64453795e-02 3.18035409e-02 5.50432086e-01
-9.23722446e-01 8.53859603e-01 5.57922459e+00 1.18601298e+00
-8.28164637e-01 -6.66956836e-03 5.72632551e-01 -3.29373986e-01
-3.37813377e-01 -3.69893014e-01 -8.32400143e-01 9.87451792e-01
1.37479305e+00 -9.51256812e-01 2.98983399e-02 5.30712485e-01
3.81230861e-01 -1.49831653e-01 -1.11625397e+00 8.92596245e-01
5.97974896e-01 -1.85297024e+00 4.32429820e-01 -3.42649311e-01
1.04999006e+00 6.87819943e-02 -1.09184042e-01 3.38406503e-01
2.00683311e-01 -8.97492349e-01 5.82035244e-01 9.47995901e-01
7.22302556e-01 -8.69498551e-01 1.05008495e+00 5.17393172e-01
-5.84616303e-01 3.70485246e-01 -5.10680318e-01 3.42472821e-01
3.11750770e-01 1.13005400e+00 -6.54027224e-01 1.01885200e+00
2.42372379e-01 1.02645159e+00 -5.13258755e-01 1.02079105e+00
-2.29211897e-01 6.34104133e-01 1.93067357e-01 -5.38947999e-01
2.90858328e-01 -2.33144671e-01 7.22935915e-01 1.71590912e+00
8.10622513e-01 -2.28071347e-01 -1.56111300e-01 1.03071606e+00
-7.37044334e-01 1.40139043e-01 -3.95123720e-01 -3.52764249e-01
4.60018337e-01 1.33242881e+00 -2.72873491e-01 -7.92797267e-01
1.56485036e-01 1.21226454e+00 1.98598176e-01 1.09398954e-01
-5.34115493e-01 -7.42776752e-01 -4.11239192e-02 -3.97648692e-01
1.30346864e-01 4.10015769e-02 -6.29751027e-01 -1.23338592e+00
1.94533363e-01 -8.87829602e-01 1.00645624e-01 -9.86848235e-01
-1.19194412e+00 5.88451922e-01 -1.41874909e-01 -1.02501094e+00
-4.56711084e-01 4.13762853e-02 -9.87726390e-01 9.54222322e-01
-1.38595223e+00 -9.63521183e-01 5.02574928e-02 -1.86600953e-01
1.18650961e+00 -4.99179423e-01 6.63116515e-01 2.60897130e-01
-7.66207397e-01 2.82487094e-01 6.94720566e-01 -2.28127286e-01
7.31922805e-01 -1.26908398e+00 5.13971210e-01 7.97808349e-01
1.88316386e-02 4.47635978e-01 9.44160104e-01 -8.67914259e-01
-1.33535790e+00 -1.38134933e+00 1.82399726e+00 -1.28733128e-01
6.27844453e-01 -7.26355463e-02 -7.87692130e-01 1.41276285e-01
8.21812570e-01 -8.65648627e-01 4.48711216e-01 -2.18828037e-01
2.05915734e-01 -1.03789322e-01 -6.40703261e-01 6.53860033e-01
6.20138109e-01 -9.03396755e-02 -9.24957693e-01 7.59533107e-01
1.03412664e+00 -3.93058211e-01 -6.80591404e-01 1.14719216e-02
2.04770520e-01 -4.71971750e-01 7.18599081e-01 -5.26412606e-01
1.55892050e+00 -5.67968599e-02 3.32016610e-02 -1.80217016e+00
-3.83278877e-01 -5.58567405e-01 -5.65795839e-01 1.56795621e+00
7.42430329e-01 4.85409312e-02 4.48303372e-01 -1.80627167e-01
-7.08970487e-01 -6.56571925e-01 -8.07846725e-01 -4.26553965e-01
4.45437521e-01 1.88836996e-02 3.58544946e-01 6.31417751e-01
1.98532730e-01 1.02031171e+00 -3.94119859e-01 -2.25958169e-01
8.17307472e-01 4.59815055e-01 3.83152425e-01 -1.05108702e+00
3.98762301e-02 -7.68853068e-01 2.04750076e-01 -7.62871325e-01
3.87774706e-01 -1.24350023e+00 1.96825638e-01 -2.45643878e+00
6.75543427e-01 5.43821692e-01 -8.24818835e-02 -2.40744781e-02
-4.72111493e-01 -1.91710621e-01 2.27660030e-01 3.43059093e-01
-1.11384630e+00 7.91470885e-01 1.26726913e+00 -4.86446589e-01
6.98074922e-02 -1.07495353e-01 -1.06867218e+00 4.07496274e-01
6.60379529e-01 -3.77922863e-01 -9.48617831e-02 -5.50070107e-01
2.82406926e-01 4.26113158e-01 -1.76458687e-01 -1.13092744e+00
5.88148654e-01 3.66478190e-02 2.47072861e-01 -1.25508606e+00
2.34410036e-02 1.01036265e-01 -1.85138091e-01 1.61772221e-01
-1.11091149e+00 7.43287280e-02 9.50054154e-02 1.39585525e-01
-3.90760332e-01 -6.41191840e-01 2.99645841e-01 -1.98541939e-01
-2.09934577e-01 -8.35503712e-02 -5.62976062e-01 3.70940953e-01
7.55011201e-01 4.24078912e-01 -5.39774776e-01 -4.85631853e-01
-2.00786456e-01 5.98240972e-01 -2.78600365e-01 3.99816662e-01
8.84173632e-01 -1.09095514e+00 -1.62770438e+00 -5.48636675e-01
-2.21128836e-01 1.55058324e-01 4.64075416e-01 6.94334745e-01
-6.70998991e-01 9.07936752e-01 4.77956459e-02 -3.53823043e-02
-1.14749610e+00 5.61212122e-01 -2.39584878e-01 -6.56668246e-01
-4.74239916e-01 6.99350774e-01 -9.92230028e-02 -1.63241342e-01
3.00711125e-01 -1.34380624e-01 -5.66556931e-01 3.93427372e-01
9.03324008e-01 8.29324841e-01 1.75541595e-01 -4.19794470e-01
-5.37852421e-02 2.81726807e-01 -3.82930815e-01 -2.27521226e-01
1.48635161e+00 -1.80454642e-01 -6.02895021e-01 4.78915006e-01
1.32236147e+00 -1.49277568e-01 -5.39383292e-01 -1.99334864e-02
3.85484695e-01 4.35747683e-01 2.44825706e-01 -1.05880463e+00
-6.72656417e-01 1.00906265e+00 -2.66287625e-01 1.40297115e-01
7.75213838e-01 -5.05435765e-02 1.16616642e+00 5.26686728e-01
-2.03935802e-01 -1.23941672e+00 -1.34114653e-01 7.75712311e-01
1.43885148e+00 -1.07503355e+00 2.96682954e-01 3.05228662e-02
-8.03574562e-01 1.06959224e+00 1.91658720e-01 -3.19995545e-02
-2.17250418e-02 2.40441769e-01 -2.76517868e-01 -1.25959948e-01
-1.02609575e+00 2.15069443e-01 6.00584447e-01 3.08692493e-02
7.71186113e-01 8.60671997e-02 -1.02309513e+00 9.07568455e-01
-4.97179508e-01 1.13096640e-01 7.77097344e-01 4.33479995e-01
-6.82705939e-01 -6.92305565e-01 -2.01947823e-01 8.50340188e-01
-7.31709659e-01 -3.76951128e-01 -8.07558954e-01 7.85784051e-02
-2.00857058e-01 1.21995711e+00 7.04680309e-02 -1.74980476e-01
1.18277363e-01 1.26878649e-01 1.96918279e-01 -6.11846030e-01
-5.71712613e-01 5.50726652e-02 3.74553084e-01 4.46491987e-02
-2.05655009e-01 -7.35106528e-01 -1.25319445e+00 -3.60932827e-01
1.24556534e-01 6.39752507e-01 8.89473319e-01 9.25490201e-01
6.85072303e-01 1.08781636e+00 6.22177482e-01 -9.12357867e-01
-8.20402265e-01 -1.52569151e+00 -3.55488509e-01 8.98253247e-02
1.73675478e-01 2.27901369e-01 -3.52390230e-01 2.14568615e-01] | [12.50955867767334, 9.556293487548828] |
7f76c06d-a3f8-41d1-8e40-7108e073cbd4 | reducing-the-gap-between-streaming-and-non | 2306.15171 | null | https://arxiv.org/abs/2306.15171v1 | https://arxiv.org/pdf/2306.15171v1.pdf | Reducing the gap between streaming and non-streaming Transducer-based ASR by adaptive two-stage knowledge distillation | Transducer is one of the mainstream frameworks for streaming speech recognition. There is a performance gap between the streaming and non-streaming transducer models due to limited context. To reduce this gap, an effective way is to ensure that their hidden and output distributions are consistent, which can be achieved by hierarchical knowledge distillation. However, it is difficult to ensure the distribution consistency simultaneously because the learning of the output distribution depends on the hidden one. In this paper, we propose an adaptive two-stage knowledge distillation method consisting of hidden layer learning and output layer learning. In the former stage, we learn hidden representation with full context by applying mean square error loss function. In the latter stage, we design a power transformation based adaptive smoothness method to learn stable output distribution. It achieved 19\% relative reduction in word error rate, and a faster response for the first token compared with the original streaming model in LibriSpeech corpus. | ["Ming'en Zhao", 'Genshun Wan', 'Jia Pan', 'Minghui Wu', 'Zhiqiang Ma', 'Yongchao Li', 'Dan Liu', 'Jiabin Xue', 'Lei Sun', 'Yu Fu', 'Haitao Tang'] | 2023-06-27 | null | null | null | null | ['speech-recognition'] | ['speech'] | [ 3.69040579e-01 9.74113867e-02 -6.49198368e-02 -4.36698914e-01
-1.09280288e+00 -2.53700495e-01 2.05964148e-01 2.06304207e-01
-4.96928543e-01 5.31811893e-01 4.64556247e-01 -2.53250539e-01
1.56370047e-02 -5.80505311e-01 -5.42361319e-01 -9.77497637e-01
1.28660128e-01 1.94313824e-01 6.32239103e-01 -9.65558439e-02
1.18662074e-01 -2.37770975e-01 -1.48001838e+00 6.03657305e-01
7.27171600e-01 1.07186115e+00 4.82414603e-01 9.66010988e-01
-5.35058439e-01 8.54045331e-01 -5.59475601e-01 -1.90735132e-01
-3.65437679e-02 -7.24623501e-01 -7.38585949e-01 -1.94234803e-01
1.19888159e-02 -8.87477621e-02 -2.82730013e-01 1.26746237e+00
7.95702457e-01 2.36953497e-01 6.55398786e-01 -9.35818434e-01
-4.67054069e-01 1.08481324e+00 -3.27119231e-01 5.44754446e-01
2.26910174e-01 -2.26161167e-01 1.13720143e+00 -6.00233853e-01
6.91259727e-02 1.20459688e+00 5.18814743e-01 5.56824386e-01
-8.93303096e-01 -7.40868568e-01 2.67352313e-01 5.32388508e-01
-1.36736262e+00 -7.77653515e-01 7.44039834e-01 -2.05162778e-01
9.65189219e-01 3.66576165e-01 2.58294791e-01 6.53266728e-01
6.20555058e-02 9.99978960e-01 8.09340179e-01 -6.28489614e-01
3.90113533e-01 2.56456494e-01 1.29826471e-01 3.01500052e-01
-3.08463871e-01 -1.62843287e-01 -6.90981269e-01 -1.37812003e-01
3.98963630e-01 -1.14927344e-01 -3.95114273e-01 1.10133134e-01
-9.82288897e-01 6.84234440e-01 -5.86925410e-02 4.76254702e-01
-3.03172737e-01 -6.73303530e-02 4.78137136e-01 5.09462893e-01
4.14019018e-01 -2.80930877e-01 -4.78012443e-01 -6.23889387e-01
-1.05171692e+00 7.72676468e-02 1.07553351e+00 9.71439302e-01
4.15924072e-01 2.22369581e-01 -2.68858612e-01 9.52200472e-01
5.53658247e-01 4.27599996e-01 9.22394693e-01 -5.62650144e-01
7.57310331e-01 1.26899764e-01 -2.37833261e-01 -8.08657944e-01
6.07620478e-02 -2.40701154e-01 -8.98885131e-01 -1.24559455e-01
1.88561961e-01 -2.46833235e-01 -9.34590697e-01 1.61095452e+00
3.63699347e-01 4.51298356e-01 3.91722322e-01 7.69549370e-01
7.36282051e-01 1.19858229e+00 -3.00442576e-02 -7.62875795e-01
1.11386430e+00 -8.37166011e-01 -1.18168926e+00 1.45104349e-01
5.53072870e-01 -7.34472096e-01 9.60434020e-01 5.11254311e-01
-1.23721361e+00 -3.68812442e-01 -1.02977204e+00 -4.01895121e-02
5.21756150e-02 -2.17914581e-01 1.70017675e-01 4.95900810e-01
-8.36707592e-01 3.96759689e-01 -8.62886906e-01 2.51985155e-02
1.12936094e-01 3.18146467e-01 -4.17293906e-02 1.72276378e-01
-1.48115814e+00 4.80097145e-01 6.88201010e-01 2.07574978e-01
-7.67699122e-01 -4.93251175e-01 -7.71922946e-01 3.90841961e-01
2.15272173e-01 -1.06333949e-01 1.59528077e+00 -1.07341456e+00
-2.08175397e+00 2.27160856e-01 -3.28213483e-01 -4.32012796e-01
2.63780951e-01 -1.27306879e-01 -5.25880396e-01 -5.68831526e-02
-2.84616560e-01 -3.31678614e-03 1.06656706e+00 -7.81717479e-01
-1.10125947e+00 -1.35239616e-01 -4.00715023e-01 4.88783717e-01
-5.40038228e-01 1.43389314e-01 -5.09692013e-01 -5.43272316e-01
3.22782397e-01 -3.98132771e-01 -7.28491917e-02 -5.30825257e-01
2.14620400e-02 -5.05505979e-01 8.13217223e-01 -1.06684685e+00
1.77173245e+00 -2.57417774e+00 1.45535856e-01 1.43241122e-01
-1.01865180e-01 3.76164556e-01 5.20263389e-02 4.46544588e-01
1.96393594e-01 -1.53188750e-01 -4.25971121e-01 -3.38228405e-01
-5.89065552e-02 2.90839046e-01 -5.20930707e-01 2.97729373e-01
-3.26609723e-02 4.12532121e-01 -9.22989607e-01 -6.63039565e-01
-2.36718774e-01 4.09677923e-01 -5.27289510e-01 4.98413414e-01
-3.23415607e-01 1.67361647e-01 -1.79773837e-01 2.83827722e-01
5.29520452e-01 2.00542986e-01 1.24034680e-01 2.02791095e-01
-1.65012151e-01 5.55992186e-01 -1.53689265e+00 1.61660361e+00
-3.85929912e-01 3.61881226e-01 2.05375716e-01 -1.21083808e+00
1.19395256e+00 8.64363730e-01 2.29643717e-01 -5.35380661e-01
1.08986147e-01 3.10509026e-01 -1.43176608e-03 -7.56854892e-01
4.90318179e-01 -5.95934927e-01 1.39078990e-01 3.03627163e-01
-2.63203867e-02 -3.41110900e-02 -3.21673512e-01 -9.53021049e-02
9.13816869e-01 -1.64111540e-01 -1.25863617e-02 -1.37309879e-02
5.99979877e-01 -2.44145274e-01 8.92994702e-01 4.29188043e-01
-9.38625261e-02 7.78955460e-01 5.80592036e-01 -4.14126068e-02
-8.70473623e-01 -8.10317814e-01 1.20347396e-01 1.17483294e+00
-1.00041758e-02 -1.87078387e-01 -6.81772947e-01 -5.13503373e-01
-1.83053926e-01 6.60562992e-01 -2.74277151e-01 -1.94459930e-01
-7.66866088e-01 -3.89855444e-01 6.61604881e-01 5.29975235e-01
5.18499374e-01 -1.07333672e+00 -9.96282250e-02 4.81510997e-01
-1.86511502e-01 -8.60533953e-01 -8.99311066e-01 2.71092266e-01
-8.10611188e-01 -4.67380285e-01 -7.14764774e-01 -1.17858529e+00
3.14304352e-01 2.80108638e-02 5.49019039e-01 -5.92766516e-02
3.47809702e-01 -2.15478241e-01 -6.40157998e-01 -3.42386842e-01
-4.38544929e-01 4.00206864e-01 7.96951652e-02 2.98884928e-01
3.45241457e-01 -4.98828918e-01 -3.92918438e-01 4.60032374e-03
-1.06458080e+00 -2.04840712e-02 6.17246687e-01 7.01057494e-01
5.80556273e-01 3.73479098e-01 8.56298745e-01 -7.86903739e-01
7.57173657e-01 -5.56672394e-01 -4.29683268e-01 1.87785178e-01
-7.36526430e-01 2.48196647e-01 6.28909349e-01 -6.25822842e-01
-1.19345772e+00 -1.04248216e-02 -5.03173292e-01 -2.75357366e-01
8.13455954e-02 7.11720169e-01 -3.28467906e-01 5.55920243e-01
2.56237179e-01 5.61746061e-01 3.95726934e-02 -6.81706488e-01
2.81873435e-01 1.35923100e+00 4.98926759e-01 -2.11589962e-01
3.88173461e-01 -1.06683254e-01 -6.39675677e-01 -9.03363585e-01
-5.57759464e-01 -6.71546340e-01 -3.59524012e-01 5.87457046e-02
8.32093656e-01 -7.57826209e-01 -5.73733151e-01 4.67823863e-01
-1.05588663e+00 -3.59619260e-01 -3.90474051e-01 8.55232954e-01
-4.69444513e-01 6.00709736e-01 -5.96137941e-01 -1.14315259e+00
-6.78447485e-01 -7.11055994e-01 7.76417017e-01 2.25799486e-01
8.98613334e-02 -9.27540839e-01 2.53701329e-01 4.64350432e-02
5.27268410e-01 -2.60530680e-01 8.54161203e-01 -9.05862808e-01
-1.98390573e-01 -1.10547826e-01 2.24444200e-03 6.84473276e-01
1.01993233e-01 -1.95979893e-01 -9.67145026e-01 -3.74175698e-01
4.61891919e-01 -1.55130371e-01 7.24432528e-01 2.16072574e-01
8.82937849e-01 -6.62919641e-01 3.66598107e-02 5.58316469e-01
1.23856294e+00 3.64496410e-01 6.29167199e-01 -1.26995429e-01
5.69714010e-01 4.30700362e-01 4.86698836e-01 3.09636116e-01
4.02132034e-01 4.59230363e-01 -2.10743994e-01 3.85219097e-01
-1.29134998e-01 -4.71654564e-01 6.93856955e-01 1.98404825e+00
2.69450784e-01 -1.67705163e-01 -7.52984345e-01 7.06757665e-01
-2.05580640e+00 -1.06345987e+00 1.12932324e-01 2.40225816e+00
1.36345017e+00 4.36600059e-01 1.01687245e-01 4.21624005e-01
7.98650086e-01 7.84593299e-02 -3.07500511e-01 -6.33054554e-01
6.56196699e-02 6.92988783e-02 2.52499163e-01 9.10605192e-01
-5.27063489e-01 8.19613874e-01 6.43043184e+00 1.13539016e+00
-1.21887517e+00 2.78062671e-01 2.96001285e-01 -7.44667277e-02
-6.13319218e-01 1.09828748e-01 -9.94525909e-01 7.45740294e-01
1.29078436e+00 -2.08406255e-01 1.90479115e-01 5.21815956e-01
2.84493029e-01 7.44016990e-02 -9.03480709e-01 1.04461181e+00
8.12932104e-02 -9.48699892e-01 -1.18107997e-01 -2.77081221e-01
4.37274635e-01 -2.47919738e-01 -2.74201334e-01 6.08815134e-01
3.07380911e-02 -9.47908759e-01 6.12982273e-01 6.16973221e-01
8.40312302e-01 -7.65950024e-01 8.34438562e-01 8.25194657e-01
-1.33138239e+00 -1.84124545e-03 -4.39883649e-01 -2.45491803e-01
1.71544224e-01 4.26524192e-01 -8.45131457e-01 4.13425386e-01
5.65693200e-01 2.43972167e-01 2.22138539e-01 1.03005040e+00
-1.53013080e-01 1.27618551e+00 -4.53349382e-01 -3.60166490e-01
9.29638445e-02 -1.00542769e-01 3.87199521e-01 1.48417258e+00
1.54107258e-01 3.85637403e-01 2.92066038e-01 2.03585148e-01
1.24565214e-01 4.03494477e-01 -2.29240209e-01 9.63103324e-02
6.94284797e-01 7.60434568e-01 -2.29210779e-01 -3.35975558e-01
-4.42546487e-01 9.05156493e-01 3.06253254e-01 4.22081739e-01
-7.01723874e-01 -8.41722727e-01 1.88196212e-01 3.73663642e-02
4.74358410e-01 -3.10221493e-01 -1.87399641e-01 -1.07851243e+00
1.20507419e-01 -8.07236969e-01 4.79919344e-01 -2.53575146e-01
-1.05982280e+00 5.69390416e-01 -3.46448831e-03 -9.21658456e-01
-3.93258154e-01 -3.90042290e-02 -8.22669923e-01 1.09471714e+00
-1.71603525e+00 -9.07123744e-01 3.63261141e-02 6.62111819e-01
8.93089771e-01 -1.23315759e-01 7.30723977e-01 4.69371647e-01
-4.70317125e-01 9.85792220e-01 2.49298334e-01 9.67067182e-02
4.75209683e-01 -1.19183719e+00 9.08417925e-02 7.29525805e-01
5.38808592e-02 1.63537830e-01 7.12770343e-01 -5.90488672e-01
-1.28547919e+00 -8.11454952e-01 1.19569123e+00 1.98606979e-02
5.22941887e-01 -3.30252975e-01 -1.42059517e+00 4.95231181e-01
2.80345559e-01 -1.97855622e-01 6.13493502e-01 -5.78775965e-02
-6.83011934e-02 -3.20458114e-01 -8.34974408e-01 2.67563224e-01
5.54073751e-01 -5.89209318e-01 -9.23089802e-01 3.93954813e-02
1.22491539e+00 -3.69107485e-01 -9.01036739e-01 3.57117444e-01
4.83973265e-01 -4.80355889e-01 2.56409019e-01 -4.32939410e-01
2.86876298e-02 -2.78954834e-01 -3.12936187e-01 -1.11695015e+00
-3.56126577e-01 -9.92413759e-01 -4.13713843e-01 1.53894794e+00
6.66236281e-01 -5.69529116e-01 7.67367899e-01 4.98766452e-01
-2.74060816e-01 -7.86414921e-01 -1.14170468e+00 -7.06470370e-01
1.96719289e-01 -5.40265143e-01 5.28285027e-01 6.99341655e-01
4.13905084e-01 5.12012839e-01 -3.20493817e-01 2.97571063e-01
2.68141031e-01 -1.56696990e-01 4.65677589e-01 -6.03353798e-01
-6.59330964e-01 -2.69236505e-01 -2.92557955e-01 -1.43218374e+00
-7.53325969e-02 -8.14886510e-01 5.15783966e-01 -1.42179859e+00
-2.81307362e-02 -4.22500253e-01 -5.72773576e-01 3.55963558e-01
-3.32343847e-01 -4.15233314e-01 -5.18086888e-02 3.67728889e-01
-5.57296634e-01 7.81112015e-01 9.02484238e-01 7.61479139e-02
-5.62377453e-01 3.14915389e-01 -3.64579141e-01 3.54574203e-01
9.84167278e-01 -5.18274069e-01 -6.02465868e-01 -6.54170156e-01
1.26098797e-01 2.90278584e-01 -2.65582353e-01 -9.25521433e-01
7.59383440e-01 -1.83261767e-01 -2.36462235e-01 -5.35782158e-01
1.92190573e-01 -6.86053097e-01 -8.46733376e-02 3.32081527e-01
-5.82610011e-01 7.56660243e-03 4.47319560e-02 6.68464959e-01
-5.33584774e-01 -4.61921245e-01 5.79231322e-01 1.34239689e-01
-5.29188693e-01 2.80968577e-01 -3.14001799e-01 2.62395322e-01
7.11970270e-01 -1.41174449e-02 1.18686818e-01 -5.00021219e-01
-6.51652336e-01 2.80719191e-01 -2.26638556e-01 4.65006143e-01
8.15757692e-01 -1.31223655e+00 -1.00485575e+00 4.82958496e-01
-1.72194034e-01 2.50476927e-01 5.78927025e-02 7.74970233e-01
-3.26175243e-01 5.58239445e-02 4.30986583e-01 -5.21273255e-01
-1.27019310e+00 3.09924752e-01 1.82203948e-01 -1.33918300e-01
-5.89541614e-01 9.10595417e-01 -2.29059130e-01 -3.19413781e-01
7.00363874e-01 -4.44595039e-01 -3.02903801e-01 -1.12223230e-01
8.04593027e-01 3.02803904e-01 1.44823417e-01 -3.97962779e-01
-1.89007536e-01 2.55539864e-01 -1.84798867e-01 -6.29890740e-01
1.09198880e+00 -2.98466921e-01 3.64709198e-02 8.49381983e-01
1.27064729e+00 1.37868449e-01 -1.23963714e+00 -6.43349648e-01
1.21781975e-01 -3.05984527e-01 2.08776742e-01 -3.47519726e-01
-8.59238088e-01 9.07438397e-01 5.80050945e-01 6.42250121e-01
1.03564894e+00 -1.04679376e-01 1.30441964e+00 1.19617887e-01
-1.09412093e-02 -1.21368408e+00 -2.72631526e-01 8.21775556e-01
6.66237473e-01 -1.03392196e+00 -3.09814274e-01 -2.62240618e-01
-6.99511170e-01 9.35213149e-01 3.95621687e-01 3.02109849e-02
9.96088266e-01 4.10508007e-01 2.33810350e-01 1.37934104e-01
-1.04650664e+00 -6.36703670e-02 1.09700352e-01 4.51210141e-01
5.24596870e-01 6.99188635e-02 -4.17044163e-01 8.46114874e-01
-2.97921330e-01 -1.31305903e-01 7.74812475e-02 9.47564185e-01
-9.21270192e-01 -9.02265847e-01 -2.83170372e-01 2.95412272e-01
-6.31096840e-01 -2.33713016e-01 2.05275878e-01 6.30479380e-02
-1.02791607e-01 1.17981458e+00 1.93339158e-02 -5.38479030e-01
4.62805837e-01 3.18257511e-01 1.46559343e-01 -8.06127846e-01
-5.19518614e-01 3.03390592e-01 -5.21310652e-03 -8.82540420e-02
-1.47352695e-01 -7.49090433e-01 -1.49609590e+00 -2.87524797e-02
-7.00535834e-01 6.89685404e-01 7.30866969e-01 1.03227007e+00
1.79475710e-01 4.62110966e-01 8.54129553e-01 -2.67916471e-01
-1.16240585e+00 -1.18625891e+00 -6.73790097e-01 2.94171035e-01
3.25036764e-01 2.41171732e-03 -5.53292155e-01 3.61591101e-01] | [14.54891586303711, 6.7562994956970215] |
789b39e0-597f-4c4a-9210-456087c4edfb | modeling-multi-scale-data-via-a-network-of | 2105.12226 | null | https://arxiv.org/abs/2105.12226v1 | https://arxiv.org/pdf/2105.12226v1.pdf | Modeling multi-scale data via a network of networks | Prediction of node and graph labels are prominent network science tasks. Data analyzed in these tasks are sometimes related: entities represented by nodes in a higher-level (higher-scale) network can themselves be modeled as networks at a lower level. We argue that systems involving such entities should be integrated with a "network of networks" (NoN) representation. Then, we ask whether entity label prediction using multi-level NoN data via our proposed approaches is more accurate than using each of single-level node and graph data alone, i.e., than traditional node label prediction on the higher-level network and graph label prediction on the lower-level networks. To obtain data, we develop the first synthetic NoN generator and construct a real biological NoN. We evaluate accuracy of considered approaches when predicting artificial labels from the synthetic NoNs and proteins' functions from the biological NoN. For the synthetic NoNs, our NoN approaches outperform or are as good as node- and network-level ones depending on the NoN properties. For the biological NoN, our NoN approaches outperform the single-level approaches for just under half of the protein functions, and for 30% of the functions, only our NoN approaches make meaningful predictions, while node- and network-level ones achieve random accuracy. So, NoN-based data integration is important. | ['Tijana Milenkovic', 'Pietro Hiram Guzzi', 'Meng Jiang', 'Shawn Gu'] | 2021-05-25 | null | null | null | null | ['data-integration'] | ['knowledge-base'] | [ 3.95248532e-01 1.02640808e+00 -2.26747274e-01 -3.59829605e-01
1.13806538e-01 -6.30346537e-01 5.56431472e-01 6.77937984e-01
1.14014685e-01 1.11303747e+00 -2.82397866e-01 -3.34709138e-01
-3.69905084e-01 -1.26768804e+00 -8.66444707e-01 -4.38461423e-01
-2.29105964e-01 8.77079785e-01 7.45881677e-01 -2.83255130e-01
-2.06679970e-01 5.94946921e-01 -1.44104826e+00 1.07150324e-01
9.45034742e-01 7.25243568e-01 -2.24650115e-01 4.45313662e-01
-2.25000396e-01 9.58337307e-01 -6.49926126e-01 -5.85441411e-01
1.88713744e-02 -4.13750678e-01 -1.03007305e+00 -2.75165588e-01
1.24835387e-01 5.83674431e-01 -1.12644300e-01 1.15895915e+00
1.60224900e-01 -1.92789853e-01 8.92868102e-01 -1.95222723e+00
-3.91281664e-01 9.37907040e-01 -2.33371288e-01 -4.45574731e-01
4.82812673e-01 -1.84058383e-01 1.30342209e+00 -5.60344219e-01
1.15540266e+00 1.39494896e+00 1.01536810e+00 4.08589870e-01
-1.70080352e+00 -4.56776500e-01 5.65263964e-02 -2.11118251e-01
-1.34860003e+00 3.38683240e-02 5.72618186e-01 -6.45748734e-01
7.39855826e-01 1.99564174e-01 6.63849294e-01 9.14611816e-01
1.16409808e-01 3.54484133e-02 9.65337515e-01 -2.58948654e-01
1.36596099e-01 1.94947630e-01 5.67946732e-01 8.83815825e-01
7.25736439e-01 -3.12051833e-01 -2.60383636e-01 -4.09120917e-01
4.79735166e-01 -1.31759763e-01 -1.67447552e-01 -2.51276702e-01
-1.52140415e+00 4.75253165e-01 6.48022175e-01 7.47416794e-01
-2.96856195e-01 1.64094165e-01 3.17026496e-01 4.77863997e-01
7.41625905e-01 7.73768842e-01 -1.02203715e+00 3.24379683e-01
-5.22547305e-01 -1.34713545e-01 1.58063161e+00 1.26189649e+00
1.15765107e+00 -1.58651397e-01 1.29099553e-02 5.16894102e-01
1.67534560e-01 -1.75980181e-01 1.60271451e-01 -7.07138538e-01
-1.05454482e-01 1.28068173e+00 -9.45841745e-02 -1.27559996e+00
-9.21496809e-01 -5.32140613e-01 -1.09173965e+00 -5.28593198e-04
6.73464239e-01 -1.36467576e-01 -7.33726621e-01 1.93354475e+00
4.76579875e-01 3.28065157e-01 9.53938141e-02 1.78173959e-01
1.33319545e+00 5.07916808e-01 3.22466731e-01 -4.45201248e-01
1.48751056e+00 -1.02667475e+00 -7.07358003e-01 1.88830093e-01
1.21946716e+00 -4.04455662e-01 5.78554511e-01 1.26661792e-01
-6.86466098e-01 -6.11146986e-01 -8.39753091e-01 2.29931518e-01
-1.08998513e+00 -4.62485775e-02 7.07651317e-01 4.19532478e-01
-1.34199178e+00 1.01665306e+00 -2.20254034e-01 -6.81163073e-01
4.07303981e-02 6.31169438e-01 -8.54545176e-01 3.00853044e-01
-1.43643200e+00 9.08265829e-01 6.86937094e-01 -4.13381189e-01
-5.17070413e-01 -7.70066977e-01 -7.79271662e-01 1.46926790e-01
4.37075555e-01 -7.57406414e-01 5.06736934e-01 -9.47120965e-01
-8.66871774e-01 1.08850944e+00 -6.06611669e-02 -2.01187715e-01
2.35654503e-01 5.56804597e-01 -6.07683539e-01 1.43150659e-02
5.09004705e-02 7.65927434e-01 1.96682289e-01 -1.44396424e+00
-2.65705436e-01 -6.72671422e-02 3.50763470e-01 -3.16721261e-01
-2.32632279e-01 -1.85437247e-01 7.59381279e-02 -4.59158897e-01
2.81476825e-01 -1.01486695e+00 -2.03370661e-01 1.15692168e-01
-1.02317286e+00 -6.27792120e-01 6.50535047e-01 -2.98531324e-01
1.11098409e+00 -1.66419029e+00 1.01347491e-01 6.04508221e-01
1.01226008e+00 1.06959350e-01 -3.72522593e-01 7.81520247e-01
-7.29512513e-01 8.69960785e-01 -2.28923541e-02 1.05281673e-01
-9.36722085e-02 1.19624503e-01 4.81130123e-01 1.77123860e-01
2.38010913e-01 9.22074020e-01 -9.80398536e-01 -5.95174849e-01
-2.11847469e-01 2.29197294e-01 -2.77072906e-01 -2.41730064e-02
-5.44687152e-01 5.00844538e-01 -3.20450753e-01 7.75857627e-01
4.35028851e-01 -9.55516636e-01 7.21964896e-01 -6.02184772e-01
4.06154066e-01 1.71069384e-01 -9.46040034e-01 8.92440617e-01
-3.03131156e-02 4.44817632e-01 -5.61463507e-03 -1.26386058e+00
1.26605713e+00 4.44308788e-01 7.38657355e-01 9.77017805e-02
-9.29058120e-02 2.84322053e-01 5.82079291e-01 -2.19065741e-01
-1.10134922e-01 2.40427449e-01 1.68079510e-01 2.96652406e-01
3.36062402e-01 6.51857108e-02 5.06686628e-01 4.68187988e-01
1.69173741e+00 -2.94918846e-02 6.64176524e-01 -5.02667546e-01
7.05921412e-01 3.81287411e-02 7.42600083e-01 4.43253338e-01
-4.02059630e-02 3.76133025e-02 1.42178285e+00 -3.46534610e-01
-9.48680520e-01 -5.58902800e-01 -2.46799126e-01 1.09605026e+00
2.21722156e-01 -8.32868934e-01 -8.35354626e-01 -1.12544405e+00
1.93919644e-01 2.39013463e-01 -6.51543975e-01 -1.38430387e-01
-1.85050815e-01 -7.49024332e-01 6.73981845e-01 -1.88774198e-01
8.65381658e-02 -1.06240141e+00 7.77200401e-01 2.54237920e-01
1.93328578e-02 -1.30281043e+00 1.28043637e-01 3.25012207e-01
-7.32242882e-01 -1.58179867e+00 -2.44514853e-01 -1.01024032e+00
1.06465900e+00 -2.54715443e-01 1.54532444e+00 5.64520955e-01
1.28820226e-01 -1.76136911e-01 -4.93460834e-01 -1.53942049e-01
-9.56153452e-01 3.69006813e-01 1.78029060e-01 5.30642606e-02
2.53086984e-01 -1.16905499e+00 -2.57123500e-01 7.15963781e-01
-7.99082935e-01 2.24225014e-01 6.63975537e-01 7.64995813e-01
4.50112373e-01 1.91974819e-01 1.14189672e+00 -1.64420247e+00
6.61457658e-01 -8.09613585e-01 -3.22930127e-01 4.72977728e-01
-1.02380586e+00 5.28499372e-02 8.38105500e-01 -5.76073050e-01
-2.70412982e-01 1.08702257e-01 -1.18223786e-01 1.95747703e-01
-3.02092522e-01 7.51379311e-01 -4.29072171e-01 -3.58815938e-01
8.64205182e-01 -2.28895590e-01 -1.68224871e-01 -3.33742291e-01
1.95114315e-01 3.60966414e-01 -1.25821650e-01 -6.77553952e-01
7.87762046e-01 -2.19024383e-02 8.08994710e-01 -5.96952796e-01
-7.34320402e-01 -2.35431418e-01 -8.81274700e-01 -3.87444109e-01
7.40499020e-01 -5.63746929e-01 -1.13960755e+00 1.49753645e-01
-1.33981311e+00 -3.39426287e-02 -3.60681891e-01 1.65198699e-01
-3.42087567e-01 1.67748928e-01 -7.39699960e-01 -5.80305099e-01
-2.15984546e-02 -9.81143177e-01 9.47080851e-01 -7.45506510e-02
-4.43594009e-01 -1.39770758e+00 7.61391446e-02 6.15721010e-02
1.27812326e-01 8.66210043e-01 1.42795765e+00 -1.51167846e+00
-4.60089684e-01 -3.26947421e-01 -5.30713201e-01 1.59190714e-01
3.99584025e-01 3.33319604e-01 -8.03678870e-01 1.25749812e-05
-6.31718099e-01 -1.49092481e-01 4.32201862e-01 -4.41957265e-02
9.33109343e-01 -2.82542080e-01 -8.74203324e-01 1.29414827e-01
1.25635481e+00 -1.24868311e-01 3.47607225e-01 -5.76138869e-02
9.03918684e-01 1.14198828e+00 3.73689204e-01 5.95069490e-02
2.59675175e-01 5.14999568e-01 5.52879214e-01 -3.17620575e-01
-8.63648504e-02 -3.12228590e-01 1.36026874e-01 1.02382362e+00
-2.29640111e-01 -5.98673940e-01 -1.17110372e+00 2.10478067e-01
-2.08365154e+00 -6.04128361e-01 -9.56792474e-01 1.69793212e+00
1.07584524e+00 3.21156889e-01 1.83691934e-01 1.37392357e-01
1.22435796e+00 -1.51795462e-01 -6.96332455e-01 9.48140100e-02
-4.39629376e-01 4.75507267e-02 2.62960553e-01 4.03651208e-01
-8.74027133e-01 9.34656680e-01 6.33265114e+00 1.00319600e+00
-3.81838560e-01 4.36341874e-02 7.79225528e-01 8.14449310e-01
-4.34561223e-01 3.77526551e-01 -7.35434115e-01 5.06679714e-01
1.32241595e+00 -2.37288266e-01 5.76787628e-02 8.58004272e-01
9.20316279e-02 1.67643771e-01 -1.50677514e+00 5.22425473e-01
-5.20318806e-01 -1.41071451e+00 3.78712304e-02 3.42545152e-01
8.01945269e-01 -9.24656466e-02 -8.75319779e-01 2.39372864e-01
7.61119843e-01 -1.08670354e+00 1.71294153e-01 1.04883730e+00
6.97633803e-01 -3.28296423e-01 1.00070572e+00 7.21850276e-01
-1.34447145e+00 4.42294538e-01 -2.87715644e-01 -2.23389771e-02
-6.60162568e-02 1.10835099e+00 -7.63778985e-01 8.23205352e-01
2.05689266e-01 1.01250732e+00 -7.27706730e-01 7.16249347e-01
-2.71937251e-01 7.00837076e-01 -3.87802184e-01 -2.71168411e-01
-2.40643844e-01 -5.57896197e-01 4.23467189e-01 1.12691319e+00
7.54189193e-02 -9.42009389e-02 2.93707669e-01 8.00806642e-01
-6.77676558e-01 2.03084186e-01 -8.95514250e-01 -3.08522135e-01
5.49244344e-01 1.70752513e+00 -9.60212469e-01 -5.21167159e-01
-3.59057754e-01 4.44301724e-01 4.66105193e-01 4.08942163e-01
-6.85736537e-01 -4.53149527e-01 5.73192358e-01 3.99925202e-01
-4.03670758e-01 1.92703918e-01 9.32332203e-02 -1.07052159e+00
-2.59933352e-01 -6.79055810e-01 1.56648800e-01 -9.11161244e-01
-1.78679633e+00 8.31431627e-01 -3.11489403e-01 -1.24545026e+00
-3.79917994e-02 -7.10603237e-01 -2.93263167e-01 6.88371003e-01
-1.29369879e+00 -1.33876681e+00 -3.57281119e-01 3.28300565e-01
-2.11242244e-01 -2.80616637e-02 1.15973139e+00 3.02498072e-01
-5.33902764e-01 3.56334567e-01 1.21391648e-02 1.93198368e-01
6.23529315e-01 -1.29056752e+00 3.49840552e-01 1.20021403e-01
2.94969827e-02 5.49748123e-01 6.34285867e-01 -1.00721550e+00
-8.41461539e-01 -1.28663182e+00 1.23070800e+00 -5.41470110e-01
9.89480078e-01 -7.23796546e-01 -8.95930111e-01 7.15433478e-01
-9.62046397e-05 3.07612687e-01 7.75469065e-01 1.41675264e-01
-2.66792655e-01 3.54181342e-02 -1.50510156e+00 4.63714510e-01
1.65868461e+00 -4.34158206e-01 -2.59302467e-01 1.03939342e+00
1.25413871e+00 3.57017130e-01 -1.69265616e+00 7.68849194e-01
2.58140415e-01 -6.77288830e-01 8.87030840e-01 -1.04399943e+00
3.18242580e-01 -5.19090772e-01 8.82792547e-02 -1.20309293e+00
-4.19336051e-01 -5.41596591e-01 -1.92049339e-01 1.54241860e+00
9.78727818e-01 -1.08529627e+00 1.05378222e+00 3.98449481e-01
1.22230269e-01 -6.89619303e-01 -4.86422688e-01 -9.24372256e-01
-2.43185580e-01 -6.26793727e-02 5.00237525e-01 1.56150842e+00
1.40688792e-01 8.47146094e-01 -1.55910835e-01 1.78462674e-03
5.35550177e-01 -1.52702183e-01 6.81751013e-01 -2.18937683e+00
-1.50092199e-01 -2.28936404e-01 -7.90969431e-01 -6.71600938e-01
6.29946053e-01 -1.18278289e+00 -4.30906922e-01 -1.76642203e+00
4.02790233e-02 -6.61626756e-01 -1.20158717e-01 7.60079324e-01
8.70561041e-03 1.94767758e-01 -2.54147947e-01 5.06644070e-01
-5.28376102e-01 1.83661282e-01 1.29356599e+00 -2.06117570e-01
1.85986400e-01 9.96071566e-03 -5.54597914e-01 9.84090269e-01
6.81094468e-01 -6.94786310e-01 -2.47980163e-01 6.28400207e-01
8.40703905e-01 2.29650691e-01 5.12463264e-02 -9.53466773e-01
3.38746399e-01 -8.00969377e-02 -8.42107367e-03 -3.57367456e-01
3.27164352e-01 -1.08473539e+00 6.45317197e-01 5.24659812e-01
-4.01828766e-01 -2.29865331e-02 -3.51017684e-01 8.68154466e-01
-3.47368240e-01 -3.10166419e-01 5.28411925e-01 -1.26321375e-01
-1.13851056e-01 4.21217591e-01 -2.43724003e-01 1.42697208e-02
1.10792387e+00 -2.99681634e-01 -7.47898400e-01 -3.97873789e-01
-1.41143799e+00 1.38420820e-01 5.40451169e-01 -4.11798730e-02
1.79546550e-01 -1.10149443e+00 -5.44726133e-01 -1.60962164e-01
2.31362239e-01 -1.46703154e-01 -3.68299961e-01 9.26907241e-01
-5.92183053e-01 3.05006951e-01 -1.28912687e-01 -4.31089133e-01
-1.18754089e+00 5.29082894e-01 3.27746838e-01 -6.39582813e-01
1.21544704e-01 6.04356766e-01 4.21667904e-01 -1.19452322e+00
-5.60917035e-02 -2.10550651e-01 -5.24512529e-01 2.44487345e-01
-2.09715173e-01 4.77851570e-01 -3.85976247e-02 -7.76079059e-01
-1.92299068e-01 5.45175374e-01 2.93117523e-01 5.01723886e-01
1.33117366e+00 2.03312308e-01 -8.51167798e-01 6.46085262e-01
9.60858047e-01 -5.96841611e-03 -4.23381895e-01 -2.18061551e-01
4.70536917e-01 2.35775784e-01 -7.21042097e-01 -7.71897852e-01
-8.39755535e-01 4.77054596e-01 -1.64975628e-01 1.10664880e+00
7.38811731e-01 2.88366914e-01 2.14368805e-01 5.90055823e-01
5.65508962e-01 -4.70290422e-01 -1.06721476e-01 3.69814068e-01
7.06032395e-01 -1.12330365e+00 -5.26787899e-02 -1.40012729e+00
-9.64707136e-02 1.14483225e+00 8.11463118e-01 -1.37547135e-01
1.20516217e+00 1.65435523e-01 -5.45356631e-01 -5.36059499e-01
-9.40309227e-01 -3.59453291e-01 3.31839740e-01 7.15114772e-01
6.35281920e-01 1.03465676e-01 -4.97899204e-01 5.44985354e-01
-6.82923943e-02 -2.81952679e-01 4.97620434e-01 4.86137658e-01
-4.05232400e-01 -1.57733846e+00 6.44728243e-02 9.67156112e-01
-3.26585352e-01 -2.81145751e-01 -1.04549432e+00 9.79525387e-01
4.04680461e-01 9.51244831e-01 -2.83334881e-01 -7.89984882e-01
3.96063626e-01 2.23625943e-01 -1.70211777e-01 -9.23285127e-01
-8.15471292e-01 -3.36791217e-01 9.15454984e-01 -3.84728253e-01
-7.65178323e-01 -2.07634002e-01 -1.39372027e+00 -7.25349605e-01
-7.33149648e-01 5.71418941e-01 4.64351922e-01 7.83222556e-01
6.16592646e-01 7.21905947e-01 4.15165544e-01 -4.87062395e-01
6.99435622e-02 -1.28648221e+00 -9.47154284e-01 5.72554529e-01
-2.76614487e-01 -5.90344071e-01 -6.60776138e-01 5.71647771e-02] | [6.814568519592285, 5.625408172607422] |
a12e5696-84e6-4189-b072-8f3f8258dd6d | sequence-to-sequence-models-for-extracting | 2201.05658 | null | https://arxiv.org/abs/2201.05658v1 | https://arxiv.org/pdf/2201.05658v1.pdf | Sequence-to-Sequence Models for Extracting Information from Registration and Legal Documents | A typical information extraction pipeline consists of token- or span-level classification models coupled with a series of pre- and post-processing scripts. In a production pipeline, requirements often change, with classes being added and removed, which leads to nontrivial modifications to the source code and the possible introduction of bugs. In this work, we evaluate sequence-to-sequence models as an alternative to token-level classification methods for information extraction of legal and registration documents. We finetune models that jointly extract the information and generate the output already in a structured format. Post-processing steps are learned during training, thus eliminating the need for rule-based methods and simplifying the pipeline. Furthermore, we propose a novel method to align the output with the input text, thus facilitating system inspection and auditing. Our experiments on four real-world datasets show that the proposed method is an alternative to classical pipelines. | ['Rodrigo Nogueira', 'Roberto A. Lotufo', 'Guilherme Rosa', 'Fábio C. de Souza', 'Ramon Pires'] | 2022-01-14 | null | null | null | null | ['open-information-extraction'] | ['natural-language-processing'] | [ 5.79337001e-01 2.02747747e-01 -1.57864783e-02 -4.93016481e-01
-6.95417106e-01 -1.03620696e+00 6.04606628e-01 1.01431406e+00
-5.14098942e-01 5.96380234e-01 -2.44058549e-01 -6.67364061e-01
1.13942148e-02 -8.94178510e-01 -6.24676049e-01 -8.67373496e-02
2.25058511e-01 1.88854471e-01 5.44314802e-01 6.92284405e-02
5.89767873e-01 4.07609910e-01 -1.59927154e+00 6.69358969e-01
1.18135130e+00 6.29264474e-01 1.65138349e-01 5.05446970e-01
-5.73185384e-01 7.73432076e-01 -1.04153836e+00 -7.88543940e-01
2.29663432e-01 -3.57807338e-01 -8.37808907e-01 5.78442626e-02
6.77571818e-02 -1.03731111e-01 4.94535834e-01 1.09680951e+00
7.07330927e-02 -3.96974653e-01 4.23298568e-01 -1.05366731e+00
2.54658889e-02 8.05948555e-01 -2.92162269e-01 -4.09994841e-01
5.59807062e-01 2.50102967e-01 8.51403952e-01 -5.63785613e-01
7.46147513e-01 7.50649035e-01 5.88982344e-01 2.87973106e-01
-9.87008691e-01 -2.11440533e-01 -2.61678901e-02 1.87494278e-01
-1.06452453e+00 -3.72528613e-01 4.76169080e-01 -7.55487084e-01
1.15785062e+00 3.76798749e-01 4.77083534e-01 7.87136018e-01
2.89418697e-01 5.88533044e-01 9.54640806e-01 -8.35177779e-01
4.46459830e-01 5.06982923e-01 4.06597525e-01 6.37901604e-01
7.47889519e-01 -4.20178294e-01 -1.65703878e-01 -4.07363713e-01
1.29620910e-01 -1.11573324e-01 2.02454969e-01 -1.52483240e-01
-6.72077417e-01 4.57935244e-01 -4.32368308e-01 4.92433161e-01
-3.62200201e-01 -3.22664589e-01 6.54233277e-01 1.74709082e-01
1.44012302e-01 4.99035567e-01 -5.60051978e-01 -2.27799863e-01
-1.24732804e+00 1.54206038e-01 9.48311806e-01 1.17371213e+00
1.00188172e+00 -2.90496409e-01 -2.51981258e-01 3.92910928e-01
2.29077786e-01 7.20105469e-02 4.73347008e-01 -3.06631267e-01
6.84140086e-01 1.15667856e+00 4.20481861e-01 -6.98671460e-01
-2.97574222e-01 -4.42498624e-01 -2.61702091e-01 2.05411062e-01
5.04153132e-01 -3.58765759e-02 -7.54112065e-01 9.94512856e-01
4.11932826e-01 -3.57008666e-01 -5.04010767e-02 2.53187656e-01
3.20948988e-01 3.41669708e-01 -5.72272390e-02 -3.72868180e-01
1.43016922e+00 -5.58261275e-01 -7.63361454e-01 -1.62372023e-01
9.94450212e-01 -9.39214468e-01 8.99369657e-01 7.03174472e-01
-9.10280526e-01 -4.76554364e-01 -1.01178706e+00 4.15803269e-02
-4.97521341e-01 4.09163594e-01 4.13800061e-01 8.85540187e-01
-4.48150545e-01 8.57100010e-01 -6.46522462e-01 -1.93637684e-01
1.55801788e-01 1.30347818e-01 -3.42160046e-01 1.57867357e-01
-9.02854741e-01 6.77812815e-01 7.07560360e-01 2.35093579e-01
-3.49594355e-01 -6.05346799e-01 -9.23116028e-01 1.29264519e-01
7.35746741e-01 -1.65768310e-01 1.47679448e+00 -5.09284794e-01
-1.52716136e+00 5.08634210e-01 -2.22741485e-01 -2.60920644e-01
6.56883240e-01 -3.91279846e-01 -2.94083089e-01 -1.16831854e-01
1.40483856e-01 -1.67043000e-01 8.65541935e-01 -9.30199921e-01
-9.22356308e-01 -1.67254359e-01 1.09073512e-01 -6.86691344e-01
-3.37618053e-01 4.66784924e-01 -4.25140589e-01 -4.39146936e-01
-4.79059339e-01 -5.82868814e-01 -3.19595456e-01 -6.19483709e-01
-6.38145864e-01 -2.85107881e-01 4.15713072e-01 -9.62639093e-01
1.53790855e+00 -2.02593613e+00 -2.53885508e-01 4.65471476e-01
-1.22975051e-01 5.51586807e-01 2.23589391e-02 8.16492617e-01
-1.80134669e-01 4.48207676e-01 -4.15703237e-01 -2.02412620e-01
1.31235793e-01 -6.47451654e-02 -2.41314590e-01 8.05285871e-02
6.81962609e-01 4.93658990e-01 -9.27254975e-01 -6.36509895e-01
7.62540521e-03 -7.50714839e-02 -5.59103847e-01 3.52567852e-01
-6.44549429e-01 3.56465876e-01 -3.25273842e-01 5.49386024e-01
7.20177710e-01 1.53514847e-01 6.22628331e-01 6.54693693e-02
-5.52993238e-01 8.61722112e-01 -1.44845498e+00 1.59593570e+00
-6.40125275e-01 1.55410364e-01 -1.44862041e-01 -1.07771575e+00
1.02486432e+00 3.25805098e-01 2.66427789e-02 -4.92605716e-01
1.80942535e-01 2.77732670e-01 -1.69262141e-01 -8.66686344e-01
7.37823665e-01 3.95912588e-01 -3.52815151e-01 5.99285126e-01
8.39432050e-03 -8.99011735e-03 9.03942347e-01 7.22960383e-02
1.00991905e+00 7.16477275e-01 6.06113791e-01 2.51720071e-01
6.46513283e-01 3.02670270e-01 8.06914508e-01 6.90431416e-01
5.27781606e-01 1.41163886e-01 1.05644894e+00 -2.66215503e-01
-1.02495790e+00 -3.22393537e-01 1.45450190e-01 5.94583452e-01
-4.89115983e-01 -1.10520101e+00 -1.19295812e+00 -9.75242257e-01
-3.36597174e-01 9.24639583e-01 -2.58156598e-01 4.87540029e-02
-6.64240241e-01 -5.05583525e-01 7.96142638e-01 3.68468642e-01
-9.47490409e-02 -8.63588214e-01 -8.60869288e-01 5.38791418e-01
-1.37119740e-01 -1.10018206e+00 -4.57986034e-02 3.17950189e-01
-8.34578037e-01 -1.37773335e+00 5.63740209e-02 -3.86049300e-01
8.98693323e-01 -4.19448763e-01 7.09031105e-01 3.53163898e-01
-3.71885598e-01 -4.18666363e-01 -7.46299148e-01 -4.13844436e-01
-1.00695550e+00 1.83651984e-01 -2.78896958e-01 9.90578234e-02
3.88673246e-01 -2.18203694e-01 -2.41450197e-03 8.05618241e-02
-1.31818473e+00 -2.16803886e-02 7.79884279e-01 6.38522863e-01
3.42907667e-01 2.88195789e-01 2.83763081e-01 -1.49737382e+00
6.54427230e-01 -1.66129142e-01 -1.28684437e+00 6.11377060e-01
-8.90094757e-01 4.02572751e-01 1.18830240e+00 -1.64032146e-01
-1.16951478e+00 4.84012425e-01 -2.22591251e-01 2.12705910e-01
-5.54553747e-01 8.28060389e-01 -4.36699897e-01 4.14869845e-01
5.10636330e-01 2.31227487e-01 -2.83500135e-01 -9.32145000e-01
2.64005989e-01 9.64052260e-01 3.47415417e-01 -6.34088457e-01
1.14025545e+00 -1.15561403e-01 -3.45115095e-01 -6.44315779e-01
-3.72520179e-01 -1.92888856e-01 -9.97364700e-01 1.50705069e-01
4.09039438e-01 -3.36675107e-01 -5.38176954e-01 4.01172370e-01
-1.51501465e+00 -1.96471617e-01 -3.19036424e-01 1.67044342e-01
-1.82757825e-01 7.31613219e-01 -4.01496917e-01 -8.40293229e-01
-3.26795876e-01 -1.22922790e+00 8.44763696e-01 1.54447168e-01
-4.63578284e-01 -4.40059245e-01 4.04108316e-02 3.15920651e-01
7.39555806e-02 1.64992824e-01 1.10605454e+00 -9.76774096e-01
-5.55853724e-01 -5.83378077e-01 -4.31164801e-02 3.55945289e-01
1.95510328e-01 6.04869843e-01 -8.74712706e-01 3.06075551e-02
-6.61504408e-03 6.03931732e-02 3.04840714e-01 -3.79495144e-01
1.00652599e+00 -5.75522006e-01 -2.62238950e-01 2.76904583e-01
1.24539316e+00 3.80851746e-01 7.45234370e-01 5.62954128e-01
3.52392137e-01 9.58338261e-01 9.51591849e-01 5.21055758e-01
4.45558056e-02 7.48075545e-01 1.55793935e-01 1.80371970e-01
2.30420962e-01 -4.69472080e-01 3.95376503e-01 5.30555904e-01
4.69078869e-02 -1.23367839e-01 -8.59942436e-01 5.64478159e-01
-1.72903359e+00 -9.05473232e-01 -4.89791155e-01 2.35825682e+00
1.13601351e+00 3.51199865e-01 2.57304370e-01 6.10052049e-01
5.63756227e-01 -3.15489739e-01 -2.07479298e-02 -6.07679844e-01
4.88605767e-01 5.01471102e-01 4.38210130e-01 5.04383504e-01
-9.04949129e-01 8.91539752e-01 5.52048254e+00 6.99736774e-01
-1.16300976e+00 -4.93654981e-02 -6.06428683e-02 1.58740833e-01
-3.83773118e-01 4.56554234e-01 -1.20035350e+00 5.36236167e-01
1.22063565e+00 -1.07405432e-01 -1.77952461e-02 9.45577264e-01
3.21566671e-01 -1.54534176e-01 -1.33835125e+00 4.24366921e-01
-2.13077396e-01 -1.28192127e+00 -6.29053563e-02 -2.12442111e-02
1.69395491e-01 -4.30250466e-01 -6.29841268e-01 7.14401454e-02
2.23091036e-01 -5.92065930e-01 1.14981961e+00 4.19915020e-01
6.43258631e-01 -6.86405182e-01 1.03652239e+00 4.63385612e-01
-9.47625458e-01 -5.28047122e-02 -2.08670869e-01 -2.16802210e-02
1.53225794e-01 9.17991281e-01 -1.20901179e+00 1.11967516e+00
3.88620347e-01 3.65879238e-01 -8.43476117e-01 1.22692275e+00
-6.63283229e-01 4.88887519e-01 -1.68629333e-01 -2.97345072e-01
-1.15712695e-01 -5.78883551e-02 2.87716419e-01 1.80423343e+00
1.94532692e-01 -3.94609779e-01 1.10203147e-01 8.32957149e-01
2.31190786e-01 3.30257356e-01 -3.77488106e-01 -2.68897176e-01
2.71074265e-01 1.42358387e+00 -8.75947237e-01 -5.14057815e-01
-4.55375552e-01 7.85963297e-01 1.07484691e-01 1.00767307e-01
-6.41037583e-01 -1.14176929e+00 2.56084472e-01 2.90955216e-01
5.39464414e-01 -2.75842488e-01 -2.52037913e-01 -1.23527098e+00
5.79246759e-01 -1.17068517e+00 2.47863173e-01 -2.88318247e-01
-6.20375335e-01 6.61930740e-01 1.09977677e-01 -1.36357832e+00
-5.91007769e-01 -4.91386324e-01 -4.78115082e-01 7.46197999e-01
-1.46106923e+00 -8.31036031e-01 -2.78218426e-02 2.14406282e-01
3.20088029e-01 1.19719505e-01 7.52944112e-01 4.19211507e-01
-6.29838645e-01 7.53314197e-01 -1.23409823e-01 3.88216406e-01
6.93535268e-01 -1.19040191e+00 5.56406379e-01 1.34683883e+00
5.03730178e-02 1.16232669e+00 6.15951896e-01 -8.40451360e-01
-1.30892551e+00 -1.19708133e+00 1.31377709e+00 -2.15835020e-01
7.82409608e-01 -7.93468416e-01 -9.59675908e-01 4.54076290e-01
2.93112278e-01 -5.27531266e-01 8.37761760e-01 -2.79718470e-02
-6.15986943e-01 -2.78128147e-01 -1.07136452e+00 2.48615950e-01
5.39701641e-01 -5.65208256e-01 -6.57640874e-01 2.76646435e-01
4.74707276e-01 -2.79028952e-01 -8.51523697e-01 -5.81333898e-02
5.47646523e-01 -7.43439674e-01 2.47298285e-01 -5.89859068e-01
3.41957510e-01 -7.27762163e-01 3.36839318e-01 -1.00400984e+00
8.28043967e-02 -9.15038884e-01 1.29456833e-01 1.70813072e+00
6.61094844e-01 -6.13862038e-01 3.84777904e-01 6.18188322e-01
-8.63718167e-02 -2.31093496e-01 -4.97637182e-01 -8.94658387e-01
-3.83126587e-01 -5.75801075e-01 7.22519338e-01 7.81010509e-01
4.30729240e-01 2.52735049e-01 -1.67105630e-01 2.43998334e-01
4.15951073e-01 2.23107234e-01 9.77139294e-01 -1.31254017e+00
-5.59967637e-01 -1.72161683e-01 -2.18301073e-01 -5.79140782e-01
-6.22998886e-02 -7.41445601e-01 1.93261087e-01 -1.53489995e+00
3.21667232e-02 -4.57830459e-01 -9.56127362e-04 9.47099984e-01
-3.19228210e-02 -2.30813384e-01 1.05504813e-02 3.82375829e-02
-2.90531993e-01 -1.48906663e-01 3.73804778e-01 -2.00875700e-01
-4.74397779e-01 1.68374345e-01 -6.26416266e-01 5.71061492e-01
7.94323206e-01 -9.61049497e-01 -1.87732950e-01 -2.74841845e-01
5.65754771e-01 -2.17814535e-01 6.69430047e-02 -7.75299728e-01
2.11473882e-01 -3.64586085e-01 -1.17215207e-02 -5.96057296e-01
-3.58580530e-01 -9.03863072e-01 1.79503128e-01 5.90577185e-01
-2.14218035e-01 -6.78757951e-02 2.29478046e-01 1.75612837e-01
-2.30056897e-01 -1.12692654e+00 4.76760358e-01 -3.06772068e-02
-2.76696920e-01 -1.96807772e-01 -7.57350504e-01 -2.86964327e-01
1.03473735e+00 -2.51688510e-01 -4.55606669e-01 2.43544772e-01
-4.30561453e-01 -2.23104581e-02 6.28235459e-01 3.33883643e-01
1.41308397e-01 -5.10372400e-01 -4.43995327e-01 4.47959810e-01
2.66888648e-01 6.33085221e-02 -2.63008535e-01 6.82441294e-01
-6.59042597e-01 5.47343969e-01 -1.52292043e-01 -3.55115414e-01
-1.37811351e+00 6.49771214e-01 -1.81986317e-01 -6.96215987e-01
-3.54805261e-01 1.24997884e-01 -4.52156246e-01 -4.39710319e-01
5.93224429e-02 -8.32782328e-01 -2.81369925e-01 2.52840102e-01
7.94925332e-01 1.38573855e-01 7.02070475e-01 -1.43680125e-01
-5.01594901e-01 3.64205331e-01 -3.65568697e-01 -7.68180415e-02
1.53504932e+00 3.33181381e-01 -4.55693871e-01 2.26994157e-01
5.70860445e-01 4.59508002e-01 -6.32570326e-01 -6.80166259e-02
7.29469657e-01 -2.71075100e-01 -4.61239994e-01 -7.21424878e-01
-5.76108813e-01 9.77219403e-01 5.49546145e-02 4.43023145e-01
1.02078068e+00 -3.02620709e-01 5.66379726e-01 5.69779873e-01
5.12967706e-01 -1.18471718e+00 -4.12537456e-01 3.47506195e-01
4.54703182e-01 -6.65446341e-01 7.80031085e-02 -6.39870942e-01
-2.56614894e-01 1.40118766e+00 4.55323786e-01 3.92322719e-01
2.11241528e-01 6.40055954e-01 -9.48981047e-02 1.83744922e-01
-7.04035640e-01 1.87761206e-02 -1.45837843e-01 5.00631034e-01
5.91393054e-01 -2.08818480e-01 -8.69197488e-01 9.16994750e-01
-1.64705336e-01 4.34131652e-01 8.50592315e-01 1.37281442e+00
-4.12393034e-01 -2.07649851e+00 -4.78543013e-01 3.10529083e-01
-7.25519121e-01 -1.91197962e-01 -6.12692595e-01 5.81994891e-01
3.18296134e-01 1.15740931e+00 -1.11483321e-01 -4.12671864e-01
5.31399429e-01 4.31712598e-01 3.69983077e-01 -1.05173528e+00
-1.05270815e+00 2.61495054e-01 6.09939516e-01 -5.07295072e-01
-1.39137581e-01 -7.51738071e-01 -1.20439136e+00 1.46808937e-01
-5.51492810e-01 5.17580390e-01 8.94820988e-01 1.13652635e+00
5.31132817e-01 4.86377448e-01 5.35158873e-01 -2.82608569e-01
-4.75820065e-01 -9.45710897e-01 -3.44384350e-02 2.83708900e-01
8.41049924e-02 -2.54633069e-01 9.63747427e-02 6.09716892e-01] | [7.952281951904297, 7.827744960784912] |
054060d2-de39-4dee-99f9-6a34f95ef503 | pizza-a-new-benchmark-for-complex-end-to-end | 2212.00265 | null | https://arxiv.org/abs/2212.00265v1 | https://arxiv.org/pdf/2212.00265v1.pdf | PIZZA: A new benchmark for complex end-to-end task-oriented parsing | Much recent work in task-oriented parsing has focused on finding a middle ground between flat slots and intents, which are inexpressive but easy to annotate, and powerful representations such as the lambda calculus, which are expressive but costly to annotate. This paper continues the exploration of task-oriented parsing by introducing a new dataset for parsing pizza and drink orders, whose semantics cannot be captured by flat slots and intents. We perform an extensive evaluation of deep-learning techniques for task-oriented parsing on this dataset, including different flavors of seq2seq systems and RNNGs. The dataset comes in two main versions, one in a recently introduced utterance-level hierarchical notation that we call TOP, and one whose targets are executable representations (EXR). We demonstrate empirically that training the parser to directly generate EXR notation not only solves the problem of entity resolution in one fell swoop and overcomes a number of expressive limitations of TOP notation, but also results in significantly greater parsing accuracy. | ['Khan Haidar', 'Weiqi Sun', 'Saarthak Khanna', 'Sandesh Swamy', 'Melanie Rubino', 'Nicolas Guenon des Mesnards', 'Konstantine Arkoudas'] | 2022-12-01 | null | null | null | null | ['entity-resolution'] | ['natural-language-processing'] | [ 4.16422486e-01 7.52881289e-01 -6.40662089e-02 -6.99069619e-01
-1.06275022e+00 -8.45128238e-01 4.03243870e-01 1.14393560e-02
-1.47695810e-01 8.58506083e-01 6.20759130e-01 -8.52224886e-01
7.72232236e-03 -8.31398427e-01 -6.92809880e-01 -8.98247212e-02
-2.25957483e-01 7.39721894e-01 9.54093933e-02 -5.12970984e-01
-7.23580420e-02 5.60977086e-02 -1.35363817e+00 6.18545771e-01
5.97300053e-01 8.38267803e-01 8.76650885e-02 1.02977693e+00
-7.52360761e-01 1.49935901e+00 -8.42081785e-01 -8.74903500e-01
5.33809103e-02 -2.50647813e-01 -1.42669892e+00 -3.95471424e-01
4.32518184e-01 -2.23592222e-01 1.45411626e-01 7.64723778e-01
1.89432830e-01 4.00970392e-02 9.13749263e-02 -9.98593688e-01
-6.03817403e-01 1.39075160e+00 2.85654068e-02 1.61848545e-01
4.42042470e-01 -2.54660975e-02 1.58458221e+00 -1.47436380e-01
1.02989650e+00 1.42811477e+00 9.07037616e-01 9.63429272e-01
-1.25464618e+00 -4.57828760e-01 1.31217256e-01 -4.93866503e-01
-8.13118041e-01 -4.05910164e-01 3.80366445e-01 -3.46485406e-01
1.41797912e+00 4.27196592e-01 5.54408543e-02 1.18076432e+00
-9.59123205e-03 8.72590363e-01 8.88127506e-01 -5.38435102e-01
5.27689978e-02 -3.33425105e-01 7.08886206e-01 8.69269311e-01
8.66887718e-02 -3.59937064e-02 -3.52514535e-01 -1.54265389e-01
6.79512680e-01 -4.78903472e-01 8.10039341e-02 -1.23872899e-01
-1.14336920e+00 1.09469306e+00 6.76831231e-02 3.04326445e-01
3.48724648e-02 3.67506474e-01 9.42386568e-01 3.28577578e-01
2.24159390e-01 8.39078426e-01 -1.05992019e+00 -4.33157414e-01
-5.18187225e-01 4.45431411e-01 1.33459866e+00 1.18596137e+00
4.75291163e-01 1.15760220e-02 -2.06234038e-01 8.65334272e-01
-1.70874712e-03 3.04474812e-02 5.88358283e-01 -1.39254451e+00
9.31679487e-01 4.91143227e-01 2.76261687e-01 -2.73612380e-01
-8.13536227e-01 -4.85515781e-02 -3.14094633e-01 -1.78428814e-01
8.51006746e-01 -5.61010718e-01 -7.19895840e-01 2.09982657e+00
1.41303912e-01 -3.29784662e-01 5.53840816e-01 5.37072182e-01
9.32077050e-01 6.29739285e-01 7.36389756e-01 1.30403847e-01
1.86809075e+00 -9.59071636e-01 -5.61087012e-01 -4.01332200e-01
1.33576345e+00 -4.56452459e-01 1.17759430e+00 1.90272644e-01
-1.44583714e+00 -3.09797764e-01 -8.30606163e-01 -8.05319488e-01
-6.86217070e-01 -3.79340611e-02 1.36435390e+00 6.69871211e-01
-1.05937696e+00 7.39674211e-01 -8.64948511e-01 -2.96306700e-01
2.13804856e-01 2.11490408e-01 -4.07165825e-01 1.96332812e-01
-1.38966763e+00 8.35785687e-01 7.76019335e-01 -9.81653258e-02
-2.44456068e-01 -8.11672330e-01 -1.28559792e+00 1.84092849e-01
5.86525738e-01 -6.11907721e-01 2.08205962e+00 -8.51242006e-01
-1.64155042e+00 1.03073823e+00 -2.83462591e-02 -6.28250837e-01
3.47065806e-01 -5.37115037e-01 -3.18319172e-01 -2.12827981e-01
4.09356356e-01 6.84984088e-01 9.07501802e-02 -6.93669498e-01
-8.54713619e-01 -3.65046978e-01 6.80858076e-01 3.41620110e-02
2.29191393e-01 5.75891137e-01 -1.97827682e-01 -5.10331571e-01
8.22439417e-02 -7.78099239e-01 -3.69134396e-01 -5.26222944e-01
-7.23070025e-01 -6.22923374e-01 1.59362212e-01 -8.27582419e-01
1.12033582e+00 -1.92837334e+00 1.59532800e-01 -3.33083212e-01
1.96692750e-01 7.58851469e-02 -1.23271264e-01 3.69116545e-01
-2.26417556e-01 6.50135934e-01 -2.70413607e-01 -2.61102140e-01
4.59793627e-01 7.30254471e-01 -5.11726558e-01 -3.77646357e-01
5.40947139e-01 1.07599974e+00 -9.98906195e-01 -5.03198504e-01
-2.45232120e-01 5.72369397e-02 -6.65037870e-01 2.50566900e-01
-7.79553294e-01 3.85683589e-02 -5.57432830e-01 6.21335745e-01
4.15401936e-01 -2.54660159e-01 7.96571016e-01 1.47463903e-01
-3.41712087e-01 1.21047473e+00 -9.15553272e-01 1.90112531e+00
-6.15752101e-01 3.46119612e-01 3.40544373e-01 -8.48822057e-01
6.45000458e-01 3.85991216e-01 4.72920127e-02 -6.74962997e-01
-2.18603238e-02 3.12109798e-01 1.27249926e-01 -4.93353844e-01
9.37573791e-01 -2.18613490e-01 -8.74478042e-01 4.71842587e-01
2.62739211e-01 -7.89265931e-02 6.34498060e-01 2.68753201e-01
1.40555787e+00 6.33414567e-01 5.65719068e-01 -2.97997624e-01
2.02352062e-01 3.28594029e-01 8.08532059e-01 9.46166217e-01
1.10361099e-01 3.00699323e-01 1.09509802e+00 -8.22157145e-01
-1.14614916e+00 -9.71501112e-01 -1.37883797e-01 1.81632447e+00
-3.83362949e-01 -7.32082546e-01 -8.55727613e-01 -9.24908042e-01
-2.91693449e-01 9.41705883e-01 -4.52612728e-01 4.27993894e-01
-1.28213787e+00 -7.21506834e-01 1.07426322e+00 1.00385368e+00
2.97996968e-01 -1.30966628e+00 -8.52494180e-01 4.91388589e-01
-3.05809468e-01 -1.31600034e+00 3.65144648e-02 9.42069232e-01
-9.15638566e-01 -1.09499192e+00 -1.25249416e-01 -8.70220006e-01
1.50287911e-01 -5.61703086e-01 1.88291478e+00 2.87552588e-02
-1.18275419e-01 1.32440841e-02 -4.93231475e-01 -5.19754469e-01
-7.56294489e-01 4.44121897e-01 -5.76374471e-01 -8.39015245e-01
6.07098937e-01 -3.68394285e-01 2.62734503e-01 -1.59533039e-01
-7.51379907e-01 2.76917547e-01 4.27325755e-01 8.02633703e-01
2.89694250e-01 -3.74788105e-01 4.27271426e-01 -1.84970140e+00
6.05782568e-01 -4.30928230e-01 -6.81417823e-01 2.92070210e-01
1.16597347e-01 5.78699112e-01 9.57637548e-01 9.24869850e-02
-1.22175336e+00 9.13140252e-02 -5.82700849e-01 3.02870661e-01
-3.44565630e-01 4.47365046e-01 -2.68078089e-01 5.86956799e-01
5.59194386e-01 -2.06224486e-01 -3.22870940e-01 -6.83901489e-01
7.35303462e-01 3.07323426e-01 7.86827028e-01 -1.25828254e+00
4.78716612e-01 -1.71699911e-01 -2.32936046e-03 -4.96725827e-01
-1.15212858e+00 -1.52199954e-01 -6.25169754e-01 6.26037717e-01
1.05801427e+00 -8.02062035e-01 -6.66683555e-01 7.08555430e-02
-1.46021557e+00 -7.71202207e-01 -5.70623040e-01 -5.26238792e-02
-7.65730560e-01 2.60193199e-01 -1.17256200e+00 -4.86470312e-01
-4.24302310e-01 -9.31090415e-01 1.23297787e+00 -2.61407830e-02
-5.78054190e-01 -8.81339371e-01 -1.16107643e-01 3.35638702e-01
3.30539048e-01 3.18462491e-01 1.65481675e+00 -1.09986627e+00
-6.16184890e-01 1.98621646e-01 -3.03119242e-01 7.38775805e-02
-3.64154905e-01 -2.59368926e-01 -7.28380263e-01 3.88274372e-01
-2.71675974e-01 -6.30251825e-01 5.15685558e-01 3.76690701e-02
1.07837081e+00 -3.83346945e-01 -1.29180744e-01 6.33354366e-01
1.23383796e+00 2.32536301e-01 6.98224604e-01 5.05775273e-01
5.46489596e-01 7.64165878e-01 3.68699551e-01 1.21680079e-02
6.59969687e-01 5.47168136e-01 2.33348906e-01 1.24283738e-01
-2.90217511e-02 -4.15624499e-01 2.67175227e-01 6.45869672e-01
-5.83286285e-02 -1.57245800e-01 -9.71985340e-01 4.08884197e-01
-1.77526402e+00 -7.76749611e-01 -2.45685637e-01 1.64788139e+00
1.12635958e+00 1.17986776e-01 7.29560256e-02 -1.97194234e-01
5.22268951e-01 1.94621310e-01 -2.26861924e-01 -1.27420855e+00
2.77083665e-02 7.27093399e-01 3.61344576e-01 5.54022491e-01
-1.50355053e+00 1.35810781e+00 6.94851637e+00 1.88343212e-01
-6.67636693e-01 8.38200971e-02 4.33520347e-01 1.89590827e-01
-2.46012926e-01 -3.70263830e-02 -1.06171966e+00 2.51979142e-01
1.45119786e+00 1.91087171e-01 3.75789374e-01 1.30482662e+00
-4.05094504e-01 3.31180811e-01 -1.62560761e+00 4.35791254e-01
-3.63375813e-01 -1.37692785e+00 -1.15342990e-01 -2.15697721e-01
1.68167531e-01 1.81838363e-01 -3.54466707e-01 1.04198706e+00
1.08325469e+00 -1.04480028e+00 8.90996933e-01 -6.88006431e-02
5.88472903e-01 -3.46444428e-01 7.10321307e-01 2.96440005e-01
-1.04497290e+00 -3.46827835e-01 -4.85212356e-01 -2.52496392e-01
3.09414655e-01 8.49132612e-02 -7.47468472e-01 5.03710210e-01
5.89692056e-01 3.33371162e-01 -3.63567710e-01 3.67784172e-01
-5.12629092e-01 5.04145563e-01 -1.77395299e-01 1.31056635e-02
5.31707942e-01 -1.17426567e-01 3.69248569e-01 1.68681419e+00
4.17282358e-02 4.39603835e-01 2.18855083e-01 8.81109059e-01
-4.47393179e-01 7.51148490e-03 -5.76749325e-01 -2.11635500e-01
3.30836356e-01 1.17173302e+00 -5.61271191e-01 -6.24928474e-01
-5.50527632e-01 6.52956009e-01 7.85899937e-01 1.16435789e-01
-7.50094235e-01 -5.37988603e-01 7.53553808e-01 -1.45415202e-01
4.19350296e-01 -1.68196619e-01 -2.10790753e-01 -1.28769922e+00
2.56677754e-02 -1.05064631e+00 7.75733471e-01 -8.70612204e-01
-1.25251842e+00 7.17911422e-01 8.28568935e-02 -3.75486404e-01
-8.21812272e-01 -1.13041651e+00 -4.21518773e-01 8.55913758e-01
-1.50888526e+00 -1.10185659e+00 9.04725716e-02 1.54657915e-01
7.15271056e-01 3.03613007e-01 1.35873008e+00 2.45534167e-01
-6.14666939e-01 4.30561095e-01 -2.00799480e-01 6.87160254e-01
1.90597013e-01 -1.77813029e+00 9.95396674e-01 6.66093051e-01
7.10344836e-02 8.39013517e-01 5.76361537e-01 -4.29669052e-01
-1.44672155e+00 -1.11362135e+00 1.41048729e+00 -8.31804454e-01
8.55773926e-01 -7.31467426e-01 -8.28862250e-01 1.52581859e+00
2.76700151e-03 -1.15201436e-01 5.80275118e-01 6.70050323e-01
-8.08861017e-01 2.71473169e-01 -8.97129953e-01 4.79264885e-01
1.30222654e+00 -5.20738482e-01 -9.90273178e-01 4.58022326e-01
1.16621327e+00 -8.74141753e-01 -9.87191319e-01 1.64038941e-01
4.64478523e-01 -8.19786608e-01 8.68318319e-01 -1.30657685e+00
6.98505402e-01 1.82171568e-01 -4.24723834e-01 -1.02074265e+00
-1.31021664e-01 -7.57340133e-01 -3.47226746e-02 1.44288099e+00
7.49267101e-01 -5.16992450e-01 9.14468825e-01 1.09486914e+00
-5.77427924e-01 -4.35796350e-01 -6.16140306e-01 -6.48725688e-01
5.16247153e-01 -6.39037013e-01 8.87690783e-01 7.75514364e-01
2.49294594e-01 6.95930541e-01 2.40828972e-02 -6.06485941e-02
2.67123520e-01 2.82116622e-01 6.54677212e-01 -1.37101352e+00
-4.47921664e-01 -2.58108407e-01 -1.93725318e-01 -1.08795357e+00
4.31127608e-01 -1.19357002e+00 2.09914878e-01 -1.48731959e+00
-6.10304289e-02 -7.42982328e-01 2.42625736e-03 8.24241996e-01
1.13453567e-01 -9.83909294e-02 3.24339807e-01 -7.06406310e-02
-8.42882752e-01 -2.92592645e-01 7.45039999e-01 1.37604415e-01
-9.07215700e-02 -2.86029607e-01 -1.07282245e+00 9.95051324e-01
7.27616310e-01 -4.49125856e-01 -2.07799867e-01 -9.53826725e-01
4.63038921e-01 3.81951630e-01 7.39597008e-02 -6.97032452e-01
-1.23729274e-01 4.37323153e-02 1.87504590e-02 -2.43683457e-01
1.70299008e-01 -3.99732172e-01 -9.58259851e-02 1.05100304e-01
-8.42275262e-01 2.32654259e-01 2.55784571e-01 1.70476466e-01
-8.47495422e-02 -4.97099638e-01 4.47661459e-01 -7.54436672e-01
-1.03514850e+00 -2.18306512e-01 -2.54345328e-01 7.21566021e-01
5.81913114e-01 1.14683107e-01 -7.29388237e-01 -2.29681842e-02
-8.93991232e-01 9.93445888e-02 1.27952889e-01 3.60800594e-01
-1.55995622e-01 -7.57057548e-01 -5.35233617e-01 1.90860853e-02
-4.17927653e-02 8.99508074e-02 -1.65022507e-01 3.20772827e-01
-7.55937576e-01 7.24214911e-01 -2.03400403e-01 -2.00768158e-01
-1.03831899e+00 4.00347173e-01 2.05477297e-01 -8.65601361e-01
-8.67687583e-01 9.64239836e-01 2.27252692e-01 -1.04592299e+00
1.75321981e-01 -8.64942133e-01 -1.85437322e-01 -1.16406113e-01
5.40515542e-01 5.39230444e-02 7.38439411e-02 -5.43906271e-01
-2.89699942e-01 1.69210091e-01 -2.23073125e-01 1.51121333e-01
1.35501313e+00 2.92901635e-01 -2.33300358e-01 2.11968482e-01
9.74390388e-01 -9.42269489e-02 -8.42123926e-01 -7.52481923e-04
6.95482612e-01 5.60500510e-02 -5.46223104e-01 -1.22465014e+00
-5.41500807e-01 9.46658134e-01 -1.43468916e-01 5.54280639e-01
6.67041659e-01 2.21380979e-01 1.10146260e+00 7.86531627e-01
4.76118535e-01 -9.43012714e-01 -3.46502751e-01 9.55006480e-01
4.69965547e-01 -8.95725787e-01 -4.87968415e-01 -5.64662397e-01
-6.65417612e-01 1.16562414e+00 6.63593471e-01 6.88370736e-03
6.60163611e-02 7.50348032e-01 7.45600536e-02 -2.95122653e-01
-8.89337718e-01 -2.30331868e-01 -4.02741343e-01 7.54094601e-01
9.50717628e-01 4.09686536e-01 -2.53924727e-01 1.26444757e+00
-7.78040409e-01 -1.04820684e-01 5.71474373e-01 1.12043679e+00
-3.08554590e-01 -1.49506426e+00 -5.71983419e-02 3.63128483e-01
-1.18466759e+00 -4.69599932e-01 -3.85001659e-01 1.35629380e+00
1.35033280e-01 7.87341058e-01 3.00937533e-01 1.31460890e-01
4.82639253e-01 7.61083901e-01 4.50271577e-01 -1.00010550e+00
-8.39403152e-01 -4.08300400e-01 1.02399564e+00 -7.98009932e-01
-1.83138445e-01 -5.74429691e-01 -1.57085609e+00 -1.54994830e-01
1.73061248e-02 3.08360815e-01 6.88536525e-01 8.04983079e-01
3.04905206e-01 6.37488127e-01 -8.64212364e-02 -5.05831957e-01
-8.78654003e-01 -7.24311888e-01 -5.14451742e-01 4.82558906e-01
4.23049890e-02 -2.90619314e-01 1.63116343e-02 1.16236009e-01] | [10.4133939743042, 9.321539878845215] |
a948681a-b9fd-4f38-a260-a5bf1e928a50 | knowledge-enhanced-iterative-instruction | 2209.03005 | null | https://arxiv.org/abs/2209.03005v1 | https://arxiv.org/pdf/2209.03005v1.pdf | Knowledge-enhanced Iterative Instruction Generation and Reasoning for Knowledge Base Question Answering | Multi-hop Knowledge Base Question Answering(KBQA) aims to find the answer entity in a knowledge base which is several hops from the topic entity mentioned in the question. Existing Retrieval-based approaches first generate instructions from the question and then use them to guide the multi-hop reasoning on the knowledge graph. As the instructions are fixed during the whole reasoning procedure and the knowledge graph is not considered in instruction generation, the model cannot revise its mistake once it predicts an intermediate entity incorrectly. To handle this, we propose KBIGER(Knowledge Base Iterative Instruction GEnerating and Reasoning), a novel and efficient approach to generate the instructions dynamically with the help of reasoning graph. Instead of generating all the instructions before reasoning, we take the (k-1)-th reasoning graph into consideration to build the k-th instruction. In this way, the model could check the prediction from the graph and generate new instructions to revise the incorrect prediction of intermediate entities. We do experiments on two multi-hop KBQA benchmarks and outperform the existing approaches, becoming the new-state-of-the-art. Further experiments show our method does detect the incorrect prediction of intermediate entities and has the ability to revise such errors. | ['Dongyan Zhao', 'Chen Zhang', 'Quzhe Huang', 'Haowei Du'] | 2022-09-07 | null | null | null | null | ['knowledge-base-question-answering'] | ['natural-language-processing'] | [-1.09303556e-01 7.64311850e-01 -9.06296447e-02 -9.47616100e-02
-1.01851809e+00 -5.80003798e-01 1.66256353e-01 7.07186460e-01
-2.62278140e-01 8.84959221e-01 1.34002911e-02 -6.22150898e-01
-3.67541194e-01 -1.25175273e+00 -1.07015920e+00 -1.10742420e-01
2.86697954e-01 9.49622929e-01 1.36589468e+00 -6.53604329e-01
6.21020675e-01 6.20885938e-02 -1.50893319e+00 5.92585206e-01
1.39439058e+00 7.25393891e-01 1.05747566e-01 7.61786759e-01
-5.06030977e-01 1.17829502e+00 -7.57001936e-01 -7.21887410e-01
-1.50229141e-01 -5.89596868e-01 -1.44700408e+00 -6.88069642e-01
5.06608248e-01 -5.54380357e-01 -1.45020619e-01 9.99114096e-01
2.01983243e-01 3.08432609e-01 3.84464115e-01 -1.20604897e+00
-9.37814653e-01 9.22003508e-01 -7.97600076e-02 4.57514971e-01
7.68064439e-01 -2.47552618e-01 1.01280522e+00 -7.76472569e-01
7.57852256e-01 1.08211422e+00 2.94415534e-01 4.66065526e-01
-3.62291038e-01 -2.56708026e-01 3.27246279e-01 9.77876008e-01
-1.50100648e+00 -1.45730913e-01 4.87689406e-01 -1.16617128e-01
1.09266353e+00 3.95064443e-01 4.45397317e-01 4.29483913e-02
1.57429188e-01 6.41151786e-01 5.82154691e-01 -6.46550000e-01
2.11884454e-01 1.13378607e-01 6.85113728e-01 1.29301846e+00
3.23130965e-01 -4.62608814e-01 -6.07237816e-01 -4.01943743e-01
8.68563950e-02 -4.24528331e-01 -3.91841441e-01 -6.63597956e-02
-1.05866563e+00 6.46042407e-01 7.14774430e-01 1.27587974e-01
-5.89589775e-01 1.16923116e-01 1.16919175e-01 1.65488869e-01
-2.57151663e-01 6.89126909e-01 -5.81660330e-01 -7.26884305e-02
-7.58847475e-01 3.09606671e-01 1.14252949e+00 8.85255277e-01
1.21681499e+00 -5.56223631e-01 -6.31719768e-01 3.32901031e-01
2.18714178e-01 2.98468590e-01 4.97375876e-01 -7.76896536e-01
6.13702118e-01 1.18293190e+00 2.30941221e-01 -1.07365394e+00
-1.21665657e-01 -1.71913579e-01 -7.94797316e-02 -4.59853619e-01
4.31371808e-01 5.32570817e-02 -8.86577010e-01 1.27569854e+00
7.98263371e-01 4.55292225e-01 3.92634988e-01 8.12254488e-01
1.10182261e+00 7.54604280e-01 -3.29202823e-02 -1.23051234e-01
1.51360965e+00 -1.26801336e+00 -6.11200571e-01 -1.43381387e-01
9.88630176e-01 -6.70508564e-01 7.02926457e-01 2.73135543e-01
-1.00784492e+00 -5.87657452e-01 -9.80503500e-01 -1.90742865e-01
-6.16620481e-01 -5.39380349e-02 2.63672709e-01 1.36231378e-01
-1.09274685e+00 2.87811846e-01 -3.98534477e-01 -1.21576294e-01
-2.42390916e-01 1.40705720e-01 7.63124004e-02 -5.06025195e-01
-1.64311206e+00 7.98277199e-01 7.86467612e-01 1.59665257e-01
-5.92868924e-01 -7.65340865e-01 -7.62033820e-01 1.19678169e-01
9.98546004e-01 -9.89407957e-01 1.36037529e+00 -5.53448558e-01
-1.22693908e+00 2.92980492e-01 -4.79194790e-01 -4.29810762e-01
5.55297472e-02 -2.20438138e-01 -4.50775325e-01 3.51995140e-01
2.29713157e-01 6.18125618e-01 6.14759803e-01 -1.12955248e+00
-9.02892947e-01 -2.47666404e-01 6.86617792e-01 4.21666265e-01
1.22272313e-01 -4.81907129e-01 -9.72152531e-01 1.03893511e-01
2.50017613e-01 -9.20446634e-01 8.35617706e-02 -6.25586808e-01
-5.39528847e-01 -6.59833014e-01 6.81390405e-01 -8.75774801e-01
1.69715476e+00 -1.77310812e+00 1.52426004e-01 3.07579488e-01
2.08066702e-01 4.42789108e-01 2.42122114e-02 3.90040457e-01
3.40909511e-01 2.04453528e-01 1.22187212e-01 5.48371851e-01
-9.39091891e-02 2.15560600e-01 -5.19965291e-01 -2.89741516e-01
2.01829188e-02 1.11164081e+00 -1.19250584e+00 -6.65729344e-01
-3.03182662e-01 -5.86232357e-02 -8.06587458e-01 1.94088086e-01
-9.03712273e-01 5.00053838e-02 -9.07435477e-01 5.55499792e-01
3.58486116e-01 -4.56610858e-01 8.31149220e-02 -4.61139649e-01
2.39368945e-01 4.83026803e-01 -1.16131616e+00 1.34705544e+00
-3.66536617e-01 1.61206648e-01 -5.39223194e-01 -5.96400261e-01
7.69872248e-01 1.26444912e-02 -9.06326622e-02 -9.99425650e-01
-4.90594983e-01 4.77137148e-01 6.45020083e-02 -7.14724064e-01
7.84690320e-01 3.09395999e-01 -1.15449153e-01 3.09298605e-01
-2.77484328e-01 -7.32791349e-02 4.03771460e-01 7.12166846e-01
1.44498062e+00 -5.77646010e-02 2.01460063e-01 1.91953197e-01
1.14681780e+00 5.04371941e-01 2.71272689e-01 9.93346393e-01
8.61988440e-02 -8.00760537e-02 3.46177965e-01 -3.20636988e-01
-3.27589095e-01 -8.56446147e-01 4.19271916e-01 1.13436329e+00
5.97533584e-01 -8.39868546e-01 -8.50868642e-01 -1.25492084e+00
5.81523664e-02 1.05723083e+00 -5.03021181e-01 -5.18643081e-01
-7.79423296e-01 5.62990643e-03 3.23849916e-01 2.98930585e-01
6.64084792e-01 -1.18555903e+00 -4.58137542e-01 3.22573662e-01
-5.99623680e-01 -9.65580761e-01 -3.71199191e-01 -2.22235754e-01
-5.13896406e-01 -1.63087702e+00 -2.17930391e-01 -6.92984641e-01
9.91699815e-01 9.16218981e-02 1.28430402e+00 9.71793413e-01
2.87694126e-01 6.56289220e-01 -6.35087729e-01 -2.31400341e-01
-6.80498660e-01 3.37508470e-01 -4.94768292e-01 -4.03493911e-01
5.38887680e-01 2.13776976e-01 -6.56541646e-01 3.77611905e-01
-9.94919896e-01 -1.05602600e-01 4.90825474e-01 5.13978064e-01
8.23115230e-01 3.33693177e-01 8.13179493e-01 -1.09457290e+00
8.12781632e-01 -6.19449675e-01 -7.38731682e-01 9.70990717e-01
-6.19863153e-01 6.64912522e-01 8.60401690e-01 4.27020080e-02
-1.05076730e+00 -3.35930020e-01 -1.92942843e-01 -1.74981534e-01
2.00495243e-01 9.22849953e-01 1.70544639e-01 2.28340700e-02
7.72956073e-01 3.07469159e-01 -6.67408168e-01 -1.54112116e-01
4.79933739e-01 3.47239971e-01 4.79349196e-01 -7.54181087e-01
7.06126451e-01 -2.77758949e-02 -8.54177997e-02 -1.80198953e-01
-1.00755799e+00 -3.84963155e-01 -4.86016780e-01 -2.12044641e-01
7.76928484e-01 -6.26274586e-01 -7.80455530e-01 1.84193347e-02
-1.32281601e+00 -2.68654525e-01 -2.16973186e-01 8.54415745e-02
-1.65813223e-01 4.20109212e-01 -4.05977577e-01 -5.10664165e-01
-4.46487486e-01 -1.25518417e+00 1.00023520e+00 4.43187505e-01
-9.83310491e-02 -6.66673541e-01 1.74666092e-01 5.54729700e-01
1.00213826e-01 -3.12729955e-01 1.54572821e+00 -9.04758334e-01
-1.19343543e+00 -1.59998149e-01 -1.03245266e-02 -1.16110340e-01
-1.08692609e-01 5.44035668e-03 -3.80618095e-01 1.03677943e-01
-5.26531160e-01 -1.38430804e-01 6.10321522e-01 -3.26571792e-01
1.01143372e+00 -5.23242891e-01 -3.91739547e-01 -8.97637755e-02
1.29044318e+00 3.34882021e-01 8.27784836e-01 4.44017798e-01
6.60012960e-01 4.62611556e-01 9.86742437e-01 -1.19926162e-01
1.00062668e+00 6.76051378e-01 3.87239575e-01 6.37099087e-01
-2.64542192e-01 -5.19246221e-01 3.34471852e-01 1.16527343e+00
2.83018261e-01 -3.70603979e-01 -1.08326197e+00 7.37157822e-01
-1.99169171e+00 -5.32484889e-01 -4.40703541e-01 2.23007870e+00
9.00261045e-01 1.73299953e-01 -4.07642424e-01 -1.24783325e-03
4.90994304e-01 -3.32174301e-01 -6.51659667e-01 -6.49315834e-01
5.03354371e-01 2.04220027e-01 2.10047260e-01 7.68616319e-01
-3.17982197e-01 1.37087965e+00 5.02115059e+00 7.36541808e-01
-7.44685352e-01 -9.62078124e-02 2.03874141e-01 4.09820408e-01
-5.92063069e-01 3.49351406e-01 -1.36108077e+00 2.25094095e-01
1.04111803e+00 -4.07853842e-01 4.91198391e-01 6.83057129e-01
-4.28612024e-01 -6.09028816e-01 -1.10713959e+00 3.68720680e-01
3.01529229e-01 -1.37582576e+00 4.84023541e-01 -5.61603606e-01
6.86068535e-01 -3.76206964e-01 -2.76128471e-01 8.82138610e-01
3.80468935e-01 -4.38200295e-01 5.90259433e-01 8.85884881e-01
1.72896922e-01 -8.91866207e-01 8.16613913e-01 5.84356189e-01
-1.25299644e+00 -4.73090671e-02 -4.17814106e-01 3.87660623e-01
5.80711924e-02 5.14803529e-01 -1.31968474e+00 9.46075916e-01
7.10002363e-01 4.29399312e-02 -9.62609828e-01 9.49467242e-01
-7.58738339e-01 4.55997825e-01 -2.09072784e-01 -2.27764919e-01
2.43335336e-01 1.15424931e-01 4.16823477e-01 6.84055746e-01
5.52784324e-01 5.43867171e-01 9.94966626e-02 5.32934427e-01
-3.17699939e-01 2.90708631e-01 -2.22788408e-01 1.05477519e-01
6.10145986e-01 8.57658029e-01 -5.19981623e-01 -8.48547459e-01
-4.26185459e-01 1.08287549e+00 7.41022706e-01 3.10184836e-01
-8.53501320e-01 -6.19428635e-01 1.27330542e-01 3.64242680e-03
5.94055951e-01 1.65474743e-01 4.49774235e-01 -8.73123467e-01
3.88358504e-01 -9.75664556e-01 8.15802872e-01 -1.29380250e+00
-8.13026547e-01 7.02743471e-01 3.27935703e-02 -6.01010799e-01
-5.68564296e-01 -1.66602373e-01 -4.13310349e-01 7.75630116e-01
-1.93784881e+00 -6.12380326e-01 -4.99517113e-01 6.19875550e-01
4.15072829e-01 3.35125387e-01 6.80462420e-01 1.32374108e-01
-3.97518486e-01 5.71552515e-01 -4.07298237e-01 -8.59101638e-02
8.01096439e-01 -1.24597013e+00 2.02079266e-02 8.30062985e-01
5.25497682e-02 7.67875850e-01 4.66531485e-01 -1.05603850e+00
-1.55385315e+00 -1.06496358e+00 1.12386382e+00 -6.06270671e-01
7.65388548e-01 3.40059608e-01 -1.46983469e+00 6.84698880e-01
4.72283028e-02 -1.98561892e-01 5.46220899e-01 -1.73787057e-01
-2.35894501e-01 -1.54185429e-01 -9.09059942e-01 6.95675850e-01
7.17863679e-01 -5.04880190e-01 -1.01760828e+00 4.01850194e-01
1.02102435e+00 -7.71816850e-01 -8.29411447e-01 5.50555408e-01
1.50926888e-01 -7.94765770e-01 7.90416241e-01 -6.25540376e-01
3.09483021e-01 -7.82058775e-01 2.00474143e-01 -1.22509515e+00
-1.20044209e-01 -3.14125299e-01 -8.12516928e-01 1.08230424e+00
8.63076568e-01 -4.88772154e-01 7.74100900e-01 7.52954483e-01
-3.92895699e-01 -9.55265045e-01 -8.08205724e-01 -2.70576507e-01
-2.49259621e-01 -4.33054306e-02 1.01490414e+00 6.61683738e-01
1.58340767e-01 3.69782299e-01 1.71141207e-01 7.83530951e-01
1.50279298e-01 4.14205253e-01 9.39687669e-01 -8.96107793e-01
-3.47194195e-01 -1.64101198e-02 -1.01154186e-01 -1.37097216e+00
1.95372075e-01 -9.15398121e-01 1.97430626e-01 -1.97008753e+00
-3.49135846e-02 -3.96397442e-01 -2.26381391e-01 5.60375988e-01
-8.92766178e-01 -2.26112112e-01 -1.44996447e-02 1.04457341e-01
-1.21536851e+00 1.81835949e-01 1.45275211e+00 -1.34850189e-01
-2.45197371e-01 -1.99504152e-01 -8.00649345e-01 5.40133297e-01
5.56690514e-01 -4.79296118e-01 -7.52722800e-01 -3.76767278e-01
8.54584098e-01 3.76046389e-01 2.48254672e-01 -9.35205102e-01
9.65454280e-01 2.22141463e-02 2.78243050e-02 -8.14384341e-01
6.77714273e-02 -8.14772904e-01 -1.51661374e-02 5.51998258e-01
-3.41840178e-01 3.20330322e-01 3.47950697e-01 6.26898468e-01
-3.18134964e-01 -5.90088487e-01 1.90745354e-01 -1.73642471e-01
-1.16585350e+00 8.85269791e-02 -1.97095811e-01 2.59640425e-01
9.97761726e-01 2.70929392e-02 -9.47422206e-01 -2.33163252e-01
-6.33715451e-01 6.04972899e-01 1.53229237e-01 2.67430156e-01
7.02172101e-01 -1.03974819e+00 -4.28893119e-01 -1.62481949e-01
3.87148798e-01 2.40992308e-01 3.30122918e-01 6.77109063e-01
-6.72693610e-01 7.10170925e-01 3.01126122e-01 -2.97471344e-01
-1.01697564e+00 7.64504135e-01 4.19062823e-01 -7.56066144e-01
-2.38137722e-01 7.37351120e-01 -1.62030250e-01 -3.46310288e-01
-4.31913175e-02 -5.14981210e-01 -4.51362014e-01 -1.23471282e-01
6.41789377e-01 1.96817666e-01 5.57254493e-01 -3.64244789e-01
-3.69084716e-01 4.92395610e-01 -4.22952890e-01 1.49111286e-01
6.88803196e-01 -6.75295070e-02 -3.70168567e-01 1.73069149e-01
9.03109968e-01 2.86116302e-01 -4.66881663e-01 -3.26584727e-01
1.83302030e-01 -2.74407297e-01 -8.31427425e-02 -1.09323013e+00
-7.83857942e-01 3.45004886e-01 1.27186447e-01 3.68876427e-01
9.65859652e-01 2.40065858e-01 1.09484863e+00 1.07051551e+00
6.00866556e-01 -9.64271307e-01 2.44337916e-01 8.15130651e-01
8.44464660e-01 -8.26911747e-01 -6.51667416e-02 -7.34553635e-01
-3.84748876e-01 1.01955974e+00 1.20126748e+00 3.24294955e-01
4.03275281e-01 -3.12939137e-01 -6.85157999e-02 -5.98463535e-01
-7.66794801e-01 -1.47486970e-01 6.19810045e-01 2.79020905e-01
5.47499992e-02 -2.50735760e-01 -4.12626415e-01 4.40423518e-01
-3.38976175e-01 1.98246807e-01 3.52279335e-01 1.12020969e+00
-7.97794342e-01 -1.05261278e+00 -3.85364026e-01 6.33344114e-01
-2.30928227e-01 -3.77129883e-01 -6.20514214e-01 6.67315662e-01
4.15478833e-02 1.04561341e+00 -1.32416755e-01 -3.07911277e-01
7.05521762e-01 5.21947324e-01 3.84932280e-01 -7.70862401e-01
-7.33801663e-01 -6.90812469e-01 2.94555426e-01 -6.67717934e-01
2.72899307e-02 -8.39326978e-02 -1.86088741e+00 -2.47964397e-01
-6.39927745e-01 8.78200591e-01 3.32883865e-01 1.07391131e+00
7.10295558e-01 7.85133481e-01 2.41370007e-01 2.88183272e-01
-1.95045233e-01 -5.00361145e-01 1.00395717e-01 3.07502121e-01
2.18162537e-01 -5.52672148e-01 -2.11381719e-01 -6.26890361e-02] | [10.625969886779785, 7.956364154815674] |
bdecca1c-89bd-4467-acf0-2a79619971ae | warp-consistency-for-unsupervised-learning-of | 2104.03308 | null | https://arxiv.org/abs/2104.03308v3 | https://arxiv.org/pdf/2104.03308v3.pdf | Warp Consistency for Unsupervised Learning of Dense Correspondences | The key challenge in learning dense correspondences lies in the lack of ground-truth matches for real image pairs. While photometric consistency losses provide unsupervised alternatives, they struggle with large appearance changes, which are ubiquitous in geometric and semantic matching tasks. Moreover, methods relying on synthetic training pairs often suffer from poor generalisation to real data. We propose Warp Consistency, an unsupervised learning objective for dense correspondence regression. Our objective is effective even in settings with large appearance and view-point changes. Given a pair of real images, we first construct an image triplet by applying a randomly sampled warp to one of the original images. We derive and analyze all flow-consistency constraints arising between the triplet. From our observations and empirical results, we design a general unsupervised objective employing two of the derived constraints. We validate our warp consistency loss by training three recent dense correspondence networks for the geometric and semantic matching tasks. Our approach sets a new state-of-the-art on several challenging benchmarks, including MegaDepth, RobotCar and TSS. Code and models are at github.com/PruneTruong/DenseMatching. | ['Luc van Gool', 'Fisher Yu', 'Martin Danelljan', 'Prune Truong'] | 2021-04-07 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Truong_Warp_Consistency_for_Unsupervised_Learning_of_Dense_Correspondences_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Truong_Warp_Consistency_for_Unsupervised_Learning_of_Dense_Correspondences_ICCV_2021_paper.pdf | iccv-2021-1 | ['dense-pixel-correspondence-estimation'] | ['computer-vision'] | [ 1.66367710e-01 -6.85740588e-03 -1.88832924e-01 -4.84533638e-01
-8.25767934e-01 -5.15965939e-01 7.06340313e-01 -6.29123002e-02
-3.06082249e-01 5.27412295e-01 -6.60753623e-02 1.19095191e-01
-6.57952130e-02 -6.44341648e-01 -1.08397150e+00 -5.15338123e-01
1.18833408e-01 8.91004801e-01 2.77499825e-01 -2.84330130e-01
2.49556705e-01 6.48168862e-01 -1.59178579e+00 3.55554596e-02
7.88789272e-01 9.42298174e-01 2.09148973e-01 3.70854765e-01
1.04969025e-01 5.65224230e-01 -5.59753459e-03 -4.58292305e-01
9.18650687e-01 -3.68060768e-01 -1.04839253e+00 1.90358207e-01
1.20644963e+00 -3.88719559e-01 -5.70473373e-01 1.11026073e+00
3.72926056e-01 3.36481392e-01 3.39709282e-01 -1.57243407e+00
-5.74527919e-01 1.34407088e-01 -7.37937331e-01 -6.41919896e-02
4.89000171e-01 2.77857125e-01 1.19400907e+00 -1.00106919e+00
1.10253823e+00 1.26636016e+00 9.18784022e-01 3.51212025e-01
-1.48413599e+00 -5.17460823e-01 -1.59059197e-01 1.59370318e-01
-1.23822224e+00 -7.00234652e-01 8.63206446e-01 -3.75758499e-01
7.65436172e-01 -1.75739825e-02 4.46082562e-01 1.10830176e+00
-4.51972969e-02 5.16818821e-01 9.99659717e-01 -3.83396685e-01
-5.71269020e-02 -8.14908668e-02 -2.19186008e-01 7.39037097e-01
1.08200975e-01 5.61161220e-01 -6.21390045e-01 1.11316806e-02
9.44258630e-01 1.76141108e-03 -2.31627107e-01 -1.20830750e+00
-1.44606650e+00 8.41218531e-01 8.83714855e-01 -3.47528122e-02
-9.04113799e-02 3.76573622e-01 1.17362007e-01 6.30808830e-01
3.52400094e-01 5.21990359e-01 -2.79383421e-01 5.12127094e-02
-8.06005478e-01 4.63723779e-01 7.39180326e-01 1.25497699e+00
1.21069932e+00 -3.35423440e-01 2.53883600e-01 8.18325341e-01
7.95563161e-02 4.29375499e-01 1.08078532e-01 -1.38181174e+00
5.22171795e-01 2.92880356e-01 8.70268866e-02 -1.43250322e+00
-2.10586935e-01 -2.53747553e-01 -6.79967701e-01 4.15822446e-01
7.05722392e-01 4.92716610e-01 -7.72942185e-01 1.71380150e+00
4.12731320e-01 3.35287213e-01 -1.27463266e-01 1.06063998e+00
6.41063631e-01 3.58209670e-01 -3.07161331e-01 1.33394986e-01
7.93241024e-01 -1.27015817e+00 -2.29921877e-01 -3.25933427e-01
5.08675694e-01 -1.02095389e+00 9.04179692e-01 1.36085674e-01
-1.21583009e+00 -5.43961406e-01 -9.27653372e-01 -5.61512291e-01
-2.50796288e-01 -2.76134491e-01 4.69124317e-01 4.17177826e-02
-1.09440780e+00 9.64496791e-01 -8.03445697e-01 -6.57346845e-01
2.82255441e-01 2.92450368e-01 -6.67411685e-01 -4.73509371e-01
-8.29423547e-01 9.93076503e-01 9.81492177e-02 1.34102702e-01
-6.04651332e-01 -9.45884645e-01 -1.03168130e+00 -3.65614325e-01
2.02679887e-01 -9.67020631e-01 1.08928871e+00 -1.11594617e+00
-1.27897501e+00 1.42908037e+00 1.76595956e-01 -3.98858309e-01
1.04210758e+00 -1.81913316e-01 1.29463866e-01 1.46667868e-01
2.09584311e-01 1.04717517e+00 7.48341322e-01 -1.43546772e+00
-3.91128957e-01 -2.49603808e-01 1.99625622e-02 8.84495825e-02
2.10024878e-01 -2.81568050e-01 -4.37495530e-01 -4.11235690e-01
2.45940253e-01 -1.02098393e+00 -2.12853223e-01 6.53789103e-01
-4.21109021e-01 2.07808390e-01 7.75787294e-01 -5.34600914e-01
2.65764147e-01 -2.08890939e+00 3.80583942e-01 3.12234402e-01
2.11419150e-01 -4.82131094e-02 -5.00363171e-01 4.07438755e-01
-1.08888090e-01 -3.92096877e-01 -4.67523664e-01 -7.95166194e-01
9.58290920e-02 3.41879904e-01 -3.20958853e-01 8.98052216e-01
2.69306034e-01 1.04493833e+00 -1.12491941e+00 -4.54167008e-01
4.98530418e-01 4.94884431e-01 -7.22980201e-01 4.36573476e-01
-2.00714156e-01 6.76776826e-01 -9.18404981e-02 4.48018193e-01
7.71586478e-01 -3.19081217e-01 -1.83275521e-01 -4.17220742e-01
-6.18143287e-03 2.95982122e-01 -1.06728423e+00 2.59342957e+00
-4.50152725e-01 6.77139342e-01 -6.02270588e-02 -1.14621103e+00
9.53132689e-01 -2.09411547e-01 6.01953924e-01 -8.13258886e-01
1.68924093e-01 4.66606438e-01 -2.17968553e-01 -3.46968383e-01
5.11151850e-01 -5.21626286e-02 1.93752512e-01 2.78549969e-01
1.88375786e-01 -6.25680089e-01 3.84283438e-02 1.25547484e-01
1.03407133e+00 4.94977206e-01 7.69144669e-02 -3.44583422e-01
3.00028950e-01 6.29004836e-02 5.01276791e-01 5.43986499e-01
-8.64603594e-02 1.17512083e+00 3.33458990e-01 -7.23475218e-01
-1.58871615e+00 -1.21914017e+00 -1.44717261e-01 5.93092561e-01
5.89127362e-01 -1.75130755e-01 -3.47350657e-01 -4.30146962e-01
2.45043010e-01 2.12347001e-01 -5.98799467e-01 -3.23788188e-02
-6.55498087e-01 -1.99647427e-01 2.98302084e-01 3.54995877e-01
5.90367436e-01 -8.73421550e-01 -5.17994106e-01 8.29861984e-02
-2.17360869e-01 -1.47393334e+00 -5.84421575e-01 -1.90451890e-01
-7.35149205e-01 -1.27426517e+00 -5.08584678e-01 -8.62343788e-01
8.08448493e-01 4.49899763e-01 1.50884926e+00 3.50080162e-01
-5.47287643e-01 3.92452627e-01 -1.96578220e-01 2.22375348e-01
-4.43516940e-01 1.71179637e-01 -5.17745726e-02 -1.26948014e-01
2.56731752e-02 -8.97741318e-01 -7.23531485e-01 6.56352699e-01
-8.28719199e-01 2.02151477e-01 3.78198445e-01 9.22838926e-01
7.24763751e-01 -5.95404625e-01 -2.01306827e-02 -7.81300545e-01
3.96923535e-02 -2.44189501e-01 -8.71638358e-01 1.09958954e-01
-4.17560220e-01 2.27420822e-01 3.65541995e-01 -1.99341476e-01
-6.33394957e-01 2.91311949e-01 -4.19429094e-02 -8.93604755e-01
-1.52815590e-02 9.14671551e-03 -2.20827013e-02 -5.15313864e-01
6.64148510e-01 -1.90019891e-01 4.40287620e-01 -3.52223098e-01
5.64497292e-01 -3.53078954e-02 1.00427306e+00 -9.21454132e-01
1.28020024e+00 8.51741493e-01 1.82238474e-01 -6.68945432e-01
-8.52190971e-01 -6.24421120e-01 -9.32455301e-01 3.77504826e-02
6.19722068e-01 -8.21323454e-01 -5.45110345e-01 4.87623215e-01
-1.21159530e+00 -4.89878833e-01 -4.33781981e-01 3.05041641e-01
-1.02162933e+00 5.94698191e-01 -4.72229213e-01 -1.53953105e-01
-1.26362070e-01 -1.19236839e+00 1.28935409e+00 2.95348354e-02
-1.34173140e-01 -1.11130750e+00 3.49667847e-01 1.95740715e-01
3.81249249e-01 6.41265929e-01 5.54529846e-01 -1.42739192e-01
-9.33979332e-01 -1.15739219e-01 -4.56257999e-01 2.90644675e-01
1.09076619e-01 7.47274607e-02 -7.94792831e-01 -4.77708817e-01
-2.53017813e-01 -6.89394414e-01 7.70532906e-01 8.63748416e-02
1.08696628e+00 -2.53310055e-02 -2.77009010e-01 1.19341290e+00
1.54451823e+00 -4.33515817e-01 6.89719915e-01 5.16733646e-01
9.30486798e-01 9.22506392e-01 6.41754031e-01 5.19536845e-02
3.40831727e-01 1.00906837e+00 7.32684433e-01 -3.62843156e-01
-1.67515740e-01 -4.62095410e-01 5.96367940e-02 6.48208439e-01
-3.15327160e-02 1.41029665e-02 -8.72932553e-01 5.89504659e-01
-1.97808671e+00 -8.05149615e-01 -1.20915085e-01 2.37493110e+00
7.73276508e-01 -5.65786473e-02 7.67918825e-02 -1.78580970e-01
6.07170343e-01 2.61279136e-01 -4.27702934e-01 3.57162133e-02
-1.65000409e-01 2.80360878e-01 6.55302584e-01 7.75096714e-01
-1.01384199e+00 9.11572874e-01 5.28710079e+00 3.80458772e-01
-1.01013768e+00 9.77584440e-03 6.04633987e-01 -8.83607119e-02
-4.24749404e-01 2.75525391e-01 -4.19563383e-01 2.17719808e-01
4.00927663e-01 7.16499388e-02 4.62791920e-01 5.53225577e-01
-4.82199527e-02 -6.85453787e-02 -1.76594889e+00 1.21948552e+00
9.42925289e-02 -1.44776607e+00 -4.72117402e-02 1.15376506e-02
9.55832064e-01 3.09467971e-01 1.54745013e-01 -2.18093216e-01
4.92335945e-01 -1.03313482e+00 6.53857887e-01 3.56106341e-01
8.60328853e-01 -4.65549380e-01 4.72383052e-01 1.15973940e-02
-9.47120905e-01 3.71160239e-01 -5.18326283e-01 1.56055585e-01
2.06312910e-01 5.12419522e-01 -3.66710275e-01 7.20982730e-01
6.65017903e-01 1.17419004e+00 -4.21601236e-01 1.08335102e+00
-1.20539919e-01 -2.18831554e-01 -6.02097034e-01 6.48722351e-01
2.93313473e-01 -4.41751987e-01 5.10885775e-01 9.41113293e-01
2.68531233e-01 -2.38300607e-01 1.02468282e-01 1.19554305e+00
-2.57748783e-01 -1.82134807e-01 -8.46576869e-01 3.53782386e-01
5.15806735e-01 1.17884398e+00 -5.84170282e-01 -1.71402562e-02
-3.70744377e-01 1.14658070e+00 6.18472338e-01 2.87368089e-01
-6.83508515e-01 -7.45748505e-02 9.67820585e-01 2.15733185e-01
1.43759221e-01 -3.26298088e-01 -1.91078782e-01 -1.29723418e+00
3.30082059e-01 -8.23964596e-01 1.20188661e-01 -7.63274670e-01
-1.50346220e+00 4.59042937e-01 1.02733001e-01 -1.45870793e+00
-3.12720269e-01 -5.31417966e-01 -6.32373035e-01 7.49906838e-01
-1.73968518e+00 -1.24827397e+00 -7.73864448e-01 5.95798731e-01
4.87938583e-01 1.01281047e-01 5.79034686e-01 3.91072839e-01
-2.36274913e-01 6.39613330e-01 5.46077117e-02 1.47387922e-01
9.87333179e-01 -1.21555936e+00 8.73357773e-01 8.35396171e-01
2.94372052e-01 4.07241970e-01 8.26025784e-01 -1.99215874e-01
-1.45950556e+00 -9.16302621e-01 7.26610363e-01 -5.46840429e-01
7.47586370e-01 -5.54295063e-01 -1.02327144e+00 8.22483599e-01
1.61177069e-01 5.89635134e-01 -7.10907355e-02 -2.03127965e-01
-7.21505582e-01 -1.33933529e-01 -1.15805316e+00 4.14451838e-01
1.54984498e+00 -5.18639982e-01 -4.27946091e-01 4.48976249e-01
6.79041862e-01 -8.39504778e-01 -8.73293161e-01 4.25566345e-01
6.19607568e-01 -1.27304471e+00 1.29656160e+00 -4.69329745e-01
8.04870903e-01 -2.34394774e-01 -1.73888505e-01 -1.42765033e+00
-1.75859518e-02 -8.01563025e-01 2.97572643e-01 1.09823275e+00
2.27646172e-01 -7.30957687e-01 8.05006087e-01 5.52706659e-01
-7.98557550e-02 -4.80790317e-01 -8.89492095e-01 -9.69322622e-01
2.35117614e-01 -1.62573397e-01 4.99643445e-01 1.22278047e+00
-3.41540903e-01 6.37298822e-02 -4.08762217e-01 8.44393969e-02
1.09002745e+00 1.20960787e-01 1.19947553e+00 -1.11103547e+00
-4.30508941e-01 -3.29474241e-01 -7.74972141e-01 -1.16444480e+00
3.47007781e-01 -8.48146796e-01 2.16386527e-01 -1.20785046e+00
1.96564216e-02 -7.81856537e-01 1.68944404e-01 2.81832963e-01
2.46513158e-01 4.37247664e-01 2.39252985e-01 3.95732611e-01
-2.96240628e-01 6.37973785e-01 1.27535379e+00 -2.21714512e-01
4.94957864e-02 -2.75415748e-01 -6.95869848e-02 5.64248860e-01
6.06185555e-01 -4.17251557e-01 -3.86143655e-01 -8.42336178e-01
2.92857975e-01 -8.20027366e-02 7.52486408e-01 -8.94581556e-01
3.25344890e-01 -1.42919734e-01 6.52697086e-02 -3.92146885e-01
3.88593942e-01 -9.40562606e-01 4.00008589e-01 2.64242083e-01
-3.78330320e-01 3.82615447e-01 9.35740322e-02 4.25545871e-01
-2.44038939e-01 -5.87654114e-03 1.07440317e+00 -1.04627341e-01
-7.48829067e-01 6.47175312e-01 6.03331029e-01 4.04049397e-01
7.84464657e-01 -2.73260772e-01 -4.26403522e-01 -2.88432896e-01
-4.64165449e-01 3.07633519e-01 1.17240167e+00 7.07110941e-01
6.29308343e-01 -1.53209460e+00 -7.96777368e-01 3.52112025e-01
4.33396876e-01 6.03361011e-01 -8.50704312e-02 1.01532829e+00
-1.01409256e+00 8.12944993e-02 -5.49477398e-01 -9.03245866e-01
-9.73642468e-01 4.25248593e-01 6.36596799e-01 -5.54247163e-02
-9.26140010e-01 7.87787974e-01 4.30528134e-01 -6.67276084e-01
2.89723307e-01 -3.50331634e-01 5.35604537e-01 -4.03192848e-01
1.14554174e-01 2.71829069e-01 1.94813430e-01 -7.64840066e-01
-3.19608569e-01 9.33258057e-01 -9.39096063e-02 1.03796432e-02
1.33312035e+00 -7.61780813e-02 -2.05421850e-01 1.90689459e-01
1.78711760e+00 -3.93825680e-01 -1.51824343e+00 -5.89404583e-01
-4.19647992e-02 -1.02954233e+00 -2.35195771e-01 -2.35628486e-01
-1.32921934e+00 6.83199942e-01 4.26900506e-01 -1.28934756e-01
8.00412893e-01 6.90874532e-02 9.08127964e-01 3.69405687e-01
5.21146059e-01 -8.36857617e-01 1.86132029e-01 3.75380933e-01
9.92282510e-01 -1.58465004e+00 2.48594452e-02 -5.44184566e-01
-1.52805060e-01 1.00943124e+00 5.93600392e-01 -6.54369175e-01
4.09759849e-01 1.92833692e-01 2.39813216e-02 -2.63506591e-01
-4.27544028e-01 -4.36262302e-02 2.62113452e-01 7.39837527e-01
1.67818531e-01 -2.93529660e-01 -3.65127325e-02 -6.30475819e-01
-4.22278613e-01 -2.57722855e-01 3.77589494e-01 7.89098561e-01
-1.88946314e-02 -1.29188776e+00 -1.67808637e-01 7.90454075e-02
4.11992818e-02 5.32267541e-02 -4.75717634e-01 1.03104115e+00
-1.77831098e-01 4.88798231e-01 3.85408789e-01 -2.32742637e-01
4.72945839e-01 -3.96559656e-01 8.88173163e-01 -3.80648166e-01
-2.95355082e-01 -2.71502405e-01 -5.10992445e-02 -1.08875716e+00
-7.99078107e-01 -6.91982448e-01 -9.81837690e-01 -6.86180294e-01
-1.49532426e-02 -1.25917450e-01 5.04412234e-01 8.06949914e-01
1.95174292e-01 -2.25729961e-02 7.24328995e-01 -1.22071946e+00
-5.74671090e-01 -6.47292197e-01 -3.47525656e-01 9.70276892e-01
5.48943520e-01 -8.74849677e-01 -5.92350304e-01 1.31502107e-01] | [8.49324893951416, -2.207772731781006] |
cb312384-7c51-4594-ba88-b8692998bb28 | s-net-from-answer-extraction-to-answer | 1706.04815 | null | http://arxiv.org/abs/1706.04815v6 | http://arxiv.org/pdf/1706.04815v6.pdf | S-Net: From Answer Extraction to Answer Generation for Machine Reading Comprehension | In this paper, we present a novel approach to machine reading comprehension
for the MS-MARCO dataset. Unlike the SQuAD dataset that aims to answer a
question with exact text spans in a passage, the MS-MARCO dataset defines the
task as answering a question from multiple passages and the words in the answer
are not necessary in the passages. We therefore develop an
extraction-then-synthesis framework to synthesize answers from extraction
results. Specifically, the answer extraction model is first employed to predict
the most important sub-spans from the passage as evidence, and the answer
synthesis model takes the evidence as additional features along with the
question and passage to further elaborate the final answers. We build the
answer extraction model with state-of-the-art neural networks for single
passage reading comprehension, and propose an additional task of passage
ranking to help answer extraction in multiple passages. The answer synthesis
model is based on the sequence-to-sequence neural networks with extracted
evidences as features. Experiments show that our extraction-then-synthesis
method outperforms state-of-the-art methods. | ['Nan Yang', 'Furu Wei', 'Chuanqi Tan', 'Weifeng Lv', 'Ming Zhou', 'Bowen Du'] | 2017-06-15 | null | null | null | null | ['passage-ranking'] | ['natural-language-processing'] | [ 3.92325252e-01 1.86258391e-01 -5.37109748e-02 -3.57172996e-01
-1.41734600e+00 -6.72543108e-01 4.53193635e-01 6.55301988e-01
-6.35779142e-01 8.05578649e-01 6.74187422e-01 -5.77040434e-01
-2.61882901e-01 -9.48466837e-01 -7.86224365e-01 8.93239528e-02
3.10310721e-01 3.88976067e-01 7.28705764e-01 -5.86087465e-01
7.64727890e-01 -2.50059605e-01 -1.55181885e+00 8.78492832e-01
1.17515731e+00 1.14061904e+00 2.64997989e-01 1.41897249e+00
-6.38370931e-01 1.42919552e+00 -8.60988140e-01 -4.88723099e-01
-7.20657110e-02 -1.01554251e+00 -1.67169905e+00 -3.96378517e-01
5.26178598e-01 -6.26560807e-01 -2.92028219e-01 7.16681778e-01
4.19166952e-01 2.77490109e-01 6.40195489e-01 -7.85926104e-01
-8.96967113e-01 7.28069842e-01 1.55953601e-01 6.31388485e-01
1.29782963e+00 -1.81292314e-02 1.36102223e+00 -1.14726019e+00
3.95251662e-01 9.99059737e-01 3.48675519e-01 4.27975386e-01
-5.97115755e-01 4.24949713e-02 2.06266522e-01 8.00898910e-01
-6.03208482e-01 -2.63341755e-01 5.31477451e-01 -2.00240940e-01
1.36714339e+00 4.89037693e-01 5.02399027e-01 7.21542358e-01
2.00551718e-01 1.30584717e+00 7.73101926e-01 -6.89943373e-01
5.27055115e-02 -4.57136095e-01 8.20745111e-01 7.71619022e-01
-3.71897042e-01 -2.95310229e-01 -6.84849381e-01 1.52038246e-01
-4.24404666e-02 -3.05940002e-01 -5.00559449e-01 4.73623127e-01
-1.22207475e+00 9.87770498e-01 4.14147556e-01 2.52571143e-02
-4.16352034e-01 -2.27267012e-01 5.57086170e-01 7.34124720e-01
3.38548571e-01 9.16647673e-01 -7.65357912e-01 -1.59803301e-01
-1.11077690e+00 7.28560805e-01 1.38457429e+00 7.12331116e-01
5.16755521e-01 -6.82585001e-01 -8.70155394e-01 7.38960207e-01
2.42156722e-02 8.79248157e-02 4.41668153e-01 -1.00586319e+00
1.08316350e+00 7.63005674e-01 3.04815084e-01 -1.04474008e+00
-4.75055516e-01 -4.04611528e-01 -4.47837681e-01 -4.02803332e-01
5.41495800e-01 -2.24085942e-01 -5.19365072e-01 1.41797137e+00
3.06267440e-01 -3.55194211e-02 1.88371003e-01 7.54314959e-01
1.68097699e+00 9.23592269e-01 -3.16633284e-01 -1.51383385e-01
1.49296522e+00 -1.82943821e+00 -8.27908754e-01 -3.33609641e-01
6.25529766e-01 -7.71541059e-01 1.14974642e+00 2.23189428e-01
-1.37295187e+00 -8.93670321e-01 -1.21824086e+00 -7.13922322e-01
-3.59416842e-01 2.50738949e-01 -1.64325923e-01 -3.45848650e-01
-8.94770443e-01 4.74045217e-01 7.79411048e-02 -1.15582459e-01
-4.01160419e-02 -1.14166401e-02 -4.88418490e-02 -1.98149636e-01
-1.60075390e+00 1.22216535e+00 5.89014292e-01 1.82503276e-02
-5.43747425e-01 -7.26934195e-01 -9.35030639e-01 4.32774156e-01
5.44933677e-01 -1.17291570e+00 1.76691282e+00 -4.67601031e-01
-1.45779836e+00 7.66716957e-01 -5.55554211e-01 -5.34063339e-01
9.24738422e-02 -6.87070906e-01 -4.05925423e-01 7.47375190e-01
3.81199986e-01 5.86440265e-01 7.28130579e-01 -8.28080952e-01
-9.70426738e-01 1.70655157e-02 6.98986709e-01 3.08208346e-01
1.04909934e-01 1.46044925e-01 -4.25299734e-01 -2.50060052e-01
1.39006197e-01 -3.40332240e-01 2.81206612e-02 -6.52679741e-01
-5.06724477e-01 -6.82404339e-01 1.77021950e-01 -1.28205633e+00
1.78110528e+00 -1.36874604e+00 2.96683520e-01 -3.05227190e-01
4.10522312e-01 2.44368806e-01 -7.01785028e-01 7.64324784e-01
1.62989929e-01 1.90797281e-02 -1.12343729e-01 5.94009608e-02
2.11631745e-01 -2.59311795e-01 -6.23186767e-01 -3.77787024e-01
4.40023482e-01 1.14302552e+00 -1.18905079e+00 -5.01643479e-01
-3.79881710e-01 -3.08217436e-01 -5.25785208e-01 8.58657479e-01
-8.77727926e-01 3.30286175e-01 -5.97802043e-01 2.98812836e-01
2.82520771e-01 -4.66941535e-01 -4.14287984e-01 1.13955438e-01
5.95163219e-02 1.08149266e+00 -5.37384510e-01 1.59620857e+00
-4.44360256e-01 4.75813955e-01 -3.31761926e-01 -6.79162025e-01
8.46379459e-01 3.73144269e-01 -1.55056670e-01 -9.70866859e-01
1.91903189e-02 4.28604424e-01 -1.93309356e-02 -1.39585531e+00
9.84786510e-01 2.45945871e-01 -2.43418634e-01 5.44780850e-01
2.80296743e-01 -3.87454629e-01 8.95298481e-01 4.98326182e-01
1.32928658e+00 4.74768907e-01 6.07293963e-01 1.44485444e-01
1.16332197e+00 1.22891806e-01 -4.16026413e-02 1.05147147e+00
1.09700263e-01 6.69096172e-01 6.54683888e-01 -3.35888475e-01
-7.94320345e-01 -8.70415866e-01 4.49079365e-01 1.33316183e+00
-1.28654405e-01 -6.33954585e-01 -1.02163529e+00 -1.09986746e+00
-1.49264738e-01 8.46647739e-01 -7.04968452e-01 5.29258978e-03
-7.51836002e-01 7.03235716e-02 3.54834020e-01 5.91269255e-01
6.05866194e-01 -1.23044062e+00 -5.99715590e-01 5.30014157e-01
-1.06583321e+00 -9.67301428e-01 -6.66380823e-01 -7.28150178e-03
-5.22249401e-01 -1.21506083e+00 -7.00473011e-01 -1.05724156e+00
3.31936836e-01 -5.91700152e-03 1.65555680e+00 6.64180577e-01
3.59715432e-01 3.62502217e-01 -8.63969803e-01 -3.44391525e-01
-3.58321041e-01 5.16312838e-01 -5.82759261e-01 -3.20566475e-01
6.37004256e-01 -3.11574817e-01 -7.91686356e-01 3.63290273e-02
-7.76569545e-01 7.89291710e-02 2.94799656e-01 9.34097886e-01
4.03921396e-01 -5.16820669e-01 1.24239767e+00 -6.72609985e-01
1.33558381e+00 -6.55928969e-01 -9.64919329e-02 6.24065816e-01
-1.37059137e-01 1.27110243e-01 8.47038031e-01 -1.25989884e-01
-9.89296734e-01 -5.37450850e-01 -8.04811001e-01 4.49406058e-01
-3.10954228e-02 1.05921412e+00 -3.34148183e-02 5.83079517e-01
7.77996600e-01 4.06991750e-01 -4.18017238e-01 -4.70296293e-01
5.77559769e-01 6.36718631e-01 6.39184117e-01 -7.08207309e-01
5.55245578e-01 -2.54311025e-01 -2.17584461e-01 -5.04360378e-01
-1.98440206e+00 -7.55929530e-01 -9.06250358e-01 -4.18330491e-01
9.58164632e-01 -4.25594866e-01 -7.11103678e-01 2.44693428e-01
-1.63192952e+00 -5.79523779e-02 -2.05938324e-01 1.95414260e-01
-6.44063652e-01 4.61623162e-01 -8.98095250e-01 -5.74927866e-01
-6.78757071e-01 -7.42011726e-01 8.19456995e-01 6.28294051e-01
-6.30933523e-01 -8.01628113e-01 1.45287290e-01 8.77921402e-01
2.15626910e-01 -2.12871376e-02 1.36126471e+00 -1.10649395e+00
-6.15288973e-01 -2.40867510e-01 -2.08292127e-01 2.94629633e-01
-1.75227597e-01 -4.29618239e-01 -8.32270324e-01 9.47961509e-02
1.51262477e-01 -8.05367589e-01 1.17981017e+00 -2.23033223e-02
1.14420438e+00 -5.64175189e-01 2.06945479e-01 -4.80615497e-02
1.02486193e+00 -2.87149828e-02 7.00171649e-01 5.90118825e-01
3.77251893e-01 9.35379088e-01 6.64579749e-01 6.44332692e-02
8.38142812e-01 4.18435670e-02 3.42936933e-01 2.84906000e-01
-2.02537104e-01 -6.25769794e-01 2.09980771e-01 1.58597732e+00
4.06742573e-01 -4.78759408e-01 -5.71673155e-01 8.40291560e-01
-1.91844773e+00 -9.95853007e-01 -5.92522144e-01 1.86260104e+00
1.18843544e+00 2.74974674e-01 1.05132416e-01 2.34103784e-01
1.55459106e-01 3.14109117e-01 -3.35488349e-01 -6.10029697e-01
2.15326734e-02 5.64910114e-01 -5.39642394e-01 8.56159806e-01
-8.42967808e-01 6.53793037e-01 6.08972263e+00 6.84363127e-01
-3.52738857e-01 -1.31313935e-01 3.38154376e-01 1.51487052e-01
-3.57421815e-01 8.99504349e-02 -1.03285122e+00 2.07819402e-01
9.14490938e-01 -1.17729239e-01 1.99901551e-01 3.29684108e-01
-7.95882568e-02 -4.04196769e-01 -1.42239153e+00 1.86822310e-01
6.03344440e-01 -1.37183583e+00 1.87927678e-01 -8.01425815e-01
6.30700707e-01 -2.76459992e-01 -2.79121429e-01 7.24507391e-01
-1.07755087e-01 -1.05145657e+00 7.23931015e-01 9.99548435e-01
3.17358702e-01 -6.40977800e-01 9.45842683e-01 8.07472169e-01
-1.11778629e+00 -2.44949415e-01 -3.17997456e-01 -5.62569857e-01
4.04156297e-01 3.14293325e-01 -6.81039572e-01 7.53110647e-01
4.77619648e-01 4.35330272e-01 -9.40085888e-01 1.18713570e+00
-1.05734098e+00 8.19408953e-01 7.16410801e-02 -6.45860612e-01
3.63078862e-01 1.08061597e-01 4.09083605e-01 1.07578111e+00
2.18491212e-01 3.51923496e-01 1.88602597e-01 7.30378926e-01
-4.06805724e-01 4.19685602e-01 -3.55536453e-02 1.24638118e-01
3.19531947e-01 1.00442958e+00 -3.73308837e-01 -6.74705029e-01
-5.32652020e-01 9.95837927e-01 7.36760080e-01 4.32953030e-01
-3.41977268e-01 -8.77345979e-01 -2.34054685e-01 -2.01508284e-01
3.61695170e-01 4.51951511e-02 -1.78214103e-01 -1.28061986e+00
3.97954792e-01 -1.25962996e+00 6.52407050e-01 -1.09908485e+00
-1.42825687e+00 5.98965228e-01 -2.31895819e-01 -9.78399277e-01
-4.55762565e-01 -6.17323518e-01 -9.57373857e-01 1.11486614e+00
-1.99059963e+00 -8.91575634e-01 -6.95375726e-02 1.90564364e-01
1.04421699e+00 1.04084805e-01 8.15783143e-01 2.58752517e-02
-1.22168891e-01 3.27369332e-01 -2.28642493e-01 4.02434349e-01
4.70231354e-01 -1.45112097e+00 4.36362833e-01 8.59683931e-01
5.89581467e-02 8.27221453e-01 6.72250688e-01 -5.69214702e-01
-9.41194177e-01 -6.26004815e-01 1.82951081e+00 -7.24992990e-01
6.81506336e-01 5.06346785e-02 -1.11375940e+00 5.04786670e-01
8.31366181e-01 -6.49421215e-01 7.62412727e-01 2.45112449e-01
-3.00521970e-01 1.94345012e-01 -6.63259506e-01 6.79182231e-01
6.46157086e-01 -8.26239526e-01 -1.65269709e+00 3.08280349e-01
1.28630579e+00 -6.06169522e-01 -6.92069709e-01 2.90940404e-01
3.69757175e-01 -7.69124568e-01 8.50268543e-01 -1.04706669e+00
1.35258925e+00 -3.93677264e-01 -1.23779513e-01 -1.29285812e+00
-1.67525962e-01 -3.28571290e-01 -7.34379590e-01 1.18074918e+00
9.59063172e-01 -6.01173155e-02 3.68945122e-01 1.15255408e-01
-2.20508784e-01 -1.22403085e+00 -7.74227977e-01 -2.96607524e-01
3.87238413e-01 -1.83539346e-01 8.66197348e-01 3.35219920e-01
3.44591022e-01 1.05822194e+00 -1.29030347e-01 -1.25145048e-01
1.63162246e-01 4.55479771e-01 6.66573703e-01 -1.06964576e+00
-4.10616904e-01 -3.75817329e-01 4.72548872e-01 -1.93039513e+00
2.65933067e-01 -9.43743765e-01 3.72772217e-01 -2.24926043e+00
2.51760006e-01 5.62192321e-01 -1.70347005e-01 2.79844534e-02
-1.00511718e+00 -3.20402026e-01 2.96545595e-01 -1.76957205e-01
-1.14700842e+00 5.52702069e-01 1.56607521e+00 -3.02861601e-01
1.14321969e-01 2.50946790e-01 -7.78338730e-01 7.15729773e-01
7.03843355e-01 -3.34327340e-01 -3.90908390e-01 -5.69812477e-01
9.96653378e-01 6.99959874e-01 1.69568673e-01 -7.71849036e-01
5.09931386e-01 1.13628805e-01 4.68988836e-01 -1.19904971e+00
1.31641150e-01 -2.74483204e-01 -1.06857359e+00 1.62434950e-01
-1.04478157e+00 3.95610482e-01 -6.09374940e-02 3.95725638e-01
-4.60224688e-01 -9.59069252e-01 1.10677123e-01 -3.53462249e-01
-4.64183629e-01 -1.26070127e-01 -5.11622965e-01 7.54669785e-01
2.75698364e-01 2.93476246e-02 -6.33674681e-01 -8.06252182e-01
-6.06652081e-01 8.36779594e-01 -2.68948495e-01 6.40427053e-01
9.98682022e-01 -1.18878257e+00 -1.21200275e+00 -3.10324758e-01
2.70442367e-01 -3.49963754e-02 3.67331386e-01 6.71704590e-01
-4.12219256e-01 6.50143623e-01 1.48823768e-01 -2.05926225e-01
-1.29406679e+00 6.66085303e-01 1.63060978e-01 -8.99450421e-01
-3.34269315e-01 1.06364512e+00 -3.20191950e-01 -8.79693866e-01
2.11836413e-01 -6.43517494e-01 -1.09152472e+00 3.65176469e-01
9.71927822e-01 3.29933286e-01 1.05890781e-01 -9.27919596e-02
1.94351271e-01 5.49128592e-01 -1.91388473e-01 -2.00303122e-01
1.09587038e+00 -3.29539955e-01 -4.30821568e-01 4.40843135e-01
1.10621047e+00 4.70396914e-02 -8.89936984e-01 -4.88161713e-01
5.02589166e-01 -9.96301249e-02 -5.30917585e-01 -1.19601667e+00
-1.60621673e-01 1.01164544e+00 -3.15636188e-01 3.44223022e-01
1.08635879e+00 -8.33194144e-03 1.40954101e+00 8.05079520e-01
7.08996952e-02 -1.15964317e+00 5.51063955e-01 1.33628905e+00
1.27917767e+00 -1.07056439e+00 -1.08665004e-01 -1.99952498e-01
-2.94129550e-01 1.43462431e+00 7.86920965e-01 1.58499891e-03
1.93071738e-01 -1.78321451e-01 4.53174207e-03 -2.75213003e-01
-9.88205671e-01 -2.88710326e-01 8.66782904e-01 9.26248059e-02
5.14463365e-01 -3.10528159e-01 -6.54027522e-01 9.88041520e-01
-5.51224411e-01 1.36250615e-01 5.82400143e-01 8.75779629e-01
-1.03208923e+00 -9.87961411e-01 -2.56797463e-01 7.58375764e-01
-4.91207242e-01 -4.19080228e-01 -6.28786147e-01 2.61268854e-01
-8.88688490e-02 1.68924987e+00 -4.10476387e-01 -4.05217379e-01
4.48059022e-01 5.40099144e-01 5.53429842e-01 -7.06053376e-01
-1.04948473e+00 -5.97732127e-01 6.27675295e-01 -2.84210593e-01
-2.68094182e-01 -2.20541477e-01 -1.33832216e+00 2.37844549e-02
-3.45077008e-01 6.42312229e-01 2.10068107e-01 1.47809756e+00
2.17325702e-01 9.67386663e-01 6.08020842e-01 -1.55927926e-01
-8.60934198e-01 -1.30753565e+00 1.61731437e-01 3.11682045e-01
8.18331063e-01 -2.29599234e-02 -2.65170097e-01 7.91661162e-03] | [11.377016067504883, 8.057703018188477] |
a0e6b073-3e4a-4215-a085-d12da64e8089 | improving-encoder-by-auxiliary-supervision | null | null | https://aclanthology.org/2021.acl-long.466 | https://aclanthology.org/2021.acl-long.466.pdf | Improving Encoder by Auxiliary Supervision Tasks for Table-to-Text Generation | Table-to-text generation aims at automatically generating natural text to help people conveniently obtain salient information in tables. Although neural models for table-to-text have achieved remarkable progress, some problems are still overlooked. Previous methods cannot deduce the factual results from the entity{'}s (player or team) performance and the relations between entities. To solve this issue, we first build an entity graph from the input tables and introduce a reasoning module to perform reasoning on the graph. Moreover, there are different relations (e.g., the numeric size relation and the importance relation) between records in different dimensions. And these relations may contribute to the data-to-text generation. However, it is hard for a vanilla encoder to capture these. Consequently, we propose to utilize two auxiliary tasks, Number Ranking (NR) and Importance Ranking (IR), to supervise the encoder to capture the different relations. Experimental results on ROTOWIRE and RW-FG show that our method not only has a good generalization but also outperforms previous methods on several metrics: BLEU, Content Selection, Content Ordering. | ['Dayong Hu', 'Yinliang Yue', 'Can Ma', 'Liang Li'] | 2021-08-01 | null | null | null | acl-2021-5 | ['table-to-text-generation', 'data-to-text-generation'] | ['natural-language-processing', 'natural-language-processing'] | [ 0.07101449 0.45584884 -0.2719588 -0.60746944 -0.6495738 -0.45486692
0.53584856 0.37886974 -0.19090001 1.1771047 0.6429435 -0.31371623
-0.13112876 -1.335747 -0.8113982 -0.31700778 0.20348895 0.75790846
0.03694075 -0.52060294 0.286199 -0.0599088 -1.5151111 0.57330483
1.1238253 0.99231637 0.14053681 0.40021992 -0.5428389 1.4010947
-0.953765 -1.1972759 0.05032035 -0.68119144 -0.84011793 -0.05045793
0.03729657 -0.5437302 -0.40131932 1.1308075 0.4467345 0.16435671
0.81056595 -1.2513875 -1.04839 1.5065337 -0.5810633 0.10896377
0.5226017 -0.30303207 1.481355 -0.63546604 0.7641659 1.287049
0.2708645 0.47045007 -0.7136912 -0.5198111 0.25448236 0.400755
-1.2848198 -0.17686002 0.6916004 -0.06176538 0.59170324 0.5866394
0.59264266 1.0727948 -0.03860856 0.8654946 0.7515192 -0.3377376
0.06203892 0.39982802 -0.18129396 0.55660206 0.6104244 -0.48911968
-0.6835069 0.25797263 0.7983424 -0.25264794 -0.21025768 -0.0457727
-1.4414854 0.7787377 0.56056017 0.06009024 -0.37567064 -0.09455194
0.5092618 0.15045404 0.1973341 0.71208185 -0.44728717 -0.09351567
-0.4821437 0.61535865 1.0564808 1.4862556 0.4753661 -0.1885979
-0.6673516 0.66045105 0.04793283 0.18690433 0.49054846 -0.458739
1.2336336 0.8507176 0.2866557 -1.2659217 -0.28844237 -0.26935896
-1.3608938 -0.55636144 0.38300854 -0.300772 -0.7442735 1.5492222
0.19609894 -0.55646247 0.15390788 0.86971897 1.1543133 0.72075284
0.0396904 -0.16824362 1.499747 -0.73175037 -1.1373696 -0.09627051
0.5834802 -0.4919558 1.0737926 0.13211104 -1.1395563 -0.58940166
-0.9346712 -0.48550838 -0.6992246 0.5985159 0.8516658 0.32356167
-0.60643494 0.56513375 -0.24984741 0.0546872 0.34899214 0.21449916
-0.15608676 0.16532305 -1.9218298 0.8073716 0.87152165 0.29613394
-0.12994264 -0.4189701 -0.94176227 0.28802976 0.7693755 -0.94574887
1.0981286 -0.47229943 -1.1390765 0.5642938 0.06345425 -0.44628996
0.61826706 -0.10660993 -0.48410314 -0.2256095 0.2917197 0.597421
0.19249487 -1.1661838 -0.9223457 -0.28732914 0.39147145 0.6287614
-0.27465326 -0.18124186 -0.61413825 -0.8081579 0.09895349 -0.4407181
-0.08523747 -0.4537022 -1.1414498 -0.5797931 -0.01469243 -0.9123656
1.6992971 -1.6862224 0.03932957 0.11273833 0.3601468 -0.15430532
0.24275109 0.4855658 0.05537973 0.35259596 0.08126169 0.15204212
0.21901473 -0.0131277 -0.5299425 -0.31829366 0.18897846 1.0804466
-0.8370925 -0.7481586 -0.27171883 0.12261535 -0.488051 0.31186935
-0.38900846 -0.1154816 -0.6201619 0.54063463 0.37600678 -0.3549251
0.29672247 -0.34850875 0.16309845 0.8269166 -1.3298029 1.3282639
-0.23848875 0.34210274 -0.5339254 -0.7594968 1.0689949 0.13776906
0.14707701 -0.78803396 0.18233068 0.09403773 0.11195396 -0.63165486
0.9773869 0.05112854 -0.3595306 0.29469416 -0.21050003 0.05611218
0.7258702 0.4331092 0.83310616 0.06101584 0.3723585 -0.00665025
0.5189889 -0.00973739 0.6392158 0.53601295 0.42250368 0.5232511
1.1197679 -0.4894654 -1.1174599 -0.96571034 0.18169475 0.8925029
0.24037445 -0.8298012 -0.7438387 -0.8097911 -0.06061571 1.0184115
-0.88815075 -0.28179097 -0.5529657 -0.8290197 0.29726502 0.8283435
0.5180635 -1.204871 -0.21986453 0.3529932 -0.7675549 -1.1350871
-0.41047183 0.19318016 -0.6150585 -0.8455778 -0.43059477 -0.6264397
0.86882454 -0.08017235 1.5157697 0.03452378 0.07695056 -0.53571546
-0.47007936 -0.82835597 -0.32320616 0.40159252 -0.1564139 -0.23124526
0.53648394 -0.25342518 -0.36462954 0.15610851 -0.84122306 0.50506604
0.76718944 0.856188 0.56754833 0.65846294 0.72657746 -1.3319267
1.1411884 -0.46164206 -0.41898033 0.52698606 -0.6906427 0.36671495
0.97915834 -0.13424084 -1.0866272 -0.2469311 -0.04264522 0.05707604
0.05539215 0.71980107 -0.77121013 1.085883 0.5014301 0.3711154
-0.5146712 -0.19561204 0.45222515 0.69338185 0.482148 -0.6571444
0.9003596 0.0499494 -0.11308744 -0.06840575 -1.1715477 0.04266786
-0.5730688 0.05135952 0.81409466 -1.0211077 -0.7922109 0.05495619
-1.2504387 0.14103524 -0.40583578 0.29305503 -0.37021253 -0.09353949
-0.63192123 -0.7281479 -0.4331544 -0.83664525 0.9205855 0.4236204
-0.25142127 -0.76927745 -0.53087395 0.37644312 0.1189243 0.23446277
1.3437322 -0.6149575 -0.80841243 -0.18536915 -0.45939884 -0.0998373
0.1885853 0.01273029 -0.52177966 0.2817436 -0.33960956 -0.36200398
0.79274315 0.03679847 1.5100445 -0.7806123 -0.23111796 0.2573114
1.1719593 0.37651655 0.8753935 0.28198615 0.94651157 0.8383345
0.7212457 0.618765 0.9539998 0.6028272 0.140768 0.01683864
0.06429034 -0.9357863 0.11017971 0.8865941 -0.345764 -0.56484264
-0.50212353 0.34068385 -1.8074807 -0.95980793 -0.18523012 2.0518641
1.2842418 0.6070915 0.06580485 0.25752416 0.72998965 0.00605431
-0.5703699 -0.27516422 -0.13731301 -0.26453215 0.3450921 0.1096578
-0.7817858 0.9145369 5.2497735 0.84392565 -0.5181074 -0.4943796
1.0321983 0.09263591 -0.7505342 -0.18231522 -1.2006325 0.5392886
0.63329756 -0.592471 0.4561529 0.72940093 -0.13036299 0.06309576
-1.1505133 1.0191476 0.16723886 -1.1872461 0.62124735 -0.08915576
0.57094747 -0.7801424 -0.23274884 0.6542522 0.59877056 -1.0818796
0.86639416 0.6000139 0.8384356 -1.108186 0.8381635 0.42446098
-1.2295445 0.1115612 -0.7837839 0.04107825 -0.1009442 0.81841
-0.90780056 0.8229589 0.5526253 0.46362305 -0.7605942 0.5178989
-0.6868563 0.05242416 0.15877557 -0.5578445 -0.16591552 -0.16178775
0.09263127 0.81318885 0.39858273 0.07521469 -0.21439506 1.2547241
-0.77393097 0.32681876 -0.5167702 -0.18623869 0.62774223 1.2725542
-0.73101765 -0.60170233 -0.28605464 0.89371777 0.59232557 0.26345167
-0.8916868 -0.8204799 0.40048394 0.26562804 0.21854177 0.1364987
-0.46229047 -1.381473 0.38004377 -0.9790805 0.514892 -0.95858306
-1.1803406 0.767419 -0.05925094 -1.3190954 -0.516989 -0.37595257
-0.36366647 0.95640796 -1.2667158 -0.7344385 -0.2072122 0.3272145
0.53611785 -0.18597876 0.5698486 0.24819171 -0.70425254 0.7906223
-0.27102342 0.6002401 0.6008005 -1.7296978 0.68014973 0.74741334
0.3145166 0.71692246 0.57381946 -0.7550993 -1.2356964 -1.0174063
1.3684946 -0.54144907 0.4787406 -0.5404973 -0.7026607 0.6197606
0.28174564 -0.4763219 0.61880344 0.3115603 -0.30155522 -0.2168272
-0.7278289 0.9566487 1.1641144 -0.27669892 -0.65745145 0.35778356
0.8780586 -0.86190385 -0.7626301 0.14815286 0.3811662 -0.8135929
0.83988464 -0.79264015 1.2256824 -0.3051596 0.06684501 -1.5454981
-0.3990264 -0.36295623 -0.33633056 1.5775658 0.8354882 -0.14292724
0.8387441 0.9457665 0.10513137 -0.99210364 -0.38524517 -0.45119756
-0.09022201 0.00690019 1.312756 0.96094495 0.18265109 0.8601852
-0.57567275 -0.15820739 0.31607875 0.37001243 0.7108611 -1.28284
-0.18674922 -0.4455717 -0.04314957 -1.0595572 -0.07895957 -0.8476871
-0.08570987 -2.0493872 0.30083576 -0.3203802 -0.14331052 0.41243213
-0.758998 -0.39193773 0.11287111 0.00806653 -0.6679944 0.5763756
1.7686815 -0.08271746 -0.05930242 0.04132714 -1.5674856 0.39769354
0.8070308 -0.29086423 -0.591796 -0.4205753 0.9124623 0.36048883
-0.04456989 -0.9426707 0.2530079 -0.33059707 0.7886828 -0.7747961
-0.07019722 -0.58869827 -0.17218736 0.09839103 -0.7982116 0.4177043
-0.19957563 0.4548958 -0.377883 -0.19169776 0.14531434 -0.28221965
-0.40390512 0.3080623 0.01044086 0.45333558 0.5525526 0.20336604
-0.5928096 -0.5708231 -0.25194854 0.31092772 0.04764067 0.70515716
0.5322212 -1.6332332 -0.881838 0.15309016 0.2461581 0.45845988
0.1336772 0.32846683 -0.40119198 0.5703751 -0.2632825 0.23941585
-0.86488944 0.57993 0.00972927 -0.7519063 -0.44275188 0.9457849
-0.01291196 -0.34890318 0.3698351 -0.64830846 -0.68248194 0.47438455
0.7217681 0.18774812 0.1202376 -0.31401166 0.09320254 -0.0471251
-0.43698755 0.07459641 1.0461428 -0.09765702 -0.02612949 0.49635163
0.76042813 0.08510957 -0.8361544 -0.2366509 0.03256396 -0.37252024
-0.31236187 -0.9496206 -1.1047163 0.7403432 -0.19557153 0.6313206
1.0800699 -0.14817098 0.99309015 0.6035198 0.26554304 -1.3115705
-0.18042967 0.4169232 1.0806961 -1.1848999 0.13344395 -0.7062893
-0.9904252 1.0870935 1.0147971 0.46164465 0.15181436 0.19597603
-0.06844328 -0.1372093 -0.9526488 -0.3002604 0.43944654 0.33836988
0.81559426 0.23980653 -0.44912815 0.99722904 -0.9635108 -0.14913885
0.731559 0.45475006 -0.23346318 -1.141825 -0.0845934 1.0420986
-0.5493859 -0.2955994 -0.6559477 0.72038066 0.08344128 0.7928098
0.00978949 -0.66219777 0.6421973 -0.03026074 0.24161713 -0.5513157
-0.4575083 -0.30633083 0.3828661 -0.18601055 -0.0155836 -0.5222155
-1.3622888 -0.44505244 -0.23000857 0.4181361 0.41525832 0.6520567
0.16453694 1.0615097 0.607164 0.00585629 -0.5004594 -1.1541072
-0.8615937 0.64331543 -0.1186206 -0.52363 -0.06957126 0.06755925] | [11.666080474853516, 8.821560859680176] |
7198677c-97e9-4390-825d-4e8a3e94c0da | detr-taming-your-multi-scale-detection | 2206.02977 | null | https://arxiv.org/abs/2206.02977v1 | https://arxiv.org/pdf/2206.02977v1.pdf | DETR++: Taming Your Multi-Scale Detection Transformer | Convolutional Neural Networks (CNN) have dominated the field of detection ever since the success of AlexNet in ImageNet classification [12]. With the sweeping reform of Transformers [27] in natural language processing, Carion et al. [2] introduce the Transformer-based detection method, i.e., DETR. However, due to the quadratic complexity in the self-attention mechanism in the Transformer, DETR is never able to incorporate multi-scale features as performed in existing CNN-based detectors, leading to inferior results in small object detection. To mitigate this issue and further improve performance of DETR, in this work, we investigate different methods to incorporate multi-scale features and find that a Bi-directional Feature Pyramid (BiFPN) works best with DETR in further raising the detection precision. With this discovery, we propose DETR++, a new architecture that improves detection results by 1.9% AP on MS COCO 2017, 11.5% AP on RICO icon detection, and 9.1% AP on RICO layout extraction over existing baselines. | ['Jindong Chen', 'Xinying Song', 'Hao Zhang', 'Frederick Liu', 'Xiaoxue Zang', 'Lijuan Liu', 'Chi Zhang'] | 2022-06-07 | null | null | null | null | ['small-object-detection'] | ['computer-vision'] | [ 9.72532555e-02 -1.40012205e-01 -2.36244783e-01 -4.53089811e-02
-5.43844163e-01 -6.48920894e-01 6.53807044e-01 -1.27014723e-02
-9.11000609e-01 1.84081256e-01 -3.51019427e-02 -5.22185981e-01
4.36600000e-01 -7.28989422e-01 -6.68442607e-01 -2.73591131e-02
7.60264173e-02 -1.36480927e-01 6.16476476e-01 -2.91513443e-01
2.50319660e-01 5.59804738e-01 -1.23529065e+00 6.02813661e-01
5.68336606e-01 1.14214706e+00 1.04140937e-01 6.15313888e-01
-2.63885912e-02 1.02825260e+00 -7.93299258e-01 -5.58772862e-01
4.15564448e-01 -2.82224100e-02 -5.89079261e-01 -3.68449837e-01
8.79028082e-01 -7.44013071e-01 -6.78581238e-01 1.26114559e+00
6.38655782e-01 -3.44799697e-01 5.05681932e-01 -1.08623087e+00
-6.10211372e-01 9.65829551e-01 -9.01668906e-01 6.57709241e-01
2.61799507e-02 2.75233626e-01 1.12060428e+00 -1.32075882e+00
4.71430451e-01 1.43278348e+00 7.21823752e-01 4.05092418e-01
-1.01518679e+00 -8.96456540e-01 6.40769079e-02 4.65687484e-01
-1.35008430e+00 -4.24841106e-01 4.11446869e-01 -1.16685815e-01
1.22635865e+00 7.78730139e-02 5.11493027e-01 1.13440931e+00
1.35765344e-01 1.36104369e+00 9.62287664e-01 -4.93739754e-01
-9.72935706e-02 -2.87055429e-02 1.24513060e-02 8.90246749e-01
3.91067624e-01 -7.80424327e-02 -5.55472016e-01 1.86754346e-01
6.17212355e-01 2.39899512e-02 8.41068402e-02 1.80619769e-02
-9.66928542e-01 8.27349782e-01 9.41433430e-01 4.44674641e-01
-2.03795955e-01 3.61570179e-01 6.48334205e-01 4.54030326e-03
2.58129686e-01 8.45246911e-01 -2.57451743e-01 -3.39583047e-02
-1.19783342e+00 2.06257597e-01 3.94333273e-01 9.96087968e-01
3.46489549e-01 1.49172828e-01 -4.37371284e-01 7.65023470e-01
-2.67297141e-02 3.37315649e-01 4.15505856e-01 -6.14909649e-01
5.54888666e-01 9.74483609e-01 -1.85501799e-01 -9.30482507e-01
-5.14755487e-01 -8.50741982e-01 -6.52107358e-01 1.67879492e-01
6.03880584e-01 -7.16357678e-02 -1.17560935e+00 1.49460721e+00
-1.67955533e-01 -1.30260959e-01 -3.21210533e-01 9.63254273e-01
8.19700778e-01 3.66085291e-01 9.99015719e-02 4.77464348e-01
1.78806365e+00 -1.22356927e+00 -4.86216635e-01 -5.25892973e-01
7.08937943e-01 -9.89735305e-01 9.72883761e-01 6.03982508e-01
-9.34061468e-01 -5.77162921e-01 -1.51834536e+00 -2.83224702e-01
-5.01228869e-01 5.65396786e-01 7.10766852e-01 8.56548905e-01
-8.71067703e-01 4.89734083e-01 -9.42005515e-01 -5.49634993e-01
8.91048193e-01 3.50341916e-01 -2.46660635e-02 2.42471807e-02
-1.05719101e+00 9.26208377e-01 3.21699888e-01 4.65690568e-02
-9.24245477e-01 -6.18033528e-01 -4.94166791e-01 2.32520819e-01
6.52343452e-01 -2.60868341e-01 1.35046864e+00 -7.48531699e-01
-1.09432495e+00 6.61860347e-01 7.28284642e-02 -9.19446349e-01
7.37256527e-01 -6.85707986e-01 -3.22804749e-01 2.01582879e-01
2.62205899e-01 1.01260817e+00 8.23285878e-01 -6.66726708e-01
-9.98272836e-01 -1.95566162e-01 5.97312689e-01 -1.67050421e-01
-7.14115798e-01 1.69942528e-01 -7.16910958e-01 -6.77053213e-01
7.44142011e-03 -8.65883768e-01 -3.05139184e-01 6.22339323e-02
-6.68320835e-01 -2.53972352e-01 1.05235338e+00 -5.23054063e-01
1.40927386e+00 -2.17871475e+00 -4.21991676e-01 -6.61600307e-02
7.43433952e-01 6.95612192e-01 -2.39413723e-01 5.71070127e-02
1.16619408e-01 3.18050593e-01 1.00634232e-01 -4.48463172e-01
-1.51590547e-02 -2.51106232e-01 -1.87791914e-01 4.74562138e-01
5.29808581e-01 1.13126898e+00 -8.79008114e-01 -4.18892592e-01
1.96785495e-01 4.53797162e-01 -6.62135124e-01 -2.51273274e-01
7.71457180e-02 -2.45516822e-01 -2.55990237e-01 8.72925103e-01
7.08937466e-01 -2.65447468e-01 4.64987606e-02 -5.77751935e-01
-3.84841323e-01 4.24120367e-01 -9.58880126e-01 1.39502597e+00
-3.05283040e-01 1.10538208e+00 7.96256214e-02 -7.22089827e-01
6.89874291e-01 8.82430002e-02 1.67356268e-01 -1.04310608e+00
4.01096731e-01 2.65328407e-01 3.87714922e-01 -5.02490923e-02
7.66959190e-01 4.66385931e-01 -1.07042026e-02 -1.07868649e-01
4.90690097e-02 3.84594828e-01 4.47017789e-01 5.20573616e-01
1.48256111e+00 -2.53327399e-01 1.98602304e-01 -3.73939097e-01
3.45672131e-01 1.03332056e-02 4.25209731e-01 1.16914594e+00
-4.32679504e-01 6.97915673e-01 5.91381550e-01 -5.13827622e-01
-1.00605404e+00 -8.65657568e-01 -3.41164470e-01 1.15568757e+00
1.90825369e-02 -7.88404882e-01 -7.81031549e-01 -8.62178445e-01
-5.98721653e-02 3.59915227e-01 -5.94830692e-01 -1.71020284e-01
-8.77037048e-01 -9.26792681e-01 1.25761902e+00 8.10098588e-01
8.83319259e-01 -9.12640333e-01 -9.94297564e-01 3.39663595e-01
-6.79841414e-02 -1.29661214e+00 -4.79691535e-01 4.80243742e-01
-5.51890790e-01 -9.33260441e-01 -6.49499476e-01 -6.30618453e-01
3.20860893e-01 2.16789350e-01 9.17634010e-01 7.81342667e-03
-6.46996319e-01 -1.43872842e-01 -3.19967568e-01 -5.44089675e-01
-1.31350726e-01 4.05312657e-01 -1.36084870e-01 -7.17821643e-02
5.09947538e-01 -1.92105249e-01 -7.78288603e-01 2.02457950e-01
-7.40765095e-01 -1.02388941e-01 8.85377526e-01 7.80268967e-01
4.76158783e-02 -2.70420015e-01 2.47265533e-01 -7.09676266e-01
4.98234600e-01 -1.07957050e-01 -6.99771285e-01 -7.31235966e-02
-6.51207626e-01 4.20808345e-02 7.26390541e-01 -5.82951069e-01
-5.14972568e-01 3.94200496e-02 -1.73696712e-01 -4.63403195e-01
2.07677901e-01 3.46793771e-01 -1.27502820e-02 -2.76724815e-01
8.67053092e-01 2.77524125e-02 -5.29326141e-01 -5.84168434e-01
2.96315253e-01 5.10853469e-01 5.12844920e-01 -5.67290373e-02
6.67607844e-01 4.65811491e-01 -1.22477099e-01 -6.76638663e-01
-9.46451306e-01 -4.47314888e-01 -3.64669114e-01 6.09023198e-02
8.18835914e-01 -1.06447220e+00 -6.31689489e-01 4.31115270e-01
-1.26912761e+00 1.21838890e-01 4.03618403e-02 3.03436279e-01
1.37821108e-01 1.84672982e-01 -8.75166714e-01 -6.78864479e-01
-4.52082366e-01 -1.40726244e+00 1.11862898e+00 1.29039049e-01
-2.37981960e-01 -4.24162149e-01 -5.20601869e-01 1.02889918e-01
7.01662898e-01 1.72827303e-01 6.45149469e-01 -5.27894199e-01
-7.32608676e-01 -3.84247243e-01 -9.00470197e-01 3.94086599e-01
-1.93396628e-01 -5.09758946e-03 -1.07824564e+00 -4.43629086e-01
-2.46196449e-01 -1.08315475e-01 1.32383263e+00 2.83164829e-01
9.61777806e-01 -1.18875146e-01 -5.11285961e-01 6.17419183e-01
1.42409384e+00 6.26958460e-02 5.41719973e-01 6.67691588e-01
8.35335910e-01 7.69455954e-02 2.99456686e-01 3.60157818e-01
1.83178455e-01 7.41814494e-01 5.91745555e-01 -3.81360054e-01
-5.05018473e-01 -2.34291241e-01 4.61927712e-01 3.05985093e-01
9.53581370e-03 -2.31832281e-01 -1.21129620e+00 6.83141291e-01
-1.60340881e+00 -7.00705588e-01 -2.10531771e-01 1.67737210e+00
5.32450259e-01 7.79728055e-01 2.67533571e-01 9.28127840e-02
6.81594431e-01 9.11307037e-02 -5.19645989e-01 -3.34885418e-01
-3.81944418e-01 3.79804134e-01 1.09674346e+00 1.40467733e-01
-1.43844450e+00 1.19041610e+00 6.01463604e+00 9.74417746e-01
-1.52489138e+00 1.61429986e-01 5.41254938e-01 -2.42288366e-01
4.12113428e-01 -1.66821301e-01 -1.20462513e+00 2.55712956e-01
7.05542326e-01 6.67904466e-02 1.30772665e-01 1.07791722e+00
-1.47284627e-01 -4.02946211e-02 -1.05378425e+00 1.10578144e+00
8.89566615e-02 -1.37060404e+00 9.62158591e-02 4.13866229e-02
5.17796516e-01 3.65176350e-01 2.07486585e-01 6.47080541e-01
8.08374658e-02 -1.12749422e+00 1.12817919e+00 -2.28966087e-01
6.48420274e-01 -8.54192257e-01 8.28895926e-01 1.48592532e-01
-1.38216686e+00 -2.87806869e-01 -3.81814390e-01 -1.00381911e-01
-2.37154365e-02 5.45340538e-01 -1.01460373e+00 1.22673519e-01
8.85000408e-01 6.47027969e-01 -1.05036020e+00 1.23685622e+00
-3.35648924e-01 8.48484278e-01 -5.39626241e-01 -2.20291659e-01
6.67069793e-01 5.02380729e-01 6.85450554e-01 1.58286178e+00
6.94244653e-02 -4.06277627e-01 2.59342819e-01 9.45864201e-01
-5.40366292e-01 7.08273426e-02 -2.07724005e-01 -9.22654495e-02
3.89769733e-01 1.43142700e+00 -9.45882797e-01 -4.24772978e-01
-4.38065261e-01 9.30990815e-01 3.93778682e-01 1.20062947e-01
-1.08311474e+00 -6.21091843e-01 3.12408656e-01 3.19164038e-01
5.33323824e-01 -1.52322292e-01 -4.70145494e-01 -1.03061247e+00
2.05246449e-01 -9.91115928e-01 1.94062993e-01 -1.72346592e-01
-9.37851489e-01 7.32522964e-01 -2.65301585e-01 -1.14875531e+00
2.23282695e-01 -8.58777642e-01 -4.87465560e-01 5.73806286e-01
-1.35235620e+00 -1.20571733e+00 -1.11853279e-01 2.36676201e-01
6.18774891e-01 -8.73973221e-02 4.41023856e-01 8.87903392e-01
-7.35669494e-01 1.03071463e+00 -1.45035833e-01 5.90312004e-01
7.37864912e-01 -1.22606611e+00 8.76396954e-01 1.05808830e+00
2.41179019e-01 8.24156165e-01 4.70400900e-01 -3.95066649e-01
-1.41881216e+00 -1.22969842e+00 6.52685761e-01 -4.42586362e-01
7.33439445e-01 -8.26292694e-01 -6.54599190e-01 6.07094884e-01
1.63854837e-01 1.98305741e-01 -2.18324326e-02 4.74871881e-02
-8.57635200e-01 -4.40696441e-02 -9.23017740e-01 8.91402543e-01
1.05848587e+00 -4.10743773e-01 -3.20659310e-01 5.03013432e-02
5.87370455e-01 -3.54756773e-01 -4.10291195e-01 5.90376556e-01
5.97315669e-01 -9.31551576e-01 9.83043730e-01 -2.66054064e-01
4.80938494e-01 -3.17397803e-01 -5.43096252e-02 -7.72933304e-01
-4.72073615e-01 -5.07538557e-01 -7.47944191e-02 9.72880483e-01
6.65133893e-01 -4.21351433e-01 9.75933552e-01 -5.30954935e-02
-1.52452156e-01 -7.08886087e-01 -9.26957667e-01 -8.63911569e-01
1.55420811e-03 -5.89757800e-01 3.68354172e-01 6.61682129e-01
-1.35497972e-01 6.35729313e-01 -3.72126192e-01 5.80162778e-02
4.36070293e-01 -2.70640999e-01 6.76664412e-01 -8.73842716e-01
-2.90723890e-01 -7.62082875e-01 -5.67410469e-01 -1.47187114e+00
-3.71102363e-01 -9.23511446e-01 -7.79599100e-02 -1.21806920e+00
2.83715427e-01 -1.55154496e-01 -4.03558016e-01 6.58734918e-01
-2.07849830e-01 8.32720399e-01 6.52082503e-01 4.78061028e-02
-8.79286528e-01 1.86856523e-01 9.50617671e-01 -3.75530511e-01
-2.35748775e-02 -4.74909157e-01 -6.41222477e-01 9.25289512e-01
6.85756564e-01 -5.57040453e-01 6.09271452e-02 -6.50832355e-01
1.44364268e-01 -5.62689900e-01 4.86636579e-01 -1.34428871e+00
3.35701048e-01 5.69350243e-01 6.34984314e-01 -8.56741548e-01
1.98007569e-01 -4.61976081e-01 -5.86660683e-01 8.07124972e-01
-1.31753251e-01 2.33396649e-01 5.92146277e-01 1.89837575e-01
-1.94380909e-01 -3.73046175e-02 9.80848134e-01 -4.42354009e-02
-9.70616817e-01 5.85396960e-02 -3.93308729e-01 4.92801927e-02
6.29640639e-01 -1.75339416e-01 -5.54777801e-01 -4.58744131e-02
-4.01207745e-01 1.87121570e-01 9.68174711e-02 7.08934844e-01
5.22992313e-01 -1.13056946e+00 -6.99063778e-01 -1.33589432e-02
1.33446753e-01 -3.39357793e-01 -1.49795994e-01 9.29362416e-01
-6.09282374e-01 5.94904006e-01 -2.22114503e-01 -7.12305188e-01
-1.11773407e+00 3.42374772e-01 3.65224779e-01 -5.27334213e-01
-5.74733317e-01 1.08052313e+00 3.40260506e-01 -1.04193486e-01
3.14028561e-01 -4.49948132e-01 -1.54603899e-01 9.20381099e-02
7.26047695e-01 3.78953993e-01 3.03632975e-01 -3.52859914e-01
-6.45712197e-01 2.63675392e-01 -6.16525412e-01 6.93665177e-04
1.06492341e+00 3.18186551e-01 4.36758921e-02 6.83574155e-02
1.24121845e+00 5.23065329e-02 -1.09001148e+00 -1.87832162e-01
3.25809866e-01 -2.65982926e-01 3.51499081e-01 -8.45559299e-01
-1.17400122e+00 8.81295919e-01 9.02500451e-01 1.46701247e-01
9.56862569e-01 -4.95254695e-02 8.87664437e-01 3.47693563e-01
1.84072480e-01 -9.06069696e-01 1.98878378e-01 7.87808061e-01
7.47121394e-01 -1.20245183e+00 4.42661047e-02 -3.61227334e-01
-1.01253472e-01 1.16098726e+00 8.22162867e-01 -3.39193791e-01
3.85169595e-01 6.75305843e-01 -1.48366153e-01 -2.22230509e-01
-5.37582457e-01 -3.83079171e-01 3.72757554e-01 1.91905171e-01
4.98289466e-01 -3.47537524e-03 -2.87632138e-01 4.00179029e-01
-2.14454070e-01 -1.17445543e-01 4.19754982e-01 9.72503543e-01
-3.96545082e-01 -9.84721780e-01 -3.60829979e-01 5.03286421e-01
-7.71681607e-01 -6.01747990e-01 -5.04547656e-01 8.81492376e-01
3.93830419e-01 9.99313891e-01 1.97786465e-03 -5.29673398e-01
6.68187201e-01 -3.67504150e-01 2.82120943e-01 -6.19652569e-01
-8.97579670e-01 2.15476811e-01 5.82461171e-02 -6.77441776e-01
4.35330719e-02 -4.66496170e-01 -1.12022781e+00 -3.00798059e-01
-6.05319023e-01 -3.40302974e-01 6.28507495e-01 7.24751115e-01
2.71684587e-01 7.47956812e-01 6.63886815e-02 -6.59768879e-01
-7.53102422e-01 -1.10576677e+00 -1.97593793e-01 2.11230218e-01
3.81020606e-01 -5.12665093e-01 -2.93392807e-01 -2.22972438e-01] | [8.843342781066895, 0.06927932053804398] |
986f13e8-f170-4bdd-a9bf-a3cba4ee7be2 | handling-heavy-occlusion-in-dense-crowd | 2304.07705 | null | https://arxiv.org/abs/2304.07705v2 | https://arxiv.org/pdf/2304.07705v2.pdf | Handling Heavy Occlusion in Dense Crowd Tracking by Focusing on the Heads | With the rapid development of deep learning, object detection and tracking play a vital role in today's society. Being able to identify and track all the pedestrians in the dense crowd scene with computer vision approaches is a typical challenge in this field, also known as the Multiple Object Tracking (MOT) challenge. Modern trackers are required to operate on more and more complicated scenes. According to the MOT20 challenge result, the pedestrian is 4 times denser than the MOT17 challenge. Hence, improving the ability to detect and track in extremely crowded scenes is the aim of this work. In light of the occlusion issue with the human body, the heads are usually easier to identify. In this work, we have designed a joint head and body detector in an anchor-free style to boost the detection recall and precision performance of pedestrians in both small and medium sizes. Innovatively, our model does not require information on the statistical head-body ratio for common pedestrians detection for training. Instead, the proposed model learns the ratio dynamically. To verify the effectiveness of the proposed model, we evaluate the model with extensive experiments on different datasets, including MOT20, Crowdhuman, and HT21 datasets. As a result, our proposed method significantly improves both the recall and precision rate on small & medium sized pedestrians and achieves state-of-the-art results in these challenging datasets. | ['Zao Zhang', 'Dong Yuan', 'Zhongzheng Lai', 'Wei Bao', 'Huaming Chen', 'Yu Zhang'] | 2023-04-16 | null | null | null | null | ['multiple-object-tracking'] | ['computer-vision'] | [-4.49333459e-01 -4.00584757e-01 1.57578602e-01 -7.42819011e-02
-2.06130430e-01 -1.83750764e-01 4.84346062e-01 4.87497309e-03
-7.73464501e-01 6.28167093e-01 -1.51674360e-01 1.30644739e-01
5.78710020e-01 -6.31432116e-01 -6.47152066e-01 -7.82091320e-01
-5.22722490e-02 3.42104465e-01 1.07717812e+00 1.37490416e-02
-2.76923805e-01 2.06101179e-01 -1.74752688e+00 -2.60783672e-01
7.54625618e-01 9.62465584e-01 4.42189574e-01 7.41221607e-01
2.22571805e-01 3.62797290e-01 -6.51040494e-01 -6.34810686e-01
4.35457528e-01 9.02843475e-02 -6.12429492e-02 7.52248690e-02
8.94033968e-01 -4.99792308e-01 -4.49574530e-01 1.07526708e+00
8.68653119e-01 8.11023265e-02 3.03668767e-01 -1.32685363e+00
-7.86591247e-02 2.18304694e-01 -1.25115347e+00 5.53975582e-01
2.20896125e-01 5.07343113e-01 4.89035130e-01 -8.18902314e-01
1.16266839e-01 1.46430147e+00 8.24649274e-01 4.95402873e-01
-6.72889769e-01 -9.98199463e-01 3.27890694e-01 4.93932039e-01
-1.59057486e+00 -4.84692127e-01 2.99479127e-01 -4.98518527e-01
2.69837916e-01 9.47134495e-02 8.68898451e-01 8.75938475e-01
4.41548862e-02 9.90464449e-01 8.13294411e-01 -1.96166977e-01
-5.36002070e-02 1.38968468e-01 1.33121461e-01 7.51856983e-01
1.00898635e+00 1.41585916e-01 -2.10528880e-01 2.40685940e-01
5.13880730e-01 -1.22722657e-02 -3.83251309e-02 -4.30795312e-01
-1.07046151e+00 6.11969531e-01 5.37973225e-01 3.66969407e-02
-1.10226296e-01 1.93327352e-01 3.66650432e-01 -5.14348745e-01
1.18791156e-01 -1.72150180e-01 -1.76102370e-01 -2.98209433e-02
-7.65534222e-01 3.25609624e-01 4.46851283e-01 1.15031922e+00
2.32329011e-01 -9.02409479e-02 -3.52502763e-01 5.85121810e-01
6.20929301e-01 1.17886615e+00 1.38145089e-01 -5.52316964e-01
6.25722051e-01 6.07567310e-01 4.84352171e-01 -1.09611130e+00
-5.21060944e-01 -6.68618143e-01 -9.85085249e-01 9.04444531e-02
9.24485981e-01 -3.42521727e-01 -8.62884879e-01 1.50210369e+00
9.23124492e-01 3.88860732e-01 -3.63077551e-01 1.06056774e+00
1.04868042e+00 5.14843524e-01 3.75048399e-01 -1.11491099e-01
1.78696704e+00 -8.77637029e-01 -4.93040174e-01 -5.82649529e-01
3.14924568e-01 -6.91472113e-01 5.21408498e-01 3.87374759e-02
-6.73213243e-01 -8.42578650e-01 -1.02572846e+00 1.59215912e-01
-1.75154060e-01 3.31316262e-01 5.37320018e-01 8.93293023e-01
-6.48568809e-01 -7.85118863e-02 -8.14065158e-01 -6.45854950e-01
6.67041302e-01 4.11799103e-01 -1.16576158e-01 -2.53566295e-01
-9.71785963e-01 8.60324919e-01 4.00945276e-01 1.06566995e-01
-7.64570832e-01 -5.01041949e-01 -6.46877289e-01 -3.18772793e-02
4.62454498e-01 -8.12080562e-01 1.14672077e+00 -1.98254138e-01
-1.00516081e+00 7.67774284e-01 -2.23321021e-01 -5.68716347e-01
8.62423658e-01 -3.27410400e-01 -4.25982237e-01 -1.27852753e-01
2.99162656e-01 7.61310697e-01 7.67847180e-01 -1.12540948e+00
-1.40427220e+00 -3.74354452e-01 -1.84303876e-02 -4.24637571e-02
-1.87615573e-01 2.34548226e-01 -8.43996763e-01 -2.31696784e-01
-2.75888115e-01 -1.07686996e+00 -2.40532741e-01 3.51576239e-01
-3.88141036e-01 -3.35974783e-01 1.05863547e+00 -5.92044532e-01
9.49525476e-01 -1.98898554e+00 -2.46817797e-01 -2.08135307e-01
4.93996769e-01 8.47254455e-01 1.71744153e-01 -3.39353710e-01
4.90290791e-01 -3.54254425e-01 2.90042251e-01 -5.27823865e-01
-1.08880781e-01 9.00848862e-03 3.08089674e-01 7.58112490e-01
-5.70341535e-02 7.00730741e-01 -9.49385405e-01 -9.95052993e-01
5.31389058e-01 6.21483505e-01 -2.94430792e-01 1.75063699e-01
1.28836289e-01 6.66246235e-01 -4.57029045e-01 8.62195253e-01
9.49741721e-01 -2.95034826e-01 -2.28085816e-01 -1.99284077e-01
-3.10544252e-01 -5.29132307e-01 -1.48894358e+00 1.08685207e+00
-1.16199359e-01 6.74532175e-01 1.57102659e-01 -5.82541704e-01
6.77996218e-01 2.07419008e-01 3.42032850e-01 -7.74093390e-01
3.27315122e-01 8.75442177e-02 2.15387806e-01 -4.33659434e-01
4.35736030e-01 2.09503293e-01 1.47130281e-01 -2.82734156e-01
-2.51100272e-01 6.08656824e-01 6.19389951e-01 6.97651058e-02
8.79380167e-01 -2.38508210e-01 5.08694291e-01 -9.59937945e-02
7.79484808e-01 -2.94565171e-01 8.59847784e-01 8.00234020e-01
-7.69561350e-01 3.90241235e-01 -7.23519847e-02 -5.21929502e-01
-8.84366691e-01 -1.05236399e+00 -2.69225061e-01 9.67840850e-01
5.07004857e-01 -2.49732837e-01 -6.17747486e-01 -4.83413965e-01
1.08501434e-01 4.41912271e-04 -4.03411984e-01 1.04690164e-01
-8.50320995e-01 -9.52345073e-01 4.97294635e-01 6.57495677e-01
9.49682117e-01 -5.98069012e-01 -8.47168982e-01 1.16470009e-01
-2.77754307e-01 -1.57881331e+00 -4.55897748e-01 -4.68400538e-01
-2.71803528e-01 -1.32424748e+00 -1.10263240e+00 -6.60328984e-01
4.88301277e-01 7.96615124e-01 7.89393604e-01 1.84367448e-01
-5.01886249e-01 1.13695092e-01 -1.82849213e-01 -7.77392149e-01
7.46737793e-02 -7.00346380e-02 2.82573074e-01 2.06529424e-01
2.37724110e-01 -2.37067819e-01 -1.04869175e+00 6.31266057e-01
-3.24487150e-01 9.43320710e-03 5.93257546e-01 3.17444265e-01
4.06774789e-01 1.51133880e-01 1.93630159e-01 -1.84402823e-01
-2.87437290e-02 -4.08898234e-01 -1.06486356e+00 1.54605001e-01
-7.93208554e-02 -3.21094543e-01 4.76377130e-01 -6.50860250e-01
-8.92190635e-01 3.64027053e-01 -1.65812802e-02 -1.95415914e-01
-1.94505200e-01 -3.68076980e-01 -5.14614105e-01 -2.91096926e-01
4.47892547e-01 9.82284024e-02 -2.70306706e-01 -5.14686286e-01
1.54941887e-01 6.27914667e-01 7.70578802e-01 -3.58473837e-01
1.07925797e+00 6.70871675e-01 3.29997599e-01 -9.48117256e-01
-8.46359611e-01 -8.91383767e-01 -6.03446364e-01 -6.55006170e-01
9.31632042e-01 -1.30251801e+00 -1.08234727e+00 7.17254400e-01
-1.27288401e+00 1.64057359e-01 1.34993955e-01 5.67417324e-01
2.26812661e-01 4.89084363e-01 -2.83326834e-01 -1.06659317e+00
-4.63010669e-01 -1.19571865e+00 1.13250411e+00 7.50802755e-01
1.99744910e-01 -6.41686082e-01 -1.71707124e-01 3.71558428e-01
3.65812659e-01 4.14265245e-01 -1.18962333e-01 -3.35057646e-01
-8.25663149e-01 -4.56411779e-01 -6.53096378e-01 -4.84870188e-02
-7.54476488e-02 -1.39333159e-01 -9.05981064e-01 -4.93485600e-01
-5.35891473e-01 1.60985246e-01 9.38135922e-01 5.95382452e-01
8.07823598e-01 6.42970353e-02 -8.46823692e-01 5.34602106e-01
1.10043550e+00 5.89800961e-02 3.18114549e-01 4.06612575e-01
1.01565444e+00 3.96539360e-01 6.54337287e-01 6.67275667e-01
7.86062539e-01 1.00599468e+00 4.47773665e-01 -2.82456819e-02
-5.65337062e-01 -7.82976523e-02 4.52875584e-01 4.36269939e-01
-2.31472969e-01 -2.20767587e-01 -9.51698363e-01 4.58461612e-01
-1.91781020e+00 -1.03953528e+00 -5.57142079e-01 2.18970919e+00
4.09201086e-01 2.34443337e-01 6.28401816e-01 7.68822897e-03
1.04849839e+00 -8.23395252e-02 -4.85888541e-01 6.75642788e-01
2.28131898e-02 -3.26409250e-01 8.13093305e-01 2.00882167e-01
-1.59360611e+00 7.49349117e-01 5.43383408e+00 8.40964079e-01
-8.77126038e-01 1.48001969e-01 2.79025346e-01 -1.82043001e-01
8.56966376e-01 -3.75831068e-01 -1.66719282e+00 9.84577954e-01
5.60778737e-01 -6.49148226e-02 -4.94806841e-02 9.27434683e-01
2.54297167e-01 -3.50606084e-01 -8.31210375e-01 1.25360739e+00
-9.35653523e-02 -8.94414127e-01 -4.25500661e-01 1.44920379e-01
4.77118403e-01 -2.53201313e-02 6.90879822e-02 4.58762288e-01
2.44297937e-01 -6.51741385e-01 7.93322206e-01 2.81349689e-01
4.83538359e-01 -6.89889133e-01 1.04374397e+00 4.93214875e-01
-1.89327478e+00 -5.51603921e-02 -5.70432246e-01 -3.28811468e-03
4.03235883e-01 5.71554065e-01 -7.05851138e-01 3.16684425e-01
1.00137818e+00 6.06754899e-01 -1.05382395e+00 1.93116856e+00
-9.93737727e-02 3.12942564e-01 -6.87337756e-01 -1.20596878e-01
-1.18374586e-01 2.72586286e-01 5.91220438e-01 1.34058499e+00
3.61429721e-01 8.91695917e-02 5.74188650e-01 3.39750439e-01
5.56267835e-02 -8.32139552e-02 -3.34620118e-01 5.30333281e-01
4.98767823e-01 1.42465389e+00 -7.71961212e-01 -3.76881182e-01
-5.43994069e-01 3.50746214e-01 1.44624427e-01 2.53135040e-02
-1.19214046e+00 -4.19165976e-02 6.50011778e-01 4.28188145e-01
7.07383931e-01 -2.93349057e-01 1.16172053e-01 -1.11793065e+00
2.09760904e-01 -6.66881263e-01 3.21280092e-01 -2.08850756e-01
-1.13519287e+00 3.47644299e-01 9.02723372e-02 -1.32636225e+00
2.28991970e-01 -5.38250685e-01 -5.60028315e-01 4.50918794e-01
-1.33825707e+00 -1.42936087e+00 -8.22279036e-01 5.89563370e-01
3.47873837e-01 -3.17399427e-02 7.28852898e-02 9.20862973e-01
-9.33600247e-01 7.86787868e-01 -1.41992450e-01 4.71987069e-01
6.12845182e-01 -9.22517061e-01 4.21566695e-01 1.20291317e+00
-2.94662595e-01 2.93843508e-01 9.64957833e-01 -6.97644651e-01
-1.22253406e+00 -1.39158964e+00 3.86235058e-01 -4.90651965e-01
3.74225318e-01 -4.21665490e-01 -5.77992082e-01 3.68536532e-01
-8.62935632e-02 5.58051705e-01 3.45229506e-01 -7.18319714e-02
-1.44151030e-02 -3.34416240e-01 -9.98961508e-01 3.58250529e-01
1.13577604e+00 3.41684192e-01 -2.71480739e-01 3.63473594e-01
6.09456480e-01 -5.97154498e-01 -5.39758921e-01 3.93413037e-01
6.73625946e-01 -8.34551036e-01 1.35787058e+00 -4.90180217e-02
-3.01465362e-01 -7.94553518e-01 -1.08550154e-01 -9.71328974e-01
-3.59914601e-01 -1.49565026e-01 -5.81713676e-01 1.36360884e+00
-1.70779645e-01 -4.53691870e-01 8.74867082e-01 4.19686228e-01
1.08998328e-01 -4.55474973e-01 -1.19369340e+00 -9.94014800e-01
-3.30246478e-01 6.24722894e-03 5.83900809e-01 3.52460474e-01
-5.67541540e-01 2.61444479e-01 -6.95290983e-01 5.21244526e-01
1.30263293e+00 -1.17669061e-01 1.28310120e+00 -1.51882625e+00
-1.56624913e-01 -2.61546195e-01 -8.39524627e-01 -1.33237040e+00
-3.08002919e-01 -4.04181987e-01 1.03050649e-01 -1.38184917e+00
5.23229361e-01 -3.69130522e-01 -1.41151864e-02 9.45524648e-02
-5.50456941e-01 3.06565702e-01 5.74711800e-01 1.09150596e-01
-1.31098616e+00 5.11106730e-01 1.23924947e+00 -2.41841763e-01
1.18992321e-01 3.05988193e-01 -4.98800755e-01 8.67046893e-01
5.99897981e-01 -4.75759834e-01 1.70074761e-01 -3.31748843e-01
-2.91713536e-01 -3.28635395e-01 7.27223635e-01 -1.69324350e+00
6.04663670e-01 1.26650572e-01 8.13623667e-01 -8.45192730e-01
5.19218862e-01 -7.10168183e-01 1.81834064e-02 7.17939615e-01
4.64193940e-01 -7.66683519e-02 3.17988485e-01 6.91933155e-01
1.45685643e-01 1.93715960e-01 1.15943658e+00 -1.46002080e-02
-1.02894354e+00 6.42302573e-01 1.26010431e-02 1.49947017e-01
1.39297986e+00 -2.18287989e-01 -4.68631774e-01 -3.35264369e-03
-2.64018357e-01 6.85468078e-01 2.41472915e-01 5.53701580e-01
3.97801131e-01 -1.29358387e+00 -8.51341426e-01 8.97409618e-02
2.59114981e-01 1.99504957e-01 1.79566815e-01 1.06997240e+00
-2.70277500e-01 5.56356430e-01 -1.16658792e-01 -9.93342757e-01
-1.65608990e+00 5.11247933e-01 3.81426722e-01 -2.13843256e-01
-6.45330548e-01 7.52505898e-01 3.76918733e-01 -4.40376885e-02
5.03733575e-01 -1.98072925e-01 -3.57066423e-01 -8.96835178e-02
1.01129651e+00 6.78780437e-01 -4.08828646e-01 -1.04856527e+00
-5.85963249e-01 6.69863760e-01 -1.83620796e-01 3.77681881e-01
8.67992997e-01 -2.75548160e-01 3.43875647e-01 -1.01922750e-01
7.44603395e-01 -6.86952323e-02 -1.48215377e+00 -3.77238095e-01
-2.00013816e-01 -7.95398653e-01 -5.20741232e-02 -3.70099515e-01
-1.23096323e+00 7.89463162e-01 1.10020411e+00 -1.33512750e-01
4.73675668e-01 5.50797880e-02 1.08815074e+00 2.05097631e-01
5.82920849e-01 -8.46383095e-01 -4.10092436e-02 3.56261760e-01
2.99502969e-01 -1.72407424e+00 1.82729915e-01 -4.30557698e-01
-2.99570769e-01 7.22868800e-01 1.12995219e+00 2.38326445e-01
5.14170527e-01 2.81210929e-01 -1.38282269e-01 -6.35789558e-02
-1.13905594e-01 -6.56802952e-01 3.41048837e-01 8.55345845e-01
1.23189799e-01 2.90733010e-01 -9.61973295e-02 4.02030587e-01
-6.96625561e-02 -1.79564103e-01 8.17669928e-02 7.54418433e-01
-8.47596884e-01 -8.36656213e-01 -1.00667918e+00 4.21024859e-01
-2.67308176e-01 3.68878871e-01 -4.71075512e-02 8.48592043e-01
6.61477208e-01 1.26378250e+00 -1.67367175e-01 -3.12488645e-01
5.30305207e-01 -5.63858688e-01 4.76744950e-01 -2.58151442e-01
-2.39009008e-01 -1.89765215e-01 -1.02871135e-01 -4.30126995e-01
-4.50289935e-01 -7.85765886e-01 -1.11733103e+00 -7.10236490e-01
-5.68073332e-01 -1.80247307e-01 5.33476353e-01 9.86307919e-01
2.15538412e-01 4.63150948e-01 1.91177443e-01 -1.11392677e+00
-4.00803387e-01 -9.08501089e-01 -3.99268210e-01 3.41857791e-01
2.97149479e-01 -1.23034465e+00 7.67967701e-02 -1.43039927e-01] | [7.910381317138672, -0.6702067255973816] |
9454efcf-7a85-4979-887b-1ebd01f0f88d | anomalous-sound-detection-based-on-machine | 2204.07353 | null | https://arxiv.org/abs/2204.07353v1 | https://arxiv.org/pdf/2204.07353v1.pdf | Anomalous Sound Detection Based on Machine Activity Detection | We have developed an unsupervised anomalous sound detection method for machine condition monitoring that utilizes an auxiliary task -- detecting when the target machine is active. First, we train a model that detects machine activity by using normal data with machine activity labels and then use the activity-detection error as the anomaly score for a given sound clip if we have access to the ground-truth activity labels in the inference phase. If these labels are not available, the anomaly score is calculated through outlier detection on the embedding vectors obtained by the activity-detection model. Solving this auxiliary task enables the model to learn the difference between the target machine sounds and similar background noise, which makes it possible to identify small deviations in the target sounds. Experimental results showed that the proposed method improves the anomaly-detection performance of the conventional method complementarily by means of an ensemble. | ['Yohei Kawaguchi', 'Masaaki Yamamoto', 'Takashi Endo', 'Kota Dohi', 'Tomoya Nishida'] | 2022-04-15 | null | null | null | null | ['activity-detection'] | ['computer-vision'] | [ 5.28795004e-01 1.13858528e-01 3.08665156e-01 1.14004305e-02
-5.99978328e-01 -1.81582451e-01 3.36405337e-01 3.92252207e-01
-2.36972213e-01 3.70154560e-01 -1.79527758e-03 -1.96200117e-01
-5.06930649e-02 -8.45954239e-01 -4.29084927e-01 -9.93265092e-01
-2.56521642e-01 1.74846575e-01 4.52015847e-01 3.24090332e-01
3.73384923e-01 4.59104121e-01 -1.85959816e+00 3.38902146e-01
5.29825568e-01 1.28449893e+00 -2.42809296e-01 1.09894836e+00
-2.27921396e-01 8.68429601e-01 -9.87138391e-01 3.32212448e-01
1.95627108e-01 -7.78888643e-01 -4.43914801e-01 -1.16814733e-01
2.61609942e-01 -3.67364734e-02 -2.90055163e-02 1.22947049e+00
2.92160571e-01 3.76942605e-01 6.30189121e-01 -1.37755978e+00
1.30214155e-01 2.94079036e-01 -1.67911261e-01 6.59374475e-01
5.48686087e-01 1.08574495e-01 9.67329025e-01 -7.90040731e-01
9.73853935e-03 7.55774438e-01 5.94519794e-01 5.01973331e-01
-1.23117709e+00 -5.18542469e-01 1.92568656e-02 3.59696746e-01
-1.25881302e+00 -3.60793024e-01 1.11299539e+00 -6.14723980e-01
8.96597922e-01 4.49797809e-01 5.68189025e-01 1.04137731e+00
2.78045721e-02 1.80935904e-01 5.78217745e-01 -5.84327042e-01
6.30772591e-01 1.68878268e-02 2.25262150e-01 7.44762838e-01
1.47011966e-01 4.29238081e-02 -5.69770396e-01 -5.39743006e-01
1.07415684e-01 1.86570719e-01 -1.59307018e-01 1.65235046e-02
-1.12507570e+00 5.11048615e-01 3.49121541e-02 8.33569944e-01
-8.16345692e-01 1.00758947e-01 4.53140318e-01 4.19205725e-01
7.18133390e-01 6.12419724e-01 -3.59861434e-01 -2.03298166e-01
-9.20416534e-01 -2.63098478e-01 1.00285757e+00 1.14707753e-01
7.30522394e-01 5.41839480e-01 -1.17521822e-01 4.00319070e-01
3.16404969e-01 4.81161594e-01 5.65841019e-01 -6.64175987e-01
2.81532764e-01 8.71338785e-01 1.10310495e-01 -1.30456007e+00
-1.77895486e-01 -4.54061180e-01 -6.48621321e-01 2.76632905e-01
5.48922241e-01 -3.05082276e-02 -6.09051168e-01 1.48564970e+00
2.65117258e-01 1.02747679e+00 2.38931216e-02 3.86532992e-01
1.65195987e-01 8.59866858e-01 3.47496457e-02 -6.35273516e-01
7.04962194e-01 -5.63682199e-01 -9.47580278e-01 -7.52676465e-03
8.01854551e-01 -3.14161003e-01 1.13667881e+00 7.41951466e-01
-4.70250845e-01 -6.57944798e-01 -1.25341630e+00 9.23328340e-01
-5.17733276e-01 -8.98151100e-03 5.22747822e-02 7.06924140e-01
-6.79610312e-01 7.64119864e-01 -1.02222025e+00 -2.75518566e-01
8.37878957e-02 8.59685093e-02 -2.45630160e-01 4.36398000e-01
-9.29523408e-01 6.65892780e-01 3.36847037e-01 4.82424945e-02
-1.10311854e+00 -4.37393546e-01 -6.67429090e-01 1.44912630e-01
3.34416062e-01 -3.69693786e-02 1.04608881e+00 -9.83761191e-01
-1.40361655e+00 4.71500635e-01 -1.67492867e-01 -6.03691101e-01
5.12767673e-01 -2.72354484e-01 -1.06885922e+00 2.28000611e-01
-1.07440010e-01 -5.15850902e-01 1.21579552e+00 -1.07069504e+00
-6.51926756e-01 -2.81524479e-01 -5.83871126e-01 -3.68792951e-01
-4.80221957e-01 -1.34731367e-01 6.87841326e-02 -3.52915138e-01
3.61629277e-01 -5.81418335e-01 9.73523110e-02 -2.14187711e-01
-6.08758330e-01 -1.54193223e-01 1.03059506e+00 -7.40243852e-01
1.73128235e+00 -2.60307693e+00 -4.69415128e-01 8.87910962e-01
1.42446738e-02 2.45697334e-01 1.01573169e-01 2.98809528e-01
-2.63434619e-01 -1.50025636e-01 -3.83423477e-01 -1.21516645e-01
-1.94728717e-01 1.20205350e-01 -5.39048135e-01 5.28994620e-01
8.81545618e-02 1.30123645e-01 -9.48685884e-01 -3.71156543e-01
2.23578677e-01 7.87381381e-02 -4.78860676e-01 6.75862908e-01
-3.24369669e-01 4.33156192e-01 -2.99283355e-01 3.76285076e-01
9.81403515e-02 2.59632587e-01 -2.69314945e-01 2.95051709e-02
-4.37074490e-02 2.61844546e-01 -1.55767095e+00 1.22899461e+00
-4.54090565e-01 5.20987749e-01 -4.37233508e-01 -1.23130131e+00
1.08115411e+00 5.84648252e-01 7.50895500e-01 -1.97562531e-01
-1.52915522e-01 3.13324600e-01 2.03597192e-02 -8.13112915e-01
-1.57722265e-01 1.39912561e-01 1.08159974e-01 5.42487621e-01
6.20204657e-02 2.77279228e-01 -6.19516820e-02 -1.60761192e-01
1.59110761e+00 -1.59007102e-01 2.75752217e-01 2.51558095e-01
9.04205322e-01 -3.55156392e-01 6.02969110e-01 9.04370070e-01
1.89373288e-02 1.38823807e-01 5.42163312e-01 -4.51695710e-01
-5.72982490e-01 -1.40932941e+00 9.71973240e-02 1.14492977e+00
-3.31238538e-01 -3.31374377e-01 -7.32946038e-01 -1.16344619e+00
-1.10014096e-01 9.31967735e-01 -7.06577599e-01 -4.94241863e-01
-4.34358001e-01 -6.62865341e-01 6.79067910e-01 4.13828790e-01
3.02691191e-01 -1.22690809e+00 -6.72352672e-01 3.22026968e-01
-1.04068868e-01 -6.07544482e-01 -6.75497577e-02 6.72299147e-01
-8.81632566e-01 -1.17099845e+00 2.48687148e-01 -5.15649378e-01
6.47030473e-01 -3.44277650e-01 7.48638749e-01 2.26361349e-01
-3.36270720e-01 4.41376835e-01 -1.91227585e-01 -6.07880533e-01
-8.29742789e-01 -4.17411447e-01 3.70203733e-01 6.79984927e-01
5.70101678e-01 -7.20119476e-01 -2.08764955e-01 1.57163590e-01
-7.60013580e-01 -8.08578491e-01 2.09699601e-01 5.62178075e-01
6.40365124e-01 4.47121769e-01 9.41703200e-01 -7.90109336e-01
4.11373079e-01 -5.36260724e-01 -4.76495266e-01 6.02011196e-02
-7.35614002e-01 1.96333691e-01 7.20941663e-01 -6.35550380e-01
-8.13252628e-01 4.11296450e-02 -1.83503240e-01 -4.09641415e-01
-5.54531276e-01 2.19848514e-01 -5.29536366e-01 3.91504079e-01
8.31792057e-01 3.19841683e-01 -1.96084306e-01 -7.46589363e-01
1.21059418e-02 8.93856645e-01 8.92013669e-01 -1.79636016e-01
9.16950166e-01 2.14419380e-01 4.87467423e-02 -1.07347989e+00
-7.92677104e-01 -7.52567887e-01 -6.05230570e-01 -3.90074909e-01
8.64915431e-01 -4.82811928e-01 -3.78705680e-01 5.49717128e-01
-1.01858521e+00 1.01235352e-01 -6.79278374e-01 6.78404391e-01
-2.86853135e-01 4.34490740e-01 -1.91475049e-01 -1.43222141e+00
-2.56938219e-01 -4.51698124e-01 6.14087939e-01 -1.21791132e-01
-4.85023260e-01 -1.00153661e+00 5.37389755e-01 -1.50583595e-01
2.79934108e-01 4.19140190e-01 1.03840065e+00 -1.63654661e+00
-1.56384110e-01 -8.19360614e-01 5.67801714e-01 9.28219378e-01
3.71635437e-01 1.49730176e-01 -1.54610717e+00 -7.82736093e-02
4.19348985e-01 2.81135976e-01 6.28154099e-01 -1.29952691e-02
1.30603004e+00 -4.88410801e-01 -2.48631179e-01 4.24650669e-01
1.16751909e+00 4.56872046e-01 4.56876785e-01 2.60680556e-01
6.17274821e-01 1.66844964e-01 6.14257634e-01 5.27188122e-01
-5.52154541e-01 5.12209296e-01 5.67887545e-01 9.51850116e-02
3.53635877e-01 -4.32580829e-01 6.83420956e-01 1.07356453e+00
-2.29129083e-02 1.19151836e-02 -7.52415538e-01 4.25224394e-01
-1.73067307e+00 -1.33841014e+00 -1.52208423e-02 2.63656878e+00
5.92057586e-01 5.36687255e-01 8.29863325e-02 1.03124630e+00
7.34356165e-01 1.29495040e-01 -4.80513245e-01 -4.37080711e-01
3.39007795e-01 3.18257898e-01 -7.00660869e-02 3.91797006e-01
-1.15548360e+00 4.32542175e-01 5.99824142e+00 3.36380661e-01
-9.89875376e-01 2.26517186e-01 9.00209248e-02 -1.25754207e-01
8.19781050e-02 -2.50915617e-01 -3.59446138e-01 7.22804904e-01
1.22212517e+00 1.08754098e-01 2.36537769e-01 9.73127365e-01
4.12156016e-01 -1.54111058e-01 -1.41731167e+00 9.33219135e-01
3.19469154e-01 -5.73183417e-01 -1.90560117e-01 -2.50012726e-02
3.77979487e-01 -2.95552939e-01 -1.39419079e-01 3.48121315e-01
-2.43945643e-01 -6.27424479e-01 3.90139222e-01 9.24880862e-01
2.99814314e-01 -7.67020762e-01 9.62950408e-01 6.59112275e-01
-1.04229248e+00 -3.87019098e-01 -8.45305026e-02 -3.09059232e-01
-1.94162980e-01 8.93175840e-01 -1.11998224e+00 1.47306457e-01
4.97991353e-01 4.09980834e-01 -5.60390711e-01 1.04831064e+00
-2.98395157e-01 1.16951358e+00 -4.88068551e-01 1.08821847e-01
-1.85194045e-01 -2.39151165e-01 8.59215260e-01 1.18346930e+00
4.88412231e-01 -5.44069469e-01 2.56896198e-01 7.37311184e-01
2.69164622e-01 1.59356058e-01 -1.01338398e+00 -6.69218823e-02
5.06545007e-01 8.79068911e-01 -5.52614093e-01 -2.90364414e-01
-3.36762041e-01 1.00937569e+00 -9.54447910e-02 3.42732817e-01
-7.08186686e-01 -5.46022356e-01 5.81383646e-01 -2.83164158e-02
2.71535292e-02 1.42648280e-01 -1.98601574e-01 -9.08256114e-01
1.88565671e-01 -6.24550641e-01 5.58059990e-01 -5.28089702e-01
-1.08276153e+00 4.94706243e-01 -3.37131888e-01 -1.62888801e+00
-6.80038095e-01 -3.65620583e-01 -1.19478881e+00 6.05629206e-01
-8.56439352e-01 -4.90666777e-01 -1.58627957e-01 6.93271458e-01
2.19446883e-01 -4.13369536e-01 1.14243245e+00 1.15179278e-01
-7.41945922e-01 5.13731182e-01 -7.52548948e-02 2.79911399e-01
4.94530261e-01 -1.39911234e+00 -4.47023287e-03 1.39124155e+00
7.58980870e-01 1.43746570e-01 8.30962658e-01 -5.46589494e-01
-8.43412101e-01 -1.18905187e+00 9.15411294e-01 -6.72099948e-01
8.35541487e-01 -2.08430350e-01 -1.21565163e+00 4.46386278e-01
-3.18753183e-01 2.38107786e-01 8.93942714e-01 -7.68494830e-02
-2.68211424e-01 -3.56176585e-01 -1.26779759e+00 1.63646728e-01
7.03865767e-01 -7.85494506e-01 -1.08317041e+00 1.01520449e-01
5.60450077e-01 1.82850897e-01 -5.58932900e-01 1.80457816e-01
2.56752461e-01 -8.66337240e-01 6.38693571e-01 -7.67030478e-01
-2.87534028e-01 -6.53472006e-01 -3.23621690e-01 -1.35791755e+00
-1.68826655e-01 -3.67348760e-01 -7.06669629e-01 1.31866300e+00
4.22158331e-01 -7.40100861e-01 5.25550842e-01 1.23545935e-03
3.56915817e-02 -4.07375008e-01 -1.10833049e+00 -8.06816220e-01
-7.51168072e-01 -8.99148881e-01 5.73700070e-01 8.36391807e-01
7.67515749e-02 1.13629691e-01 -1.61881089e-01 7.09122002e-01
4.99798805e-01 -2.07618535e-01 6.13007963e-01 -1.76489878e+00
-4.97610718e-01 -9.25872549e-02 -8.28043878e-01 -3.06879699e-01
3.05721670e-01 -7.63122022e-01 3.80329847e-01 -1.07636034e+00
-2.99461454e-01 1.80864945e-01 -1.03840232e+00 5.57068288e-01
-1.60970032e-01 3.13004702e-01 -2.07590714e-01 -7.89414570e-02
-5.45032144e-01 2.72348493e-01 3.61122012e-01 -5.22423983e-02
-5.72794080e-01 5.65420806e-01 -1.68385163e-01 1.07911551e+00
8.68554056e-01 -8.24365497e-01 -2.27033839e-01 2.74744004e-01
1.88679352e-01 -1.90161198e-01 4.97029155e-01 -1.68055737e+00
8.07962790e-02 4.55911346e-02 3.92851055e-01 -4.86166030e-01
-4.23507858e-03 -1.25324655e+00 -4.57961336e-02 6.04002893e-01
-4.84730482e-01 -1.13630155e-03 -1.10102475e-01 1.05128050e+00
-2.39758149e-01 -3.02797526e-01 6.77616239e-01 2.79161751e-01
-5.66479325e-01 -7.51296757e-03 -7.34011292e-01 -4.90027666e-02
9.78074670e-01 -2.61382043e-01 1.41221985e-01 -2.55598783e-01
-1.03970671e+00 -4.26941186e-01 -7.92841986e-03 3.54125559e-01
6.95777297e-01 -1.38217854e+00 -4.33708429e-01 7.16096699e-01
4.11092877e-01 -3.38265002e-01 -1.31765261e-01 8.95673215e-01
-1.89911485e-01 -8.23729709e-02 1.31426081e-01 -6.68590426e-01
-1.24041247e+00 5.96317112e-01 4.62719947e-01 -1.09954447e-01
-5.86166561e-01 4.80631173e-01 -2.97619611e-01 -1.04580306e-01
6.21545553e-01 -5.07726073e-01 -2.75536031e-01 -1.81816891e-02
7.59819448e-01 6.59526885e-01 3.95733327e-01 -4.55741197e-01
-4.43026423e-01 3.01587820e-01 4.12384123e-01 -1.35153994e-01
1.08173609e+00 2.82480299e-01 -2.30688304e-02 1.22585452e+00
1.14420378e+00 3.00303310e-01 -9.15548563e-01 -2.68262476e-01
4.53420430e-01 -4.38282847e-01 8.17776546e-02 -7.81169116e-01
-8.94242048e-01 8.86556089e-01 1.21348107e+00 5.94387829e-01
1.25691903e+00 -9.23039019e-02 3.79252315e-01 7.99915254e-01
1.18792206e-01 -1.24625957e+00 3.77981663e-01 5.21167278e-01
4.45360094e-01 -8.44055533e-01 -4.43228990e-01 1.56857565e-01
-3.81005973e-01 9.56159472e-01 5.24493456e-01 -4.04101200e-02
8.69383276e-01 2.77091354e-01 2.36977905e-01 -6.40170127e-02
-6.62650466e-01 -1.49648771e-01 2.83664823e-01 7.76709020e-01
-1.82976481e-02 4.17610407e-02 1.28331900e-01 5.64699829e-01
-6.09159395e-02 -2.94169456e-01 2.48859107e-01 7.50612557e-01
-8.82750928e-01 -5.75773776e-01 -4.96804029e-01 6.27081215e-01
-6.38451099e-01 1.62633732e-01 -4.87341225e-01 2.85479963e-01
3.91453177e-01 1.32886159e+00 4.46983784e-01 -7.37015486e-01
6.12131655e-01 1.04419112e+00 4.73789545e-03 -8.62042665e-01
-4.05648023e-01 -6.27410486e-02 1.98178291e-02 -6.87274456e-01
-1.12866156e-01 -5.55581927e-01 -1.25498533e+00 3.55710864e-01
-4.60728407e-01 4.52090740e-01 5.81326962e-01 1.12388551e+00
4.01644409e-02 6.47217095e-01 1.16711700e+00 -4.07851905e-01
-6.70646966e-01 -1.22156465e+00 -7.88873553e-01 6.18849039e-01
7.49578714e-01 -5.42619586e-01 -1.14388239e+00 1.63913220e-01] | [7.593717575073242, 2.4294686317443848] |
ca7504ba-34d3-4030-a39c-78b7afccf987 | saliency-based-sequential-image-attention | 1711.05165 | null | http://arxiv.org/abs/1711.05165v1 | http://arxiv.org/pdf/1711.05165v1.pdf | Saliency-based Sequential Image Attention with Multiset Prediction | Humans process visual scenes selectively and sequentially using attention.
Central to models of human visual attention is the saliency map. We propose a
hierarchical visual architecture that operates on a saliency map and uses a
novel attention mechanism to sequentially focus on salient regions and take
additional glimpses within those regions. The architecture is motivated by
human visual attention, and is used for multi-label image classification on a
novel multiset task, demonstrating that it achieves high precision and recall
while localizing objects with its attention. Unlike conventional multi-label
image classification models, the model supports multiset prediction due to a
reinforcement-learning based training process that allows for arbitrary label
permutation and multiple instances per label. | ['Sean Welleck', 'Kyunghyun Cho', 'Zheng Zhang', 'Jialin Mao'] | 2017-11-14 | saliency-based-sequential-image-attention-1 | http://papers.nips.cc/paper/7102-saliency-based-sequential-image-attention-with-multiset-prediction | http://papers.nips.cc/paper/7102-saliency-based-sequential-image-attention-with-multiset-prediction.pdf | neurips-2017-12 | ['multi-label-image-classification'] | ['computer-vision'] | [ 5.96269190e-01 1.44185781e-01 -5.53487718e-01 -4.18569833e-01
-5.40317774e-01 -4.09869790e-01 4.78396952e-01 6.25781059e-01
-4.20902491e-01 4.17597651e-01 2.23150671e-01 -1.31156698e-01
-7.68745393e-02 -3.06582510e-01 -4.66531008e-01 -3.61020535e-01
7.47407274e-03 5.25431156e-01 4.57917303e-01 -6.31227270e-02
7.48811901e-01 3.57166827e-01 -1.82687938e+00 7.63892353e-01
3.55986148e-01 5.16721725e-01 6.63726807e-01 8.58201563e-01
-3.26642096e-02 1.11275482e+00 -5.04486501e-01 -6.83004484e-02
1.44736534e-02 -5.13879359e-01 -1.15767646e+00 1.61776274e-01
8.76570880e-01 -1.90309919e-02 1.73742339e-01 1.12331223e+00
1.06394321e-01 3.22774768e-01 8.40593815e-01 -1.43669176e+00
-1.11119568e+00 3.78601775e-02 -8.24713349e-01 9.24916148e-01
4.37248170e-01 9.05345529e-02 1.58299124e+00 -1.13947344e+00
7.52345562e-01 1.46709883e+00 5.41206419e-01 5.74220955e-01
-1.69098938e+00 -3.65298748e-01 3.98889184e-01 4.28505331e-01
-9.65414345e-01 -9.98198912e-02 4.79824662e-01 -5.91006339e-01
1.19092155e+00 4.88694191e-01 8.09732914e-01 7.66862452e-01
6.79468095e-01 9.08816099e-01 1.35332036e+00 -5.50078809e-01
1.76628590e-01 2.13097557e-01 3.92399609e-01 7.08881974e-01
-1.59879699e-01 2.04994425e-01 -7.56925166e-01 -1.39843896e-01
7.74410903e-01 3.53753090e-01 1.80953201e-02 -6.79358363e-01
-1.27791667e+00 1.03305674e+00 9.52495456e-01 2.94135988e-01
-4.70068991e-01 2.47689247e-01 3.49612772e-01 9.43877622e-02
3.46954584e-01 8.70828867e-01 -2.80444503e-01 5.14544785e-01
-9.16790903e-01 5.69492504e-02 8.42964277e-02 7.86013544e-01
9.43720281e-01 -1.74221665e-01 -6.17118537e-01 7.56470978e-01
3.45112532e-01 4.24350470e-01 5.89240491e-01 -9.76323128e-01
-5.46632819e-02 5.08991718e-01 6.57985881e-02 -1.06125653e+00
-7.82244802e-01 -2.18664274e-01 -2.41512388e-01 7.15664685e-01
1.26764001e-02 4.36581463e-01 -1.12114549e+00 1.59772360e+00
9.62510109e-02 -8.60333443e-02 -2.90803403e-01 9.95881021e-01
6.99609756e-01 4.02299047e-01 8.51894975e-01 -1.14138439e-01
1.63459814e+00 -1.13855994e+00 -5.16202927e-01 -6.54451311e-01
3.04449946e-01 -4.86909598e-01 1.23389411e+00 5.33971703e-03
-1.08795512e+00 -7.11871982e-01 -9.98378754e-01 -4.66706634e-01
-5.42149127e-01 -2.89870143e-01 5.73909402e-01 1.20439276e-01
-1.45135534e+00 3.24381620e-01 -1.84853017e-01 -6.27447486e-01
6.32121801e-01 3.57768536e-01 -3.14062506e-01 1.40465572e-01
-9.19037998e-01 1.36767900e+00 4.84717190e-01 -5.19102514e-01
-1.03343797e+00 -6.55699432e-01 -9.98626828e-01 3.21693748e-01
-2.95452122e-02 -7.70938218e-01 1.46396875e+00 -1.47508121e+00
-9.12950993e-01 1.40234983e+00 -4.96239841e-01 -4.47837114e-01
-2.74463445e-01 9.33939219e-02 -3.44054908e-01 8.31517637e-01
6.48909152e-01 1.49698353e+00 1.21652174e+00 -1.43963313e+00
-9.55173492e-01 -1.75399318e-01 2.06599049e-02 5.47205687e-01
2.08988078e-02 5.05152822e-01 1.82931349e-01 -7.43579447e-01
-6.82827309e-02 -7.76003778e-01 -3.16614300e-01 -9.82375368e-02
-7.13606253e-02 -3.50263536e-01 6.11618578e-01 -2.96713769e-01
7.51609504e-01 -1.99482620e+00 8.34414810e-02 3.93335223e-02
7.03766584e-01 -1.77221984e-01 -3.05145711e-01 1.13915972e-01
-5.63188255e-01 1.12001196e-01 1.15337215e-01 -1.23416729e-01
-2.36255735e-01 -1.54866204e-01 -2.94516474e-01 3.05059284e-01
4.79611784e-01 1.20462656e+00 -1.30540359e+00 -7.33394802e-01
3.24898064e-01 2.10236832e-02 -4.04974818e-01 2.47414131e-02
-1.87181979e-01 2.18150705e-01 -6.00303076e-02 6.76295578e-01
3.25995773e-01 -1.00584471e+00 2.14511007e-01 -3.06716561e-02
-5.33241443e-02 2.09342316e-02 -5.42344153e-01 1.38470554e+00
-9.22245756e-02 6.75416231e-01 -2.37937227e-01 -5.85293353e-01
6.82047844e-01 2.66164035e-01 2.64632195e-01 -7.61238515e-01
1.25366822e-01 -2.04921380e-01 -1.31917149e-01 -3.09929818e-01
5.79771578e-01 -5.13578534e-01 -9.54054222e-02 7.88784027e-01
4.10923600e-01 2.07597073e-02 1.32879838e-01 4.88955408e-01
7.86880255e-01 -1.66262075e-01 7.98914611e-01 -7.12840021e-01
1.75311312e-01 2.63033450e-01 3.11318219e-01 9.70423937e-01
-5.82910717e-01 5.79787076e-01 2.39529252e-01 -5.83906412e-01
-1.12309361e+00 -1.14905345e+00 -9.79644060e-03 2.15018511e+00
4.91655707e-01 -2.96879798e-01 -3.80797356e-01 -9.23897088e-01
7.41802081e-02 7.32491136e-01 -1.18729448e+00 -2.10942701e-01
-1.01872429e-01 -4.21507984e-01 -8.11065435e-02 6.83605373e-01
-1.99517384e-02 -1.79587317e+00 -1.22863066e+00 8.40129033e-02
2.28526946e-02 -4.60964262e-01 -6.13571763e-01 5.76053321e-01
-4.89938140e-01 -1.21057463e+00 -6.71759725e-01 -1.29552889e+00
7.76853621e-01 7.85445631e-01 1.38425207e+00 2.63397723e-01
-4.92516369e-01 6.57541871e-01 -8.01483318e-02 -5.38682938e-01
-2.03903005e-01 -1.47199243e-01 -1.22299023e-01 -1.76924080e-01
6.04632914e-01 -7.02524558e-02 -6.13054752e-01 2.19767168e-01
-5.29495239e-01 3.89899641e-01 5.06097972e-01 1.01440954e+00
6.40759170e-01 -6.59435987e-01 8.40822339e-01 -8.43988419e-01
4.95536804e-01 -7.40142226e-01 -3.80291194e-01 4.32584643e-01
-4.27669048e-01 -6.01927526e-02 2.04141259e-01 -2.31471017e-01
-8.08212340e-01 8.49290565e-02 3.79757136e-01 -4.45309788e-01
-3.12703967e-01 9.13956463e-02 4.69175071e-01 -2.87371069e-01
7.83540606e-01 1.87698439e-01 2.74446816e-03 -9.25932080e-03
5.10116935e-01 1.82649478e-01 4.28604066e-01 1.00654505e-01
3.47348601e-01 4.77162898e-01 7.40462020e-02 -4.26020026e-01
-1.34524691e+00 -6.95866883e-01 -7.76148677e-01 -4.83117521e-01
1.28250325e+00 -8.26963425e-01 -1.12668216e+00 -2.60142535e-02
-9.90806997e-01 -3.07567179e-01 -3.24635267e-01 1.90273732e-01
-7.98011959e-01 5.10129444e-02 -6.44887507e-01 -4.88125384e-01
-3.06072891e-01 -8.48934174e-01 1.09753346e+00 4.12011147e-01
-6.70830607e-01 -1.09353983e+00 -8.29213113e-03 7.43087754e-02
2.96536356e-01 -1.02614433e-01 9.79432940e-01 -6.34075880e-01
-5.58381200e-01 2.81741619e-01 -6.18676543e-01 -2.80086190e-01
1.07653767e-01 -3.17600906e-01 -9.77646530e-01 -4.63941127e-01
-1.82646856e-01 -7.24374413e-01 1.13356662e+00 4.80011553e-01
1.07017577e+00 8.60700570e-03 -6.47311985e-01 1.46156505e-01
1.39205277e+00 2.23442197e-01 1.75970107e-01 3.69376600e-01
6.39403164e-01 7.10812986e-01 6.74933195e-01 1.29070535e-01
4.89543766e-01 3.33527774e-01 5.39411306e-01 -4.06848282e-01
-8.82610008e-02 -1.21508099e-01 -2.05574200e-01 1.04219094e-01
4.12230074e-01 1.32887885e-01 -9.40688610e-01 7.86263466e-01
-1.75313699e+00 -1.17445076e+00 -8.99594836e-03 1.96077001e+00
8.30720246e-01 2.11536720e-01 2.31605783e-01 -2.45055780e-01
1.01183105e+00 1.96380764e-01 -7.03690886e-01 -6.18310928e-01
-5.54997809e-02 2.43325755e-01 3.39092463e-01 6.81810260e-01
-1.34093821e+00 1.12959397e+00 8.99708080e+00 4.29787934e-01
-9.09382999e-01 1.48960590e-01 7.18036771e-01 -2.25709140e-01
-2.64392972e-01 -9.38831046e-02 -8.18318129e-01 1.78701192e-01
6.67043507e-01 -3.27098161e-01 2.55800992e-01 7.70246565e-01
-1.91807836e-01 -3.36531907e-01 -1.11534846e+00 8.27907324e-01
4.33966219e-01 -1.42462265e+00 3.00838739e-01 -1.59145474e-01
7.15788782e-01 -3.26010436e-02 4.17052984e-01 2.32736200e-01
6.20368361e-01 -1.24391365e+00 7.72545159e-01 5.62399983e-01
6.69418097e-01 -5.99133432e-01 8.39553922e-02 1.09891728e-01
-9.91337061e-01 -4.89915520e-01 -3.32603365e-01 -2.34310716e-01
-7.44670033e-02 -1.34446234e-01 -1.04562604e+00 -3.91671330e-01
7.51685798e-01 7.94587433e-01 -1.09233272e+00 1.15056872e+00
-3.22319537e-01 1.29029930e-01 3.93149316e-01 -1.97833434e-01
3.10860485e-01 3.97851408e-01 4.93704379e-01 1.28958631e+00
-1.96651578e-01 -2.68630702e-02 5.00600278e-01 6.84047103e-01
1.17922388e-01 2.45286226e-01 -7.20839202e-01 5.67610919e-01
5.17726183e-01 1.38047051e+00 -1.12937558e+00 -7.62824237e-01
-4.48006362e-01 1.06224537e+00 7.55321681e-01 4.15583193e-01
-6.83569133e-01 -1.70614988e-01 3.66950721e-01 -1.06581077e-01
5.69425464e-01 1.70502394e-01 -6.59841239e-01 -6.55766368e-01
-8.06908786e-01 -6.95453346e-01 7.49682605e-01 -1.25200045e+00
-1.21099782e+00 4.11228001e-01 -1.73363581e-01 -9.09617484e-01
-1.49858266e-01 -5.05268693e-01 -4.25912201e-01 9.40077782e-01
-1.31080091e+00 -1.25115180e+00 -6.49554431e-02 5.36995471e-01
6.88265622e-01 -6.72510043e-02 1.06097782e+00 -3.62873673e-01
-1.08585684e-02 5.24930537e-01 -2.26889521e-01 -3.21214527e-01
8.09361219e-01 -1.72163260e+00 3.78863633e-01 4.39553648e-01
3.73737484e-01 5.94428957e-01 6.43094003e-01 -6.50888622e-01
-3.80226761e-01 -9.31769192e-01 1.02280939e+00 -7.13707030e-01
5.34804642e-01 -3.51061910e-01 -9.70481694e-01 9.73282337e-01
6.99206412e-01 -1.06577992e-01 1.03074455e+00 3.50710213e-01
-5.46147823e-01 3.81927222e-01 -1.29724228e+00 7.22286165e-01
7.55214095e-01 -6.65453255e-01 -1.03405213e+00 4.54107493e-01
6.56542003e-01 -8.78790393e-02 -5.20117342e-01 1.58133194e-01
4.10742700e-01 -8.89595985e-01 1.08593357e+00 -8.37951124e-01
3.03142965e-01 -4.10931408e-01 7.22861290e-02 -1.38012362e+00
-1.36908937e+00 5.60834110e-02 8.24102908e-02 5.17172992e-01
4.60759252e-01 -3.88955235e-01 4.60983545e-01 3.53689373e-01
-4.77761813e-02 -3.46525013e-01 -6.76864266e-01 -3.12009752e-01
-2.44828999e-01 1.61166802e-01 2.08650351e-01 1.09601378e+00
3.25538248e-01 7.81718612e-01 -3.27248245e-01 -1.83450727e-04
6.69105828e-01 4.84042078e-01 8.40310156e-02 -1.32971704e+00
-2.23889112e-01 -6.41098797e-01 -4.24754977e-01 -7.64624417e-01
2.71534413e-01 -1.05950606e+00 1.25116691e-01 -1.45558250e+00
7.37592399e-01 -4.28429805e-02 -9.19856191e-01 9.15598333e-01
-5.68295956e-01 7.61306345e-01 5.62837303e-01 2.65220821e-01
-1.32961059e+00 2.35950261e-01 1.21879172e+00 -2.43913978e-01
2.20107101e-02 -3.34104538e-01 -1.14635742e+00 7.46992409e-01
8.22803736e-01 -5.80093741e-01 -2.66014159e-01 -1.28996775e-01
1.14597067e-01 -8.52745995e-02 6.95863068e-01 -1.00453877e+00
1.63208947e-01 -4.30835456e-01 9.15810823e-01 -6.65456235e-01
2.62583852e-01 -5.14126301e-01 -3.43065828e-01 5.38989007e-01
-8.68970633e-01 1.98038846e-01 3.46494973e-01 5.73911786e-01
5.41376555e-03 -2.44813144e-01 1.22553599e+00 -4.76059407e-01
-1.35172224e+00 -3.69370878e-02 -7.94226527e-01 -1.09286490e-03
1.29983854e+00 -1.83486775e-01 -4.31863695e-01 -1.28219664e-01
-1.22576046e+00 3.18648010e-01 6.05015039e-01 6.21811271e-01
7.70580173e-01 -1.51403439e+00 -4.80121970e-01 2.39453971e-01
5.49333692e-01 -7.23903179e-01 1.66455731e-01 4.26182717e-01
-1.76282868e-01 3.74214888e-01 -7.19550729e-01 -7.23226488e-01
-1.55513537e+00 1.29931009e+00 2.65386134e-01 -1.48195595e-01
-4.73297358e-01 9.77298856e-01 6.82002664e-01 1.92055684e-02
-1.63741484e-01 -1.08221710e-01 -6.09272599e-01 1.60411984e-01
8.16937149e-01 1.40298426e-01 -4.08196121e-01 -8.48731279e-01
-4.16258484e-01 5.38622141e-01 -4.03515071e-01 1.90166235e-02
8.06276798e-01 -3.66359711e-01 -1.93709612e-01 8.76972020e-01
9.08955514e-01 -3.91920745e-01 -1.20621562e+00 -1.67551681e-01
1.49954498e-01 -4.60824162e-01 -1.82027161e-01 -1.08319449e+00
-4.29867089e-01 7.95852244e-01 6.82375193e-01 2.95375556e-01
9.57661033e-01 3.49433213e-01 1.33839190e-01 1.10982560e-01
2.73891419e-01 -1.11234510e+00 6.27200902e-01 5.40756702e-01
1.01567924e+00 -1.53176117e+00 2.15475056e-02 -2.93936908e-01
-9.35220540e-01 8.06679964e-01 9.46276963e-01 -3.20075154e-01
5.88764906e-01 -1.80937245e-01 1.66899085e-01 -6.10468686e-01
-1.00984025e+00 -4.24522430e-01 4.47937518e-01 7.74406910e-01
3.36241931e-01 1.66939020e-01 7.53726065e-02 -5.38779004e-03
2.37583250e-01 -2.93171525e-01 3.36103201e-01 8.93631041e-01
-1.17244291e+00 -4.75284040e-01 -4.32586223e-01 6.70953453e-01
-3.67107123e-01 -3.75769377e-01 -2.01549113e-01 4.32411760e-01
1.44180804e-01 8.23410690e-01 3.56609821e-01 -3.96081358e-02
-1.19974166e-01 2.00720131e-01 5.23525238e-01 -1.13771141e+00
-8.95041466e-01 -1.70096338e-01 -4.35659081e-01 -4.52122211e-01
-3.89318168e-01 -8.17996144e-01 -1.24524927e+00 2.43533924e-01
-5.68911247e-02 7.54862791e-03 1.62743017e-01 7.18064964e-01
3.41194957e-01 6.24254167e-01 5.86344481e-01 -1.02283442e+00
-1.33353680e-01 -1.03742468e+00 -6.74551368e-01 8.08008373e-01
5.58995068e-01 -8.61669600e-01 -9.62700844e-02 1.39534861e-01] | [9.945141792297363, 1.6879757642745972] |
b41881c3-00b8-4a80-ba7f-58efcd32ada9 | deep-angiogram-trivializing-retinal-vessel | 2307.00245 | null | https://arxiv.org/abs/2307.00245v1 | https://arxiv.org/pdf/2307.00245v1.pdf | Deep Angiogram: Trivializing Retinal Vessel Segmentation | Among the research efforts to segment the retinal vasculature from fundus images, deep learning models consistently achieve superior performance. However, this data-driven approach is very sensitive to domain shifts. For fundus images, such data distribution changes can easily be caused by variations in illumination conditions as well as the presence of disease-related features such as hemorrhages and drusen. Since the source domain may not include all possible types of pathological cases, a model that can robustly recognize vessels on unseen domains is desirable but remains elusive, despite many proposed segmentation networks of ever-increasing complexity. In this work, we propose a contrastive variational auto-encoder that can filter out irrelevant features and synthesize a latent image, named deep angiogram, representing only the retinal vessels. Then segmentation can be readily accomplished by thresholding the deep angiogram. The generalizability of the synthetic network is improved by the contrastive loss that makes the model less sensitive to variations of image contrast and noisy features. Compared to baseline deep segmentation networks, our model achieves higher segmentation performance via simple thresholding. Our experiments show that the model can generate stable angiograms on different target domains, providing excellent visualization of vessels and a non-invasive, safe alternative to fluorescein angiography. | ['Ipek Oguz', 'Yuankai K. Tao', 'Jiacheng Wang', 'Xing Yao', 'Dewei Hu'] | 2023-07-01 | null | null | null | null | ['retinal-vessel-segmentation'] | ['medical'] | [ 1.84624404e-01 4.81804162e-02 -3.39643992e-02 -3.83975476e-01
-5.48385739e-01 -7.59305775e-01 1.58210173e-01 -3.39267433e-01
-2.35269234e-01 8.89963984e-01 1.18848838e-01 -2.81716138e-01
1.17342561e-01 -7.03653812e-01 -6.23216987e-01 -9.16189253e-01
3.08778703e-01 6.49038628e-02 5.16385555e-01 2.14150056e-01
8.18578973e-02 6.26036823e-01 -1.52213943e+00 1.93866193e-01
1.18577456e+00 1.00240719e+00 7.91538283e-02 5.66289306e-01
-1.07303128e-01 5.74686468e-01 -6.29704237e-01 -2.74388701e-01
6.89991295e-01 -7.02249348e-01 -7.12125301e-01 4.81970727e-01
8.30723345e-01 -6.29251957e-01 -2.72023737e-01 1.39992511e+00
5.10412097e-01 -1.29746184e-01 6.78048134e-01 -8.65891635e-01
-7.44594693e-01 1.74946234e-01 -6.62728250e-01 3.06274921e-01
-3.21422577e-01 4.45975542e-01 8.47497523e-01 -4.08038586e-01
7.60505319e-01 1.14623523e+00 2.45618507e-01 5.84543884e-01
-1.58833826e+00 -3.75746667e-01 1.47113651e-01 -6.79340586e-02
-9.85417962e-01 -3.80177647e-01 5.94102502e-01 -7.36347139e-01
2.24842504e-01 9.79472846e-02 6.46644235e-01 9.85119224e-01
2.24777192e-01 5.29798269e-01 1.36776817e+00 -1.78221270e-01
1.90813735e-01 2.72534162e-01 -8.02736431e-02 7.55616963e-01
4.19084579e-01 1.43375143e-01 9.57171693e-02 7.12575391e-02
1.30546117e+00 -1.16385102e-01 -5.24495780e-01 -4.64813322e-01
-9.88282084e-01 6.73302054e-01 5.48709035e-01 -8.04504827e-02
-3.92835885e-01 -5.14914915e-02 2.11392656e-01 1.12651125e-01
3.09937090e-01 7.11877108e-01 -3.12304944e-01 2.35074386e-01
-9.61754978e-01 -1.86793394e-02 4.96661484e-01 7.14813530e-01
6.59053445e-01 8.44874308e-02 -4.43648368e-01 8.02602947e-01
1.83600217e-01 4.01343048e-01 4.84290630e-01 -1.33590424e+00
-5.92261739e-02 5.95419049e-01 2.16045231e-01 -6.32497489e-01
-3.47698718e-01 -5.97772777e-01 -8.29124808e-01 5.73956966e-01
7.93334365e-01 -3.74919444e-01 -1.33968341e+00 1.55520666e+00
3.57052565e-01 2.07773745e-01 -1.59223437e-01 1.21117318e+00
9.11816180e-01 2.13856950e-01 2.92515904e-02 -2.93667287e-01
1.27325618e+00 -7.90826201e-01 -6.10141039e-01 -9.77909043e-02
2.42856294e-01 -7.81977355e-01 1.02341259e+00 2.46306017e-01
-1.23538280e+00 -5.63584864e-01 -7.64371693e-01 -3.16395432e-01
1.79890804e-02 4.30938564e-02 5.98611534e-01 5.94272077e-01
-1.02684093e+00 2.78247595e-01 -8.97604644e-01 -2.38239914e-01
8.75642359e-01 2.19405979e-01 -4.21500914e-02 1.30833656e-01
-7.92739153e-01 7.72761583e-01 2.89337747e-02 1.12902142e-01
-6.99741960e-01 -8.70525658e-01 -6.22690439e-01 -7.76318088e-02
2.54649282e-01 -1.02573609e+00 1.09864998e+00 -1.12263596e+00
-1.65305054e+00 1.00743806e+00 -5.43791831e-01 -6.32435381e-01
7.66463161e-01 2.36169159e-01 -1.96330830e-01 5.55173397e-01
4.91902642e-02 7.59695888e-01 1.04583240e+00 -1.07939649e+00
-6.70606375e-01 -2.17350751e-01 6.26136586e-02 -1.47954062e-01
8.03791210e-02 -8.06278437e-02 -3.21987212e-01 -4.15710181e-01
4.56884503e-02 -8.67906034e-01 -4.28282559e-01 5.51515043e-01
-5.61286867e-01 2.06478998e-01 3.87252867e-01 -3.89251411e-01
9.63332534e-01 -2.07774329e+00 -1.38014093e-01 -4.46946884e-04
6.08529270e-01 4.75506932e-01 -2.85511434e-01 -3.02175105e-01
6.97901621e-02 3.19055051e-01 -2.39111066e-01 1.25033408e-01
-3.50405902e-01 1.55818284e-01 -3.51647824e-01 4.88631815e-01
4.31555182e-01 1.01046669e+00 -7.42675841e-01 -6.68132007e-01
3.55171651e-01 3.77125263e-01 -6.31227136e-01 4.94506210e-02
-3.00858796e-01 8.66147041e-01 -5.11493862e-01 6.30058050e-01
7.81013668e-01 -7.01831520e-01 -5.75346649e-02 -3.71043712e-01
-1.82149246e-01 -4.45668437e-02 -8.66302848e-01 1.22569680e+00
-2.73377895e-01 8.80326092e-01 -5.78627661e-02 -8.68236959e-01
7.93880701e-01 1.26111254e-01 5.03579974e-01 -6.13555551e-01
3.67804974e-01 2.45893046e-01 3.85726660e-01 -6.77892804e-01
3.31225363e-03 -1.48313105e-01 5.74659646e-01 9.48342532e-02
-7.21187741e-02 -3.28478627e-02 2.92613208e-01 -1.93875551e-01
7.47076392e-01 6.76475540e-02 5.58816046e-02 -7.49107301e-02
3.44150722e-01 5.93583696e-02 7.69294977e-01 5.84801257e-01
-6.83798015e-01 8.87921929e-01 8.58646393e-01 -4.36230093e-01
-1.04410386e+00 -1.18710530e+00 -6.94921255e-01 3.39313328e-01
4.31102365e-01 1.25246227e-01 -7.56756425e-01 -5.65289736e-01
-7.00094625e-02 2.55466551e-01 -6.30855262e-01 -5.36853001e-02
-2.90076375e-01 -9.17197227e-01 3.52034420e-01 3.34679902e-01
6.84959173e-01 -6.79076374e-01 -7.05582261e-01 2.86742300e-01
-2.70457089e-01 -1.26240861e+00 -3.32376271e-01 -2.77605474e-01
-9.09217954e-01 -1.19738948e+00 -9.73594069e-01 -9.04637635e-01
8.89533818e-01 1.79051578e-01 9.36600208e-01 -2.27554515e-01
-8.07740390e-01 -6.88551590e-02 8.08652639e-02 -3.78310740e-01
-4.51837629e-01 -2.83013403e-01 -2.55384624e-01 4.87623245e-01
3.55820715e-01 -4.96062368e-01 -1.17088389e+00 3.09598923e-01
-7.57056177e-01 -1.16955683e-01 6.62085593e-01 9.17993128e-01
1.04431486e+00 5.09368554e-02 4.41192061e-01 -1.05327594e+00
4.22915995e-01 -1.19273998e-01 -9.79129732e-01 1.19896702e-01
-4.12858605e-01 7.28658168e-03 6.70178771e-01 -5.59634149e-01
-1.12459278e+00 5.65748438e-02 2.37059548e-01 -5.25693178e-01
-4.02070642e-01 1.93956941e-01 2.00558171e-01 -2.48000666e-01
9.42055166e-01 1.32994846e-01 4.66972530e-01 -3.21732134e-01
4.18769836e-01 5.66063941e-01 5.47441185e-01 -2.33384639e-01
6.30401492e-01 9.82822955e-01 2.56606460e-01 -9.86841798e-01
-1.01031101e+00 -2.04010889e-01 -5.93893468e-01 -1.76524557e-02
9.95010912e-01 -7.94924378e-01 -4.72618043e-01 4.58287954e-01
-1.04713619e+00 -3.43923658e-01 -4.27726328e-01 4.20069009e-01
-4.19568509e-01 4.45833564e-01 -5.76198399e-01 -4.49552000e-01
4.90921922e-02 -1.58455181e+00 7.71128774e-01 5.66452980e-01
1.65015068e-02 -1.09518123e+00 -1.84430867e-01 2.29326442e-01
4.55733091e-01 5.84053934e-01 1.11912811e+00 -1.12077102e-01
-1.00675631e+00 1.25232294e-01 -5.18257439e-01 7.92729139e-01
3.73919010e-01 4.56899375e-01 -1.02814126e+00 -8.14296529e-02
-2.20852554e-01 -2.00317636e-01 9.97845531e-01 1.17738891e+00
1.28385067e+00 -1.38489306e-01 -2.61424720e-01 1.07922006e+00
1.35384929e+00 1.84820920e-01 7.65770555e-01 1.29372001e-01
5.02014935e-01 7.18383670e-01 2.11855277e-01 1.45759895e-01
1.37074068e-01 3.59836340e-01 4.55873609e-01 -7.96083212e-01
-6.05644941e-01 5.68947494e-02 -7.10884854e-02 1.63882360e-01
-1.22347124e-01 -8.70595351e-02 -5.41105807e-01 6.10329390e-01
-1.50035000e+00 -8.09495389e-01 -4.01760906e-01 2.21211076e+00
8.76857460e-01 -1.41982406e-01 2.02093244e-01 -5.13028562e-01
6.60514295e-01 -1.66702494e-02 -1.01049685e+00 -1.33025721e-01
-3.26141566e-01 2.93358028e-01 6.04537368e-01 4.62884039e-01
-1.13156927e+00 8.69357586e-01 7.09346342e+00 2.10746855e-01
-1.53291535e+00 -1.86105952e-01 8.28891575e-01 -2.08818257e-01
-1.59291819e-01 -1.24480307e-01 -6.17426872e-01 7.34152317e-01
5.51052988e-01 -1.59055665e-01 8.85033682e-02 4.71540928e-01
4.62629259e-01 -2.49753539e-02 -8.80106926e-01 8.92172873e-01
-2.45975509e-01 -1.39323127e+00 1.36902556e-01 2.68020123e-01
9.03333306e-01 7.17687756e-02 4.29910809e-01 -2.10469916e-01
2.34889492e-01 -9.80277240e-01 1.75875813e-01 6.43894494e-01
9.40625548e-01 -3.29030395e-01 6.00555658e-01 -6.59222826e-02
-5.18678606e-01 -4.00366969e-02 -5.74750066e-01 3.05411398e-01
9.26740468e-02 6.75935447e-01 -7.47194648e-01 -2.55510658e-02
5.91691315e-01 7.63229251e-01 -6.03518009e-01 1.54060209e+00
-1.84411004e-01 7.04109192e-01 -1.32401347e-01 2.22145706e-01
1.17898531e-01 -4.88708586e-01 7.58951783e-01 7.24818110e-01
1.87951177e-01 -7.00209588e-02 -9.69062150e-02 1.31365800e+00
5.66686429e-02 1.13542965e-02 -4.17090386e-01 8.12678561e-02
3.88487905e-01 1.14451623e+00 -6.00793421e-01 -1.58428594e-01
-5.79308331e-01 8.22869301e-01 5.64645045e-02 9.12657320e-01
-7.43189514e-01 -3.46163005e-01 8.19381714e-01 4.04511124e-01
3.32100570e-01 8.80198404e-02 -4.10636216e-01 -1.29652405e+00
1.19476192e-01 -7.07043648e-01 1.15648694e-01 -6.98322237e-01
-1.30184639e+00 6.68206275e-01 -3.17225784e-01 -1.43789947e+00
9.37237069e-02 -7.63267398e-01 -4.84612316e-01 1.05582941e+00
-2.04752660e+00 -9.14331019e-01 -4.80825782e-01 5.53233802e-01
2.05406725e-01 -2.21550196e-01 5.06370664e-01 1.91524908e-01
-6.60345316e-01 4.75696445e-01 1.19394533e-01 1.45427912e-01
1.03759480e+00 -1.40410590e+00 1.87585466e-02 1.01158464e+00
-1.08179450e-01 6.51960671e-01 5.56016505e-01 -3.12419325e-01
-5.88593781e-01 -1.33659267e+00 4.44049597e-01 -4.36799854e-01
6.36315823e-01 -1.77589804e-02 -9.60735500e-01 6.15543127e-01
7.51414374e-02 4.49781388e-01 6.00803912e-01 -2.65126586e-01
-2.45249227e-01 -3.03196967e-01 -1.09086525e+00 8.05765629e-01
7.79261947e-01 -1.95625603e-01 -3.86827976e-01 4.47395474e-01
4.38264668e-01 -5.78417480e-01 -7.78575003e-01 1.52787209e-01
4.01662916e-01 -1.07071018e+00 8.94017994e-01 -7.78768599e-01
6.47774220e-01 -3.40623200e-01 3.20093632e-01 -1.15547550e+00
-3.91865790e-01 -8.73642981e-01 1.22033946e-01 8.92357469e-01
4.76982802e-01 -1.05233717e+00 7.53986895e-01 7.24482417e-01
1.24781644e-02 -5.40956914e-01 -7.75111198e-01 -6.20459795e-01
2.39773303e-01 2.21150577e-01 4.83167797e-01 6.82183623e-01
-5.28172255e-01 9.49801058e-02 -1.65796742e-01 2.89481431e-01
8.50400746e-01 4.30360645e-01 6.39834881e-01 -1.39757955e+00
-1.96224481e-01 -6.44528627e-01 -5.68743587e-01 -1.25843513e+00
4.02693264e-02 -8.51449907e-01 -1.19083434e-01 -1.55048203e+00
-2.79049762e-02 -3.49334538e-01 -2.58215308e-01 3.13495338e-01
-1.71830699e-01 3.78136814e-01 -1.82755262e-01 2.48241052e-01
-1.24660492e-01 1.94345370e-01 2.14115238e+00 -5.45667596e-02
-4.30270702e-01 1.21173881e-01 -9.32723463e-01 7.86466897e-01
7.51304030e-01 -2.20535874e-01 -5.62422216e-01 -4.39923376e-01
-1.30469874e-01 2.98894085e-02 6.11851871e-01 -6.18770182e-01
2.02739120e-01 -2.11639971e-01 3.53178531e-01 -1.17163733e-01
4.35579419e-02 -5.65488517e-01 -2.57697403e-01 2.44430661e-01
-2.73429334e-01 -6.82156920e-01 8.44277069e-02 6.04795218e-01
-4.94581759e-01 7.35698268e-03 1.28658962e+00 -2.86108047e-01
-4.40821111e-01 5.40301740e-01 -4.25880253e-01 3.41447324e-01
9.90658522e-01 -5.61741292e-01 -5.41984141e-01 -2.41291255e-01
-8.84771109e-01 1.41065478e-01 5.16956508e-01 1.07968912e-01
4.17804927e-01 -8.48480582e-01 -7.97948003e-01 4.29786563e-01
-1.01027869e-01 3.02818239e-01 4.11201656e-01 1.08743966e+00
-7.12025523e-01 2.31663644e-01 -3.53527337e-01 -7.82673001e-01
-1.06594288e+00 3.11617285e-01 8.85059536e-01 2.47051120e-01
-7.44042695e-01 9.68148768e-01 5.69879353e-01 1.95060045e-01
1.31734788e-01 -7.50900269e-01 -1.37414753e-01 1.52145419e-02
4.82146084e-01 1.87562346e-01 -1.96007058e-01 -3.94239157e-01
2.55925581e-03 8.24850321e-01 -2.16878802e-01 3.24704856e-01
1.05747461e+00 -2.69012004e-01 -1.06730625e-01 9.20747519e-02
1.05365217e+00 -1.58780038e-01 -1.74553216e+00 -2.66874343e-01
-5.06531656e-01 -6.07329011e-01 3.42299163e-01 -9.22051787e-01
-1.40167642e+00 1.06089389e+00 7.35459924e-01 2.85759747e-01
1.25059938e+00 -1.69454947e-01 8.85254383e-01 -2.25995928e-01
1.26062796e-01 -8.51337910e-01 -4.94742207e-02 -8.10725167e-02
5.81582367e-01 -1.44708979e+00 -1.86794028e-01 -6.64315999e-01
-7.51397550e-01 1.14114630e+00 5.57096839e-01 -2.01777071e-01
6.34161174e-01 1.11259438e-01 6.18889630e-01 -3.06755826e-02
-5.01260281e-01 -5.35751224e-01 5.29992223e-01 8.83822918e-01
1.71985343e-01 -2.60307398e-02 -1.60114944e-01 2.55360872e-01
-1.51485786e-01 1.16934150e-01 8.58600855e-01 3.47166449e-01
-3.72290701e-01 -9.40802693e-01 4.33756970e-02 6.31309271e-01
-6.28338754e-01 4.27748337e-02 -2.11230487e-01 8.15090656e-01
1.48558438e-01 7.50232816e-01 2.94173807e-01 2.99071699e-01
3.43590230e-01 -1.63157985e-01 5.87055326e-01 -6.33444250e-01
-1.27137765e-01 3.91079128e-01 -2.50127077e-01 -5.95146954e-01
-5.20208657e-01 -6.29337251e-01 -1.08294022e+00 1.42655000e-01
-2.30840817e-01 -3.72915924e-01 2.19361603e-01 5.50789893e-01
4.94068772e-01 5.91941237e-01 5.37699223e-01 -3.07654917e-01
-4.92822617e-01 -4.98262137e-01 -8.61328661e-01 5.70107877e-01
8.52086306e-01 -7.13671565e-01 -5.85180640e-01 4.69972193e-01] | [15.733181953430176, -3.916961908340454] |
fc6da0dc-fd99-41f3-a2cb-659bcb8a229c | autoregressive-visual-tracking | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Wei_Autoregressive_Visual_Tracking_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Wei_Autoregressive_Visual_Tracking_CVPR_2023_paper.pdf | Autoregressive Visual Tracking | We present ARTrack, an autoregressive framework for visual object tracking. ARTrack tackles tracking as a coordinate sequence interpretation task that estimates object trajectories progressively, where the current estimate is induced by previous states and in turn affects subsequences. This time-autoregressive approach models the sequential evolution of trajectories to keep tracing the object across frames, making it superior to existing template matching based trackers that only consider the per-frame localization accuracy. ARTrack is simple and direct, eliminating customized localization heads and post-processings. Despite its simplicity, ARTrack achieves state-of-the-art performance on prevailing benchmark datasets. | ['Yihong Gong', 'Dahu Shi', 'Yongchao Zheng', 'Yifan Bai', 'Xing Wei'] | 2023-01-01 | autoregressive-visual-tracking-1 | https://openaccess.thecvf.com/content/CVPR2023/papers/Wei_Autoregressive_Visual_Tracking_CVPR_2023_paper.pdf | https://openaccess.thecvf.com/content/CVPR2023/papers/Wei_Autoregressive_Visual_Tracking_CVPR_2023_paper.pdf | cvpr-2023-2023-2 | ['template-matching', 'visual-tracking', 'visual-object-tracking'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-4.04369712e-01 -5.17607450e-01 -2.85428375e-01 9.30602625e-02
-5.56340814e-01 -8.94800723e-01 7.69696653e-01 -2.05133572e-01
-3.13072503e-01 3.60781342e-01 1.31866097e-01 -1.15783542e-01
2.23854724e-02 -1.13331676e-01 -8.23346376e-01 -7.77345479e-01
-1.32292330e-01 5.74275315e-01 8.43252301e-01 3.70709985e-01
-7.23169744e-02 8.18107486e-01 -1.39496768e+00 -3.02127395e-02
1.50196224e-01 9.39578176e-01 1.33123323e-01 1.17688251e+00
6.37388229e-03 1.07239914e+00 -4.80956137e-01 -5.04575312e-01
1.37741998e-01 -1.75189897e-02 -3.88599634e-01 8.40593353e-02
8.66300821e-01 -7.71262795e-02 -7.81002522e-01 1.06164932e+00
1.99182659e-01 1.57142028e-01 5.17457485e-01 -1.54925394e+00
-8.34569871e-01 1.61739886e-01 -5.10438621e-01 5.34539938e-01
1.91273347e-01 4.97144014e-01 4.79801953e-01 -9.27178383e-01
9.55473185e-01 1.42400205e+00 1.15470207e+00 5.04948676e-01
-1.29104877e+00 -6.48480237e-01 7.85393119e-01 2.05574125e-01
-1.21030557e+00 -4.69805509e-01 3.42756808e-01 -9.84348238e-01
9.62080479e-01 1.52714849e-01 8.87207091e-01 1.15022302e+00
4.85970676e-01 9.19585347e-01 6.46887481e-01 -1.36289634e-02
4.43696678e-02 -2.66368181e-01 5.42232096e-01 6.86863959e-01
2.38292769e-01 5.14012098e-01 -4.50178057e-01 -2.66009659e-01
7.35160232e-01 2.21829519e-01 -1.14836819e-01 -8.75715911e-01
-1.51081240e+00 2.59350061e-01 3.42543215e-01 2.21855909e-01
-5.73434770e-01 8.29519212e-01 4.64521527e-01 6.74450919e-02
3.04990232e-01 -1.01734854e-01 -5.18049717e-01 -4.38389778e-01
-1.04541814e+00 2.98170060e-01 2.18514949e-01 1.44881546e+00
3.04749012e-01 2.70893335e-01 -7.80757010e-01 8.99864361e-02
6.38533771e-01 8.20196867e-01 3.66083294e-01 -9.60258961e-01
1.54569037e-02 3.79877657e-01 4.70531791e-01 -7.80653834e-01
-2.98755646e-01 -5.89774489e-01 -3.89136076e-01 5.34434378e-01
6.23885155e-01 -7.34300092e-02 -1.15676844e+00 1.57557631e+00
6.34814680e-01 9.82105911e-01 -2.28363395e-01 8.00378978e-01
6.81490123e-01 5.18589556e-01 4.86181110e-01 -3.72112811e-01
1.42029071e+00 -1.26076806e+00 -9.78041291e-01 1.19919643e-01
2.36798540e-01 -7.68004835e-01 5.10066748e-01 1.94232360e-01
-1.07744074e+00 -8.62429798e-01 -5.74611485e-01 1.94799870e-01
-4.85087335e-01 2.95656949e-01 3.55004340e-01 7.53730476e-01
-1.23631358e+00 5.74453831e-01 -1.47038507e+00 -1.83703169e-01
4.35564727e-01 3.75234544e-01 -2.08446980e-01 4.89876926e-01
-5.38017869e-01 8.72523725e-01 2.58145481e-01 2.34870717e-01
-1.06850898e+00 -1.27341008e+00 -7.84896612e-01 -3.00901800e-01
4.40876156e-01 -6.82506740e-01 1.55428779e+00 -4.83736575e-01
-1.58388388e+00 5.96640348e-01 -7.58536100e-01 -8.51900518e-01
7.35217869e-01 -5.97647011e-01 -5.29436529e-01 -2.79320568e-01
-7.95149282e-02 5.00292122e-01 1.24721766e+00 -1.02333963e+00
-6.91857517e-01 -3.48491430e-01 -6.33595049e-01 -2.39420265e-01
1.31382287e-01 4.83033061e-01 -1.24134874e+00 -6.60660625e-01
-1.27157152e-01 -1.34639966e+00 -4.43598926e-01 4.39827919e-01
-3.20863962e-01 -3.61342967e-01 1.30668318e+00 -4.84635562e-01
1.25836325e+00 -1.77067101e+00 1.75560638e-01 -1.77415729e-01
2.73485631e-01 6.01213992e-01 -1.95528325e-02 3.66395600e-02
-3.22768129e-02 -4.07121390e-01 2.91540325e-01 -5.43422878e-01
1.04185075e-01 -1.66441321e-01 -5.51388562e-01 9.77945745e-01
1.26303211e-01 1.37339997e+00 -1.10725129e+00 -6.91068590e-01
6.57139242e-01 7.65833259e-01 -1.17414698e-01 2.63101071e-01
-5.25918841e-01 6.25488997e-01 -4.04028773e-01 6.00134254e-01
6.59922123e-01 -4.21471357e-01 -1.31803632e-01 -3.28550041e-01
-6.06806576e-01 -3.02016437e-01 -1.11161339e+00 1.58592558e+00
3.79155204e-02 1.05111217e+00 -3.91853303e-01 -2.90246278e-01
6.02064788e-01 1.85476154e-01 8.54450703e-01 -1.69782668e-01
1.77760035e-01 -3.47815186e-01 -2.29052067e-01 -2.56689698e-01
6.52530372e-01 5.80402672e-01 9.58741158e-02 1.11886516e-01
1.07711367e-01 4.94781256e-01 9.00122747e-02 2.43022799e-01
1.08659446e+00 9.23713088e-01 3.17775786e-01 -5.57188839e-02
5.95272601e-01 3.09563786e-01 6.00146174e-01 7.93687224e-01
-6.02282047e-01 3.87264431e-01 -3.16948071e-02 -7.18624294e-01
-8.72470081e-01 -1.21371174e+00 8.80791573e-04 1.25142968e+00
4.33756351e-01 -4.33377594e-01 -4.31781411e-01 -7.65493393e-01
3.18877220e-01 5.13842762e-01 -9.07426715e-01 3.13604735e-02
-7.74370253e-01 -1.69675440e-01 5.42741895e-01 8.47339213e-01
1.24261221e-02 -1.08751559e+00 -1.02631271e+00 4.95727420e-01
3.29875827e-01 -1.23664367e+00 -9.79571164e-01 -2.04708293e-01
-5.88980913e-01 -1.30195117e+00 -8.59834433e-01 -3.26097369e-01
3.93298119e-01 3.28646898e-01 1.27215528e+00 -1.07216187e-01
-3.88618946e-01 8.93783748e-01 3.89203280e-02 -6.44465327e-01
-3.59803021e-01 -2.61633068e-01 1.95320949e-01 1.13901094e-01
3.09495658e-01 8.28417242e-02 -4.22983438e-01 4.60261196e-01
-1.32841751e-01 -3.23717088e-01 2.06755176e-01 3.70124042e-01
9.82968092e-01 -4.69520539e-01 3.85958441e-02 -5.18673778e-01
3.90530601e-02 -2.05737367e-01 -1.36055791e+00 5.78852117e-01
-3.52991372e-01 5.60240485e-02 2.54850656e-01 -1.03655326e+00
-1.08128893e+00 6.64760172e-01 2.93471217e-01 -1.45933867e+00
4.74349549e-03 -3.82317603e-01 2.62196690e-01 -3.51873010e-01
2.76192814e-01 4.12079811e-01 -5.34888245e-02 -5.65569460e-01
6.57944441e-01 -4.74821329e-02 1.13942254e+00 -1.53394893e-01
1.15647149e+00 6.69546127e-01 7.32577220e-02 -6.83531106e-01
-6.13261282e-01 -9.64096546e-01 -1.08175993e+00 -9.10969436e-01
1.06613445e+00 -8.40714812e-01 -1.41975904e+00 5.26370168e-01
-1.29512584e+00 -2.77583301e-01 -3.37916523e-01 5.76290309e-01
-6.23882711e-01 3.03255022e-01 -4.72419858e-01 -1.15284240e+00
-3.05957675e-01 -1.02055407e+00 1.35099530e+00 2.97631979e-01
-2.20621377e-01 -1.16587782e+00 4.57747191e-01 -4.05836374e-01
3.71489733e-01 3.98200095e-01 -1.38994232e-01 -4.52364057e-01
-1.03922439e+00 -2.40819827e-01 -4.43155468e-02 -4.25778776e-01
2.69050058e-02 3.86588991e-01 -7.46349871e-01 -5.23952067e-01
-3.33467066e-01 2.44545579e-01 6.28557265e-01 8.15393329e-01
9.45150673e-01 -1.00300007e-01 -1.00289309e+00 5.44992089e-01
1.27678573e+00 3.84823054e-01 4.79595900e-01 2.94280946e-01
8.44369888e-01 -1.71976145e-02 9.76438105e-01 1.96603477e-01
2.44176105e-01 1.23452353e+00 4.01149899e-01 2.80586462e-02
-4.01738852e-01 -3.05140704e-01 5.08751631e-01 4.18792427e-01
-6.98562041e-02 -4.59420793e-02 -7.92501688e-01 7.21645594e-01
-2.26322961e+00 -1.28508699e+00 -5.74960172e-01 1.97590661e+00
2.13614434e-01 -4.13895175e-02 5.51105201e-01 -6.13841474e-01
1.02628648e+00 1.64743841e-01 -8.21559608e-01 2.17874184e-01
1.02260895e-01 -1.75693482e-01 7.87487209e-01 3.78446817e-01
-1.43831778e+00 1.33829796e+00 7.49725533e+00 4.48914170e-01
-8.77005517e-01 3.26542586e-01 -6.74174362e-06 -1.58881783e-01
4.79469299e-01 -9.32663381e-02 -1.29591727e+00 6.33581758e-01
9.10690844e-01 -2.43871987e-01 5.12485579e-02 1.03570604e+00
2.36051977e-01 2.21870095e-01 -1.04364789e+00 1.08209085e+00
-1.01917565e-01 -1.62700379e+00 -2.78652310e-01 -2.75904201e-02
5.80844939e-01 4.13199902e-01 7.36812577e-02 3.02990019e-01
8.13337445e-01 -6.89859569e-01 1.16809201e+00 9.90200341e-01
3.90622795e-01 -3.32565069e-01 2.84697235e-01 1.47906259e-01
-1.94341052e+00 1.79969575e-02 -2.65061706e-01 4.62691933e-01
6.05163097e-01 1.27577916e-01 -6.18118107e-01 4.96274203e-01
8.41065347e-01 9.83908892e-01 -7.43064046e-01 1.57114983e+00
3.54239903e-02 6.11603379e-01 -2.40646839e-01 1.33230895e-01
2.38976806e-01 -1.57860503e-01 1.07769668e+00 1.60931468e+00
2.16295332e-01 -2.03408614e-01 5.56460917e-01 6.58954322e-01
2.50051290e-01 -3.18477631e-01 -5.48060358e-01 2.50034451e-01
5.52079618e-01 1.16315746e+00 -8.70859444e-01 -8.22369635e-01
-2.86370456e-01 6.99270427e-01 2.66212702e-01 2.69496948e-01
-1.31185949e+00 2.19470277e-01 1.09787583e+00 -5.81608079e-02
7.91031122e-01 -4.37556922e-01 1.99123353e-01 -9.89528477e-01
-2.03819096e-01 -5.14539778e-01 5.32976687e-01 -8.48024786e-01
-9.97266769e-01 6.54279470e-01 -2.81806160e-02 -1.61345625e+00
-3.56722027e-01 -5.96681595e-01 -4.77240711e-01 5.96212089e-01
-1.26981771e+00 -1.47755647e+00 -3.34059328e-01 6.20076060e-01
9.26656246e-01 5.73440874e-03 3.40519518e-01 8.82398337e-02
-5.92092395e-01 4.66983885e-01 8.24371446e-03 1.65390626e-01
5.41001678e-01 -1.17712390e+00 9.77138221e-01 1.20828557e+00
4.52437341e-01 7.82303810e-01 9.76239622e-01 -1.03075325e+00
-1.71615613e+00 -1.62156403e+00 4.16198969e-01 -1.13609195e+00
9.09469008e-01 -1.07041441e-01 -1.08254099e+00 1.16350722e+00
2.03411713e-01 6.44109964e-01 -2.48174220e-02 -1.49066538e-01
-4.01577652e-01 2.03206822e-01 -6.24780357e-01 5.49230039e-01
1.25657272e+00 -3.48148584e-01 -4.92225826e-01 2.08755597e-01
8.44177246e-01 -8.90133977e-01 -9.72504377e-01 2.03081757e-01
8.43451262e-01 -6.31139994e-01 1.25407088e+00 -9.45534289e-01
-6.36569083e-01 -1.01483333e+00 3.92203450e-01 -8.69253993e-01
-8.39843810e-01 -1.11942482e+00 -9.18176174e-01 1.07912600e+00
6.21484220e-03 -2.92819589e-01 9.91084993e-01 5.36705911e-01
-7.31096938e-02 -3.91605049e-01 -8.64452183e-01 -1.29030395e+00
-3.88515323e-01 -6.48929417e-01 6.13351703e-01 5.55584729e-01
-5.78176498e-01 -2.29924411e-01 -6.20519757e-01 5.92268705e-01
1.04328310e+00 2.67243743e-01 1.02733135e+00 -1.26830328e+00
-5.34787355e-03 -2.41920173e-01 -7.48498261e-01 -1.26087224e+00
3.12605292e-01 -5.69406688e-01 3.33164871e-01 -1.11381507e+00
-1.64664928e-02 -4.56164181e-02 -2.57673472e-01 3.31926316e-01
-2.45310828e-01 2.82454461e-01 5.52691281e-01 4.10571754e-01
-1.16949832e+00 4.72943395e-01 8.49426866e-01 -1.65410906e-01
-3.44371825e-01 2.87095249e-01 -1.92080513e-02 8.27994287e-01
2.73309708e-01 -8.02073658e-01 -1.31061539e-01 -1.71908692e-01
-3.85604113e-01 -9.98725295e-02 6.65761173e-01 -1.13433230e+00
8.12935293e-01 -4.60014567e-02 6.82901680e-01 -1.44414926e+00
5.39390385e-01 -1.11411214e+00 8.56606424e-01 8.54646206e-01
-2.20630825e-01 6.19006097e-01 3.96387398e-01 1.11921155e+00
2.04631165e-01 1.43498227e-01 8.56363654e-01 1.86037704e-01
-1.17640352e+00 6.90881610e-01 -5.85863173e-01 -2.28535905e-01
1.41707766e+00 -3.96374017e-01 -1.25630453e-01 7.74552524e-02
-8.33107650e-01 3.16487312e-01 5.55940390e-01 8.31875980e-01
3.98916453e-01 -1.51126671e+00 -4.52374279e-01 -2.33019859e-01
7.67423809e-02 -3.40873361e-01 1.41023144e-01 9.19604003e-01
-5.78351915e-02 6.89523757e-01 2.23336034e-02 -1.24239504e+00
-1.88987446e+00 1.23729551e+00 4.17587876e-01 -3.28903675e-01
-1.16405690e+00 5.14606059e-01 1.54010490e-01 -6.24406077e-02
3.79228264e-01 -4.31809634e-01 -3.89798164e-01 -8.28608647e-02
8.57047915e-01 6.46494448e-01 -1.69413567e-01 -1.04627466e+00
-6.40287101e-01 1.00987267e+00 2.41029877e-02 9.67362598e-02
9.98451233e-01 -2.42366821e-01 2.62299955e-01 6.36064529e-01
7.79962599e-01 -1.90812573e-01 -1.93027318e+00 -2.41437793e-01
4.40109372e-01 -5.48773706e-01 -1.15947008e-01 -4.49384600e-01
-9.84749794e-01 3.35449129e-01 8.71225119e-01 -1.27384245e-01
6.89184546e-01 1.34525418e-01 5.11245608e-01 2.74059117e-01
4.59972352e-01 -6.88551545e-01 -7.88537264e-02 5.81024766e-01
6.68436229e-01 -7.95405149e-01 5.12555651e-02 -4.42552596e-01
-6.15080595e-01 8.90724361e-01 6.10864162e-01 -2.27067769e-01
6.35031879e-01 5.13520360e-01 6.66993484e-02 -2.58060366e-01
-9.77812052e-01 -4.86024231e-01 5.66474259e-01 9.03265953e-01
1.27532214e-01 -2.68499106e-01 4.40778047e-01 9.50970650e-02
2.07983255e-01 1.43667519e-01 -1.00970000e-01 7.02425539e-01
-2.29281709e-01 -7.47557342e-01 -8.05550456e-01 1.67108793e-02
-3.16049367e-01 3.58818769e-01 -1.90647796e-01 9.71762419e-01
-9.68310833e-02 6.35819435e-01 2.85370976e-01 -2.11582892e-02
4.16200489e-01 -6.65372983e-02 4.75230515e-01 -3.44721861e-02
-7.84943163e-01 1.39651686e-01 -2.68268824e-01 -1.00071335e+00
-6.28430068e-01 -1.09906650e+00 -1.03867352e+00 -7.51193194e-03
-3.94228190e-01 1.78167075e-02 5.90515614e-01 6.69750869e-01
7.52592325e-01 7.68610239e-01 1.68937564e-01 -1.13906360e+00
-3.27371955e-01 -7.12118149e-01 5.84451668e-02 2.15013012e-01
7.03558445e-01 -9.83267963e-01 2.15355799e-01 5.72531343e-01] | [6.26374626159668, -2.1049695014953613] |
7e9d9924-a445-4e31-93c4-0971837d1ced | unsupervised-open-domain-keyphrase-generation | 2306.10755 | null | https://arxiv.org/abs/2306.10755v1 | https://arxiv.org/pdf/2306.10755v1.pdf | Unsupervised Open-domain Keyphrase Generation | In this work, we study the problem of unsupervised open-domain keyphrase generation, where the objective is a keyphrase generation model that can be built without using human-labeled data and can perform consistently across domains. To solve this problem, we propose a seq2seq model that consists of two modules, namely \textit{phraseness} and \textit{informativeness} module, both of which can be built in an unsupervised and open-domain fashion. The phraseness module generates phrases, while the informativeness module guides the generation towards those that represent the core concepts of the text. We thoroughly evaluate our proposed method using eight benchmark datasets from different domains. Results on in-domain datasets show that our approach achieves state-of-the-art results compared with existing unsupervised models, and overall narrows the gap between supervised and unsupervised methods down to about 16\%. Furthermore, we demonstrate that our model performs consistently across domains, as it overall surpasses the baselines on out-of-domain datasets. | ['Kevin Chen-Chuan Chang', 'Pritom Saha Akash', 'Lam Thanh Do'] | 2023-06-19 | null | null | null | null | ['keyphrase-generation'] | ['natural-language-processing'] | [ 4.10315096e-01 3.32474977e-01 -4.41735685e-01 -4.12466079e-02
-1.20978892e+00 -1.08203614e+00 1.12856162e+00 2.47884005e-01
-3.93798530e-01 1.05493343e+00 8.45382988e-01 -1.85480639e-01
1.02011032e-01 -7.10565388e-01 -6.85179651e-01 -3.86019796e-01
1.64270729e-01 9.27022099e-01 3.60847056e-01 -6.33394778e-01
3.61997247e-01 -3.77452493e-01 -1.36694276e+00 4.36907858e-01
1.07886755e+00 7.69827843e-01 6.72813877e-02 5.47120810e-01
-5.13370931e-01 8.64383459e-01 -6.77753508e-01 -3.72205973e-01
2.51421362e-01 -2.84009010e-01 -1.16491342e+00 9.64309555e-03
3.82928312e-01 -3.15269828e-01 -2.75852323e-01 9.39005196e-01
4.75907296e-01 4.76853475e-02 1.04574955e+00 -1.14730918e+00
-7.77473807e-01 1.12707436e+00 -4.13038462e-01 1.12739697e-01
6.03836417e-01 -1.41006500e-01 1.67206478e+00 -9.28430438e-01
9.78366196e-01 1.13403857e+00 2.21423388e-01 4.46236700e-01
-1.43302822e+00 -7.95462549e-01 3.65471900e-01 -2.15587497e-01
-1.11769402e+00 -2.81236112e-01 6.24857366e-01 -4.85766947e-01
8.84224951e-01 -1.44122407e-01 1.81895703e-01 1.38383341e+00
4.71005179e-02 1.14186609e+00 9.77607727e-01 -4.09207672e-01
2.61189699e-01 -7.86100700e-03 1.19043984e-01 2.08497718e-01
3.56439292e-01 -1.84096321e-02 -8.63063037e-01 -5.18508434e-01
5.94566464e-01 -2.60521352e-01 -1.13495305e-01 -2.42071912e-01
-1.67877257e+00 9.72935140e-01 -1.60702869e-01 1.51601478e-01
-4.49915439e-01 -3.49582024e-02 5.06644726e-01 4.36177343e-01
8.92125666e-01 9.40892756e-01 -8.25887084e-01 -2.40157902e-01
-1.15165532e+00 8.10953021e-01 1.22734785e+00 1.25810933e+00
6.10783100e-01 -3.16417843e-01 -5.16213298e-01 1.01288307e+00
-1.03069268e-01 5.70884466e-01 5.67787290e-01 -6.31020606e-01
8.11182678e-01 5.97269595e-01 3.16700965e-01 -7.24617362e-01
1.22867692e-02 -1.17187370e-02 -7.78413951e-01 -6.02385342e-01
2.42936075e-01 -5.33286333e-01 -9.73414719e-01 1.67577362e+00
8.93033370e-02 -6.44255010e-03 2.65772343e-01 4.80761439e-01
1.17391121e+00 8.92407894e-01 1.54082999e-01 -6.15175553e-02
1.49243987e+00 -1.09394217e+00 -7.44759023e-01 -4.93897676e-01
3.36259156e-01 -9.06950057e-01 1.00990307e+00 4.98068243e-01
-8.98781955e-01 -4.35548156e-01 -9.48311448e-01 -1.80926844e-01
-6.10101521e-01 2.43669510e-01 4.07283992e-01 1.08929016e-01
-8.79233479e-01 3.77293855e-01 -2.06181183e-01 -3.97267103e-01
1.01742402e-01 -3.73779386e-02 -2.54181057e-01 6.41757697e-02
-1.68784773e+00 3.95515800e-01 9.52685833e-01 -8.13169062e-01
-1.12267745e+00 -9.64692593e-01 -8.76518846e-01 -8.25576261e-02
6.56609297e-01 -8.91796470e-01 1.65693164e+00 -5.60354114e-01
-1.38320577e+00 9.44871426e-01 -2.12547109e-01 -6.48277640e-01
2.02643022e-01 -6.83868349e-01 -4.25335079e-01 2.59731859e-01
6.27238512e-01 9.84392107e-01 1.01097965e+00 -1.16071451e+00
-1.03496361e+00 1.17272481e-01 9.43652838e-02 2.00586781e-01
-5.13029754e-01 8.15795660e-02 -5.20036876e-01 -1.37953889e+00
-2.98187137e-01 -7.99160600e-01 -2.76871145e-01 -5.66016436e-01
-7.12502956e-01 -5.33079267e-01 5.13453782e-01 -5.06489575e-01
1.56256104e+00 -1.88730431e+00 1.93638220e-01 -2.29299515e-02
3.08187693e-01 6.34762459e-04 -2.68265873e-01 9.58672583e-01
-1.81074351e-01 1.50475383e-01 -2.87468880e-01 -1.30771384e-01
2.72036225e-01 1.10671259e-01 -1.13701570e+00 -2.37965390e-01
4.50957507e-01 8.51648927e-01 -1.16924226e+00 -5.11159599e-01
-3.92960519e-01 -2.14466766e-01 -5.81297219e-01 3.67913783e-01
-9.88886833e-01 2.96218306e-01 -6.88503921e-01 3.12944531e-01
5.42807937e-01 -4.79898930e-01 3.89333904e-01 1.34770572e-01
4.18777242e-02 9.12697971e-01 -1.31197643e+00 1.95723772e+00
-2.50581235e-01 2.89956272e-01 -2.58947223e-01 -8.19927573e-01
1.08985853e+00 4.56407160e-01 6.07469082e-01 -4.35448438e-01
-1.81635797e-01 2.17506409e-01 -4.35794681e-01 -1.88987385e-02
1.05678773e+00 3.57818231e-02 -7.08382249e-01 9.88129258e-01
4.98132944e-01 -4.48142499e-01 7.70445645e-01 8.08284760e-01
1.17310667e+00 6.47968352e-02 4.47646469e-01 -4.78701383e-01
4.60350513e-01 3.45976502e-01 1.80732474e-01 8.76777589e-01
3.39911193e-01 5.02810359e-01 8.28038633e-01 -2.24384189e-01
-1.12498641e+00 -1.12853456e+00 1.42460451e-01 1.60953593e+00
7.86755756e-02 -9.61716115e-01 -6.30305231e-01 -1.05556059e+00
1.64369509e-01 7.06499636e-01 -5.17585039e-01 2.40182201e-03
-2.86738873e-01 -3.95369977e-01 7.14781106e-01 4.81819719e-01
5.86729825e-01 -1.05443096e+00 7.26056620e-02 3.82941693e-01
-6.64273083e-01 -1.43057036e+00 -6.40245020e-01 2.16908947e-01
-5.74761331e-01 -9.10843313e-01 -9.48147893e-01 -9.30487514e-01
6.06192410e-01 9.23247039e-02 1.71481097e+00 -6.05798483e-01
2.01118186e-01 3.51908267e-01 -7.66832829e-01 -5.95440626e-01
-5.30875027e-01 6.49015367e-01 3.90610732e-02 -2.39370599e-01
5.80697775e-01 -5.69277048e-01 -4.85007018e-01 1.55301057e-02
-1.11434484e+00 -7.39385784e-02 6.79377496e-01 9.63221014e-01
5.85677743e-01 2.48393685e-01 6.90699875e-01 -1.36199057e+00
1.27555442e+00 -6.26183152e-01 -3.87081563e-01 9.94725972e-02
-6.75052941e-01 4.99194592e-01 8.29647541e-01 -3.71058196e-01
-1.16519630e+00 -1.41130328e-01 1.66115656e-01 1.83950394e-01
-2.45988876e-01 5.12085319e-01 1.58011094e-01 5.68098068e-01
8.56062233e-01 6.37567639e-01 -3.75615835e-01 -6.03716195e-01
8.21367204e-01 7.31782436e-01 4.68677223e-01 -1.03591549e+00
1.16790485e+00 4.81971264e-01 -5.09453833e-01 -8.17187905e-01
-1.20742297e+00 -8.10961246e-01 -5.49888015e-01 3.96125436e-01
5.59123158e-01 -1.42701054e+00 4.55480516e-02 3.60436916e-01
-1.15387213e+00 -2.15028301e-01 -4.53946501e-01 -4.06375118e-02
-4.62319434e-01 5.04371047e-01 -5.37185073e-01 -3.58493328e-01
-8.60844016e-01 -7.59107769e-01 1.51021945e+00 1.19645283e-01
-6.18610740e-01 -1.10899019e+00 3.47627223e-01 1.59694567e-01
3.82773392e-02 -3.02098207e-02 8.73493314e-01 -1.27712107e+00
-2.99186319e-01 9.58044603e-02 -1.34214222e-01 4.62021977e-01
2.24907964e-01 -4.05480504e-01 -7.03382611e-01 -1.28338650e-01
-6.13026798e-01 -8.17492425e-01 1.18038332e+00 -9.86716338e-03
9.49783444e-01 -5.57005882e-01 -3.25739771e-01 1.16503380e-01
1.03476930e+00 -7.82104358e-02 4.96528804e-01 4.46203947e-01
2.34818771e-01 4.07819003e-01 8.86541247e-01 7.34938264e-01
6.28152013e-01 2.92370021e-01 -2.24064186e-01 -9.94438976e-02
-1.77536737e-02 -8.43004704e-01 6.27108693e-01 6.39757097e-01
2.96314180e-01 -2.54926652e-01 -8.38303387e-01 9.66137588e-01
-1.84049392e+00 -9.36444819e-01 5.80384256e-03 1.76327586e+00
1.58194292e+00 4.16985601e-01 4.11251426e-01 -2.69436333e-02
3.45123589e-01 4.53604251e-01 -1.97685242e-01 -3.64121854e-01
-1.43739194e-01 7.59680390e-01 3.26225340e-01 1.89527944e-01
-1.39121449e+00 1.41109860e+00 7.04504633e+00 1.06038344e+00
-6.84923649e-01 -1.80034965e-01 2.73091167e-01 1.77380472e-01
-5.78553736e-01 8.78187791e-02 -1.25513244e+00 5.42119145e-01
7.34131873e-01 -5.39876461e-01 1.07617699e-01 9.78745818e-01
-6.68543652e-02 -1.11080013e-01 -1.27463782e+00 6.33947670e-01
5.43364473e-02 -1.33783734e+00 5.12207866e-01 -4.88302298e-02
1.20026398e+00 -1.15887009e-01 2.24965401e-02 5.90008080e-01
1.07440484e+00 -7.52543151e-01 5.62193871e-01 1.57515645e-01
9.07786071e-01 -6.57948136e-01 4.60179657e-01 5.47910213e-01
-1.27922392e+00 1.98125333e-01 -2.77899683e-01 4.63913344e-02
7.26759285e-02 8.20939362e-01 -1.21428132e+00 6.79465830e-01
5.13329744e-01 8.17336559e-01 -4.72716421e-01 4.92203116e-01
-7.93436527e-01 6.89649820e-01 -1.85222730e-01 -1.76858753e-01
3.97128552e-01 1.16544850e-01 5.75646877e-01 1.45874727e+00
1.18077090e-02 5.14305085e-02 6.98630452e-01 9.86796081e-01
-3.69771838e-01 1.13309048e-01 -6.54612660e-01 -5.90385318e-01
5.67296386e-01 1.12572706e+00 -3.35438937e-01 -6.54088080e-01
-2.87145823e-01 1.12530363e+00 8.96474719e-02 4.06205028e-01
-5.98104417e-01 -6.90278828e-01 7.56637931e-01 1.11529455e-01
5.92570484e-01 -2.29671389e-01 -1.17654786e-01 -1.41379642e+00
4.40205708e-02 -1.29480910e+00 7.69988298e-01 -5.36866128e-01
-1.56028485e+00 4.03243452e-01 2.41007566e-01 -1.23642373e+00
-6.19363546e-01 -4.72166181e-01 -3.76031458e-01 8.44102263e-01
-1.76124072e+00 -1.08609617e+00 7.60230497e-02 6.71231389e-01
8.22045326e-01 -3.15150380e-01 1.00072253e+00 -7.70896971e-02
-8.44309404e-02 4.97646660e-01 2.79220730e-01 5.30542970e-01
1.15372252e+00 -1.58936989e+00 9.98618066e-01 9.23621118e-01
3.57832313e-01 8.29502463e-01 5.86463749e-01 -9.02521670e-01
-1.03866971e+00 -1.18901455e+00 1.17950761e+00 -6.24239564e-01
1.12848639e+00 -5.22497118e-01 -4.33435142e-01 5.85209310e-01
5.87871790e-01 -4.75675941e-01 8.16532493e-01 2.08566129e-01
-7.21061826e-01 1.50884703e-01 -6.73239827e-01 6.32295132e-01
8.83331418e-01 -4.77238625e-01 -1.25019050e+00 4.18096691e-01
1.16054940e+00 -3.86726052e-01 -7.66856372e-01 2.33210653e-01
1.51838556e-01 -4.26878959e-01 9.10155237e-01 -8.62160325e-01
1.01807737e+00 -1.71219409e-01 7.77844638e-02 -1.63646710e+00
-3.41568559e-01 -1.09211576e+00 -3.97005111e-01 1.46568930e+00
8.10929954e-01 -4.01184112e-01 5.54138660e-01 -4.02448997e-02
2.52720684e-01 -2.55325437e-01 -5.15870631e-01 -7.67027318e-01
3.36947531e-01 -3.88621122e-01 3.71681422e-01 8.65923584e-01
3.08590055e-01 1.03052580e+00 -2.82712221e-01 -1.22516640e-01
6.15147114e-01 2.36214325e-01 1.06989288e+00 -1.17724240e+00
-3.09661061e-01 -2.63689876e-01 2.44257182e-01 -1.72039521e+00
3.44413459e-01 -9.33752835e-01 5.53713143e-02 -1.63188815e+00
3.19318801e-01 -2.27313235e-01 -1.52361523e-02 5.50560474e-01
-5.00191808e-01 -1.18148334e-01 -4.33051623e-02 2.04898581e-01
-8.93213928e-01 5.63873708e-01 1.18403006e+00 -3.71442050e-01
-1.71113431e-01 -1.52265146e-01 -1.33858073e+00 5.79411685e-01
5.28987288e-01 -5.10726094e-01 -5.74063957e-01 -3.10182393e-01
3.30585331e-01 -3.03963780e-01 3.44832577e-02 -7.58406103e-01
2.71461785e-01 -1.80994198e-01 1.35245621e-01 -7.19868064e-01
4.10909019e-03 -4.55032289e-01 -5.70220828e-01 6.91893697e-02
-6.51766360e-01 -1.40500277e-01 2.36832127e-01 7.32647181e-01
-4.57423359e-01 3.45129147e-02 4.73015547e-01 -3.26621354e-01
-9.61470187e-01 2.58304596e-01 -3.77574980e-01 9.22096491e-01
7.70814836e-01 2.48901859e-01 -3.39497328e-01 -5.45934260e-01
-3.48459631e-01 4.76342529e-01 7.70198554e-02 7.31836557e-01
5.27409256e-01 -1.34931731e+00 -1.05636966e+00 1.09327801e-01
6.58190727e-01 2.98942596e-01 -2.59013832e-01 2.47514918e-01
-1.38282016e-01 8.01277220e-01 3.00270438e-01 -3.11587900e-01
-7.73907125e-01 4.53363955e-01 -2.83410162e-01 -1.09718454e+00
-3.66115093e-01 8.69721174e-01 2.47137755e-01 -6.19078815e-01
2.89069176e-01 -4.32044417e-01 -2.90789992e-01 3.34326893e-01
7.78973103e-01 4.26857509e-02 -7.88621083e-02 -2.26910770e-01
-1.25006184e-01 3.31197053e-01 -6.34381115e-01 -4.47044164e-01
1.31228089e+00 1.70158356e-01 -1.71265095e-01 1.78943574e-01
8.12599123e-01 1.31280690e-01 -8.24809372e-01 -7.58104980e-01
2.50582099e-01 -4.06217575e-02 -2.26020053e-01 -1.10950303e+00
-3.48505527e-01 5.14899194e-01 -1.80698216e-01 3.04529518e-01
1.10936654e+00 2.50351846e-01 1.20691967e+00 4.80934501e-01
4.19142008e-01 -1.48025048e+00 5.89977205e-01 1.11703157e+00
7.66192257e-01 -1.10014606e+00 6.17975555e-02 -4.12191153e-01
-1.00850356e+00 1.00098193e+00 5.41688800e-01 -2.28316337e-01
6.06273890e-01 2.49321416e-01 -2.77154803e-01 -1.15162082e-01
-1.12005377e+00 -4.83517557e-01 5.14410853e-01 6.84624970e-01
5.30490100e-01 2.13494450e-02 -5.31302154e-01 1.02980661e+00
-4.98407394e-01 9.20973718e-02 3.26079607e-01 9.86514807e-01
-6.22426867e-01 -1.45774126e+00 -2.71563381e-01 2.57814318e-01
-6.34816468e-01 -5.47896683e-01 -1.02629101e+00 6.69627547e-01
-6.68487251e-02 1.05943692e+00 -5.86526394e-02 -4.47287709e-01
3.58860165e-01 3.09176236e-01 8.20266530e-02 -1.16646898e+00
-5.23695707e-01 1.08168557e-01 4.55592275e-01 -2.76170343e-01
-3.02286029e-01 -4.20261353e-01 -1.02322459e+00 -2.27497637e-01
-2.37302892e-02 5.41516423e-01 8.85710027e-03 8.47757578e-01
6.33943141e-01 4.26337838e-01 6.54585183e-01 -3.48870695e-01
-8.86758626e-01 -1.25847614e+00 -6.90540671e-01 4.84442294e-01
1.04454122e-01 -4.33217585e-01 -1.66212231e-01 5.82722306e-01] | [12.240622520446777, 8.852910041809082] |
6ee861a4-998e-4fce-a089-fc46c928dd7a | predicting-heart-disease-and-reducing-survey | 2306.00023 | null | https://arxiv.org/abs/2306.00023v1 | https://arxiv.org/pdf/2306.00023v1.pdf | Predicting Heart Disease and Reducing Survey Time Using Machine Learning Algorithms | Currently, many researchers and analysts are working toward medical diagnosis enhancement for various diseases. Heart disease is one of the common diseases that can be considered a significant cause of mortality worldwide. Early detection of heart disease significantly helps in reducing the risk of heart failure. Consequently, the Centers for Disease Control and Prevention (CDC) conducts a health-related telephone survey yearly from over 400,000 participants. However, several concerns arise regarding the reliability of the data in predicting heart disease and whether all of the survey questions are strongly related. This study aims to utilize several machine learning techniques, such as support vector machines and logistic regression, to investigate the accuracy of the CDC's heart disease survey in the United States. Furthermore, we use various feature selection methods to identify the most relevant subset of questions that can be utilized to forecast heart conditions. To reach a robust conclusion, we perform stability analysis by randomly sampling the data 300 times. The experimental results show that the survey data can be useful up to 80% in terms of predicting heart disease, which significantly improves the diagnostic process before bloodwork and tests. In addition, the amount of time spent conducting the survey can be reduced by 77% while maintaining the same level of performance. | ['Ashraf Obaidat', 'Shuxia Lu', 'Asma Yamin', 'Salahaldeen Rababa'] | 2023-05-30 | null | null | null | null | ['medical-diagnosis'] | ['medical'] | [-9.99854282e-02 -2.57939667e-01 -5.09759307e-01 -4.92833555e-01
-6.01556599e-01 -1.62685141e-01 -1.06777787e-01 6.99059129e-01
-2.21116528e-01 6.61615968e-01 2.07825676e-01 -7.43688107e-01
-2.18459517e-01 -9.13177788e-01 1.08894102e-01 -3.59277755e-01
1.01033837e-01 4.34076190e-01 3.74090038e-02 1.81238398e-01
2.14718163e-01 5.65681934e-01 -1.20510495e+00 -8.68845079e-03
1.11335337e+00 8.30136836e-01 -1.11524537e-01 4.86992657e-01
-3.31278779e-02 6.59585059e-01 -5.50025523e-01 -2.81939954e-01
-8.86806995e-02 -7.99383223e-01 -7.66238928e-01 -1.54840006e-02
-3.76964957e-02 -5.88793993e-01 4.92960885e-02 6.37328804e-01
6.02087319e-01 -1.32579535e-01 4.79475111e-01 -9.41892266e-01
-1.68167889e-01 2.83774763e-01 -3.18853855e-01 3.15827698e-01
4.92752373e-01 -9.05919969e-02 7.22460687e-01 -6.88039660e-01
2.39336267e-01 9.04394627e-01 6.46793783e-01 4.13581550e-01
-9.62614775e-01 -6.97789550e-01 -2.37041190e-01 2.57490039e-01
-1.44165015e+00 -2.71510899e-01 7.89277971e-01 -5.40241361e-01
2.90410012e-01 6.37014449e-01 7.34695435e-01 3.45394701e-01
1.34546995e-01 1.42771646e-01 1.05144358e+00 -5.42388737e-01
-1.27337803e-03 6.00173473e-01 6.24353766e-01 7.96807230e-01
4.11466241e-01 -9.74891856e-02 3.97794731e-02 -7.78514802e-01
7.57855415e-01 3.90254110e-01 -1.37723729e-01 1.49514982e-02
-1.08942163e+00 1.10865784e+00 8.71835127e-02 5.04570127e-01
-4.49272394e-01 -5.65041304e-01 3.53789657e-01 1.90024480e-01
2.63514429e-01 2.91374594e-01 -5.38199246e-01 9.46994051e-02
-7.39568353e-01 1.11661822e-01 9.36416745e-01 9.05653089e-02
5.18379390e-01 -1.38534516e-01 -1.87126905e-01 9.52369988e-01
2.52621919e-01 7.78397143e-01 3.05890262e-01 -1.07949007e+00
2.05698013e-01 9.31038082e-01 8.51149485e-02 -1.38494372e+00
-6.07698798e-01 -5.10888219e-01 -1.23231590e+00 -2.35080719e-01
5.06900072e-01 -5.12553453e-01 -3.37427109e-01 1.39255214e+00
5.01867414e-01 6.05254024e-02 -1.72441915e-01 9.24336851e-01
7.71750033e-01 5.61324835e-01 2.29266942e-01 -3.97180647e-01
1.59256160e+00 -2.47399285e-01 -5.45889020e-01 9.01433900e-02
6.45309627e-01 -6.30926430e-01 7.41813362e-01 2.71789163e-01
-7.10297048e-01 -5.53601146e-01 -4.76108640e-01 4.09121066e-01
4.32891846e-02 3.97077471e-01 6.18047237e-01 1.00786197e+00
-5.05055308e-01 4.64427024e-01 -7.43615448e-01 -5.11540592e-01
9.56170112e-02 -6.50856793e-02 -1.31642431e-01 -3.54121238e-01
-1.33164823e+00 7.06578612e-01 3.14583629e-02 -1.86693087e-01
-2.44121000e-01 -6.33255124e-01 -6.15447640e-01 2.17626959e-01
9.49461684e-02 -8.51675570e-01 7.02881753e-01 -3.55327219e-01
-7.81171381e-01 6.75137281e-01 -5.24190247e-01 -1.97976381e-01
5.31664230e-02 2.40604766e-02 -5.78972042e-01 4.40032780e-01
2.35799417e-01 1.46170601e-01 3.24306101e-01 -4.70295817e-01
-7.46405065e-01 -5.59834599e-01 -4.18462634e-01 -5.30294999e-02
-4.66805756e-01 3.61093163e-01 -2.53547549e-01 -6.01598680e-01
2.22736329e-01 -9.83779132e-01 -3.93969774e-01 -1.08256526e-01
-3.02213520e-01 -4.58083093e-01 3.30821425e-01 -1.05762005e+00
1.53399074e+00 -2.04501438e+00 -4.19150442e-01 4.49870139e-01
3.68561327e-01 1.79246619e-01 3.34754020e-01 3.75689149e-01
6.35408238e-02 4.11400467e-01 3.70794460e-02 3.96397352e-01
-5.93628228e-01 1.43619075e-01 4.38120309e-03 3.94675642e-01
2.27597848e-01 5.81555128e-01 -5.31352997e-01 -4.99985486e-01
5.21055937e-01 5.03914416e-01 -4.38633263e-01 2.48004183e-01
4.98359799e-01 6.10218108e-01 -7.63521791e-01 5.88605464e-01
2.94490278e-01 -6.74078584e-01 4.93872911e-01 -8.71123672e-02
3.82037535e-02 3.86663973e-01 -1.03848910e+00 8.51717412e-01
-1.99714806e-02 3.26497942e-01 -3.73061776e-01 -1.32577264e+00
1.20861363e+00 5.66278398e-01 7.99473703e-01 -5.54978251e-01
-6.22348934e-02 -5.24023026e-02 1.01756072e-02 -8.44951749e-01
-2.79851824e-01 -2.59937763e-01 -5.83336083e-03 4.23170656e-01
-7.84866929e-01 4.30082053e-01 2.17883632e-01 1.98412351e-02
9.24030185e-01 -8.01089346e-01 6.44906521e-01 -1.91046819e-01
6.62269354e-01 1.97660655e-01 9.15908754e-01 5.87932825e-01
-3.14990550e-01 3.58000547e-01 4.73953307e-01 -7.50404596e-01
-5.89117587e-01 -8.21997046e-01 -3.87114912e-01 6.37564361e-01
-2.16412872e-01 -1.71116129e-01 -3.56577814e-01 -4.83505756e-01
8.84466097e-02 5.68516314e-01 -2.63609320e-01 -6.86506331e-02
-2.83425629e-01 -1.06590056e+00 5.04035413e-01 4.35563207e-01
7.00346410e-01 -5.99522591e-01 -7.20730126e-01 2.19351977e-01
-6.00814581e-01 -5.33530056e-01 -7.93802664e-02 -3.47585589e-01
-1.21933675e+00 -1.45211494e+00 -9.68752980e-01 -7.37649620e-01
5.49227118e-01 3.13467473e-01 9.73273218e-01 4.46170151e-01
-4.98499513e-01 2.48394474e-01 -3.97570759e-01 -2.28623748e-01
-4.25067931e-01 1.14882968e-01 -6.35454878e-02 -1.77485332e-01
5.47453463e-01 -2.31738701e-01 -8.21999490e-01 4.07283098e-01
-3.47192734e-01 -2.61557966e-01 5.75823843e-01 5.24393916e-01
2.80275941e-01 4.11695629e-01 1.04132771e+00 -1.05824554e+00
4.77319747e-01 -6.59862697e-01 -2.16553867e-01 3.12091261e-01
-1.14568317e+00 -2.63614446e-01 2.61181235e-01 -1.98655814e-01
-7.38885999e-01 -2.55376268e-02 -2.16557011e-01 2.09772184e-01
-3.89666319e-01 7.05189288e-01 1.67259306e-01 1.47584319e-01
5.65455675e-01 1.18758261e-01 2.92180985e-01 -6.31571054e-01
-4.33405727e-01 7.93046832e-01 3.70163843e-02 -1.28651217e-01
5.36481142e-01 2.79362090e-02 1.41452700e-01 -9.96482491e-01
-7.70598173e-01 -7.72831798e-01 -9.81127098e-02 -1.97732642e-01
7.57443726e-01 -8.52492690e-01 -7.28383005e-01 3.03365350e-01
-7.71487653e-01 3.17165107e-01 3.15400928e-01 9.26607668e-01
3.34703296e-01 5.26385963e-01 -7.54565358e-01 -7.68846333e-01
-4.11323309e-01 -6.26132667e-01 4.13762808e-01 2.26226926e-01
-5.98211467e-01 -1.05660474e+00 2.19530001e-01 5.45093834e-01
4.76979047e-01 2.70864397e-01 1.44605970e+00 -5.76172769e-01
-1.65033847e-01 -4.28471357e-01 -3.52642477e-01 2.95670390e-01
4.95773196e-01 -3.87619212e-02 -4.24579412e-01 -2.20379934e-01
9.14749578e-02 1.23830818e-01 6.10706806e-01 7.97148108e-01
8.38038325e-01 -2.82797605e-01 -3.80878061e-01 1.07694060e-01
1.15710878e+00 4.87796694e-01 5.04831731e-01 -8.82059522e-03
2.68467396e-01 7.27673888e-01 5.61259806e-01 7.43461370e-01
6.38202488e-01 2.72085547e-01 -6.20662123e-02 -4.82488245e-01
3.13802510e-01 4.28487435e-02 -1.57666847e-01 6.77457869e-01
-6.20472096e-02 2.40207776e-01 -1.10119653e+00 3.18932146e-01
-1.30959213e+00 -8.54829073e-01 -4.11888301e-01 2.37695003e+00
6.67467356e-01 -1.31697103e-01 4.44112986e-01 6.07114196e-01
8.01352084e-01 -3.44100088e-01 -3.61268342e-01 -5.82397170e-02
3.64295661e-01 1.44271180e-01 1.24755926e-01 3.16337138e-01
-8.56184125e-01 -3.71121243e-02 7.05864000e+00 -2.24386156e-02
-1.19697869e+00 -3.25585246e-01 1.18839371e+00 3.50163162e-01
-3.64098437e-02 -2.06173509e-02 -6.28839791e-01 4.23933685e-01
9.47556436e-01 -3.43111157e-01 1.05141215e-01 8.71382535e-01
6.46378696e-01 -1.98490143e-01 -5.53222299e-01 7.62048900e-01
-6.83335513e-02 -1.14705777e+00 -1.66819781e-01 -7.79230744e-02
3.91450942e-01 -4.73832458e-01 -2.42876485e-01 2.57525593e-01
-2.05112681e-01 -6.85657442e-01 -1.81920990e-01 6.02821589e-01
8.10774267e-01 -6.54191196e-01 8.91499162e-01 4.65559810e-01
-1.02729535e+00 -1.33575462e-02 -1.66667789e-01 -3.73481452e-01
1.12535864e-01 1.31870687e+00 -1.02605546e+00 2.91321039e-01
7.15342760e-01 3.76181543e-01 -6.10423088e-01 9.99695420e-01
1.76845267e-01 1.22755682e+00 -2.94837087e-01 -1.16134033e-01
-3.83573353e-01 -1.08587496e-01 1.35003924e-01 8.68183553e-01
4.66869980e-01 3.12785923e-01 3.91766518e-01 5.00729978e-01
1.83178887e-01 5.65877676e-01 -5.01904130e-01 -2.71209050e-03
5.71970940e-01 9.55426931e-01 -7.50122428e-01 -5.65026700e-01
-5.11590064e-01 4.09103513e-01 -2.11403266e-01 2.16279387e-01
-7.62125254e-01 -5.22317886e-01 2.30518520e-01 5.86324215e-01
-2.06730857e-01 1.61139846e-01 -4.18219775e-01 -9.20538664e-01
-2.73093104e-01 -8.20963562e-01 6.67474568e-01 -3.21310639e-01
-1.13356042e+00 2.65704095e-01 -1.69067264e-01 -9.09814656e-01
-4.31808770e-01 -1.12363629e-01 -5.26375175e-01 1.17865872e+00
-1.40428388e+00 -4.44520116e-01 -3.75011951e-01 5.02368748e-01
1.87866241e-01 -8.86146817e-03 1.09639049e+00 5.81150830e-01
-8.29562604e-01 3.49385381e-01 -1.85896114e-01 4.09728080e-01
7.70742714e-01 -9.18604732e-01 -1.39657781e-01 4.03969944e-01
-3.75931054e-01 8.12220335e-01 3.04076225e-01 -7.89745271e-01
-9.42458212e-01 -1.02217197e+00 1.65391994e+00 -1.60629489e-02
-1.77336074e-02 5.00366986e-01 -9.34937656e-01 1.67013913e-01
-3.94641459e-01 -2.84970611e-01 1.05671036e+00 4.10402507e-01
6.31935969e-02 -3.45579803e-01 -1.35451746e+00 1.95496112e-01
1.45438954e-01 -4.43223476e-01 -4.16092038e-01 2.20506638e-01
2.27687582e-01 1.50518224e-01 -1.19071972e+00 5.03962159e-01
7.54427373e-01 -1.08654857e+00 9.48373437e-01 -6.19630396e-01
3.40401620e-01 -2.86840228e-03 5.86805381e-02 -8.93218517e-01
-5.98330021e-01 -1.06137365e-01 1.28490031e-01 1.24095190e+00
4.65988040e-01 -8.85978341e-01 8.53632331e-01 1.06397998e+00
6.03474855e-01 -8.04604828e-01 -6.36914313e-01 -4.46430326e-01
-2.22482115e-01 -2.37277493e-01 2.71174043e-01 1.15151989e+00
6.44542277e-02 4.65041757e-01 -4.08802450e-01 5.01833782e-02
5.23569286e-01 3.35504681e-01 5.78166425e-01 -1.59984303e+00
-1.30886495e-01 -5.22049032e-02 -1.17557079e-01 -5.95474243e-01
-4.42618608e-01 -5.44710755e-01 -5.66511095e-01 -1.68914139e+00
4.83801097e-01 -5.68391919e-01 -4.24039483e-01 4.15063322e-01
-5.68427444e-01 -4.59779650e-02 -1.10024154e-01 1.74955428e-01
7.91632235e-02 -9.37061757e-02 9.11838949e-01 8.66298079e-02
-3.77458125e-01 6.55736566e-01 -9.99020457e-01 6.03419840e-01
1.20868373e+00 -5.43936968e-01 -3.06336254e-01 -3.35118659e-02
2.60091852e-02 7.01534212e-01 3.12240303e-01 -8.29319835e-01
-9.62309167e-02 -3.80543888e-01 6.04101300e-01 -5.79058826e-01
-4.31995951e-02 -8.40869069e-01 4.19282049e-01 1.17148018e+00
-3.26614410e-01 3.75765413e-02 -9.46135446e-02 3.19481999e-01
-2.11223975e-01 -2.18465462e-01 7.27465272e-01 -5.52170053e-02
-1.45960823e-01 -6.33707270e-02 -7.03142583e-01 2.85409968e-02
8.76844168e-01 1.33842692e-01 -1.10490583e-01 -5.24962246e-01
-5.65917373e-01 1.82972044e-01 4.19674627e-03 -3.02479267e-02
5.39035439e-01 -1.09650517e+00 -9.47010458e-01 2.87674010e-01
-6.43910840e-02 -5.65597713e-01 3.77310157e-01 1.04775238e+00
-6.33029222e-01 6.14193141e-01 6.07205592e-02 -5.41456938e-01
-1.42075944e+00 5.63248515e-01 4.83339429e-02 -4.11907971e-01
-6.79299593e-01 1.90539867e-01 -2.56976783e-01 4.34699794e-03
3.00524324e-01 -3.90686601e-01 -4.58981007e-01 1.00387998e-01
7.22874582e-01 9.65522945e-01 -2.55133033e-01 -9.14639533e-02
-5.35082817e-01 4.97620255e-01 1.97160430e-02 1.55177847e-01
1.08952153e+00 -1.98861271e-01 -1.19944535e-01 2.82404810e-01
9.82848644e-01 1.50798574e-01 -3.46386641e-01 -1.47793278e-01
-3.74048427e-02 -2.96047211e-01 7.80252367e-02 -8.11739922e-01
-1.08122551e+00 7.77090549e-01 7.76649237e-01 5.83641946e-01
1.20264888e+00 -8.18322301e-02 8.54592741e-01 3.52676958e-01
1.99573152e-02 -6.09715819e-01 -3.62710893e-01 -6.59915656e-02
3.39187384e-01 -1.35608256e+00 -2.75490582e-02 -5.22661448e-01
-6.16512358e-01 1.04312313e+00 4.89127971e-02 1.31432544e-02
9.18474078e-01 -7.43372962e-02 2.82713056e-01 2.17233375e-02
-5.88178694e-01 1.32719949e-01 1.99267760e-01 5.44379354e-01
6.02274597e-01 2.98347712e-01 -7.83528924e-01 6.13157213e-01
3.29517812e-01 3.35653514e-01 1.38741076e-01 6.74930871e-01
-7.68897235e-01 -1.01804280e+00 -5.55755377e-01 1.08185720e+00
-7.49312043e-01 -6.62014540e-03 -3.64091396e-01 4.89353001e-01
-1.51285544e-01 1.25095034e+00 -2.88656831e-01 -2.47853816e-01
2.94772714e-01 4.37851012e-01 -1.06298819e-01 -5.02488732e-01
-3.17054570e-01 5.08747064e-02 2.43678182e-01 -1.55216619e-01
-5.37712157e-01 -5.94628990e-01 -1.19085872e+00 -5.86425185e-01
-2.23983943e-01 4.36562657e-01 4.25770700e-01 9.12064135e-01
3.60570878e-01 3.96226257e-01 1.03078318e+00 9.80651900e-02
-6.79368138e-01 -8.85076702e-01 -7.07007885e-01 3.19261700e-01
1.07694238e-01 -4.59913641e-01 -3.58029664e-01 9.50857103e-02] | [8.450149536132812, 4.901702880859375] |
e3b77a71-8f2d-4429-b1fd-d7f95a6a9bfa | not-all-claims-are-created-equal-choosing-the | 1911.03850 | null | https://arxiv.org/abs/1911.03850v3 | https://arxiv.org/pdf/1911.03850v3.pdf | Not All Claims are Created Equal: Choosing the Right Statistical Approach to Assess Hypotheses | Empirical research in Natural Language Processing (NLP) has adopted a narrow set of principles for assessing hypotheses, relying mainly on p-value computation, which suffers from several known issues. While alternative proposals have been well-debated and adopted in other fields, they remain rarely discussed or used within the NLP community. We address this gap by contrasting various hypothesis assessment techniques, especially those not commonly used in the field (such as evaluations based on Bayesian inference). Since these statistical techniques differ in the hypotheses they can support, we argue that practitioners should first decide their target hypothesis before choosing an assessment method. This is crucial because common fallacies, misconceptions, and misinterpretation surrounding hypothesis assessment methods often stem from a discrepancy between what one would like to claim versus what the method used actually assesses. Our survey reveals that these issues are omnipresent in the NLP research community. As a step forward, we provide best practices and guidelines tailored to NLP research, as well as an easy-to-use package called 'HyBayes' for Bayesian assessment of hypotheses, complementing existing tools. | ['Erfan Sadeqi Azer', 'Ashish Sabharwal', 'Dan Roth', 'Daniel Khashabi'] | 2019-11-10 | not-all-claims-are-created-equal-choosing-the-1 | https://aclanthology.org/2020.acl-main.506 | https://aclanthology.org/2020.acl-main.506.pdf | acl-2020-6 | ['misconceptions'] | ['miscellaneous'] | [ 3.01815987e-01 2.32144281e-01 -3.52180451e-01 -6.15011752e-01
-7.28111148e-01 -8.41131806e-01 6.01812184e-01 8.62631857e-01
-7.71314681e-01 7.77345896e-01 5.18369198e-01 -9.17988718e-01
-2.81970650e-01 -6.25040293e-01 -3.99185777e-01 -6.73492312e-01
5.09885848e-01 4.66623098e-01 2.31495872e-01 2.20916439e-02
9.51074779e-01 2.79917210e-01 -1.64324522e+00 3.93129364e-02
7.72997737e-01 3.16555679e-01 4.86939624e-02 3.17006230e-01
-4.23813909e-01 4.94919300e-01 -6.08440816e-01 -7.52507865e-01
-1.67150766e-01 -5.46834469e-01 -9.78767157e-01 -4.05678719e-01
5.95128573e-02 -9.83372182e-02 3.51745754e-01 1.09013128e+00
5.85564613e-01 -6.75852597e-02 7.06887841e-01 -1.08039141e+00
-3.97753119e-01 8.80152166e-01 -3.15511972e-01 3.91909033e-01
6.11987114e-01 1.65155843e-01 1.13258135e+00 -6.85173213e-01
5.08125365e-01 1.57299936e+00 6.63374364e-01 1.83549613e-01
-1.22643852e+00 -6.27052546e-01 6.61366582e-02 2.82357782e-01
-1.22115636e+00 -5.14238775e-01 3.24986339e-01 -6.34428561e-01
6.10466838e-01 2.81135768e-01 4.92839456e-01 1.17750227e+00
3.58639449e-01 4.95736837e-01 1.71005583e+00 -7.90450692e-01
5.40308237e-01 3.97223681e-01 3.38050872e-01 1.74884826e-01
7.46249259e-01 6.41357079e-02 -7.14055240e-01 -5.30378699e-01
4.37036246e-01 -5.50319493e-01 -2.43002728e-01 8.23901519e-02
-1.21428823e+00 9.88637209e-01 -3.91945124e-01 6.27307177e-01
-6.14233673e-01 -2.75645792e-01 4.81483370e-01 1.21797301e-01
4.51655060e-01 3.05077344e-01 -6.72559023e-01 -6.47324502e-01
-1.00126290e+00 5.63324153e-01 1.04221869e+00 2.38862947e-01
3.76018643e-01 -4.11082506e-01 -5.64383417e-02 8.82425249e-01
9.21612144e-01 2.81768411e-01 4.63098615e-01 -9.40815032e-01
1.58100352e-01 3.52028877e-01 2.69325793e-01 -1.26371312e+00
-1.75704658e-01 -1.40672371e-01 -2.51721025e-01 1.56164080e-01
6.95404410e-01 -1.69069842e-01 -2.95170844e-01 1.55403650e+00
3.44850451e-01 -4.96590108e-01 -8.92606080e-02 7.85573781e-01
4.95470673e-01 4.40614909e-01 5.67359805e-01 -5.62023163e-01
1.56498837e+00 -1.47169501e-01 -8.54934752e-01 -3.28136176e-01
5.00043154e-01 -8.87782812e-01 1.31331909e+00 7.66780853e-01
-1.08234346e+00 -2.09158763e-01 -7.80799627e-01 -6.26607239e-02
-3.10235769e-01 -4.35299277e-01 6.28815234e-01 1.09668493e+00
-7.83768713e-01 4.68886912e-01 -6.31492376e-01 -7.11282015e-01
9.54673067e-02 -4.10685129e-02 -1.27834365e-01 9.47149321e-02
-1.08212757e+00 1.31433463e+00 6.14054739e-01 2.24660948e-01
-2.15959772e-01 -4.65960205e-01 -5.64301550e-01 -9.31435674e-02
6.32306993e-01 -5.31394362e-01 1.37724280e+00 -8.22813869e-01
-1.51965511e+00 9.54618335e-01 -1.63198650e-01 -1.64101794e-01
4.17993367e-01 -1.77145824e-01 -3.45152050e-01 1.44010276e-01
2.61368573e-01 3.82387608e-01 3.13944489e-01 -1.19369221e+00
-7.26008177e-01 -4.75412190e-01 3.70986620e-03 2.18491957e-01
9.47976187e-02 7.81832278e-01 1.81157012e-02 -4.95876938e-01
3.53362352e-01 -5.25195599e-01 -1.36553600e-01 3.21636721e-02
-7.46870711e-02 -5.04549861e-01 2.44813725e-01 -5.96425772e-01
1.43518364e+00 -1.83641016e+00 -5.07457733e-01 1.52637111e-02
4.48744372e-02 9.12875533e-02 2.46554881e-01 7.58895874e-01
1.69354483e-01 6.55192852e-01 -3.76196593e-01 2.48577148e-01
4.19557571e-01 2.38499045e-01 -4.53890711e-01 5.89371622e-01
6.05272762e-02 5.51512480e-01 -1.20567179e+00 -7.61065781e-01
2.05076158e-01 4.52467471e-01 -2.69857347e-01 6.49021938e-02
4.21218984e-02 1.06453642e-01 -3.44465792e-01 6.18363082e-01
6.63246691e-01 2.59666536e-02 5.78583121e-01 1.56887650e-01
-7.19318449e-01 1.05832922e+00 -1.15469027e+00 1.07385445e+00
1.15571849e-01 5.50745487e-01 3.66217494e-02 -1.00019956e+00
9.60151315e-01 3.87205392e-01 -6.25023693e-02 -1.26157984e-01
1.47228837e-01 3.63251120e-01 4.92938191e-01 -6.78915143e-01
3.26705337e-01 -6.41811669e-01 1.63762078e-01 5.47910690e-01
-1.50762156e-01 -2.99823195e-01 2.46920824e-01 -1.41208827e-01
8.93728137e-01 3.87343466e-01 8.59256446e-01 -5.61808884e-01
2.40146413e-01 5.71017601e-02 8.05975318e-01 9.44865167e-01
-4.11076307e-01 1.96477786e-01 8.36929917e-01 -1.46002367e-01
-7.58075237e-01 -1.01311338e+00 -5.65579712e-01 8.72060001e-01
-3.94501746e-01 -4.24505621e-01 -5.87591231e-01 -6.39445662e-01
-2.85371304e-01 1.24690163e+00 -5.02128541e-01 2.25945100e-01
-9.68183801e-02 -1.01380372e+00 5.33172131e-01 2.33252421e-01
1.39467761e-01 -1.13345742e+00 -1.17399359e+00 3.74671757e-01
-2.86958218e-01 -7.92270780e-01 3.47537041e-01 7.59931803e-02
-8.28315139e-01 -1.07077086e+00 -3.76019537e-01 -2.87957281e-01
4.19190049e-01 3.11203450e-01 1.05721140e+00 9.29110572e-02
2.94586957e-01 3.98084968e-01 -6.56373858e-01 -7.83138037e-01
-7.72316873e-01 -3.77498448e-01 -2.17253156e-02 -4.62620318e-01
1.14387751e+00 -4.80192900e-01 -1.97781801e-01 7.88332596e-02
-1.04983807e+00 -2.59088755e-01 8.09996784e-01 6.28411293e-01
2.11925358e-01 2.28344854e-02 5.99551082e-01 -9.55883980e-01
9.65355873e-01 -5.46825469e-01 -4.82853800e-01 2.86317348e-01
-9.23998773e-01 -1.65281191e-01 8.63958150e-02 -1.50842890e-01
-1.06780910e+00 -8.06334376e-01 -4.74437386e-01 3.37045938e-01
-7.31537521e-01 1.12874055e+00 -7.42837787e-02 2.26064831e-01
7.37238169e-01 -1.32842809e-01 1.59290537e-01 -3.20704341e-01
4.52624857e-02 6.82898700e-01 8.91804919e-02 -8.14281285e-01
5.12559891e-01 1.30940303e-01 -3.36649090e-01 -9.09783959e-01
-8.72139513e-01 -3.37453693e-01 -4.71185029e-01 -4.55775827e-01
5.75542688e-01 -4.04263228e-01 -7.81730771e-01 -1.25705272e-01
-1.19392061e+00 -2.60267138e-01 -1.06851168e-01 8.62550616e-01
-3.22186470e-01 7.74181187e-01 -1.19297571e-01 -1.15028334e+00
-6.33569434e-02 -1.05114734e+00 6.79596782e-01 9.56285745e-02
-8.20232332e-01 -1.03266716e+00 3.43254238e-01 5.48885047e-01
3.59205186e-01 1.59478083e-01 1.09174728e+00 -9.64093268e-01
1.94236964e-01 -9.59571078e-02 -7.91187510e-02 1.70632124e-01
-8.10317229e-03 5.19903302e-01 -8.36572051e-01 1.56784087e-01
4.74161237e-01 -4.26696211e-01 3.43519121e-01 4.18740779e-01
8.30280066e-01 -5.56202948e-01 -1.25727296e-01 -3.56822103e-01
1.39494312e+00 5.10335751e-02 5.52024066e-01 4.27825093e-01
-5.43284193e-02 1.35568202e+00 7.13822484e-01 4.21369821e-01
4.32499081e-01 5.11921644e-01 2.48270128e-02 5.95356107e-01
4.94323343e-01 -2.61509895e-01 6.04938328e-01 8.23206544e-01
1.14530295e-01 -2.64682651e-01 -1.34427941e+00 5.05202174e-01
-1.81482029e+00 -9.38274622e-01 -3.07214707e-01 2.43422532e+00
9.27933276e-01 5.88601351e-01 2.07287222e-01 2.71465838e-01
6.19617641e-01 8.06479529e-02 8.13133419e-02 -6.80896103e-01
-1.69305071e-01 7.16945380e-02 -6.75110286e-03 5.49818158e-01
-5.82514405e-01 8.29100370e-01 7.15308571e+00 7.18349159e-01
-8.80703151e-01 7.50844181e-02 4.91783321e-01 4.50583935e-01
-5.70096135e-01 5.11925161e-01 -7.13663340e-01 5.45876384e-01
1.25461805e+00 -3.35420072e-01 -7.94266015e-02 4.58140194e-01
7.12878525e-01 -6.01065516e-01 -7.98586547e-01 4.42968577e-01
-8.32073465e-02 -8.83675814e-01 -1.97656289e-01 2.30885983e-01
5.88602488e-05 -1.13666281e-01 -2.83921927e-01 2.60636479e-01
4.32519287e-01 -9.01805699e-01 8.42409790e-01 2.94142962e-01
1.19482875e-01 -4.27102596e-01 1.05623877e+00 2.54567921e-01
-3.91673386e-01 2.51475304e-01 -6.69473171e-01 -5.56917667e-01
2.63803840e-01 1.01517010e+00 -1.03733182e+00 2.88505495e-01
7.29980350e-01 1.36040151e-01 -3.43771935e-01 1.03114641e+00
-6.84995413e-01 1.25512505e+00 -3.06796700e-01 -3.27530712e-01
1.80586457e-01 -3.34661186e-01 5.55435836e-01 1.37513912e+00
7.93324262e-02 1.28955260e-01 -1.51389837e-01 8.56713295e-01
4.83847022e-01 5.64830601e-01 -6.95157886e-01 -5.38833737e-01
8.18480968e-01 1.17750311e+00 -1.12337899e+00 -1.98608503e-01
-6.88243806e-01 2.34101698e-01 2.07157731e-01 1.57810822e-01
-5.59401095e-01 -1.18274160e-01 3.74314874e-01 4.32821028e-02
1.14910208e-01 -1.69998392e-01 -5.49784601e-01 -9.17378008e-01
9.82149765e-02 -1.23644662e+00 3.20096642e-01 -6.57166958e-01
-1.30972064e+00 9.35183540e-02 6.27727270e-01 -7.11869776e-01
-2.06940755e-01 -8.08411181e-01 -4.32416171e-01 7.66888976e-01
-1.12290490e+00 -6.23622477e-01 2.36481696e-01 -2.37324908e-01
2.80089498e-01 4.59652901e-01 7.95129538e-01 -1.16039760e-01
-4.07107621e-01 1.25084981e-01 -2.03459948e-01 -2.49951541e-01
8.92512143e-01 -1.16715443e+00 1.15094386e-01 8.31740856e-01
3.11025977e-02 9.62991178e-01 1.22937858e+00 -6.85994923e-01
-1.10422790e+00 -4.78139706e-02 1.45035172e+00 -4.21162456e-01
9.26473916e-01 -8.43515322e-02 -1.10849571e+00 4.92904723e-01
3.37291867e-01 -6.56742275e-01 1.27236223e+00 5.34165919e-01
-3.76707673e-01 3.20814371e-01 -1.19579601e+00 9.51645672e-01
5.24648488e-01 -3.52764666e-01 -1.30641401e+00 1.99414566e-01
3.44003707e-01 7.90704042e-02 -9.87118721e-01 4.12479699e-01
8.06115031e-01 -1.13247669e+00 4.30827826e-01 -4.89408642e-01
1.96353927e-01 -2.13461265e-01 -2.46174887e-01 -9.28752422e-01
-2.04975903e-01 -3.80205423e-01 5.84787190e-01 1.40612388e+00
4.80775297e-01 -8.31715345e-01 5.41704237e-01 9.79631126e-01
-4.61528497e-03 -7.06569195e-01 -9.16227102e-01 -2.64520913e-01
4.06959802e-01 -8.67757618e-01 2.32213199e-01 1.10656869e+00
3.44692111e-01 2.78357059e-01 2.32843757e-01 1.71940308e-02
4.34610814e-01 -2.84173369e-01 7.79422998e-01 -1.49311650e+00
-1.11061536e-01 -6.77961409e-01 -1.74021691e-01 -6.25820994e-01
5.13916872e-02 -4.16432798e-01 2.60015160e-01 -1.54881144e+00
3.35448563e-01 -9.71452296e-02 -9.98095348e-02 4.83990043e-01
-2.30936229e-01 -1.90384671e-01 -2.15451904e-02 1.53516755e-01
-1.31174892e-01 3.25073838e-01 7.08979666e-01 4.42488134e-01
-4.76043150e-02 -2.67439753e-01 -1.28719532e+00 1.09580946e+00
8.57310057e-01 -8.92361581e-01 -1.90270215e-01 -7.94024915e-02
8.74435306e-01 -2.37817749e-01 3.80019724e-01 -6.66055918e-01
1.37399599e-01 -5.93641877e-01 -7.84936547e-02 -5.44272065e-01
-1.96976215e-01 -4.92456943e-01 1.24081746e-01 3.90832573e-01
-4.31907564e-01 2.58975267e-01 3.79766077e-01 2.61961430e-01
-1.63825348e-01 -8.62394750e-01 4.09636229e-01 -2.74271548e-01
-2.48399556e-01 -2.60697633e-01 -8.70181501e-01 4.71953005e-02
6.61234379e-01 -3.17247927e-01 -2.92005092e-01 -3.59218746e-01
-3.46110612e-01 -6.35931874e-03 6.08655572e-01 8.27427879e-02
5.32364726e-01 -6.28939629e-01 -1.02287555e+00 -4.96542305e-01
-3.59220393e-02 -3.02357346e-01 9.21172947e-02 1.30937529e+00
-5.63310683e-01 5.65782905e-01 -7.14901388e-02 -1.87120035e-01
-8.38473737e-01 4.91187751e-01 -2.44945616e-01 -8.70635808e-02
-5.01264095e-01 4.96404141e-01 -7.34196901e-02 -4.27797794e-01
1.21686958e-01 -2.86123455e-01 -3.52426082e-01 4.06361073e-01
5.67003846e-01 4.79282975e-01 -9.34866741e-02 -4.60223377e-01
-4.76737946e-01 2.90467832e-02 -1.87738702e-01 -7.48661995e-01
1.14071023e+00 -2.05359802e-01 -6.31786525e-01 1.11419785e+00
5.07642388e-01 3.44885051e-01 -6.28111362e-01 -2.88841456e-05
3.64414811e-01 -5.02540886e-01 6.52705655e-02 -9.68506277e-01
-6.08838070e-03 8.05335462e-01 8.54939595e-02 6.13997281e-01
7.09935427e-01 -2.03729123e-01 -3.06741130e-02 3.98983330e-01
4.45903510e-01 -1.29840791e+00 -4.13710862e-01 2.38695920e-01
1.00754738e+00 -1.03750360e+00 3.37491125e-01 -4.84701365e-01
-5.53090334e-01 1.13657629e+00 1.68208569e-01 2.57593483e-01
6.30670667e-01 -7.81899877e-03 2.37434909e-01 -2.59035766e-01
-9.37191069e-01 -3.39001939e-02 -1.06053643e-01 4.89831716e-01
1.10856259e+00 -4.68953587e-02 -1.57335234e+00 5.82867146e-01
-3.70708674e-01 5.72103374e-02 5.72404921e-01 9.77925777e-01
-4.81423557e-01 -1.35122120e+00 -7.74768293e-01 4.62783843e-01
-8.85852337e-01 -2.18059525e-01 -5.86969554e-01 1.04196823e+00
-2.55760159e-02 1.30602610e+00 -1.70921624e-01 -4.02530981e-03
-2.04716608e-01 2.82552958e-01 4.13940907e-01 -6.79263771e-01
-7.11156189e-01 1.82386875e-01 4.84794378e-01 -2.32673109e-01
-8.93816411e-01 -1.18918526e+00 -7.04442382e-01 -2.76719868e-01
-3.97083551e-01 6.10832334e-01 7.72550166e-01 1.33897042e+00
7.31560737e-02 4.54708328e-03 -1.15413286e-01 -4.10696745e-01
-7.92081833e-01 -1.15201306e+00 -4.34855074e-01 -1.67082340e-01
-2.08905116e-01 -7.99756408e-01 -5.40536821e-01 -2.70455301e-01] | [10.07168960571289, 8.383220672607422] |
56bef685-2c5b-4e72-a2b2-11da204617d8 | script-induction-as-association-rule-mining | null | null | https://aclanthology.org/2020.nuse-1.7 | https://aclanthology.org/2020.nuse-1.7.pdf | Script Induction as Association Rule Mining | We show that the count-based Script Induction models of Chambers and Jurafsky (2008) and Jans et al. (2012) can be unified in a general framework of narrative chain likelihood maximization. We provide efficient algorithms based on Association Rule Mining (ARM) and weighted set cover that can discover interesting patterns in the training data and combine them in a reliable and explainable way to predict the missing event. The proposed method, unlike the prior work, does not assume full conditional independence and makes use of higher-order count statistics. We perform the ablation study and conclude that the inductive biases introduced by ARM are conducive to better performance on the narrative cloze test. | ['Benjamin Van Durme', 'Anton Belyy'] | 2020-07-01 | null | null | null | ws-2020-7 | ['cloze-test'] | ['natural-language-processing'] | [ 2.13199869e-01 3.44692945e-01 -1.03675675e+00 -4.85268712e-01
-3.14999729e-01 -6.56751752e-01 9.06126559e-01 2.98347026e-01
-4.04240310e-01 1.28089213e+00 3.52147520e-01 -7.46511877e-01
-7.35064387e-01 -1.05990076e+00 -3.84952605e-01 -3.59838933e-01
-3.44221443e-01 5.00741780e-01 -7.26668611e-02 1.73003495e-01
3.80439639e-01 3.27475518e-01 -1.47479165e+00 4.95629430e-01
1.10589731e+00 4.33269441e-01 -8.46739858e-02 3.41478705e-01
2.12258086e-01 1.57225406e+00 -1.44704148e-01 -7.08632886e-01
-3.72548461e-01 -6.54038370e-01 -7.95299053e-01 -5.45486398e-02
-1.21268719e-01 -3.49879801e-01 -2.27481306e-01 4.97741252e-01
1.64789245e-01 1.98629603e-01 1.12984324e+00 -1.16472888e+00
-3.34721059e-01 1.22467816e+00 -5.56373775e-01 1.74939588e-01
4.53451514e-01 -6.81603700e-02 1.09193456e+00 -9.84830499e-01
9.25035179e-01 9.48088169e-01 5.45766056e-01 2.37410679e-01
-1.06494486e+00 -1.02730858e+00 4.66668308e-01 7.37563133e-01
-1.17826402e+00 -2.36706957e-01 6.40096366e-01 -4.10719365e-01
9.67895985e-01 5.64354181e-01 7.92648435e-01 1.57639074e+00
-1.69816285e-01 8.46415102e-01 1.52919137e+00 -7.89032638e-01
4.60611492e-01 4.18268949e-01 4.68593627e-01 5.46744347e-01
8.32306802e-01 4.39043105e-01 -9.17219937e-01 -3.61027569e-01
5.61351061e-01 -1.62536308e-01 2.37347633e-01 8.94820914e-02
-8.82658422e-01 1.39538467e+00 -1.95234597e-01 3.75918925e-01
-4.67451930e-01 -4.65084836e-02 1.89103350e-01 -7.43460134e-02
6.01097167e-01 4.47405368e-01 -2.09061742e-01 -1.00902647e-01
-1.21091330e+00 5.88894308e-01 9.59630787e-01 7.99246848e-01
8.57537612e-02 -8.61754268e-03 -5.03443182e-01 4.02641058e-01
5.31538963e-01 2.58743793e-01 2.63049662e-01 -8.85481358e-01
1.74014792e-01 3.63340735e-01 -6.72546157e-04 -8.20251107e-01
-3.62968892e-01 -5.64988077e-01 -3.64096642e-01 2.31517795e-02
3.50489199e-01 -3.68750721e-01 -5.34229636e-01 1.77711725e+00
1.97005883e-01 1.99397519e-01 -8.54342580e-02 4.96061981e-01
6.98217988e-01 2.07340270e-01 6.11835718e-01 -8.07308912e-01
1.28257167e+00 -4.09474939e-01 -1.28867960e+00 -3.77066016e-01
5.57058513e-01 -3.75344366e-01 8.08641016e-01 8.42825294e-01
-1.17695892e+00 6.55985996e-02 -1.23154199e+00 2.36531645e-01
-6.71898276e-02 -8.42431262e-02 1.68651557e+00 9.94904459e-01
-4.38057452e-01 5.52135468e-01 -6.74368620e-01 -1.82061523e-01
7.14146256e-01 5.63490763e-03 -1.24741765e-02 1.37149347e-02
-1.19323123e+00 1.06243145e+00 6.13419056e-01 -2.59531945e-01
-8.66071045e-01 -5.89791656e-01 -7.78898478e-01 -1.13701615e-02
7.74293482e-01 -5.70526838e-01 1.02358234e+00 -5.08992910e-01
-1.06801212e+00 6.54642940e-01 -4.94022787e-01 -3.80530775e-01
4.62470680e-01 -2.80554473e-01 -1.82455972e-01 1.26983687e-01
2.00327322e-01 3.37780342e-02 2.56456256e-01 -1.22296143e+00
-5.95570207e-01 -2.81993121e-01 -3.08794975e-01 -1.40969614e-02
-3.01670432e-01 4.52405453e-01 3.65890741e-01 -6.17811322e-01
9.15506780e-02 -5.18423915e-01 -3.47740501e-01 -7.90209413e-01
-6.16655469e-01 -4.89213407e-01 7.58086592e-02 -4.87403125e-01
1.71846461e+00 -1.48820007e+00 -6.46938905e-02 5.00038266e-01
8.57952088e-02 -3.87873322e-01 5.12345731e-01 4.73396838e-01
-3.07342350e-01 4.47074503e-01 -2.74598241e-01 -1.01960965e-01
-8.42507258e-02 1.88855127e-01 -4.43069875e-01 4.74300116e-01
1.83565378e-01 7.48931348e-01 -8.72634232e-01 -8.61089170e-01
4.91185784e-02 -1.26479730e-01 -4.55237299e-01 -1.06125616e-01
-2.54738986e-01 1.91781446e-02 -2.47836366e-01 6.12578571e-01
3.99991810e-01 -8.77618566e-02 5.68011522e-01 4.02512074e-01
-4.85216826e-01 7.26840854e-01 -1.22844088e+00 1.06433654e+00
-1.38649881e-01 5.91921329e-01 -5.43041766e-01 -7.61385858e-01
7.35081792e-01 4.49406326e-01 4.37704585e-02 -1.98115721e-01
1.24538876e-01 -6.78637996e-03 9.81514975e-02 -6.98101103e-01
2.23197713e-01 -5.68677127e-01 -1.30767092e-01 7.65309215e-01
4.51055653e-02 1.28298521e-01 4.67395574e-01 4.92072731e-01
1.00754118e+00 1.55282393e-01 7.23985314e-01 -3.06215763e-01
-5.05678635e-03 2.85945982e-01 6.67184234e-01 1.13474774e+00
4.50841993e-01 1.41984552e-01 6.44353032e-01 -9.91517976e-02
-9.09699082e-01 -1.13965225e+00 -6.76350653e-01 1.02802169e+00
-4.08577949e-01 -4.86800343e-01 -2.64043897e-01 -6.65103257e-01
-4.12808597e-01 1.72264397e+00 -8.46415401e-01 1.34290472e-01
-1.86062217e-01 -1.22624350e+00 4.89696711e-01 5.73683023e-01
9.07078460e-02 -1.09546077e+00 -6.64226055e-01 2.66672760e-01
-3.44721347e-01 -9.44808960e-01 4.61706668e-01 6.04753613e-01
-9.75048363e-01 -9.97180521e-01 2.72377014e-01 -4.05758083e-01
4.22647268e-01 -2.62046635e-01 9.90004838e-01 -1.94197699e-01
-1.71195388e-01 2.74644405e-01 -6.86685860e-01 -1.06319332e+00
-2.97272682e-01 -9.05967429e-02 -5.61191477e-02 -2.78481692e-01
8.80642593e-01 -9.47818100e-01 -9.14206505e-02 -1.68164879e-01
-7.90803671e-01 1.87811822e-01 4.60694849e-01 7.46895373e-01
3.21153224e-01 1.71699047e-01 9.31859195e-01 -1.20461428e+00
5.98753214e-01 -1.11741114e+00 -1.81971818e-01 1.06957503e-01
-1.12713575e+00 -1.23947367e-01 -3.44148129e-02 -6.11833930e-01
-1.50189817e+00 -2.44227409e-01 2.14806989e-01 3.27775955e-01
-3.43091547e-01 9.24307644e-01 -9.91131887e-02 4.68920857e-01
7.81395257e-01 8.30511190e-03 -4.88871425e-01 -2.84743041e-01
1.99216411e-01 4.86935824e-01 2.20058069e-01 -5.36345065e-01
7.22788453e-01 5.12639940e-01 1.05183780e-01 -3.00870776e-01
-1.05827165e+00 -1.07115045e-01 -5.07335722e-01 -4.83578831e-01
7.06456661e-01 -1.00973535e+00 -7.30118513e-01 -1.22789904e-01
-1.12854123e+00 -2.10420340e-01 -2.76996881e-01 1.03704441e+00
-5.66338122e-01 6.28480911e-02 -4.61884856e-01 -1.68706787e+00
2.22718120e-02 -2.93673366e-01 3.23409319e-01 1.18198372e-01
-9.43307579e-01 -9.47851479e-01 3.41322601e-01 2.43221596e-01
-2.68057406e-01 5.70735812e-01 1.10772276e+00 -1.04575384e+00
-3.16663772e-01 -2.66656846e-01 2.32790768e-01 -3.44537050e-01
-3.63403559e-01 1.32439896e-01 -9.59271371e-01 3.81060988e-01
3.61693025e-01 -4.52507824e-01 1.02348065e+00 5.12564957e-01
8.93396556e-01 -6.53192580e-01 -3.34962577e-01 2.31817231e-01
1.40130067e+00 2.55138010e-01 8.87540460e-01 4.76401895e-01
2.93660074e-01 9.12066638e-01 7.64434040e-01 7.90519536e-01
3.89099479e-01 2.65013814e-01 1.09705381e-01 2.49277785e-01
3.54288638e-01 -5.31201839e-01 4.51362073e-01 3.43634427e-01
-5.40421069e-01 -1.50328889e-01 -5.39942205e-01 8.43321979e-01
-2.04122996e+00 -1.59012902e+00 -5.52031219e-01 2.01297045e+00
1.21137094e+00 2.65578508e-01 2.61233956e-01 4.96815652e-01
5.10479033e-01 -1.41135529e-01 -4.10131365e-02 -7.53821373e-01
-4.21003938e-01 5.93680143e-01 3.24775159e-01 5.34235954e-01
-8.23423207e-01 7.25363016e-01 7.64088011e+00 9.59750354e-01
3.80365760e-03 1.73015192e-01 7.75117218e-01 -5.80460250e-01
-7.84308434e-01 2.68865854e-01 -8.14242780e-01 1.99980840e-01
9.92392063e-01 -2.30674118e-01 1.38003707e-01 8.18364322e-01
2.91232169e-01 -7.13741362e-01 -1.18717611e+00 3.34980309e-01
1.59880802e-01 -1.22296286e+00 -1.06489085e-01 2.69969642e-01
8.24637651e-01 -6.40951514e-01 -9.17988494e-02 2.48928487e-01
8.88630629e-01 -1.50613976e+00 8.65121245e-01 6.11097693e-01
5.86005867e-01 -7.79874563e-01 6.52093410e-01 5.24351418e-01
-3.25092673e-01 -2.43917286e-01 -1.13080055e-01 -8.20830941e-01
9.52420831e-02 9.71968830e-01 -1.12258458e+00 4.39726323e-01
4.82466131e-01 3.96524161e-01 -2.46813178e-01 7.32513428e-01
-9.21192169e-01 1.51289535e+00 -3.16067696e-01 -4.47741210e-01
9.25655365e-02 2.72807367e-02 5.66845179e-01 1.22278225e+00
1.96949273e-01 2.97297657e-01 -2.70797342e-01 1.10554492e+00
3.10136914e-01 3.11645269e-01 -6.63390160e-01 9.39908177e-02
6.63136363e-01 1.14668381e+00 -9.22755718e-01 -3.04282755e-01
-4.14536983e-01 2.18873203e-01 1.54838085e-01 9.49563235e-02
-6.32710993e-01 1.98345289e-01 3.00373994e-02 2.20249012e-01
4.34133977e-01 1.98248886e-02 -1.28725290e+00 -9.00419056e-01
-1.84220374e-01 -6.75294578e-01 8.16660702e-01 -3.81266236e-01
-1.25253654e+00 1.65163632e-02 6.68457031e-01 -6.60020947e-01
-6.43960893e-01 -3.36399376e-01 -9.34737384e-01 7.68051386e-01
-1.15123451e+00 -1.20681441e+00 8.14486220e-02 4.80770409e-01
2.94720531e-01 -1.06993228e-01 9.38707352e-01 -2.20339850e-01
-5.95967233e-01 5.06511748e-01 -2.43406624e-01 1.09340884e-02
3.80853504e-01 -1.24422765e+00 -2.61802226e-01 7.88430452e-01
1.61892682e-01 5.81370175e-01 9.03286517e-01 -1.16563857e+00
-8.18964541e-01 -7.06687987e-01 1.25485432e+00 -7.38605917e-01
6.22783661e-01 -4.15183246e-01 -6.24808192e-01 9.56389606e-01
3.32523704e-01 -7.40751624e-01 1.37546098e+00 7.42819309e-01
-4.06787336e-01 4.09360647e-01 -1.08417881e+00 7.47802496e-01
1.16584158e+00 6.96167722e-02 -8.96108091e-01 2.45808885e-01
3.14890146e-01 3.31432432e-01 -7.82388389e-01 5.42479694e-01
5.33430636e-01 -8.81989539e-01 7.43089080e-01 -1.00996101e+00
9.53090549e-01 8.76731202e-02 1.51959080e-02 -7.84092724e-01
-5.74881673e-01 -4.09345895e-01 -2.36365736e-01 1.25636387e+00
8.13715935e-01 -1.46036536e-01 5.71038008e-01 5.63863218e-01
2.01749191e-01 -7.10510731e-01 -8.66573036e-01 -6.87399089e-01
2.30027378e-01 -1.02603555e+00 3.60557169e-01 1.27455223e+00
7.95470655e-01 5.51695287e-01 -5.33895075e-01 -1.76021546e-01
9.06664491e-01 1.22006044e-01 2.37632886e-01 -1.40772295e+00
-4.10179824e-01 -3.28444928e-01 1.43583968e-01 2.50585191e-02
1.95001528e-01 -9.31314111e-01 -2.54245222e-01 -1.46324706e+00
8.96306038e-01 -2.09733635e-01 -7.12926760e-02 4.67306703e-01
-3.39062721e-01 8.82633775e-02 -2.60293894e-02 -1.45437241e-01
-6.65510416e-01 2.58376002e-01 7.13182509e-01 2.54176110e-01
-3.09316009e-01 -1.05519652e-01 -1.29180634e+00 9.49804902e-01
9.16447043e-01 -8.80561769e-01 -4.71702665e-01 2.93167293e-01
9.11959767e-01 7.57126212e-02 3.65440547e-01 -4.70784843e-01
1.71261683e-01 -6.02944016e-01 7.19042540e-01 -6.71296775e-01
-9.60862935e-02 -3.95811826e-01 2.96116173e-01 3.17785084e-01
-5.88284612e-01 -4.23467278e-01 2.55788267e-01 3.51492584e-01
1.76993549e-01 -8.82375658e-01 2.63973773e-01 -1.94444627e-01
-8.56423154e-02 -2.31729090e-01 -9.23845172e-01 -2.02752665e-01
9.82492626e-01 -2.62526810e-01 -1.82325438e-01 -6.45833194e-01
-8.27888131e-01 2.17599105e-02 -1.02085240e-01 -9.23151374e-02
6.06785893e-01 -1.44964278e+00 -1.32244492e+00 -4.84043837e-01
-5.62390573e-02 -3.60405862e-01 1.35862827e-01 1.05824113e+00
1.12486809e-01 3.53100181e-01 6.22721836e-02 -2.82552501e-05
-9.11767602e-01 5.47336519e-01 -2.82195240e-01 -7.35534370e-01
-2.23543957e-01 4.93304551e-01 -4.48068380e-02 2.24073976e-01
4.21474241e-02 2.59708315e-01 -5.75725853e-01 4.84552383e-01
8.03988099e-01 6.77919924e-01 -2.66116828e-01 6.23831302e-02
-3.90446752e-01 -2.14520887e-01 -1.57953098e-01 -6.64729714e-01
1.77673852e+00 1.21507883e-01 -2.00967804e-01 8.52281928e-01
3.53579938e-01 2.83547759e-01 -5.40411174e-01 -6.40847236e-02
3.43319654e-01 -2.53142357e-01 2.16565952e-01 -1.20763862e+00
-2.92032331e-01 3.85149747e-01 5.79335075e-03 1.07060313e-01
9.34473634e-01 2.84258187e-01 -2.57445723e-01 1.08093835e-01
1.40947744e-01 -1.11739588e+00 -1.83852047e-01 1.71190813e-01
8.05447519e-01 -9.08217490e-01 5.60675323e-01 -6.88713610e-01
-7.85655618e-01 8.94928813e-01 4.25667018e-01 3.18420678e-02
3.61595929e-01 6.56262696e-01 -6.72688246e-01 -3.25931877e-01
-1.23727596e+00 -3.46260071e-01 9.06128809e-02 6.79866433e-01
5.71059048e-01 1.67258903e-01 -1.09992945e+00 1.57843494e+00
-3.96319807e-01 7.75358006e-02 8.06847155e-01 8.19957197e-01
-3.13821524e-01 -7.99530923e-01 -6.70962214e-01 8.52254927e-01
-7.68693030e-01 -4.65783536e-01 -5.89437842e-01 9.29780960e-01
2.05008894e-01 1.27322507e+00 9.72769335e-02 -4.05404299e-01
-5.42878360e-02 4.87846375e-01 7.96862900e-01 -5.91967225e-01
-4.68592882e-01 1.91265523e-01 7.64604449e-01 -1.19655326e-01
-6.34948969e-01 -1.29780841e+00 -7.92670190e-01 -4.99302566e-01
-5.01274824e-01 2.27906585e-01 2.95189530e-01 1.20234585e+00
-3.34092826e-01 4.27282512e-01 5.35162449e-01 -5.43331727e-02
-1.98877051e-01 -1.43364847e+00 -6.40624583e-01 8.62099789e-03
-3.52471441e-01 -8.49296868e-01 -5.52303374e-01 4.88346592e-02] | [8.085378646850586, 5.4573163986206055] |
430cff9c-aa34-45bb-a04b-0e5e302c4171 | fine-grained-text-style-transfer-with | 2305.19512 | null | https://arxiv.org/abs/2305.19512v2 | https://arxiv.org/pdf/2305.19512v2.pdf | Fine-grained Text Style Transfer with Diffusion-Based Language Models | Diffusion probabilistic models have shown great success in generating high-quality images controllably, and researchers have tried to utilize this controllability into text generation domain. Previous works on diffusion-based language models have shown that they can be trained without external knowledge (such as pre-trained weights) and still achieve stable performance and controllability. In this paper, we trained a diffusion-based model on StylePTB dataset, the standard benchmark for fine-grained text style transfers. The tasks in StylePTB requires much more refined control over the output text compared to tasks evaluated in previous works, and our model was able to achieve state-of-the-art performance on StylePTB on both individual and compositional transfers. Moreover, our model, trained on limited data from StylePTB without external knowledge, outperforms previous works that utilized pretrained weights, embeddings, and external grammar parsers, and this may indicate that diffusion-based language models have great potential under low-resource settings. | ['Honglak Lee', 'Todd C. Hollon', 'Jiacheng Shi', 'Tiange Luo', 'Yiwei Lyu'] | 2023-05-31 | null | null | null | null | ['style-transfer', 'text-style-transfoer'] | ['computer-vision', 'natural-language-processing'] | [ 2.42700994e-01 4.43048328e-01 -1.43049553e-01 -1.22457929e-01
-6.23459935e-01 -6.85246527e-01 1.18301618e+00 -4.17854339e-01
-3.86118174e-01 6.91858172e-01 5.89495003e-01 -2.85508394e-01
3.74091208e-01 -1.10340714e+00 -7.46649444e-01 -4.17738795e-01
4.83917236e-01 8.72820675e-01 3.24731082e-01 -5.72279990e-01
1.67366296e-01 2.47853443e-01 -1.05453920e+00 4.90038902e-01
8.17485690e-01 6.22407675e-01 5.30162215e-01 8.56130421e-01
-4.94719505e-01 7.88184941e-01 -7.96168447e-01 -6.04435980e-01
1.01106547e-01 -5.36927760e-01 -8.05778801e-01 -1.00327127e-01
4.88689035e-01 -3.98484349e-01 -5.45018852e-01 7.62095213e-01
7.49801159e-01 -6.12402260e-02 9.49596226e-01 -1.09725833e+00
-1.64481699e+00 1.13652670e+00 -1.63403109e-01 -9.80163217e-02
1.62272602e-01 5.39042473e-01 9.07205701e-01 -8.51595819e-01
1.05934048e+00 1.61611187e+00 4.94985521e-01 1.36989021e+00
-1.57347476e+00 -7.14294493e-01 1.27992943e-01 -2.82006323e-01
-1.03427756e+00 -4.48147714e-01 4.57898229e-01 -4.21617299e-01
1.41848719e+00 -6.06812835e-02 6.03475511e-01 2.01551986e+00
3.13647330e-01 1.06144881e+00 1.17110360e+00 -5.64622223e-01
2.87552970e-03 3.33664715e-01 -4.58484471e-01 6.61919177e-01
7.62470290e-02 1.77868932e-01 -7.76660085e-01 1.74793363e-01
9.96574879e-01 -5.43421984e-01 -2.17866719e-01 -1.83119550e-01
-1.53917241e+00 1.00041401e+00 4.24630314e-01 3.81995380e-01
-6.27854913e-02 4.82433736e-01 2.61860162e-01 2.65670627e-01
7.42997229e-01 7.58235276e-01 -3.98157984e-01 -3.33374500e-01
-9.12213266e-01 2.17960283e-01 8.00753355e-01 1.49186802e+00
3.51719409e-01 3.46366972e-01 -8.10927570e-01 8.62105072e-01
7.32518360e-02 8.77487957e-01 8.02835345e-01 -9.04749334e-01
8.55183005e-01 2.85396963e-01 1.79211572e-02 -6.64199769e-01
-6.38487190e-02 -7.26118982e-02 -8.37005138e-01 3.13782007e-01
4.65055883e-01 -5.33783317e-01 -1.18934476e+00 1.87274361e+00
-2.88295031e-01 -2.70983338e-01 1.94338888e-01 6.39249265e-01
6.16596401e-01 9.29042637e-01 1.16426699e-01 3.08247089e-01
9.38951254e-01 -1.33682156e+00 -6.93342626e-01 -3.06226820e-01
5.58722734e-01 -1.01104438e+00 1.55754662e+00 2.23402739e-01
-1.30485368e+00 -5.63253164e-01 -8.87513041e-01 -1.73523828e-01
-3.70271981e-01 4.68899459e-02 6.79383874e-01 7.55373240e-01
-1.57200980e+00 8.10915649e-01 -7.04589069e-01 -6.48022413e-01
6.23042107e-01 7.13527724e-02 -1.99414402e-01 -2.23238729e-02
-1.28771544e+00 1.06425226e+00 1.48454905e-01 -3.33116055e-01
-1.20654786e+00 -7.23635912e-01 -7.01483965e-01 -1.06601879e-01
8.91005620e-02 -1.05210447e+00 1.25408733e+00 -9.22838748e-01
-2.05763912e+00 7.87080228e-01 2.78124567e-02 -5.23923159e-01
8.28479052e-01 -2.77476877e-01 -1.31296858e-01 -1.27255753e-01
-3.65204364e-02 1.48227596e+00 1.18669319e+00 -9.76561904e-01
-3.79498422e-01 2.11586535e-01 1.58805266e-01 1.57764092e-01
-9.32087779e-01 -9.97776613e-02 -4.71252143e-01 -1.03459907e+00
-6.19835138e-01 -1.10033607e+00 -2.67393649e-01 1.25914693e-01
-4.32837427e-01 -3.47265452e-01 7.66782820e-01 -3.01848531e-01
1.00428784e+00 -1.84501004e+00 5.54219484e-01 -3.15900683e-01
-9.75510404e-02 3.27784479e-01 -6.53604448e-01 4.78064805e-01
2.96816975e-01 6.80512726e-01 -5.15792929e-02 -5.78692615e-01
1.76084712e-01 1.52280688e-01 -4.98366922e-01 -2.59820729e-01
5.96678376e-01 1.14965546e+00 -9.21625137e-01 -6.21329546e-01
-1.48114366e-02 5.95672905e-01 -6.48086548e-01 3.39819521e-01
-6.88255668e-01 2.18725577e-01 -4.37573224e-01 3.21927786e-01
3.39231998e-01 -1.51663169e-01 2.63928976e-02 1.18030589e-02
1.96163490e-01 8.06729421e-02 -5.92318118e-01 2.02968550e+00
-8.51669490e-01 1.04496801e+00 -3.15424889e-01 -2.87681758e-01
8.35106432e-01 3.56291801e-01 -4.53985445e-02 -6.04799151e-01
-1.87736705e-01 2.23463230e-05 -1.54670298e-01 -2.45980963e-01
7.87877381e-01 -2.62063980e-01 -9.30795968e-02 7.89469838e-01
3.78019780e-01 -9.08202887e-01 3.36916685e-01 4.92516458e-01
1.08760643e+00 5.08992493e-01 -3.87692422e-01 -3.48726094e-01
1.63876489e-01 4.09105085e-02 1.55127943e-01 1.00235283e+00
1.17815807e-01 1.00957060e+00 4.72548962e-01 -2.39318907e-01
-1.31288660e+00 -1.06172168e+00 1.50451601e-01 1.19863367e+00
-2.10885063e-01 -4.66825247e-01 -1.06276500e+00 -8.44773650e-01
-1.37221888e-01 1.10157514e+00 -7.19561040e-01 -1.44606858e-01
-5.92790544e-01 -8.24791908e-01 8.74623597e-01 7.72857189e-01
4.53858316e-01 -1.40260959e+00 -1.31084755e-01 4.44869846e-01
-1.06412224e-01 -1.09728694e+00 -8.09642136e-01 -2.91942153e-02
-9.82826591e-01 -3.33699882e-01 -1.20952952e+00 -6.90695703e-01
6.07919395e-01 -1.57608971e-01 1.47009850e+00 -1.57362729e-01
-3.18178870e-02 3.86184305e-01 -3.00195903e-01 -5.84205985e-01
-8.75325382e-01 6.86721623e-01 -1.62366718e-01 -3.00511211e-01
4.85104555e-03 -3.58220637e-01 -4.72547978e-01 3.27086389e-01
-1.07632351e+00 3.51076156e-01 6.79820299e-01 1.01111042e+00
2.07774073e-01 -2.37884179e-01 5.94361842e-01 -1.05599093e+00
1.18452156e+00 -7.90021494e-02 -4.17775601e-01 2.47118518e-01
-7.86739290e-01 4.66625720e-01 5.35143673e-01 -7.91684210e-01
-1.50797665e+00 -2.31269449e-01 -1.66974619e-01 -2.79149383e-01
-1.85085330e-02 7.36207068e-02 -1.90706588e-02 5.15948310e-02
9.70050395e-01 1.24187373e-01 -9.02481750e-02 -2.61628151e-01
8.94656718e-01 5.47645271e-01 1.39782047e-02 -9.54836786e-01
8.70172083e-01 3.72693837e-01 -1.77301168e-01 -4.56223190e-01
-7.85125196e-01 3.68650734e-01 -4.78396833e-01 6.82428032e-02
1.03758538e+00 -9.35736477e-01 -1.19891755e-01 7.12011755e-01
-1.29338777e+00 -1.18135059e+00 -5.94606936e-01 9.68100429e-02
-7.93030322e-01 -3.46936248e-02 -1.15020251e+00 -5.18227220e-01
-4.76764977e-01 -1.22445250e+00 1.13316929e+00 2.80250050e-02
-5.92043161e-01 -1.39482629e+00 -1.01906672e-01 1.37019068e-01
1.06658351e+00 -9.56688225e-02 1.04004836e+00 -2.56791323e-01
-7.99788475e-01 1.34559229e-01 -2.01684475e-01 4.23760384e-01
1.86354414e-01 6.87074289e-02 -8.66983950e-01 -1.22388020e-01
-3.17399174e-01 -6.49722636e-01 9.87971365e-01 1.11589953e-01
9.35055792e-01 -1.36534169e-01 -4.26743656e-01 5.55020750e-01
1.08590066e+00 -2.59469479e-01 6.05355322e-01 3.30016613e-01
9.69056249e-01 4.87668961e-01 2.40131333e-01 1.10753134e-01
3.49130839e-01 7.06292212e-01 -2.64245812e-02 -2.41007790e-01
-9.23919678e-01 -7.18622983e-01 7.90163755e-01 8.44168603e-01
2.43055504e-02 -7.40918398e-01 -7.25265801e-01 6.47254407e-01
-1.72792494e+00 -7.87971556e-01 1.23480856e-02 1.55428708e+00
1.30408394e+00 3.99648368e-01 -1.69288725e-01 -3.05010051e-01
5.86377323e-01 2.81049490e-01 -4.79578912e-01 -7.59747803e-01
-5.25984228e-01 3.19402188e-01 5.86131573e-01 3.86379898e-01
-7.15078652e-01 1.64541459e+00 7.51605034e+00 1.21024203e+00
-1.17885029e+00 3.16042751e-01 6.80984437e-01 -3.06317836e-01
-7.76895165e-01 -7.76774734e-02 -9.70583618e-01 3.36606532e-01
9.81185079e-01 -1.01135813e-01 6.28676116e-01 6.88985586e-01
3.75340022e-02 3.56295913e-01 -1.23087764e+00 4.83315170e-01
2.27708295e-01 -1.59060240e+00 6.55682027e-01 2.61763129e-02
1.22116697e+00 1.41752720e-01 4.79026794e-01 5.29232383e-01
9.96930540e-01 -1.19461954e+00 1.16477740e+00 5.48516393e-01
1.05012476e+00 -3.86037022e-01 4.47292835e-01 2.34934911e-01
-6.53961658e-01 6.40113726e-02 -5.32809913e-01 -4.56122085e-02
4.67437446e-01 5.39630711e-01 -8.08509111e-01 7.48315603e-02
4.89937574e-01 7.63047099e-01 -6.88052297e-01 2.18651891e-01
-6.24963999e-01 7.97567904e-01 -2.35916674e-02 -3.82608503e-01
2.55361497e-01 3.32767069e-02 4.36089516e-01 1.45044255e+00
6.00160241e-01 -4.20312136e-01 -1.02833673e-01 1.15454042e+00
-4.16590869e-01 4.32436578e-02 -8.09450269e-01 -4.24239039e-01
1.26967868e-02 1.07323050e+00 -6.06627882e-01 -5.64531088e-01
-2.32921988e-01 1.27543342e+00 5.65860569e-01 4.30276811e-01
-9.02165294e-01 -1.61164463e-01 5.20435035e-01 1.53649777e-01
4.40659851e-01 -3.94684136e-01 -5.73497415e-01 -1.31331694e+00
-4.07164216e-01 -9.28237081e-01 -1.54533107e-02 -1.18284237e+00
-1.48107421e+00 9.71177101e-01 -1.06446587e-01 -7.48551369e-01
-2.34744161e-01 -7.69555628e-01 -5.05720794e-01 1.15140188e+00
-1.26509607e+00 -1.57748985e+00 3.08662001e-03 5.10360897e-01
9.02455032e-01 -4.14925694e-01 1.07733583e+00 -1.19427346e-01
-3.97410452e-01 8.09155703e-01 -7.69429356e-02 -6.41161203e-02
1.15003872e+00 -1.43525028e+00 9.11803484e-01 7.05444932e-01
2.38672167e-01 6.12005591e-01 6.40090466e-01 -7.94765651e-01
-1.32070923e+00 -1.17485428e+00 9.49809372e-01 -1.10226429e+00
8.01308036e-01 -7.33217657e-01 -4.84736055e-01 6.73139453e-01
8.09705973e-01 -3.38843614e-01 2.27376565e-01 8.61980394e-02
-3.07934523e-01 1.41504496e-01 -9.28458333e-01 1.04973650e+00
1.53890371e+00 -5.10816693e-01 -4.50840354e-01 4.84191805e-01
1.05453491e+00 -4.84520733e-01 -8.12122405e-01 -2.68381042e-03
3.49773675e-01 -6.96534038e-01 8.77201140e-01 -5.55982351e-01
9.08128440e-01 6.38875216e-02 1.36884838e-01 -1.70526004e+00
-4.35017377e-01 -8.61105621e-01 6.97847530e-02 1.69603848e+00
9.60512817e-01 -4.67816830e-01 6.94147587e-01 6.70013368e-01
9.89480689e-03 -4.84263957e-01 -4.26970601e-01 -8.80776048e-01
8.15252304e-01 -3.29303592e-01 6.67115033e-01 6.36107802e-01
-1.89109936e-01 5.86525142e-01 -6.02752626e-01 -5.38700700e-01
3.18985492e-01 -1.12702854e-01 7.68027425e-01 -7.67245054e-01
-2.23704830e-01 -7.16584504e-01 8.73835385e-02 -1.22670186e+00
5.22013366e-01 -1.09131169e+00 1.21703170e-01 -1.68362916e+00
1.43919855e-01 -6.85623646e-01 1.48144262e-02 5.04873812e-01
-3.49106312e-01 3.56357157e-01 3.65201712e-01 1.71426252e-01
-3.30103487e-01 7.42021501e-01 1.74487936e+00 -5.31353354e-01
-5.59542365e-02 -4.13472265e-01 -8.96986961e-01 4.24518853e-01
8.98054421e-01 -5.73954344e-01 -6.60926998e-01 -1.13154447e+00
2.58834273e-01 -3.08569312e-01 -2.02230766e-01 -7.58532524e-01
1.83313698e-01 -2.28607908e-01 4.67469573e-01 5.26255555e-02
2.32647032e-01 -3.86651635e-01 -1.29589736e-01 2.31410250e-01
-6.72906876e-01 8.11374485e-02 2.79056966e-01 4.74932283e-01
1.67213082e-02 -2.54218429e-01 4.92488354e-01 -2.59088010e-01
-4.39945072e-01 2.76467860e-01 -8.10264349e-01 2.04938024e-01
7.23659337e-01 8.54785591e-02 -3.58444601e-01 -5.25805771e-01
-5.02336621e-01 -1.73354760e-01 8.06720555e-01 8.41498971e-01
5.23585677e-01 -1.49192429e+00 -8.19839001e-01 4.99294437e-02
8.32406655e-02 -1.09012239e-01 -4.69944850e-02 2.17922673e-01
-4.83237982e-01 3.52843463e-01 -2.31174752e-01 -4.82058585e-01
-6.07006192e-01 6.52063251e-01 1.16466813e-01 -5.48351645e-01
-6.38249874e-01 1.11938667e+00 2.76560664e-01 -4.77849096e-01
-2.76191626e-02 -3.67523223e-01 1.58776551e-01 -1.27085894e-01
2.97089636e-01 -5.91349266e-02 -2.99538404e-01 -2.85313457e-01
5.45441546e-02 6.56852007e-01 -1.70262918e-01 -6.67954862e-01
1.16798973e+00 -6.85563609e-02 1.90189838e-01 2.93040007e-01
7.10371792e-01 9.69518796e-02 -1.66279483e+00 -2.58677341e-02
-1.44638181e-01 -2.13879243e-01 7.91666005e-03 -9.91268575e-01
-1.05138230e+00 1.22075129e+00 2.40272939e-01 -6.59541488e-02
7.85711110e-01 -1.07848354e-01 1.01872075e+00 2.73852795e-01
3.63840193e-01 -1.19347358e+00 6.38112247e-01 7.62675345e-01
1.18889606e+00 -1.13989139e+00 -4.55300063e-01 -1.34983242e-01
-9.93571877e-01 1.07671738e+00 6.95303380e-01 -1.30359095e-03
2.84278184e-01 5.56798398e-01 3.01481724e-01 2.90967166e-01
-1.15677619e+00 7.73680210e-02 1.16979033e-01 9.60432708e-01
6.33368075e-01 7.43157268e-02 -6.31019250e-02 3.37414443e-01
-4.58280087e-01 9.51956958e-02 6.08749449e-01 7.23342061e-01
-5.26245162e-02 -1.59417975e+00 -6.97082281e-02 1.86643869e-01
-3.62739474e-01 -5.66394985e-01 -5.98481774e-01 6.42545283e-01
-9.54434499e-02 9.95914161e-01 -7.69629627e-02 -3.47343177e-01
3.92022222e-01 1.30625173e-01 7.72292793e-01 -7.38833547e-01
-7.33097255e-01 -1.87728032e-01 2.83823639e-01 -4.46695745e-01
-1.16126820e-01 -4.47478950e-01 -9.16734636e-01 -4.50053006e-01
-2.26869136e-01 -1.54823124e-01 7.20994174e-01 6.54416084e-01
4.88309056e-01 7.00475216e-01 2.42539480e-01 -8.56271148e-01
-4.91918623e-01 -1.29873872e+00 -2.03347683e-01 6.58729315e-01
-2.03395709e-01 -2.12782025e-01 -1.12167008e-01 4.80446160e-01] | [11.398355484008789, -0.030244501307606697] |
6bdf0aa0-cd8c-47a8-a409-f27ffd8f994d | mobile-networks-for-computer-go | 2008.10080 | null | https://arxiv.org/abs/2008.10080v1 | https://arxiv.org/pdf/2008.10080v1.pdf | Mobile Networks for Computer Go | The architecture of the neural networks used in Deep Reinforcement Learning programs such as Alpha Zero or Polygames has been shown to have a great impact on the performances of the resulting playing engines. For example the use of residual networks gave a 600 ELO increase in the strength of Alpha Go. This paper proposes to evaluate the interest of Mobile Network for the game of Go using supervised learning as well as the use of a policy head and a value head different from the Alpha Zero heads. The accuracy of the policy, the mean squared error of the value, the efficiency of the networks with the number of parameters, the playing speed and strength of the trained networks are evaluated. | ['Tristan Cazenave'] | 2020-08-23 | null | null | null | null | ['game-of-go'] | ['playing-games'] | [-4.70354527e-01 4.51506495e-01 -2.84173995e-01 -2.49573775e-02
3.09611380e-01 -3.33699793e-01 6.63789928e-01 -3.64719182e-01
-1.05581629e+00 8.26377690e-01 -7.98408836e-02 -5.90437353e-01
-3.69934112e-01 -9.48637366e-01 -7.26013541e-01 -7.60312140e-01
-1.51711777e-01 3.48183095e-01 7.68780708e-01 -8.86362970e-01
3.69178981e-01 2.66395926e-01 -1.49714804e+00 1.22267790e-01
4.08468693e-01 1.19097805e+00 1.42679632e-01 7.75049210e-01
1.27357304e-01 1.15566123e+00 -1.06180775e+00 -1.25833526e-01
3.40025067e-01 -2.18786299e-01 -6.70899689e-01 -5.67483664e-01
-3.02437067e-01 -2.76134133e-01 -4.95375901e-01 7.69276261e-01
6.66624844e-01 4.77692693e-01 4.12958533e-01 -1.00142503e+00
4.01124775e-01 8.35034370e-01 -1.01440713e-01 6.74970269e-01
-5.64685464e-02 3.18317235e-01 4.88883525e-01 -6.33692071e-02
5.31592906e-01 8.64032209e-01 5.69693446e-01 3.44645679e-01
-6.35412931e-01 -6.66490853e-01 -3.58808696e-01 -3.08958208e-03
-1.04673874e+00 -5.18616676e-01 3.21001917e-01 -2.35858977e-01
1.22173059e+00 -3.88246864e-01 6.69395924e-01 6.78294659e-01
6.47649348e-01 -7.44997486e-02 8.03514242e-01 -7.38973975e-01
4.37963337e-01 3.73836368e-01 -2.57654756e-01 8.44979048e-01
1.67391077e-01 6.25881791e-01 -1.16047055e-01 3.31201553e-01
8.81563067e-01 -7.78502285e-01 1.90468431e-01 -1.83311999e-01
-4.16802585e-01 1.03675175e+00 5.83262205e-01 7.39896715e-01
-4.68955755e-01 7.30739713e-01 6.42442226e-01 3.81664127e-01
-2.31048260e-02 7.73140252e-01 -5.51058888e-01 -5.25278330e-01
-6.58783436e-01 4.30255175e-01 6.67726576e-01 3.46706718e-01
2.65078038e-01 6.60039127e-01 1.04935765e-01 5.13441980e-01
8.56657103e-02 5.37893251e-02 9.43552613e-01 -1.07434154e+00
4.37080652e-01 4.89131451e-01 -5.42581007e-02 -8.02461565e-01
-5.93911290e-01 -5.22878408e-01 -1.04856111e-01 8.01193595e-01
5.96445441e-01 -9.64791119e-01 -7.37119496e-01 1.43302238e+00
-5.71832322e-02 -1.29737318e-01 1.17447816e-01 4.55063403e-01
7.10846722e-01 5.00244677e-01 1.41868100e-01 2.78487295e-01
6.98439181e-01 -6.95978820e-01 -4.90511268e-01 -3.79473805e-01
6.02040887e-01 -3.22674394e-01 7.63076425e-01 4.50229973e-01
-1.28534734e+00 -6.94526613e-01 -1.16789401e+00 5.74908972e-01
-4.87211525e-01 1.25961751e-01 4.54136878e-01 7.00468004e-01
-1.00656271e+00 1.20156193e+00 -6.82952404e-01 1.18006542e-01
1.41639844e-01 1.07181907e+00 -1.12008952e-01 5.37609220e-01
-1.31582940e+00 1.26727223e+00 1.23898590e+00 -1.92710027e-01
-6.93640113e-01 -3.10487837e-01 -3.60292077e-01 4.19147968e-01
4.47690487e-01 -3.25905859e-01 1.08383727e+00 -1.24243355e+00
-1.90109134e+00 4.54361707e-01 8.36053610e-01 -7.95095921e-01
6.65765524e-01 -6.73629418e-02 -2.93632299e-01 2.34266426e-02
-1.39570966e-01 6.32776797e-01 6.60459042e-01 -4.57081228e-01
-9.87974703e-01 -6.06183782e-02 4.13156182e-01 1.20816611e-01
6.58309460e-02 -1.66529179e-01 -2.33669356e-02 -1.55260593e-01
-5.64319134e-01 -8.96428823e-01 -2.18938574e-01 -6.53999031e-01
2.67311394e-01 -5.17757237e-01 6.18817568e-01 -6.91547811e-01
1.07629192e+00 -2.45309329e+00 -1.00848623e-01 4.63954479e-01
-2.13028476e-01 4.19360906e-01 1.63546652e-01 -8.78216699e-02
-1.97932109e-01 2.42256057e-02 5.51693201e-01 8.54050696e-01
-2.82695115e-01 4.15539443e-01 2.90433705e-01 3.91210169e-02
-1.38038829e-01 5.03579736e-01 -6.85453057e-01 -5.19140102e-02
1.31938711e-01 1.81075990e-01 -7.00676918e-01 2.28755221e-01
-1.91787466e-01 -4.41066995e-02 -3.22831243e-01 -1.73479408e-01
-2.00133562e-01 2.29975179e-01 1.92999452e-01 2.08298415e-01
-2.19166398e-01 3.86720031e-01 -9.02078509e-01 8.68690312e-01
-4.82634366e-01 6.24033093e-01 -2.49650478e-01 -8.29001307e-01
1.11098742e+00 4.85793889e-01 3.54913861e-01 -1.12995255e+00
6.15989983e-01 1.99090183e-01 6.36719227e-01 -5.52673221e-01
6.27603948e-01 -1.27851322e-01 1.54788136e-01 -4.38244417e-02
2.40641251e-01 3.55380438e-02 3.12758744e-01 -1.70610353e-01
9.47217584e-01 2.83232391e-01 7.14379996e-02 -4.50082153e-01
2.29470223e-01 5.22336224e-03 2.54345834e-01 5.88898599e-01
-9.02669802e-02 -7.54486918e-02 9.81480896e-01 -1.28265351e-01
-1.19842231e+00 -3.27424616e-01 2.38629192e-01 1.44069886e+00
-1.01293221e-01 4.87714149e-02 -7.89187193e-01 -7.00614989e-01
-1.54890016e-01 9.03602421e-01 -5.26021600e-01 -7.02535331e-01
-7.48815358e-01 -5.05192935e-01 7.49761939e-01 5.31002283e-01
8.51264417e-01 -1.28229821e+00 -1.03905606e+00 2.70237178e-01
4.11651045e-01 -9.80669975e-01 1.91484168e-01 7.46275008e-01
-9.89063919e-01 -9.46526229e-01 -3.82235408e-01 -4.84510094e-01
1.55729219e-01 -5.62074363e-01 8.64709496e-01 1.79727942e-01
4.50150102e-01 5.57118794e-04 -2.32063919e-01 -5.69331944e-01
-5.66859365e-01 5.96295893e-01 -1.03922933e-01 -7.10060060e-01
-1.99580997e-01 -3.99560004e-01 -3.13465953e-01 1.68859780e-01
-7.11044788e-01 -4.44687724e-01 3.15247089e-01 5.58638930e-01
-2.37402901e-01 5.75792491e-01 5.77966332e-01 -7.70159900e-01
1.28077459e+00 -3.61429721e-01 -6.93973541e-01 -2.76209027e-01
-8.88806880e-01 5.74982166e-01 7.32014596e-01 -3.49496573e-01
-7.60851562e-01 -3.37910086e-01 -3.03329974e-01 -9.13636237e-02
2.36838117e-01 3.36275965e-01 1.42569900e-01 -1.70923740e-01
9.78800237e-01 -3.67138863e-01 1.78071588e-01 -1.93414703e-01
2.04946194e-02 2.52054125e-01 4.01530206e-01 -2.99106687e-01
2.14391604e-01 -3.23555708e-01 1.71844468e-01 -6.34496391e-01
-7.23182261e-02 -4.76115942e-02 -1.41612306e-01 -5.23640215e-01
6.84337735e-01 -4.72231776e-01 -7.73302913e-01 5.39973676e-01
-6.15047216e-01 -7.43029356e-01 -4.48788822e-01 4.63056535e-01
-4.95109111e-01 -3.55608374e-01 -2.69393682e-01 -4.65969086e-01
-2.28501365e-01 -1.01157427e+00 -1.39071107e-01 6.33147180e-01
-1.38124123e-01 -9.36273634e-01 -3.78728360e-02 8.61977488e-02
6.53473198e-01 9.43563804e-02 1.18005645e+00 -5.31652212e-01
2.30828241e-01 -3.46570760e-01 2.09077179e-01 5.17943740e-01
-4.81463790e-01 3.73853371e-02 -5.75055420e-01 5.59906624e-02
-3.39252979e-01 -1.89369440e-01 6.09542608e-01 6.04479611e-01
6.03776038e-01 -1.73391446e-01 6.64829509e-03 3.72189343e-01
1.41881299e+00 7.84596503e-01 8.56704593e-01 9.80285525e-01
2.62527466e-01 2.99821496e-01 4.50508147e-01 3.16133231e-01
-4.82524157e-01 6.54483438e-01 6.29867792e-01 3.16183746e-01
1.58845648e-01 -2.48283565e-01 4.59755599e-01 2.36710280e-01
-7.70359278e-01 -2.98551861e-02 -7.54518270e-01 2.88186193e-01
-1.40739286e+00 -1.00384593e+00 3.57068479e-01 2.14929104e+00
3.66678566e-01 8.45605731e-01 5.94100893e-01 5.88664949e-01
4.89341348e-01 5.92113361e-02 -2.38330409e-01 -1.12763619e+00
3.17016989e-01 3.89983952e-01 1.09036553e+00 4.66929376e-01
-6.71442211e-01 9.35064197e-01 7.23363972e+00 8.89144182e-01
-1.35741889e+00 -5.26894107e-02 4.55974758e-01 -9.56038982e-02
4.53759074e-01 -2.23561093e-01 -6.84688032e-01 6.99703813e-01
1.51947296e+00 -7.44711200e-04 6.52782500e-01 1.13562655e+00
3.09898496e-01 -6.18464708e-01 -4.11309659e-01 3.97679269e-01
-3.34154904e-01 -1.18292975e+00 -3.14940780e-01 2.00023368e-01
5.74860394e-01 1.65326059e-01 -9.12328511e-02 7.09458172e-01
5.16743362e-01 -1.06506169e+00 8.48725677e-01 5.51435769e-01
4.34098661e-01 -1.32914686e+00 1.14535248e+00 3.86590481e-01
-4.44998473e-01 -5.27497709e-01 -2.06449300e-01 -3.21626753e-01
-4.55731302e-01 -1.04537882e-01 -1.03602219e+00 -2.04994492e-02
6.39510751e-01 -2.80951679e-01 -5.93683481e-01 9.17613328e-01
-2.80452460e-01 8.46920133e-01 -4.87354368e-01 -3.93726259e-01
6.97615564e-01 -1.06220059e-01 2.86397576e-01 8.64803314e-01
1.51648134e-01 -3.10249869e-02 -3.61421853e-01 2.70257980e-01
-6.62644729e-02 -7.25804418e-02 -4.61961746e-01 -1.52795181e-01
1.41882479e-01 6.37616515e-01 -6.57038987e-01 -1.78904086e-01
1.67113826e-01 3.40188205e-01 1.91770807e-01 1.18892774e-01
-7.62415826e-01 -5.89482725e-01 3.00663412e-01 6.40404403e-01
6.00048006e-01 3.62743512e-02 -8.13509617e-03 -2.51074702e-01
-2.26792499e-01 -9.58975077e-01 1.48808479e-01 -7.07626820e-01
-1.56318843e-01 4.95014101e-01 5.76581769e-02 -6.96632922e-01
-8.72390866e-01 -7.68595099e-01 -7.86014080e-01 7.42393672e-01
-7.72620142e-01 -8.40888843e-02 1.42738551e-01 4.26813692e-01
2.13304773e-01 -8.66151989e-01 3.14598948e-01 1.42250627e-01
-4.93540436e-01 4.28618819e-01 3.47265482e-01 2.81050831e-01
7.26203993e-02 -1.16000164e+00 -1.91565290e-01 2.62917608e-01
-3.62171561e-01 1.02926053e-01 9.89651918e-01 -2.56136686e-01
-6.59405172e-01 -3.99218231e-01 3.45157892e-01 8.13393816e-02
3.66286606e-01 2.89413303e-01 -2.49031618e-01 3.77612621e-01
2.66109556e-01 -2.43358582e-01 -2.44554326e-01 -3.02802846e-02
3.67872834e-01 -3.49718660e-01 -1.16371036e+00 4.01697665e-01
4.14176315e-01 -5.73962070e-02 -3.40225786e-01 -3.44946027e-01
2.49943987e-01 -4.96918857e-01 -5.15815914e-01 2.69539386e-01
5.94178617e-01 -1.20914578e+00 7.84102619e-01 -7.95036733e-01
7.34918177e-01 2.86332667e-01 3.30962569e-01 -1.52230656e+00
-4.69813168e-01 1.94524676e-02 -3.03146243e-02 6.35799229e-01
6.21233344e-01 -5.00946283e-01 1.04055715e+00 4.52901930e-01
1.95532128e-01 -1.01849043e+00 -8.56456935e-01 -4.16587085e-01
1.33839488e-01 1.08742468e-01 2.40648657e-01 5.35833359e-01
7.99772814e-02 4.81454849e-01 -1.23140708e-01 -3.99115652e-01
-2.15791479e-01 -8.00878286e-01 4.58685815e-01 -1.10457158e+00
-7.07647324e-01 -4.95331049e-01 -7.19667375e-01 -4.85006183e-01
-1.52987063e-01 -4.81992573e-01 -2.04671785e-01 -1.18715358e+00
-5.99503934e-01 -3.74307334e-01 -2.72867203e-01 4.49353993e-01
1.52486488e-01 -2.65462756e-01 2.53360659e-01 -1.86200619e-01
-1.77548945e-01 2.49236338e-02 8.65986824e-01 6.74350411e-02
-5.92733622e-01 3.50362599e-01 -3.43460768e-01 9.71403778e-01
1.09939945e+00 -6.00151837e-01 -3.88089746e-01 -1.44912660e-01
6.31182015e-01 3.62271696e-01 -1.46355242e-01 -1.31996000e+00
-6.64856434e-02 -1.34280622e-02 5.14638722e-01 1.49070516e-01
1.22645516e-02 -9.10113454e-01 5.75848222e-02 9.07356739e-01
-3.64320099e-01 4.64860499e-01 4.34301078e-01 1.42554745e-01
-2.97169127e-02 -1.00282574e+00 7.85627902e-01 -3.58888268e-01
-8.35466623e-01 -4.27522063e-01 -5.66421807e-01 3.14033367e-02
9.85252082e-01 -3.17629755e-01 1.18308231e-01 -5.11232078e-01
-6.72886431e-01 3.39420550e-02 -8.82420614e-02 2.96344429e-01
-1.13823064e-01 -7.36462593e-01 -1.53375611e-01 -8.67320225e-02
-6.57929420e-01 -4.25834864e-01 -2.90458977e-01 4.88734871e-01
-9.42714036e-01 4.81822312e-01 -9.24299538e-01 8.94187242e-02
-1.43020332e+00 2.55697906e-01 1.12353802e+00 -6.93521559e-01
-3.35327536e-01 4.39295202e-01 -6.52027309e-01 -1.74491882e-01
3.96106631e-01 -1.36162177e-01 -6.53329194e-01 -4.53336835e-02
-6.47993002e-04 7.14044690e-01 1.72627494e-01 -1.02521427e-01
1.36423841e-01 5.64174578e-02 -5.16475700e-02 -4.10329372e-01
1.31694758e+00 5.15682161e-01 4.06429440e-01 3.83267820e-01
5.05767941e-01 -2.16459379e-01 -1.14227843e+00 4.22646940e-01
1.44877076e-01 -6.26102537e-02 7.14781046e-01 -1.07205188e+00
-1.30102563e+00 6.25216365e-01 9.35731947e-01 7.79062286e-02
7.87035465e-01 -5.55251956e-01 2.20516697e-01 3.61389101e-01
4.09990221e-01 -1.46310365e+00 -2.92210821e-02 6.04945600e-01
4.41040397e-01 -8.78019512e-01 7.97011182e-02 2.99430549e-01
-6.20229304e-01 1.24143302e+00 6.84894204e-01 -4.99988407e-01
4.36594278e-01 3.54790986e-01 -1.70739010e-01 -5.61877117e-02
-5.25536537e-01 -1.78468449e-03 -2.05656528e-01 3.74410242e-01
3.28947365e-01 5.82670197e-02 -8.04830670e-01 4.19399798e-01
-3.33441675e-01 4.54073220e-01 4.97140348e-01 8.85255873e-01
-7.51896203e-01 -9.42065716e-01 -3.60921383e-01 4.58358139e-01
-8.11256707e-01 2.54039258e-01 -2.86302686e-01 1.09149563e+00
6.71651363e-01 5.78122675e-01 -2.19614320e-02 -4.26749617e-01
5.52763045e-01 2.07399458e-01 4.69201863e-01 -1.81646347e-01
-1.23101687e+00 -3.64711761e-01 5.24351299e-01 -2.27362379e-01
1.07535154e-01 -1.31989419e-01 -1.48473501e+00 -7.03749537e-01
-5.18311679e-01 4.32203770e-01 8.37928057e-01 1.16168785e+00
-4.06544246e-02 9.11977708e-01 2.79611290e-01 -4.00447875e-01
-6.39084220e-01 -1.13315880e+00 -6.82331502e-01 -7.23726973e-02
-1.42861173e-01 -5.93331039e-01 -2.70333439e-01 -5.68671107e-01] | [3.511537551879883, 1.476969838142395] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.