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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
b1e0604b-291f-48a5-bab0-616a6ee547bd | deep-generative-models-for-decision-making | 2306.08810 | null | https://arxiv.org/abs/2306.08810v1 | https://arxiv.org/pdf/2306.08810v1.pdf | Deep Generative Models for Decision-Making and Control | Deep model-based reinforcement learning methods offer a conceptually simple approach to the decision-making and control problem: use learning for the purpose of estimating an approximate dynamics model, and offload the rest of the work to classical trajectory optimization. However, this combination has a number of empirical shortcomings, limiting the usefulness of model-based methods in practice. The dual purpose of this thesis is to study the reasons for these shortcomings and to propose solutions for the uncovered problems. Along the way, we highlight how inference techniques from the contemporary generative modeling toolbox, including beam search, classifier-guided sampling, and image inpainting, can be reinterpreted as viable planning strategies for reinforcement learning problems. | ['Michael Janner'] | 2023-06-15 | null | null | null | null | ['image-inpainting', 'model-based-reinforcement-learning'] | ['computer-vision', 'reasoning'] | [ 4.50362749e-02 1.41234249e-01 -4.99482423e-01 1.45672336e-01
-5.77397645e-01 -3.65243554e-01 8.14637959e-01 -2.58096337e-01
-1.59484565e-01 1.07288361e+00 4.18349504e-02 -5.21765172e-01
-7.56367266e-01 -7.41129577e-01 -3.77068609e-01 -9.24056232e-01
2.03205153e-01 6.55314803e-01 -3.16597551e-01 -2.84444034e-01
6.79101706e-01 7.02583075e-01 -1.32830441e+00 -1.25151917e-01
6.13424301e-01 5.55148244e-01 3.13364476e-01 6.61516309e-01
-7.06566572e-02 1.20944786e+00 -5.07970273e-01 -6.37473492e-03
-1.17349036e-01 -9.25026000e-01 -7.79560804e-01 4.08632398e-01
-1.17316432e-01 -6.25382423e-01 -3.60334098e-01 5.25790155e-01
5.67978621e-01 5.41563869e-01 8.18750858e-01 -1.16701353e+00
-1.53589472e-01 2.56416202e-01 -2.06397295e-01 2.94635952e-01
3.83226693e-01 5.10746241e-01 5.42952061e-01 -5.17114699e-01
6.76306844e-01 1.22125924e+00 5.17016649e-01 6.38511717e-01
-1.38485324e+00 -3.36542249e-01 7.28068203e-02 3.13880980e-01
-1.06144464e+00 -4.83369797e-01 7.13693738e-01 -5.68008065e-01
1.17370403e+00 7.25645870e-02 1.15469444e+00 1.22046316e+00
4.86710131e-01 7.75004566e-01 1.41295290e+00 -5.20758212e-01
5.32243252e-01 5.20343035e-02 -4.43555206e-01 8.09906900e-01
-1.47005811e-01 9.52454388e-01 -5.38340509e-01 -2.22682789e-01
1.11763608e+00 -2.90819168e-01 4.58764806e-02 -5.48087597e-01
-7.77631640e-01 1.30860996e+00 1.33847535e-01 3.31546292e-02
-3.67835790e-01 4.72712576e-01 1.81184232e-01 2.48834178e-01
4.74661738e-01 6.47911787e-01 -4.56874333e-02 -4.09128040e-01
-1.11867058e+00 1.03472316e+00 6.44768715e-01 5.09929240e-01
6.43905103e-01 5.95817864e-01 -3.68855558e-02 4.90042597e-01
6.81819618e-01 2.12363094e-01 3.90970588e-01 -1.58857107e+00
2.41479147e-02 -3.91386859e-02 3.84640366e-01 -6.37572169e-01
-2.91473567e-01 -4.45084989e-01 -3.07043254e-01 8.52852345e-01
4.96477842e-01 -3.10731113e-01 -6.90135062e-01 1.34358597e+00
4.33124751e-01 1.48963690e-01 -1.47928029e-01 5.81935942e-01
7.93323964e-02 6.56495512e-01 6.83290735e-02 -5.97595572e-01
4.60651308e-01 -1.05052423e+00 -6.90813541e-01 -1.48160025e-01
4.67033327e-01 -6.01174355e-01 9.47881460e-01 4.84776884e-01
-1.37104177e+00 -5.19605577e-01 -8.90600622e-01 3.65685105e-01
-1.29661813e-01 -1.83068156e-01 8.88105452e-01 5.67776084e-01
-9.59722281e-01 1.22626638e+00 -1.23788321e+00 -2.36731634e-01
3.15221161e-01 2.63622522e-01 3.67662817e-01 1.86335608e-01
-6.82629764e-01 1.39554977e+00 2.76905715e-01 6.84574991e-02
-1.17528021e+00 -6.71747208e-01 -6.47256255e-01 -1.47017270e-01
4.26723719e-01 -9.05542910e-01 1.57888377e+00 -7.03020394e-01
-1.93817484e+00 2.69690931e-01 -1.74421638e-01 -4.38268244e-01
7.77784526e-01 -2.24141907e-02 2.65094668e-01 5.71661890e-02
7.63051286e-02 6.18681788e-01 1.04179728e+00 -1.19924295e+00
-5.99215150e-01 -4.95246649e-02 1.61327813e-02 4.53330070e-01
2.66303390e-01 -1.96328282e-01 1.87453970e-01 -5.93998075e-01
-7.04545900e-02 -9.67459500e-01 -5.13850033e-01 -1.55302241e-01
4.32664715e-02 -2.99901873e-01 5.62071741e-01 -4.38618302e-01
1.14451063e+00 -1.64845538e+00 5.34994423e-01 4.52483036e-02
-2.07269758e-01 -1.92544069e-02 3.39462310e-02 8.70196462e-01
8.73264372e-02 -8.42010379e-02 -1.84407756e-01 -2.77848601e-01
3.41807902e-02 4.70887989e-01 -6.73576117e-01 3.55735809e-01
1.14748850e-01 9.06797945e-01 -1.04610062e+00 -4.84009773e-01
7.30042398e-01 2.11545527e-01 -3.80014211e-01 2.16195919e-03
-4.96569186e-01 6.63001657e-01 -4.60353106e-01 5.00843942e-01
5.89979552e-02 2.97501206e-01 2.34200791e-01 3.17949444e-01
-1.84924141e-01 4.41859424e-01 -1.04587519e+00 1.58868814e+00
-3.35807264e-01 6.73225164e-01 4.57971916e-02 -1.24180889e+00
5.82827151e-01 2.43323565e-01 7.46425390e-01 -5.22786260e-01
2.50317045e-02 -1.09587135e-02 1.81721523e-02 -6.59615993e-01
3.37886631e-01 -5.13136923e-01 3.59741360e-01 6.89363539e-01
6.72249943e-02 -6.59652710e-01 3.04709733e-01 -2.14719921e-01
8.12851965e-01 1.06882417e+00 7.04604238e-02 -1.28669187e-01
1.35694575e-02 4.68643725e-01 2.27569342e-01 8.49757016e-01
-8.46680701e-02 1.39028147e-01 4.20877784e-01 -4.50830013e-01
-1.15135550e+00 -9.65861917e-01 1.01094935e-02 8.26642454e-01
-3.46408397e-01 -3.49913746e-01 -8.00246418e-01 -3.89451802e-01
1.39728442e-01 1.05959344e+00 -5.68265557e-01 -9.44240317e-02
-7.15684175e-01 -8.08198571e-01 1.88179210e-01 5.10559916e-01
1.82880148e-01 -1.17262614e+00 -9.39602613e-01 5.95239699e-01
4.46578972e-02 -4.20891434e-01 2.82082915e-01 2.02809334e-01
-1.23809469e+00 -1.04495966e+00 -4.34743375e-01 -2.60378271e-01
2.75581628e-01 -1.06734186e-02 1.02212262e+00 -1.79196540e-02
-2.10817769e-01 7.37610638e-01 -4.17581312e-02 -3.36898714e-01
-6.41701579e-01 -9.01899114e-02 -1.94404826e-01 -5.93852162e-01
-2.42640272e-01 -5.47931850e-01 -3.75141919e-01 1.06238108e-02
-5.91651738e-01 5.08918986e-02 1.99238181e-01 1.05812514e+00
6.33412480e-01 2.00685203e-01 5.48729241e-01 -6.35506511e-01
1.03376365e+00 -3.47405344e-01 -6.19295120e-01 1.87138543e-02
-1.04075468e+00 5.17817028e-02 2.19192460e-01 -5.45943201e-01
-1.12465847e+00 -2.94233877e-02 -2.23462850e-01 -5.66139758e-01
-7.34978989e-02 5.47852337e-01 3.08555305e-01 -1.65320337e-01
4.53082770e-01 4.84583586e-01 4.45455849e-01 -1.80502117e-01
2.51498163e-01 -1.70576256e-02 8.66241530e-02 -6.64213359e-01
5.40211439e-01 3.03235143e-01 3.56147051e-01 -8.88891995e-01
-6.48478210e-01 1.43123269e-01 -3.32479388e-01 -4.33527589e-01
5.73658705e-01 -6.30499840e-01 -5.98839343e-01 1.46914065e-01
-7.45614886e-01 -9.16725099e-01 -6.17961168e-01 4.61881995e-01
-1.49754059e+00 8.66703615e-02 -6.12638533e-01 -1.15062463e+00
1.24197073e-01 -1.31990516e+00 7.23580480e-01 2.92576641e-01
-5.23675978e-01 -1.30966365e+00 3.68978858e-01 2.72922367e-01
4.22574371e-01 3.24570030e-01 9.73973334e-01 -8.19394290e-02
-6.88284039e-01 2.57357005e-02 5.41355908e-01 3.72714251e-02
-1.46199897e-01 2.49243289e-01 -6.92322254e-01 -3.13915074e-01
2.39581242e-01 -4.88442957e-01 6.25398278e-01 9.16288257e-01
8.62042069e-01 -2.54263967e-01 -3.40561330e-01 4.19071496e-01
1.42190707e+00 5.27380526e-01 5.15927196e-01 6.33712232e-01
2.01787010e-01 5.88811100e-01 8.64298701e-01 6.20299399e-01
2.36565620e-01 6.30408585e-01 3.72656643e-01 3.62493336e-01
4.61109132e-02 -4.27961618e-01 2.39766106e-01 3.10087562e-01
-1.90302297e-01 -2.53509264e-04 -6.50348783e-01 1.84366375e-01
-1.94130945e+00 -1.39625931e+00 2.16854349e-01 1.84393358e+00
6.02155566e-01 1.12761512e-01 3.66688311e-01 5.50048016e-02
1.56305760e-01 1.01890609e-01 -5.93610227e-01 -5.52184463e-01
5.02281249e-01 3.61236632e-01 2.78495789e-01 6.95910096e-01
-8.38755727e-01 8.58246207e-01 8.16760826e+00 9.12434876e-01
-7.86121011e-01 2.78376993e-02 5.43828666e-01 -3.10812682e-01
-2.18261644e-01 3.46825153e-01 -7.04235137e-01 4.65428084e-01
1.13231826e+00 -1.36253625e-01 1.15087783e+00 7.81020403e-01
7.89297581e-01 -5.31867504e-01 -1.03320575e+00 7.01752901e-01
-1.72223762e-01 -1.54291606e+00 -2.31747851e-01 2.11248785e-01
7.53132284e-01 -3.54540557e-01 2.01725379e-01 5.07843554e-01
3.31411928e-01 -1.15495062e+00 8.99859726e-01 7.38218069e-01
1.15647420e-01 -1.00653279e+00 1.52785107e-01 6.06465280e-01
-6.74231350e-01 -2.94617236e-01 -2.74539828e-01 -6.22944176e-01
2.22042054e-01 -5.82079850e-02 -5.65481961e-01 2.59438515e-01
2.71814644e-01 4.62766796e-01 3.04961484e-02 1.05545664e+00
-2.90872574e-01 6.42507732e-01 -1.30634144e-01 -1.49913326e-01
4.87840831e-01 -5.21412551e-01 5.31093776e-01 8.05272281e-01
1.27446517e-01 -1.16468772e-01 2.48953640e-01 1.25984859e+00
6.88498616e-01 -4.06423777e-01 -8.12806308e-01 -1.68267250e-01
4.45079505e-01 8.55169713e-01 -8.92227411e-01 -2.07243145e-01
-1.26192318e-02 6.19491398e-01 3.08554113e-01 4.79102552e-01
-8.74910653e-01 4.18762537e-03 6.55277371e-01 6.14797547e-02
1.51717648e-01 -5.70130706e-01 -4.11031216e-01 -8.27182710e-01
-5.54230511e-01 -1.03084350e+00 -4.01231647e-02 -7.68066049e-01
-5.43583512e-01 -1.43900514e-01 5.97591102e-01 -6.73037887e-01
-1.05725491e+00 -4.47727710e-01 -7.57119298e-01 8.41743112e-01
-1.22343516e+00 -8.29717398e-01 1.58543929e-01 2.30114341e-01
1.01775718e+00 -1.66751325e-01 8.67618918e-01 -2.40248263e-01
-5.63170314e-01 -7.33825192e-02 3.84068161e-01 -4.32494402e-01
5.09260148e-02 -1.33548951e+00 6.67438880e-02 4.78688419e-01
1.82264641e-01 3.82188141e-01 9.85824168e-01 -5.63539207e-01
-1.50603890e+00 -5.58965445e-01 4.14688051e-01 -3.65065426e-01
6.18665159e-01 2.28441134e-01 -4.48470712e-01 5.93733370e-01
3.58541965e-01 -7.15268373e-01 3.82001370e-01 -1.71123400e-01
5.31485915e-01 2.00038299e-01 -1.00354600e+00 6.36579394e-01
5.78107357e-01 -3.47706318e-01 -3.44319344e-01 4.39763427e-01
2.48167813e-02 -6.21184528e-01 -7.52761304e-01 -1.24743991e-02
5.69675803e-01 -9.93103147e-01 1.10097992e+00 -6.51330769e-01
4.85721946e-01 3.32288183e-02 2.35498443e-01 -1.55507815e+00
-3.71928155e-01 -9.88097072e-01 -3.86884034e-01 9.38089490e-01
-3.37633193e-02 -4.13565695e-01 9.26933348e-01 5.62867939e-01
-1.70318022e-01 -1.18472552e+00 -9.26367104e-01 -6.49581969e-01
4.93766189e-01 -4.16718513e-01 2.58488685e-01 6.77044630e-01
-6.78051636e-02 2.17335433e-01 -3.31135213e-01 -4.79840517e-01
7.54612982e-01 2.80050159e-01 6.44437790e-01 -7.54179835e-01
-5.68630397e-01 -6.21330917e-01 1.30713627e-01 -8.59474540e-01
2.45990455e-01 -4.24355000e-01 -1.22378757e-02 -1.75851107e+00
-1.80938274e-01 -4.06465501e-01 1.13206081e-01 2.67851353e-01
2.62816310e-01 -1.65652603e-01 3.03214490e-01 1.97883144e-01
6.03857115e-02 6.36296272e-01 1.45266497e+00 -4.47607599e-02
-3.64093393e-01 2.76210368e-01 -3.89636517e-01 8.31583381e-01
1.06036532e+00 -4.38309878e-01 -8.88384044e-01 -2.10252345e-01
2.28876263e-01 6.78361714e-01 6.75969303e-01 -7.75818110e-01
-2.96800286e-02 -8.56674969e-01 5.47577918e-01 -5.29257894e-01
4.94262904e-01 -5.35568833e-01 1.75324589e-01 8.00037742e-01
-5.06698489e-01 3.16560149e-01 2.39048958e-01 5.64301372e-01
1.34560972e-01 -7.14346230e-01 7.78622031e-01 -5.17976046e-01
-5.29706359e-01 -7.29528591e-02 -1.07128358e+00 -1.40779942e-01
1.05927432e+00 -4.30221111e-01 1.72879845e-01 -5.77629685e-01
-1.00104654e+00 1.27931191e-02 3.51533562e-01 -8.03051814e-02
5.38846016e-01 -1.08643198e+00 -3.50152940e-01 1.09791001e-02
-6.79092526e-01 -2.77392566e-01 1.55147100e-02 8.04188311e-01
-5.79010427e-01 5.29455364e-01 -3.41244102e-01 -2.77253628e-01
-7.79853523e-01 4.72565472e-01 6.85976565e-01 -4.26452309e-01
-5.49578369e-01 4.88848567e-01 -3.49219292e-01 -7.05355108e-02
1.93480790e-01 -1.58968091e-01 3.18280160e-02 7.77191892e-02
8.89153853e-02 7.53568411e-01 -2.65531898e-01 -1.35468498e-01
6.77888840e-02 2.34501019e-01 2.13781446e-01 -5.85439682e-01
1.22288191e+00 -1.82483181e-01 3.00028294e-01 3.24499786e-01
4.37776148e-01 -5.43571472e-01 -1.60830164e+00 3.78260672e-01
-7.95963481e-02 -4.08148468e-01 3.04078281e-01 -7.66772091e-01
-6.42008185e-01 8.49617004e-01 4.73737359e-01 1.84642851e-01
9.16706800e-01 -2.96844274e-01 2.20738992e-01 3.75759006e-01
1.99748367e-01 -1.33858097e+00 1.84164196e-01 3.89342248e-01
8.30137849e-01 -7.29450047e-01 4.01755065e-01 1.49206311e-01
-6.68313384e-01 1.23496938e+00 6.27569258e-01 -4.52822864e-01
3.80810857e-01 3.56929958e-01 -2.72231013e-01 -4.20182943e-02
-9.52206969e-01 -2.32229933e-01 -1.83120698e-01 8.47785413e-01
2.80631572e-01 -2.32229903e-01 -5.65237939e-01 -1.91774055e-01
-1.92238286e-01 1.59119770e-01 4.67418760e-01 1.17857909e+00
-5.78858435e-01 -1.36892438e+00 -3.51599813e-01 2.86357105e-01
-1.87196746e-01 3.08859468e-01 -4.33279760e-02 1.12169337e+00
3.37387659e-02 9.11493421e-01 6.42273249e-03 -1.36991382e-01
-4.34638076e-02 1.98267788e-01 1.11808193e+00 -3.08398902e-01
-5.09625971e-01 5.44318676e-01 1.39676807e-02 -6.09829664e-01
-4.86832917e-01 -9.49658394e-01 -9.58871543e-01 -3.46001446e-01
-2.12499171e-01 -1.28639173e-02 6.16686583e-01 1.23802531e+00
1.19206652e-01 6.32429600e-01 3.79222780e-01 -1.20300055e+00
-1.24974322e+00 -9.08384144e-01 -4.27984387e-01 -1.48182049e-01
9.20830891e-02 -1.02051353e+00 7.78523311e-02 -9.12101418e-02] | [4.190046787261963, 2.147531509399414] |
0211c9c7-83c3-474b-879b-0902d1fcd130 | automating-vitiligo-skin-lesion-segmentation | 1912.08350 | null | https://arxiv.org/abs/1912.08350v1 | https://arxiv.org/pdf/1912.08350v1.pdf | Automating Vitiligo Skin Lesion Segmentation Using Convolutional Neural Networks | For several skin conditions such as vitiligo, accurate segmentation of lesions from skin images is the primary measure of disease progression and severity. Existing methods for vitiligo lesion segmentation require manual intervention. Unfortunately, manual segmentation is time and labor-intensive, as well as irreproducible between physicians. We introduce a convolutional neural network (CNN) that quickly and robustly performs vitiligo skin lesion segmentation. Our CNN has a U-Net architecture with a modified contracting path. We use the CNN to generate an initial segmentation of the lesion, then refine it by running the watershed algorithm on high-confidence pixels. We train the network on 247 images with a variety of lesion sizes, complexity, and anatomical sites. The network with our modifications noticeably outperforms the state-of-the-art U-Net, with a Jaccard Index (JI) score of 73.6% (compared to 36.7%). Moreover, our method requires only a few seconds for segmentation, in contrast with the previously proposed semi-autonomous watershed approach, which requires 2-29 minutes per image. | ['Makena Low', 'Priyanka Raina'] | 2019-12-16 | null | null | null | null | ['skin-lesion-segmentation'] | ['medical'] | [ 6.83914602e-01 1.74977079e-01 -3.21689367e-01 -1.37507126e-01
-8.72066498e-01 -4.95097488e-01 9.96498093e-02 2.68274873e-01
-5.24056077e-01 5.62092006e-01 -2.75056511e-01 -3.95591110e-01
1.09903164e-01 -9.43232238e-01 -3.58515590e-01 -8.15149784e-01
1.30055025e-01 5.46250880e-01 3.50355208e-01 -2.53122821e-02
3.24260354e-01 5.34580767e-01 -1.05309808e+00 6.56024143e-02
1.38057721e+00 9.97884512e-01 -8.84542614e-02 1.03106117e+00
-2.90322006e-01 3.52799326e-01 -5.57278931e-01 -3.83223504e-01
2.47295246e-01 -7.39103556e-01 -9.14166510e-01 2.76724905e-01
7.14086294e-01 -2.91121602e-01 -2.18507558e-01 9.38282192e-01
4.99572247e-01 -2.18273818e-01 7.62281954e-01 -6.55768156e-01
-5.14751196e-01 2.45997444e-01 -9.97418284e-01 4.13782522e-02
-1.14901282e-01 2.53167331e-01 6.21926188e-01 -2.92430490e-01
9.69852388e-01 7.13574469e-01 1.00371826e+00 7.05228925e-01
-1.16613102e+00 -3.33455354e-01 -4.00516093e-01 8.76490995e-02
-1.41870713e+00 2.60626692e-02 1.94620132e-01 -5.55964351e-01
5.89405835e-01 3.38013530e-01 9.31052208e-01 5.09524345e-01
3.30849081e-01 7.37378895e-01 1.23536015e+00 -3.65610361e-01
2.90630817e-01 -3.74062061e-01 -1.70896322e-01 9.03758943e-01
1.87745601e-01 -3.94282699e-01 5.37635498e-02 8.63056853e-02
1.21171415e+00 1.59454886e-02 -9.72277746e-02 -4.36976738e-02
-9.20533597e-01 8.34221363e-01 6.74446046e-01 1.97120771e-01
-5.33700168e-01 2.06187770e-01 3.41583014e-01 -6.57306239e-02
3.92081738e-01 3.29959601e-01 8.12980309e-02 1.19000236e-02
-1.42636049e+00 -2.06244171e-01 5.11911333e-01 3.00365239e-01
4.35410738e-01 -2.73003846e-01 -2.21961901e-01 8.62928331e-01
-1.59475178e-01 7.55620450e-02 2.58850664e-01 -8.82822752e-01
-3.98781262e-02 7.49017358e-01 -3.60219121e-01 -6.63317204e-01
-3.81714255e-01 -4.09499526e-01 -1.22595119e+00 4.94467199e-01
9.41274822e-01 -3.50071818e-01 -1.74199188e+00 1.04039061e+00
5.08722603e-01 2.04895198e-01 -4.53360498e-01 9.82569039e-01
7.11194813e-01 1.53152436e-01 1.03258632e-01 1.57594327e-02
1.17512405e+00 -1.27116978e+00 -5.82426310e-01 1.64425634e-02
5.68814635e-01 -7.99873471e-01 6.97864592e-01 4.89754856e-01
-1.23216021e+00 -4.40225564e-02 -8.85447860e-01 -1.61893815e-01
-2.22655192e-01 1.95830092e-01 5.70379674e-01 7.77329087e-01
-1.33077145e+00 8.36659074e-01 -9.08071458e-01 -7.01811612e-01
9.43907440e-01 4.23114717e-01 -4.48270291e-01 -1.86277777e-01
-6.58465147e-01 8.96483958e-01 2.47561753e-01 7.10729510e-02
-4.69146043e-01 -8.47861826e-01 -7.57852197e-01 -8.21019933e-02
4.59856629e-01 -6.40956402e-01 1.12479472e+00 -8.22593808e-01
-1.51474607e+00 1.04783285e+00 -2.34931812e-01 -4.13427949e-01
9.84064579e-01 2.33352736e-01 -1.13470428e-01 7.47756481e-01
-9.68356058e-02 8.11147034e-01 5.98907650e-01 -9.42650735e-01
-7.55472898e-01 -1.86861157e-01 -2.55434096e-01 5.56286760e-02
1.95763499e-01 -2.79495846e-02 -8.05077195e-01 -2.87027121e-01
2.14140475e-01 -7.02050924e-01 -7.39966452e-01 6.51241660e-01
-5.19597948e-01 -9.69178900e-02 4.03715730e-01 -1.10301757e+00
1.15543079e+00 -1.63219929e+00 -2.15212584e-01 6.08298719e-01
6.65076613e-01 6.47019684e-01 -5.89098223e-02 1.58098951e-01
-7.35059828e-02 5.42252481e-01 -6.96102917e-01 -6.91192821e-02
-4.50814068e-01 -5.07916436e-02 4.75387096e-01 6.36790931e-01
2.34304681e-01 1.08033705e+00 -9.18842793e-01 -8.20704818e-01
3.80980223e-01 6.34882152e-01 -5.10772705e-01 -1.47720546e-01
-1.94912210e-01 2.09556341e-01 -2.39109322e-01 1.08632767e+00
6.79834425e-01 -6.08274996e-01 2.92766213e-01 1.11284228e-02
-2.10914705e-02 -1.50128216e-01 -1.04666340e+00 1.64391923e+00
-2.49921709e-01 6.84930682e-01 3.20948303e-01 -7.27707684e-01
6.71891034e-01 2.27643058e-01 7.69338965e-01 -5.29372811e-01
3.36171389e-01 3.40932637e-01 4.38095108e-02 -5.16880929e-01
1.07177198e-01 -5.89427091e-02 5.17895639e-01 3.98826808e-01
-1.38766035e-01 -2.40905046e-01 6.31139576e-01 9.01200920e-02
1.14909017e+00 -9.69626829e-02 4.76066530e-01 -2.77295988e-03
4.01650786e-01 3.82718325e-01 3.58190566e-01 6.50306582e-01
-6.31992519e-01 1.02901578e+00 8.44478786e-01 -5.45295894e-01
-1.11350834e+00 -9.19683158e-01 -2.75791407e-01 4.75547820e-01
6.42029569e-02 -5.33843115e-02 -1.22152102e+00 -7.98994184e-01
-1.51579399e-02 1.24527290e-01 -8.74858022e-01 3.90895277e-01
-4.83217806e-01 -7.96791375e-01 5.53386569e-01 4.81179833e-01
5.07569551e-01 -9.10616875e-01 -2.40666211e-01 3.34068716e-01
-5.35381697e-02 -5.83850265e-01 -5.18413544e-01 -2.68869579e-01
-7.42835462e-01 -1.53407931e+00 -1.17593300e+00 -9.25602913e-01
1.13383174e+00 6.79945946e-03 8.66810858e-01 3.81151110e-01
-1.13701391e+00 -1.44135594e-01 6.27600923e-02 -1.17395185e-02
-3.62927467e-01 1.30628377e-01 -4.40875441e-01 -6.87959269e-02
2.74258703e-01 -2.81894505e-01 -1.11212409e+00 8.18498433e-02
-1.09843493e+00 1.33893192e-01 8.53793919e-01 8.84282827e-01
1.04185820e+00 -7.82444030e-02 3.16445440e-01 -1.04461801e+00
6.41327858e-01 -2.71756440e-01 -4.25400287e-01 4.53253388e-01
-5.55784941e-01 -5.82597196e-01 3.14489335e-01 -2.51548350e-01
-8.38612795e-01 3.81486177e-01 -3.79859924e-01 -2.39679515e-01
-4.60657597e-01 3.61050487e-01 6.48177147e-01 -5.84370852e-01
8.93317461e-01 2.61917394e-02 4.69755679e-01 -2.65803993e-01
4.50493157e-01 6.25376403e-01 7.30830371e-01 7.70141836e-04
5.77348173e-01 6.94390476e-01 2.47571662e-01 -8.02888930e-01
-6.73045516e-01 -7.02051699e-01 -8.26065898e-01 -3.53042305e-01
9.94116485e-01 -3.00868481e-01 -6.41066253e-01 7.15443015e-01
-9.37380433e-01 -4.97107536e-01 -1.11988120e-01 2.15543434e-01
-3.93207133e-01 6.14528596e-01 -1.03334188e+00 -3.04097205e-01
-6.46308124e-01 -1.07936871e+00 7.90026903e-01 7.46969342e-01
-1.16994500e-01 -1.25603426e+00 1.45519748e-01 6.74532235e-01
5.31705141e-01 8.04751873e-01 7.34914243e-01 -1.74995229e-01
-3.40113878e-01 -5.81400931e-01 -6.52810812e-01 2.11551651e-01
2.66889215e-01 4.46449697e-01 -6.88362479e-01 -1.60047352e-01
-5.91553926e-01 -2.90107787e-01 1.14742839e+00 9.49225605e-01
1.43148279e+00 -3.78408358e-02 -3.85337472e-01 8.77213478e-01
1.70537055e+00 2.12569430e-01 8.46970141e-01 1.05671316e-01
5.45188367e-01 6.04745924e-01 4.11350489e-01 7.09524453e-02
1.74789146e-01 -1.70045644e-02 4.14685756e-01 -9.69478130e-01
-3.97046268e-01 1.96475461e-02 -4.64199126e-01 3.36795568e-01
-4.20016736e-01 -6.16304390e-02 -1.11712110e+00 8.38441908e-01
-1.54242218e+00 -6.42456889e-01 -3.11916739e-01 1.93351316e+00
1.05476141e+00 -5.38616963e-02 2.53828973e-01 -1.81562468e-01
1.00646853e+00 -1.19558014e-01 -6.31523907e-01 -4.54258174e-01
1.56256646e-01 7.41553307e-01 8.06465268e-01 4.98964757e-01
-1.02629626e+00 1.08311796e+00 7.20777988e+00 9.48926389e-01
-1.48280835e+00 -1.66411102e-01 9.84262764e-01 -8.22703447e-03
5.76434955e-02 -2.45469227e-01 -2.68752545e-01 1.87103197e-01
4.13260013e-01 -3.37836370e-02 2.25155070e-01 4.76871312e-01
1.65684402e-01 -3.73527080e-01 -5.92011869e-01 6.52767003e-01
8.63427669e-02 -1.81205082e+00 -2.29080707e-01 1.80564895e-01
1.02534211e+00 -2.46422775e-02 -4.98659648e-02 -3.00781518e-01
2.95117080e-01 -1.51053965e+00 -2.84246862e-01 5.86220562e-01
1.36357605e+00 -6.66044772e-01 8.41669858e-01 -7.05114827e-02
-9.12527621e-01 5.03929496e-01 -1.82984903e-01 3.52945298e-01
2.76979089e-01 6.66347563e-01 -1.12781811e+00 2.54267663e-01
3.94665122e-01 3.02275598e-01 -5.70327699e-01 1.69585061e+00
-2.87730277e-01 8.20630014e-01 -3.90114933e-01 8.24612156e-02
5.41995049e-01 -3.46574187e-01 1.82314664e-01 1.17657959e+00
1.98023155e-01 -1.34145636e-02 -7.72215575e-02 8.40323448e-01
-6.03372492e-02 4.90789801e-01 -1.62557214e-01 -2.53279090e-01
4.84877937e-02 1.61460900e+00 -1.13260508e+00 -4.12573844e-01
-1.39968887e-01 1.00364041e+00 5.36526069e-02 3.61935139e-01
-5.49255013e-01 -8.27097833e-01 3.60623300e-01 1.06134295e-01
2.18776941e-01 5.27032539e-02 -6.08611107e-01 -7.29313493e-01
-1.81072772e-01 -5.56671739e-01 5.89340687e-01 -3.90680254e-01
-1.11061323e+00 4.20412362e-01 -8.07711780e-01 -8.70759785e-01
-9.38666612e-02 -6.71420157e-01 -1.05444682e+00 9.89406288e-01
-1.77911353e+00 -1.15375710e+00 -5.59369981e-01 3.47136557e-01
3.33043814e-01 2.24033281e-01 7.07102537e-01 1.38270244e-01
-7.83621013e-01 5.40919840e-01 5.96847087e-02 3.64753664e-01
9.07541573e-01 -1.49192894e+00 4.02373701e-01 7.53989875e-01
-5.69406688e-01 4.43133801e-01 2.03889012e-01 -8.51470530e-01
-8.05054128e-01 -1.14315879e+00 8.24009061e-01 7.75997266e-02
6.70411587e-01 1.46037430e-01 -8.29841733e-01 3.89266968e-01
3.58105481e-01 1.29123881e-01 9.14502501e-01 -1.54288292e-01
-1.22327276e-01 1.58639163e-01 -1.58778691e+00 8.07097018e-01
5.80475569e-01 -9.51968953e-02 -5.07091433e-02 7.45875359e-01
2.71842659e-01 -7.07045376e-01 -1.23726654e+00 1.76691800e-01
5.84425926e-01 -8.53597999e-01 8.64454091e-01 -5.90744734e-01
8.15383911e-01 -1.40090197e-01 6.00741863e-01 -1.12679088e+00
-4.09272492e-01 -6.18201494e-01 1.58983946e-01 6.09977305e-01
4.63446379e-01 -5.57219267e-01 1.26635492e+00 5.70643842e-01
2.19786223e-02 -1.12538505e+00 -8.22508872e-01 -2.69003630e-01
2.96625584e-01 -7.00880364e-02 1.86753288e-01 8.91316891e-01
-6.91155493e-02 -3.60749185e-01 4.56624106e-02 -1.58449516e-01
8.60215485e-01 -3.18304487e-02 4.06222671e-01 -1.00715947e+00
1.41145721e-01 -8.61041844e-01 -3.40851754e-01 -5.54548860e-01
-3.76432389e-01 -8.85187328e-01 -2.41522174e-02 -2.11275244e+00
1.42174602e-01 -3.68709177e-01 -1.60357937e-01 7.84945309e-01
-2.93617189e-01 7.87915707e-01 -1.68534949e-01 4.24163900e-02
-2.94809133e-01 -1.82933107e-01 1.82868934e+00 -2.52356231e-01
-3.25542659e-01 8.89635980e-02 -6.79396272e-01 7.65891910e-01
1.15020764e+00 -9.44837257e-02 1.07317179e-01 -6.99740574e-02
-1.29468098e-01 7.23389909e-02 1.53228253e-01 -8.62152219e-01
5.16055882e-01 -4.63549078e-01 5.30830562e-01 -6.07034922e-01
-1.12501060e-04 -2.78012961e-01 -2.39239722e-01 8.90830934e-01
-1.83571249e-01 -5.13709009e-01 1.11929819e-01 2.03871936e-01
-1.35288507e-01 -3.65402460e-01 1.18584251e+00 -3.64367723e-01
-5.45801759e-01 5.55594862e-01 -4.65163946e-01 -1.51556313e-01
1.33000970e+00 -4.96471226e-01 -5.65913796e-01 -1.69762492e-01
-8.52392375e-01 3.33218694e-01 6.19752526e-01 -5.25270104e-02
5.72843313e-01 -8.80609155e-01 -8.43864739e-01 1.96771584e-02
-8.94064382e-02 3.36627752e-01 6.66528881e-01 1.23148525e+00
-1.50952172e+00 3.02255481e-01 -3.69775742e-01 -6.85639143e-01
-1.35938668e+00 -2.04707645e-02 6.87024415e-01 -2.62115002e-01
-7.34304190e-01 8.53334188e-01 -2.51793265e-01 -3.42628926e-01
8.65541026e-02 -2.04349488e-01 -2.42437989e-01 -2.37381011e-02
5.05178750e-01 5.88877916e-01 4.57764454e-02 -2.25932047e-01
-9.52591524e-02 7.52366662e-01 -2.81797260e-01 1.63937718e-01
1.11268401e+00 2.01020807e-01 -4.22095776e-01 -4.14118886e-01
9.24881160e-01 -2.78244376e-01 -1.21207392e+00 -1.67540193e-01
-2.99064845e-01 -4.41358417e-01 3.23920131e-01 -1.06096756e+00
-1.29124844e+00 8.65956724e-01 6.10998511e-01 1.09899819e-01
1.08425248e+00 -3.30143750e-01 1.16302562e+00 -9.14502237e-03
-9.62223932e-02 -1.40200818e+00 -1.04595408e-01 5.87021522e-02
5.18021047e-01 -1.24638784e+00 8.61806870e-02 -7.22743928e-01
-5.64715087e-01 1.23114717e+00 7.30635464e-01 -2.27517009e-01
3.96594107e-01 1.61169812e-01 5.33276439e-01 -2.33585671e-01
-3.11340988e-01 -4.62248564e-01 4.13375288e-01 7.51496732e-01
3.54291648e-01 2.90329635e-01 -6.20808005e-01 -3.81439291e-02
1.88993365e-02 1.44008726e-01 7.06748784e-01 8.40361416e-01
-4.73285764e-01 -9.81118560e-01 -2.37630814e-01 8.65758419e-01
-7.15704560e-01 -2.85631865e-02 -7.03761160e-01 8.77602994e-01
2.16096357e-01 6.70491040e-01 1.14982307e-01 8.64452273e-02
-1.23193942e-01 -1.68516412e-01 4.13439691e-01 -4.80933636e-01
-6.42509997e-01 2.84462839e-01 -8.05225968e-02 -7.26843476e-01
-2.01383084e-01 -4.30427194e-01 -1.31561112e+00 -3.82892042e-01
-2.27876857e-01 -4.01129872e-01 9.73521829e-01 7.81171858e-01
7.51636848e-02 5.70331573e-01 4.11418766e-01 -3.87518376e-01
-1.21998459e-01 -8.20556104e-01 -6.66922569e-01 2.42815465e-01
2.11542249e-01 -1.83291212e-01 -1.66484311e-01 1.55509161e-02] | [15.611638069152832, -2.9559383392333984] |
91c9960c-7209-438d-81f2-ba13d42616da | causal-discovery-performance-of-chatgpt-in | 2301.13819 | null | https://arxiv.org/abs/2301.13819v2 | https://arxiv.org/pdf/2301.13819v2.pdf | Causal-Discovery Performance of ChatGPT in the context of Neuropathic Pain Diagnosis | ChatGPT has demonstrated exceptional proficiency in natural language conversation, e.g., it can answer a wide range of questions while no previous large language models can. Thus, we would like to push its limit and explore its ability to answer causal discovery questions by using a medical benchmark (Tu et al. 2019) in causal discovery. | ['Cheng Zhang', 'Chao Ma', 'Ruibo Tu'] | 2023-01-24 | null | null | null | null | ['causal-discovery'] | ['knowledge-base'] | [-2.00670004e-01 8.60570431e-01 -7.38294184e-01 -1.33316472e-01
-8.09889197e-01 -3.92607301e-01 8.84329677e-01 2.52186686e-01
-9.19240415e-02 1.48900652e+00 9.98621047e-01 -1.16017044e+00
-5.31791270e-01 -7.30711281e-01 -3.56425196e-01 -1.48942068e-01
-2.86067784e-01 7.51347005e-01 3.00573409e-01 -4.20700610e-01
2.63119429e-01 -1.87538639e-01 -6.19855165e-01 6.20732486e-01
7.67306745e-01 -6.75079003e-02 -2.67102756e-02 8.39041114e-01
-2.05075994e-01 1.65814626e+00 -8.88947427e-01 -4.43891913e-01
-7.06656158e-01 -7.11019218e-01 -1.75455034e+00 -7.74202645e-01
-1.83915067e-03 -1.44934714e-01 -5.75142026e-01 2.82133639e-01
6.37759626e-01 7.88233206e-02 3.45662177e-01 -1.04910696e+00
-5.85932136e-01 1.17055798e+00 -1.89145282e-01 7.34592915e-01
1.02584279e+00 2.46112168e-01 1.38273919e+00 -3.05523455e-01
1.03820992e+00 1.54188263e+00 5.84090114e-01 1.07464409e+00
-1.18493223e+00 -6.35771036e-01 2.52226721e-02 3.28060925e-01
-6.96077824e-01 -1.31923005e-01 1.68740511e-01 -2.42641285e-01
1.18096483e+00 4.59000021e-01 6.55817091e-01 1.75956702e+00
5.79362810e-01 6.94159210e-01 1.13070989e+00 -1.89148128e-01
1.27565190e-01 -3.95546526e-01 8.56125653e-02 8.01872015e-01
-1.81966886e-01 2.37952888e-01 -8.81442487e-01 -7.61416197e-01
5.81640124e-01 -5.96417904e-01 -2.44377896e-01 6.76598549e-01
-1.52141941e+00 1.13262641e+00 3.27328205e-01 4.66024280e-01
-1.97734907e-01 5.67455113e-01 3.31168741e-01 5.96225321e-01
2.57357627e-01 9.30944145e-01 -6.01289153e-01 -6.01766109e-01
-2.97928214e-01 4.15361494e-01 1.33730817e+00 5.75826645e-01
-2.12337255e-01 -6.44048631e-01 -4.76656824e-01 7.28262782e-01
3.81462097e-01 2.21633390e-02 3.66137087e-01 -1.06464076e+00
5.51980495e-01 2.64074057e-01 -2.35015213e-01 -6.69823527e-01
-7.90238917e-01 1.14213005e-01 -5.72347164e-01 -5.83682656e-01
6.58633232e-01 -7.05779195e-01 -1.89684421e-01 1.68717217e+00
1.30462781e-01 2.97490865e-01 1.00010134e-01 4.93022799e-01
1.31466210e+00 3.54247749e-01 6.18245900e-01 -3.41512322e-01
1.47798944e+00 -4.10300940e-01 -9.21878695e-01 -2.60132492e-01
9.10936415e-01 -8.50423634e-01 8.60315502e-01 3.46025884e-01
-9.82155740e-01 1.72655955e-01 -4.65389371e-01 -7.31375366e-02
1.05491377e-01 -7.25649476e-01 1.19717324e+00 6.80854142e-01
-1.04808593e+00 2.47133479e-01 -4.42678303e-01 -5.44336379e-01
2.27520913e-01 2.06344634e-01 -1.45679936e-01 4.58695441e-02
-2.14951777e+00 1.02197945e+00 2.91752309e-01 -1.87844709e-01
-8.60062063e-01 -1.01254213e+00 -3.08237016e-01 -1.16458103e-01
5.89619696e-01 -1.17035091e+00 1.68752682e+00 3.83896321e-01
-1.43381310e+00 6.67707622e-01 -2.69447476e-01 -4.32598472e-01
3.79135907e-01 -6.58549136e-03 -5.62440634e-01 1.92496568e-01
1.45309374e-01 4.55660999e-01 3.81602794e-02 -5.87982953e-01
-6.49262547e-01 1.67764813e-01 2.90943980e-01 -7.97251463e-02
3.21349539e-02 3.66502017e-01 3.40991467e-02 -4.62849855e-01
-5.67736149e-01 -8.50276530e-01 -4.14728940e-01 -5.85103035e-01
-6.37002707e-01 -7.73329496e-01 5.07781506e-01 -3.83777410e-01
1.19543934e+00 -1.56206453e+00 -1.41532496e-01 -1.38694242e-01
7.09173918e-01 -1.18158408e-01 -1.52057514e-01 8.03066611e-01
-2.09553823e-01 8.03586900e-01 5.76487966e-02 2.86498159e-01
-2.49646917e-01 4.34926629e-01 -3.47005039e-01 1.63210079e-01
2.61051714e-01 1.22517240e+00 -1.50170457e+00 -9.64326501e-01
3.36916633e-02 -5.15768342e-02 -1.08607328e+00 1.48052081e-01
-7.32406437e-01 7.16364861e-01 -8.97336125e-01 2.62318313e-01
-1.70756742e-01 -6.08945131e-01 4.30433989e-01 7.48029232e-01
-1.56918600e-01 1.03887773e+00 -2.73689598e-01 1.29094541e+00
-5.20940661e-01 8.64918888e-01 -2.80084074e-01 -7.34232008e-01
4.78400320e-01 1.09010422e+00 5.24693727e-01 -2.53389478e-01
3.51679325e-02 8.72269496e-02 7.79930413e-01 -8.48617673e-01
-1.28828958e-01 -5.12561858e-01 -2.25127712e-01 5.45897543e-01
-2.08919287e-01 -2.87984729e-01 -3.55941802e-02 5.09518325e-01
1.77599776e+00 -5.68692327e-01 5.15316248e-01 -5.36452711e-01
1.97556823e-01 4.07102972e-01 1.23049065e-01 1.08890271e+00
-1.64998665e-01 2.42895722e-01 1.17607629e+00 -2.83016980e-01
-1.79963991e-01 -1.09793591e+00 -3.18593115e-01 9.13735151e-01
-2.35851884e-01 -5.51330686e-01 -4.20597881e-01 -8.59350622e-01
-2.09210828e-01 1.24045050e+00 -9.15759802e-01 -6.22332506e-02
-6.59332991e-01 -1.06310117e+00 8.99296999e-01 4.35383208e-02
2.97916591e-01 -1.50894511e+00 -2.93972105e-01 3.82296979e-01
-6.83340371e-01 -8.62191200e-01 -3.15310478e-01 3.75839463e-03
-8.13996017e-01 -1.45393896e+00 -2.64270693e-01 -4.39405292e-01
-1.58472732e-01 -3.08111221e-01 1.51658285e+00 2.35290691e-01
-3.21223408e-01 3.97892833e-01 -2.76922971e-01 -4.60759729e-01
-9.27440107e-01 3.16136271e-01 -2.25129485e-01 -8.29820454e-01
5.45441091e-01 -5.29127777e-01 -5.19106686e-01 3.11394840e-01
-3.73887181e-01 -1.63351253e-01 5.12071609e-01 1.03834188e+00
-3.12636375e-01 -5.33992887e-01 1.23742485e+00 -1.17179596e+00
1.39867616e+00 -1.24529231e+00 2.70928219e-02 8.66351426e-02
-4.09701526e-01 -3.85400318e-02 4.77647632e-01 -3.83343846e-01
-1.23869610e+00 -7.41043985e-01 -5.32889485e-01 6.49245799e-01
-8.89227614e-02 9.28865492e-01 3.68219525e-01 2.79715180e-01
9.04651046e-01 -2.16408402e-01 -1.47884920e-01 -1.77989259e-01
4.14471745e-01 4.26807284e-01 3.42685908e-01 -7.49820411e-01
4.07917291e-01 -4.93108528e-03 -1.10580608e-01 -9.42653000e-01
-9.74960744e-01 -2.94029564e-01 -6.59707338e-02 -2.15591788e-01
1.08259404e+00 -5.27700126e-01 -1.45420444e+00 -3.98274064e-01
-1.51828051e+00 -6.87453568e-01 2.01186482e-02 6.73090696e-01
-4.85149235e-01 2.46209037e-02 -9.76016343e-01 -4.04309750e-01
-1.50627673e-01 -7.06222236e-01 5.75715244e-01 -6.07329607e-03
-1.03024733e+00 -1.50955594e+00 5.09686708e-01 4.61022615e-01
2.58513391e-01 2.64051706e-01 1.46391737e+00 -5.98288715e-01
-3.68644834e-01 1.18781649e-01 -4.43225503e-02 -6.60121381e-01
1.77834287e-01 -3.34757683e-03 -5.98348916e-01 3.88382941e-01
-4.92880680e-02 -4.74002123e-01 6.29582465e-01 4.54103053e-01
1.31849146e+00 -3.02022785e-01 -7.43284762e-01 -1.59823224e-01
5.33796251e-01 1.72773883e-01 6.03335440e-01 -7.41675904e-04
3.74182105e-01 7.17327297e-01 1.66317344e-01 2.28722647e-01
6.10764742e-01 3.41942936e-01 6.48500845e-02 1.55947283e-01
-7.80241117e-02 -5.15883207e-01 -4.31312844e-02 7.81281292e-01
-1.14728220e-01 -5.91199696e-01 -1.33788335e+00 5.24935722e-01
-1.84889174e+00 -9.60402369e-01 -7.28482068e-01 1.37608933e+00
1.36380076e+00 2.88389653e-01 -1.11726031e-01 -4.70196635e-01
2.72671312e-01 1.67624548e-01 -3.33256096e-01 -5.47055364e-01
1.43889919e-01 1.60372332e-01 6.44965097e-02 6.42088652e-01
-4.79140043e-01 9.06698167e-01 8.14310646e+00 7.58914649e-01
-5.07156909e-01 2.80354261e-01 8.51126611e-01 7.12536126e-02
-6.20773256e-01 6.13330081e-02 -3.06491524e-01 1.57548681e-01
1.41596174e+00 -3.79611552e-01 1.04308128e-01 8.36036429e-02
4.94319439e-01 -2.75813013e-01 -1.25779629e+00 3.72175068e-01
-3.94478709e-01 -1.93419600e+00 2.66665313e-02 3.83856483e-02
5.74861050e-01 8.39205757e-02 -3.57872814e-01 4.22676623e-01
1.08821416e+00 -1.54554355e+00 -7.07335919e-02 5.07331431e-01
4.62904572e-01 -1.98630691e-01 6.56069279e-01 5.81835926e-01
-4.55944121e-01 1.24908961e-01 -2.18606424e-02 -5.07045746e-01
2.88088590e-01 4.94892657e-01 -1.45066738e+00 4.33923334e-01
4.29859281e-01 3.21924537e-01 -2.21780360e-01 8.63809884e-01
-6.89393163e-01 1.32803261e+00 -3.61291207e-02 -7.79992878e-01
2.24608794e-01 5.36770642e-01 4.12966251e-01 1.17263234e+00
-1.10924495e-02 9.31065559e-01 -2.70362087e-02 7.29201376e-01
-3.26094776e-01 1.34583741e-01 -3.60540867e-01 -2.35339627e-01
3.60846609e-01 5.93757391e-01 -2.33275473e-01 -1.54324308e-01
-5.43029189e-01 3.36271882e-01 2.51871586e-01 1.20259844e-01
-7.99467564e-01 1.75167382e-01 5.07024169e-01 -3.50395276e-04
-6.65247142e-01 1.28051296e-01 -4.95576054e-01 -8.88202608e-01
-8.04651737e-01 -9.42344010e-01 8.89866829e-01 -6.91578984e-01
-1.51708162e+00 3.93746585e-01 4.89749759e-02 -4.78368104e-01
-4.61460650e-01 -3.64733726e-01 -8.90489995e-01 9.47731435e-01
-8.78755510e-01 -6.39717340e-01 2.16694906e-01 6.51290536e-01
4.91892844e-01 1.06746078e-01 9.64600503e-01 1.19326919e-01
-4.66562808e-01 3.55846107e-01 -6.42192245e-01 8.81691650e-02
9.32048380e-01 -1.45721877e+00 5.89539289e-01 2.15488467e-02
-8.11738670e-02 9.32009578e-01 9.46611106e-01 -8.90380085e-01
-1.24153447e+00 -6.01052225e-01 1.30398738e+00 -1.03608942e+00
1.23030436e+00 -2.52795964e-01 -6.51127756e-01 7.03334391e-01
5.73584080e-01 -5.00936806e-01 9.38527107e-01 6.52689517e-01
-6.69631213e-02 6.65010154e-01 -9.96629536e-01 9.53927398e-01
1.26141465e+00 -3.74593228e-01 -1.09789622e+00 8.74953568e-01
1.09387410e+00 -2.62621403e-01 -1.14991188e+00 1.18236490e-01
2.52185762e-01 -6.12671196e-01 9.09006655e-01 -1.09912348e+00
1.13006270e+00 3.29354107e-01 5.08412600e-01 -1.41141856e+00
-2.84541845e-01 -1.20677614e+00 -7.15397149e-02 8.11167479e-01
1.14209926e+00 -8.96878839e-01 6.69106662e-01 7.38687456e-01
9.19862837e-02 -8.16060424e-01 -1.11532485e+00 -2.65623033e-01
6.38267219e-01 -4.74098742e-01 2.30263934e-01 1.26580381e+00
9.88344371e-01 8.19514930e-01 -4.76816326e-01 7.98387546e-03
3.73741686e-01 -5.20076193e-02 5.11919975e-01 -1.34092534e+00
-3.79366696e-01 -5.86273491e-01 4.43984449e-01 -8.73161495e-01
2.28168383e-01 -8.88118327e-01 2.61511318e-02 -1.74455667e+00
3.68856370e-01 -4.06697631e-01 1.81858718e-01 4.86290425e-01
-5.84478557e-01 -2.40364671e-01 -2.29356050e-01 1.10068927e-02
-3.02747518e-01 4.21032816e-01 1.64368546e+00 1.07804544e-01
-1.86160151e-02 1.21410400e-01 -9.73193705e-01 5.33035636e-01
1.04368281e+00 -6.30658567e-01 -4.00646180e-01 -4.92155142e-02
4.16427255e-01 1.00837982e+00 2.79995412e-01 -2.10283399e-01
4.03724343e-01 -5.08185148e-01 -6.15576888e-03 -2.51532365e-02
-1.53014615e-01 7.15103447e-02 -2.15984896e-01 8.49242449e-01
-1.00264716e+00 -9.01044440e-03 3.39707464e-01 6.13419533e-01
-2.88326919e-01 -1.82369396e-01 1.68636620e-01 -3.19199979e-01
-2.84483224e-01 9.76989865e-02 -1.06833589e+00 7.07639039e-01
7.02821076e-01 4.43594396e-01 -1.00610399e+00 -8.35796773e-01
-6.54930770e-01 6.78637028e-01 -3.96941811e-01 7.67861366e-01
5.15616834e-01 -7.76963890e-01 -1.09590101e+00 -7.64969647e-01
-6.74067512e-02 -5.38143992e-01 1.44478410e-01 1.03180730e+00
-4.31196749e-01 9.13742721e-01 3.49616408e-01 -3.47040370e-02
-1.01279092e+00 5.42658806e-01 3.56251925e-01 -5.72845161e-01
-6.98486626e-01 9.76728380e-01 1.93480566e-01 -5.92109680e-01
-9.16067734e-02 -2.12056711e-01 -3.01705927e-01 -6.91219568e-02
4.26360041e-01 1.29121318e-01 -3.91767383e-01 3.22099298e-01
-5.98238051e-01 -2.15349853e-01 2.08531111e-01 -2.45040268e-01
1.21497536e+00 2.16611475e-01 -6.49580181e-01 4.76994604e-01
1.03286111e+00 2.06874833e-01 -3.12310845e-01 1.09627672e-01
4.51633900e-01 -5.54880910e-02 -1.23279691e-01 -1.31427133e+00
-2.80858487e-01 5.73766410e-01 -3.45189869e-02 6.65069699e-01
4.14141834e-01 5.81510544e-01 6.08689129e-01 5.76744318e-01
1.79714769e-01 -5.32731950e-01 1.63748503e-01 7.21220732e-01
1.13576460e+00 -1.06572938e+00 -1.36969760e-01 -5.83768904e-01
-2.25437522e-01 1.06477284e+00 4.04812932e-01 9.21230316e-02
5.33114433e-01 3.39929819e-01 7.84744248e-02 -9.04361010e-01
-1.61992610e+00 -3.89365084e-03 1.40109196e-01 1.84482977e-01
1.18250299e+00 3.77209693e-01 -9.40018356e-01 1.80573076e-01
-5.86999774e-01 1.42733872e-01 9.09395814e-01 2.55710095e-01
1.04328040e-02 -1.12474775e+00 -2.00085580e-01 7.82710373e-01
-7.56301224e-01 -5.62576413e-01 -8.55920434e-01 9.50493395e-01
-2.87012756e-01 1.61245775e+00 -2.63561308e-01 -2.44496271e-01
1.64571896e-01 1.52291000e-01 2.79163480e-01 -6.89784050e-01
-8.92368317e-01 -3.92850608e-01 8.78466427e-01 -5.90985954e-01
-4.64315802e-01 -9.58379924e-01 -1.22211146e+00 -5.41389704e-01
-3.17386240e-02 5.60185790e-01 -1.24512084e-01 1.02806985e+00
-3.03323716e-02 8.23760450e-01 -3.54028828e-02 4.03904766e-01
-2.83758730e-01 -1.13175511e+00 6.07051775e-02 -1.06254384e-01
2.81028956e-01 -3.30338776e-01 -4.48300570e-01 -3.69790763e-01] | [9.293963432312012, 8.025586128234863] |
dd98a329-eb84-40aa-b363-f2df2d1bfc3e | emotion-recognition-techniques-with-rule | 2103.00658 | null | https://arxiv.org/abs/2103.00658v1 | https://arxiv.org/pdf/2103.00658v1.pdf | Emotion recognition techniques with rule based and machine learning approaches | Emotion recognition using digital image processing is a multifarious task because facial emotions depend on warped facial features as well as on gender, age, and culture. Furthermore, there are several factors such as varied illumination and intricate settings that increase complexity in facial emotion recognition. In this paper, we used four salient facial features, Eyebrows, Mouth opening, Mouth corners, and Forehead wrinkles to identifying emotions from normal, occluded and partially-occluded images. We have employed rule-based approach and developed new methods to extract aforementioned facial features similar to local bit patterns using novel techniques. We propose new methods to detect eye location, eyebrow contraction, and mouth corners. For eye detection, the proposed methods are Enhancement of Cr Red (ECrR) and Suppression of Cr Blue (SCrB) which results in 98% accuracy. Additionally, for eyebrow contraction detection, we propose two techniques (1) Morphological Gradient Image Intensity (MGII) and (2) Degree of Curvature Line (DCL). Additionally, we present a new method for mouth corners detection. For classification purpose, we use an individual classifier, majority voting (MV) and weighted majority voting (WMV) methods which mimic Human Emotions Sensitivity (HES). These methods are straightforward to implement, improve the accuracy of results, and work best for emotion recognition using partially occluded images. It is ascertained from the results that our method outperforms previous approaches. Overall accuracy rates are around 94%. The processing time on one image using processor core i5 is ~0.12 sec. | ['Babar Hussian', 'Aasma Aslam'] | 2021-02-28 | null | null | null | null | ['facial-emotion-recognition'] | ['computer-vision'] | [ 2.72849441e-01 -3.91331375e-01 -3.31476107e-02 -4.89547938e-01
-3.64571176e-02 -4.48926002e-01 3.63883048e-01 -8.33481178e-02
-4.72771138e-01 6.17725313e-01 8.64400994e-03 6.19319864e-02
1.77471980e-01 -4.38481092e-01 -4.68421169e-02 -8.02307606e-01
1.19570062e-01 -6.51832998e-01 9.25370827e-02 -2.72354901e-01
6.33053839e-01 1.06352067e+00 -1.84910297e+00 2.97188938e-01
5.70141375e-01 1.37371826e+00 -4.10059005e-01 6.85598314e-01
-5.77644072e-02 5.18968403e-01 -4.92451161e-01 -2.50926167e-01
1.46929905e-01 -3.07943881e-01 -3.09337109e-01 3.88446480e-01
1.76490292e-01 -2.70339668e-01 3.79930615e-01 1.06675875e+00
7.38130212e-01 1.63214684e-01 9.13370430e-01 -1.40567493e+00
-3.30032140e-01 -2.46015102e-01 -1.37699127e+00 1.35382906e-01
4.80101377e-01 -1.01781726e-01 2.35438794e-01 -1.10306835e+00
5.81205726e-01 1.20351195e+00 5.59072018e-01 6.72995865e-01
-8.43324900e-01 -1.10947680e+00 -1.24733478e-01 5.24060488e-01
-1.57275605e+00 -7.22331643e-01 1.02436829e+00 -1.86529532e-01
6.52809381e-01 5.16918421e-01 4.73348051e-01 5.96753240e-01
4.10153538e-01 4.29820746e-01 1.74023473e+00 -7.02529788e-01
1.44060165e-01 5.80106378e-01 6.29011989e-02 9.47811365e-01
-1.10014826e-01 -1.25725254e-01 -4.36848640e-01 -2.76004851e-01
3.73301566e-01 7.01913610e-02 -1.52443588e-01 2.69247800e-01
-4.12122458e-01 7.22714245e-01 -1.18761986e-01 3.43977720e-01
-4.11983222e-01 -2.64143080e-01 3.18763852e-01 2.65923470e-01
2.26983041e-01 -2.04722553e-01 -3.14164579e-01 -2.21789293e-02
-7.27946103e-01 -1.99656203e-01 5.65137327e-01 4.26563710e-01
7.91751862e-01 1.39695276e-02 2.35357061e-01 1.07890368e+00
5.49801588e-01 5.21070600e-01 6.43129289e-01 -7.54129112e-01
-2.35647798e-01 7.35198975e-01 2.10776739e-02 -1.44130397e+00
-5.57936788e-01 3.62433672e-01 -7.35548198e-01 7.57196426e-01
1.85472324e-01 -2.82842875e-01 -9.26387310e-01 1.41612446e+00
5.72961926e-01 -1.23136595e-01 6.21683411e-02 7.53153682e-01
9.16347444e-01 6.08484864e-01 1.78785264e-01 -5.93231201e-01
1.61354578e+00 -5.11163592e-01 -9.68688667e-01 1.63193733e-01
1.26367182e-01 -1.26875567e+00 6.43003345e-01 6.55300558e-01
-8.73090982e-01 -5.57130277e-01 -8.89226437e-01 1.89080641e-01
-6.32035196e-01 5.94083488e-01 5.33044636e-01 1.29278743e+00
-1.00575161e+00 2.21195385e-01 -4.42044675e-01 -5.68748295e-01
1.59011140e-01 6.35482669e-01 -6.71571970e-01 3.10473472e-01
-6.18667245e-01 6.77728474e-01 -2.83344060e-01 3.05338562e-01
1.41104147e-01 -1.29413158e-01 -8.84019911e-01 -3.27173889e-01
-1.81403562e-01 1.44660294e-01 6.10944390e-01 -1.48204792e+00
-2.08004189e+00 1.13556564e+00 -5.85986197e-01 2.75353730e-01
5.91987669e-02 2.34992296e-01 -8.21776688e-01 4.99074847e-01
-5.35078287e-01 6.42427266e-01 1.16463006e+00 -1.13196337e+00
-4.78874058e-01 -7.06240594e-01 -6.85452819e-01 1.06803244e-02
-3.02629113e-01 7.67635226e-01 -6.36696741e-02 -4.70301539e-01
2.46263325e-01 -7.44822502e-01 2.27327719e-01 2.90833056e-01
4.06297520e-02 -2.74331748e-01 1.25886178e+00 -8.58110666e-01
1.20182967e+00 -2.34561658e+00 -6.56504393e-01 6.35910153e-01
-9.59783569e-02 4.67412472e-01 1.43870609e-02 1.28848538e-01
-2.08856151e-01 1.76815525e-01 7.77086988e-02 2.13010609e-02
-1.44732088e-01 3.45606729e-02 9.06412378e-02 7.44890332e-01
3.50800961e-01 3.41133177e-01 -1.63413316e-01 -8.15778613e-01
2.66881377e-01 7.89767921e-01 -2.73126304e-01 -1.55673444e-01
6.95224345e-01 2.66924407e-03 -1.70651525e-01 1.28213573e+00
1.17211902e+00 6.03385091e-01 -1.89830065e-02 -6.28453135e-01
-3.52830768e-01 -5.83352864e-01 -1.58501339e+00 5.74841678e-01
-2.50222504e-01 6.70040786e-01 4.20715094e-01 -7.07873106e-01
1.47003400e+00 5.01183212e-01 1.84621632e-01 -5.10345161e-01
7.05048025e-01 1.11001678e-01 -1.36824340e-01 -8.08355153e-01
2.97418356e-01 -2.74860620e-01 5.17115831e-01 2.99012393e-01
-8.21387023e-02 1.62393451e-01 -6.62988275e-02 -3.31258386e-01
4.70590979e-01 -8.24801326e-02 7.12393582e-01 -6.17882311e-02
1.00267315e+00 -5.13234258e-01 6.10088646e-01 -1.15073808e-02
-7.86342800e-01 2.30707183e-01 4.66645449e-01 -2.92474478e-01
-5.61710835e-01 -6.00162745e-01 -3.32153618e-01 1.00204992e+00
1.32265344e-01 2.28017159e-02 -6.72248960e-01 -2.31286213e-01
-2.21652374e-01 8.53536949e-02 -5.94291568e-01 2.38764808e-02
-3.59937131e-01 -8.15779448e-01 4.05532748e-01 1.99412271e-01
6.82264686e-01 -1.09054089e+00 -6.95124626e-01 -1.13929383e-01
3.70675355e-01 -9.06837821e-01 -2.96634376e-01 -2.78178334e-01
-7.49322414e-01 -9.95151579e-01 -5.94263375e-01 -7.74566472e-01
8.44700098e-01 9.50323716e-02 4.20290917e-01 9.64985415e-02
-8.62436652e-01 3.46184105e-01 -4.39828962e-01 -4.82378006e-01
-8.72008801e-02 -6.96043372e-01 8.78317431e-02 6.48499191e-01
6.48374498e-01 -3.02789867e-01 -7.34464049e-01 4.03663218e-01
-6.43725514e-01 -4.84335333e-01 5.36665618e-01 6.07241988e-01
3.82507652e-01 3.85748334e-02 4.59037721e-01 -4.57956403e-01
7.81914473e-01 -1.97912142e-01 -3.21117908e-01 1.40658513e-01
-4.45443749e-01 -4.23740596e-01 3.55367243e-01 -6.11382842e-01
-1.24249482e+00 1.97897434e-01 -1.12930506e-01 -1.04699120e-01
-5.01392365e-01 -4.70636897e-02 9.97574404e-02 -7.22745597e-01
4.99069303e-01 7.46163055e-02 4.32032675e-01 -2.20156819e-01
-1.12630382e-01 1.24880242e+00 2.67923594e-01 -1.80708364e-01
2.55670220e-01 7.49943256e-01 7.49506950e-02 -1.25034761e+00
-1.53039759e-02 -5.27691245e-01 -3.37717980e-01 -5.82048297e-01
7.98718810e-01 -5.35519958e-01 -1.33402777e+00 6.87583029e-01
-8.28983784e-01 3.49411696e-01 4.81449008e-01 5.11146963e-01
2.87050046e-02 5.00341415e-01 -4.36496764e-01 -1.46922874e+00
-7.82968462e-01 -9.85502720e-01 8.84232938e-01 7.67112911e-01
-2.52542973e-01 -7.37985492e-01 -3.39321852e-01 1.41938135e-01
4.93697137e-01 6.07421398e-01 6.04199350e-01 -8.84206444e-02
1.97279558e-01 -4.86924589e-01 -2.57613748e-01 4.18579012e-01
4.05356914e-01 8.53409886e-01 -9.74047899e-01 7.42805973e-02
-1.18568651e-02 -2.18854412e-01 4.78188932e-01 3.13420385e-01
9.37789738e-01 -8.10677111e-02 -2.21737564e-01 4.49555665e-01
1.50630105e+00 6.30908847e-01 7.22194910e-01 1.48534253e-01
-1.64962485e-02 7.69510329e-01 6.34224117e-01 6.94960654e-01
7.71099608e-03 3.95765752e-01 3.04454565e-01 -3.88208389e-01
1.78747717e-02 3.37440729e-01 5.70838213e-01 4.84573662e-01
-4.26902831e-01 1.79466367e-01 -4.60187644e-01 3.33576769e-01
-1.22254288e+00 -1.03883076e+00 -2.01581478e-01 1.99561894e+00
7.79895008e-01 -4.19020236e-01 8.41942802e-02 4.76431876e-01
9.42035258e-01 -1.20771326e-01 -2.91963339e-01 -1.24421251e+00
-1.50215790e-01 7.71693110e-01 3.30956876e-01 4.19239402e-01
-9.72200513e-01 8.37162256e-01 5.97071457e+00 6.56758487e-01
-1.59034967e+00 -1.23813398e-01 7.89948404e-01 6.62494600e-02
2.02037245e-01 -3.38204950e-01 -9.34280872e-01 4.96301651e-01
3.59673530e-01 2.66642272e-01 2.23595291e-01 7.15442479e-01
3.02974522e-01 -5.94819069e-01 -2.93881267e-01 1.38971508e+00
4.64517832e-01 -6.51942074e-01 -2.32315183e-01 -1.98826447e-01
4.53538746e-01 -6.18820846e-01 9.78785232e-02 -6.29503652e-02
-3.58023614e-01 -8.76811028e-01 1.96260110e-01 4.15060043e-01
7.64616549e-01 -8.93475950e-01 7.32316494e-01 -2.12141678e-01
-1.20537674e+00 -6.56971559e-02 -3.56530935e-01 -1.21879190e-01
-2.37321228e-01 1.30979791e-01 -6.04782403e-01 -9.65970084e-02
7.73896635e-01 1.20436683e-01 -3.92375141e-01 8.08414161e-01
-1.31703734e-01 3.03293347e-01 -6.30593121e-01 -3.20715576e-01
6.65513799e-02 -2.54904389e-01 2.56149888e-01 1.15364623e+00
2.81082779e-01 3.25190514e-01 -3.89665216e-01 4.94005620e-01
1.49009615e-01 7.66741931e-01 -3.64997655e-01 2.57755160e-01
2.82937407e-01 1.81642103e+00 -7.79641509e-01 -4.01734672e-02
-4.94928241e-01 9.19293046e-01 -3.41966093e-01 3.18879306e-01
-5.61469018e-01 -7.53770292e-01 7.46096849e-01 -2.14449227e-01
7.65941143e-02 1.08539328e-01 -1.79586112e-01 -5.89142442e-01
1.49149040e-03 -8.63234818e-01 3.77460837e-01 -7.43298948e-01
-7.52423823e-01 5.51332116e-01 -3.49658281e-01 -9.21834350e-01
9.33635980e-02 -8.65505159e-01 -8.33216190e-01 8.06324840e-01
-1.64026964e+00 -9.56572711e-01 -6.05447114e-01 9.90769506e-01
3.61357123e-01 -1.07403845e-01 7.22473145e-01 1.67349711e-01
-7.28014231e-01 8.55944037e-01 -2.43967533e-01 8.19723308e-02
9.17233050e-01 -6.74168944e-01 -5.76773167e-01 5.61313212e-01
-3.87330323e-01 5.32642841e-01 4.99809027e-01 -3.12429547e-01
-1.29197776e+00 -4.76984292e-01 8.30734134e-01 3.18466514e-01
3.08494717e-01 -1.14233211e-01 -5.26759803e-01 1.65561363e-01
2.94202149e-01 6.47393838e-02 1.06754458e+00 -2.80810863e-01
-2.69659102e-01 -3.26917768e-01 -1.71252358e+00 6.09005213e-01
1.80409372e-01 -1.92430362e-01 -2.09458470e-01 7.11822510e-03
-3.28809112e-01 -1.34476367e-02 -9.48437929e-01 4.58865523e-01
1.17097247e+00 -1.19701278e+00 4.83405143e-01 -1.58911362e-01
-3.09204850e-02 -3.94376755e-01 -5.26463501e-02 -6.42975807e-01
-2.02622190e-02 -8.34490538e-01 5.06399989e-01 1.25606799e+00
2.92665094e-01 -8.84804487e-01 6.50728822e-01 7.37562716e-01
4.64307845e-01 -8.86782229e-01 -7.00991988e-01 -2.29175687e-01
-5.81839561e-01 1.51894065e-02 3.18300277e-01 8.98308039e-01
3.00586492e-01 -7.86113087e-03 -2.34866709e-01 1.71151415e-01
3.49298030e-01 4.02975939e-02 6.13781095e-01 -1.05702233e+00
3.17365408e-01 -4.29501623e-01 -8.05725515e-01 -1.88485503e-01
-4.69137840e-02 -1.64233848e-01 -4.99025971e-01 -9.02738512e-01
6.78245425e-02 -1.07447609e-01 -2.92177677e-01 6.23699963e-01
-9.38462690e-02 7.85394073e-01 2.11104918e-02 -1.44563541e-01
1.15263097e-01 1.01446681e-01 9.46933627e-01 2.24205941e-01
-4.60178256e-01 -1.54063940e-01 -4.71485317e-01 9.96980488e-01
1.05672610e+00 -4.10148650e-02 -7.54116848e-02 2.66080707e-01
1.14120461e-01 -1.56780630e-01 8.11616108e-02 -7.03975737e-01
2.08461657e-01 -2.03866363e-01 8.15791428e-01 -5.03534198e-01
5.83442867e-01 -7.61623919e-01 -1.80095434e-01 4.59737062e-01
2.23558649e-01 1.60440966e-01 3.02826285e-01 4.04361114e-02
-3.16342413e-01 -2.81327933e-01 1.32266736e+00 -5.20111211e-02
-9.26966190e-01 -8.51482749e-02 -6.98900342e-01 -7.14312971e-01
1.28977978e+00 -8.91010284e-01 -2.77678132e-01 -4.17469442e-01
-7.05065370e-01 -2.30425432e-01 4.03487742e-01 1.79698125e-01
8.24235260e-01 -1.01864898e+00 -5.19384801e-01 7.02025115e-01
-1.17047012e-01 -9.48952734e-01 3.22020113e-01 1.20472693e+00
-7.25909591e-01 -7.31024891e-02 -6.53584361e-01 -1.94940343e-01
-2.17548656e+00 1.96151122e-01 2.12280199e-01 4.74213034e-01
1.32608227e-02 9.18169260e-01 -2.55823821e-01 5.45244887e-02
9.88915786e-02 -1.97652057e-01 -7.52826333e-01 3.01772177e-01
7.62883782e-01 6.65513277e-01 -4.49255519e-02 -8.81403446e-01
-5.54941535e-01 1.25095165e+00 -1.34056702e-01 4.89398800e-02
9.58879828e-01 -1.64785773e-01 -4.91450906e-01 3.63973640e-02
1.34570229e+00 5.11058629e-01 -6.80264533e-01 2.03591615e-01
-3.89080554e-01 -5.01139939e-01 1.99060902e-01 -6.99679613e-01
-1.02316415e+00 7.37662435e-01 1.10436380e+00 -5.34914841e-04
1.76855862e+00 -3.79848450e-01 5.05745351e-01 2.64535807e-02
-1.02495156e-01 -1.47353899e+00 -9.21907574e-02 -1.26309752e-01
8.10140014e-01 -1.17922020e+00 1.35043979e-01 -6.54501557e-01
-7.61057079e-01 1.46706462e+00 5.12351573e-01 -4.40745875e-02
9.03701305e-01 3.45273435e-01 3.61598551e-01 2.90275775e-02
-6.87264383e-01 -3.74395341e-01 2.18136564e-01 4.64235544e-01
5.05823791e-01 -9.32526216e-02 -7.50740409e-01 3.15056175e-01
2.72904243e-03 1.54430091e-01 3.69443804e-01 1.04066169e+00
-7.76866078e-01 -8.73755217e-01 -7.86797225e-01 3.21132869e-01
-8.38358104e-01 1.46554694e-01 -5.43694675e-01 8.24669778e-01
3.73786598e-01 1.24712896e+00 1.14517987e-01 -4.68910545e-01
5.96268959e-02 1.79095775e-01 5.69581509e-01 2.79013459e-02
-5.40589273e-01 2.45967761e-01 -8.76207277e-02 -5.12986600e-01
-7.42705226e-01 -5.51724076e-01 -1.16301906e+00 -3.50299090e-01
-4.28983361e-01 -1.12922281e-01 1.24971807e+00 6.94271922e-01
3.36436629e-01 -2.63752699e-01 8.82047236e-01 -5.85611880e-01
-3.64704728e-02 -1.05030346e+00 -8.62452626e-01 3.05215567e-01
1.42528042e-01 -7.53788054e-01 -5.84878325e-01 2.52325982e-01] | [13.272515296936035, 0.8868977427482605] |
3af71392-1c9f-40b2-b60c-874c5dbc9f31 | deep-level-set-for-box-supervised-instance | 2112.03451 | null | https://arxiv.org/abs/2112.03451v1 | https://arxiv.org/pdf/2112.03451v1.pdf | Deep Level Set for Box-supervised Instance Segmentation in Aerial Images | Box-supervised instance segmentation has recently attracted lots of research efforts while little attention is received in aerial image domain. In contrast to the general object collections, aerial objects have large intra-class variances and inter-class similarity with complex background. Moreover, there are many tiny objects in the high-resolution satellite images. This makes the recent pairwise affinity modeling method inevitably to involve the noisy supervision with the inferior results. To tackle these problems, we propose a novel aerial instance segmentation approach, which drives the network to learn a series of level set functions for the aerial objects with only box annotations in an end-to-end fashion. Instead of learning the pairwise affinity, the level set method with the carefully designed energy functions treats the object segmentation as curve evolution, which is able to accurately recover the object's boundaries and prevent the interference from the indistinguishable background and similar objects. The experimental results demonstrate that the proposed approach outperforms the state-of-the-art box-supervised instance segmentation methods. The source code is available at https://github.com/LiWentomng/boxlevelset. | ['Jianke Zhu', 'Wenyu Liu', 'Yijie Chen', 'Wentong Li'] | 2021-12-07 | null | null | null | null | ['box-supervised-instance-segmentation'] | ['computer-vision'] | [ 1.43495247e-01 5.71395233e-02 -6.87011927e-02 -5.61787784e-01
-7.25369990e-01 -4.69814271e-01 1.81811184e-01 -2.13145912e-02
-3.16100597e-01 5.19865572e-01 -4.92623746e-01 1.36424974e-01
-3.11334610e-01 -8.14946353e-01 -6.09679639e-01 -1.04196680e+00
-4.80671637e-02 6.03228867e-01 6.03855729e-01 -1.15138656e-02
7.64648244e-02 1.43483907e-01 -1.50562859e+00 -1.01587668e-01
1.28367031e+00 1.11529183e+00 2.68668562e-01 3.57697338e-01
-2.81440407e-01 3.63920212e-01 -4.33666855e-01 -2.30910182e-01
6.62617564e-01 -4.34607655e-01 -5.30940831e-01 4.84804839e-01
6.49264455e-01 -3.70911151e-01 -2.09227458e-01 1.56348205e+00
3.83718759e-01 3.12664986e-01 5.37479520e-01 -1.24163449e+00
-4.90443081e-01 4.38884884e-01 -1.11252475e+00 1.85675412e-01
-4.19943184e-01 2.02058941e-01 1.00651252e+00 -6.31427109e-01
2.12149993e-01 1.02153873e+00 5.90943158e-01 2.57892996e-01
-1.25325775e+00 -7.14866459e-01 6.36555672e-01 -9.67311338e-02
-1.63165987e+00 9.04570669e-02 9.16787982e-01 -4.92767990e-01
1.69958565e-02 2.19141752e-01 7.84459293e-01 5.40758669e-01
-3.50579441e-01 9.77315485e-01 1.05024505e+00 1.50414497e-01
1.61183909e-01 1.98215321e-01 4.29530710e-01 6.94496214e-01
3.19942772e-01 -1.17776021e-01 1.84051022e-01 -1.02844320e-01
7.18505681e-01 4.39015538e-01 -5.66355288e-01 -5.85534394e-01
-1.15259433e+00 5.69592178e-01 7.99138486e-01 1.51344314e-01
-1.48297444e-01 -3.70318778e-02 2.97356874e-01 -1.73635840e-01
7.95655668e-01 5.46862260e-02 -6.00917459e-01 4.57955837e-01
-1.18387759e+00 2.65616208e-01 5.19589305e-01 1.10248506e+00
8.63114238e-01 -1.42188892e-01 -4.35039662e-02 6.56922996e-01
4.87668812e-01 5.20435274e-01 1.90861493e-01 -6.36308968e-01
4.93593693e-01 9.67079043e-01 2.33982116e-01 -1.19379699e+00
-1.85109630e-01 -6.83303535e-01 -9.86880422e-01 2.39922896e-01
5.88871181e-01 -1.50996923e-01 -9.24122691e-01 1.31687653e+00
8.43399584e-01 3.59163076e-01 -1.41459435e-01 1.21225274e+00
8.73693287e-01 9.71355736e-01 2.15425398e-02 -1.02171741e-01
1.16915917e+00 -1.09278262e+00 -5.64486623e-01 -2.61780351e-01
3.38243693e-01 -4.51545626e-01 1.16720879e+00 2.99549401e-01
-8.52961838e-01 -6.82430089e-01 -1.00887525e+00 2.77796209e-01
-3.28540146e-01 4.63480532e-01 5.17482281e-01 4.18461382e-01
-4.78992403e-01 5.94679058e-01 -9.39814866e-01 -1.34816155e-01
9.06968236e-01 3.81838292e-01 -3.48894186e-02 3.21018510e-02
-8.00141990e-01 2.23799422e-01 7.26488948e-01 4.80816185e-01
-7.37334549e-01 -6.68832481e-01 -5.29987693e-01 -1.09817728e-01
8.30546498e-01 -2.54157633e-01 9.53659356e-01 -1.30801165e+00
-1.25299239e+00 8.20751011e-01 2.61681080e-01 -1.42335162e-01
7.16942489e-01 -1.87681228e-01 4.64850254e-02 1.13042474e-01
1.63892269e-01 6.46138966e-01 8.82614970e-01 -1.45772934e+00
-8.30619693e-01 -6.89503551e-01 -6.82551935e-02 3.50943118e-01
-2.47603863e-01 -1.79246426e-01 -6.17570043e-01 -7.00539708e-01
3.67997169e-01 -8.23681831e-01 -2.61746645e-01 4.12109047e-01
-4.08225834e-01 -7.35603049e-02 1.06850445e+00 -5.86349785e-01
1.04911339e+00 -2.15740776e+00 2.90406495e-01 1.18457511e-01
4.84622195e-02 1.96806550e-01 1.62089229e-01 -5.41174747e-02
6.94358498e-02 1.96068957e-01 -1.11494005e+00 -1.09483562e-01
-6.16048947e-02 1.23547010e-01 -1.66208670e-01 7.27994382e-01
2.91693360e-01 5.23568809e-01 -9.63320315e-01 -7.46964157e-01
1.45945445e-01 3.72638285e-01 -2.51203865e-01 3.69116157e-01
-4.71938550e-01 5.19691706e-01 -7.28180766e-01 9.64768052e-01
1.02527773e+00 -2.11896077e-01 -2.12119907e-01 -1.67739555e-01
-1.85648099e-01 -4.91794258e-01 -1.31498551e+00 1.63444817e+00
2.50011653e-01 2.25578576e-01 3.70438904e-01 -1.01997590e+00
9.93395627e-01 6.11209087e-02 5.10444045e-01 -1.26522169e-01
2.99862444e-01 1.47012234e-01 6.36047572e-02 -4.50749248e-01
1.40995711e-01 -8.64168406e-02 1.02786861e-01 4.01132461e-03
-1.22592635e-01 -4.99146044e-01 2.19986066e-01 -1.42712658e-02
4.95017469e-01 4.00048882e-01 3.05395611e-02 -4.93112624e-01
6.41169906e-01 2.48725489e-01 8.41974258e-01 3.36238176e-01
-3.40831310e-01 6.43109322e-01 2.05341369e-01 -2.73702055e-01
-7.54094720e-01 -8.05316567e-01 -3.45076084e-01 7.00671494e-01
8.05018008e-01 5.07479813e-03 -1.06965148e+00 -8.53558004e-01
3.83361764e-02 2.42216140e-01 -6.07607663e-01 9.25982818e-02
-2.72075176e-01 -1.08805203e+00 3.95426959e-01 4.43916857e-01
1.05708921e+00 -7.72309601e-01 -3.88276637e-01 -1.45609444e-02
-2.23938063e-01 -8.46911132e-01 -5.22884548e-01 3.75544056e-02
-1.12293327e+00 -1.07290471e+00 -9.57640111e-01 -7.79158115e-01
8.78816009e-01 3.44782442e-01 8.42343450e-01 2.27683663e-01
-5.51911831e-01 6.37433082e-02 -3.94086033e-01 -3.84552389e-01
2.53854275e-01 1.55128226e-01 -1.39658183e-01 2.69322246e-01
3.26051146e-01 -2.95152813e-01 -8.26180756e-01 6.57642543e-01
-1.11188591e+00 -7.93420672e-02 5.41896641e-01 7.84169614e-01
1.11932397e+00 6.28181398e-01 8.96928683e-02 -7.51689076e-01
-1.73207908e-03 -4.30131227e-01 -1.08494747e+00 1.50518402e-01
-3.27231944e-01 -4.40415382e-01 4.82472360e-01 -4.74732071e-01
-9.38843608e-01 2.58084595e-01 2.13407859e-01 -4.17788446e-01
-5.09947181e-01 1.91121936e-01 -4.96307433e-01 -2.13387296e-01
3.11315805e-01 1.85645983e-01 -1.74276799e-01 -4.05292779e-01
5.58156148e-02 4.87659991e-01 3.90273362e-01 -5.25440574e-01
1.20117116e+00 7.10856974e-01 -1.86806768e-01 -9.15151656e-01
-1.11055720e+00 -7.63489425e-01 -8.40587735e-01 -3.49426538e-01
1.14604747e+00 -1.00350761e+00 -2.69086808e-01 7.12444842e-01
-9.76136029e-01 -5.57512343e-01 -2.94830173e-01 3.68416786e-01
-2.59121984e-01 4.32974190e-01 -3.20391387e-01 -9.44725990e-01
-1.83415875e-01 -1.13976955e+00 1.20942092e+00 6.29270017e-01
5.00615418e-01 -6.84603751e-01 -1.29972085e-01 3.78744394e-01
-9.26765949e-02 4.51682299e-01 5.52398860e-01 -4.59922403e-01
-9.88473058e-01 -1.83352813e-01 -4.43055093e-01 5.62347353e-01
1.00931086e-01 2.53126293e-01 -8.27497363e-01 -2.27097243e-01
1.03672899e-01 -2.61219561e-01 6.88562512e-01 5.65573514e-01
1.34034467e+00 -1.20554738e-01 -3.41445774e-01 9.28126633e-01
1.58822143e+00 1.99037686e-01 4.74406451e-01 2.04676971e-01
9.45824027e-01 8.64603221e-01 1.19378686e+00 2.75871545e-01
1.45213053e-01 4.22925711e-01 7.21628845e-01 -4.79207516e-01
4.13760424e-01 -1.03365503e-01 5.21115586e-02 4.52800721e-01
-1.59764767e-01 -3.00642163e-01 -9.45374310e-01 4.55787510e-01
-1.95145321e+00 -8.69221032e-01 -4.95174736e-01 2.07339191e+00
5.90795636e-01 7.98359066e-02 1.60030484e-01 1.11319507e-02
9.36831474e-01 3.52224767e-01 -8.89829934e-01 6.01856172e-01
-9.37489513e-03 -3.68886739e-01 8.29524100e-01 2.66012102e-01
-1.52739799e+00 9.76883769e-01 4.62923193e+00 8.07496190e-01
-7.01175272e-01 -5.25232696e-04 9.66924131e-01 1.81054309e-01
2.87571400e-01 9.38459765e-03 -8.81428778e-01 6.78044021e-01
1.86434463e-01 2.29737565e-01 1.63052633e-01 1.01533389e+00
1.59310982e-01 -2.00561121e-01 -6.86534762e-01 8.80504727e-01
-4.45511937e-02 -7.25776136e-01 -1.21393993e-01 2.70638708e-02
8.68481755e-01 -7.26834685e-03 -9.01011303e-02 8.99021477e-02
1.48160448e-02 -6.96405351e-01 7.08480358e-01 6.03713691e-01
4.37065661e-01 -8.12163651e-01 9.33102429e-01 5.91157496e-01
-1.52198398e+00 -2.05038205e-01 -6.64938092e-01 1.73744559e-01
2.50512045e-02 6.00606561e-01 -1.69249669e-01 5.37121296e-01
1.02139020e+00 7.68512726e-01 -5.67402959e-01 1.24835515e+00
-1.41447932e-01 6.84571624e-01 -3.83591026e-01 1.07013047e-01
4.61275309e-01 -9.53552127e-01 6.11888945e-01 9.23421681e-01
1.78305730e-01 6.06732547e-01 7.27413595e-01 1.09198689e+00
-2.23870371e-02 3.10946465e-01 -3.45164150e-01 -6.46817684e-02
1.01017274e-01 1.37365568e+00 -1.29973328e+00 -3.80453140e-01
-2.07024381e-01 1.04547524e+00 -2.69279759e-02 2.87215471e-01
-1.06244540e+00 -5.22707045e-01 4.47727442e-01 1.96875900e-01
3.71248096e-01 -1.15364745e-01 -1.57334685e-01 -1.05336535e+00
1.66195080e-01 -8.15057576e-01 2.21181765e-01 -6.95085347e-01
-1.44157445e+00 4.98196781e-01 2.31568530e-01 -1.41939294e+00
7.47804642e-01 -6.53329492e-01 -8.15189660e-01 6.38969421e-01
-1.49067092e+00 -1.05493581e+00 -8.44311357e-01 3.82445067e-01
7.64202952e-01 1.43542081e-01 2.38951370e-01 2.24256590e-01
-8.71823311e-01 6.73478767e-02 3.21891218e-01 3.62549722e-01
5.29254735e-01 -1.40312135e+00 2.04150807e-02 7.88277924e-01
-2.36660585e-01 2.81966418e-01 6.61913335e-01 -6.97182477e-01
-9.90294635e-01 -1.36041605e+00 2.16779076e-02 -2.99732715e-01
6.71152234e-01 -3.46503884e-01 -1.19066322e+00 3.89568001e-01
2.44752746e-02 4.27580267e-01 5.18553436e-01 -4.60551083e-01
8.83608535e-02 -3.31642866e-01 -1.16415286e+00 4.08127248e-01
1.06689441e+00 2.64547206e-02 -4.21035081e-01 4.99653608e-01
7.17531741e-01 -3.90245438e-01 -7.73271501e-01 5.50824285e-01
1.40335351e-01 -9.20641780e-01 8.66299152e-01 -2.83686906e-01
3.78060132e-01 -9.06784177e-01 -3.15350145e-02 -1.04527164e+00
-2.10911244e-01 -2.16883630e-01 2.24998027e-01 1.41445243e+00
2.00505629e-01 -5.66735089e-01 8.29219341e-01 6.32659495e-01
-2.36683153e-02 -8.55791032e-01 -5.43911397e-01 -7.97577560e-01
-8.75631198e-02 1.16572775e-01 6.68875217e-01 8.35481286e-01
-6.09082520e-01 4.24955226e-02 -1.81196317e-01 6.48628831e-01
1.04223943e+00 4.37029153e-01 9.19137895e-01 -1.60736525e+00
-3.01986009e-01 -3.91045421e-01 -4.24937874e-01 -1.12791514e+00
1.00904897e-01 -8.31848443e-01 5.30307412e-01 -1.44414556e+00
2.08826438e-01 -6.83215022e-01 -3.49124856e-02 2.19754323e-01
-4.33539927e-01 2.38697916e-01 7.60848029e-03 3.52035940e-01
-6.55554771e-01 9.19843018e-01 1.30224395e+00 -5.07603824e-01
-3.34183544e-01 1.71712264e-01 -4.37148571e-01 1.01758158e+00
9.68944252e-01 -7.15727389e-01 -5.19547045e-01 -4.48036611e-01
-2.50822604e-01 -2.68256485e-01 5.40021598e-01 -1.11859679e+00
1.66646257e-01 -1.78855538e-01 4.86455441e-01 -8.71973693e-01
1.16491355e-01 -1.21361935e+00 -1.07753994e-02 2.37120897e-01
-1.11693539e-01 -1.12904266e-01 1.69725522e-01 9.90519464e-01
-2.65425593e-01 -3.21079105e-01 1.04868841e+00 -4.08278137e-01
-4.26437318e-01 9.05477822e-01 -1.91820133e-02 2.16377392e-01
1.28083527e+00 -3.66679221e-01 -3.00770570e-02 1.55259624e-01
-4.61370885e-01 5.91161072e-01 5.87053299e-01 1.33513687e-02
3.97702634e-01 -8.93211424e-01 -7.01122761e-01 -2.91877259e-02
7.10729584e-02 9.75918114e-01 4.06476974e-01 8.70464325e-01
-5.40735960e-01 -7.51912221e-02 2.02481747e-02 -9.96563494e-01
-1.32438529e+00 3.51233214e-01 6.06096745e-01 -9.62382406e-02
-6.85452342e-01 1.02638185e+00 4.10442263e-01 -6.50548875e-01
4.49126720e-01 -3.76142442e-01 -7.80807063e-02 2.09192947e-01
3.06849897e-01 2.20475674e-01 -4.09355074e-01 -6.82129264e-01
-2.11864680e-01 9.03864324e-01 -5.18013164e-02 2.21558765e-01
1.31835580e+00 -2.31104400e-02 -1.39559105e-01 5.75851321e-01
9.54971015e-01 -4.05152649e-01 -1.62856090e+00 -1.56037673e-01
-8.96356255e-02 -7.54092515e-01 3.30840915e-01 -5.17564237e-01
-1.26571190e+00 9.00506198e-01 9.00874078e-01 2.33558282e-01
1.08673668e+00 -2.22104758e-01 7.58089006e-01 4.20156181e-01
3.80929530e-01 -1.17998397e+00 -3.00256610e-02 1.25782818e-01
6.12404406e-01 -1.61071575e+00 1.67333737e-01 -6.45128250e-01
-7.01160550e-01 7.89228439e-01 9.00844455e-01 -4.28227246e-01
7.24818766e-01 1.13934353e-01 1.03317037e-01 -3.47409010e-01
-3.17896605e-02 -3.88575613e-01 2.52304465e-01 4.67035383e-01
1.30772009e-01 1.33889643e-02 -5.54911792e-02 6.19028389e-01
2.94274628e-01 -2.29193091e-01 6.08688174e-03 7.30385184e-01
-5.88826060e-01 -7.68175125e-01 -6.24305725e-01 5.02281606e-01
-2.90736437e-01 -8.22441839e-03 -5.52481294e-01 1.07389987e+00
3.57014239e-01 5.51344991e-01 5.53235523e-02 -4.16399427e-02
3.96413743e-01 -9.02092084e-02 1.33506015e-01 -5.65540612e-01
-5.26144087e-01 1.52684852e-01 -5.14113843e-01 -4.20875490e-01
-6.99702859e-01 -7.15196192e-01 -1.44258988e+00 1.68866396e-01
-8.00282657e-01 2.45707944e-01 3.83225918e-01 6.68006957e-01
-3.58783789e-02 5.24879575e-01 5.72478652e-01 -1.15903258e+00
-4.76692051e-01 -8.48971307e-01 -7.38576233e-01 2.16929093e-01
2.20637664e-01 -7.21729934e-01 -5.85986674e-01 1.45032436e-01] | [9.621332168579102, 0.28875720500946045] |
c393a15e-c3c9-4aab-b357-3bce28ff454a | combined-learning-of-salient-local | 1303.02783 | null | http://arxiv.org/abs/1303.2783v1 | http://arxiv.org/pdf/1303.2783v1.pdf | Combined Learning of Salient Local Descriptors and Distance Metrics for Image Set Face Verification | In contrast to comparing faces via single exemplars, matching sets of face
images increases robustness and discrimination performance. Recent image set
matching approaches typically measure similarities between subspaces or
manifolds, while representing faces in a rigid and holistic manner. Such
representations are easily affected by variations in terms of alignment,
illumination, pose and expression. While local feature based representations
are considerably more robust to such variations, they have received little
attention within the image set matching area. We propose a novel image set
matching technique, comprised of three aspects: (i) robust descriptors of face
regions based on local features, partly inspired by the hierarchy in the human
visual system, (ii) use of several subspace and exemplar metrics to compare
corresponding face regions, (iii) jointly learning which regions are the most
discriminative while finding the optimal mixing weights for combining metrics.
Face recognition experiments on LFW, PIE and MOBIO face datasets show that the
proposed algorithm obtains considerably better performance than several recent
state-of-the-art techniques, such as Local Principal Angle and the Kernel
Affine Hull Method. | ['Mehrtash T. Harandi', 'Conrad Sanderson', 'Yongkang Wong', 'Brian C. Lovell'] | 2013-03-12 | null | null | null | null | ['set-matching'] | ['computer-vision'] | [ 2.85399761e-02 -4.82245147e-01 -2.28565097e-01 -7.03054607e-01
-5.05517542e-01 -5.70555210e-01 8.91246259e-01 1.47073800e-02
-1.39960513e-01 2.56204665e-01 1.39071688e-01 4.03962076e-01
-6.94562137e-01 -5.97890019e-01 -3.51865083e-01 -8.11896741e-01
-2.58200735e-01 4.58084494e-01 -4.28309254e-02 -2.24772081e-01
3.95812362e-01 1.23151231e+00 -2.09181309e+00 1.26012892e-01
3.28893304e-01 1.04931962e+00 -2.76314735e-01 1.02097280e-01
-6.75309375e-02 3.05001140e-01 -2.63691604e-01 -3.48618776e-01
4.61577713e-01 -5.56605279e-01 -6.92052543e-01 4.86010373e-01
1.00144565e+00 1.53241709e-01 -2.81077802e-01 1.25032258e+00
5.11374533e-01 1.97412610e-01 1.03162241e+00 -1.29262602e+00
-6.25750065e-01 -6.44281646e-03 -6.80618882e-01 1.55822322e-01
7.81962693e-01 -2.16327265e-01 6.87225938e-01 -1.06813765e+00
8.01556170e-01 1.74572206e+00 5.25462747e-01 4.55836654e-01
-1.43133605e+00 -6.74275696e-01 -2.05060229e-01 3.02133501e-01
-1.81159484e+00 -9.29298878e-01 9.30289447e-01 -4.36950833e-01
4.89948571e-01 5.15109658e-01 2.72017926e-01 6.26182675e-01
5.84238507e-02 1.89975142e-01 1.33780503e+00 -5.42729557e-01
1.68705024e-02 1.53953478e-01 3.73682082e-02 9.53282356e-01
3.42792794e-02 -7.13347867e-02 -4.39350605e-01 -5.84790647e-01
9.53039944e-01 3.07442904e-01 -3.76826286e-01 -9.26748872e-01
-1.20440078e+00 7.12194920e-01 3.90618950e-01 6.64187729e-01
-3.02907348e-01 -2.17828438e-01 1.86921507e-01 5.70748210e-01
3.34608972e-01 1.40558138e-01 -7.79984472e-03 5.56472182e-01
-9.85184669e-01 3.20686013e-01 6.91644609e-01 8.18082809e-01
1.13901031e+00 -1.28871560e-01 -2.50611603e-01 1.02139902e+00
5.84688604e-01 4.98521268e-01 4.23450381e-01 -8.22200596e-01
1.47756889e-01 6.99527264e-01 -2.09459528e-01 -1.45626950e+00
-3.11322510e-01 -8.37494656e-02 -7.18584955e-01 2.67325550e-01
4.87119615e-01 4.36728865e-01 -6.34173930e-01 1.62285566e+00
5.36926568e-01 2.04782262e-01 -1.78511098e-01 7.87308216e-01
7.64871597e-01 1.49152324e-01 -1.52334392e-01 -4.14065361e-01
1.46974802e+00 -3.81144911e-01 -4.93151128e-01 3.17467481e-01
3.54422837e-01 -9.91064847e-01 3.49275202e-01 1.31919030e-02
-9.94444668e-01 -7.13663518e-01 -8.83608997e-01 2.27367386e-01
-4.99362707e-01 7.21463189e-02 2.68739849e-01 8.76286805e-01
-1.26880312e+00 6.81053996e-01 -3.10129464e-01 -7.59997606e-01
4.05469805e-01 7.13149726e-01 -8.92594874e-01 -8.02997202e-02
-5.76434910e-01 7.99633563e-01 2.20363200e-01 -9.66145843e-02
-6.11926854e-01 -4.89469856e-01 -1.04413056e+00 -1.08275190e-01
4.70780656e-02 -3.63942653e-01 5.06817222e-01 -1.08659780e+00
-1.46447778e+00 1.34616971e+00 -3.99910718e-01 -5.99788539e-02
2.32428774e-01 3.81074011e-01 -6.52663529e-01 2.98259735e-01
-1.75436586e-01 6.61632597e-01 1.25062132e+00 -1.28919125e+00
-1.39429420e-01 -9.76390898e-01 -2.23327547e-01 1.20247386e-01
-4.24119860e-01 5.49571991e-01 -2.36041829e-01 -5.44459462e-01
3.16677004e-01 -9.29576039e-01 9.90052000e-02 3.93204540e-01
8.23034644e-02 -4.30046707e-01 1.20453703e+00 -4.27981615e-01
9.03557122e-01 -2.01242542e+00 4.40513104e-01 5.98491728e-01
2.89172046e-02 2.99951971e-01 -3.66614103e-01 4.34127659e-01
-3.81905109e-01 -1.00797921e-01 -6.56505302e-02 -1.39452502e-01
-1.02376930e-01 1.03139512e-01 1.13706132e-02 1.05926168e+00
2.57706314e-01 6.41340911e-01 -5.99094272e-01 -8.20214927e-01
3.74343693e-01 6.84872985e-01 -3.04994047e-01 1.23210646e-01
4.55173582e-01 4.26402271e-01 -4.58609045e-01 8.55403364e-01
8.02397549e-01 3.46779346e-01 6.76866472e-02 -5.08595586e-01
7.25583453e-03 -3.17409813e-01 -1.51932490e+00 1.43514967e+00
-1.54331088e-01 3.67595226e-01 2.43600041e-01 -1.21360362e+00
1.32842088e+00 5.32684684e-01 5.83729148e-01 -4.86339927e-01
2.58277416e-01 1.92610785e-01 -6.48620576e-02 -3.72951627e-01
5.14484979e-02 -1.38388406e-02 3.21665496e-01 2.65824378e-01
5.52179694e-01 7.23319054e-02 2.54409969e-01 -2.74360955e-01
6.72910810e-01 9.43431556e-02 6.55139208e-01 -7.68098950e-01
1.22192919e+00 -7.04523802e-01 4.04956609e-01 1.57286257e-01
-3.31669092e-01 6.34106815e-01 1.11400969e-01 -5.93147397e-01
-7.43939459e-01 -9.70405459e-01 -7.65500426e-01 7.35164702e-01
1.93671808e-01 -2.55852461e-01 -9.09200668e-01 -4.76169139e-01
2.59000838e-01 2.99317427e-02 -7.70534992e-01 -5.13663776e-02
-5.54158390e-01 -4.60576087e-01 4.55986261e-01 2.12288663e-01
4.71799344e-01 -9.09221649e-01 -4.34527010e-01 -1.45607427e-01
1.87191248e-01 -8.79209518e-01 -6.55131698e-01 -4.20187891e-01
-9.94200408e-01 -1.19579959e+00 -8.44201982e-01 -8.72709394e-01
8.66885662e-01 5.88449478e-01 1.02360320e+00 2.34863311e-01
-6.01981699e-01 9.26560223e-01 -2.12986663e-01 -1.40343815e-01
-2.44863972e-01 -5.75572789e-01 5.07502615e-01 7.04517186e-01
5.75571656e-01 -4.72096741e-01 -5.56873083e-01 7.22185910e-01
-8.09319615e-01 -7.40124762e-01 4.39507008e-01 7.34978616e-01
6.12666965e-01 -1.24324404e-01 2.12536097e-01 -5.21178961e-01
4.33792591e-01 -3.60727906e-01 -3.57415110e-01 5.27839124e-01
-2.71051973e-01 4.83659469e-02 4.18141574e-01 -3.32520902e-01
-8.27657640e-01 1.44506872e-01 3.31180453e-01 -5.92884302e-01
-5.07496893e-01 -1.08522043e-01 -3.47563475e-01 -8.87692273e-01
7.10555673e-01 1.33537054e-01 3.28033447e-01 -4.03909981e-01
2.55283386e-01 6.02767467e-01 4.06864822e-01 -7.17261374e-01
1.10221636e+00 6.79595053e-01 3.25613737e-01 -1.23453462e+00
-1.49326399e-01 -8.09935212e-01 -1.21231854e+00 -3.88952106e-01
7.55516231e-01 -6.69899762e-01 -8.11032712e-01 3.21094811e-01
-9.36757147e-01 6.28528178e-01 -1.17559783e-01 4.21680093e-01
-6.10968709e-01 6.59264028e-01 -1.52721971e-01 -8.16261113e-01
-1.69835940e-01 -1.21405375e+00 1.27841711e+00 2.89229691e-01
-6.84179589e-02 -1.00793576e+00 -2.87786685e-02 2.15848401e-01
2.99334615e-01 5.66080272e-01 8.44486594e-01 -4.88538265e-01
-2.23214492e-01 -2.75914639e-01 -2.33853702e-02 1.05717622e-01
4.92606252e-01 2.27818161e-01 -1.06703615e+00 -6.78986728e-01
6.88011125e-02 -1.15883075e-01 6.12969160e-01 1.63153738e-01
1.00606894e+00 -3.27488005e-01 -4.89015549e-01 7.40767777e-01
1.48281336e+00 1.86210781e-01 6.43239379e-01 8.04514214e-02
4.57027346e-01 1.09071589e+00 3.85596186e-01 3.00779402e-01
-1.21042393e-01 9.53444839e-01 2.77534336e-01 -2.02312134e-02
-4.64413688e-02 1.88823134e-01 3.65278989e-01 5.41494548e-01
-4.31616277e-01 3.41435760e-01 -5.87083042e-01 3.09735984e-01
-1.65731859e+00 -1.28916168e+00 2.02788815e-01 2.59706616e+00
3.69950652e-01 -4.94127840e-01 4.05105799e-01 3.59367132e-01
9.14488435e-01 2.48813257e-01 -7.90689662e-02 -3.72057140e-01
-2.36488059e-01 2.59649009e-01 1.37408137e-01 1.71853185e-01
-1.19700170e+00 6.78826988e-01 6.15267372e+00 8.62653911e-01
-9.13399518e-01 1.29315304e-03 5.87860882e-01 3.32924455e-01
-5.97884655e-02 -1.37323951e-02 -7.33631432e-01 1.54958412e-01
6.25991106e-01 -1.25056610e-01 5.18272996e-01 6.19619787e-01
-2.52677858e-01 3.18169892e-01 -1.24604714e+00 1.33526409e+00
6.47265375e-01 -1.03573704e+00 3.94399345e-01 2.61827290e-01
6.62792206e-01 -5.02208591e-01 1.83242977e-01 -8.35780576e-02
-4.11171645e-01 -1.29908264e+00 4.96577412e-01 7.25939631e-01
6.05848730e-01 -8.18510711e-01 4.48884130e-01 -9.12275612e-02
-1.50173056e+00 -2.50177681e-02 -6.97216153e-01 2.62345612e-01
-4.63717520e-01 5.33564165e-02 -3.37572575e-01 7.06537843e-01
6.48489654e-01 6.21060967e-01 -7.86486924e-01 9.53321397e-01
4.16811943e-01 7.11817592e-02 -1.20355889e-01 1.72524333e-01
7.86067694e-02 -6.85930729e-01 6.48215175e-01 1.09149754e+00
2.64153570e-01 5.73504120e-02 1.38967186e-01 8.19281578e-01
1.77880749e-02 5.80090880e-01 -1.00768077e+00 1.42591909e-01
4.34387386e-01 1.61506343e+00 -7.32107580e-01 8.50255191e-02
-6.41833484e-01 8.43703747e-01 2.65137732e-01 2.77719051e-01
-3.68744195e-01 -1.71743467e-01 9.72629905e-01 1.15150779e-01
2.98965037e-01 -2.50150532e-01 4.68358576e-01 -8.37319732e-01
-6.12443089e-02 -1.00397491e+00 4.52847689e-01 -2.89188683e-01
-1.27039671e+00 7.17544496e-01 1.94867864e-01 -1.44301546e+00
-2.65045285e-01 -9.40466106e-01 -7.19438732e-01 7.54207313e-01
-1.30958772e+00 -1.39831269e+00 -3.04599792e-01 1.11321199e+00
3.22952479e-01 -5.78734756e-01 1.01069200e+00 2.27988362e-01
-3.04558814e-01 8.15303445e-01 3.09765577e-01 1.22924916e-01
8.85026813e-01 -9.19320226e-01 -1.28744513e-01 5.06129742e-01
5.64629793e-01 9.14193153e-01 4.71285611e-01 -6.94544986e-02
-1.74322128e+00 -8.78558278e-01 6.66369677e-01 -4.87252831e-01
2.45477974e-01 -2.50058264e-01 -9.08567607e-01 3.88486803e-01
4.52626050e-02 3.65112931e-01 7.63465703e-01 -4.86812601e-03
-7.29286790e-01 -4.67200905e-01 -1.63854575e+00 4.73288774e-01
1.01730096e+00 -7.37498760e-01 -5.30620694e-01 3.48881364e-01
-1.22444585e-01 8.11743364e-02 -1.28395295e+00 4.39221412e-01
7.71773040e-01 -1.17490983e+00 1.20200264e+00 -5.22587955e-01
-2.63408810e-01 -3.83825958e-01 -3.87369782e-01 -9.82293606e-01
-6.90267980e-01 -5.90133727e-01 2.07624733e-01 1.52863348e+00
-1.59454316e-01 -7.26423085e-01 3.67677450e-01 2.37707198e-01
4.21747357e-01 -6.24899089e-01 -1.21743870e+00 -8.92943919e-01
-1.53848216e-01 2.33286127e-01 8.18772018e-01 1.02516866e+00
-7.74014667e-02 2.07339954e-02 -1.51208848e-01 3.52338515e-02
9.25131917e-01 3.36411029e-01 7.25010097e-01 -1.65260100e+00
8.43281895e-02 -7.32893288e-01 -1.28071976e+00 -2.88991600e-01
5.02118051e-01 -1.07289886e+00 -3.46980661e-01 -8.04917634e-01
3.20149273e-01 -1.61077023e-01 -4.24707651e-01 2.94013619e-01
1.71393469e-01 5.77735066e-01 2.46779233e-01 3.96328449e-01
-4.38007504e-01 4.70268279e-01 9.89358187e-01 -2.20434338e-01
1.00379504e-01 -1.18539676e-01 -3.59543502e-01 6.97217166e-01
4.79299396e-01 -1.88990846e-01 -1.94084138e-01 -1.09378673e-01
-5.58020830e-01 -1.08859673e-01 3.64253551e-01 -1.16014600e+00
2.48078510e-01 -8.59116316e-02 7.09898412e-01 -4.12563607e-02
4.70197827e-01 -1.02051651e+00 2.49430582e-01 4.43151176e-01
-9.18143019e-02 2.09885642e-01 1.01253301e-01 5.56113958e-01
-4.36016351e-01 -3.13495696e-01 1.28321874e+00 -1.41335830e-01
-8.44893813e-01 5.79470754e-01 1.07569043e-02 -3.22253406e-01
1.21545267e+00 -7.45596766e-01 6.68389723e-02 -1.02673255e-01
-3.41912568e-01 -1.95731640e-01 6.17354393e-01 6.53535724e-01
5.95540941e-01 -1.65117538e+00 -8.64538848e-01 5.79530537e-01
3.92935365e-01 -7.51472056e-01 1.04092240e-01 8.92399967e-01
-2.38826990e-01 5.01490772e-01 -6.16118312e-01 -7.49968767e-01
-1.84702528e+00 6.39373779e-01 5.04620373e-01 2.75664300e-01
-1.70621902e-01 7.17728019e-01 2.71805525e-01 -5.06164610e-01
7.67005458e-02 2.02045515e-01 -5.68122089e-01 2.97887594e-01
5.47370613e-01 5.54297507e-01 5.06068915e-02 -1.64734352e+00
-7.60515571e-01 1.42143512e+00 6.38003787e-03 2.73420960e-01
1.03381944e+00 -5.80305718e-02 -3.99047136e-01 1.34888619e-01
1.57530606e+00 -1.25225186e-01 -7.47175157e-01 -5.04582942e-01
1.39444351e-01 -8.84014666e-01 -1.23349264e-01 -9.40541327e-02
-1.06011856e+00 7.49263942e-01 1.21205962e+00 1.01083428e-01
1.19313037e+00 1.89258195e-02 1.68998048e-01 2.62190551e-01
6.47221327e-01 -8.46371233e-01 6.29200414e-03 1.47861004e-01
1.18925858e+00 -1.26163340e+00 1.18568152e-01 -5.29073656e-01
-2.10328907e-01 1.42217529e+00 3.31268847e-01 -3.71002227e-01
8.81575644e-01 -6.40103966e-02 -3.96153927e-02 -3.11416447e-01
-2.81029224e-01 -2.86946833e-01 8.60899270e-01 7.17410564e-01
6.96169615e-01 -1.03601463e-01 -3.79794806e-01 -1.91068128e-01
1.15977302e-01 -5.42608023e-01 -1.96547985e-01 7.15848505e-01
-4.76036459e-01 -1.15805817e+00 -8.49390268e-01 3.75119478e-01
-4.27937508e-01 4.91830558e-01 -5.23215353e-01 8.76870811e-01
1.87365130e-01 9.14163768e-01 9.26464796e-02 -2.84572512e-01
3.67335409e-01 1.40996650e-01 1.00641835e+00 -3.66847754e-01
-5.61979830e-01 -4.35363948e-02 -4.00407940e-01 -7.24686027e-01
-8.35329890e-01 -9.10051703e-01 -8.08391988e-01 -1.94975138e-01
-3.68565530e-01 1.49272442e-01 5.37651539e-01 8.87172580e-01
2.82630116e-01 -1.83534086e-01 9.30380881e-01 -1.39054370e+00
-6.72318995e-01 -9.12502706e-01 -9.35356975e-01 8.54700863e-01
1.09974071e-01 -1.05053997e+00 -3.18642706e-01 1.53220296e-02] | [13.135575294494629, 0.5601373314857483] |
a7555109-d2db-4379-b3cf-1feefe0d1f74 | energy-efficient-vehicular-edge-computing | 2301.13460 | null | https://arxiv.org/abs/2301.13460v1 | https://arxiv.org/pdf/2301.13460v1.pdf | Energy-Efficient Vehicular Edge Computing with One-by-one Access Scheme | With the advent of ever-growing vehicular applications, vehicular edge computing (VEC) has been a promising solution to augment the computing capacity of future smart vehicles. The ultimate challenge to fulfill the quality of service (QoS) is increasingly prominent with constrained computing and communication resources of vehicles. In this paper, we propose an energy-efficient task offloading strategy for VEC system with one-by-one scheduling mechanism, where only one vehicle wakes up at a time to offload with a road side unit (RSU). The goal of system is to minimize the total energy consumption of vehicles by jointly optimizing user scheduling, offloading ratio and bit allocation within a given mission time. To this end, the non-convex and mixed-integer optimization problem is formulated and solved by adopting Lagrange dual problem, whose superior performances are verified via numerical results, as compared to other benchmark schemes. | ['Joonhyuk Kang', 'Seongah Jeong', 'Youngsu Jang'] | 2023-01-31 | null | null | null | null | ['total-energy'] | ['miscellaneous'] | [-8.40103179e-02 -2.95449141e-02 -6.70898974e-01 -3.15713547e-02
-1.82439119e-01 -3.84589851e-01 2.97764093e-01 -2.98846304e-01
-3.30796897e-01 9.92321372e-01 -3.95801514e-01 -6.84675336e-01
1.85966324e-02 -7.81329930e-01 -3.38263035e-01 -1.02137613e+00
-2.68852055e-01 3.14369291e-01 8.77038985e-02 -4.35568020e-02
-4.30464111e-02 3.76195222e-01 -1.45778036e+00 -4.98483896e-01
1.35100770e+00 1.33225930e+00 5.93777418e-01 2.85279572e-01
1.85583815e-01 6.33841306e-02 1.42768174e-01 -2.46655434e-01
8.31090733e-02 2.39147525e-02 -2.26951033e-01 1.83864459e-01
-5.81940114e-01 -2.49285847e-01 -5.91925979e-01 8.74613523e-01
5.05059779e-01 2.95719892e-01 5.59994459e-01 -2.02940893e+00
-2.66248792e-01 -1.70340657e-01 -8.15206468e-01 1.89559340e-01
-3.16518694e-01 8.30478966e-02 5.39737582e-01 -9.64279413e-01
5.59951961e-01 5.35532951e-01 -6.17188662e-02 5.13122916e-01
-7.85252094e-01 -5.86303949e-01 -2.14321166e-02 1.09681058e+00
-1.45463502e+00 -7.83849537e-01 5.53258955e-01 -1.37759537e-01
8.74164999e-01 3.17904502e-01 7.99520731e-01 -9.17780399e-02
3.53059590e-01 7.12106943e-01 6.13648593e-01 7.45673664e-04
3.78084809e-01 3.64223756e-02 -4.00842607e-01 3.57552737e-01
6.91301227e-01 -9.24924463e-02 -1.27202973e-01 1.44782618e-01
-1.21427335e-01 -9.42109302e-02 -2.69379497e-01 -3.42678040e-01
-6.20401263e-01 4.81735915e-01 2.93631524e-01 -8.24691728e-02
-8.08900416e-01 5.30952632e-01 6.10552192e-01 -3.18887904e-02
6.23137474e-01 -4.17987317e-01 -2.74825275e-01 -2.71130264e-01
-9.02713478e-01 -8.44931230e-02 4.59108561e-01 1.26770782e+00
6.89909160e-01 5.07517934e-01 -2.20875233e-01 5.18652320e-01
4.76163357e-01 8.91748786e-01 -1.32485971e-01 -9.47799146e-01
4.25628364e-01 3.15292001e-01 2.66039997e-01 -5.84795952e-01
-3.78728658e-01 -6.18704021e-01 -8.48719358e-01 1.59046769e-01
-4.04378355e-01 -5.78933299e-01 -7.81015337e-01 1.46537149e+00
5.57791173e-01 6.25076830e-01 1.58682972e-01 1.16651547e+00
5.92475355e-01 1.10414124e+00 1.59354359e-01 -1.05207777e+00
1.50826442e+00 -8.49798322e-01 -1.02676499e+00 -3.50714892e-01
6.59435809e-01 -7.41609573e-01 1.20923102e-01 -7.91450031e-04
-1.15988243e+00 5.29484078e-02 -1.23001778e+00 2.34097362e-01
-1.08580440e-01 3.37062329e-01 5.97282708e-01 8.77725482e-01
-9.15914953e-01 -2.47221753e-01 -4.85742658e-01 5.81242256e-02
4.08819675e-01 4.72370684e-01 1.29607748e-02 -3.53891999e-01
-1.04203641e+00 8.70200038e-01 5.67047521e-02 1.87508583e-01
-8.76592457e-01 -5.80607593e-01 -6.35183752e-01 1.98481292e-01
7.49661386e-01 -6.81150436e-01 1.09077156e+00 -4.88833934e-01
-1.35984159e+00 4.81266856e-01 -5.11299372e-01 -2.11230934e-01
3.57266515e-01 4.23379749e-01 -4.96826530e-01 -8.02641585e-02
5.09247221e-02 6.12067953e-02 5.75690687e-01 -1.39000273e+00
-8.42461526e-01 -3.77355337e-01 -3.53675038e-02 3.95326346e-01
-2.94388741e-01 -2.12936010e-02 -8.71696711e-01 5.72772585e-02
-3.24436218e-01 -1.29776812e+00 -3.75979275e-01 -3.14957947e-01
-1.75677881e-01 -3.97986889e-01 1.41211176e+00 -5.23878753e-01
1.46580482e+00 -2.11611199e+00 9.51624289e-02 1.48776174e-01
2.20131308e-01 4.64715213e-01 3.25461254e-02 2.77430385e-01
5.62410533e-01 -4.02721725e-02 5.91213629e-02 -4.20524240e-01
-1.02955513e-01 4.37720567e-01 1.66000739e-01 8.19501102e-01
-3.46569896e-01 7.45800495e-01 -9.86425638e-01 -5.14486253e-01
3.29158664e-01 4.86621618e-01 -4.16779697e-01 1.55774817e-01
6.76270202e-02 2.05550194e-01 -7.22472548e-01 6.79071307e-01
1.20446670e+00 -8.22613463e-02 3.52017969e-01 -2.11674377e-01
-3.74131024e-01 -2.80432105e-01 -9.24475968e-01 1.02358568e+00
-9.12111878e-01 1.03049028e+00 8.14273119e-01 -1.19307554e+00
2.89902508e-01 5.41235387e-01 9.61860955e-01 -1.00418258e+00
3.12368482e-01 4.04733032e-01 2.72885757e-03 -7.98473895e-01
7.59090781e-01 1.13612577e-01 5.07044829e-02 1.54082537e-01
-5.23171484e-01 4.29176897e-01 2.61324316e-01 4.01076913e-01
8.05843949e-01 -3.16598237e-01 9.78193879e-02 -2.97555387e-01
6.12343311e-01 -8.88910424e-03 8.98291111e-01 -2.35413373e-01
-4.15665716e-01 -4.09891367e-01 2.93437779e-01 1.31595284e-01
-1.20836413e+00 -6.62462115e-01 -8.83831978e-02 8.60842586e-01
9.64821577e-01 4.66554835e-02 -5.30683458e-01 -1.12410396e-01
1.76527858e-01 8.79569709e-01 -1.06287614e-01 -8.94459933e-02
-4.41906333e-01 -5.92480063e-01 -7.72551447e-02 7.15814382e-02
4.19864446e-01 -2.52694786e-01 -6.68608129e-01 2.15324268e-01
-2.03763947e-01 -1.55654442e+00 -7.06937432e-01 -2.01245666e-01
-3.69758308e-01 -8.05811405e-01 -5.70941031e-01 -8.25080991e-01
8.94803047e-01 1.09364641e+00 6.35525107e-01 2.91342288e-01
-6.60667345e-02 4.50057723e-02 -1.71224490e-01 -3.21785569e-01
1.87343791e-01 -6.52657822e-02 3.07677031e-01 3.80822420e-01
-3.26919951e-03 -2.35711828e-01 -1.00676537e+00 3.87643427e-01
-6.11479104e-01 3.72015744e-01 4.18469071e-01 3.51240546e-01
7.00772643e-01 2.71627963e-01 1.11517537e+00 -4.15925741e-01
4.17844474e-01 -1.20435917e+00 -7.48577714e-01 8.12303498e-02
-7.04973578e-01 -2.26987287e-01 8.43169749e-01 1.18201584e-01
-1.00816178e+00 4.76148948e-02 1.47359252e-01 -2.25383773e-01
7.32030153e-01 4.51395661e-01 -6.02451146e-01 -4.08082426e-01
-1.98469371e-01 3.77510279e-01 -1.21960789e-01 1.49060905e-01
4.10949796e-01 9.97833490e-01 2.70151436e-01 -1.64178833e-01
7.06219971e-01 4.92431462e-01 5.91488957e-01 -8.11386168e-01
6.40523657e-02 -5.36445081e-01 2.61620820e-01 -1.11010039e+00
6.27593458e-01 -1.05995691e+00 -1.29789209e+00 1.08432725e-01
-1.06599879e+00 -1.48238838e-01 5.05972922e-01 6.07376695e-01
-4.17501062e-01 3.02314013e-01 9.84656364e-02 -1.20343208e+00
-3.52403492e-01 -1.36876428e+00 6.13218129e-01 5.32189488e-01
8.43116701e-01 -5.72707534e-01 -4.53316808e-01 3.70918483e-01
5.80872774e-01 2.35373110e-01 5.65597177e-01 3.25367332e-01
-8.65317106e-01 -1.29716963e-01 -7.02990472e-01 -6.18696250e-02
-2.67771967e-02 -1.23815410e-01 -5.67218304e-01 -5.92311084e-01
-4.41066772e-01 3.15195590e-01 4.44272578e-01 4.73098099e-01
1.23321176e+00 -4.43987280e-01 -9.39163983e-01 6.28549874e-01
1.80909908e+00 4.72851694e-01 5.79650462e-01 -7.51782730e-02
5.55917859e-01 2.20231131e-01 9.20211554e-01 7.06873119e-01
9.45473015e-01 9.16036665e-01 1.08920968e+00 1.78034469e-01
3.83681022e-02 3.41975033e-01 1.63070858e-01 7.29758739e-01
-3.67759019e-01 -8.48325491e-01 -5.92337847e-01 8.42830002e-01
-2.06637788e+00 -7.52757788e-01 -6.13112032e-01 2.26454425e+00
2.32752845e-01 -9.91571099e-02 -1.94221437e-01 1.44932419e-01
8.24168146e-01 5.23478389e-02 -8.18196416e-01 -6.80976629e-01
2.15985447e-01 -4.73467290e-01 1.22179568e+00 2.58959472e-01
-3.63375187e-01 6.15381181e-01 5.14092922e+00 1.19122636e+00
-1.10796738e+00 5.17055690e-01 5.64578474e-01 -2.94325829e-01
-5.42556286e-01 -5.83665930e-02 -3.22205007e-01 1.02768636e+00
1.06593335e+00 -9.66211617e-01 8.29197943e-01 8.19494367e-01
1.06144238e+00 -3.76166642e-01 -5.61941445e-01 1.01009512e+00
-2.41288304e-01 -1.50364971e+00 -7.03267515e-01 5.16306639e-01
1.02263272e+00 2.42551088e-01 -1.44863278e-01 1.02473542e-01
-1.83038890e-01 -6.37034595e-01 5.12952626e-01 2.62553811e-01
1.23312187e+00 -1.24419034e+00 5.34283757e-01 3.93720865e-01
-1.77904963e+00 -1.90022826e-01 -1.15975462e-01 -7.05783367e-02
9.14948463e-01 8.70505452e-01 -3.99342507e-01 6.70495570e-01
2.28412479e-01 1.71991959e-01 4.72831428e-01 1.34591842e+00
1.22223143e-02 3.04941267e-01 -2.17517987e-01 -3.55923951e-01
1.61052689e-01 -3.77460450e-01 8.08885396e-01 8.93836737e-01
6.21457160e-01 6.06119454e-01 2.85696775e-01 4.68195304e-02
-3.80430192e-01 1.25828758e-01 -5.67722142e-01 2.11657599e-01
1.05360138e+00 1.69678819e+00 -6.62320197e-01 -3.97389740e-01
-5.51736116e-01 6.13568664e-01 -1.90126747e-01 6.18265271e-01
-1.15316164e+00 -5.41674137e-01 1.15255535e+00 1.07773669e-01
2.25340035e-02 -7.17135787e-01 -7.23422468e-01 -5.98263383e-01
1.47013947e-01 1.63379863e-01 -1.84532255e-01 -5.91031373e-01
-4.20263052e-01 2.23398104e-01 -2.96168596e-01 -1.41215599e+00
2.61033297e-01 -2.02005163e-01 -7.78544784e-01 7.30822444e-01
-2.21683240e+00 -7.44729936e-01 -6.58374965e-01 3.18483561e-01
5.13024211e-01 -5.25561608e-02 -3.98948304e-02 1.16570318e+00
-8.32231879e-01 4.64299679e-01 6.42107368e-01 -7.28525341e-01
3.81237715e-02 -5.67156196e-01 -1.13430694e-01 1.01049006e+00
-1.00005949e+00 -2.96907514e-01 9.72739697e-01 -5.99121690e-01
-2.40389132e+00 -1.17304194e+00 8.47136915e-01 6.09611988e-01
3.30018848e-01 -3.27000096e-02 -2.91226387e-01 -2.88502220e-02
3.98653448e-01 3.58857930e-01 5.02659500e-01 -7.26894736e-01
5.19422829e-01 -2.40858719e-01 -1.24635637e+00 6.55593812e-01
1.03706574e+00 -3.03729624e-02 7.72074044e-01 3.99895519e-01
5.95213652e-01 -4.93576914e-01 -4.29788679e-01 5.15254438e-01
4.06957328e-01 -1.69148803e-01 5.52661955e-01 -3.38699967e-01
-3.81284803e-02 -6.53080344e-01 -1.99500859e-01 -1.41572940e+00
-2.61746347e-01 -8.30358207e-01 -5.02333581e-01 8.85825455e-01
2.38302201e-01 -6.61959529e-01 9.73907471e-01 6.03250086e-01
-6.50740862e-01 -9.80557382e-01 -1.48485291e+00 -8.09577823e-01
-6.74764574e-01 -5.81917524e-01 5.38168788e-01 4.82635111e-01
1.98068202e-01 2.74964839e-01 -6.28834426e-01 3.93865466e-01
7.92789340e-01 -5.33003956e-02 6.71518087e-01 -9.10614729e-01
3.44742984e-01 -2.20975369e-01 -3.39420170e-01 -1.11264455e+00
2.29664564e-01 -1.02909744e+00 3.28183740e-01 -2.02128339e+00
1.12655595e-01 -7.63293624e-01 -3.27933133e-02 8.79083350e-02
-1.06632784e-01 4.81543541e-01 2.31815204e-01 6.17191195e-02
-7.93745577e-01 8.47767651e-01 1.33134019e+00 -8.16544294e-02
-8.18445161e-03 1.66916788e-01 -4.46184665e-01 1.97606236e-01
9.23695385e-01 -2.29587361e-01 -7.33457386e-01 -5.27405500e-01
4.50466603e-01 5.65168023e-01 -1.23117611e-01 -6.18028104e-01
4.46406215e-01 -7.13458359e-01 -4.68120605e-01 -6.60788774e-01
5.08771300e-01 -1.36346865e+00 4.74007636e-01 6.76803291e-01
6.34976923e-01 -6.60756836e-03 -5.43616787e-02 7.71854103e-01
1.37820050e-01 7.71156119e-05 6.41548574e-01 7.12321222e-01
-1.15094197e+00 8.31261694e-01 -8.17055047e-01 -2.45310292e-01
2.06957030e+00 -2.92802930e-01 -5.35378218e-01 -2.92850465e-01
-2.98453212e-01 1.15424097e+00 1.85553342e-01 3.90317231e-01
4.02066082e-01 -1.56485796e+00 -6.73497498e-01 -2.65824348e-01
1.91419590e-02 -5.10857522e-01 8.43212724e-01 1.03473163e+00
-4.29908365e-01 7.15960264e-01 -1.18695773e-01 -4.12883103e-01
-1.30380559e+00 4.14663881e-01 3.49690542e-02 1.50340304e-01
-1.20093971e-02 3.61425400e-01 -1.99303836e-01 3.63413632e-01
-1.78294793e-01 4.21650171e-01 -1.78977475e-01 7.10713677e-03
1.65351346e-01 1.14968526e+00 1.34978339e-01 -1.01442838e+00
-5.96972108e-01 1.94503054e-01 5.15352726e-01 1.54839963e-01
1.08654761e+00 -6.84118450e-01 -7.19763860e-02 -4.97447312e-01
1.17852044e+00 -4.95509245e-02 -1.14816082e+00 -5.86552918e-02
-4.81487542e-01 -6.73683226e-01 7.30373740e-01 -2.71002948e-01
-1.55381024e+00 3.61294389e-01 4.76346970e-01 4.40286279e-01
1.24035096e+00 -2.27009743e-01 1.38548803e+00 -2.59157240e-01
7.15325773e-01 -1.55069423e+00 -3.68709922e-01 2.86094874e-01
2.70994663e-01 -1.04730678e+00 7.37323686e-02 -8.25112879e-01
-4.98761505e-01 7.11689115e-01 7.05552161e-01 1.28534883e-01
6.83838725e-01 1.77210569e-01 -5.64385772e-01 7.24375322e-02
-8.68371487e-01 -5.41691482e-01 -6.27076328e-02 4.80901808e-01
7.62762427e-02 7.88011134e-01 -1.14763319e+00 3.80921096e-01
3.45501542e-01 1.11415736e-01 6.11097932e-01 7.27231324e-01
-7.30713427e-01 -1.08152294e+00 -5.63217215e-02 7.48380601e-01
-3.16614091e-01 1.53219104e-01 5.83893597e-01 1.32497683e-01
2.56565720e-01 1.34806907e+00 -1.52008524e-02 -4.50824589e-01
-7.77193299e-03 -5.55444062e-01 6.20718934e-02 -2.26160958e-01
3.85192007e-01 -2.50590563e-01 6.30428076e-01 -2.32523814e-01
-1.96669757e-01 -3.87557089e-01 -1.64671731e+00 -7.66473651e-01
-4.68681455e-01 3.83991659e-01 1.42562139e+00 1.05261183e+00
8.00245285e-01 8.26469481e-01 1.40802515e+00 -1.00874543e+00
1.44399315e-01 -2.45892242e-01 -4.39399123e-01 -4.63161141e-01
2.47495621e-01 -8.19746196e-01 -1.32206157e-02 -3.39433044e-01] | [5.841587543487549, 1.6014593839645386] |
d27ced44-ff71-4969-b5ac-dd4566224780 | asyncval-a-toolkit-for-asynchronously | 2202.12510 | null | https://arxiv.org/abs/2202.12510v2 | https://arxiv.org/pdf/2202.12510v2.pdf | Asyncval: A Toolkit for Asynchronously Validating Dense Retriever Checkpoints during Training | The process of model checkpoint validation refers to the evaluation of the performance of a model checkpoint executed on a held-out portion of the training data while learning the hyperparameters of the model, and is used to avoid over-fitting and determine when the model has converged so as to stop training. A simple and efficient strategy to validate deep learning checkpoints is the addition of validation loops to execute during training. However, the validation of dense retrievers (DR) checkpoints is not as trivial -- and the addition of validation loops is not efficient. This is because, in order to accurately evaluate the performance of a DR checkpoint, the whole document corpus needs to be encoded into vectors using the current checkpoint before any actual retrieval operation for checkpoint validation can be performed. This corpus encoding process can be very time-consuming if the document corpus contains millions of documents (e.g., 8.8m for MS MARCO and 21m for Natural Questions). Thus, a naive use of validation loops during training will significantly increase training time. To address this issue, in this demo paper, we propose Asyncval: a Python-based toolkit for efficiently validating DR checkpoints during training. Instead of pausing the training loop for validating DR checkpoints, Asyncval decouples the validation loop from the training loop, uses another GPU to automatically validate new DR checkpoints and thus permits to perform validation asynchronously from training. Asyncval also implements a range of different corpus subset sampling strategies for validating DR checkpoints; these strategies allow to further speed up the validation process. We provide an investigation of these methods in terms of their impact on validation time and validation fidelity. Asyncval is made available as an open-source project at https://github.com/ielab/asyncval. | ['Guido Zuccon', 'Shengyao Zhuang'] | 2022-02-25 | null | null | null | null | ['natural-questions', 'passage-retrieval'] | ['miscellaneous', 'natural-language-processing'] | [-1.99156076e-01 -3.50020409e-01 -8.14844072e-02 -3.42942357e-01
-1.13444543e+00 -7.73289919e-01 6.23242557e-01 5.68099141e-01
-6.70040965e-01 5.36877990e-01 -3.56418163e-01 -7.40270555e-01
1.73122600e-01 -6.97697997e-01 -8.78539979e-01 -5.87354243e-01
3.74669395e-03 9.57128644e-01 1.12926945e-01 -2.31371745e-02
2.52898455e-01 5.75467587e-01 -1.63414407e+00 3.84254515e-01
5.82832456e-01 6.35111749e-01 1.32211328e-01 1.02989769e+00
-1.46951869e-01 6.32261455e-01 -1.02340698e+00 -1.58348143e-01
6.71461150e-02 -1.93259478e-01 -9.02430594e-01 -4.18655515e-01
1.86394110e-01 -5.09543002e-01 1.12807490e-01 8.30509484e-01
4.60364193e-01 1.25505969e-01 5.78872085e-01 -1.00369239e+00
1.50764555e-01 5.41519284e-01 -1.94293201e-01 4.81276810e-02
2.03592300e-01 1.08063050e-01 7.99783111e-01 -7.06532300e-01
4.64505553e-01 6.63792431e-01 7.33091772e-01 2.88779169e-01
-1.32244968e+00 -7.86937475e-01 -3.90458524e-01 1.73488501e-02
-1.55757868e+00 -5.41169763e-01 4.47935611e-01 -3.80387008e-01
1.18372583e+00 4.66434777e-01 5.65953314e-01 1.13528383e+00
1.54111460e-01 5.68500757e-01 9.91693854e-01 -6.42784774e-01
5.54473817e-01 3.01340520e-01 6.23836875e-01 5.79945922e-01
2.74597645e-01 1.69097647e-01 -4.50389594e-01 -4.65337127e-01
3.86729211e-01 -4.87475663e-01 -2.26986766e-01 -5.72167821e-02
-8.42151403e-01 7.89827824e-01 -9.66802239e-02 2.83725142e-01
-2.69858539e-01 1.99294999e-01 8.14442515e-01 5.46896279e-01
4.19892788e-01 4.24342304e-01 -4.87717658e-01 -6.27698898e-01
-1.56300652e+00 2.16587886e-01 9.12458479e-01 5.85507870e-01
9.07782316e-01 -2.18897104e-01 6.29296899e-02 8.50278974e-01
1.72267333e-01 3.98850441e-01 5.38583279e-01 -8.03307533e-01
1.30705073e-01 6.58084333e-01 1.14442021e-01 -4.66702759e-01
-5.04170597e-01 -3.42732817e-01 -7.05767930e-01 3.83613527e-01
4.50109243e-01 -7.45723620e-02 -9.12627101e-01 1.21067357e+00
4.42036420e-01 9.85362083e-02 7.27080926e-02 8.75697196e-01
5.63612521e-01 7.44599760e-01 -1.52814100e-04 -1.24950409e-01
1.16962659e+00 -6.36410177e-01 -3.90047401e-01 -1.55522421e-01
1.29409552e+00 -9.17533278e-01 1.34150541e+00 6.54884756e-01
-9.49254453e-01 -5.00602543e-01 -1.23233867e+00 1.56501040e-01
-3.95341188e-01 3.33434939e-01 5.89995444e-01 5.33120811e-01
-1.10006177e+00 7.10115731e-01 -1.05644679e+00 -2.84329444e-01
-4.92921546e-02 4.75048095e-01 -3.79795074e-01 -6.68409169e-02
-1.15535486e+00 8.48271191e-01 6.74003780e-01 1.80281445e-01
-8.00270319e-01 -3.05225581e-01 -7.73046255e-01 1.27848580e-01
1.15864038e-01 -3.01074594e-01 1.43428504e+00 -9.00179863e-01
-1.28782618e+00 8.18458736e-01 -1.10881463e-01 -3.17538023e-01
5.09130359e-01 -1.87200844e-01 -1.72542065e-01 -1.84857309e-01
-2.03512684e-01 2.77782500e-01 6.90752149e-01 -1.04192615e+00
-2.42536604e-01 -2.29647249e-01 -1.10546742e-02 -3.70111167e-02
-2.76324689e-01 -1.61469597e-02 -9.72418964e-01 -2.70852238e-01
-2.01068923e-01 -1.18049788e+00 2.23109964e-02 -8.07194889e-01
-1.96382850e-01 -3.68425474e-02 4.19626683e-01 -5.90965390e-01
1.42222404e+00 -2.14595771e+00 -2.20268026e-01 5.64096391e-01
-1.58488318e-01 4.87719387e-01 -3.04610074e-01 5.98315597e-01
2.88302694e-02 9.97337922e-02 -9.99092385e-02 -5.15051305e-01
-1.86876982e-01 2.47225702e-01 -3.00801486e-01 4.88336563e-01
1.40134897e-02 5.59748590e-01 -5.70177913e-01 -5.29641271e-01
5.64095750e-02 5.99221051e-01 -3.73644799e-01 3.18065435e-01
-3.46052706e-01 2.58686334e-01 -3.84542793e-01 3.57905298e-01
6.22791052e-01 -3.97281528e-01 4.22988921e-01 1.46534622e-01
-8.29636678e-02 4.90925997e-01 -1.20990109e+00 1.51054358e+00
-6.01150811e-01 7.45067418e-01 -2.68814832e-01 -9.29906785e-01
9.76243973e-01 3.15570265e-01 1.48781627e-01 -8.17204833e-01
1.39563173e-01 4.09072220e-01 -5.23832500e-01 -3.74471009e-01
8.26462269e-01 1.57611430e-01 -1.24807462e-01 7.81125367e-01
-7.07415640e-02 -6.49126768e-02 3.08942139e-01 1.70392334e-01
1.17951882e+00 2.27178529e-01 -4.07879315e-02 -7.24565834e-02
4.38076794e-01 4.38710660e-01 3.15107644e-01 8.99039984e-01
3.91424537e-01 4.25267547e-01 4.97989744e-01 -5.91533899e-01
-1.22733438e+00 -6.05225623e-01 -8.19874853e-02 9.49620664e-01
-2.77696729e-01 -7.06102312e-01 -9.96453047e-01 -7.23859191e-01
-3.70087847e-02 7.60927439e-01 -5.40277839e-01 -1.17673665e-01
-6.15804434e-01 -6.35635912e-01 8.53687048e-01 3.83847088e-01
1.29475117e-01 -1.02393866e+00 -6.12599075e-01 1.09180667e-01
-7.74196535e-02 -6.72787309e-01 -7.89943635e-02 6.30077124e-01
-9.57322955e-01 -1.26208591e+00 -4.03060287e-01 -4.93868679e-01
6.10218167e-01 -6.45179078e-02 1.21360683e+00 5.65022588e-01
-1.35059401e-01 2.43745208e-01 -4.50828612e-01 -2.07332760e-01
-9.17037249e-01 3.63850892e-01 -8.15264657e-02 -4.92748082e-01
5.42926550e-01 -2.49503180e-01 -2.61020392e-01 1.03419639e-01
-1.10164273e+00 3.09382454e-02 3.80703449e-01 9.44202781e-01
5.79713702e-01 -8.62914547e-02 2.46863976e-01 -1.09084034e+00
8.57591629e-01 -4.36214387e-01 -1.16226876e+00 3.81350547e-01
-8.96614313e-01 2.88364887e-01 7.49959469e-01 -4.41860437e-01
-3.90209258e-01 6.94967285e-02 -3.05728108e-01 -5.63566625e-01
-7.07837641e-02 9.45694685e-01 4.59900528e-01 2.40451857e-01
8.21333647e-01 2.76485771e-01 1.65201798e-01 -5.61352193e-01
-8.89216959e-02 9.83691752e-01 1.48281142e-01 -5.98840654e-01
4.27144915e-01 -4.25138958e-02 -3.19927543e-01 -8.35197091e-01
-3.87065768e-01 -4.68135983e-01 -2.06554770e-01 -3.39960873e-01
3.88452739e-01 -7.48484910e-01 -8.46270204e-01 5.24882436e-01
-1.17343080e+00 -9.42508817e-01 -7.90178403e-02 4.82121795e-01
-2.96084017e-01 2.63662398e-01 -6.47335410e-01 -6.78262293e-01
-5.74419975e-01 -1.25164628e+00 1.06928754e+00 -1.97333340e-02
-4.57822412e-01 -8.31293046e-01 4.88837063e-01 2.91932791e-01
3.68010640e-01 -8.47008601e-02 9.51238751e-01 -8.06163788e-01
-4.05213296e-01 -6.73039556e-01 1.75248310e-01 5.24125159e-01
-5.27861655e-01 2.85917222e-01 -1.14054871e+00 -7.75047958e-01
-1.61967874e-01 -5.56921721e-01 6.56692386e-01 5.86249568e-02
1.00772655e+00 -2.77177542e-02 -2.17849299e-01 4.77533728e-01
1.39520824e+00 2.56621814e-03 6.63135707e-01 6.94867313e-01
1.02452546e-01 3.64367783e-01 8.28409791e-01 4.29687411e-01
1.29657641e-01 5.94350517e-01 2.33175501e-01 1.01817384e-01
2.15631455e-01 3.60400937e-02 5.51984668e-01 8.99132967e-01
1.03846364e-01 -3.26859325e-01 -1.44400752e+00 3.83257806e-01
-1.56695569e+00 -5.66789389e-01 -2.74435073e-01 2.73345876e+00
9.46175873e-01 2.49771640e-01 -1.02697611e-01 2.33315274e-01
4.01830912e-01 -6.58985078e-02 -2.90496469e-01 -7.39265859e-01
2.49172896e-01 3.81165057e-01 4.65947658e-01 7.85673082e-01
-8.02820265e-01 9.87436950e-01 5.84204626e+00 7.90752053e-01
-1.52521360e+00 9.18105096e-02 5.80382705e-01 -3.02780628e-01
-2.02682331e-01 2.15750411e-01 -9.24416780e-01 4.39842850e-01
1.51828885e+00 2.59700447e-01 5.46525419e-01 8.35205555e-01
1.58910334e-01 -5.87880850e-01 -1.10201275e+00 8.01119506e-01
-1.99301988e-01 -1.33857822e+00 -2.82105148e-01 -4.55486663e-02
2.04167366e-01 5.42634368e-01 -3.09961677e-01 6.38505578e-01
2.03888685e-01 -1.00838959e+00 6.58793151e-01 4.48665082e-01
9.44039464e-01 -8.22948217e-01 9.34790432e-01 6.45010471e-01
-6.97939456e-01 1.47685021e-01 -4.87803996e-01 1.19291976e-01
-1.90804079e-01 7.96277821e-01 -1.06678236e+00 4.14238125e-01
7.73931026e-01 3.25113162e-02 -8.19939077e-01 7.17847168e-01
-1.29644386e-02 1.05634260e+00 -5.25571585e-01 -1.04343757e-01
2.04420879e-01 -2.83135977e-02 2.26695180e-01 1.44451630e+00
2.35986859e-01 -3.63784224e-01 -4.97883707e-02 5.46666086e-01
1.73602983e-01 2.23308772e-01 -3.67558897e-01 -4.00133245e-02
5.45219421e-01 1.09262156e+00 -6.24314070e-01 -4.29969817e-01
-1.06853493e-01 9.61164176e-01 5.49755633e-01 4.38919365e-01
-9.48809624e-01 -4.06618387e-01 4.81579840e-01 2.14893192e-01
2.54700124e-01 -3.35393786e-01 -1.69276670e-01 -9.12038624e-01
1.31198630e-01 -1.18373263e+00 2.83403903e-01 -6.72980309e-01
-6.55028701e-01 7.91471481e-01 -1.33951664e-01 -1.05124998e+00
-5.71319044e-01 -2.85540938e-01 -5.03989637e-01 9.89216328e-01
-1.43760002e+00 -8.58193278e-01 -4.76923972e-01 4.35786813e-01
1.22801110e-01 -4.36800420e-02 1.21780431e+00 3.72163922e-01
-5.62938213e-01 7.47599065e-01 4.80918825e-01 7.92522356e-02
9.36776936e-01 -1.02571619e+00 4.52805519e-01 4.40922946e-01
1.72810048e-01 7.94922888e-01 7.21966624e-01 -6.77806616e-01
-1.35246229e+00 -9.62001324e-01 9.39881623e-01 -3.28532815e-01
4.36063617e-01 -5.04568100e-01 -1.28600395e+00 5.91626823e-01
-4.55615036e-02 -1.66614458e-01 7.55158544e-01 3.29703033e-01
-3.43963176e-01 -1.57157183e-01 -6.20702267e-01 3.07941705e-01
1.15474321e-01 -6.37970328e-01 -9.86314565e-02 3.00371647e-01
2.18784571e-01 -5.77602029e-01 -9.50666964e-01 2.48283431e-01
6.49666607e-01 -1.04995143e+00 5.55884540e-01 -3.73130947e-01
1.22474365e-01 -1.79587290e-01 -1.02835777e-03 -9.32183444e-01
1.36249959e-01 -3.96087825e-01 -1.17231444e-01 1.16869152e+00
5.70563316e-01 -6.78481340e-01 8.65091085e-01 4.20339912e-01
6.00060709e-02 -8.98609936e-01 -8.91038537e-01 -6.52402163e-01
6.78868368e-02 -8.16758811e-01 7.11311877e-01 1.03863406e+00
-1.42150074e-01 1.48916215e-01 -4.44888100e-02 2.70696908e-01
2.51411736e-01 3.19011100e-02 1.22551489e+00 -1.09742749e+00
-5.88908017e-01 -1.63787067e-01 -1.43665686e-01 -7.69283473e-01
1.16762206e-01 -9.33028638e-01 1.12211622e-01 -1.24383616e+00
2.20189951e-02 -8.21204662e-01 -2.27329463e-01 6.66559696e-01
5.01774959e-02 2.52740592e-01 -6.80432245e-02 5.22280276e-01
-4.38250601e-01 -3.84950601e-02 5.30705631e-01 6.66046143e-03
-3.30946386e-01 4.60215323e-02 -8.41124579e-02 3.15137804e-01
1.04084432e+00 -7.73078859e-01 -3.19747508e-01 -4.13788080e-01
6.16093874e-01 3.13834459e-01 4.11568940e-01 -1.10703528e+00
2.87827820e-01 2.09669247e-01 4.00917828e-01 -6.42178178e-01
3.04049104e-01 -5.82668364e-01 4.67940748e-01 4.02818143e-01
-2.26034552e-01 3.81662607e-01 5.14259100e-01 2.45818183e-01
-2.74401724e-01 -6.60365701e-01 6.03133976e-01 9.69626978e-02
-3.94743264e-01 -2.42746398e-01 -5.87292790e-01 -7.63348043e-02
8.40879858e-01 -2.22716689e-01 -3.56119245e-01 -2.05229938e-01
-6.37758076e-01 2.25336701e-01 9.52508867e-01 1.99770659e-01
4.36934561e-01 -7.68913031e-01 -4.35533106e-01 5.05833328e-01
1.25645533e-01 1.51476428e-01 2.14129135e-01 8.62290561e-01
-9.37831819e-01 3.69818836e-01 2.32884169e-01 -6.66930318e-01
-1.48247683e+00 3.46786708e-01 2.63315499e-01 -6.72850966e-01
-5.21935463e-01 6.54542327e-01 -5.22628725e-01 -5.06916523e-01
3.41570139e-01 -1.34106278e-01 -8.26682001e-02 4.78510596e-02
5.11708617e-01 7.05551282e-02 6.88261986e-01 -1.89657614e-01
-2.94229805e-01 2.49284610e-01 -2.91361690e-01 -1.91190794e-01
1.31511939e+00 4.50452954e-01 -2.78583735e-01 4.96961743e-01
1.37762928e+00 -3.42260748e-02 -8.98135364e-01 1.55368477e-01
1.27937615e-01 -6.78359419e-02 2.26503432e-01 -8.74508679e-01
-7.24777937e-01 8.42493951e-01 6.74610078e-01 4.88012731e-02
1.03174901e+00 -2.04648241e-01 6.24464154e-01 6.18775547e-01
2.56081372e-01 -1.26021528e+00 -3.27898324e-01 6.30125701e-01
5.37074924e-01 -1.17834568e+00 1.05519891e-01 3.32088083e-01
-4.91391122e-01 1.20965850e+00 3.15167248e-01 4.00135141e-05
4.14014697e-01 4.44114000e-01 2.68135637e-01 -2.96399117e-01
-7.89847791e-01 3.30852598e-01 6.00042157e-02 3.49935330e-02
5.94208896e-01 -3.22392024e-02 -2.62820601e-01 2.67530799e-01
-1.60754874e-01 2.42564470e-01 4.91395593e-01 1.10206604e+00
-1.81112722e-01 -1.39121115e+00 -4.68898505e-01 5.30097902e-01
-4.90986854e-01 -2.23064572e-01 -3.26045781e-01 8.13098907e-01
-3.41563225e-01 6.95571542e-01 8.30541775e-02 -5.73423445e-01
1.03420980e-01 4.01783705e-01 2.72785902e-01 -5.10818899e-01
-9.85502005e-01 -9.45891887e-02 2.62249917e-01 -5.56635916e-01
9.58641320e-02 -4.85693008e-01 -1.36239660e+00 -5.96746624e-01
-6.40275419e-01 6.82839572e-01 1.24294853e+00 6.13701999e-01
4.74474758e-01 2.82814980e-01 3.55130792e-01 -6.15705371e-01
-7.50627279e-01 -9.64508474e-01 -3.49336416e-01 1.57783300e-01
1.68523028e-01 -2.15701163e-01 -3.44349980e-01 -1.66828632e-01] | [8.609722137451172, 3.6960935592651367] |
c3a644aa-993b-4422-9601-53356cbf385e | agconv-adaptive-graph-convolution-on-3d-point | 2206.04665 | null | https://arxiv.org/abs/2206.04665v2 | https://arxiv.org/pdf/2206.04665v2.pdf | AGConv: Adaptive Graph Convolution on 3D Point Clouds | Convolution on 3D point clouds is widely researched yet far from perfect in geometric deep learning. The traditional wisdom of convolution characterises feature correspondences indistinguishably among 3D points, arising an intrinsic limitation of poor distinctive feature learning. In this paper, we propose Adaptive Graph Convolution (AGConv) for wide applications of point cloud analysis. AGConv generates adaptive kernels for points according to their dynamically learned features. Compared with the solution of using fixed/isotropic kernels, AGConv improves the flexibility of point cloud convolutions, effectively and precisely capturing the diverse relations between points from different semantic parts. Unlike the popular attentional weight schemes, AGConv implements the adaptiveness inside the convolution operation instead of simply assigning different weights to the neighboring points. Extensive evaluations clearly show that our method outperforms state-of-the-arts of point cloud classification and segmentation on various benchmark datasets.Meanwhile, AGConv can flexibly serve more point cloud analysis approaches to boost their performance. To validate its flexibility and effectiveness, we explore AGConv-based paradigms of completion, denoising, upsampling, registration and circle extraction, which are comparable or even superior to their competitors. Our code is available at https://github.com/hrzhou2/AdaptConv-master. | ['Jing Qin', 'Jun Wang', 'Yanwen Guo', 'Xuefeng Yan', 'Jingbo Qiu', 'Zhe Zhu', 'Zhilei Chen', 'Huajian Si', 'Fei Hu', 'Haoran Zhou', 'Zeyong Wei', 'Mingqiang Wei'] | 2022-06-09 | null | null | null | null | ['point-cloud-classification'] | ['computer-vision'] | [-4.14699167e-01 -2.78606325e-01 2.06565350e-01 -1.85497493e-01
-4.04341221e-01 -5.96201956e-01 5.63063383e-01 -2.91699357e-02
-1.34756848e-01 8.29676092e-02 -2.47890785e-01 -3.49732220e-01
-2.88053691e-01 -9.16872382e-01 -8.66897166e-01 -5.67707717e-01
-7.20631378e-03 4.53968585e-01 1.65147632e-01 -2.36743391e-01
3.46812546e-01 1.20770609e+00 -1.47511685e+00 5.98670393e-02
9.66224313e-01 1.07279325e+00 1.91842273e-01 3.56866151e-01
-5.57814896e-01 -2.82190554e-03 -2.05942109e-01 -5.16352832e-01
5.21831751e-01 3.98905486e-01 -6.11625850e-01 -1.99765302e-02
9.35580254e-01 -1.80641949e-01 -5.57998717e-01 1.19449377e+00
4.18590784e-01 4.97628003e-02 6.74080789e-01 -1.34139967e+00
-1.20428455e+00 2.80282348e-01 -7.63826847e-01 2.92521507e-01
-7.74282888e-02 1.96347192e-01 9.69309509e-01 -1.34611654e+00
3.43285978e-01 1.26741719e+00 1.09429002e+00 3.64473164e-01
-1.08767307e+00 -8.19306135e-01 3.47535402e-01 1.04563870e-01
-1.65944314e+00 -1.50951132e-01 9.90956008e-01 -4.81092393e-01
9.01085317e-01 3.45009774e-01 7.24779069e-01 7.00256228e-01
-7.22289607e-02 5.81181526e-01 6.19252980e-01 -6.82648197e-02
8.99991691e-02 -3.93596649e-01 2.23568067e-01 6.74529374e-01
3.83053184e-01 4.31307591e-02 -2.90751249e-01 -3.00359994e-01
1.41548765e+00 5.14357388e-01 -3.84321362e-01 -5.65174282e-01
-1.30949712e+00 6.69596076e-01 8.01418424e-01 2.00470030e-01
-5.05130410e-01 5.18188477e-01 1.71529293e-01 4.01100338e-01
5.93559742e-01 1.78464636e-01 -4.91583794e-01 1.50989607e-01
-6.91174507e-01 3.41701150e-01 3.47798079e-01 1.38204169e+00
1.03043854e+00 1.47965044e-01 -2.43701175e-01 6.47342741e-01
3.11505288e-01 6.35727704e-01 1.09196424e-01 -9.76892471e-01
2.81738669e-01 9.12753463e-01 -5.47211505e-02 -1.13337862e+00
-3.46424162e-01 -6.11842096e-01 -9.35953557e-01 5.55761576e-01
4.18639153e-01 1.09005958e-01 -1.02739406e+00 1.27952933e+00
5.13172686e-01 8.13184083e-01 -3.39525551e-01 1.00811779e+00
1.14021444e+00 3.14818501e-01 -1.65276423e-01 5.16605437e-01
1.30684173e+00 -6.38297439e-01 -1.55936033e-01 -9.05351713e-03
3.46690744e-01 -8.97955418e-01 1.08027720e+00 8.95039588e-02
-9.58348989e-01 -5.68728924e-01 -7.42727935e-01 -2.24804237e-01
-4.13498580e-01 7.68843889e-02 9.68316793e-01 2.99463183e-01
-1.13117313e+00 9.57959116e-01 -9.11393046e-01 -1.97373167e-01
9.54658508e-01 4.79918301e-01 -3.28678995e-01 -1.36099784e-02
-6.05660021e-01 4.65096056e-01 -4.95978817e-02 1.82288602e-01
-3.31245869e-01 -1.23022449e+00 -7.30423152e-01 2.12874457e-01
6.18309788e-02 -1.15352523e+00 1.11539781e+00 -6.26097441e-01
-1.22961736e+00 9.48350906e-01 -8.69646072e-02 -2.75742114e-01
5.18387914e-01 -3.34251851e-01 -1.11194991e-01 1.38418734e-01
3.59595940e-02 6.08808517e-01 1.08729923e+00 -1.18971145e+00
-4.08835441e-01 -6.25622571e-01 4.33747098e-02 1.92565441e-01
-1.29430983e-02 -2.89355814e-01 -7.63320684e-01 -6.57577991e-01
5.52194536e-01 -8.23269069e-01 -2.16236219e-01 4.41393733e-01
-3.33408505e-01 -5.20730197e-01 1.02706993e+00 -3.71645987e-02
7.20787525e-01 -2.33456206e+00 -2.12546557e-01 3.95647258e-01
7.26025522e-01 2.81864911e-01 -2.33225808e-01 2.44675323e-01
-2.20164046e-01 5.27465530e-02 -1.42609775e-01 -3.74177992e-01
1.32849604e-01 1.82606205e-01 -3.32083791e-01 7.68842399e-01
3.96962225e-01 1.43238139e+00 -7.59303153e-01 -2.45928749e-01
5.92528343e-01 9.06962752e-01 -4.86255884e-01 -2.90498197e-01
-2.91630141e-02 2.69328266e-01 -7.14590728e-01 8.97758603e-01
1.27352607e+00 -5.63040614e-01 -5.15217006e-01 -3.96651149e-01
-2.40130257e-02 -1.40087172e-01 -1.15744448e+00 1.83752632e+00
-2.14366660e-01 5.89131832e-01 2.23030057e-02 -8.28354180e-01
1.09976065e+00 5.25891446e-02 6.18346453e-01 -4.17079538e-01
2.02222526e-01 1.85469180e-01 -1.96720317e-01 -2.11326182e-01
4.34525430e-01 1.96568549e-01 3.52996290e-01 -3.71735431e-02
1.12121359e-01 -3.67664039e-01 -4.58334506e-01 1.42156050e-01
1.12084234e+00 5.28423488e-02 -1.27301335e-01 -3.63268375e-01
4.61070478e-01 -7.17511773e-02 2.75458306e-01 7.49361038e-01
-1.64009690e-01 9.65058923e-01 1.55483887e-01 -5.85815847e-01
-9.18607473e-01 -1.28414857e+00 -2.89289594e-01 6.74846828e-01
5.16624928e-01 -2.96313912e-01 -6.00010991e-01 -6.83074713e-01
6.97432756e-01 4.24561709e-01 -6.11200094e-01 1.30984724e-01
-6.04764521e-01 -3.56362432e-01 4.60538387e-01 6.31179988e-01
5.26681840e-01 -8.48369122e-01 -2.41380423e-01 -6.52303696e-02
2.50315309e-01 -1.09405828e+00 -5.68568170e-01 -1.89241022e-01
-1.18218327e+00 -1.15956652e+00 -6.49879873e-01 -7.11884856e-01
7.56998599e-01 8.83225441e-01 1.37304735e+00 3.11429590e-01
-1.20477341e-01 6.81744158e-01 -2.59661227e-01 -5.73568225e-01
2.29854003e-01 1.41691029e-01 -1.61637500e-01 -2.15975847e-02
6.82718694e-01 -9.93425965e-01 -9.36256468e-01 2.25826457e-01
-8.45708668e-01 -2.29110792e-01 5.66444099e-01 5.64034879e-01
8.94845426e-01 -2.96338946e-01 2.67466903e-01 -7.32589066e-01
5.04770339e-01 -3.61742496e-01 -6.77101851e-01 1.63546503e-02
-4.54550892e-01 -1.28963724e-01 3.99277478e-01 -3.37100804e-01
-5.63527167e-01 4.91024181e-02 -2.48925596e-01 -1.23772776e+00
-2.59492040e-01 8.64706784e-02 4.15637121e-02 -6.50768220e-01
4.74060982e-01 2.02103868e-01 1.94641292e-01 -5.65352619e-01
5.95221877e-01 2.36127555e-01 5.58117688e-01 -5.74009538e-01
1.20269537e+00 9.40480769e-01 4.17116731e-02 -8.51555407e-01
-4.98435527e-01 -6.98249817e-01 -8.45468700e-01 -2.05907091e-01
7.40071118e-01 -8.46403003e-01 -8.47590625e-01 5.91085970e-01
-1.38170719e+00 -1.03146031e-01 -5.85500658e-01 2.69897968e-01
-4.75887120e-01 3.66649330e-01 -5.36786675e-01 -3.60309184e-01
-5.09593189e-01 -1.00480151e+00 1.41551816e+00 2.02056050e-01
1.76908270e-01 -1.07534039e+00 -2.10278645e-01 -8.09354533e-04
4.54028666e-01 3.26942325e-01 6.64679229e-01 -4.55186784e-01
-9.70786333e-01 -2.05115736e-01 -5.18957853e-01 2.65060604e-01
6.59037456e-02 1.31006047e-01 -1.05015039e+00 -2.19167724e-01
-8.88374299e-02 2.44848236e-01 8.56994033e-01 5.36884189e-01
1.56826937e+00 4.43981774e-02 -4.17012721e-01 1.20666635e+00
1.48428190e+00 -2.64368147e-01 8.13068330e-01 2.49669626e-01
1.10322869e+00 1.38868898e-01 2.36255646e-01 2.84923792e-01
3.88706714e-01 3.99241865e-01 8.83020759e-01 -2.90225089e-01
-1.61975369e-01 -7.06303492e-02 -1.74417093e-01 7.03829348e-01
-5.09051681e-01 1.06087871e-01 -9.73888040e-01 3.38224709e-01
-1.86868298e+00 -8.57985079e-01 -6.19947076e-01 1.91174746e+00
1.26498803e-01 3.04957889e-02 -2.01709583e-01 -1.88004136e-01
7.89054096e-01 1.15831949e-01 -6.65560365e-01 -2.08748858e-02
-2.24876374e-01 5.68064034e-01 8.44526112e-01 2.35215396e-01
-1.22284842e+00 1.01578224e+00 5.40459108e+00 1.10232544e+00
-1.21068227e+00 1.68751955e-01 3.93917620e-01 5.51728643e-02
-4.25717205e-01 -1.11370869e-01 -6.15456164e-01 3.59337240e-01
1.10384144e-01 -8.70613456e-02 2.85139143e-01 9.15984929e-01
-2.44236682e-02 5.31514585e-01 -9.76221442e-01 1.37386382e+00
-2.42183566e-01 -1.80989170e+00 3.14985693e-01 9.34823900e-02
6.83145761e-01 6.14811003e-01 2.37561181e-01 1.41300946e-01
1.70315191e-01 -1.08300197e+00 7.13172555e-01 6.85595155e-01
7.02261388e-01 -7.63814390e-01 5.99697113e-01 6.46437034e-02
-1.34581256e+00 2.11545229e-01 -6.76443696e-01 -6.43272027e-02
-1.09296374e-01 6.46935642e-01 -5.06779253e-01 8.16979527e-01
9.79473710e-01 8.68704677e-01 -5.80669582e-01 1.27364397e+00
-3.44205610e-02 3.56225222e-01 -4.50168699e-01 2.11635903e-01
4.67297763e-01 -5.04362583e-01 7.24157572e-01 1.05067432e+00
4.85457361e-01 1.84649095e-01 -8.56789500e-02 1.25713706e+00
-1.47270262e-01 6.09043203e-02 -7.83774078e-01 2.75694132e-01
7.11096823e-01 1.38238585e+00 -8.00758660e-01 -1.98636279e-01
-7.07983315e-01 8.59407067e-01 4.42440003e-01 5.62603235e-01
-7.45482504e-01 -4.10954982e-01 1.25655854e+00 3.27798098e-01
6.31605208e-01 -5.93104005e-01 -6.45393372e-01 -1.06069469e+00
1.18577562e-01 -3.04739267e-01 1.05116382e-01 -8.20797920e-01
-1.74075913e+00 6.61352873e-01 -2.18833417e-01 -1.54608870e+00
5.77406704e-01 -8.57653320e-01 -1.01779640e+00 9.77092326e-01
-1.68321729e+00 -1.34710872e+00 -7.09891260e-01 7.97367454e-01
3.86958122e-01 -6.04549162e-02 4.25696522e-01 4.06140745e-01
-2.71986485e-01 5.34244418e-01 9.45251584e-02 2.73214906e-01
3.78933728e-01 -1.19494092e+00 9.88172233e-01 5.83410621e-01
1.47360146e-01 7.20113814e-01 1.72265321e-01 -6.01122558e-01
-1.52934885e+00 -1.35167599e+00 3.92853290e-01 -5.71920097e-01
6.13991737e-01 -2.64045566e-01 -1.24055314e+00 6.55725956e-01
-6.73136115e-02 4.18915510e-01 2.93990076e-01 -8.59092164e-04
-3.71123374e-01 -5.61736301e-02 -1.07199824e+00 5.62370002e-01
1.44928288e+00 -3.47252220e-01 -4.31220680e-01 3.33257973e-01
9.28120315e-01 -5.65759182e-01 -8.50897789e-01 5.72613239e-01
1.94221616e-01 -9.73193467e-01 1.48842287e+00 -4.90033031e-01
2.46154428e-01 -4.83047128e-01 -1.30422965e-01 -1.08029187e+00
-7.66136229e-01 -4.76652712e-01 -1.52865335e-01 1.01781404e+00
5.43247163e-03 -9.37705994e-01 1.03174794e+00 3.42051178e-01
-6.39638782e-01 -9.54106569e-01 -1.12030995e+00 -8.11308146e-01
3.27975869e-01 -6.75848007e-01 1.18404448e+00 1.25227392e+00
-7.24398136e-01 -7.23656341e-02 3.11843872e-01 6.28544092e-01
7.97277093e-01 3.22826117e-01 8.83311629e-01 -1.57769442e+00
6.03547767e-02 -9.22974527e-01 -8.33415568e-01 -1.17447937e+00
2.64521483e-02 -1.22228718e+00 -4.66782928e-01 -1.54924834e+00
-2.88886875e-01 -8.14650595e-01 -2.20561892e-01 3.58055145e-01
-1.69040576e-01 4.82411355e-01 1.53737664e-01 4.46889430e-01
-3.96634161e-01 6.14487052e-01 1.56594336e+00 -2.21325383e-01
-1.93264887e-01 1.17827952e-01 -7.15275288e-01 8.96120548e-01
8.57063711e-01 -1.31228700e-01 -3.07473093e-01 -8.34792376e-01
6.21191449e-02 -4.41519052e-01 8.95551443e-01 -1.02861714e+00
3.06123167e-01 2.87275948e-02 5.60336709e-01 -9.24657106e-01
3.52400929e-01 -8.85257959e-01 3.28425497e-01 7.08092973e-02
3.45731139e-01 3.28236550e-01 2.84487933e-01 6.35853708e-01
-7.19212890e-02 3.35470252e-02 5.34124255e-01 -2.99702227e-01
-8.72075558e-01 9.61353958e-01 2.41872296e-01 -1.30670533e-01
9.56584632e-01 -5.80321312e-01 -3.39264542e-01 -1.47862304e-02
-6.60269201e-01 2.13138565e-01 7.39222229e-01 3.63554507e-01
1.07344854e+00 -1.53135240e+00 -7.09912837e-01 3.74756515e-01
8.18221793e-02 6.22388899e-01 4.66690630e-01 8.90778601e-01
-8.61376345e-01 1.24192171e-01 -4.31122035e-02 -1.09312129e+00
-1.07127833e+00 5.91279387e-01 5.12168884e-01 2.92387664e-01
-1.15545118e+00 9.94939983e-01 4.53095496e-01 -5.52382112e-01
1.18278056e-01 -6.37883484e-01 4.65800874e-02 -3.09781641e-01
1.41413674e-01 3.72598082e-01 3.55449528e-01 -5.16656339e-01
-4.30957019e-01 9.29601133e-01 -2.66247708e-02 4.65782911e-01
1.45280886e+00 1.55382663e-01 -1.06060244e-01 1.03994183e-01
1.08410764e+00 3.88532616e-02 -1.32942748e+00 -3.69780481e-01
-1.75748929e-01 -7.57178783e-01 1.19494326e-01 -2.57625759e-01
-1.51286161e+00 7.63192475e-01 6.49629772e-01 1.69454724e-01
9.03158247e-01 3.18441212e-01 7.49246836e-01 2.45569378e-01
4.10978734e-01 -5.66957235e-01 -8.04145336e-02 6.11531079e-01
1.00289118e+00 -1.15940034e+00 -7.91605115e-02 -7.39821970e-01
-3.50753546e-01 1.17944634e+00 6.31841123e-01 -8.17096174e-01
9.96422648e-01 1.69454902e-01 1.28313988e-01 -7.69674122e-01
-2.97517776e-01 -3.02545905e-01 5.24154484e-01 7.01465428e-01
1.53721973e-01 9.27899107e-02 9.37102512e-02 3.09690595e-01
-3.62181664e-01 -2.47039482e-01 6.19808547e-02 7.09687352e-01
-2.14105129e-01 -8.10932815e-01 -5.31749427e-01 6.84484363e-01
-1.61519185e-01 -1.84898615e-01 -2.71575332e-01 9.41347361e-01
2.11549923e-01 3.95643830e-01 5.65807760e-01 -4.03200567e-01
5.72888553e-01 -2.00567886e-01 4.19456691e-01 -3.78612876e-01
-5.51992238e-01 -1.07474104e-01 -5.62507629e-01 -7.06823707e-01
-2.83806771e-01 -7.12266445e-01 -1.30364513e+00 -6.13010585e-01
-3.25919896e-01 -3.31120491e-02 5.99383652e-01 5.33850133e-01
8.53801072e-01 5.70542157e-01 5.15591323e-01 -1.25516474e+00
-4.92623717e-01 -6.61861837e-01 -5.42471707e-01 5.47816753e-01
3.43585849e-01 -9.34893429e-01 -3.01512897e-01 -3.87395650e-01] | [7.927119731903076, -3.5872700214385986] |
7d9bf0aa-86a8-4dcd-b126-75b927da5ddd | lidar-iris-for-loop-closure-detection | 1912.03825 | null | https://arxiv.org/abs/1912.03825v3 | https://arxiv.org/pdf/1912.03825v3.pdf | LiDAR Iris for Loop-Closure Detection | In this paper, a global descriptor for a LiDAR point cloud, called LiDAR Iris, is proposed for fast and accurate loop-closure detection. A binary signature image can be obtained for each point cloud after several LoG-Gabor filtering and thresholding operations on the LiDAR-Iris image representation. Given two point clouds, their similarities can be calculated as the Hamming distance of two corresponding binary signature images extracted from the two point clouds, respectively. Our LiDAR-Iris method can achieve a pose-invariant loop-closure detection at a descriptor level with the Fourier transform of the LiDAR-Iris representation if assuming a 3D (x,y,yaw) pose space, although our method can generally be applied to a 6D pose space by re-aligning point clouds with an additional IMU sensor. Experimental results on five road-scene sequences demonstrate its excellent performance in loop-closure detection. | ['Jian Yang', 'Cheng-Zhong Xu', 'Ying Wang', 'Hui Kong', 'Sanjay Sarma', 'Zezhou Sun'] | 2019-12-09 | null | null | null | null | ['loop-closure-detection'] | ['computer-vision'] | [ 4.32592660e-01 -5.18236578e-01 -4.34699744e-01 -2.12971509e-01
-8.05039465e-01 -6.58855855e-01 4.67574894e-01 2.57812947e-01
-3.81563544e-01 1.49493972e-02 -5.02634466e-01 -2.59459674e-01
-2.66882062e-01 -6.14436805e-01 -5.55263042e-01 -4.23388094e-01
-7.08335638e-02 5.78983963e-01 1.88729241e-01 1.77072033e-01
6.74130499e-01 1.02478707e+00 -1.82776606e+00 -4.99016315e-01
7.78724909e-01 1.09433258e+00 4.83571552e-02 6.87428713e-01
1.89641729e-01 -4.45250154e-01 -3.82630318e-01 -1.38072714e-01
7.73294449e-01 -1.40386065e-02 -2.97449380e-01 2.81025231e-01
1.33083451e+00 -2.05504283e-01 -1.33330286e-01 1.35914993e+00
2.23956838e-01 4.57362197e-02 5.17223775e-01 -1.07463789e+00
-1.85352221e-01 -4.41981673e-01 -8.20080519e-01 -2.09388644e-01
6.43484950e-01 -1.23315506e-01 9.06472981e-01 -1.12754452e+00
6.05529726e-01 1.29845858e+00 9.25768793e-01 -9.49627683e-02
-1.40615547e+00 -6.50347173e-01 -6.02132797e-01 1.56000391e-01
-1.95528364e+00 -2.18789168e-02 5.50371587e-01 -4.80512619e-01
8.25794578e-01 4.49406266e-01 6.37812853e-01 6.81837499e-02
2.36331835e-01 1.56405509e-01 1.07691646e+00 -4.45114821e-01
-6.93310574e-02 -1.79846793e-01 -5.81417582e-04 7.66548693e-01
6.06540442e-01 4.86478806e-01 -3.89631540e-01 -3.69495690e-01
8.62121761e-01 2.44690239e-01 -2.16263354e-01 -8.60441267e-01
-1.43630230e+00 5.74000895e-01 5.21350443e-01 -7.70135373e-02
-4.87364493e-02 -8.25312978e-04 1.30539045e-01 2.65893519e-01
1.21022433e-01 3.49495977e-01 3.95933911e-02 -1.65916264e-01
-9.08491313e-01 9.43769962e-02 4.20372009e-01 1.17935169e+00
1.23166525e+00 -3.96835536e-01 3.03168178e-01 3.36579978e-01
4.34599787e-01 1.11400628e+00 1.53269097e-01 -6.79299772e-01
3.47243816e-01 7.60760367e-01 -1.50873391e-02 -1.28434598e+00
-2.63222128e-01 -7.78451636e-02 -6.88028395e-01 3.77151459e-01
2.44276866e-01 2.85520554e-01 -7.22288072e-01 9.57006812e-01
6.91416204e-01 6.00567043e-01 -3.28493793e-03 8.41558099e-01
4.24345493e-01 1.74216494e-01 -6.65834010e-01 -1.32785559e-01
1.39488995e+00 -2.83883065e-01 -3.39742690e-01 -2.91122228e-01
4.52486813e-01 -1.18051255e+00 7.36452520e-01 1.78646222e-01
-4.92979556e-01 -7.24523425e-01 -1.46853220e+00 -7.38933980e-02
-1.62700474e-01 5.01400530e-01 1.82318483e-02 6.19750381e-01
-6.75062120e-01 5.17143011e-01 -8.95038247e-01 -4.74296927e-01
-2.84386188e-01 4.06256497e-01 -4.38024640e-01 1.32739216e-01
-6.17968380e-01 9.24919367e-01 2.71068901e-01 5.64911356e-03
-9.98702049e-02 -2.19074115e-01 -1.09299958e+00 -4.13127452e-01
1.27355114e-01 -3.26014727e-01 8.10360491e-01 -4.18499969e-02
-1.23068285e+00 1.31221402e+00 -4.64740843e-01 -3.46798211e-01
2.41474926e-01 -1.52504399e-01 -2.07855925e-01 3.00178468e-01
2.98228413e-01 2.54460126e-01 1.30270553e+00 -1.10309267e+00
-7.55408645e-01 -8.49241674e-01 -5.69068670e-01 2.17219695e-01
1.96269527e-01 -6.39576986e-02 -4.37018037e-01 -3.38039905e-01
1.11069405e+00 -1.38711560e+00 -1.19082211e-02 1.78690150e-01
-3.62728566e-01 -3.22900638e-02 1.21295917e+00 -2.05521822e-01
1.00807905e+00 -2.46940613e+00 -1.49595752e-01 5.07327795e-01
-1.10784337e-01 3.78250211e-01 -6.11563176e-02 2.24334061e-01
-8.20837095e-02 -5.33865914e-02 -1.67548597e-01 -3.36268455e-01
-2.75939912e-01 3.07891041e-01 -5.84909916e-01 1.15382850e+00
1.17889471e-01 4.10840720e-01 -8.40871871e-01 -6.18215561e-01
5.47165155e-01 3.84651273e-01 -1.41695067e-01 -1.38002083e-01
3.60153794e-01 3.29255760e-01 -3.14584225e-01 8.27882826e-01
1.19459236e+00 2.85130203e-01 -2.03856990e-01 -2.36711398e-01
-3.79596412e-01 2.98406243e-01 -1.55703008e+00 1.66086400e+00
-1.75083295e-01 8.54644716e-01 1.14751756e-02 -6.17242277e-01
1.55812657e+00 1.17863212e-02 4.74054456e-01 -3.98018569e-01
-3.79929878e-02 4.93703336e-01 -3.00774097e-01 -2.39584371e-02
7.85109460e-01 8.40805918e-02 -2.12632358e-01 2.04880238e-01
-3.06636214e-01 -8.08806658e-01 3.07823792e-02 -3.88379604e-01
4.92971867e-01 -3.42081413e-02 3.29777747e-01 -1.99768953e-02
7.87807524e-01 -4.31152619e-03 5.75412691e-01 5.33265173e-01
-2.81806976e-01 8.26642632e-01 -1.12466644e-02 -2.56800205e-01
-8.13039243e-01 -1.24387586e+00 -9.30235088e-01 1.60096288e-01
6.79344416e-01 -6.48312390e-01 -3.27635407e-01 -3.78716379e-01
5.97809494e-01 1.27811342e-01 -9.39738452e-02 -3.08844224e-02
-5.50355494e-01 5.54213375e-02 6.46520555e-01 2.24719524e-01
5.50573170e-01 -7.46802166e-02 -9.75523710e-01 -4.50142995e-02
5.34656495e-02 -1.19401693e+00 -6.85246110e-01 -8.84386450e-02
-1.28480887e+00 -1.36779094e+00 -1.76327378e-01 -8.61230135e-01
7.01542318e-01 7.32419729e-01 4.65157151e-01 1.48264453e-01
-5.34785688e-01 4.59205180e-01 1.19419433e-01 -2.98010230e-01
-5.36742844e-02 -5.26006162e-01 5.73119342e-01 1.48051277e-01
7.20707297e-01 -3.57307196e-01 -3.39784086e-01 7.99384177e-01
-4.11602616e-01 -5.01215219e-01 2.30861187e-01 7.54635870e-01
1.11182487e+00 -6.36645779e-02 -4.41846758e-01 1.34274825e-01
2.39802822e-01 3.84529203e-01 -1.18993306e+00 8.21716711e-02
-7.03409553e-01 3.19437608e-02 -1.65101483e-01 -3.50635886e-01
-2.39770472e-01 5.49447775e-01 4.67574477e-01 -9.01647985e-01
-1.43455699e-01 4.30623919e-01 2.30184898e-01 -7.23069191e-01
7.53539443e-01 2.91533470e-01 3.46931666e-01 -3.93587708e-01
4.42026824e-01 1.00662220e+00 1.11850965e+00 -5.46272695e-01
1.40913892e+00 6.69410527e-01 7.05215693e-01 -1.20627558e+00
-1.90678075e-01 -1.18484020e+00 -1.31274045e+00 -1.85541231e-02
5.98183572e-01 -1.04125035e+00 -1.00764465e+00 3.99251938e-01
-1.29461825e+00 7.84700334e-01 -1.48131043e-01 8.15794170e-01
-7.09524632e-01 8.64468515e-01 3.97994788e-03 -7.81165719e-01
-1.99501649e-01 -1.18715787e+00 1.61170959e+00 8.37479532e-02
-1.46340668e-01 -5.62049091e-01 3.65492612e-01 4.10336256e-03
-2.48236984e-01 4.82350379e-01 4.37768638e-01 -2.07336232e-01
-8.99421692e-01 -8.28256190e-01 -2.20748469e-01 3.03688347e-01
2.39113569e-01 4.83384401e-01 -7.52682626e-01 -5.11023879e-01
8.18863809e-02 8.02244917e-02 5.25753081e-01 6.48085847e-02
5.50203919e-01 6.70254081e-02 -6.05370939e-01 9.69879329e-01
1.49571228e+00 4.26033139e-02 5.69579363e-01 4.06181425e-01
4.89468306e-01 1.84256598e-01 1.23145664e+00 2.23966062e-01
1.62397474e-01 9.48485613e-01 5.40558577e-01 1.54999986e-01
1.01588648e-02 -4.73349601e-01 3.39719594e-01 5.70207655e-01
-1.07200123e-01 7.98048258e-01 -1.14790154e+00 4.75951672e-01
-1.66905868e+00 -7.95767725e-01 -4.15354729e-01 3.05472255e+00
5.08297682e-01 -3.34978104e-03 -1.78708270e-01 2.90653020e-01
1.01019681e+00 3.39166038e-02 -4.60624963e-01 -2.98159480e-01
3.30625623e-02 1.83771074e-01 9.59537506e-01 5.37991524e-01
-1.31289113e+00 7.10873306e-01 6.46745014e+00 5.67847252e-01
-1.35006750e+00 -3.94590020e-01 -4.60106641e-01 2.49904543e-01
2.73126811e-01 3.91292959e-01 -1.08836293e+00 1.67801484e-01
6.02101564e-01 -3.94592255e-01 6.90332428e-02 9.53136802e-01
-5.45475744e-02 -4.20056611e-01 -1.11703765e+00 1.51343918e+00
7.24271461e-02 -1.06170881e+00 -2.26616398e-01 4.00229454e-01
5.36907434e-01 2.67812580e-01 3.28774065e-01 -2.56588578e-01
-4.06170636e-01 -6.18294716e-01 4.09923106e-01 2.91597158e-01
1.29790771e+00 -7.63431191e-01 1.89996898e-01 5.13092399e-01
-1.70071530e+00 1.52861610e-01 -6.69008672e-01 -5.79099022e-02
-1.41185030e-01 3.53665233e-01 -1.23566020e+00 5.74898541e-01
6.16347611e-01 9.20647204e-01 -5.57307303e-01 1.34235179e+00
-1.27101019e-01 -9.46031418e-03 -7.61868536e-01 2.63991147e-01
-6.85409978e-02 -8.91140163e-01 1.07658041e+00 8.27815115e-01
7.12801754e-01 -2.26138532e-02 3.81942779e-01 6.75556362e-01
4.88933742e-01 1.37960732e-01 -1.03314793e+00 6.39870716e-03
5.90574265e-01 1.08882093e+00 -4.68781024e-01 -1.14535242e-01
-2.31250376e-01 6.81774318e-01 -2.87898898e-01 -1.08482046e-02
-4.85545605e-01 -9.48084593e-01 1.21570623e+00 1.59607515e-01
4.30596948e-01 -8.62684965e-01 -3.79210472e-01 -1.12231171e+00
4.00816239e-02 -5.95733583e-01 2.82544992e-04 -7.17235982e-01
-7.24803388e-01 1.41358241e-01 -3.50105576e-02 -2.16517544e+00
-3.22423667e-01 -6.76502049e-01 -5.28324068e-01 9.73744452e-01
-1.31790209e+00 -9.62753117e-01 -2.20746234e-01 7.78591156e-01
-1.85861066e-01 6.42827228e-02 9.64060187e-01 7.50149786e-02
-1.48754090e-01 5.32665014e-01 2.88429856e-01 2.84930199e-01
7.50737965e-01 -1.09641373e+00 6.67132318e-01 9.58630979e-01
4.56311911e-01 8.69693398e-01 7.20764220e-01 -8.05441618e-01
-1.75929594e+00 -9.96724486e-01 9.87522721e-01 -6.69227839e-01
5.76854765e-01 -1.07793637e-01 -7.45725632e-01 6.97525442e-01
-3.66346002e-01 1.81001395e-01 3.46297532e-01 -9.75945219e-02
-5.34587502e-01 -1.60316229e-01 -1.24850667e+00 2.57509023e-01
9.26618934e-01 -1.05204248e+00 -9.87910986e-01 4.17140663e-01
3.37882876e-01 -1.04840326e+00 -1.04422128e+00 6.68888807e-01
4.44992632e-01 -8.04245353e-01 1.30290318e+00 -3.96566056e-02
-4.47527617e-01 -1.01614666e+00 -1.91019446e-01 -8.35173666e-01
-6.64789416e-03 -7.44789839e-01 2.38370806e-01 7.00188458e-01
-3.39299113e-01 -7.68552363e-01 6.55964077e-01 -1.63631424e-01
2.14023978e-01 -2.14548409e-01 -1.49430311e+00 -1.12775505e+00
-4.70883340e-01 -5.05252063e-01 6.96038067e-01 5.17021358e-01
9.21169817e-02 2.16972262e-01 2.99571324e-02 8.67417276e-01
1.06771612e+00 6.55833483e-01 1.34291136e+00 -1.71595311e+00
2.83898264e-01 -1.85377300e-01 -1.35866654e+00 -1.37173951e+00
6.49784356e-02 -9.48667765e-01 1.07290484e-01 -9.13106740e-01
-3.87966365e-01 -3.14938664e-01 3.36968273e-01 1.27148047e-01
2.33232364e-01 3.88846040e-01 2.28182614e-01 5.27445614e-01
4.64326255e-02 2.98305809e-01 1.06900489e+00 -2.33269647e-01
-1.83313936e-01 3.73447418e-01 2.06465587e-01 6.91306531e-01
3.38962048e-01 -3.64777684e-01 -8.04807642e-04 -4.39004079e-02
-6.74030483e-02 1.43415472e-02 4.17981505e-01 -1.34663022e+00
3.99634629e-01 -8.75935256e-02 6.60498217e-02 -1.38060009e+00
5.83099127e-01 -1.04239810e+00 1.72588006e-01 6.81875944e-01
3.35010439e-01 2.36830860e-01 3.10992122e-01 7.26422131e-01
-5.71448565e-01 -1.63861319e-01 9.24447656e-01 3.67014378e-01
-6.81075096e-01 3.31061333e-01 2.11532116e-01 -5.22500098e-01
1.14452291e+00 -8.68072748e-01 -1.23092726e-01 -5.78516498e-02
-4.10602689e-01 1.01305313e-01 1.12462366e+00 3.42571378e-01
1.03253698e+00 -1.66215408e+00 -5.16985476e-01 1.04305279e+00
6.62312567e-01 1.62287518e-01 -2.63353795e-01 9.53901291e-01
-7.08372951e-01 6.69050932e-01 -6.93925396e-02 -1.55976343e+00
-1.91169810e+00 5.33027589e-01 4.08490360e-01 2.72981375e-01
-4.88376230e-01 5.66297352e-01 -3.41900468e-01 -5.49106419e-01
1.65890828e-01 -8.14089358e-01 3.11484784e-01 1.43332377e-01
5.11702061e-01 3.49260211e-01 2.70188332e-01 -1.28707802e+00
-5.54718018e-01 1.79304159e+00 3.67637634e-01 -4.23717238e-02
5.55209935e-01 -1.41094118e-01 -3.29051465e-01 2.35546216e-01
1.50655174e+00 2.25224361e-01 -9.57259893e-01 -4.29873437e-01
1.43105254e-01 -1.12469566e+00 -9.32288691e-02 2.46216282e-01
-6.17243230e-01 9.68555689e-01 9.46006715e-01 -3.48168574e-02
9.22121227e-01 -2.22047105e-01 5.13619840e-01 5.93184412e-01
5.76700747e-01 -8.88200045e-01 -3.86131495e-01 6.76287591e-01
9.00634229e-01 -1.14831197e+00 3.72970104e-01 -4.49949145e-01
-1.55350104e-01 1.32125652e+00 3.48399132e-02 -3.07918042e-01
7.18500376e-01 -1.28269583e-01 1.55561104e-01 -9.88252386e-02
-1.10070338e-03 -5.54248631e-01 6.42867446e-01 8.57245624e-01
-9.48773548e-02 2.32790232e-01 -1.87456369e-01 -5.59903383e-01
-3.47243011e-01 -2.27262676e-01 3.57423186e-01 9.92735803e-01
-8.01929653e-01 -1.18297017e+00 -1.15299273e+00 -1.67829450e-02
1.84750602e-01 2.10785523e-01 -3.33852977e-01 7.47033179e-01
2.15672031e-02 7.55603790e-01 5.18633902e-01 -7.54048944e-01
6.12645507e-01 -7.23095536e-02 5.90139329e-01 -6.89277232e-01
-1.44382805e-01 1.70529932e-01 -3.81212622e-01 -7.42774248e-01
-3.09709460e-01 -9.32352781e-01 -1.29663229e+00 -2.09945783e-01
-5.36003172e-01 1.56893656e-02 1.15113485e+00 3.81065667e-01
4.69714612e-01 -5.82475364e-01 9.13174033e-01 -9.62340772e-01
-9.70797360e-01 -4.88411129e-01 -8.39393198e-01 2.22290650e-01
7.68227339e-01 -9.21499789e-01 -5.90345502e-01 -1.74956471e-01] | [7.567543983459473, -2.501615047454834] |
a11dc51c-bd4b-4f8c-8de8-faf996f40c97 | learning-to-predict-indoor-illumination-from | 1704.00090 | null | http://arxiv.org/abs/1704.00090v3 | http://arxiv.org/pdf/1704.00090v3.pdf | Learning to Predict Indoor Illumination from a Single Image | We propose an automatic method to infer high dynamic range illumination from
a single, limited field-of-view, low dynamic range photograph of an indoor
scene. In contrast to previous work that relies on specialized image capture,
user input, and/or simple scene models, we train an end-to-end deep neural
network that directly regresses a limited field-of-view photo to HDR
illumination, without strong assumptions on scene geometry, material
properties, or lighting. We show that this can be accomplished in a three step
process: 1) we train a robust lighting classifier to automatically annotate the
location of light sources in a large dataset of LDR environment maps, 2) we use
these annotations to train a deep neural network that predicts the location of
lights in a scene from a single limited field-of-view photo, and 3) we
fine-tune this network using a small dataset of HDR environment maps to predict
light intensities. This allows us to automatically recover high-quality HDR
illumination estimates that significantly outperform previous state-of-the-art
methods. Consequently, using our illumination estimates for applications like
3D object insertion, we can achieve results that are photo-realistic, which is
validated via a perceptual user study. | ['Jean-François Lalonde', 'Christian Gagné', 'Kalyan Sunkavalli', 'Marc-André Gardner', 'Ersin Yumer', 'Xiaohui Shen', 'Emiliano Gambaretto'] | 2017-04-01 | null | null | null | null | ['lighting-estimation'] | ['computer-vision'] | [ 4.16460127e-01 -2.72991627e-01 5.29125512e-01 -7.17631459e-01
-6.00559413e-01 -7.96606421e-01 3.67034048e-01 -1.15209259e-01
-3.66630763e-01 4.47901160e-01 2.20258236e-01 -2.56135523e-01
3.51343751e-01 -7.91331351e-01 -1.11641264e+00 -3.46853405e-01
3.91014278e-01 2.06345305e-01 1.87498733e-01 -4.05726545e-02
1.71465456e-01 7.39345789e-01 -1.80704486e+00 7.58518949e-02
6.49855554e-01 1.14695013e+00 3.81072074e-01 1.09567881e+00
2.40162075e-01 8.11438859e-01 -3.72869283e-01 3.01403999e-02
6.92050397e-01 -3.11652899e-01 -5.08529782e-01 4.02909309e-01
9.41283345e-01 -9.91939008e-01 -2.82197386e-01 7.24074364e-01
7.19429731e-01 1.75347418e-01 3.62130314e-01 -8.09603035e-01
-5.56257725e-01 -2.04931691e-01 -5.17088473e-01 -9.83338654e-02
5.73663712e-01 6.03291750e-01 5.03954291e-01 -5.89883268e-01
4.69103396e-01 9.76487935e-01 7.10248291e-01 7.10535869e-02
-1.27521098e+00 -5.49118280e-01 1.24427326e-01 -2.34164953e-01
-1.29360092e+00 -5.96702039e-01 9.10527706e-01 -4.34793144e-01
1.04249036e+00 1.33167645e-02 7.64792383e-01 9.32563007e-01
-7.68601745e-02 1.78415984e-01 1.46632302e+00 -5.12477398e-01
3.77503425e-01 5.87445199e-02 -3.68554205e-01 6.28834903e-01
-3.43150526e-01 6.07206561e-02 -2.55000770e-01 2.09791169e-01
8.03404272e-01 -9.81508661e-03 -4.87426847e-01 -3.92414540e-01
-1.10724831e+00 2.61280745e-01 8.33028078e-01 -2.34971911e-01
-2.55570322e-01 1.76900670e-01 -4.36736569e-02 7.43658170e-02
7.08303094e-01 4.07679021e-01 -5.56048334e-01 6.94800019e-02
-6.98447704e-01 -7.98184052e-02 4.69635129e-01 6.03072047e-01
1.06836045e+00 -1.85645655e-01 2.26576641e-01 8.64102304e-01
2.24500820e-01 9.20599997e-01 -1.45162329e-01 -1.38854444e+00
2.85157770e-01 3.69053036e-01 4.45700914e-01 -4.88562614e-01
-4.42925304e-01 -1.40566692e-01 -4.79281992e-01 6.13884330e-01
3.06193084e-01 -2.67255068e-01 -1.19302833e+00 1.54945922e+00
3.79020751e-01 4.29585800e-02 -2.20907658e-01 1.21115041e+00
3.98566961e-01 6.17732346e-01 -3.96307409e-01 8.20827931e-02
9.40476120e-01 -6.84933245e-01 -2.88994640e-01 -3.88866276e-01
1.75717920e-01 -9.25397456e-01 1.42943263e+00 5.37448227e-01
-1.14007246e+00 -7.16722488e-01 -9.60318327e-01 -4.86701041e-01
-3.59570444e-01 3.35803419e-01 6.29812479e-01 5.12863517e-01
-1.22534263e+00 3.89222831e-01 -5.83173931e-01 -5.59158564e-01
2.80297965e-01 1.68728948e-01 -2.16986731e-01 -5.37016392e-01
-7.36195803e-01 8.04897606e-01 5.54617755e-02 1.90597251e-01
-1.10211897e+00 -8.69577229e-01 -8.69121969e-01 -1.22577474e-02
3.37867141e-01 -8.72217834e-01 1.09160745e+00 -9.81489480e-01
-1.74632549e+00 8.10778201e-01 -1.99904755e-01 3.10589671e-02
3.05464685e-01 -2.83917785e-01 -4.46937419e-02 1.74803287e-01
-1.05710283e-01 6.17209792e-01 6.25332654e-01 -1.68683755e+00
-5.83439827e-01 -3.41017216e-01 4.10698712e-01 3.69922727e-01
-1.10777862e-01 5.77111989e-02 -6.16370976e-01 5.01603968e-02
-2.31562648e-02 -7.80274332e-01 -1.82723492e-01 3.62199336e-01
-4.41350460e-01 4.13344949e-01 7.75608659e-01 -5.45880139e-01
4.57798570e-01 -2.03939080e+00 -2.56169319e-01 1.14324629e-01
-5.17508388e-02 -1.27646048e-02 -5.58877476e-02 1.20561041e-01
-2.89898645e-02 -2.53568918e-01 -7.72341713e-02 -5.21676600e-01
-1.73717842e-01 -1.01216644e-01 -4.27390248e-01 5.33860922e-01
-1.38785345e-02 5.26685774e-01 -8.91730487e-01 -6.00838661e-03
9.29603040e-01 7.99431860e-01 -4.98412609e-01 7.26607919e-01
-4.94502693e-01 6.31911576e-01 1.14485428e-01 5.82232654e-01
8.17091405e-01 -1.99453339e-01 3.63859162e-03 -5.76036274e-01
-2.75873959e-01 1.84108034e-01 -1.17512822e+00 1.94455683e+00
-1.15080130e+00 8.68358433e-01 -4.46982961e-03 -2.34745964e-01
8.35744977e-01 -7.91994929e-02 4.66209352e-01 -8.83941412e-01
2.08289802e-01 -4.43772525e-02 -6.14761055e-01 -5.71263790e-01
3.00864011e-01 2.29124762e-02 1.30969167e-01 6.49811506e-01
-1.66383624e-01 -4.37850207e-01 -2.14939326e-01 -4.34957594e-02
1.13559961e+00 4.89340633e-01 -5.43959700e-02 3.76115352e-01
1.90976098e-01 -3.50158334e-01 3.42539072e-01 5.31358600e-01
1.15675047e-01 1.06831169e+00 -8.48275572e-02 -5.54244518e-01
-1.41788280e+00 -1.27596891e+00 -3.92455198e-02 9.68841434e-01
1.88097060e-01 6.94683939e-02 -6.34647846e-01 -3.44350159e-01
-1.82314143e-01 7.38581181e-01 -4.05727237e-01 2.77936220e-01
-4.94428068e-01 -5.28563678e-01 3.16760056e-02 5.41188240e-01
7.69624412e-01 -7.69113779e-01 -9.38921869e-01 -1.57344863e-01
-1.94170594e-01 -1.54502153e+00 -3.46642047e-01 3.03917080e-01
-2.85636365e-01 -1.08076274e+00 -4.20401812e-01 -2.87699342e-01
7.19925225e-01 6.27788961e-01 1.54277885e+00 -2.24446923e-01
-5.35624623e-01 5.75848281e-01 -1.31689519e-01 -3.95092696e-01
-1.71180785e-01 -3.29838574e-01 -8.41593817e-02 7.49418838e-03
1.12757266e-01 -6.97732687e-01 -1.14574468e+00 4.16686922e-01
-7.52932072e-01 4.07451153e-01 4.56997812e-01 2.53213584e-01
6.79572403e-01 2.10393265e-01 -5.91294020e-02 -6.69969857e-01
-1.80330686e-02 -6.18062690e-02 -1.14891315e+00 1.26379758e-01
-4.26509023e-01 -2.15916321e-01 7.40417719e-01 -2.55397499e-01
-1.45273685e+00 5.81628859e-01 -1.85613677e-01 -7.56262600e-01
-6.24113023e-01 -2.60982007e-01 -4.11496103e-01 -1.86596587e-01
8.72575223e-01 8.10200348e-03 -4.71532643e-01 -2.74605691e-01
6.08959854e-01 7.76075959e-01 7.81964660e-01 -5.06848454e-01
9.62384760e-01 8.53907466e-01 7.85137117e-02 -8.58902454e-01
-1.14021611e+00 -5.03528655e-01 -8.29019845e-01 -4.15485114e-01
9.94641960e-01 -1.26812744e+00 -9.37829196e-01 5.41299284e-01
-1.00205553e+00 -1.05983675e+00 -1.31330714e-01 4.27719235e-01
-6.37519002e-01 7.47744180e-03 -4.20084894e-01 -9.40248668e-01
-9.83221754e-02 -9.68167007e-01 1.60671043e+00 2.97913998e-01
2.41329685e-01 -8.38067174e-01 7.65264556e-02 5.53683579e-01
3.90170246e-01 3.89951766e-01 7.04682946e-01 4.47579712e-01
-9.79381859e-01 3.10052373e-02 -5.38607061e-01 5.02714217e-01
2.27415621e-01 1.30937442e-01 -1.61843002e+00 -1.99677497e-01
-1.43385679e-01 -6.06976688e-01 5.46474874e-01 6.32867515e-01
1.43599427e+00 9.59523674e-03 -8.19664001e-02 1.13426387e+00
1.85013545e+00 -2.74285257e-01 7.35721290e-01 2.77720422e-01
9.81105864e-01 3.91526014e-01 5.85283756e-01 4.03465062e-01
6.63506210e-01 8.83441627e-01 6.96930110e-01 -7.44198859e-01
-3.91458482e-01 -2.71467596e-01 1.60228208e-01 1.54385358e-01
-2.55840365e-02 -3.13351125e-01 -7.52243042e-01 4.46213514e-01
-1.43399727e+00 -6.75099254e-01 9.49868634e-02 2.39612293e+00
8.01079333e-01 -9.25665870e-02 1.06074035e-01 -1.09086342e-01
3.85785848e-01 1.01835117e-01 -8.30422103e-01 -2.72089064e-01
-1.13336191e-01 9.48643778e-03 9.76129651e-01 6.28676832e-01
-8.85932863e-01 8.55705857e-01 6.38026285e+00 1.77122261e-02
-1.50461435e+00 -9.40387696e-02 7.70395219e-01 -3.81793499e-01
-3.87492746e-01 -8.55651721e-02 -5.22418320e-01 4.26005006e-01
8.88719380e-01 6.27064705e-01 1.15011442e+00 6.82218671e-01
5.87988496e-01 -3.83441269e-01 -1.24762309e+00 1.13101923e+00
3.84675950e-01 -9.80205834e-01 -4.89938319e-01 1.92135215e-01
7.83593535e-01 4.93425846e-01 4.39767316e-02 -1.98105589e-01
7.34460413e-01 -8.40829909e-01 5.78401268e-01 6.12862647e-01
1.06414139e+00 -4.93467957e-01 2.76210427e-01 2.69511849e-01
-9.26748335e-01 -2.00868472e-01 -5.28201520e-01 3.32992785e-02
3.77077669e-01 7.99250603e-01 -9.27020967e-01 2.83301890e-01
9.95659769e-01 4.77939874e-01 -7.22265959e-01 1.03711331e+00
-5.15938699e-01 3.39425772e-01 -5.49402714e-01 5.57185709e-01
-3.08196545e-01 -7.29191005e-02 -5.31420074e-02 9.08768415e-01
2.80164093e-01 1.80693895e-01 1.16882861e-01 1.01514840e+00
-2.39978001e-01 -3.76254827e-01 -7.26703405e-01 5.87032616e-01
2.31257319e-01 1.50777936e+00 -6.00499690e-01 4.80102003e-02
-4.08283740e-01 1.25276923e+00 3.17994088e-01 7.02290177e-01
-7.28806853e-01 -3.45413625e-01 5.30189097e-01 3.56898278e-01
3.49199951e-01 -3.15154940e-01 -2.36971334e-01 -1.17448509e+00
5.59721626e-02 -5.35659969e-01 -2.30523929e-01 -1.75475979e+00
-1.18188679e+00 5.02846956e-01 -1.99006930e-01 -9.58870709e-01
-1.87079594e-01 -6.13795340e-01 -5.34416795e-01 9.89459813e-01
-1.85831559e+00 -1.39398468e+00 -9.55074847e-01 6.24392748e-01
3.85223567e-01 4.16929066e-01 7.67770171e-01 2.75615871e-01
-4.25857246e-01 1.20181896e-01 2.41309274e-02 8.02112520e-02
9.69809413e-01 -1.41453612e+00 4.87439603e-01 1.10175276e+00
-1.27471471e-02 2.49402657e-01 7.52879202e-01 -2.70462841e-01
-1.49313283e+00 -1.21675944e+00 3.60847503e-01 -7.49333680e-01
1.62179023e-01 -8.02759886e-01 -5.31319439e-01 6.42701030e-01
1.85770571e-01 4.64518487e-01 4.68053222e-01 2.73297966e-01
-4.43181783e-01 -5.92247486e-01 -1.22635865e+00 5.24048746e-01
1.16315651e+00 -7.17622697e-01 4.65521626e-02 5.14918685e-01
7.30294049e-01 -6.81279659e-01 -8.96653652e-01 2.19406649e-01
7.20617652e-01 -1.30339766e+00 1.18580127e+00 9.68124568e-02
4.53949720e-01 -6.52217388e-01 -3.97928417e-01 -1.43112040e+00
8.68896767e-03 -4.71971571e-01 1.25853658e-01 1.21031475e+00
1.36530399e-01 -3.21311831e-01 5.25569260e-01 9.64943230e-01
-9.33631212e-02 -4.59864140e-01 -2.96128422e-01 -3.29971552e-01
-4.00714844e-01 -4.63824034e-01 6.32415950e-01 5.43600559e-01
-7.83440709e-01 4.32585120e-01 -4.30302262e-01 5.14891744e-01
8.85840714e-01 3.89693528e-01 1.33150947e+00 -1.09661162e+00
-4.08418149e-01 1.67019770e-01 -1.24224052e-01 -1.12753427e+00
1.72718138e-01 -2.81224877e-01 4.42959666e-01 -1.87686419e+00
1.84705719e-01 -6.94294870e-01 -7.92394653e-02 4.80040252e-01
-5.70938252e-02 6.91371202e-01 -7.75860325e-02 4.81804013e-02
-7.29460001e-01 4.44458157e-01 9.37089920e-01 1.29849523e-01
-3.03406000e-01 -1.90847680e-01 -5.28980374e-01 7.11136460e-01
5.70966005e-01 -4.91925590e-02 -5.33167601e-01 -8.19253862e-01
4.83302355e-01 -1.41916767e-01 5.05169272e-01 -1.16338074e+00
5.42347692e-02 -2.91569769e-01 1.09599483e+00 -5.08650362e-01
6.02295876e-01 -1.06779325e+00 4.41399813e-02 -1.42075196e-01
-2.43134648e-01 -4.38166380e-01 9.25411955e-02 3.55567187e-01
4.11086559e-01 1.14824884e-01 8.84239197e-01 -3.07451606e-01
-6.20887458e-01 2.96537071e-01 1.27189398e-01 -1.11589082e-01
8.03923130e-01 -2.07143009e-01 -4.91957963e-01 -5.24235666e-01
-9.50204581e-02 -2.93172500e-03 1.16743231e+00 2.46193364e-01
5.30516148e-01 -1.06759226e+00 -4.45235968e-01 4.37173039e-01
2.54980534e-01 3.82989317e-01 3.24828774e-01 2.58854240e-01
-7.27742910e-01 1.38346612e-01 -1.83340877e-01 -9.03840065e-01
-1.06709623e+00 6.74392641e-01 6.27380610e-01 3.00279886e-01
-8.03333759e-01 3.81010294e-01 3.61118108e-01 -4.73425925e-01
1.28117036e-02 -3.10085684e-01 2.82457590e-01 -6.55635536e-01
4.74956691e-01 3.73265264e-03 1.58566505e-01 -5.27269721e-01
-1.13829635e-01 9.36441422e-01 3.74681622e-01 -9.27674547e-02
1.56439602e+00 -7.01431453e-01 9.72449109e-02 5.19032240e-01
1.26955676e+00 2.32076108e-01 -1.96566474e+00 -1.37853250e-01
-9.96360004e-01 -1.04377151e+00 4.70557362e-01 -1.14856470e+00
-1.01152599e+00 8.80681515e-01 9.11557019e-01 -2.61701643e-01
1.63455749e+00 -1.20698974e-01 7.21678376e-01 4.47596133e-01
4.54159290e-01 -1.13045895e+00 5.14677204e-02 2.68177122e-01
5.22253036e-01 -1.53234267e+00 3.23573388e-02 -2.31728986e-01
-3.75468254e-01 1.05095840e+00 6.50451839e-01 1.12255506e-01
4.73032624e-01 4.62680787e-01 3.74716669e-01 -7.52026215e-02
-5.96429527e-01 -3.62915456e-01 7.94229433e-02 8.37220609e-01
2.29608312e-01 -3.71218443e-01 9.23345685e-01 -2.85190374e-01
-2.36860067e-01 2.54880786e-01 6.32234633e-01 5.28783679e-01
-5.51677227e-01 -6.21856987e-01 -3.64512831e-01 6.69790506e-02
-1.38187841e-01 -4.03616838e-02 -1.93944484e-01 3.36056590e-01
2.62001723e-01 1.10759377e+00 1.49206877e-01 -2.18333215e-01
4.28069413e-01 -3.31491828e-01 7.44771302e-01 -4.94922578e-01
-1.68538764e-01 4.56215627e-02 -8.05840120e-02 -8.57545435e-01
-4.03429806e-01 -3.17424834e-01 -7.25578070e-01 -1.88754588e-01
-1.37937009e-01 -6.92204714e-01 1.19462705e+00 7.90619314e-01
4.24950570e-01 5.15152693e-01 1.07113981e+00 -1.32566607e+00
1.47439599e-01 -6.47925317e-01 -4.61207300e-01 5.66134632e-01
5.79246998e-01 -4.34620500e-01 -4.45861071e-01 1.85809821e-01] | [9.736108779907227, -2.959409475326538] |
bb0ba082-383f-4d20-9301-8252467c01d7 | response-to-moffat-s-comment-on-towards | 2212.11735 | null | https://arxiv.org/abs/2212.11735v1 | https://arxiv.org/pdf/2212.11735v1.pdf | Response to Moffat's Comment on "Towards Meaningful Statements in IR Evaluation: Mapping Evaluation Measures to Interval Scales" | Moffat recently commented on our previous work. Our work focused on how laying the foundations of our evaluation methodology into the theory of measurement can improve our knowledge and understanding of the evaluation measures we use in IR and how it can shed light on the different types of scales adopted by our evaluation measures; we also provided evidence, through extensive experimentation, on the impact of the different types of scales on the statistical analyses, as well as on the impact of departing from their assumptions. Moreover, we investigated, for the first time in IR, the concept of meaningfulness, i.e. the invariance of the experimental statements and inferences you draw, and proposed it as a way to ensure more valid and generalizabile results. Moffat's comments build on: (i) misconceptions about the representational theory of measurement, such as what an interval scale actually is and what axioms it has to comply with; (ii) they totally miss the central concept of meaningfulness. Therefore, we reply to Moffat's comments by properly framing them in the representational theory of measurement and in the concept of meaningfulness. All in all, we can only reiterate what we said several times: the goal of this research line is to theoretically ground our evaluation methodology - and IR is a field where it is extremely challenging to perform any theoretical advances - in order to aim for more robust and generalizable inferences - something we currently lack in the field. Possibly there are other and better ways to achieve this objective and these proposals could emerge from an open discussion in the field and from the work of others. On the other hand, reducing everything to a contrast on what is (or pretend to be) an interval scale or whether all or none evaluation measures are interval scales may be more a barrier from than a help in progressing towards this goal. | ['Norbert Fuhr', 'Nicola Ferro', 'Marco Ferrante'] | 2022-12-22 | null | null | null | null | ['misconceptions'] | ['miscellaneous'] | [ 2.36020610e-01 2.53896534e-01 -2.88441867e-01 -5.46883345e-01
-4.78767931e-01 -6.35250270e-01 5.86530685e-01 4.22463566e-01
-5.14671087e-01 5.30758977e-01 6.02679133e-01 -8.36784244e-01
-7.82092214e-01 -7.75994658e-01 -5.64141572e-01 -3.27958077e-01
2.74162710e-01 2.25022569e-01 8.46310705e-02 -4.14205641e-01
7.03798652e-01 4.51376617e-01 -1.47249424e+00 1.85761862e-02
5.62373340e-01 7.76859641e-01 -2.51263052e-01 2.75934070e-01
-1.16155133e-01 8.85853589e-01 -5.87683678e-01 -6.34779096e-01
1.04666889e-01 -6.10852361e-01 -1.19564426e+00 4.00292192e-04
1.02237053e-01 -2.14359671e-01 2.26013184e-01 1.20389616e+00
2.08984986e-01 8.31978954e-03 6.80693626e-01 -8.72055769e-01
-9.48088646e-01 6.48075879e-01 -2.10041538e-01 1.01100154e-01
5.08239985e-01 7.09254071e-02 9.46396589e-01 -3.75285119e-01
4.09323812e-01 1.25501871e+00 4.99220908e-01 1.24872416e-01
-1.13314879e+00 -4.62549001e-01 1.78452954e-02 6.29035570e-03
-1.27570856e+00 -4.30654466e-01 4.00981426e-01 -5.87554395e-01
2.88622230e-01 5.63674629e-01 5.21377802e-01 8.57022941e-01
1.56352147e-01 -9.27348882e-02 1.63874793e+00 -9.29914057e-01
2.71126628e-01 4.87922221e-01 4.57591146e-01 1.01869956e-01
8.25917304e-01 1.18162066e-01 6.64210394e-02 -9.32076760e-03
8.80976915e-01 -2.97395438e-01 -1.32036388e-01 -4.65737358e-02
-1.08195639e+00 9.66325641e-01 2.41891947e-02 1.06882811e+00
-3.92682225e-01 -4.65068072e-02 3.76238316e-01 6.41897559e-01
2.33304963e-01 6.37349665e-01 -2.32764855e-01 -4.64686006e-01
-7.20800757e-01 7.10530207e-02 8.15810263e-01 3.46420616e-01
3.23226511e-01 -1.63041383e-01 -9.81171150e-03 5.13691843e-01
2.67904222e-01 2.74968415e-01 3.46243888e-01 -1.10273623e+00
-1.60498526e-02 6.52132452e-01 4.45150793e-01 -1.08115041e+00
-5.05590022e-01 -2.30039611e-01 -2.78135538e-01 4.65213686e-01
6.75214112e-01 -1.99005574e-01 -2.83922702e-01 1.78774941e+00
-1.22100286e-01 -5.06051183e-01 -2.00822785e-01 1.10433793e+00
4.55524415e-01 1.94051638e-01 1.76152185e-01 -4.39725935e-01
1.72641814e+00 -5.94917871e-02 -8.85594428e-01 7.79427364e-02
8.01423669e-01 -1.08434045e+00 1.48570049e+00 4.36528772e-01
-1.18928206e+00 -4.66310859e-01 -1.10222089e+00 -1.16759706e-02
-3.25912952e-01 -1.83150277e-01 9.46814179e-01 9.86374795e-01
-9.43836987e-01 5.84774017e-01 -5.78985810e-01 -6.55207217e-01
-4.33038473e-01 1.20364577e-01 -2.27850690e-01 1.63726807e-01
-1.03046834e+00 1.25009024e+00 4.03153121e-01 2.37510681e-01
1.27124161e-01 -7.11193979e-02 -5.22616744e-01 2.99734138e-02
7.29876339e-01 -5.50061166e-01 1.16848075e+00 -1.29465580e+00
-1.30252516e+00 8.90270591e-01 2.81516742e-02 -1.64791733e-01
2.25118607e-01 4.54971083e-02 -7.78616786e-01 -4.16406654e-02
4.19717021e-02 1.38145208e-01 1.86360046e-01 -1.06968546e+00
-3.36556733e-01 -5.29270172e-01 5.53590238e-01 -7.02411234e-02
-1.74356416e-01 5.06848514e-01 2.47107208e-01 -5.56138277e-01
3.32751155e-01 -9.69315708e-01 1.22349128e-01 -3.07513416e-01
4.91539389e-02 -3.68654609e-01 7.84076899e-02 -3.23838562e-01
1.49857938e+00 -2.08561420e+00 -2.20467374e-01 2.11332500e-01
6.97872490e-02 1.66798383e-01 -2.35295724e-02 9.76909757e-01
-3.49888384e-01 4.86662358e-01 2.08354488e-01 3.52920651e-01
4.57831681e-01 1.53107688e-01 -3.11781794e-01 6.45013988e-01
4.30249721e-02 7.42653847e-01 -7.11139202e-01 -3.67535591e-01
2.83091366e-01 5.07067025e-01 -1.48828268e-01 -2.41866902e-01
1.41741112e-01 1.88530579e-01 -2.40251690e-01 1.84807032e-01
5.78122377e-01 -8.98302048e-02 3.25575501e-01 -1.31467089e-01
-5.20170987e-01 8.06059599e-01 -1.35309672e+00 1.14854121e+00
-2.80614588e-02 5.29075205e-01 -3.32852989e-01 -1.24752700e+00
9.44536626e-01 4.94539946e-01 2.92160124e-01 -1.01451647e+00
2.76052296e-01 3.17800015e-01 4.14766014e-01 -4.69905317e-01
6.70890927e-01 -6.08374119e-01 -1.06891595e-01 5.98319173e-01
-3.35481703e-01 -1.16932079e-01 2.31017366e-01 -1.56039819e-01
5.88496685e-01 -3.48426178e-02 4.29350436e-01 -5.15367866e-01
4.06169236e-01 -2.28384793e-01 3.68170470e-01 6.87701166e-01
-8.82279500e-02 3.26641142e-01 7.16949642e-01 -3.50954652e-01
-9.44260538e-01 -8.47551405e-01 -6.82308197e-01 8.95339251e-01
-2.42108315e-01 -4.52142656e-01 -5.46077192e-01 -3.41509998e-01
-4.02736276e-01 1.17705214e+00 -7.21626401e-01 6.27200678e-03
-1.29119813e-01 -5.00864446e-01 3.41666609e-01 2.96208948e-01
3.54810022e-02 -8.64223838e-01 -8.21848333e-01 -1.48127088e-02
-8.90127942e-02 -8.06483507e-01 1.58974349e-01 -3.79693182e-03
-7.99330115e-01 -9.85535085e-01 -2.50709742e-01 -2.83919841e-01
3.30447853e-01 4.73744154e-01 9.74906981e-01 3.12979132e-01
6.21084332e-01 5.25090694e-01 -7.16147125e-01 -8.62872243e-01
-6.10545337e-01 -4.30644691e-01 -1.71720237e-01 -3.75543892e-01
8.45871389e-01 -6.19337142e-01 -5.29516697e-01 3.46561402e-01
-1.26187468e+00 -4.58687842e-01 5.45417368e-01 1.96681470e-01
-2.62746867e-02 1.48294017e-01 6.57961428e-01 -9.06987906e-01
9.83479798e-01 -2.77314067e-01 -4.93546903e-01 2.16745526e-01
-8.93785238e-01 -1.44801429e-02 9.92793143e-02 -1.00435920e-01
-8.33559096e-01 -1.02651060e+00 -1.49046496e-01 3.76658916e-01
-2.23977104e-01 7.66919196e-01 1.81088194e-01 -2.26024818e-02
9.46548998e-01 -8.16211924e-02 1.70447573e-01 -2.71082491e-01
3.24225962e-01 7.79663742e-01 1.76330000e-01 -9.03501093e-01
6.91484332e-01 3.63242656e-01 8.65642205e-02 -8.27182114e-01
-9.01741385e-01 -3.78645331e-01 -3.58593047e-01 1.99204516e-02
7.49395967e-01 -6.63670063e-01 -9.16135669e-01 -2.49223411e-01
-8.18412960e-01 -9.37988013e-02 -5.52919924e-01 9.40051198e-01
-5.17051518e-01 6.12358212e-01 -4.62755471e-01 -1.12540448e+00
1.70962214e-01 -1.08538830e+00 6.01433337e-01 -4.61922362e-02
-7.98809707e-01 -1.20336759e+00 1.22290052e-01 5.18101752e-01
6.65185571e-01 1.63659900e-01 9.67450678e-01 -6.59334302e-01
-2.97019072e-02 -3.39677066e-01 -2.10973918e-01 4.58831906e-01
2.25545615e-01 1.47235274e-01 -7.07874596e-01 -1.43629953e-01
6.14759862e-01 -3.02752376e-01 3.00559729e-01 3.52020502e-01
6.92602277e-01 -4.59347755e-01 3.35771561e-01 -1.62508368e-01
1.61083984e+00 3.78782481e-01 9.77295339e-01 6.14362061e-01
-3.12685072e-02 1.08137000e+00 8.46278548e-01 3.55224192e-01
3.15034240e-01 1.01221955e+00 1.26682505e-01 -2.50724871e-02
8.74547586e-02 1.31009649e-02 4.01657999e-01 7.84012377e-01
-4.14763182e-01 1.58297479e-01 -5.90840399e-01 2.63545722e-01
-1.72222126e+00 -1.16327035e+00 -5.55908263e-01 2.81641722e+00
6.49872541e-01 3.76676440e-01 5.70892036e-01 3.23050112e-01
5.37477016e-01 1.10067137e-01 2.98009932e-01 -1.07071888e+00
7.97787085e-02 1.82315931e-01 1.12027079e-01 4.61594850e-01
-5.27051508e-01 4.10058856e-01 6.56079149e+00 4.13627028e-01
-1.20497668e+00 -1.02528304e-01 4.29909229e-01 2.74058878e-01
-7.04996526e-01 5.55998206e-01 -4.66045558e-01 3.52291048e-01
1.12461555e+00 -3.43589753e-01 2.44972751e-01 4.89225239e-01
5.98705947e-01 -2.04360783e-01 -1.10423636e+00 4.80292231e-01
-6.88204318e-02 -8.05082262e-01 -1.76669121e-01 3.78640145e-01
2.64761746e-01 -3.99807394e-01 -8.48641712e-03 3.07690263e-01
2.40827203e-01 -1.16453040e+00 9.06642199e-01 3.43034536e-01
4.07620192e-01 -4.65779632e-01 1.12187350e+00 2.82310784e-01
-4.80690151e-01 5.08328192e-02 -6.76986635e-01 -8.17533314e-01
-1.40760392e-01 5.14935613e-01 -5.13726234e-01 6.46519899e-01
1.97184414e-01 -5.43252639e-02 -5.35570979e-01 8.55216384e-01
-3.19133580e-01 8.10008764e-01 -9.28324759e-02 -6.78755641e-02
2.50994474e-01 -4.96472239e-01 3.36570859e-01 1.03785098e+00
3.06312412e-01 4.99025173e-02 -3.33724856e-01 7.74561346e-01
5.18593013e-01 4.49822694e-01 -6.58828795e-01 -2.60834783e-01
4.70894337e-01 1.07179868e+00 -8.78757358e-01 -2.81455219e-02
-8.84008527e-01 2.23451063e-01 4.44136932e-02 2.29472071e-01
-7.06031263e-01 3.60585772e-03 2.44930819e-01 3.13265234e-01
2.66748033e-02 -8.51918384e-02 -5.81684828e-01 -1.12018132e+00
1.58329561e-01 -1.07287061e+00 4.20133919e-01 -6.86010718e-01
-9.58515942e-01 1.85504913e-01 4.09380704e-01 -1.01409268e+00
-1.56928316e-01 -5.90930998e-01 -4.35704947e-01 8.42320979e-01
-1.11875272e+00 -7.64948606e-01 5.73232658e-02 1.62721500e-01
-6.13843463e-02 5.57141840e-01 9.76652265e-01 1.75875425e-02
-1.64184451e-01 4.11993682e-01 -1.12375952e-01 -6.58451580e-04
8.53720188e-01 -1.09026456e+00 -1.19561322e-01 7.94618368e-01
3.43962342e-01 9.90399897e-01 1.10854363e+00 -2.47714102e-01
-9.90842760e-01 -2.00540751e-01 1.19189835e+00 -5.63352644e-01
9.00873840e-01 2.55300831e-02 -9.09338593e-01 8.60195756e-01
3.60234052e-01 -5.60781240e-01 8.90697241e-01 7.62687385e-01
-3.28012168e-01 5.09201735e-02 -9.11263943e-01 5.79661787e-01
6.30107880e-01 -3.60976279e-01 -1.14066374e+00 2.11894363e-01
6.79080844e-01 1.39788255e-01 -1.31057394e+00 2.96996325e-01
8.41147959e-01 -1.38982594e+00 7.23503292e-01 -4.47173029e-01
2.59441793e-01 -1.60171390e-01 -3.25978428e-01 -9.82876420e-01
-5.39959192e-01 -1.68466523e-01 7.99306095e-01 1.49680424e+00
3.30512464e-01 -1.03909528e+00 3.72084558e-01 9.33719218e-01
-1.26189619e-01 -5.78195095e-01 -7.84690320e-01 -8.33072305e-01
5.81150949e-01 -8.91550839e-01 6.38338268e-01 1.22012520e+00
4.18282390e-01 5.01242936e-01 3.45517434e-02 -1.43037438e-01
9.24755409e-02 -1.07310705e-01 9.15011466e-01 -1.35660875e+00
-3.60175401e-01 -5.92880189e-01 -5.12742758e-01 -4.43311900e-01
-3.16811949e-01 -4.25060868e-01 -6.40121937e-01 -1.52990723e+00
1.86603934e-01 -3.34224522e-01 -4.40610945e-01 1.16667308e-01
1.27106920e-01 -5.96787445e-02 6.26634300e-01 2.18699172e-01
-2.96346188e-01 -9.54127312e-02 1.16404009e+00 3.53919297e-01
-1.12795137e-01 -4.66603376e-02 -1.56991756e+00 8.31320763e-01
7.07033813e-01 -4.66852784e-01 -5.64021111e-01 -7.04314858e-02
8.16159666e-01 2.62139708e-01 4.83795643e-01 -7.64175117e-01
-8.69413987e-02 -5.92526257e-01 9.32732671e-02 -1.39561564e-01
1.67882487e-01 -9.25298035e-01 3.46616924e-01 4.10491943e-01
-3.61341774e-01 1.01317883e-01 1.20708100e-01 -3.10006663e-02
-9.95149761e-02 -8.26718628e-01 4.54501241e-01 -2.70841986e-01
-4.35696036e-01 -4.65876579e-01 -2.31597900e-01 -7.12916180e-02
7.23293066e-01 -4.90158886e-01 -5.18802583e-01 -6.65618718e-01
-5.42554438e-01 -2.99245864e-01 9.63924050e-01 3.20609719e-01
8.41876268e-02 -1.16226399e+00 -7.46283293e-01 -3.07232797e-01
1.25261977e-01 -6.91621840e-01 7.03536570e-02 1.37230718e+00
-2.63781846e-01 8.97013009e-01 -1.26338780e-01 -1.10405661e-01
-1.08119333e+00 7.51218617e-01 4.00245562e-02 -1.08445503e-01
-5.29748082e-01 4.71038260e-02 2.57993966e-01 -4.83354032e-02
3.34976874e-02 -5.03938437e-01 -4.94634688e-01 1.32576436e-01
5.47953844e-01 3.97855490e-01 -4.85117584e-02 -7.62170672e-01
-1.87067926e-01 5.32838523e-01 2.37526670e-02 -3.92683715e-01
1.23428452e+00 -2.32635677e-01 -4.12045270e-01 1.02323890e+00
8.47844183e-01 5.36988139e-01 -3.60348344e-01 1.12441845e-01
2.13304818e-01 -5.89836180e-01 -1.47019103e-01 -9.61973548e-01
-1.52749434e-01 6.25737667e-01 5.22843838e-01 8.33335280e-01
1.01272202e+00 -8.01961720e-02 6.33662334e-03 1.78028777e-01
4.27760482e-01 -1.10078776e+00 -3.21630448e-01 -7.06790015e-02
1.08220971e+00 -1.00766218e+00 3.08008581e-01 -3.59342396e-01
-4.60714340e-01 1.34299898e+00 -6.60692006e-02 -1.94156112e-03
3.01010579e-01 -1.74612343e-01 1.36850104e-01 -4.68171775e-01
-6.78756237e-01 -4.18293864e-01 3.43343228e-01 4.63996291e-01
1.16092145e+00 3.76580209e-01 -1.58913326e+00 4.89609331e-01
-5.46336353e-01 3.94385189e-01 8.84627104e-01 6.49794817e-01
-6.34330988e-01 -1.41369736e+00 -7.71177769e-01 2.66052991e-01
-9.04023826e-01 2.26119667e-01 -5.60268462e-01 1.34217525e+00
8.69273767e-02 1.41252828e+00 -1.64926454e-01 -3.66224825e-01
4.44479793e-01 4.62963432e-02 6.14842951e-01 -5.77390909e-01
-4.89322364e-01 1.57648921e-01 5.10501742e-01 -1.99058577e-01
-6.99192286e-01 -7.01621294e-01 -7.20014274e-01 -6.93362117e-01
-4.60816085e-01 5.40469348e-01 7.78784215e-01 1.13679409e+00
-7.46864974e-02 3.23682547e-01 3.28546822e-01 -1.42812118e-01
-1.00488853e+00 -1.05981767e+00 -6.88189149e-01 4.88291085e-01
-1.38079792e-01 -5.43103397e-01 -5.70762038e-01 -5.48716664e-01] | [10.000914573669434, 8.437910079956055] |
68e31b29-1033-400e-a7ea-505730136b66 | condnet-conditional-classifier-for-scene | 2109.10322 | null | https://arxiv.org/abs/2109.10322v1 | https://arxiv.org/pdf/2109.10322v1.pdf | CondNet: Conditional Classifier for Scene Segmentation | The fully convolutional network (FCN) has achieved tremendous success in dense visual recognition tasks, such as scene segmentation. The last layer of FCN is typically a global classifier (1x1 convolution) to recognize each pixel to a semantic label. We empirically show that this global classifier, ignoring the intra-class distinction, may lead to sub-optimal results. In this work, we present a conditional classifier to replace the traditional global classifier, where the kernels of the classifier are generated dynamically conditioned on the input. The main advantages of the new classifier consist of: (i) it attends on the intra-class distinction, leading to stronger dense recognition capability; (ii) the conditional classifier is simple and flexible to be integrated into almost arbitrary FCN architectures to improve the prediction. Extensive experiments demonstrate that the proposed classifier performs favourably against the traditional classifier on the FCN architecture. The framework equipped with the conditional classifier (called CondNet) achieves new state-of-the-art performances on two datasets. The code and models are available at https://git.io/CondNet. | ['Nong Sang', 'Changxin Gao', 'Yuanjie Shao', 'Changqian Yu'] | 2021-09-21 | null | null | null | null | ['scene-segmentation'] | ['computer-vision'] | [ 4.15252864e-01 5.49314693e-02 -2.45809197e-01 -5.34443617e-01
-1.09982543e-01 -3.20588380e-01 6.76582754e-01 -1.80352315e-01
-5.42464018e-01 4.84864444e-01 -2.31895953e-01 -3.56027931e-01
-1.10648818e-01 -8.09095562e-01 -7.89628506e-01 -9.00988042e-01
9.60476771e-02 1.96775630e-01 7.02718616e-01 2.02907577e-01
4.96351011e-02 7.89978027e-01 -1.93550146e+00 6.45954609e-01
8.50102186e-01 1.56718004e+00 5.37186265e-01 5.56449234e-01
-1.99268728e-01 1.06355333e+00 -4.90299433e-01 -1.43328696e-01
2.22574040e-01 -2.54710346e-01 -9.90846634e-01 3.87555026e-02
4.70342636e-01 -2.51353774e-02 6.90343082e-02 1.06312156e+00
2.43624579e-02 9.45632998e-03 6.04897261e-01 -1.19618237e+00
-3.18012446e-01 4.68529463e-01 -2.88191825e-01 1.55220658e-01
-4.89089526e-02 2.79832315e-02 9.91569698e-01 -8.18046272e-01
4.96113449e-01 1.17898548e+00 6.62866712e-01 4.73928213e-01
-1.02478576e+00 -4.01746660e-01 3.89417440e-01 3.90556991e-01
-1.40511346e+00 -2.41883114e-01 5.34485161e-01 -5.43053269e-01
8.88505697e-01 2.61142075e-01 5.73556423e-01 9.74049985e-01
8.79907757e-02 9.94399130e-01 1.16620159e+00 -4.59132046e-01
3.65423977e-01 1.76419452e-01 3.20233345e-01 7.51535296e-01
1.15278130e-02 1.28099948e-01 -2.04412550e-01 1.41399711e-01
7.50978231e-01 7.37852380e-02 -2.57462561e-01 -5.24582803e-01
-8.08315396e-01 8.49931479e-01 8.95812511e-01 5.84832489e-01
-3.06184441e-01 1.70659974e-01 3.64799052e-01 -3.10052838e-02
3.36074173e-01 1.84721038e-01 -5.37679195e-01 8.15127864e-02
-8.60818744e-01 8.90958533e-02 7.19876230e-01 7.27815449e-01
9.33347106e-01 1.61605366e-02 -2.84143656e-01 7.94084072e-01
1.26705676e-01 1.07170567e-01 5.70141971e-01 -6.49677515e-01
1.50773153e-01 8.57742012e-01 -2.51778752e-01 -7.65965641e-01
-4.77236480e-01 -7.94325233e-01 -8.09278667e-01 6.02254570e-01
4.96034026e-01 -5.38470149e-02 -1.26983011e+00 1.37182927e+00
1.11869916e-01 3.81455064e-01 -9.28299055e-02 8.41547251e-01
9.12449181e-01 5.22412837e-01 3.36724192e-01 2.92294830e-01
1.06473136e+00 -1.30980718e+00 -4.96487051e-01 -2.00458109e-01
5.98883092e-01 -5.97342253e-01 8.84177446e-01 2.55175233e-01
-5.89574993e-01 -9.48692441e-01 -9.51808095e-01 1.10773779e-01
-7.80968010e-01 5.20713449e-01 7.59989738e-01 6.58955395e-01
-1.16948855e+00 5.25170565e-01 -9.01705921e-01 -5.17546177e-01
6.14993930e-01 5.66388607e-01 -2.13370681e-01 -9.32089519e-03
-8.70979130e-01 7.62475133e-01 7.98953235e-01 3.06142420e-01
-9.49173748e-01 -5.28448403e-01 -7.99985528e-01 1.91409469e-01
3.30478281e-01 -3.84667754e-01 1.16816032e+00 -1.65482211e+00
-1.57416105e+00 8.08093309e-01 -6.36134744e-02 -6.11025691e-01
6.11447334e-01 -1.77131861e-01 -1.63005903e-01 2.35759705e-01
-1.38587460e-01 8.75832021e-01 8.83072674e-01 -1.23424149e+00
-7.29041398e-01 -6.26203194e-02 1.51272595e-01 -1.02420129e-01
-1.22700721e-01 -6.02230765e-02 -4.86106634e-01 -5.85806012e-01
-7.05978880e-03 -8.05060685e-01 -3.20099533e-01 -2.35443432e-02
-4.90518510e-01 -3.28926682e-01 1.16468132e+00 -1.58402488e-01
8.62079680e-01 -2.24732184e+00 -5.34397289e-02 2.26720497e-01
1.31818131e-01 7.58636534e-01 -4.51514572e-02 -3.39458557e-03
-1.82790101e-01 -8.34549293e-02 -4.04134214e-01 -2.14982226e-01
-3.68382573e-01 2.60811567e-01 -1.53879840e-02 4.51456398e-01
4.02255416e-01 1.06121445e+00 -5.20371795e-01 -3.11162978e-01
5.58512926e-01 5.18380880e-01 -3.50058377e-01 9.58388969e-02
-2.26497084e-01 2.93595791e-01 -3.46383482e-01 5.56863785e-01
8.49148035e-01 -4.15578306e-01 2.28144720e-01 -2.27017701e-01
-2.06227526e-01 -1.60978995e-02 -1.23781765e+00 1.33055341e+00
-1.77996472e-01 5.87332904e-01 1.83915898e-01 -1.40984809e+00
1.03533125e+00 1.50622636e-01 1.88191503e-01 -4.95692104e-01
2.90667295e-01 1.20645516e-01 2.12663319e-02 -3.00309420e-01
9.75396559e-02 3.77927087e-02 1.12076320e-01 -1.82080213e-02
5.56325793e-01 3.72289896e-01 1.39504358e-01 -1.10039733e-01
8.35684061e-01 3.64836633e-01 2.75124252e-01 -7.23943651e-01
8.58921349e-01 -1.21238574e-01 5.25140584e-01 8.08097303e-01
-2.48338223e-01 5.02539515e-01 4.55335528e-01 -6.97019994e-01
-6.11566782e-01 -8.62929702e-01 -3.37882280e-01 9.49330151e-01
1.76058754e-01 -8.97372440e-02 -9.55456257e-01 -7.44708955e-01
-3.76689099e-02 2.97677904e-01 -7.49996722e-01 8.33842438e-03
-3.05735022e-01 -6.94742203e-01 4.85092163e-01 8.63071799e-01
8.81135762e-01 -1.31223977e+00 -8.06621969e-01 5.45314066e-02
1.84083402e-01 -1.31288862e+00 6.12472445e-02 5.77841163e-01
-9.71323550e-01 -1.16887319e+00 -4.48141158e-01 -1.00423789e+00
6.53690159e-01 1.54971927e-01 7.62695193e-01 2.55717456e-01
-3.44975084e-01 2.56652206e-01 -5.12932539e-01 -1.44935191e-01
-1.62098244e-01 3.88478816e-01 -3.07083338e-01 5.63528836e-01
3.97408962e-01 -3.71611387e-01 -4.57291663e-01 2.19430432e-01
-8.44983995e-01 1.49428844e-01 6.54824853e-01 6.99562371e-01
5.76451063e-01 3.95735838e-02 3.87180597e-01 -1.29112506e+00
1.34519592e-01 -2.74861783e-01 -6.11206949e-01 2.02057377e-01
-4.14933354e-01 -8.66212696e-02 7.21648753e-01 -3.21981102e-01
-1.24897730e+00 4.81044412e-01 -2.22362593e-01 -3.38899940e-01
-6.48645401e-01 3.08257431e-01 -1.75500199e-01 -3.12610298e-01
5.21148562e-01 2.65601166e-02 -1.32564768e-01 -8.35974157e-01
2.36202940e-01 4.70250994e-01 5.26162446e-01 -5.41999876e-01
5.22573531e-01 4.41244006e-01 4.44834903e-02 -9.12580848e-01
-7.57548809e-01 -5.72092772e-01 -1.08298624e+00 -3.16462487e-01
1.14476681e+00 -8.86520922e-01 -5.09688556e-01 8.30756903e-01
-9.45868790e-01 -6.31484628e-01 -2.81709760e-01 3.64133477e-01
-4.56132412e-01 3.13170552e-02 -6.35646462e-01 -7.01816022e-01
-9.24559534e-02 -1.16330969e+00 7.07376540e-01 6.26055658e-01
1.94514215e-01 -1.12016320e+00 -4.28528041e-01 2.05929741e-01
5.26050270e-01 2.70500928e-01 8.03018570e-01 -6.38733268e-01
-5.42282224e-01 -8.56735334e-02 -5.32518923e-01 7.77045846e-01
1.69259571e-02 2.76871890e-01 -1.36099207e+00 -1.69565991e-01
-2.26254731e-01 -3.51767361e-01 1.29303825e+00 4.72627252e-01
1.54401505e+00 1.05024762e-01 -3.91365886e-01 8.82139325e-01
1.73092067e+00 1.96473703e-01 8.15961778e-01 2.69836634e-01
8.24496269e-01 5.43679535e-01 3.48962635e-01 2.07033515e-01
6.00350201e-02 5.86257994e-01 6.54028714e-01 -4.16226774e-01
-3.35969329e-01 -1.08566065e-03 1.85232833e-01 4.36874717e-01
-2.54539222e-01 -1.16517507e-01 -9.04337406e-01 4.36534673e-01
-1.94201505e+00 -7.49969542e-01 -1.98472589e-01 1.86272573e+00
3.76791835e-01 2.88102180e-01 -3.32514234e-02 9.78431106e-02
8.48600805e-01 1.64082929e-01 -3.81243020e-01 -6.18201315e-01
-1.79897323e-01 5.87478697e-01 4.84169632e-01 4.11809891e-01
-1.45530999e+00 1.16296923e+00 6.13059092e+00 9.65205431e-01
-1.40869474e+00 5.47661036e-02 8.04230750e-01 2.67777145e-01
3.31823677e-01 -1.15886942e-01 -9.63517547e-01 2.99970984e-01
7.99750984e-01 5.05683303e-01 2.24161193e-01 1.09061158e+00
-1.83857739e-01 -1.71159610e-01 -9.09159541e-01 6.81305766e-01
-1.05905302e-01 -1.38468421e+00 1.79490477e-01 6.77114539e-03
5.99346340e-01 2.34163791e-01 -1.51045695e-01 3.76249224e-01
5.06156459e-02 -1.13174629e+00 8.26589525e-01 4.43439573e-01
7.80386448e-01 -8.42383683e-01 9.34760392e-01 3.25155765e-01
-1.37650073e+00 -1.88480511e-01 -5.18057227e-01 -2.42835224e-01
-3.34846199e-01 5.43804705e-01 -4.73487139e-01 6.10072851e-01
9.57850873e-01 7.23776639e-01 -8.48962605e-01 1.21854436e+00
-2.72864491e-01 6.25672221e-01 -2.31864765e-01 4.22947295e-02
5.79905987e-01 -9.77565795e-02 1.86666846e-01 1.56332326e+00
-4.49646264e-02 -1.82405204e-01 2.89158821e-01 8.32934022e-01
1.29032293e-02 9.86407995e-02 -3.41839790e-01 2.36721471e-01
9.12719518e-02 1.36607873e+00 -1.03719664e+00 -3.84957790e-01
-5.13904870e-01 9.42438483e-01 6.12558901e-01 4.51543808e-01
-6.42353833e-01 -4.40271080e-01 6.16068184e-01 -1.50602143e-02
8.85527790e-01 1.35875627e-01 -3.79519969e-01 -8.35433185e-01
-9.51451510e-02 -4.80406880e-01 3.71470213e-01 -5.41441739e-01
-1.07276583e+00 8.60359251e-01 -6.66599125e-02 -9.79529142e-01
2.59078164e-02 -1.28211069e+00 -7.75144994e-01 7.24093437e-01
-1.60727274e+00 -1.20327795e+00 -5.12427151e-01 7.87811100e-01
3.58499646e-01 -7.85446167e-02 8.64514053e-01 3.22086275e-01
-6.34876609e-01 6.00657940e-01 -1.14274867e-01 3.55573088e-01
2.19367400e-01 -1.24736714e+00 -6.49972484e-02 9.25444484e-01
-3.82825397e-02 3.46318543e-01 2.30017453e-01 -3.86364996e-01
-8.78858209e-01 -1.26760399e+00 7.14114964e-01 -1.39265627e-01
4.60995048e-01 -4.22687382e-01 -8.21162999e-01 6.45713210e-01
1.77810594e-01 4.57275659e-01 5.50402164e-01 1.34054542e-01
-4.14865524e-01 -3.37304085e-01 -1.17001653e+00 1.61989018e-01
8.77357304e-01 -4.00927067e-01 -2.96901494e-01 1.05982922e-01
6.08747244e-01 -2.82670051e-01 -6.26942754e-01 5.53799748e-01
5.18791139e-01 -1.14633250e+00 7.02132583e-01 -3.47572565e-01
3.33575696e-01 -4.31864500e-01 -1.44607395e-01 -9.00674462e-01
-5.39524078e-01 -1.26741484e-01 -1.05506055e-01 1.03784227e+00
5.26320040e-01 -7.53531992e-01 8.38636041e-01 2.86031306e-01
-2.44575635e-01 -9.83828723e-01 -1.05000556e+00 -8.37891281e-01
-3.33788246e-02 -4.78704274e-01 4.62748945e-01 7.52545953e-01
-5.51304519e-01 1.48796782e-01 -9.93428081e-02 1.44067153e-01
3.68630320e-01 2.38245100e-01 4.18703526e-01 -1.48399067e+00
-3.89369220e-01 -5.51356196e-01 -7.25819349e-01 -1.18848753e+00
9.33015421e-02 -9.57559109e-01 8.00238550e-02 -1.39984334e+00
1.48769855e-01 -6.89779580e-01 -7.08236337e-01 8.44431937e-01
-8.97356048e-02 4.59606558e-01 3.26588869e-01 -1.26784137e-02
-6.11362398e-01 3.04086864e-01 1.22884190e+00 -1.47533463e-02
-1.34569004e-01 1.86622143e-01 -4.62099910e-01 7.24724472e-01
9.76264894e-01 -2.78622538e-01 -3.50951135e-01 -2.92385370e-01
-3.25826406e-01 -5.55184007e-01 5.10068774e-01 -1.35795438e+00
1.84842870e-01 1.15208492e-01 5.79762757e-01 -4.69988704e-01
2.27537051e-01 -9.99352157e-01 6.17916882e-02 6.11615360e-01
-9.33392271e-02 -3.73227119e-01 2.46005520e-01 3.94621015e-01
-4.55495894e-01 -4.15000409e-01 1.05529010e+00 -1.47932559e-01
-1.11220121e+00 2.45516151e-01 -4.19126093e-01 -3.21013033e-01
1.17257404e+00 -3.40650827e-01 -2.90564626e-01 3.88847627e-02
-7.58603752e-01 -3.41557078e-02 2.48983249e-01 4.21010315e-01
4.47294325e-01 -1.13350523e+00 -4.01383162e-01 3.96083206e-01
8.79597887e-02 6.48563206e-02 8.49337354e-02 7.52134383e-01
-7.22320437e-01 6.18839681e-01 -2.92892665e-01 -8.90887082e-01
-1.08342266e+00 5.93322337e-01 6.03826165e-01 -1.75550207e-01
-6.94852829e-01 1.07025445e+00 4.13112193e-01 -5.38003027e-01
4.61025983e-01 -5.09767413e-01 -4.04805958e-01 5.69914728e-02
4.93659437e-01 1.61221564e-01 2.03530326e-01 -6.53739393e-01
-4.85642165e-01 5.66858053e-01 -8.11620951e-02 3.54164720e-01
1.35300732e+00 2.81263202e-01 -2.30527773e-01 3.49724472e-01
1.22761619e+00 -4.06620622e-01 -1.50156319e+00 -1.71730310e-01
5.39530516e-02 -2.53270656e-01 2.13461414e-01 -8.55655253e-01
-1.43643546e+00 8.14684510e-01 7.45103419e-01 2.54446745e-01
1.26453710e+00 4.24979590e-02 3.65582943e-01 1.97793007e-01
2.92393088e-01 -8.82791579e-01 -2.86196798e-01 5.92014849e-01
5.74993730e-01 -1.01640010e+00 -1.46935299e-01 -5.53554296e-01
-5.91470301e-01 1.37486684e+00 7.86582768e-01 -4.29469287e-01
9.44907486e-01 2.95743138e-01 2.46787563e-01 -1.72596425e-01
-6.37869954e-01 -4.68848884e-01 4.13186103e-01 6.57124400e-01
3.14743280e-01 3.36797595e-01 -7.72758946e-02 5.03590524e-01
1.14165857e-01 -6.48597777e-02 1.64898157e-01 8.42443407e-01
-4.92657930e-01 -1.08766901e+00 -6.87562302e-02 4.56785411e-01
-3.24547559e-01 5.06827794e-02 -2.92193383e-01 8.65411520e-01
6.59007967e-01 8.84571970e-01 1.86188117e-01 -4.35681432e-01
1.58699065e-01 1.48009121e-01 3.88468444e-01 -5.91199160e-01
-7.65025437e-01 -6.90711439e-02 -5.06994911e-02 -9.00507092e-01
-6.82811379e-01 -5.37317812e-01 -1.23706138e+00 -1.13664337e-01
-2.63924360e-01 -2.38763820e-02 7.92073846e-01 9.84908581e-01
2.31179133e-01 6.98636174e-01 3.58518124e-01 -1.01167214e+00
-9.48477611e-02 -9.49400902e-01 -5.88148475e-01 2.60845900e-01
4.12869096e-01 -7.05959558e-01 -4.06021863e-01 1.06052808e-01] | [9.454669952392578, 1.9335832595825195] |
5d6e6af9-ce5e-4a18-9c4f-d2cf83593a07 | real-time-indoor-scene-reconstruction-with | 1812.03015 | null | http://arxiv.org/abs/1812.03015v1 | http://arxiv.org/pdf/1812.03015v1.pdf | Real-time Indoor Scene Reconstruction with RGBD and Inertia Input | Camera motion estimation is a key technique for 3D scene reconstruction and
Simultaneous localization and mapping (SLAM). To make it be feasibly achieved,
previous works usually assume slow camera motions, which limits its usage in
many real cases. We propose an end-to-end 3D reconstruction system which
combines color, depth and inertial measurements to achieve robust
reconstruction with fast sensor motions. Our framework extends Kalman filter to
fuse the three kinds of information and involve an iterative method to jointly
optimize feature correspondences, camera poses and scene geometry. We also
propose a novel geometry-aware patch deformation technique to adapt the feature
appearance in image domain, leading to a more accurate feature matching under
fast camera motions. Experiments show that our patch deformation method
improves the accuracy of feature tracking, and our 3D reconstruction
outperforms the state-of-the-art solutions under fast camera motions. | ['Zunjie Zhu', 'Feng Xu'] | 2018-12-07 | null | null | null | null | ['3d-scene-reconstruction', 'indoor-scene-reconstruction'] | ['computer-vision', 'computer-vision'] | [-1.05058163e-01 -6.89283371e-01 -6.04201853e-02 -3.18881661e-01
-6.08451426e-01 -6.77715957e-01 5.63346028e-01 -2.85012782e-01
-4.95469272e-01 2.73828626e-01 5.50570106e-03 -8.82506296e-02
2.59020198e-02 -7.81974673e-01 -7.94739723e-01 -4.75973308e-01
4.49010342e-01 3.90502661e-01 4.90079939e-01 7.81331025e-03
4.46444035e-01 9.09309685e-01 -1.41274774e+00 -6.71833038e-01
7.80912161e-01 8.75944912e-01 6.66047990e-01 5.58175921e-01
-2.21668348e-01 2.51190096e-01 -3.49714868e-02 1.13878526e-01
6.58827662e-01 -1.05998814e-01 -2.96738535e-01 6.15460277e-01
6.32850766e-01 -4.20926660e-01 -4.77063298e-01 1.21128571e+00
6.18870914e-01 1.55190885e-01 1.35757640e-01 -8.35971713e-01
-1.19090438e-01 -2.96149909e-01 -7.49372244e-01 -3.58598888e-01
9.28145111e-01 7.34530985e-02 3.68209630e-01 -1.04638278e+00
6.47732437e-01 1.04932606e+00 9.37933445e-01 2.29018286e-01
-1.00727558e+00 -4.81516451e-01 1.68593645e-01 1.00710630e-01
-1.67893958e+00 -5.21642208e-01 9.13755894e-01 -3.07552457e-01
7.88354576e-01 4.91750650e-02 8.55885684e-01 7.27909386e-01
2.10809305e-01 3.80368441e-01 9.03835952e-01 -4.20078605e-01
6.53483868e-02 -1.66371003e-01 -3.95588130e-01 7.59625912e-01
3.77632260e-01 1.53628550e-02 -5.76659381e-01 -3.63327712e-02
1.38708186e+00 5.29403627e-01 -4.13585812e-01 -1.04540694e+00
-1.73801792e+00 4.81157243e-01 4.96351480e-01 -1.15777582e-01
-5.23228407e-01 1.86421141e-01 -1.24144681e-01 1.65288135e-01
2.79564261e-01 5.93538769e-02 -2.66106278e-01 -3.04565817e-01
-6.07583344e-01 1.42009467e-01 5.06594658e-01 1.31193388e+00
1.16009367e+00 -1.14592075e-01 4.29369360e-01 4.42930847e-01
5.27356446e-01 1.06489480e+00 3.25897485e-01 -1.07922339e+00
3.33916813e-01 6.65072918e-01 5.42607605e-01 -1.21412206e+00
-5.36715508e-01 -2.63047606e-01 -6.58665597e-01 -4.84474488e-02
1.21305987e-01 1.34360671e-01 -7.01225042e-01 1.23511314e+00
7.53443003e-01 5.21828711e-01 -3.84699777e-02 1.12894154e+00
4.98503357e-01 4.48566407e-01 -7.13490307e-01 -2.87948579e-01
9.68526304e-01 -8.06893408e-01 -7.58081496e-01 -4.59647328e-01
4.53257471e-01 -1.11201704e+00 7.21323371e-01 2.18288064e-01
-6.26338959e-01 -5.76592445e-01 -9.89871144e-01 7.25096138e-03
1.24811009e-01 -8.46287757e-02 5.29590249e-01 3.98024946e-01
-8.04740131e-01 1.31926909e-01 -1.08542526e+00 -6.07766449e-01
-1.25506610e-01 4.44484323e-01 -6.83368921e-01 -3.14002007e-01
-5.53188682e-01 9.97797608e-01 2.77804106e-01 2.42516175e-01
-4.04714167e-01 -3.04029733e-01 -1.01836479e+00 -4.98940051e-01
4.66907293e-01 -1.08578622e+00 1.06181431e+00 -3.36236507e-01
-2.03362894e+00 7.12279320e-01 -6.06982946e-01 7.80657539e-03
5.46833992e-01 -5.22105753e-01 4.15070355e-03 -2.33507454e-02
1.25060771e-02 2.72176296e-01 6.85166478e-01 -1.23109221e+00
-5.71120143e-01 -5.13149083e-01 -2.18381211e-01 6.01683617e-01
1.36772040e-02 -4.89068061e-01 -9.44051027e-01 -7.41342902e-02
1.09919727e+00 -1.12330508e+00 -4.78341222e-01 1.61107853e-01
-2.84910612e-02 3.06280434e-01 8.62525165e-01 -2.33805090e-01
8.08035433e-01 -1.95440316e+00 3.18243146e-01 2.45472610e-01
-8.48106071e-02 -1.33537918e-01 6.31250441e-02 3.49088937e-01
4.96785820e-01 -6.01845562e-01 7.96029121e-02 -6.65064454e-01
-7.81362727e-02 3.77236515e-01 -9.08382162e-02 1.02990425e+00
-3.60529274e-01 6.81564569e-01 -9.01705027e-01 -2.61556178e-01
7.76307583e-01 7.19457746e-01 -4.39209342e-01 3.15691948e-01
9.68323201e-02 8.40237200e-01 -6.31262362e-01 7.19375014e-01
1.06909740e+00 -1.41073379e-03 -2.16902807e-01 -2.59774715e-01
-4.75275606e-01 3.90980989e-02 -1.63170290e+00 2.52453041e+00
-5.59011400e-01 2.52203822e-01 2.57714033e-01 -3.23756307e-01
1.22199857e+00 8.62231827e-04 5.28401971e-01 -4.82234180e-01
1.95239007e-01 3.69798839e-01 -6.15871727e-01 -2.70911634e-01
7.26209581e-01 7.71729797e-02 -5.20721674e-02 9.91082713e-02
-2.67311662e-01 -6.61870897e-01 -5.38611591e-01 -1.43744409e-01
9.37038064e-01 5.57320774e-01 4.76822615e-01 1.84422918e-02
7.01721370e-01 3.74806556e-03 8.52890432e-01 5.47948241e-01
1.49015128e-03 6.62429571e-01 -3.80881786e-01 -5.45904636e-01
-9.65514719e-01 -1.01346648e+00 -4.39279824e-02 1.40238971e-01
9.86736476e-01 -3.09001952e-01 -3.67851555e-01 -4.68673348e-01
2.99022198e-01 1.14551619e-01 -3.74697112e-02 1.00541553e-02
-5.67619205e-01 -3.80240947e-01 -1.85596004e-01 2.79436141e-01
7.50876009e-01 -1.82320431e-01 -8.46666098e-01 2.87816286e-01
-1.76502869e-01 -1.37596548e+00 -5.68595946e-01 -2.67429382e-01
-1.05814707e+00 -9.52851832e-01 -4.41999316e-01 -6.08653605e-01
8.83518815e-01 1.00028229e+00 4.81637269e-01 -5.02694538e-03
7.96400309e-02 6.30041540e-01 -5.13498545e-01 2.62803840e-03
-5.77277914e-02 -1.70417204e-01 5.24370551e-01 1.53944418e-01
2.24112034e-01 -6.90802157e-01 -6.57467186e-01 7.22310483e-01
-6.13125622e-01 2.88305789e-01 5.71803689e-01 6.30137265e-01
1.02317166e+00 -1.99955627e-01 -3.29178333e-01 -4.77726221e-01
-1.07440829e-01 1.11562036e-01 -1.21223533e+00 -8.78734980e-03
-5.29380023e-01 -8.30833465e-02 2.84030139e-01 -3.87500376e-01
-9.17008936e-01 9.17883754e-01 -5.65689020e-02 -6.81838393e-01
-1.43943191e-01 3.39978158e-01 -3.96288753e-01 -8.42777669e-01
3.67777139e-01 4.81075525e-01 1.22398555e-01 -5.82200587e-01
3.68233651e-01 6.06008112e-01 7.82644808e-01 -3.64482075e-01
1.27421594e+00 7.57299006e-01 3.72320682e-01 -7.78085172e-01
-5.69985509e-01 -7.97060728e-01 -1.24592948e+00 -2.59943008e-01
5.56414366e-01 -1.32768822e+00 -6.71581805e-01 6.74090147e-01
-1.24617624e+00 -3.45574617e-02 1.50695935e-01 1.02487445e+00
-6.89164877e-01 6.98906779e-01 -2.19261125e-01 -6.96230054e-01
-1.13768958e-01 -1.35607386e+00 1.24723506e+00 2.76300907e-01
1.99878454e-01 -8.14486980e-01 2.63928175e-01 -8.94876197e-02
1.48495823e-01 4.50096339e-01 -7.92160444e-03 3.02287281e-01
-1.14631951e+00 -4.63576138e-01 -9.51428264e-02 -1.96332484e-01
2.51584113e-01 -1.51144728e-01 -6.53692782e-01 -5.61713457e-01
7.44913667e-02 2.70125747e-01 4.93106425e-01 2.09961206e-01
4.80953634e-01 1.15298405e-01 -4.70228821e-01 1.27064049e+00
1.67779863e+00 -5.17593957e-02 5.92617154e-01 7.35722423e-01
9.21541691e-01 1.63017273e-01 9.53234971e-01 6.29623175e-01
7.29111493e-01 1.10449958e+00 6.83186650e-01 1.19628534e-01
1.38768747e-01 -4.35496122e-01 3.75836372e-01 1.17098165e+00
-9.38206241e-02 1.91562086e-01 -7.98260629e-01 3.80449593e-01
-2.07588601e+00 -5.15330434e-01 -2.51003742e-01 2.57831097e+00
3.69504035e-01 -1.69331715e-01 -2.46502891e-01 -2.64592648e-01
6.36316478e-01 1.17209904e-01 -5.40510237e-01 2.63330877e-01
-4.33932133e-02 -3.20301443e-01 9.88481462e-01 7.29082823e-01
-9.40052688e-01 1.09743750e+00 5.67664862e+00 2.99425423e-01
-1.15907657e+00 5.63212596e-02 -5.21866441e-01 1.05567515e-01
-3.54703128e-01 3.97932470e-01 -1.00284922e+00 1.95833966e-01
4.18734640e-01 -4.30600643e-02 3.58294517e-01 8.34676504e-01
1.39478922e-01 -3.24368834e-01 -7.95478046e-01 1.53810096e+00
2.11532548e-01 -1.24827743e+00 -2.15379596e-01 2.55519331e-01
7.75686264e-01 2.49909222e-01 -2.93148369e-01 -3.61758113e-01
7.78107420e-02 -3.21791530e-01 7.74721980e-01 5.97499669e-01
5.58175862e-01 -7.12919354e-01 6.50078416e-01 6.81614280e-01
-1.33487523e+00 1.55338079e-01 -6.19704247e-01 -2.90361315e-01
5.77620924e-01 7.17513680e-01 -7.32305467e-01 8.37131917e-01
5.74234068e-01 9.79101479e-01 -4.28722888e-01 1.37089860e+00
-2.29841813e-01 -1.61014289e-01 -6.55302882e-01 5.63796535e-02
3.76727693e-02 -4.02996957e-01 8.17628980e-01 6.83735490e-01
7.75561750e-01 1.66535988e-01 6.32961214e-01 4.52951998e-01
2.72535354e-01 -4.22302308e-03 -7.80814528e-01 5.98184049e-01
7.69301057e-01 1.14165759e+00 -7.08316088e-01 -3.46450806e-02
-5.00978947e-01 1.37495911e+00 5.08901626e-02 7.48473778e-02
-5.58921754e-01 -3.11950237e-01 7.11715460e-01 8.88225511e-02
2.88685650e-01 -1.02054763e+00 -2.18953900e-02 -1.57658410e+00
1.84316978e-01 -2.98837990e-01 -1.09532960e-01 -6.99054658e-01
-7.67165720e-01 4.28264290e-01 -1.88274890e-01 -1.79698896e+00
-2.89120555e-01 -3.42492044e-01 -1.54600099e-01 7.65589356e-01
-1.73558950e+00 -1.30476189e+00 -7.54842758e-01 8.32682550e-01
3.70543063e-01 1.65944397e-01 7.61535227e-01 1.75432548e-01
-3.39741051e-01 1.45193458e-01 1.94695354e-01 -1.15302883e-01
8.76186490e-01 -8.75220299e-01 5.33805847e-01 1.25916910e+00
2.42796749e-01 6.40271723e-01 5.96423626e-01 -7.25578785e-01
-2.35164499e+00 -9.22748744e-01 6.05733633e-01 -5.92348933e-01
2.95561224e-01 -4.76732194e-01 -5.42914987e-01 7.13049650e-01
-3.31981808e-01 2.10776746e-01 1.21165425e-01 -1.99353606e-01
-1.99327946e-01 -3.50423962e-01 -1.05554819e+00 3.02790284e-01
1.31534433e+00 -5.21972418e-01 -3.94971013e-01 1.92055300e-01
6.95941448e-01 -1.18903220e+00 -8.54173243e-01 4.87544477e-01
6.44080341e-01 -9.31990087e-01 1.04302180e+00 4.24503654e-01
-5.16607046e-01 -1.02018940e+00 -4.27937061e-01 -1.14886475e+00
-4.16133344e-01 -8.69392872e-01 -1.40539736e-01 1.02095151e+00
-2.04862863e-01 -8.05922329e-01 8.39177668e-01 3.53270382e-01
-1.62183613e-01 -7.99080059e-02 -1.15251946e+00 -8.07978094e-01
-7.11598337e-01 -4.35667098e-01 7.36974180e-01 8.36761296e-01
-3.99198055e-01 2.48261541e-02 -5.00070095e-01 7.48456597e-01
8.32287014e-01 5.14520705e-01 1.51020598e+00 -1.23126554e+00
-1.89851429e-02 -1.36883378e-01 -9.31061506e-01 -1.72572410e+00
1.10535309e-01 -4.84066427e-01 2.50287443e-01 -1.51710463e+00
-4.48895171e-02 -3.94464970e-01 1.99941531e-01 6.98222816e-02
1.83915887e-02 3.58643621e-01 1.64389238e-01 3.83592188e-01
-5.93029559e-01 5.78230381e-01 1.16488838e+00 3.47538978e-01
-4.24557805e-01 5.52636087e-02 -1.59453005e-01 7.71786749e-01
2.40868807e-01 -2.54255295e-01 -1.00717895e-01 -9.73599494e-01
1.61812663e-01 1.85897127e-01 4.10507828e-01 -1.15226042e+00
6.17759407e-01 -3.20303231e-01 5.58688223e-01 -9.23658609e-01
5.10236204e-01 -1.21881676e+00 5.53529441e-01 4.57326710e-01
4.78184700e-01 3.32588196e-01 7.08105564e-02 7.11422443e-01
-1.54826507e-01 1.13412805e-01 6.07102156e-01 -1.51396737e-01
-1.12155008e+00 6.34360790e-01 3.83942276e-02 -6.57106698e-01
1.03745377e+00 -5.87295592e-01 1.43788829e-01 -5.36415339e-01
-2.74062663e-01 2.37598702e-01 1.33334875e+00 4.23423201e-01
9.18728590e-01 -1.63183963e+00 -4.56046402e-01 5.83744884e-01
3.36097002e-01 4.96187687e-01 2.81319946e-01 1.06749761e+00
-9.86633837e-01 1.41766727e-01 -1.52024245e-02 -1.23566115e+00
-1.16801393e+00 3.65604013e-01 2.33415127e-01 3.45671475e-01
-8.25248241e-01 4.96435672e-01 -5.70979081e-02 -7.34439015e-01
1.34643495e-01 -1.77607119e-01 3.43907952e-01 -4.94129151e-01
5.42212427e-01 2.52976567e-01 -3.30875851e-02 -9.76943910e-01
-5.79972804e-01 1.61022651e+00 2.58825213e-01 -9.03865919e-02
1.32204270e+00 -8.57621729e-01 -1.01468012e-01 2.82416701e-01
1.10019660e+00 2.57298470e-01 -1.59441471e+00 -3.89439315e-01
-2.09358022e-01 -1.18198454e+00 2.47027665e-01 -1.65220499e-01
-8.26818585e-01 7.51990855e-01 6.85364842e-01 -4.35622275e-01
1.09250569e+00 -2.01872781e-01 9.25717413e-01 6.49240136e-01
1.04146457e+00 -7.42170632e-01 -2.78996497e-01 7.97187567e-01
5.15415370e-01 -1.25857460e+00 4.26127076e-01 -4.77226615e-01
-1.59341156e-01 1.22770739e+00 4.27661479e-01 -1.79771319e-01
3.01855981e-01 1.32564053e-01 2.44822949e-01 1.39744163e-01
-1.12705812e-01 -2.43415967e-01 1.68933883e-01 6.41697824e-01
-1.10099822e-01 -2.12823704e-01 4.39607352e-02 -1.24270730e-01
7.08988607e-02 -1.91826180e-01 4.23405379e-01 1.01093519e+00
-5.73965609e-01 -1.31466043e+00 -6.77986562e-01 -2.43413538e-01
1.13854006e-01 2.82995939e-01 -9.80171263e-02 7.88662612e-01
-1.22448772e-01 5.84691584e-01 -3.46974819e-03 -5.75836182e-01
6.02898300e-01 -2.32753202e-01 9.32213068e-01 -4.22796935e-01
-2.51976568e-02 5.82828522e-01 -3.12762350e-01 -1.05718780e+00
-6.77588761e-01 -8.03807437e-01 -1.29797149e+00 -3.26520085e-01
-6.58795416e-01 -1.42550766e-01 1.06461513e+00 7.34545410e-01
6.01872444e-01 -1.34355426e-01 9.35521185e-01 -1.26690483e+00
-3.03578287e-01 -5.76075375e-01 -5.35629332e-01 2.91584551e-01
5.95650017e-01 -8.78759861e-01 -2.63110816e-01 -1.25413731e-01] | [7.505223274230957, -2.2569618225097656] |
2b050a85-498a-4eb0-b759-f4edde5c0f3d | atomistic-calculations-of-charged-point | 2102.01016 | null | https://arxiv.org/abs/2102.01016v2 | https://arxiv.org/pdf/2102.01016v2.pdf | Atomistic calculations of charged point defects at grain boundaries in SrTiO$_3$ | Oxygen vacancies have been identified to play an important role in accelerating grain growth in polycrystalline perovskite-oxide ceramics. In order to advance the fundamental understanding of growth mechanisms at the atomic scale, classical atomistic simulations were carried out to investigate the atomistic structures and oxygen vacancy formation energies at grain boundaries in the prototypical perovskite-oxide material SrTiO$_3$. In this work, we focus on two symmetric tilt grain boundaries, namely $\Sigma$5(310)[001] and $\Sigma$5(210)[001]. A one-dimensional continuum model is adapted to determine the electrostatic potential induced by charged lattice planes in atomistic structure models containing grain boundaries and point defects. By means of this model, electrostatic artifacts, which are inherent to supercell models with periodic or open boundary conditions, can be taken into account and corrected properly. We report calculated formation energies of oxygen vacancies on all the oxygen sites across boundaries between two misoriented grains, and we analyze and discuss the formation-energy values with respect to local charge densities at the vacant sites. | ['Christian Elsässer', 'Daniel F. Urban', 'Daniel Mutter', 'Cong Tao'] | 2021-02-01 | null | null | null | null | ['formation-energy'] | ['miscellaneous'] | [ 1.47311032e-01 3.64453569e-02 3.80493992e-04 -2.80375510e-01
-2.68253922e-01 2.93640554e-01 1.17453136e-01 3.71756822e-01
-4.24120128e-01 1.18001914e+00 -1.69666305e-01 1.49363652e-01
2.82601751e-02 -1.03905427e+00 -6.68395340e-01 -1.48439324e+00
-1.60529047e-01 9.74751353e-01 4.24582273e-01 -3.26883256e-01
7.68666863e-01 4.27616507e-01 -2.24216104e+00 1.97738148e-02
1.29527938e+00 8.61161709e-01 3.39566976e-01 1.97359324e-01
-3.01297978e-02 -2.46095896e-01 -3.96406412e-01 1.98595285e-01
-1.79711103e-01 -2.26208821e-01 -3.44032466e-01 -3.41879465e-02
1.09569751e-01 2.42359579e-01 3.96493852e-01 1.19442356e+00
4.36904848e-01 -1.05041988e-01 1.38340473e+00 -6.84687793e-01
-6.90866053e-01 4.99406844e-01 -5.30579627e-01 6.27975315e-02
-6.25985414e-02 3.64184171e-01 1.02151000e+00 -1.14249635e+00
6.79330289e-01 7.84035444e-01 2.33318254e-01 7.01741397e-01
-9.70501006e-01 -5.42237878e-01 -1.89468011e-01 3.53663057e-01
-1.73476899e+00 -5.04610538e-01 8.09647739e-01 -4.48792845e-01
1.43547809e+00 4.50661123e-01 7.69549549e-01 2.05054611e-01
1.00910914e+00 -2.29852140e-01 1.26475620e+00 -8.68756175e-01
6.04960024e-01 3.46051492e-02 6.40437782e-01 6.29205480e-02
1.13431096e+00 -8.21087956e-02 -4.27077502e-01 -2.18196988e-01
3.13540578e-01 -4.27936673e-01 -4.10921276e-01 -1.84485421e-01
-3.96200031e-01 5.61510086e-01 2.85499334e-01 5.89745343e-01
-8.02479684e-01 2.23828688e-01 4.86618839e-02 -4.12425905e-01
4.42473143e-01 3.51135820e-01 1.75969955e-02 -1.42038129e-02
-4.21503514e-01 8.58349502e-01 6.17555857e-01 6.47238910e-01
6.35302961e-01 2.45400906e-01 1.75494775e-01 4.81680602e-01
4.86842364e-01 8.13958108e-01 1.54331982e-01 -4.15764928e-01
-8.33006427e-02 3.06520343e-01 5.56151330e-01 -2.28506073e-01
1.00463428e-01 4.53011356e-02 -5.11756659e-01 4.92161274e-01
1.52112618e-01 3.18445474e-01 -8.89363408e-01 9.12815332e-01
4.71155703e-01 -6.39367282e-01 1.64980084e-01 8.30444336e-01
8.48132551e-01 5.99767387e-01 -3.27460766e-02 -7.84447372e-01
1.53870189e+00 -5.20457447e-01 -9.29594517e-01 1.62474915e-01
5.43276906e-01 -2.73405850e-01 1.10743833e+00 2.49837369e-01
-1.64718664e+00 -8.93687606e-02 -1.25546014e+00 1.54807076e-01
-2.83492446e-01 -5.71140110e-01 1.49921581e-01 6.80148482e-01
-7.04297185e-01 7.87727594e-01 -1.01373196e+00 -5.61425742e-03
6.10642582e-02 4.30660158e-01 4.21709828e-02 -6.85616806e-02
-1.15858626e+00 4.95987087e-01 1.17911197e-01 1.95682019e-01
-5.96291840e-01 -6.00316465e-01 -3.61452222e-01 -8.61323327e-02
8.62931758e-02 -3.16420794e-01 7.33231962e-01 -1.80995151e-01
-1.21221495e+00 8.57414901e-01 -8.23092818e-01 -3.03881407e-01
3.25839221e-02 2.15460658e-01 -4.22010332e-01 7.78053552e-02
3.76092106e-01 2.01290175e-01 3.10522199e-01 -1.45154965e+00
-3.54601480e-02 -4.89500403e-01 -6.49616778e-01 1.73177481e-01
-8.03800076e-02 1.14637557e-02 2.13869661e-01 -1.47110149e-01
5.32766223e-01 -6.96619809e-01 -4.61512089e-01 -8.06862354e-01
-3.53239119e-01 -6.08917654e-01 4.72733140e-01 -2.96148121e-01
1.00974047e+00 -1.92943990e+00 -4.01461823e-03 7.01936483e-01
3.84634256e-01 -1.97242916e-01 8.57217908e-01 5.27490318e-01
-1.52064279e-01 1.95871636e-01 -6.26695514e-01 1.99136436e-01
-8.32133591e-02 -1.36040777e-01 3.74416530e-01 5.51797152e-01
-3.96637283e-02 6.01066947e-01 -3.08872610e-01 -1.39349446e-01
-2.83748716e-01 6.48957431e-01 -7.98766553e-01 -4.57031935e-01
-6.07548535e-01 2.43524462e-01 -5.62582552e-01 9.96909261e-01
1.24942732e+00 -6.17171787e-02 7.73656368e-02 9.55770686e-02
-1.11591351e+00 7.17286050e-01 -9.41546202e-01 6.47156179e-01
3.23740095e-01 -1.81002870e-01 5.64126849e-01 -5.44316828e-01
1.05122411e+00 3.92528236e-01 3.82032812e-01 -1.08947611e+00
-8.93046185e-02 9.63380277e-01 4.71827477e-01 -3.23509574e-01
6.91070259e-01 -7.90845633e-01 2.28014871e-01 4.28182721e-01
-6.95384979e-01 -6.08175039e-01 1.84060633e-01 -3.02392364e-01
6.29143357e-01 -5.22607565e-01 6.55636936e-03 -1.30741179e+00
6.79622471e-01 2.87840188e-01 5.59246421e-01 5.35819054e-01
7.72408172e-02 5.38287997e-01 2.53267527e-01 -4.58808184e-01
-1.33992088e+00 -7.31957674e-01 -9.08843458e-01 2.86686391e-01
3.67185712e-01 -4.74344254e-01 -9.43939626e-01 5.57779312e-01
1.37651963e-02 8.24922085e-01 -2.98333496e-01 -2.30216146e-01
-7.42345750e-01 -1.60180831e+00 -1.62765130e-01 3.76865149e-01
4.63068545e-01 -1.47775042e+00 -6.23940706e-01 4.45877612e-01
2.34848559e-01 -4.57365900e-01 1.59502611e-01 5.13891399e-01
-8.74754310e-01 -7.75377154e-01 -4.74576950e-01 -3.12818646e-01
6.69156671e-01 1.08408295e-02 9.59692717e-01 4.23635006e-01
-7.80299366e-01 -3.91380712e-02 1.56058827e-02 -7.48808205e-01
-3.60421270e-01 -9.55452994e-02 5.08397639e-01 -5.17390132e-01
7.36574411e-01 -5.22812605e-01 -6.03653312e-01 2.78498232e-01
-4.62763250e-01 -3.16694885e-01 -6.97366567e-03 3.92612934e-01
1.12669146e+00 4.51751947e-01 4.62394655e-01 -7.82284260e-01
4.12065864e-01 -2.29508713e-01 -6.27218008e-01 -2.45034814e-01
-8.82871091e-01 -6.64571151e-02 2.76307732e-01 3.37424248e-01
-1.02140331e+00 -6.42206490e-01 -4.57500637e-01 3.14005911e-01
-2.13570043e-01 4.89062816e-01 -7.33176887e-01 -1.56510979e-01
3.82400364e-01 2.65727907e-01 -2.62696892e-01 -5.07001758e-01
-5.49915433e-01 6.69307232e-01 2.59937346e-01 -8.90458941e-01
3.82465333e-01 7.16116250e-01 2.83345371e-01 -1.37142086e+00
-1.34288639e-01 -7.25426376e-02 -5.61295390e-01 -2.41760910e-01
1.31126499e+00 -6.54613316e-01 -9.75093305e-01 6.08641922e-01
-8.27935159e-01 -1.24407411e-01 -4.50135410e-01 4.89313573e-01
-2.77269781e-01 1.98694065e-01 -5.16087532e-01 -9.13616419e-01
-5.39617360e-01 -1.35452259e+00 6.96743309e-01 4.39475983e-01
-4.13699783e-02 -1.05514705e+00 -3.81740034e-02 4.16355014e-01
6.10976636e-01 1.23479836e-01 1.61852944e+00 -2.61182874e-01
-7.30385125e-01 -6.00513630e-02 1.60635486e-01 -7.55329281e-02
7.70158246e-02 4.85718576e-03 -7.83082545e-01 -2.07626238e-01
4.63181615e-01 1.91983819e-01 9.55730915e-01 8.63435566e-01
7.93312252e-01 2.57471383e-01 -6.98435664e-01 9.38191265e-02
1.52503204e+00 6.81468010e-01 8.94667387e-01 5.80271006e-01
7.94254839e-01 4.44236457e-01 8.11932564e-01 4.84289467e-01
1.97366729e-01 4.54367787e-01 6.53384984e-01 2.69868344e-01
3.08640748e-01 3.89388233e-01 1.59675196e-01 8.69593740e-01
-9.06623781e-01 -3.15431714e-01 -1.22821927e+00 6.19926691e-01
-1.01724279e+00 -6.76329017e-01 -1.21509886e+00 2.58166838e+00
8.22558820e-01 5.80227911e-01 -2.86538959e-01 5.61454892e-01
8.10907125e-01 -3.69205810e-02 -6.38428390e-01 -4.41811323e-01
-3.59777570e-01 5.51574290e-01 6.51701093e-01 8.94266427e-01
-2.86574095e-01 8.27481985e-01 6.94527721e+00 4.29166645e-01
-1.35057831e+00 5.59583716e-02 2.54852444e-01 -3.17943543e-01
-7.59209454e-01 2.45618239e-01 -1.33420563e+00 5.71594656e-01
9.98729944e-01 -1.62697718e-01 -2.41801664e-01 4.09730405e-01
3.36757988e-01 -5.55306137e-01 -7.21100688e-01 3.57944638e-01
-1.36482641e-01 -1.38191998e+00 2.09616393e-01 5.66471696e-01
6.78700268e-01 -1.60613880e-01 -1.27517106e-02 -3.98334056e-01
-4.01435763e-01 -1.12830687e+00 6.84908986e-01 4.53927934e-01
9.36458528e-01 -6.25844836e-01 5.45139849e-01 3.20180170e-02
-1.23514843e+00 4.11395311e-01 -6.23865783e-01 -5.04371226e-01
5.20026922e-01 1.07319391e+00 -7.17520535e-01 3.70482624e-01
1.12179494e+00 3.66456747e-01 8.07091492e-05 7.44331539e-01
2.95004427e-01 4.61180121e-01 -5.45484543e-01 -1.79744571e-01
1.07226253e-01 -8.17013443e-01 5.89410484e-01 1.83848470e-01
1.51078373e-01 4.27073866e-01 -4.21907723e-01 1.22775507e+00
1.98517084e-01 9.02089328e-02 -3.02889615e-01 4.84887101e-02
4.62538064e-01 5.44183791e-01 -1.00276220e+00 -2.86824375e-01
4.63298745e-02 2.73471653e-01 -3.08590770e-01 1.68511551e-02
-5.24147689e-01 -4.14524406e-01 7.79483080e-01 1.17909634e+00
3.81175697e-01 -1.60324246e-01 -3.18257004e-01 -6.35660112e-01
2.99043089e-01 -1.60872623e-01 -3.81575823e-01 -3.27703387e-01
-8.08441281e-01 2.51318902e-01 2.44540721e-01 -1.83015332e-01
1.80096433e-01 -6.85892224e-01 -9.82073426e-01 1.09478533e+00
-1.56012249e+00 -4.40527827e-01 -1.19725600e-01 1.29358873e-01
1.33757338e-01 5.11606455e-01 6.41837776e-01 1.86372161e-01
-4.51368034e-01 1.50730029e-01 6.50715172e-01 -7.06946313e-01
3.46213937e-01 -1.06807315e+00 2.03733593e-01 2.69603372e-01
-9.65768576e-01 4.13158596e-01 1.35971260e+00 -1.02992165e+00
-1.52718461e+00 -8.95329654e-01 1.16769326e+00 3.27705666e-02
1.94402531e-01 -6.28141820e-01 -1.34337282e+00 3.43024373e-01
-1.08904671e-03 -1.49514198e-01 6.26235962e-01 -3.95849973e-01
5.28798580e-01 3.13716650e-01 -1.20944178e+00 4.29115295e-01
1.00559771e+00 -5.94386458e-02 -3.07968408e-01 5.36510706e-01
4.72856641e-01 -8.68096352e-02 -1.05543983e+00 6.78110540e-01
3.79397929e-01 -1.15551448e+00 9.02188480e-01 -9.02076215e-02
1.18330620e-01 -3.70842144e-02 -2.15954959e-01 -6.88207030e-01
-2.71800667e-01 -3.13089788e-02 6.37211740e-01 1.08928287e+00
4.54217404e-01 -9.94477808e-01 9.80092943e-01 8.37184906e-01
-6.12383842e-01 -7.05415368e-01 -1.42837632e+00 -5.24503529e-01
5.54832637e-01 1.14462249e-01 8.22514296e-01 5.20205200e-01
3.61260734e-02 -2.74261297e-03 2.70458668e-01 2.57673025e-01
8.75667036e-01 1.68084785e-01 -5.20242825e-02 -1.77122068e+00
3.03967386e-01 -1.71989620e-01 1.45356302e-04 -3.79998058e-01
2.07194760e-01 -7.14816391e-01 4.07707393e-02 -1.40770042e+00
-7.62640983e-02 -1.13088357e+00 -6.55488372e-02 -1.02621494e-02
2.15467408e-01 2.57834584e-01 -4.32231933e-01 4.67103004e-01
-1.70709178e-01 1.02272785e+00 1.12123227e+00 -3.70110869e-02
-3.42494160e-01 -3.24148446e-01 -4.65892643e-01 7.54805088e-01
6.41788960e-01 -6.85208380e-01 2.01832533e-01 -2.61988621e-02
5.22019863e-01 -1.12177439e-01 -2.00314894e-02 -1.17112279e+00
-8.38425085e-02 -1.60525665e-01 -2.18916506e-01 -7.60844409e-01
5.78655243e-01 -5.85085809e-01 5.63781917e-01 6.73525810e-01
3.51413429e-01 -1.25800610e-01 -4.01458666e-02 1.96679994e-01
-2.53708363e-01 -7.92110622e-01 8.05113077e-01 -5.01475871e-01
-1.16128944e-01 3.57677005e-02 -7.37109363e-01 -4.04551625e-01
1.32829952e+00 -7.23223686e-01 -1.47956014e-01 3.61582428e-01
-9.21609879e-01 -1.79904103e-02 1.15266883e+00 -2.23911121e-01
5.37948370e-01 -1.20351541e+00 -4.77601916e-01 4.73113596e-01
1.51433647e-01 5.03264189e-01 5.62231600e-01 8.59256804e-01
-1.12411988e+00 6.50121629e-01 -2.87555873e-01 -7.62422383e-01
-1.34189284e+00 2.48627916e-01 5.44413805e-01 2.93545574e-02
-4.98310238e-01 8.67261529e-01 3.30614239e-01 1.83171615e-01
-3.77180099e-01 -5.25145888e-01 -2.97643811e-01 -2.15186939e-01
4.18071300e-01 4.86308664e-01 5.30722558e-01 -7.42220223e-01
-3.92909139e-01 9.53445017e-01 -4.69652295e-01 5.08783087e-02
1.41799080e+00 7.28201270e-02 -7.28826761e-01 7.28146076e-01
3.75698596e-01 2.38443956e-01 -7.74584591e-01 3.57784152e-01
-1.11059912e-01 -3.02502036e-01 -2.21159309e-01 7.46830106e-02
-7.89611101e-01 6.42291665e-01 3.17517579e-01 2.48032629e-01
4.96999919e-01 3.65929484e-01 4.74625140e-01 -7.08545074e-02
4.76018548e-01 -1.07221961e+00 -4.89866138e-01 2.00276285e-01
5.46507001e-01 -8.12956214e-01 3.63391079e-02 -1.06312895e+00
-2.97076516e-02 7.96529651e-01 7.84011662e-01 -2.10404083e-01
8.23531628e-01 2.79986590e-01 -4.86764759e-01 -8.68357360e-01
-8.38280737e-01 2.60243177e-01 -3.44216257e-01 3.97195488e-01
7.60553300e-01 1.85916588e-01 -1.10396135e+00 5.09229302e-01
-3.48899692e-01 -3.04302454e-01 8.09424460e-01 1.38520193e+00
-8.67505848e-01 -1.26689577e+00 -1.00417030e+00 5.94873548e-01
-3.39875489e-01 -3.66395652e-01 -9.44355130e-02 6.49689913e-01
3.02540034e-01 5.92673302e-01 6.02485240e-01 4.00834948e-01
1.22171193e-01 4.02409792e-01 3.91347617e-01 -4.42588210e-01
-2.73230016e-01 6.40011653e-02 -1.98607314e-02 2.54582744e-02
-4.45375681e-01 -9.49488699e-01 -2.08056021e+00 -2.09308788e-01
-7.16540158e-01 1.00725508e+00 8.28477323e-01 8.65131140e-01
1.99972376e-01 4.80609417e-01 -7.59329945e-02 -1.13860679e+00
-1.88767284e-01 -8.00833821e-01 -1.37615085e+00 1.51033416e-01
-1.90541625e-01 -1.02018535e+00 -6.82182133e-01 -4.36676472e-01] | [5.3594536781311035, 4.90732479095459] |
fd609662-9d3e-4863-bb79-fb178f329f07 | attacking-and-defending-deep-learning-based | 2211.08291 | null | https://arxiv.org/abs/2211.08291v1 | https://arxiv.org/pdf/2211.08291v1.pdf | Attacking and Defending Deep-Learning-Based Off-Device Wireless Positioning Systems | Localization services for wireless devices play an increasingly important role and a plethora of emerging services and applications already rely on precise position information. Widely used on-device positioning methods, such as the global positioning system, enable accurate outdoor positioning and provide the users with full control over what services are allowed to access location information. To provide accurate positioning indoors or in cluttered urban scenarios without line-of-sight satellite connectivity, powerful off-device positioning systems, which process channel state information (CSI) with deep neural networks, have emerged recently. Such off-device positioning systems inherently link a user's data transmission with its localization, since accurate CSI measurements are necessary for reliable wireless communication -- this not only prevents the users from controlling who can access this information but also enables virtually everyone in the device's range to estimate its location, resulting in serious privacy and security concerns. We propose on-device attacks against off-device wireless positioning systems in multi-antenna orthogonal frequency-division multiplexing systems while minimizing the impact on quality-of-service, and we demonstrate their efficacy using measured datasets for outdoor and indoor scenarios. We also investigate defenses to counter such attack mechanisms, and we discuss the limitations and implications on protecting location privacy in future wireless communication systems. | ['Christoph Studer', 'Jakob Hoydis', 'K. Pavan Srinath', 'Maximilian Arnold', 'Emre Gönültaş', 'Pengzhi Huang'] | 2022-11-15 | null | null | null | null | ['outdoor-positioning'] | ['miscellaneous'] | [-1.09315291e-02 1.65493160e-01 -4.10941243e-01 -3.09842974e-01
-6.38673365e-01 -1.19833040e+00 -8.66235718e-02 -8.10114667e-02
-5.11735618e-01 8.53465497e-01 -1.26719862e-01 -8.81730080e-01
-2.84145534e-01 -6.68915570e-01 -3.97269666e-01 -8.66196513e-01
-3.41092288e-01 -2.07603239e-02 -2.88180143e-01 1.04422294e-01
1.46635115e-01 6.04854345e-01 -5.33766270e-01 -7.11369634e-01
4.75615948e-01 1.60503566e+00 -1.58107996e-01 5.74764550e-01
3.48826766e-01 -1.86727569e-01 -9.69367564e-01 -2.38870651e-01
3.02443057e-01 1.93037242e-01 3.02298479e-02 -4.21499342e-01
-3.46402079e-02 -6.57041371e-01 -7.09120512e-01 1.03300571e+00
1.00344980e+00 -2.12844178e-01 3.00427735e-01 -1.35821068e+00
-3.47145945e-01 1.51422590e-01 -2.84721911e-01 -2.03790870e-02
3.39372963e-01 -5.09198844e-01 5.13449550e-01 -6.84609869e-03
-2.86959168e-02 4.80530828e-01 1.09690201e+00 4.66742575e-01
-6.87112451e-01 -1.05929625e+00 -3.27600598e-01 -4.14235532e-01
-1.78180289e+00 -8.82404149e-01 3.90969902e-01 1.28673956e-01
3.30994785e-01 7.09345281e-01 1.84827223e-01 1.05600488e+00
7.23788559e-01 2.68493146e-01 2.73296714e-01 -9.84572619e-02
5.64069450e-01 2.01583415e-01 -3.58624876e-01 2.04587609e-01
8.52816284e-01 9.65988263e-02 -3.89571816e-01 -7.22636938e-01
8.43120754e-01 1.72767818e-01 -7.70735145e-01 -7.10838318e-01
-1.20932829e+00 2.59106219e-01 6.37478590e-01 5.62451541e-01
-2.99737424e-01 5.44658720e-01 -2.15789184e-01 2.50996649e-01
2.15452109e-02 5.51628232e-01 -5.37656248e-01 -2.47014955e-01
-7.80463696e-01 -2.05395237e-01 9.35984790e-01 1.55660558e+00
1.80788472e-01 -2.03853905e-01 6.44685328e-02 9.56294909e-02
4.83051538e-01 1.21718800e+00 2.38569319e-01 -8.80901992e-01
7.20790207e-01 -2.69615293e-01 8.38428259e-01 -1.39750183e+00
-7.21358776e-01 -1.26235712e+00 -1.29958248e+00 -4.84139383e-01
2.47880593e-01 -8.18657398e-01 -3.49420637e-01 1.89306569e+00
-8.09216872e-02 2.55055934e-01 1.43184692e-01 3.37205797e-01
8.70763958e-02 3.00759256e-01 -2.55085558e-01 -1.46876097e-01
1.23343825e+00 -2.76950032e-01 -8.16259563e-01 -2.56585956e-01
7.13500798e-01 -2.59906590e-01 7.90164247e-03 2.60670573e-01
-6.41375482e-01 -1.02867141e-01 -1.41936898e+00 5.05184889e-01
-4.79955882e-01 2.71954179e-01 7.63539135e-01 1.47512186e+00
-1.21174419e+00 1.52930170e-01 -1.03287637e+00 -2.67420024e-01
5.91829598e-01 1.21455789e+00 -2.23632872e-01 -9.75423381e-02
-1.35652339e+00 4.25291628e-01 -1.48085162e-01 4.36752826e-01
3.71142924e-02 -4.20608461e-01 -7.83371449e-01 5.08320987e-01
-1.82203680e-01 -7.01832831e-01 1.27388465e+00 -1.17777251e-01
-1.21159935e+00 2.12755218e-01 -3.38204473e-01 -5.66594183e-01
2.05249056e-01 9.63694081e-02 -1.05565345e+00 -1.66475520e-01
2.62857825e-01 7.31522068e-02 4.70127136e-01 -9.33458388e-01
-7.97123134e-01 -5.18608689e-01 6.97530881e-02 -1.16375849e-01
-4.24716949e-01 -4.17359442e-01 -3.47497553e-01 -2.45802730e-01
7.94520617e-01 -1.28869808e+00 -3.04742128e-01 2.17803970e-01
-7.45721459e-01 6.65452659e-01 9.28519309e-01 -2.31205419e-01
1.19818807e+00 -2.36796498e+00 -5.62685668e-01 6.52913868e-01
2.96968877e-01 4.94998954e-02 5.55907965e-01 3.29971135e-01
3.99943084e-01 3.63461822e-01 2.40368247e-01 -7.65128493e-01
2.81249341e-02 8.61841515e-02 -4.98881668e-01 7.85566032e-01
-7.18266189e-01 5.88677824e-01 -9.58745480e-01 3.76143634e-01
1.41410768e-01 5.59125245e-01 -4.79311913e-01 -3.25224847e-01
6.50974274e-01 9.81105447e-01 -9.13443923e-01 5.81183732e-01
1.15600562e+00 -3.42733413e-01 3.61746222e-01 4.88519631e-02
-1.36456668e-01 2.54208654e-01 -7.70538211e-01 1.44217646e+00
-7.97070146e-01 5.77603042e-01 5.64002275e-01 -4.98973966e-01
4.11528140e-01 6.12326145e-01 3.61942559e-01 -6.35984659e-01
5.42591035e-01 4.65494633e-01 -2.82503992e-01 -5.83783351e-02
3.86377603e-01 4.21099097e-01 -9.55462754e-01 2.91799784e-01
-3.67375016e-01 4.78209972e-01 -7.58997023e-01 9.32936743e-02
1.34315586e+00 -5.61294675e-01 5.01211882e-01 -1.74205512e-01
4.73048568e-01 -5.95876276e-01 4.45177704e-01 1.47758341e+00
-1.89035699e-01 2.11794287e-01 7.10626841e-02 1.34932861e-01
-3.31266493e-01 -1.00635374e+00 -4.82104480e-01 5.41132867e-01
8.51642489e-01 -2.07316741e-01 -3.01265299e-01 -4.36378241e-01
4.25605178e-01 4.39950824e-01 -2.48189017e-01 -1.35481864e-01
4.01990227e-02 -4.69714582e-01 1.07192898e+00 2.33538762e-01
9.82006490e-01 -1.31491898e-02 -3.36928926e-02 2.66380668e-01
-3.35836709e-01 -1.22116351e+00 -4.34489667e-01 3.81193012e-01
-2.36615717e-01 -4.10502404e-01 -7.94030666e-01 -4.51781392e-01
8.30700815e-01 8.28238606e-01 5.31170666e-01 -9.63928457e-03
4.58523005e-01 5.67133844e-01 1.41683994e-02 -2.13967577e-01
8.53616744e-02 5.82671404e-01 7.33694911e-01 2.39093021e-01
2.14831963e-01 -9.19711053e-01 -6.85623050e-01 8.31281781e-01
-8.29324052e-02 -6.93735778e-01 5.27451932e-01 3.02858025e-01
1.08989358e-01 6.37041628e-01 6.13871932e-01 -4.04204398e-01
2.35198528e-01 -7.80710042e-01 -8.44237804e-01 7.91502297e-02
-6.56685755e-02 -1.62035510e-01 3.14020663e-01 -4.80637923e-02
-3.75696808e-01 -1.86719410e-02 -2.98189312e-01 2.64936894e-01
1.12597138e-01 2.73512810e-01 -8.29770029e-01 -1.11578274e+00
4.34677988e-01 2.79870450e-01 -6.47630572e-01 -3.17150772e-01
7.53730610e-02 1.41211295e+00 7.19636500e-01 -1.72510639e-01
1.07043564e+00 6.75745904e-01 1.83630541e-01 -1.05166435e+00
-4.98353809e-01 -6.97111368e-01 -4.09819990e-01 2.39121422e-01
3.98951083e-01 -1.30959570e+00 -1.26088417e+00 5.90522766e-01
-1.20930362e+00 3.11302483e-01 7.57453442e-01 4.01849419e-01
7.76919425e-02 2.86718100e-01 -4.07130271e-02 -1.03650296e+00
-4.60824324e-03 -8.90238702e-01 1.09457028e+00 5.04977643e-01
-2.49177054e-01 -9.29819167e-01 -5.15756369e-01 -5.07738665e-02
1.29195416e+00 1.89904749e-01 2.31324807e-01 -4.90865201e-01
-8.99720192e-01 -1.00268841e+00 -6.81930035e-03 -2.56500572e-01
4.83657926e-01 -1.13612258e+00 -9.73830044e-01 -6.45687759e-01
1.30417258e-01 5.43378294e-01 -1.84698403e-01 8.57374728e-01
1.30642128e+00 -5.10377288e-01 -1.23260915e+00 1.11675346e+00
9.50157166e-01 1.67804733e-01 8.22899044e-01 8.13194662e-02
4.45482105e-01 -4.10394728e-01 4.21063781e-01 3.83968413e-01
3.60950828e-01 8.79980206e-01 4.78786319e-01 2.24528953e-01
7.44629979e-01 -8.51088166e-02 -1.22256145e-01 -2.23634299e-03
4.74080354e-01 -1.08887649e+00 -3.94770503e-01 6.25405535e-02
-1.65603566e+00 -6.88262999e-01 3.51668932e-02 2.67855453e+00
1.45306185e-01 4.05773699e-01 -4.33428943e-01 -1.26560643e-01
7.46750951e-01 7.72980750e-02 -6.16064787e-01 1.90835699e-01
7.22868368e-02 -1.62111551e-01 1.82624197e+00 6.43440366e-01
-1.48900831e+00 4.84101683e-01 5.41134501e+00 5.44973552e-01
-1.27787650e+00 2.69067138e-01 5.29908419e-01 -6.88018128e-02
-6.89588264e-02 -3.24185699e-01 -9.60042894e-01 8.04601490e-01
9.00920391e-01 2.18472168e-01 6.50078282e-02 1.09013283e+00
1.41743705e-01 -3.78253609e-01 -8.50566924e-01 1.36873257e+00
-3.29312146e-01 -1.34782827e+00 -8.45684826e-01 9.37108696e-01
4.94810969e-01 8.14361572e-02 4.57391649e-01 1.12256244e-01
-8.59267786e-02 -5.40191770e-01 3.84641230e-01 2.89658189e-01
9.56225574e-01 -1.01750493e+00 1.03623271e+00 4.51022983e-01
-1.32637680e+00 -3.26607972e-01 -2.43729874e-01 -1.97339773e-01
3.93286616e-01 8.04639161e-01 -4.37634557e-01 5.74118674e-01
5.59105933e-01 2.55828589e-01 -1.97831169e-01 1.36001801e+00
-3.81854624e-01 2.29417548e-01 -9.40057278e-01 -1.75807983e-01
1.99426953e-02 2.14934021e-01 5.08682847e-01 4.54455853e-01
1.10643637e+00 2.80982375e-01 -1.57113448e-01 2.22900212e-01
-2.44396418e-01 -6.76244080e-01 -7.78973460e-01 2.97932297e-01
1.52516878e+00 8.87750089e-01 -5.53815663e-01 2.46998832e-01
1.52609835e-03 1.37347519e+00 -6.57208979e-01 4.62085694e-01
-6.71387434e-01 -9.94815409e-01 1.17586946e+00 1.02483451e-01
2.63324618e-01 -1.02513695e+00 -3.20394486e-01 -1.01815283e+00
-5.01360446e-02 -1.98185205e-01 -2.82817483e-01 -4.81693089e-01
-8.12353075e-01 1.08118154e-01 -7.42604434e-01 -1.67171848e+00
-3.67616326e-01 -2.15285465e-01 -2.67189950e-01 9.23255324e-01
-1.14888799e+00 -8.40077877e-01 -1.43276513e-01 4.28710252e-01
-6.61681473e-01 8.82255957e-02 1.14469934e+00 6.68474019e-01
-2.23495454e-01 1.20125651e+00 1.00231647e+00 4.22938168e-01
3.66912067e-01 -6.90051496e-01 8.25119853e-01 8.32154512e-01
-2.36460790e-01 1.15484512e+00 4.16704237e-01 -5.39516985e-01
-1.71314800e+00 -8.49795103e-01 9.79743242e-01 -6.71424508e-01
2.96106011e-01 -8.79917443e-01 -1.68524474e-01 6.99808538e-01
-3.94993484e-01 1.91544831e-01 1.10863066e+00 5.89920133e-02
1.97589666e-01 -1.30408481e-01 -1.60230172e+00 6.32514179e-01
9.75492716e-01 -6.35717511e-01 2.22406849e-01 4.87585142e-02
4.05504465e-01 -5.68583071e-01 -3.97825092e-01 5.57949729e-02
9.39093351e-01 -6.62335753e-01 1.03829968e+00 2.68006414e-01
-8.94293785e-01 -5.26350141e-01 -4.09008354e-01 -9.84276354e-01
-3.86535227e-01 -1.25902450e+00 -2.29122162e-01 1.24100280e+00
3.24724972e-01 -1.05725181e+00 1.13049197e+00 8.04206848e-01
3.68915886e-01 -5.30361161e-02 -1.39574218e+00 -8.68559420e-01
-6.53960466e-01 -5.97755015e-01 8.91484499e-01 7.15061247e-01
-1.31194919e-01 5.50744422e-02 -7.24159956e-01 1.51381457e+00
7.55695760e-01 -6.57690823e-01 8.95750880e-01 -1.34823656e+00
-2.54985452e-01 1.09496966e-01 -7.09501982e-01 -1.97876954e+00
-1.93742812e-01 -4.09215152e-01 1.88175552e-02 -1.13910151e+00
-9.67829049e-01 -9.92137015e-01 -3.46991330e-01 4.36488420e-01
6.52821720e-01 5.22254050e-01 -4.99567956e-01 2.72004418e-02
-8.19131911e-01 3.07914734e-01 4.82323408e-01 3.72983366e-02
-3.05110335e-01 1.15973842e+00 -1.23951113e+00 3.92908216e-01
8.59839082e-01 -4.97522146e-01 -3.65988135e-01 -3.99206668e-01
3.13340634e-01 3.25723112e-01 2.21485093e-01 -1.69659686e+00
3.90510350e-01 3.90951663e-01 7.52562821e-01 -3.73626649e-01
6.29161716e-01 -1.54396534e+00 1.42023593e-01 6.08163536e-01
2.10671663e-01 -5.14431536e-01 1.04416385e-01 9.78887141e-01
4.27971274e-01 2.86127061e-01 4.39697415e-01 5.59380233e-01
-4.55580443e-01 3.92921269e-01 -7.74623215e-01 -7.50938416e-01
7.47896910e-01 -2.30082378e-01 -3.35999340e-01 -1.26090944e+00
-6.04320109e-01 5.24685979e-01 5.26711524e-01 3.27454031e-01
-1.57193150e-02 -1.35802257e+00 3.75968039e-01 3.86499405e-01
1.04429545e-02 -4.41376507e-01 2.36098036e-01 7.06670582e-01
-1.65861815e-01 1.09332252e+00 2.09893659e-01 -5.34781873e-01
-1.04452205e+00 -2.71028746e-03 5.72007358e-01 2.64582008e-01
3.74642015e-02 9.82820690e-01 -9.83116105e-02 -3.56557041e-01
6.05574191e-01 -3.28892410e-01 2.24518076e-01 -4.90059793e-01
7.57327855e-01 -1.89135656e-01 2.82199323e-01 -5.26576340e-01
-7.03660667e-01 5.09739459e-01 2.29610145e-01 -1.50668249e-01
6.85635030e-01 -8.45980406e-01 2.59937942e-01 -3.93029600e-01
1.24863255e+00 6.55036151e-01 -1.01929164e+00 -1.66865692e-01
-1.00707360e-01 -6.18249059e-01 2.32293144e-01 -6.14029825e-01
-7.67986059e-01 4.47378248e-01 8.59210789e-01 5.79251409e-01
6.59256279e-01 -1.75523877e-01 1.00017762e+00 8.85728240e-01
1.52715576e+00 -4.13714468e-01 -8.17762196e-01 2.71182299e-01
-4.62962687e-02 -1.13555837e+00 -1.07222266e-01 -3.13444734e-01
2.34247461e-01 8.64333868e-01 -1.43890530e-01 3.40090811e-01
1.08545959e+00 3.11933607e-01 4.49441634e-02 -1.79750081e-02
3.35267007e-01 2.72419542e-01 -7.73934424e-02 1.07061410e+00
-3.76155190e-02 5.64615615e-02 2.71773189e-01 1.16519928e+00
-4.48789924e-01 -2.51317531e-01 2.48563364e-01 1.28828442e+00
-5.92565775e-01 -1.39359856e+00 -4.67883646e-01 4.85405684e-01
-6.36047959e-01 6.22770526e-02 4.17755498e-03 4.90131736e-01
6.35158829e-03 1.23962498e+00 -3.11846309e-03 -5.25869966e-01
-6.32884949e-02 -4.34645802e-01 1.23708792e-01 -3.67625892e-01
2.75861174e-01 -3.42568904e-01 9.82806608e-02 -5.61199486e-01
2.52362281e-01 -3.96747053e-01 -9.66435015e-01 -4.96169716e-01
-5.33294201e-01 5.56013227e-01 1.21887779e+00 9.90747631e-01
9.15241301e-01 3.76033932e-01 1.00645983e+00 -1.14158142e+00
-2.74482548e-01 -2.16815740e-01 -8.25204909e-01 -9.27955210e-01
8.04752827e-01 -6.43633962e-01 -5.47669768e-01 -9.41209853e-01] | [6.391132354736328, 0.9222306609153748] |
641e7d20-82ce-4cd8-a8f6-e0359ac08192 | safe-reinforcement-learning-with-self | 2304.08897 | null | https://arxiv.org/abs/2304.08897v2 | https://arxiv.org/pdf/2304.08897v2.pdf | An adaptive safety layer with hard constraints for safe reinforcement learning in multi-energy management systems | Safe reinforcement learning (RL) with hard constraint guarantees is a promising optimal control direction for multi-energy management systems. It only requires the environment-specific constraint functions itself a priori and not a complete model (i.e. plant, disturbance and noise models, and prediction models for states not included in the plant model - e.g. demand forecasts, weather forecasts, price forecasts). The project-specific upfront and ongoing engineering efforts are therefore still reduced, better representations of the underlying system dynamics can still be learned and modelling bias is kept to a minimum (no model-based objective function). However, even the constraint functions alone are not always trivial to accurately provide in advance, leading to potentially unsafe behaviour. In this paper, we present two novel advancements: (I) combining the Optlayer and SafeFallback method, named OptLayerPolicy, to increase the initial utility while keeping a high sample efficiency. (II) introducing self-improving hard constraints, to increase the accuracy of the constraint functions as more data becomes available so that better policies can be learned. Both advancements keep the constraint formulation decoupled from the RL formulation, so that new (presumably better) RL algorithms can act as drop-in replacements. We have shown that, in a simulated multi-energy system case study, the initial utility is increased to 92.4% (OptLayerPolicy) compared to 86.1% (OptLayer) and that the policy after training is increased to 104.9% (GreyOptLayerPolicy) compared to 103.4% (OptLayer) - all relative to a vanilla RL benchmark. While introducing surrogate functions into the optimization problem requires special attention, we do conclude that the newly presented GreyOptLayerPolicy method is the most advantageous. | ['Maarten Messagie', 'Ann Nowé', 'Rüdiger Franke', 'Muhammad Andy Putratama', 'Glenn Ceusters'] | 2023-04-18 | null | null | null | null | ['energy-management'] | ['time-series'] | [ 6.43876866e-02 3.22355002e-01 -3.87671441e-01 1.38971508e-01
-6.83118701e-01 -6.73228383e-01 5.50752997e-01 2.02608183e-01
-3.54434252e-01 1.27646112e+00 -2.64603168e-01 -4.09538150e-01
-4.73652244e-01 -7.20760822e-01 -6.98779941e-01 -9.89399910e-01
-2.08876863e-01 4.26065832e-01 -1.52401656e-01 -2.08386257e-01
-5.17266393e-02 5.59958577e-01 -1.47399127e+00 -3.61620158e-01
1.00461006e+00 1.04252684e+00 3.70223194e-01 5.41000605e-01
3.44490230e-01 5.89199781e-01 -3.80307317e-01 2.80483246e-01
6.19310558e-01 -2.63457716e-01 -3.80397707e-01 -2.08632555e-02
-3.85809332e-01 -4.20190722e-01 1.81663573e-01 1.09036672e+00
4.08959687e-01 3.29100132e-01 4.47604388e-01 -1.33459258e+00
-1.72205344e-01 3.77329856e-01 -4.72034037e-01 -2.53574163e-01
-7.24153146e-02 6.70148373e-01 8.01405311e-01 -1.89741611e-01
3.93783569e-01 8.71484578e-01 3.83349538e-01 4.64449197e-01
-1.38686168e+00 -6.48725092e-01 4.68660295e-01 1.34543665e-02
-1.15376782e+00 -3.52354497e-01 6.13581121e-01 -2.77047724e-01
1.31105447e+00 6.03137791e-01 7.43652523e-01 9.01825726e-01
2.43601456e-01 4.37820911e-01 1.28697908e+00 -4.05681670e-01
6.82108879e-01 4.69566494e-01 1.35802925e-02 4.34579313e-01
4.67458427e-01 7.50637770e-01 6.85865134e-02 -1.57254264e-01
5.13379216e-01 -2.88077712e-01 -3.69371980e-01 -6.06190383e-01
-6.88910723e-01 7.95642197e-01 1.97063982e-01 2.72821814e-01
-6.44816637e-01 3.19421679e-01 2.77244598e-01 3.86046052e-01
4.15294230e-01 8.00824881e-01 -7.32489109e-01 -3.82666141e-02
-1.02612674e+00 3.84046912e-01 8.51074398e-01 7.75565386e-01
7.22367525e-01 8.09738934e-01 -5.04452328e-04 5.43573797e-01
1.42247006e-01 6.86182618e-01 3.10957998e-01 -1.11247396e+00
2.65619189e-01 3.16923887e-01 7.23347366e-01 -5.57080090e-01
-6.42735660e-01 -8.89599025e-01 -8.12783897e-01 7.28463590e-01
4.31479126e-01 -6.53415322e-01 -8.73284280e-01 1.77423179e+00
1.46557584e-01 2.42961384e-02 2.01955825e-01 7.30064094e-01
-1.85829669e-01 1.07843220e+00 -7.28531405e-02 -9.30025458e-01
1.01500714e+00 -6.87877536e-01 -7.23633289e-01 -2.57678837e-01
5.54611385e-01 -4.40248132e-01 8.78286123e-01 7.47963250e-01
-1.18123913e+00 -3.11275512e-01 -1.27620065e+00 7.85519361e-01
-6.07914686e-01 4.16501537e-02 4.62074727e-01 7.43449926e-01
-9.19685721e-01 8.52818668e-01 -8.56884181e-01 -1.08544827e-01
-2.21766848e-02 6.33916616e-01 1.59088179e-01 1.42544016e-01
-1.20403183e+00 1.36331594e+00 5.76069891e-01 2.01474935e-01
-9.66383398e-01 -9.03706372e-01 -7.06971228e-01 2.45140344e-01
9.60970223e-01 -4.99331355e-01 1.24956524e+00 -8.60871553e-01
-1.94178283e+00 -1.53196096e-01 1.80739477e-01 -6.04458928e-01
6.21038198e-01 2.18095072e-02 -3.64823192e-01 -1.56187847e-01
-4.54808056e-01 2.12691426e-01 7.38252640e-01 -1.56737435e+00
-7.36519456e-01 5.30188680e-02 8.29962268e-02 1.55946240e-01
-1.20066218e-01 -3.15293670e-01 1.84956089e-01 -4.14593965e-01
-4.62174296e-01 -1.09259534e+00 -5.11270344e-01 -4.06782210e-01
-2.83888340e-01 -6.60263747e-02 7.28787005e-01 -7.71657825e-01
1.15306437e+00 -1.76827192e+00 1.48269758e-01 5.01249611e-01
-3.97964776e-01 3.71228486e-01 -5.94038218e-02 5.01168072e-01
-3.17364395e-01 2.40779966e-01 -3.49466592e-01 -1.15549050e-01
2.76060998e-01 5.44997454e-01 -1.92741230e-01 4.43980753e-01
2.86135644e-01 5.89599371e-01 -9.14731205e-01 1.84107319e-01
5.77123106e-01 7.18777254e-02 -4.55223978e-01 -1.37949595e-02
-6.43822551e-01 2.56644040e-01 -4.94076401e-01 3.33474040e-01
5.07044077e-01 1.54714599e-01 2.83172995e-01 -6.61787689e-02
-4.26207393e-01 -1.35483593e-01 -1.56957912e+00 1.16579890e+00
-9.24236774e-01 1.94605783e-01 5.27952909e-01 -1.23448312e+00
7.62222648e-01 3.31066757e-01 8.16645503e-01 -7.19945490e-01
3.84780690e-02 2.82734931e-01 7.07027465e-02 -2.88872719e-01
3.62622142e-01 -2.84440935e-01 6.05150359e-03 3.40816140e-01
-2.15999305e-01 -3.67970109e-01 3.30182463e-01 -3.19122285e-01
8.25867772e-01 3.59258533e-01 4.30060446e-01 -6.59628510e-01
4.96945649e-01 7.24778324e-02 8.73875797e-01 3.70326340e-01
6.26258776e-02 -9.98526365e-02 5.52562654e-01 -3.62202004e-02
-9.82857168e-01 -5.64970434e-01 8.40297062e-03 5.25749803e-01
-1.90654665e-01 -1.08361654e-01 -4.53969657e-01 -6.53904319e-01
2.54719406e-01 1.45341551e+00 -2.77354211e-01 -3.68435740e-01
-5.28869152e-01 -9.57331538e-01 -8.46443251e-02 3.30049068e-01
2.76914358e-01 -6.73509061e-01 -7.83318460e-01 4.57608044e-01
3.84976000e-01 -7.62049317e-01 -6.41076937e-02 8.13701212e-01
-6.96747601e-01 -8.87308240e-01 -5.85673809e-01 -4.64236587e-02
6.05012596e-01 -2.90557653e-01 8.54895055e-01 -1.16319157e-01
-3.04200687e-02 3.77602875e-01 -7.05437511e-02 -4.31327343e-01
-6.38276517e-01 -2.26562396e-01 3.11825007e-01 -3.06726515e-01
-2.87884116e-01 -4.14869070e-01 -3.08532506e-01 1.99134469e-01
-6.46254063e-01 -1.96183264e-01 5.58502495e-01 1.00372648e+00
5.31267881e-01 6.76345289e-01 8.08062494e-01 -7.18188405e-01
5.35381615e-01 -3.12109858e-01 -1.36054325e+00 3.32169890e-01
-1.40927875e+00 3.16330194e-01 1.10859323e+00 -4.75211233e-01
-1.00984454e+00 1.87028512e-01 9.41652879e-02 -4.77651477e-01
4.46527973e-02 3.52192760e-01 -1.21175550e-01 -9.76840854e-02
2.15939209e-01 2.13260785e-01 1.42007574e-01 -3.84272933e-01
2.43657783e-01 3.32029015e-01 1.83008268e-01 -7.09399879e-01
9.41995442e-01 -1.57010496e-01 2.85678208e-01 -6.30182922e-01
-3.62179428e-01 -7.87675828e-02 -2.48168856e-01 -3.14273924e-01
4.98937428e-01 -7.91281819e-01 -9.84830320e-01 7.23126382e-02
-6.02060020e-01 -7.61967599e-01 -7.42151678e-01 3.91865790e-01
-7.40189910e-01 1.17661081e-01 -1.91766024e-01 -1.57934773e+00
-1.85968846e-01 -1.17856586e+00 5.72078168e-01 2.44095311e-01
-6.12572394e-02 -9.76197660e-01 -1.33453667e-01 -1.51155412e-01
5.52850664e-01 6.45308852e-01 1.06754696e+00 -4.32823718e-01
-5.09791911e-01 -4.81438339e-02 1.97368786e-01 6.66383743e-01
1.13250315e-01 1.64478391e-01 -8.46128047e-01 -7.37844169e-01
3.10029685e-01 -1.49939910e-01 5.47024369e-01 4.19852883e-01
1.01192892e+00 -6.63973868e-01 -3.02399278e-01 2.33025923e-01
1.73694825e+00 7.19100595e-01 2.47051522e-01 3.29475224e-01
1.91904068e-01 5.88342130e-01 7.49059737e-01 6.72840476e-01
1.53636545e-01 6.92207456e-01 7.45516121e-01 -1.39638826e-01
2.06074774e-01 -3.55965160e-02 5.84786296e-01 5.57747185e-01
-1.10505126e-01 -3.29186887e-01 -6.57366931e-01 3.04312766e-01
-1.89400518e+00 -8.70699644e-01 1.23844724e-02 2.61260414e+00
6.91450655e-01 4.31111306e-01 2.32725665e-01 2.40033403e-01
3.33257049e-01 -2.44708210e-02 -9.98907566e-01 -8.09704125e-01
2.16471538e-01 2.60035425e-01 9.89746153e-01 7.22736895e-01
-7.84548402e-01 3.52067918e-01 5.86824465e+00 8.35006237e-01
-1.02111864e+00 -9.50735807e-02 4.56716686e-01 -2.70222813e-01
-1.86173782e-01 9.52127129e-02 -7.07798660e-01 6.15387559e-01
1.43700194e+00 -5.33335805e-01 1.01240659e+00 8.37272048e-01
7.94995546e-01 -5.06998301e-01 -1.02450418e+00 5.26091337e-01
-5.17224014e-01 -1.02370977e+00 -5.18009007e-01 3.94953281e-01
6.49811745e-01 -2.57515043e-01 -1.78952113e-01 7.52488494e-01
5.94454825e-01 -8.73703301e-01 7.52066612e-01 6.67798698e-01
4.70430076e-01 -1.17710197e+00 7.76154935e-01 7.85838723e-01
-1.07397461e+00 -4.99687016e-01 -1.91093132e-01 -9.03592184e-02
1.56712979e-01 7.26884842e-01 -6.64370000e-01 1.14406443e+00
3.04714918e-01 4.39953774e-01 -1.82980582e-01 7.99941123e-01
-2.63708532e-02 7.02581823e-01 -6.83617175e-01 -2.17700079e-01
3.70702147e-01 -3.60411078e-01 5.29482603e-01 8.57871294e-01
2.69080728e-01 -6.54540304e-03 5.80871940e-01 9.23664331e-01
4.84646976e-01 -1.78870693e-01 -4.29407448e-01 -8.73255059e-02
2.22149089e-01 1.35717988e+00 -4.06975120e-01 -9.11921337e-02
-1.73679888e-01 5.19684196e-01 -1.19994842e-01 5.04216254e-01
-8.60867500e-01 -1.60146385e-01 7.70356297e-01 -1.43755704e-01
2.48188749e-01 -2.79705524e-01 -8.86283442e-02 -7.14677989e-01
-2.24131644e-01 -9.39909279e-01 2.81322688e-01 -5.69419324e-01
-1.04644847e+00 1.42702535e-01 2.65841514e-01 -1.06692779e+00
-8.99492919e-01 -6.34537697e-01 -4.47455257e-01 1.03277707e+00
-1.90315557e+00 -6.15890563e-01 1.98092416e-01 2.28671372e-01
6.31089449e-01 -3.11684068e-02 8.83440018e-01 2.68342830e-02
-1.01027191e+00 3.00848097e-01 6.66905284e-01 -6.68877065e-01
3.02422583e-01 -1.46296120e+00 -2.17654124e-01 7.01559544e-01
-4.40940291e-01 2.69481033e-01 1.04817832e+00 -4.68822628e-01
-1.66187918e+00 -1.07213044e+00 2.36175537e-01 -6.54501170e-02
8.13793600e-01 -2.75527537e-01 -8.46344471e-01 2.38741204e-01
4.77962941e-01 -2.05844596e-01 2.42985398e-01 -1.47914261e-01
3.78672659e-01 -3.51802111e-01 -1.39637983e+00 4.95725900e-01
4.70407158e-01 -7.44729862e-02 -1.04871862e-01 4.12255079e-01
5.38284481e-01 -2.73817390e-01 -1.08624315e+00 4.64924663e-01
2.73535341e-01 -5.16749680e-01 7.02697814e-01 -4.98497993e-01
-2.76990443e-01 -3.81840169e-01 -1.04382053e-01 -1.74180079e+00
-2.91321784e-01 -9.93753493e-01 -5.93641818e-01 1.31653690e+00
4.79269356e-01 -1.00952959e+00 5.88260055e-01 9.34889257e-01
-2.23168850e-01 -9.99104977e-01 -1.01213193e+00 -1.40504885e+00
3.71252090e-01 -3.77178758e-01 4.83085096e-01 8.64910364e-01
-1.47531539e-01 3.91865969e-02 -3.47593963e-01 3.72794777e-01
5.77478409e-01 6.22410141e-02 4.16351646e-01 -8.88917863e-01
-5.74214101e-01 -6.02996588e-01 3.17827731e-01 -5.58928668e-01
2.08860770e-01 -5.93196213e-01 1.17489681e-01 -1.43667579e+00
-3.36876452e-01 -5.68505406e-01 -5.62480688e-01 6.83418393e-01
6.67744130e-02 -4.99630690e-01 3.86182427e-01 -2.10605234e-01
1.51478723e-01 7.40068257e-01 9.85126853e-01 -2.02869847e-01
-4.00070161e-01 3.51792097e-01 -4.57167059e-01 5.09688377e-01
1.10727775e+00 -4.34428453e-01 -8.25736165e-01 1.97763726e-01
8.63069482e-03 3.91224563e-01 3.07354748e-01 -9.26110446e-01
1.71797276e-02 -6.55793905e-01 2.22983092e-01 -3.76204401e-01
2.57044464e-01 -1.20503724e+00 5.08655131e-01 8.95184219e-01
-8.26304331e-02 -2.29880307e-03 5.86890697e-01 6.31205320e-01
2.73099214e-01 -4.57786381e-01 7.34266818e-01 -1.11701086e-01
-6.46783531e-01 -4.24800180e-02 -5.22762716e-01 -2.45059431e-01
1.29179502e+00 -5.55150174e-02 -2.25427508e-01 -4.03655589e-01
-6.98170364e-01 8.09635043e-01 3.49901348e-01 1.88793451e-01
-1.72236599e-02 -1.05855823e+00 -4.29008037e-01 -1.25558108e-01
-4.08612996e-01 -1.93346232e-01 -5.46348393e-02 7.84308314e-01
1.35447472e-01 5.82993627e-01 7.94190913e-02 -3.03457648e-01
-8.01776886e-01 6.98714316e-01 5.97942173e-01 -7.14249313e-01
-4.40894037e-01 2.60863066e-01 -2.07371309e-01 -2.10429505e-01
1.63474992e-01 -6.96736932e-01 1.19794190e-01 1.06691763e-01
1.32376745e-01 5.82136989e-01 2.14925587e-01 5.04522733e-02
-2.83881456e-01 3.95826310e-01 1.85411587e-01 -4.01675552e-02
1.49452662e+00 4.69422191e-02 3.40861171e-01 4.05754328e-01
6.82819605e-01 -2.39708200e-02 -1.59299338e+00 2.68242776e-01
1.87624082e-01 -8.45889226e-02 5.63949823e-01 -1.39786959e+00
-1.14235079e+00 4.18548256e-01 7.07776308e-01 6.69056892e-01
1.38822949e+00 -7.21766293e-01 4.70770031e-01 4.28735435e-01
7.05438197e-01 -1.46056664e+00 -3.98994058e-01 2.52143681e-01
1.05523348e+00 -8.22678924e-01 3.43635827e-01 -1.64701920e-02
-4.96976227e-01 9.91523027e-01 5.04672229e-01 -1.01512149e-01
4.74206418e-01 5.79479396e-01 -4.57905471e-01 3.77269268e-01
-1.05912924e+00 -4.11530495e-01 -1.10705197e-02 4.64499235e-01
-7.95037001e-02 8.44437703e-02 -3.42831999e-01 6.86227620e-01
1.10942401e-01 -6.91598654e-02 5.78989327e-01 9.50882077e-01
-4.67643976e-01 -1.22703373e+00 -3.83942991e-01 4.55418348e-01
-2.08238158e-02 2.86928326e-01 8.79599452e-02 1.28849387e+00
1.40758529e-01 1.04854965e+00 -2.94661850e-01 -1.02978721e-02
7.73607969e-01 1.61016181e-01 3.58507484e-01 -4.75559950e-01
-6.60466075e-01 1.68337435e-01 5.00152111e-01 -6.84461415e-01
-5.12468629e-02 -7.54638731e-01 -1.30322242e+00 -2.60807663e-01
-7.82536268e-01 1.96714014e-01 8.31123054e-01 8.55990827e-01
2.75667220e-01 8.97942185e-01 9.39317286e-01 -9.08569336e-01
-1.23612130e+00 -6.20436430e-01 -6.25588775e-01 -3.75784934e-01
3.72101009e-01 -8.85081053e-01 -5.33142328e-01 -3.28883410e-01] | [5.227841854095459, 2.4435436725616455] |
3931a542-fdfc-411a-be4c-c8666a0089c7 | how-good-is-your-tokenizer-on-the-monolingual | 2012.15613 | null | https://arxiv.org/abs/2012.15613v2 | https://arxiv.org/pdf/2012.15613v2.pdf | How Good is Your Tokenizer? On the Monolingual Performance of Multilingual Language Models | In this work, we provide a systematic and comprehensive empirical comparison of pretrained multilingual language models versus their monolingual counterparts with regard to their monolingual task performance. We study a set of nine typologically diverse languages with readily available pretrained monolingual models on a set of five diverse monolingual downstream tasks. We first aim to establish, via fair and controlled comparisons, if a gap between the multilingual and the corresponding monolingual representation of that language exists, and subsequently investigate the reason for any performance difference. To disentangle conflating factors, we train new monolingual models on the same data, with monolingually and multilingually trained tokenizers. We find that while the pretraining data size is an important factor, a designated monolingual tokenizer plays an equally important role in the downstream performance. Our results show that languages that are adequately represented in the multilingual model's vocabulary exhibit negligible performance decreases over their monolingual counterparts. We further find that replacing the original multilingual tokenizer with the specialized monolingual tokenizer improves the downstream performance of the multilingual model for almost every task and language. | ['Iryna Gurevych', 'Sebastian Ruder', 'Ivan Vulić', 'Jonas Pfeiffer', 'Phillip Rust'] | 2020-12-31 | null | https://aclanthology.org/2021.acl-long.243 | https://aclanthology.org/2021.acl-long.243.pdf | acl-2021-5 | ['pretrained-multilingual-language-models'] | ['natural-language-processing'] | [-3.90922844e-01 -3.00042093e-01 -5.49063623e-01 -2.04122111e-01
-1.01909935e+00 -1.25439131e+00 8.42655003e-01 2.15884537e-01
-1.03085613e+00 9.57330644e-01 4.58364546e-01 -9.74305212e-01
2.29491085e-01 -3.28403682e-01 -8.32899451e-01 -3.64676803e-01
3.05036962e-01 7.39745140e-01 -7.72187859e-02 -4.61786777e-01
-2.18527950e-02 2.27636695e-01 -1.08318162e+00 6.66626617e-02
1.15338302e+00 8.51491094e-02 5.15350521e-01 2.90791512e-01
-4.26339246e-02 6.40490413e-01 -4.00922209e-01 -4.66656864e-01
4.30269957e-01 -1.50044039e-01 -8.60550404e-01 -4.48030233e-01
8.10058713e-01 -5.83072715e-02 -1.89432994e-01 9.81654644e-01
2.95499712e-01 -1.72762141e-01 7.91797280e-01 -6.65933192e-01
-9.99468386e-01 1.33797514e+00 -2.95284718e-01 4.93328333e-01
3.05039197e-01 2.46891707e-01 1.35576618e+00 -1.14390600e+00
9.64234054e-01 1.41539133e+00 6.69376433e-01 9.84651223e-02
-1.61072719e+00 -9.10367072e-01 5.49096048e-01 1.85802057e-02
-1.58682084e+00 -7.52927423e-01 2.92169720e-01 -7.10434973e-01
1.22922361e+00 -2.35064924e-01 4.82632697e-01 1.12749577e+00
5.21766186e-01 3.21687669e-01 1.68017268e+00 -5.71084023e-01
-5.69985569e-01 4.24151391e-01 2.70765424e-01 4.71964598e-01
4.56800789e-01 4.19124842e-01 -4.02630419e-01 1.14289522e-01
6.89314783e-01 -5.89472055e-01 -1.89428523e-01 -1.61904156e-01
-1.63393986e+00 6.90789104e-01 2.57921636e-01 7.66103208e-01
-6.68374673e-02 -1.25829369e-01 6.31022990e-01 9.90761399e-01
5.13559580e-01 8.26483488e-01 -8.90107393e-01 1.72700018e-01
-7.76371598e-01 6.67695180e-02 6.25350833e-01 8.94966960e-01
7.71093667e-01 3.35490972e-01 3.82209010e-02 1.07181513e+00
1.50194854e-01 7.87946939e-01 6.94068789e-01 -6.23632550e-01
6.62418962e-01 1.53265893e-01 -1.34813925e-02 -3.80685270e-01
-1.50577396e-01 -4.69522834e-01 -1.85660616e-01 -8.35632235e-02
9.95428681e-01 -1.81486934e-01 -6.77131414e-01 2.09244585e+00
-1.28559917e-01 -5.14681101e-01 2.55644232e-01 5.85484445e-01
3.43840271e-01 7.28421211e-01 7.22871661e-01 -1.61835119e-01
1.40618551e+00 -8.08309257e-01 -2.98423439e-01 -5.73995173e-01
9.62238729e-01 -1.30374646e+00 1.48058164e+00 1.00031428e-01
-9.24738944e-01 -8.29837322e-01 -9.96145368e-01 -4.82707411e-01
-4.64166731e-01 1.86679885e-01 5.51549792e-01 4.35516357e-01
-1.27833915e+00 1.60300836e-01 -4.53599960e-01 -4.36120450e-01
-4.90636289e-01 1.52765960e-01 -6.22176528e-01 -1.83800355e-01
-1.63439596e+00 1.55164707e+00 6.68229580e-01 -2.88576365e-01
-1.12336326e+00 -6.77431226e-01 -9.10026789e-01 -1.60042703e-01
2.93899029e-02 -3.55018675e-01 1.39894223e+00 -1.24796712e+00
-1.04839444e+00 1.23041904e+00 -8.64448622e-02 -3.58723588e-02
4.17303085e-01 -3.45521420e-02 -3.97447765e-01 -4.70940650e-01
6.47515535e-01 6.34477496e-01 5.34806371e-01 -1.22879577e+00
-8.88320029e-01 -1.93932250e-01 7.70190507e-02 6.71429992e-01
-1.76292345e-01 3.17369491e-01 -1.44720554e-01 -9.68974888e-01
-2.85981506e-01 -1.07224667e+00 9.11171064e-02 -6.92044675e-01
-1.82011694e-01 -4.96277034e-01 4.77668531e-02 -9.09909070e-01
1.16004336e+00 -2.03713822e+00 8.29628780e-02 1.50735795e-01
6.39310926e-02 -1.91036717e-03 -4.91985232e-01 4.54078645e-01
-1.81601524e-01 2.09957018e-01 1.66956022e-01 -2.82709211e-01
-2.31930986e-02 2.70975411e-01 -3.00712556e-01 5.58316588e-01
9.76290777e-02 1.03130853e+00 -1.01498687e+00 -4.84149009e-01
-1.00758940e-01 3.10037225e-01 -5.63450158e-01 -2.83892632e-01
9.30671543e-02 7.85366535e-01 5.18857315e-02 5.55656493e-01
2.24060267e-01 3.49562466e-01 5.93460441e-01 1.17240600e-01
-5.60043037e-01 7.76300132e-01 -6.17828727e-01 1.45876122e+00
-8.78095984e-01 8.10771525e-01 1.25998110e-01 -5.78303039e-01
6.80772841e-01 4.46783125e-01 -5.34561686e-02 -8.87692809e-01
-3.45450006e-02 9.82921183e-01 7.49808669e-01 -5.38948961e-02
6.87747657e-01 -6.79214120e-01 -4.09418672e-01 4.76822525e-01
2.30860636e-01 1.44657478e-01 4.40639049e-01 -1.15840070e-01
4.59851950e-01 1.49054915e-01 5.46179712e-01 -1.05370486e+00
4.93336409e-01 3.28581542e-01 5.44423163e-01 7.28952110e-01
-1.47644654e-01 4.72723208e-02 1.94981948e-01 -2.37803087e-01
-1.26883209e+00 -1.50432396e+00 -6.53997183e-01 1.71932268e+00
-2.10558385e-01 -1.36315793e-01 -3.35721403e-01 -5.05435169e-01
2.10977927e-01 8.05847883e-01 -2.65264422e-01 8.14286694e-02
-9.12316382e-01 -5.87618172e-01 7.96553016e-01 5.39782763e-01
-4.58562672e-02 -1.02433455e+00 1.94149509e-01 2.24709585e-01
-2.77058214e-01 -9.95610893e-01 -7.71359682e-01 1.84757233e-01
-7.13380456e-01 -7.92496562e-01 -6.02634370e-01 -1.27393413e+00
5.34610450e-01 1.88375324e-01 1.43564534e+00 -1.75574124e-01
4.53513324e-01 -5.69533557e-02 -3.03298496e-02 -2.56099403e-01
-8.89025092e-01 6.08950734e-01 3.83672565e-01 -6.18577182e-01
6.51267588e-01 -2.94840753e-01 -3.90367173e-02 1.96582019e-01
-4.97145236e-01 -2.73537993e-01 5.80630541e-01 7.42184937e-01
4.63232219e-01 -4.41583842e-01 8.84728789e-01 -1.01703906e+00
8.36323202e-01 -7.04196990e-01 -6.29291296e-01 2.45100051e-01
-8.59534740e-01 3.57380629e-01 8.59940171e-01 -6.50840938e-01
-9.87357914e-01 -1.94271863e-01 -9.95397493e-02 4.23485087e-03
3.64821628e-02 9.17728901e-01 -9.53572094e-02 1.25298843e-01
7.35290051e-01 1.13287434e-01 -3.61110389e-01 -6.42401993e-01
5.72357118e-01 5.48263013e-01 5.10702312e-01 -1.18827415e+00
7.38256752e-01 -2.88460441e-02 -6.82611585e-01 -6.93214118e-01
-6.74327135e-01 -3.35361570e-01 -8.33695710e-01 9.18897241e-02
7.38656044e-01 -1.47645140e+00 -1.40424877e-01 3.37939829e-01
-1.30862868e+00 -5.93377888e-01 -1.99202612e-01 8.97493601e-01
-2.61881053e-01 -3.60822491e-02 -8.98777783e-01 -5.38676754e-02
-8.28827024e-02 -1.54586780e+00 6.98584378e-01 -3.72679263e-01
-6.55211329e-01 -1.48782015e+00 4.78000253e-01 1.24713123e-01
3.07372510e-01 -3.42313200e-01 1.41587925e+00 -1.08584988e+00
-2.52528787e-01 9.25578848e-02 -2.99157917e-01 4.30050552e-01
1.74461663e-01 -1.38074070e-01 -6.64507747e-01 -5.77988386e-01
-1.41364440e-01 -4.13706988e-01 7.07242608e-01 2.72379309e-01
4.10451628e-02 -1.63621053e-01 -1.91187948e-01 6.64588809e-01
1.45268083e+00 1.68350637e-01 7.02055320e-02 4.58278060e-01
9.37690079e-01 7.25553453e-01 2.29830831e-01 -5.16588092e-01
8.05111647e-01 3.81544352e-01 -5.16933680e-01 -1.63724441e-02
-2.66375512e-01 -5.57582736e-01 9.44794774e-01 1.59198689e+00
-2.19712947e-02 5.79046942e-02 -1.23572123e+00 1.02976072e+00
-1.35269368e+00 -6.10678911e-01 -1.43994838e-01 2.35930085e+00
1.32196736e+00 1.94022626e-01 -3.67937386e-02 -3.12666625e-01
6.29027367e-01 1.71911150e-01 -3.79386812e-01 -5.73335707e-01
-3.81447285e-01 3.43450069e-01 7.71916568e-01 1.06932044e+00
-7.47788548e-01 1.69210410e+00 7.11842823e+00 8.22918892e-01
-1.34878969e+00 4.09664214e-01 2.91882306e-01 -5.45995869e-02
-5.26513278e-01 1.34513751e-01 -1.03404009e+00 1.65680036e-01
1.21673357e+00 -6.89428389e-01 6.49201870e-01 6.73018515e-01
2.28193447e-01 1.00956753e-01 -1.58237326e+00 4.24361676e-01
-7.00355321e-02 -7.24574745e-01 2.74364263e-01 9.21639428e-02
1.01268733e+00 5.77322245e-01 7.89047852e-02 7.53558576e-01
6.57832325e-01 -1.18541563e+00 1.28751540e+00 5.86195365e-02
1.29350412e+00 -6.58145368e-01 4.43412453e-01 3.09738666e-01
-1.28065538e+00 2.31410488e-01 -3.43030512e-01 -2.24824816e-01
7.56328776e-02 8.25324506e-02 -5.97344398e-01 1.88062623e-01
3.15450042e-01 5.60143709e-01 -6.88964188e-01 4.92485613e-01
-5.32632113e-01 7.34411955e-01 -3.99727747e-02 5.32450020e-01
3.94085526e-01 -5.61933853e-02 5.25302649e-01 1.39894140e+00
2.01500237e-01 -6.00173891e-01 3.83112967e-01 5.23360848e-01
-3.32713991e-01 7.27968037e-01 -9.70765531e-01 -2.57698029e-01
6.56563103e-01 7.85589337e-01 -4.32388067e-01 -3.67370963e-01
-7.97322154e-01 6.16484523e-01 7.83311427e-01 5.15215456e-01
-4.36925203e-01 -2.63446242e-01 7.55264342e-01 2.48774767e-01
-1.98517218e-01 -6.20411217e-01 -3.38940769e-01 -1.46436059e+00
-9.21296924e-02 -1.22441816e+00 1.77793711e-01 -4.08596367e-01
-1.49886966e+00 7.60789692e-01 8.82209092e-02 -8.45064223e-01
-5.18148184e-01 -8.33829641e-01 -2.15066284e-01 1.62279701e+00
-1.68614435e+00 -1.16637433e+00 4.51233327e-01 6.75295591e-01
5.79225779e-01 -4.17607635e-01 7.73760021e-01 5.34424186e-01
-3.85261089e-01 7.89698005e-01 2.78331161e-01 3.91972870e-01
1.36595166e+00 -1.21304595e+00 5.42421043e-01 9.30263102e-01
2.53967285e-01 1.23845339e+00 4.48091328e-01 -7.93620348e-01
-7.94018984e-01 -9.71748233e-01 1.66394210e+00 -7.98553586e-01
1.20063031e+00 -4.08586353e-01 -8.10741246e-01 1.47659326e+00
5.33569217e-01 -4.98262376e-01 5.36620736e-01 5.14456809e-01
-6.38817489e-01 1.00097910e-01 -4.96037632e-01 7.93075204e-01
1.01192951e+00 -1.05619860e+00 -9.07899439e-01 2.10277423e-01
8.48487496e-01 -8.60865489e-02 -1.05604184e+00 2.70129859e-01
6.52864695e-01 -5.40007055e-01 8.06677997e-01 -7.30506778e-01
3.99220943e-01 -5.81625029e-02 -1.01512574e-01 -1.83515000e+00
-4.08923894e-01 -1.33532628e-01 8.10580492e-01 1.14607859e+00
8.82384539e-01 -1.01226902e+00 -1.29185356e-02 -1.66241173e-02
-2.30241820e-01 -2.87175834e-01 -8.68504584e-01 -1.11188865e+00
1.18127930e+00 -3.67288679e-01 2.22636044e-01 1.43630290e+00
-6.82080071e-03 7.57677138e-01 2.59205736e-02 -8.17672238e-02
3.39122742e-01 -9.21834558e-02 3.85138750e-01 -1.26844013e+00
-1.39113560e-01 -6.00636780e-01 1.56527296e-01 -9.10358429e-01
8.23316872e-01 -1.73691297e+00 5.55092432e-02 -1.17664325e+00
1.60170823e-01 -8.05184245e-01 -3.07190895e-01 4.68977153e-01
-3.49750131e-01 1.60629779e-01 2.90992051e-01 5.70829928e-01
-4.94614244e-03 4.76460382e-02 1.09825540e+00 2.70921476e-02
-2.41326615e-01 -3.97363394e-01 -1.01616859e+00 7.34916627e-01
7.01887131e-01 -5.78629136e-01 -4.62441742e-01 -1.16828787e+00
2.82460421e-01 -2.40913287e-01 -2.04358652e-01 -4.94647741e-01
-7.89981782e-02 -2.86883235e-01 1.48022085e-01 2.89799631e-01
-1.82397068e-01 -4.46998179e-01 -1.40748203e-01 4.08278614e-01
-3.61173719e-01 8.54570985e-01 5.21000504e-01 -8.91007856e-02
-3.51834595e-01 -1.37519956e-01 8.20843339e-01 -2.25337937e-01
-8.33153129e-01 -7.76470229e-02 -7.28770673e-01 5.26706100e-01
7.63038874e-01 7.48763010e-02 -3.51966232e-01 2.60425601e-02
-6.47316039e-01 1.16191037e-01 7.12014496e-01 7.67976820e-01
-2.22799107e-01 -1.41654491e+00 -1.14893556e+00 3.23636264e-01
3.42080891e-01 -6.80531561e-01 -3.91929597e-01 8.11190724e-01
-5.44247925e-01 9.58184779e-01 -3.25024992e-01 -2.60804951e-01
-7.82142401e-01 4.73549068e-01 3.64313394e-01 -5.63902020e-01
2.19160710e-02 5.71997702e-01 5.40135682e-01 -9.75508213e-01
-1.36426061e-01 -3.94043177e-01 5.26957326e-02 2.42043510e-01
4.83327620e-02 1.87349230e-01 -5.33368066e-02 -1.20961225e+00
-4.09406424e-01 6.21006012e-01 -3.11080486e-01 -4.38441277e-01
7.69980252e-01 -2.51032293e-01 -2.34522358e-01 9.12660420e-01
1.11040604e+00 7.81608939e-01 -7.16310084e-01 -4.26924109e-01
1.47165850e-01 -4.87081483e-02 -1.06170729e-01 -6.78414643e-01
-6.82068467e-01 5.59608579e-01 3.62103581e-02 -2.64747918e-01
4.48483586e-01 6.62172288e-02 5.41144431e-01 2.86865979e-01
5.13685465e-01 -1.03994179e+00 -6.66996837e-01 1.15348458e+00
7.94442534e-01 -1.27087533e+00 -3.07416707e-01 -7.98901021e-02
-6.10408783e-01 5.87263048e-01 7.16678560e-01 -2.82748342e-01
5.64748704e-01 2.70667583e-01 5.61392784e-01 4.03838269e-02
-7.25138009e-01 -1.84011132e-01 3.67899150e-01 2.40264118e-01
1.15821445e+00 3.81509513e-01 -8.45176101e-01 3.81242067e-01
-8.24831009e-01 -5.10695577e-01 1.39473766e-01 5.92586339e-01
-2.00439021e-01 -1.49427295e+00 -5.16756475e-01 2.15874523e-01
-5.45034289e-01 -8.22099805e-01 -4.07114118e-01 1.33048546e+00
3.43582392e-01 7.10017264e-01 3.22928548e-01 -1.47128120e-01
2.22639024e-01 5.48283756e-01 5.80683887e-01 -1.04624248e+00
-7.92435944e-01 7.81288072e-02 2.18382597e-01 -9.61794779e-02
-4.38204594e-02 -8.54714930e-01 -8.75216305e-01 -4.54798371e-01
7.22432733e-02 1.48440957e-01 3.68294984e-01 1.09053993e+00
-1.72783643e-01 1.76185906e-01 1.52514935e-01 -7.18114913e-01
-5.94056368e-01 -1.12961686e+00 -4.24902856e-01 3.92116755e-01
2.66788036e-01 -4.31260377e-01 -4.05639887e-01 1.88108906e-02] | [10.891134262084961, 9.981413841247559] |
9c7ec19d-2ca4-4e36-9ce3-6dbf19b7e90d | clinically-inspired-multi-agent-transformers | 2210.13889 | null | https://arxiv.org/abs/2210.13889v1 | https://arxiv.org/pdf/2210.13889v1.pdf | Clinically-Inspired Multi-Agent Transformers for Disease Trajectory Forecasting from Multimodal Data | Deep neural networks are often applied to medical images to automate the problem of medical diagnosis. However, a more clinically relevant question that practitioners usually face is how to predict the future trajectory of a disease. Current methods for prognosis or disease trajectory forecasting often require domain knowledge and are complicated to apply. In this paper, we formulate the prognosis prediction problem as a one-to-many prediction problem. Inspired by a clinical decision-making process with two agents -- a radiologist and a general practitioner -- we predict prognosis with two transformer-based components that share information with each other. The first transformer in this framework aims to analyze the imaging data, and the second one leverages its internal states as inputs, also fusing them with auxiliary clinical data. The temporal nature of the problem is modeled within the transformer states, allowing us to treat the forecasting problem as a multi-task classification, for which we propose a novel loss. We show the effectiveness of our approach in predicting the development of structural knee osteoarthritis changes and forecasting Alzheimer's disease clinical status directly from raw multi-modal data. The proposed method outperforms multiple state-of-the-art baselines with respect to performance and calibration, both of which are needed for real-world applications. An open-source implementation of our method is made publicly available at \url{https://github.com/Oulu-IMEDS/CLIMATv2}. | ['Aleksei Tiulpin', 'Simo Saarakkala', 'Matthew B. Blaschko', 'Huy Hoang Nguyen'] | 2022-10-25 | null | null | null | null | ['trajectory-forecasting', 'disease-trajectory-forecasting'] | ['computer-vision', 'medical'] | [ 3.91393676e-02 -8.77433643e-02 -2.53654569e-01 -5.79385400e-01
-1.02387762e+00 -1.27400964e-01 3.93870801e-01 1.88279942e-01
-3.34643036e-01 7.80311644e-01 3.48218411e-01 -3.07281107e-01
-3.86849016e-01 -5.63317716e-01 -4.99597698e-01 -9.31866884e-01
-1.65612340e-01 9.92860436e-01 1.18709087e-01 1.33148029e-01
-2.56658554e-01 2.71833390e-01 -1.20854986e+00 6.23174131e-01
7.27525830e-01 1.46079028e+00 1.84138566e-01 4.92672771e-01
3.92980754e-01 1.23290193e+00 -1.11278757e-01 -3.05717587e-01
-6.82524815e-02 -1.40113622e-01 -9.15434301e-01 -4.15966809e-02
-1.07502285e-03 -3.84316295e-01 -2.91913480e-01 6.29820883e-01
5.64386547e-01 -3.83187801e-01 8.72719049e-01 -1.13832712e+00
-2.91888624e-01 1.80818439e-01 -2.09442601e-01 3.19520980e-01
-9.07374695e-02 7.30638206e-02 9.14580762e-01 -5.96269250e-01
6.09359205e-01 8.43633711e-01 8.15457046e-01 5.22817969e-01
-1.10646212e+00 -3.67155433e-01 1.35758817e-01 5.35241008e-01
-1.10828602e+00 -4.03407604e-01 6.67252064e-01 -9.33866501e-01
4.18457389e-01 2.23124325e-01 6.74706697e-01 1.51403046e+00
6.10060155e-01 8.97301614e-01 1.17336273e+00 1.71460375e-01
2.48446867e-01 -1.26011550e-01 3.88363749e-01 5.24643719e-01
-2.13142738e-01 2.05584183e-01 -2.99158096e-01 -4.44733411e-01
5.31527579e-01 4.06183064e-01 -3.52686137e-01 -2.33283684e-01
-1.48965466e+00 7.73359179e-01 5.03563643e-01 2.19998240e-01
-8.28989029e-01 1.67040825e-01 4.50224996e-01 3.28489840e-01
7.27874517e-01 -6.62528053e-02 -5.98705649e-01 7.51700550e-02
-1.00220859e+00 3.71801078e-01 5.83232045e-01 2.39195466e-01
-3.48964147e-03 -4.29317683e-01 -1.95113972e-01 8.57348204e-01
3.39014977e-01 3.52040619e-01 8.28993678e-01 -7.46322453e-01
5.25472406e-03 5.28764307e-01 1.58137590e-01 -6.46096110e-01
-7.72126198e-01 -8.18106472e-01 -1.18685424e+00 1.30287167e-02
5.03847241e-01 -1.44643545e-01 -9.42872465e-01 1.60949016e+00
4.44616586e-01 2.91225195e-01 -9.90209281e-02 9.44989324e-01
4.71170574e-01 4.09728438e-01 2.36929879e-01 -2.44006038e-01
1.70543182e+00 -8.49518716e-01 -4.96850431e-01 -2.54433155e-01
8.15477550e-01 -3.37903351e-01 6.84424758e-01 5.32612562e-01
-9.49197710e-01 -2.05691814e-01 -5.64823449e-01 4.17976715e-02
-1.34288222e-01 4.93201733e-01 4.71538782e-01 -7.52256438e-02
-1.04696667e+00 7.79501438e-01 -1.23904777e+00 -2.23204881e-01
4.66960937e-01 3.09467703e-01 -2.61813074e-01 -1.85491726e-01
-1.37324119e+00 1.01149797e+00 6.20120130e-02 2.40278304e-01
-1.02330446e+00 -8.37789714e-01 -3.35637778e-01 -9.26424190e-02
2.43855968e-01 -1.23759174e+00 1.56304216e+00 -8.06691706e-01
-9.67185199e-01 8.91123712e-01 -1.81847036e-01 -4.67103153e-01
8.79538894e-01 -3.52598548e-01 -4.71860677e-01 3.57843079e-02
2.87675917e-01 3.22668731e-01 6.72375083e-01 -9.17652190e-01
-7.81184256e-01 -8.61723721e-01 -4.17430639e-01 6.15941919e-02
-2.66351312e-01 -2.83554997e-02 -2.60706246e-01 -8.10533643e-01
-5.34555912e-02 -1.10657287e+00 -4.99902546e-01 2.48610392e-01
-4.51606750e-01 -3.05952460e-01 7.05054045e-01 -8.91582429e-01
1.16440725e+00 -1.98085010e+00 3.57573569e-01 -5.93862459e-02
6.29760385e-01 -5.51899709e-03 1.29730582e-01 3.62370312e-01
-3.23688895e-01 -2.64067382e-01 -1.70965850e-01 -5.22620797e-01
-3.23728859e-01 1.55104801e-01 -1.01228185e-01 5.65692782e-01
4.00349796e-02 1.01216698e+00 -7.57993519e-01 -3.77280205e-01
-1.93054363e-01 6.55069947e-01 -3.02624762e-01 2.87364632e-01
-2.51828194e-01 8.50735247e-01 -7.49854803e-01 7.32240856e-01
1.15946352e-01 -1.03187406e+00 2.33637363e-01 -3.20353687e-01
2.17802107e-01 2.46627674e-01 -7.30775237e-01 1.52015293e+00
-2.86717951e-01 2.62300950e-02 -7.84800425e-02 -1.39543390e+00
4.45367813e-01 7.22773254e-01 1.02032852e+00 -4.90076959e-01
1.67066634e-01 4.39786822e-01 6.17280900e-02 -7.19396234e-01
-2.91135728e-01 -4.01920825e-01 6.20567314e-02 5.29541671e-01
-1.16960302e-01 5.66742599e-01 -5.08373119e-02 5.18209822e-02
1.44616961e+00 -7.97317028e-02 2.18755022e-01 -3.24675925e-02
4.45580989e-01 1.34915397e-01 8.24858487e-01 4.32414681e-01
-3.14894795e-01 3.93657923e-01 6.35692298e-01 -8.11681032e-01
-9.90115225e-01 -1.05348790e+00 -2.61551976e-01 6.80580795e-01
-2.89357305e-01 1.02422778e-02 -4.22696024e-01 -7.11278200e-01
1.38838708e-01 3.81222367e-01 -8.66278470e-01 -2.19076008e-01
-5.65779269e-01 -1.08000684e+00 2.28275955e-01 6.32250786e-01
-2.87635401e-02 -9.36655521e-01 -7.36139357e-01 4.32977349e-01
-4.42063749e-01 -9.79625702e-01 -2.51455843e-01 1.52698785e-01
-1.11448824e+00 -1.10988748e+00 -1.25746787e+00 -5.02382934e-01
3.66194993e-01 -2.88654149e-01 1.16488278e+00 1.44379716e-02
-3.60575706e-01 5.34039974e-01 -2.54604906e-01 -2.87315071e-01
-3.97603899e-01 2.72001978e-03 1.65845007e-01 3.25234503e-01
8.52335021e-02 -7.57918179e-01 -1.17087066e+00 2.24596277e-01
-7.43375897e-01 3.40779126e-01 8.46599221e-01 1.02943504e+00
8.34662676e-01 -2.20215574e-01 6.95383191e-01 -8.67710292e-01
4.83360916e-01 -8.48213792e-01 -1.33058161e-01 3.35739702e-01
-7.29196370e-01 1.00524127e-01 3.91334653e-01 -5.62938809e-01
-7.72808075e-01 2.85256207e-01 -1.08433463e-01 -6.14527524e-01
-2.62553006e-01 9.88452554e-01 2.08704606e-01 3.91054124e-01
2.89900362e-01 3.51002753e-01 2.99114048e-01 -6.24494433e-01
7.39051076e-03 5.50185561e-01 5.58198154e-01 -4.59428549e-01
2.83287257e-01 5.31225383e-01 1.12637602e-01 -3.35783541e-01
-1.28478265e+00 -5.02235293e-01 -4.66797084e-01 -4.02888238e-01
8.88660312e-01 -1.04154646e+00 -6.48139298e-01 7.97603667e-01
-9.99618530e-01 -4.56735343e-01 -1.43500492e-01 4.75127012e-01
-6.20650113e-01 -5.66234663e-02 -8.58285367e-01 -3.97406906e-01
-5.20250499e-01 -1.35258007e+00 1.28093708e+00 -3.21502805e-01
-5.24117053e-02 -1.22249067e+00 2.56261379e-01 4.96031135e-01
4.08568442e-01 4.46620435e-01 1.13401341e+00 -8.25609565e-01
-3.48003983e-01 -2.87435770e-01 -1.77727133e-01 2.56972849e-01
8.69791061e-02 -5.93610942e-01 -7.49990463e-01 -2.19306320e-01
3.22524071e-01 -3.63879830e-01 8.80419374e-01 7.21238732e-01
1.34430015e+00 -2.56091028e-01 -6.80028796e-01 4.05756891e-01
1.16373217e+00 1.53374478e-01 3.90248299e-01 3.43691111e-01
5.70860386e-01 6.61233902e-01 5.10041118e-01 5.05497038e-01
8.42962623e-01 7.38995492e-01 5.66986978e-01 -1.60774530e-03
-4.74026281e-04 2.66764820e-01 1.54430836e-01 9.08212483e-01
-1.85639262e-01 -1.73653603e-01 -1.54092383e+00 6.77822113e-01
-2.20079136e+00 -6.86054528e-01 -2.17466071e-01 1.90305507e+00
9.43515420e-01 -9.82425436e-02 2.18433037e-01 5.17462716e-02
4.26804274e-01 -9.83274430e-02 -7.96737909e-01 3.95965368e-01
2.64003128e-01 -1.88597068e-01 1.35958359e-01 1.30615100e-01
-1.23668587e+00 4.22519028e-01 5.50461531e+00 5.07239819e-01
-1.44615912e+00 4.53439116e-01 1.10345984e+00 -3.03320825e-01
9.91268903e-02 -3.03243667e-01 -5.82118809e-01 5.63683569e-01
1.21313620e+00 2.26469375e-02 1.01935053e-02 5.77441692e-01
4.39540952e-01 5.94046153e-02 -1.44731092e+00 7.07955539e-01
-3.30244660e-01 -1.25529695e+00 -1.18674546e-01 1.51868463e-01
2.03182146e-01 3.54165792e-01 2.37838566e-01 3.93626131e-02
3.22997630e-01 -1.00486708e+00 6.82944059e-01 1.09051275e+00
7.39620030e-01 -1.26776651e-01 7.33047724e-01 4.52307910e-01
-9.64723170e-01 -2.51737326e-01 3.22396189e-01 1.30973637e-01
4.80004609e-01 9.11328852e-01 -6.33242667e-01 5.80414653e-01
7.72046447e-01 1.04750907e+00 -4.43247527e-01 1.03492367e+00
1.90969985e-02 7.45485246e-01 -6.89751655e-02 4.03540999e-01
1.14764072e-01 -7.28301704e-02 3.40661317e-01 7.44406462e-01
5.06680429e-01 1.64696604e-01 2.39739627e-01 7.19543755e-01
1.70635745e-01 -1.06263518e-01 -2.45237514e-01 4.70878731e-05
1.04624644e-01 1.19262028e+00 -4.58601952e-01 -5.10510683e-01
-4.16915119e-01 9.00553524e-01 3.05058688e-01 2.88355261e-01
-8.39226544e-01 3.38670939e-01 5.33448935e-01 4.24492896e-01
1.38359353e-01 1.01445429e-01 -2.67368644e-01 -1.29753900e+00
1.81206480e-01 -1.00560260e+00 5.81062675e-01 -6.91306174e-01
-1.60657513e+00 7.50124812e-01 -2.30622292e-01 -1.35893166e+00
-4.18241411e-01 -7.25955486e-01 -4.33651984e-01 6.56487107e-01
-1.56630814e+00 -1.49761212e+00 -2.40889534e-01 7.31124282e-01
2.79974163e-01 -1.11609690e-01 8.25597227e-01 6.07919514e-01
-6.25001550e-01 1.26377493e-01 2.44829819e-01 9.09806117e-02
7.51273692e-01 -1.18918324e+00 -1.42653240e-04 3.90535116e-01
-3.05115551e-01 2.08217695e-01 4.64259297e-01 -7.41182327e-01
-1.07387972e+00 -1.32742453e+00 1.08793986e+00 -3.69786084e-01
1.02773798e+00 3.71032767e-02 -9.40387607e-01 7.31650114e-01
-4.09503490e-01 3.38368535e-01 7.58309960e-01 7.38987746e-03
-1.90294683e-01 -1.17801234e-01 -8.22342396e-01 2.55425870e-01
7.69563437e-01 -3.45255643e-01 -4.51745927e-01 6.63175404e-01
3.95606518e-01 -1.99566826e-01 -1.32065856e+00 6.94071651e-01
7.22036302e-01 -7.52863884e-01 1.11618984e+00 -8.43009353e-01
8.03113937e-01 6.98139220e-02 -7.79617280e-02 -1.31069839e+00
-3.02854657e-01 3.15013677e-02 -3.03377301e-01 5.51892042e-01
3.26004803e-01 -7.00985968e-01 6.86840594e-01 4.33595687e-01
-5.16396947e-02 -1.41166377e+00 -9.81701255e-01 -6.46935284e-01
2.33327582e-01 -3.07340950e-01 1.11016922e-01 9.71096516e-01
-3.14461768e-01 2.47813255e-01 -5.11197627e-01 3.60277325e-01
6.81790173e-01 1.45641148e-01 7.43684247e-02 -1.72236228e+00
-4.30182546e-01 -3.51602376e-01 -3.14338207e-01 -6.47600055e-01
-1.41389121e-03 -8.77900064e-01 -1.55307651e-01 -1.63413692e+00
4.48545992e-01 -5.50738335e-01 -6.38808429e-01 5.97906888e-01
-1.50259614e-01 -1.85141638e-02 -2.30590943e-02 7.61694849e-01
-4.51977581e-01 6.37828946e-01 1.15850174e+00 -2.40372658e-01
9.72602591e-02 3.75768214e-01 -4.62021410e-01 7.44741023e-01
7.35205948e-01 -6.88663661e-01 -3.13094109e-01 -4.43510860e-01
2.27035791e-01 5.67361236e-01 9.90442336e-01 -1.09600925e+00
3.30783725e-01 -2.36773714e-02 2.67430902e-01 -4.51386869e-01
4.76176143e-01 -6.62741303e-01 4.22737777e-01 9.10976350e-01
-2.45515704e-01 2.02324972e-01 -2.93890446e-01 6.91594064e-01
-4.03264105e-01 1.18700370e-01 6.82389021e-01 -2.31318191e-01
-3.44688833e-01 7.64229536e-01 -3.81828606e-01 -2.83397716e-02
1.08436835e+00 3.57259065e-01 -2.63541192e-01 -3.33489239e-01
-1.30704069e+00 3.88931155e-01 -3.97337303e-02 3.10464680e-01
5.07287443e-01 -1.32005835e+00 -9.45112348e-01 -3.00303221e-01
6.02376200e-02 -5.38534559e-02 5.11818349e-01 1.61257803e+00
-1.39062315e-01 3.91377568e-01 -8.16626996e-02 -8.44023764e-01
-1.17222309e+00 5.36691666e-01 7.60581017e-01 -9.03264523e-01
-7.00864136e-01 6.71183348e-01 3.59595597e-01 -2.09579915e-01
2.82375276e-01 -2.56398648e-01 -2.85686433e-01 3.07193190e-01
5.68044901e-01 8.34154263e-02 1.89046934e-01 -4.91447181e-01
-4.20815557e-01 3.22153181e-01 -4.21586096e-01 5.04440069e-02
1.82262850e+00 3.16310115e-02 -1.90326422e-01 7.92358160e-01
1.06212592e+00 -7.97517419e-01 -8.21535587e-01 -5.05129158e-01
1.63867235e-01 2.45486975e-01 3.67182136e-01 -1.15665948e+00
-1.22860968e+00 9.54845130e-01 9.67736423e-01 1.64404139e-01
1.16528869e+00 3.36049616e-01 9.58912075e-01 1.10845357e-01
2.39318743e-01 -6.78529501e-01 1.07627682e-01 2.94024467e-01
1.08590460e+00 -1.37256706e+00 -1.38439924e-01 -7.37095028e-02
-7.30029583e-01 9.64228749e-01 3.07481289e-01 -7.00373156e-03
1.01179016e+00 1.08990766e-01 2.97233969e-01 -4.44245428e-01
-1.18482244e+00 1.24621980e-01 5.42659998e-01 2.18580201e-01
3.65695924e-01 2.07342774e-01 -2.45773256e-01 9.89715576e-01
1.78649038e-01 5.23914456e-01 1.62210748e-01 8.18848610e-01
-8.40196013e-02 -1.19870818e+00 -2.69457787e-01 9.11562800e-01
-6.03203714e-01 -8.60381722e-02 1.67532079e-02 3.75746161e-01
4.57622074e-02 4.15934116e-01 -5.92661165e-02 -3.93084824e-01
1.61338672e-01 2.51332283e-01 7.67253414e-02 -3.77629280e-01
-4.07878995e-01 2.14818344e-01 8.98593068e-02 -8.20511043e-01
-5.23354650e-01 -1.01022828e+00 -8.54630589e-01 3.75623889e-02
1.82316110e-01 -1.98348939e-01 4.36011642e-01 1.01224053e+00
5.91869652e-01 9.24747765e-01 4.03928638e-01 -7.12210834e-01
-9.39283252e-01 -9.25467432e-01 -5.13524055e-01 2.96276122e-01
5.91026068e-01 -8.79130840e-01 -1.83244616e-01 2.55494624e-01] | [14.902478218078613, -1.9632627964019775] |
01b67603-92cc-47c2-9d95-90b34b5067a3 | mect-multi-metadata-embedding-based-cross | 2107.05418 | null | https://arxiv.org/abs/2107.05418v1 | https://arxiv.org/pdf/2107.05418v1.pdf | MECT: Multi-Metadata Embedding based Cross-Transformer for Chinese Named Entity Recognition | Recently, word enhancement has become very popular for Chinese Named Entity Recognition (NER), reducing segmentation errors and increasing the semantic and boundary information of Chinese words. However, these methods tend to ignore the information of the Chinese character structure after integrating the lexical information. Chinese characters have evolved from pictographs since ancient times, and their structure often reflects more information about the characters. This paper presents a novel Multi-metadata Embedding based Cross-Transformer (MECT) to improve the performance of Chinese NER by fusing the structural information of Chinese characters. Specifically, we use multi-metadata embedding in a two-stream Transformer to integrate Chinese character features with the radical-level embedding. With the structural characteristics of Chinese characters, MECT can better capture the semantic information of Chinese characters for NER. The experimental results obtained on several well-known benchmarking datasets demonstrate the merits and superiority of the proposed MECT method.\footnote{The source code of the proposed method is publicly available at https://github.com/CoderMusou/MECT4CNER. | ['ZhenHua Feng', 'Xiaoning Song', 'Shuang Wu'] | 2021-07-12 | null | https://aclanthology.org/2021.acl-long.121 | https://aclanthology.org/2021.acl-long.121.pdf | acl-2021-5 | ['chinese-named-entity-recognition'] | ['natural-language-processing'] | [-1.72461197e-01 -5.27267933e-01 -6.46205693e-02 -2.53613591e-01
-5.64860761e-01 -7.03920007e-01 2.93289393e-01 1.57205448e-01
-8.92744303e-01 4.30264264e-01 6.38434827e-01 -1.05892994e-01
2.54179329e-01 -7.94051468e-01 -1.79915264e-01 -6.74309790e-01
4.96808648e-01 -7.54324272e-02 1.88437372e-01 -1.24164686e-01
4.18438166e-01 2.31027260e-01 -9.44051087e-01 2.02119291e-01
1.41047108e+00 7.25929320e-01 5.03119230e-01 3.74495387e-01
-7.98736453e-01 4.32576388e-01 -5.80677807e-01 -7.10040629e-01
-3.08980495e-02 -4.39529538e-01 -6.70698524e-01 -2.43625894e-01
-7.03358278e-02 2.05469970e-02 -4.18263882e-01 1.41916549e+00
6.87850118e-01 -3.57588939e-02 2.01140270e-01 -7.12032199e-01
-8.85704935e-01 1.02913785e+00 -3.34758133e-01 5.49479760e-02
4.90735509e-02 -1.75750837e-01 1.08927369e+00 -9.62127745e-01
5.68981886e-01 1.02840221e+00 6.32104933e-01 6.62965834e-01
-5.50875247e-01 -7.65292287e-01 1.57347113e-01 4.28348601e-01
-1.55168295e+00 -1.07023589e-01 5.94467163e-01 -5.74697480e-02
6.14683211e-01 3.91831845e-01 5.51293850e-01 7.94276178e-01
-1.13023072e-01 1.19147921e+00 9.82444823e-01 -3.51488322e-01
-1.71550885e-01 2.73272067e-01 3.04746062e-01 4.43734854e-01
2.66641825e-01 -3.01678926e-01 -1.47910416e-01 1.80898860e-01
5.50243735e-01 1.81154534e-01 -3.26751113e-01 2.95087606e-01
-1.31122446e+00 6.28669024e-01 3.77114892e-01 8.55720878e-01
-2.42326006e-01 -1.28708640e-02 5.37274957e-01 -1.66081890e-01
4.62654501e-01 2.73251891e-01 -5.19308746e-01 -5.24499714e-01
-7.57899702e-01 -3.11287910e-01 6.24129355e-01 1.20866334e+00
7.51682520e-01 1.86409026e-01 -1.15252227e-01 9.22008991e-01
1.69972241e-01 5.11362314e-01 8.50264013e-01 -6.01530313e-01
5.59225917e-01 6.37486517e-01 -1.85490817e-01 -8.06838155e-01
-1.10751070e-01 -3.42286617e-01 -8.25302958e-01 -5.82823157e-01
5.59914857e-02 -5.05484521e-01 -7.94907033e-01 1.39267421e+00
2.76912451e-01 2.10553303e-01 2.01061040e-01 8.12330663e-01
9.24147844e-01 1.13300729e+00 2.38563284e-01 1.45504490e-01
1.54263163e+00 -9.61286306e-01 -9.54032063e-01 -7.12931156e-02
8.35941195e-01 -1.08190680e+00 9.63463664e-01 -6.31818846e-02
-5.69519162e-01 -5.12874305e-01 -8.58158231e-01 -2.23841742e-01
-6.62083685e-01 6.35395586e-01 4.02729660e-01 8.54966879e-01
-4.99659330e-01 4.44590211e-01 -7.18432248e-01 -2.16259271e-01
3.01820606e-01 8.26967880e-02 -2.59215504e-01 -2.51061559e-01
-1.46221387e+00 6.30932331e-01 8.61681879e-01 5.00098050e-01
-1.19546466e-01 -6.46744668e-01 -9.07426715e-01 1.96984425e-01
3.92874569e-01 3.38259675e-02 8.71350884e-01 -7.97918797e-01
-1.31960237e+00 3.33957225e-01 -2.73948610e-01 7.26181939e-02
3.08942735e-01 -5.47361493e-01 -7.51198232e-01 1.42530128e-01
2.48156548e-01 5.58879077e-01 2.90975988e-01 -9.06166852e-01
-6.66727126e-01 -1.05997719e-01 -3.52833897e-01 2.07749918e-01
-9.19601858e-01 2.95693994e-01 -9.01086032e-01 -9.81895983e-01
-1.49668813e-01 -6.71775043e-01 -2.47477531e-01 -3.94113213e-01
-6.12110972e-01 -2.38665760e-01 9.14079428e-01 -8.69655192e-01
1.58886039e+00 -2.44761658e+00 -1.49998069e-01 -6.49378449e-02
-2.42332611e-02 6.48204386e-01 -1.97865203e-01 6.37646139e-01
4.91391011e-02 5.67936659e-01 -3.75462353e-01 -1.24196060e-01
1.12734139e-01 1.19522981e-01 -3.45796347e-03 2.01295301e-01
4.19852704e-01 9.63579237e-01 -8.62669408e-01 -8.21922302e-01
2.21874982e-01 5.96884847e-01 -1.67150110e-01 -3.59221199e-03
1.00371204e-01 2.25997180e-01 -8.67552161e-01 5.17331302e-01
9.47935402e-01 3.44751514e-02 1.38296530e-01 -2.94153959e-01
-4.98830557e-01 2.94520319e-01 -1.29482496e+00 1.48264229e+00
-2.75023252e-01 4.21762019e-01 -1.72797710e-01 -5.84957540e-01
8.98286879e-01 2.78110653e-01 3.01348358e-01 -7.41733551e-01
3.33801419e-01 3.69414598e-01 -1.60253160e-02 -3.86522800e-01
7.39650607e-01 1.79088619e-02 -3.08890849e-01 1.33501783e-01
-9.75406766e-02 9.11362469e-02 4.20688987e-01 2.17129424e-01
8.29604566e-01 8.59960839e-02 2.62170643e-01 -2.60269165e-01
7.46654093e-01 9.57109332e-02 1.10699224e+00 2.49216497e-01
-9.09660161e-02 5.73752224e-01 3.20405960e-01 -4.50472198e-02
-1.15032935e+00 -7.17549086e-01 -3.36609960e-01 7.19962299e-01
2.17461199e-01 -6.47414505e-01 -9.13262308e-01 -6.85921967e-01
-3.36209565e-01 7.22853899e-01 -4.48274344e-01 2.07110178e-02
-8.41958940e-01 -7.60239422e-01 9.94180202e-01 8.65842164e-01
8.31777334e-01 -1.06061149e+00 -9.05358344e-02 3.18374276e-01
-3.74474764e-01 -1.25742793e+00 -1.01639795e+00 -8.82025734e-02
-7.94666409e-01 -6.96405709e-01 -1.06888056e+00 -1.18083966e+00
5.69167197e-01 1.04172759e-01 4.77146596e-01 2.42337063e-01
-2.45505869e-01 9.88698676e-02 -9.18994427e-01 -2.74125934e-01
-1.09846964e-01 3.82081181e-01 -2.67432213e-01 1.49611607e-01
6.41754687e-01 6.01870799e-03 -3.99221689e-01 2.98458815e-01
-1.12645411e+00 6.19907454e-02 8.76240432e-01 7.54017889e-01
5.08150220e-01 2.95947611e-01 1.56853572e-01 -1.04152298e+00
4.57167953e-01 -3.98201257e-01 -3.87995392e-01 2.90986180e-01
-5.18282473e-01 2.04477251e-01 8.55964899e-01 -5.08391678e-01
-1.39909589e+00 -1.09088078e-01 -4.43922281e-01 -8.71577933e-02
-1.65257439e-01 6.60821557e-01 -7.62217045e-01 2.98475862e-01
-1.26066312e-01 6.71962738e-01 -5.10109603e-01 -9.18580115e-01
3.41790438e-01 8.75701070e-01 3.78479272e-01 -6.50621891e-01
6.53957903e-01 2.71024913e-01 -5.21937907e-01 -9.28348660e-01
-5.69649220e-01 -5.96864223e-01 -7.24805713e-01 1.35225922e-01
1.15641320e+00 -9.43780541e-01 -3.43044609e-01 7.72174716e-01
-1.25135064e+00 5.73330671e-02 -1.40178502e-01 7.15354681e-01
2.74403006e-01 8.96427870e-01 -1.05710065e+00 -5.10451794e-01
-4.96398002e-01 -9.94461060e-01 7.24342644e-01 6.91535711e-01
2.23564729e-01 -1.09105086e+00 -1.06709339e-02 1.41213670e-01
2.96230853e-01 -9.69828591e-02 8.69749606e-01 -8.07824552e-01
-3.93317282e-01 -3.88224721e-01 -4.63327616e-01 4.97580945e-01
2.74958909e-01 1.52795739e-03 -7.25198984e-01 -9.79736224e-02
-2.38938317e-01 2.26857185e-01 8.95689726e-01 -1.10396400e-01
9.44381177e-01 -1.93377271e-01 -2.19887510e-01 5.54339707e-01
1.72706461e+00 2.91732639e-01 7.53619730e-01 5.00922680e-01
1.26160431e+00 4.43891168e-01 6.20637834e-01 4.46405411e-01
5.13857186e-01 2.63169497e-01 1.01814285e-01 -1.10654742e-01
-6.11366108e-02 -2.97419816e-01 5.43671548e-01 1.60671628e+00
-4.65914495e-02 -3.03858608e-01 -9.34056759e-01 7.80556142e-01
-1.47904456e+00 -8.16806257e-01 -3.69476497e-01 1.71442938e+00
1.00542176e+00 5.20596504e-02 -3.65658224e-01 -6.46787360e-02
1.18972778e+00 1.54077470e-01 -3.88384581e-01 -2.27979302e-01
-4.88715857e-01 2.63075411e-01 7.55584776e-01 9.09571946e-02
-1.07328331e+00 1.27795029e+00 4.80597162e+00 1.29604793e+00
-1.09015286e+00 1.77377671e-01 2.09503248e-01 5.34275889e-01
-5.48606157e-01 2.46716604e-01 -1.17891908e+00 8.01198542e-01
7.68789947e-01 -2.99150527e-01 -1.47436699e-02 6.49872363e-01
5.13741337e-02 1.93196565e-01 -4.02354479e-01 6.29867852e-01
3.19886543e-02 -1.21625292e+00 1.32818386e-01 3.28811556e-02
7.11414039e-01 2.81665940e-02 -5.10567129e-02 1.23592488e-01
1.27015650e-01 -6.36940241e-01 7.53236830e-01 4.50878501e-01
7.88131416e-01 -8.40994596e-01 1.14017427e+00 1.08467728e-01
-1.74066305e+00 5.12539893e-02 -5.09882510e-01 3.90786141e-01
2.06482574e-01 5.94291747e-01 -2.11495087e-01 8.59517574e-01
8.16941559e-01 1.11127698e+00 -6.86748505e-01 1.22483063e+00
-5.02422392e-01 8.51401985e-01 -1.73269659e-01 -4.28435147e-01
5.13014734e-01 -4.59300309e-01 4.51493859e-01 1.70985436e+00
3.51684213e-01 1.81203663e-01 -1.08566493e-01 7.35048175e-01
-1.09384634e-01 6.26544297e-01 -9.60197821e-02 -6.49096966e-01
6.57696128e-01 1.34869480e+00 -8.29552114e-01 -3.45017165e-01
-5.97033024e-01 1.13422084e+00 1.07198037e-01 1.65994719e-01
-8.53586555e-01 -1.14226615e+00 5.98186016e-01 -4.06161904e-01
7.88168013e-01 -3.62964272e-01 -2.92966962e-01 -1.42924571e+00
1.35841087e-01 -7.88354158e-01 3.82408500e-01 -4.49095517e-01
-1.16032565e+00 7.85704017e-01 -3.64377379e-01 -1.43342614e+00
4.76709127e-01 -6.54459774e-01 -7.79724598e-01 8.82187188e-01
-1.53863072e+00 -1.15839481e+00 -9.16828737e-02 3.51944089e-01
5.90944350e-01 1.08435340e-01 6.87050402e-01 5.74926019e-01
-1.07062078e+00 7.37150669e-01 5.18009961e-01 9.60162282e-01
6.62800074e-01 -1.32579350e+00 4.87058640e-01 1.29383147e+00
1.13857590e-01 8.32420588e-01 1.82601646e-01 -8.35405350e-01
-1.27197039e+00 -1.04489875e+00 1.12265003e+00 -1.69040382e-01
6.36236668e-01 -3.50214422e-01 -9.60990071e-01 4.35489863e-01
3.95838588e-01 -2.21530944e-01 8.38070452e-01 -2.85907865e-01
-2.65576094e-01 -9.58678126e-02 -6.13062799e-01 6.06202602e-01
5.53053141e-01 -4.59025800e-01 -6.96519375e-01 -2.57603854e-01
9.32104468e-01 -1.80527702e-01 -1.00864947e+00 2.20985472e-01
4.07711804e-01 -5.07923782e-01 5.90790808e-01 -2.58849353e-01
2.34684169e-01 -5.58042407e-01 -8.29098672e-02 -1.14297080e+00
-2.98474073e-01 -3.90394777e-01 4.20433253e-01 1.73045456e+00
5.29078543e-01 -5.05928934e-01 5.71527898e-01 3.15636754e-01
-3.80266368e-01 -4.83835369e-01 -7.78572917e-01 -7.36920536e-01
2.93461114e-01 -4.19777721e-01 7.95933902e-01 1.04769993e+00
-7.88042098e-02 2.16579974e-01 -3.12289417e-01 7.68261328e-02
3.62693727e-01 -3.78607251e-02 2.11200297e-01 -8.92665327e-01
1.52517781e-01 -4.66719300e-01 -3.92444551e-01 -1.15857863e+00
1.19486682e-01 -8.36754382e-01 5.24772890e-02 -1.54452372e+00
3.22580695e-01 -4.90467072e-01 -4.60516900e-01 4.79988068e-01
-5.80477297e-01 1.54542970e-02 4.95972604e-01 1.42286420e-01
-6.04673326e-01 7.42294788e-01 1.31208706e+00 -6.97706044e-02
6.42840043e-02 -3.08924079e-01 -5.98552167e-01 5.15771627e-01
9.62840974e-01 -6.23347700e-01 1.16344683e-01 -8.84996295e-01
8.65709707e-02 -4.43663657e-01 -8.49654377e-02 -8.74362886e-01
5.13986647e-01 -4.31078114e-02 5.08086085e-01 -7.10973680e-01
1.37653261e-01 -8.50419343e-01 -1.39576271e-01 5.17563641e-01
-2.60749850e-02 3.08229297e-01 3.81598651e-01 4.40966010e-01
-4.34302896e-01 -4.56124544e-01 6.05100572e-01 -1.50218472e-01
-1.16631866e+00 3.20634961e-01 -4.83444810e-01 3.28583896e-01
7.50652432e-01 -2.88581938e-01 -3.03644329e-01 7.65043274e-02
-2.67903298e-01 3.20977271e-01 4.29258734e-01 5.20224154e-01
7.88636804e-01 -1.37527013e+00 -7.54561961e-01 -3.09164040e-02
1.17175229e-01 -2.19934672e-01 3.88799608e-01 7.33257532e-01
-8.82979512e-01 3.14451277e-01 -1.58834621e-01 -1.60492286e-01
-1.09462380e+00 4.39078718e-01 1.07930332e-01 -8.78586918e-02
-5.94073951e-01 7.07475960e-01 -6.85570687e-02 -5.32029569e-01
-1.16422646e-01 -1.95865899e-01 -5.52413702e-01 9.01629627e-02
6.21280313e-01 5.37842810e-01 -2.43548468e-01 -8.44202936e-01
-4.08510625e-01 8.51106226e-01 -3.01963955e-01 -1.26312628e-01
1.29509902e+00 -2.28880957e-01 -4.00037736e-01 3.52942020e-01
1.29233658e+00 4.99568224e-01 -7.03505933e-01 -4.07701045e-01
3.51002306e-01 -3.74016792e-01 -2.70627141e-02 -4.42406654e-01
-1.21030593e+00 1.06345785e+00 3.68756175e-01 -2.26609483e-01
1.11219764e+00 -2.67205089e-01 1.41758585e+00 2.30351448e-01
4.79808114e-02 -1.36225104e+00 -2.49002665e-01 7.03437328e-01
9.31127742e-02 -9.78704333e-01 -2.01288790e-01 -5.65336764e-01
-1.01538932e+00 1.26616049e+00 6.14415407e-01 5.17853536e-02
7.90829957e-01 1.93501338e-01 3.45190883e-01 1.56035289e-01
-2.01115549e-01 -4.50012922e-01 2.13945702e-01 2.82495350e-01
4.69173998e-01 1.30864203e-01 -7.46744812e-01 8.29517007e-01
-8.78603086e-02 -4.56223607e-01 5.50061047e-01 9.63765442e-01
-3.72919649e-01 -1.55477583e+00 -1.73989594e-01 1.81228191e-01
-7.46333778e-01 -6.81559384e-01 -3.14069778e-01 6.74962103e-01
3.70415002e-01 9.03515697e-01 -6.34390265e-02 -5.06653368e-01
1.63450599e-01 -9.65346210e-03 -7.97251835e-02 -6.05651319e-01
-5.60323775e-01 3.54258001e-01 -1.29283577e-01 -2.40713090e-01
-3.17579389e-01 -8.00103903e-01 -1.55125511e+00 -3.08094919e-01
-5.65000534e-01 6.62678599e-01 7.09943414e-01 7.51376271e-01
3.83380949e-01 3.66443753e-01 7.40872443e-01 -1.75864190e-01
-2.20651910e-01 -9.58079100e-01 -5.68330824e-01 2.57902980e-01
-2.20652714e-01 -1.27937958e-01 -2.73178935e-01 -7.04941452e-02] | [9.891426086425781, 9.853864669799805] |
9a79b216-d3e8-4e9d-8f10-852101c61efb | interactive-object-segmentation-in-3d-point | 2204.07183 | null | https://arxiv.org/abs/2204.07183v2 | https://arxiv.org/pdf/2204.07183v2.pdf | Interactive Object Segmentation in 3D Point Clouds | We propose an interactive approach for 3D instance segmentation, where users can iteratively collaborate with a deep learning model to segment objects in a 3D point cloud directly. Current methods for 3D instance segmentation are generally trained in a fully-supervised fashion, which requires large amounts of costly training labels, and does not generalize well to classes unseen during training. Few works have attempted to obtain 3D segmentation masks using human interactions. Existing methods rely on user feedback in the 2D image domain. As a consequence, users are required to constantly switch between 2D images and 3D representations, and custom architectures are employed to combine multiple input modalities. Therefore, integration with existing standard 3D models is not straightforward. The core idea of this work is to enable users to interact directly with 3D point clouds by clicking on desired 3D objects of interest~(or their background) to interactively segment the scene in an open-world setting. Specifically, our method does not require training data from any target domain, and can adapt to new environments where no appropriate training sets are available. Our system continuously adjusts the object segmentation based on the user feedback and achieves accurate dense 3D segmentation masks with minimal human effort (few clicks per object). Besides its potential for efficient labeling of large-scale and varied 3D datasets, our approach, where the user directly interacts with the 3D environment, enables new applications in AR/VR and human-robot interaction. | ['Konrad Schindler', 'Siyu Tang', 'Ekin Celikkan', 'Theodora Kontogianni'] | 2022-04-14 | null | null | null | null | ['3d-instance-segmentation-1'] | ['computer-vision'] | [ 1.60456613e-01 2.05504909e-01 -6.73620626e-02 -4.95029807e-01
-4.02532309e-01 -9.35677469e-01 2.36169651e-01 5.69018796e-02
-3.24721992e-01 5.82199357e-02 -8.55597198e-01 -5.22885442e-01
4.05125856e-01 -7.33380258e-01 -7.46406913e-01 -2.16187701e-01
1.09981969e-01 1.08001506e+00 7.49964595e-01 -1.65947393e-01
2.29862198e-01 1.25342250e+00 -1.76472759e+00 -2.03392759e-01
9.65264976e-01 8.09409142e-01 6.94743931e-01 6.79804802e-01
-6.81375861e-01 -1.89086094e-01 -5.74185967e-01 1.99287802e-01
6.07950032e-01 7.13228583e-02 -8.26523304e-01 6.79827273e-01
3.29728633e-01 -4.06940788e-01 1.30489841e-01 9.51780975e-01
3.61811817e-01 2.89241105e-01 6.01654351e-01 -1.21725929e+00
-3.22662085e-01 -5.82218766e-02 -6.15348756e-01 -1.30824760e-01
5.53124726e-01 4.30185050e-01 4.68339115e-01 -8.95076513e-01
6.81744635e-01 1.32271171e+00 4.37929004e-01 7.22622931e-01
-1.46883166e+00 -5.70061564e-01 6.86266303e-01 -2.94505090e-01
-1.37218571e+00 -5.67833036e-02 9.16392982e-01 -6.64208829e-01
9.55544353e-01 3.14183503e-01 9.46368039e-01 7.47449219e-01
-4.22962010e-01 9.96364951e-01 8.33488286e-01 -3.59289110e-01
2.80173063e-01 6.13808513e-01 1.22067807e-02 6.15161538e-01
-1.12567104e-01 -7.72179589e-02 -8.00216049e-02 1.31459385e-01
1.46714985e+00 1.15156129e-01 -1.76405936e-01 -1.24323452e+00
-1.26221931e+00 3.76298517e-01 5.77010512e-01 1.43523112e-01
-1.74615979e-01 -6.37949705e-02 5.36613092e-02 2.80079544e-01
5.60277760e-01 5.28034210e-01 -9.43039179e-01 -1.07502704e-02
-6.45454288e-01 2.85771668e-01 6.23405397e-01 1.37542558e+00
1.04149783e+00 -3.84898961e-01 2.49102131e-01 6.14994824e-01
3.95187378e-01 6.22613668e-01 1.98049601e-02 -1.11822963e+00
2.40873113e-01 9.73206937e-01 4.75292057e-01 -7.30685294e-01
-4.05050635e-01 -1.67003691e-01 -3.30936790e-01 8.32121670e-01
2.62043208e-01 7.67133310e-02 -1.39655936e+00 1.31034565e+00
9.86357272e-01 -1.22510791e-02 -2.81551182e-01 9.97542739e-01
7.95812786e-01 2.79251993e-01 -1.12315468e-01 2.09605813e-01
9.15580451e-01 -7.21371889e-01 -2.21831858e-01 -4.62513596e-01
6.98288560e-01 -5.76180339e-01 1.61310613e+00 2.72582829e-01
-1.01880908e+00 -7.44279802e-01 -7.98364580e-01 -1.88186496e-01
-5.93208730e-01 -1.31332248e-01 6.07987165e-01 4.61085975e-01
-8.54783773e-01 4.05840814e-01 -1.14016438e+00 -5.20871460e-01
5.66094935e-01 6.99478328e-01 -3.05801064e-01 -5.77029735e-02
-5.94479084e-01 8.82558167e-01 4.96751845e-01 2.38154158e-02
-7.03411102e-01 -5.64343631e-01 -8.81358504e-01 -3.22593570e-01
6.03839278e-01 -8.26352298e-01 1.56826460e+00 -9.08966780e-01
-1.76642907e+00 1.17473352e+00 4.43832427e-02 8.73664767e-02
6.01799965e-01 -5.13979673e-01 3.38703781e-01 1.90317720e-01
-1.00613972e-02 1.09972942e+00 7.28796124e-01 -1.75511909e+00
-5.30585170e-01 -6.18382215e-01 4.79670912e-01 6.51500702e-01
1.57541618e-01 -3.47243011e-01 -1.06490934e+00 -5.38652875e-02
5.78420937e-01 -9.41582382e-01 -6.42123520e-01 3.17465782e-01
-4.54988658e-01 -2.22941428e-01 1.31663775e+00 3.65186818e-02
4.69023734e-01 -2.21998763e+00 1.67854413e-01 3.84014040e-01
1.56723768e-01 2.25970447e-01 7.05414172e-03 -2.09181588e-02
1.41860008e-01 2.22706765e-01 -2.21656978e-01 -5.42262077e-01
1.09323367e-01 2.72558630e-01 -2.38075592e-02 1.88930124e-01
7.79768676e-02 7.90699780e-01 -9.37641680e-01 -6.22302294e-01
7.65887499e-01 4.27487016e-01 -6.58249140e-01 5.88603258e-01
-7.69589961e-01 9.01262105e-01 -6.58183157e-01 6.82350993e-01
8.44370961e-01 -5.53144753e-01 -1.00222714e-01 -3.76083329e-02
-1.00135297e-01 1.27409488e-01 -1.31458390e+00 2.18389869e+00
-5.42911589e-01 3.20795059e-01 2.78613269e-01 -9.15750563e-01
9.26375031e-01 2.15867713e-01 4.05433357e-01 -2.06148490e-01
1.85404107e-01 1.81373462e-01 -5.12491107e-01 -5.37352383e-01
2.19990492e-01 1.92855656e-01 -1.46703303e-01 4.63910103e-01
7.12018553e-03 -1.10513031e+00 -1.64543152e-01 9.59528610e-02
7.43876994e-01 5.67780912e-01 8.47916231e-02 2.32973978e-01
1.59603387e-01 3.23695123e-01 2.42209509e-01 8.10666740e-01
7.25076161e-03 6.68261230e-01 1.13434166e-01 -4.15659577e-01
-7.96288371e-01 -1.19701147e+00 -1.43845677e-01 1.00606036e+00
8.82474184e-01 1.60471484e-01 -5.64985335e-01 -1.02788353e+00
2.56652962e-02 6.12013459e-01 -2.65553772e-01 3.36114645e-01
-4.70991313e-01 -3.74850594e-02 -1.86647400e-01 3.94554108e-01
4.75342631e-01 -1.02189910e+00 -1.10591865e+00 -2.10462436e-02
1.51207387e-01 -1.05856586e+00 -3.11848313e-01 3.10928434e-01
-1.32666278e+00 -1.05858481e+00 -5.82911134e-01 -8.25657189e-01
1.15246224e+00 5.19604385e-01 1.15368319e+00 3.80788110e-02
-1.96774334e-01 8.73795867e-01 -2.88338363e-01 -3.98428112e-01
-2.48999476e-01 3.22220832e-01 -1.50375709e-01 -4.10767615e-01
3.99597228e-01 -7.18471825e-01 -6.01371169e-01 6.66354835e-01
-7.52276421e-01 3.95547569e-01 4.32033241e-01 2.75732338e-01
9.75967348e-01 -1.26679748e-01 2.06249610e-01 -1.19992018e+00
1.31509885e-01 -5.20645455e-02 -9.02884126e-01 -4.01848331e-02
-2.08697140e-01 -2.71059304e-01 1.72733888e-01 -6.76705122e-01
-1.00482225e+00 6.76401615e-01 -4.62765805e-02 -8.66284907e-01
-9.22061443e-01 1.83207795e-01 -4.60955411e-01 -1.86168537e-01
6.99762762e-01 -2.41727024e-01 -1.80226535e-01 -6.52677834e-01
6.97545588e-01 6.99913442e-01 4.38164085e-01 -4.78686959e-01
1.08576524e+00 4.47695315e-01 -3.26825649e-01 -6.19955420e-01
-7.10147262e-01 -4.89052117e-01 -1.45006573e+00 -2.66185135e-01
9.68356311e-01 -6.47457719e-01 -6.22999251e-01 3.06955814e-01
-1.15397751e+00 -7.71497726e-01 -4.78500634e-01 4.11145210e-01
-6.05400980e-01 1.21448904e-01 -1.74316332e-01 -7.57893443e-01
8.86914805e-02 -1.29219985e+00 1.38165987e+00 3.13927650e-01
-1.71278402e-01 -9.28018332e-01 -4.47015077e-01 2.48434380e-01
-1.59705710e-02 3.97494614e-01 8.48140657e-01 -5.39805174e-01
-1.04389989e+00 -5.24102092e-01 -1.88453659e-01 9.80975479e-02
4.02781397e-01 -1.98895156e-01 -8.81428063e-01 -4.82816659e-02
-1.35381147e-01 -3.04237068e-01 1.85072005e-01 1.63152128e-01
1.57492554e+00 2.82237768e-01 -6.85061514e-01 4.79246587e-01
1.02204025e+00 3.59688669e-01 1.56199217e-01 5.09887226e-02
9.51509655e-01 4.98299152e-01 9.20242667e-01 1.35896772e-01
3.69180322e-01 7.26127982e-01 7.07438767e-01 -4.96521294e-01
2.58381039e-01 -2.54913539e-01 -1.93490878e-01 3.81873757e-01
1.19818058e-02 -1.68859035e-01 -1.00862682e+00 3.91240984e-01
-1.67063367e+00 -3.78492504e-01 4.64519560e-02 2.34491158e+00
7.27765381e-01 4.44286972e-01 6.76205605e-02 -1.29728513e-02
5.96610427e-01 -2.53153205e-01 -1.15867877e+00 1.28806406e-03
5.09104729e-01 1.96585879e-01 4.02292848e-01 5.31180918e-01
-1.13340306e+00 1.22214067e+00 5.62172985e+00 3.07489991e-01
-1.24077117e+00 -1.48736551e-01 4.63378280e-01 -1.55234322e-01
-3.30308139e-01 -5.05310018e-03 -6.93910122e-01 -9.84879285e-02
1.45127177e-01 2.85673529e-01 2.85157919e-01 1.05430460e+00
3.13006669e-01 -2.30690494e-01 -1.46755016e+00 1.26304722e+00
-2.33547062e-01 -1.00668585e+00 -1.80365637e-01 6.51167110e-02
5.31774759e-01 2.14847494e-02 -2.89359372e-02 2.50771701e-01
4.46918994e-01 -8.97003293e-01 5.85902154e-01 2.73345232e-01
8.76786709e-01 -5.58136344e-01 2.90521204e-01 9.28423464e-01
-9.08026457e-01 3.22127193e-01 -4.52445038e-02 6.20123185e-02
3.33102375e-01 3.63115072e-01 -1.36380291e+00 1.22395970e-01
9.02287722e-01 4.54454064e-01 -3.37705374e-01 1.03550148e+00
-2.65595794e-01 5.54602444e-02 -7.38293529e-01 2.29593888e-02
9.43706371e-03 -1.10226221e-01 6.25496507e-01 9.53418493e-01
1.39161319e-01 2.94307828e-01 7.22081423e-01 9.83041406e-01
1.18404821e-01 -3.85150500e-02 -8.01005483e-01 1.21549703e-01
4.84325826e-01 9.78236794e-01 -1.17279637e+00 -2.87163168e-01
-2.13308111e-01 1.31546843e+00 1.35353535e-01 5.15094399e-01
-5.91741860e-01 -6.21253669e-01 5.76370060e-01 3.76991689e-01
5.02306402e-01 -7.15601325e-01 -3.81617129e-01 -1.00428665e+00
-1.73329532e-01 -5.19482553e-01 1.14972619e-02 -8.79296005e-01
-1.07112098e+00 6.06515944e-01 3.08129102e-01 -1.28644156e+00
-1.56411126e-01 -4.86863792e-01 -2.13251159e-01 8.72733057e-01
-1.22350109e+00 -9.94667649e-01 -6.03698909e-01 7.11263835e-01
8.89262855e-01 3.44044238e-01 7.39521623e-01 1.85083985e-01
-2.28924900e-01 1.28391609e-01 -5.11736274e-01 -1.56111181e-01
3.09917688e-01 -1.54230905e+00 6.86826289e-01 3.45281750e-01
2.42202550e-01 4.81047332e-01 5.57328165e-01 -6.87887549e-01
-1.16501474e+00 -9.46062446e-01 2.76406676e-01 -8.28495860e-01
-3.27869765e-02 -8.01342547e-01 -1.00255263e+00 7.18708336e-01
-1.77044109e-01 2.54340678e-01 5.72612703e-01 9.80054662e-02
6.71411157e-02 2.28509039e-01 -1.36712122e+00 6.95091367e-01
1.54187179e+00 -4.26651895e-01 -4.23870176e-01 5.93072057e-01
1.05105972e+00 -1.15773404e+00 -5.51952660e-01 5.03846645e-01
9.03220177e-02 -8.23299885e-01 1.08645988e+00 -5.32390714e-01
-2.17810988e-01 -7.24801064e-01 2.59366501e-02 -1.16035128e+00
7.21894503e-02 -5.39695680e-01 -1.06063329e-01 6.94089949e-01
4.76516575e-01 -4.30513710e-01 1.05404401e+00 1.04355371e+00
-3.73340189e-01 -5.89245737e-01 -6.60366356e-01 -4.79803205e-01
-2.81780303e-01 -8.35508287e-01 8.06185961e-01 7.63240695e-01
-3.43451530e-01 3.37498158e-01 2.06581786e-01 7.41103888e-01
5.67223191e-01 5.10132134e-01 1.43603766e+00 -1.57148409e+00
-1.24424413e-01 -2.35830724e-01 -3.35499853e-01 -1.96909535e+00
-4.53109071e-02 -8.43107283e-01 2.85262972e-01 -1.66370499e+00
-4.29614544e-01 -1.08622897e+00 1.05395593e-01 4.82942909e-01
1.15698867e-01 2.78278410e-01 1.88996091e-01 3.32861155e-01
-6.43472016e-01 4.28542018e-01 1.62148023e+00 -3.51748914e-02
-1.08866632e+00 5.96636057e-01 -3.06198031e-01 1.09534717e+00
6.75039649e-01 -2.46717647e-01 -7.40953624e-01 -7.11442888e-01
-4.03936543e-02 -3.40788551e-02 4.18267429e-01 -9.05476213e-01
8.34223926e-02 -4.79455620e-01 5.98738194e-01 -1.10631752e+00
5.58285236e-01 -1.21332657e+00 1.07389167e-01 -1.04549408e-01
-4.09596898e-02 -3.43359560e-01 2.64895767e-01 4.45124924e-01
2.15219945e-01 -3.30538511e-01 5.91852188e-01 -6.43123627e-01
-7.43711114e-01 6.05619550e-01 -2.54929304e-01 -2.92107433e-01
1.24485862e+00 -7.34949112e-01 4.68487412e-01 -7.80669302e-02
-1.22870934e+00 4.15125161e-01 9.00326252e-01 4.32299227e-01
7.64570951e-01 -1.00057471e+00 6.82784617e-02 3.97944123e-01
6.72177970e-02 1.14502621e+00 1.19947493e-01 2.82225460e-01
-6.24625564e-01 6.14991710e-02 2.12970614e-01 -1.29994011e+00
-1.11548829e+00 5.56948066e-01 5.51357925e-01 2.54955202e-01
-9.50241327e-01 9.21762109e-01 4.19902980e-01 -1.04939580e+00
5.08641183e-01 -6.17645621e-01 -5.39619029e-02 -2.96316147e-01
1.28420681e-01 -9.68610272e-02 1.75740868e-02 -2.42983282e-01
-1.68575317e-01 6.64507449e-01 -1.35263622e-01 -1.69539183e-01
1.25506878e+00 -3.93192708e-01 3.56422514e-01 8.41689825e-01
1.00826156e+00 -2.71384060e-01 -1.64270473e+00 -2.73412704e-01
-2.06203461e-01 -8.84027064e-01 -4.56199981e-02 -6.42759800e-01
-9.17041779e-01 9.80661869e-01 7.76336968e-01 3.00990701e-01
9.96140957e-01 4.72732902e-01 6.08652592e-01 5.52002966e-01
6.29520535e-01 -9.39695060e-01 3.50317836e-01 4.89157915e-01
7.34250426e-01 -1.31725597e+00 -1.89579427e-01 -6.29014134e-01
-5.11407793e-01 9.81346965e-01 1.05798864e+00 6.27988502e-02
7.48269439e-01 1.32923320e-01 3.77499044e-01 -4.43876922e-01
-2.25495532e-01 -3.00839722e-01 8.60791653e-02 9.25905406e-01
1.57371070e-03 -1.48978770e-01 5.07147968e-01 1.01846233e-01
-1.47626728e-01 -8.34427401e-02 2.44112760e-01 1.09042561e+00
-5.84464490e-01 -1.20599401e+00 -3.84096116e-01 3.84586155e-01
2.17548072e-01 4.32822019e-01 -3.89474183e-01 8.74148190e-01
2.52164394e-01 6.36616707e-01 2.72445083e-01 -1.00940198e-01
6.66482210e-01 8.97647068e-03 5.38541436e-01 -1.27371860e+00
-2.98677623e-01 1.58181086e-01 -4.07960743e-01 -6.21567547e-01
-5.25248349e-01 -6.05563939e-01 -1.49025965e+00 2.20482528e-01
-5.17554045e-01 1.51017746e-02 8.25774789e-01 7.85255015e-01
4.93905276e-01 2.05979407e-01 7.23270595e-01 -1.70957518e+00
3.89301032e-02 -6.48478031e-01 -1.70779482e-01 2.31811881e-01
3.35871577e-01 -8.50922644e-01 -2.62812197e-01 2.00573787e-01] | [8.086259841918945, -3.0363729000091553] |
0361fa99-2869-451a-9a56-f3af3cf14c8d | a-survey-on-neural-open-information | 2205.11725 | null | https://arxiv.org/abs/2205.11725v2 | https://arxiv.org/pdf/2205.11725v2.pdf | A Survey on Neural Open Information Extraction: Current Status and Future Directions | Open Information Extraction (OpenIE) facilitates domain-independent discovery of relational facts from large corpora. The technique well suits many open-world natural language understanding scenarios, such as automatic knowledge base construction, open-domain question answering, and explicit reasoning. Thanks to the rapid development in deep learning technologies, numerous neural OpenIE architectures have been proposed and achieve considerable performance improvement. In this survey, we provide an extensive overview of the-state-of-the-art neural OpenIE models, their key design decisions, strengths and weakness. Then, we discuss limitations of current solutions and the open issues in OpenIE problem itself. Finally we list recent trends that could help expand its scope and applicability, setting up promising directions for future research in OpenIE. To our best knowledge, this paper is the first review on this specific topic. | ['Yongbin Li', 'Haiyang Yu', 'Jian Sun', 'Jingyang Li', 'Cheng Long', 'Aixin Sun', 'Bowen Yu', 'Shaowen Zhou'] | 2022-05-24 | null | null | null | null | ['open-information-extraction'] | ['natural-language-processing'] | [-4.43891615e-01 8.51447344e-01 -5.50385714e-01 -4.56715763e-01
-7.17061341e-01 -6.14901245e-01 2.82565981e-01 2.76241839e-01
-2.81060070e-01 1.14156711e+00 3.01769495e-01 -3.97222340e-01
-5.51277161e-01 -1.15096569e+00 -8.28101039e-01 -1.18655853e-01
-1.65009335e-01 8.07359993e-01 6.78209960e-02 -5.34176707e-01
3.80574055e-02 1.60499364e-01 -1.40756512e+00 5.48708260e-01
9.04828846e-01 9.80077267e-01 -6.15611970e-02 3.59075010e-01
-9.25487161e-01 1.38092828e+00 -4.57316548e-01 -9.26343560e-01
-4.24321443e-02 3.66858244e-01 -1.59813023e+00 -6.05249703e-01
5.30514717e-01 -3.12388182e-01 -5.89590967e-01 7.54029393e-01
3.72995555e-01 1.62357643e-01 3.29609424e-01 -1.05726373e+00
-1.40351427e+00 9.66120422e-01 -1.24232396e-01 5.53957462e-01
4.85407412e-01 -3.60689282e-01 1.67330873e+00 -7.29064763e-01
1.06548250e+00 9.17833984e-01 6.97498024e-01 5.57876587e-01
-7.43000329e-01 -5.73404789e-01 5.62164605e-01 7.26172507e-01
-1.34291661e+00 -3.23362857e-01 6.67827725e-01 -1.68261692e-01
1.82764852e+00 1.89940981e-03 4.49238002e-01 9.07479763e-01
2.01876208e-01 1.07201552e+00 5.98593473e-01 -5.12303174e-01
1.25520714e-02 -3.26859532e-03 6.88845515e-01 5.68146825e-01
5.50645232e-01 -6.11579977e-02 -4.16935384e-01 -8.58250782e-02
6.34037256e-01 -3.70674193e-01 -2.93927733e-02 -4.40410703e-01
-1.10943604e+00 9.64010358e-01 7.59185314e-01 6.26537979e-01
-3.13352257e-01 4.59363032e-03 5.71494222e-01 6.26068473e-01
4.20738935e-01 8.10606658e-01 -1.04561782e+00 -4.64660022e-03
-4.40225542e-01 6.05957150e-01 1.48243988e+00 1.01502299e+00
7.33950913e-01 -2.49218851e-01 1.58028901e-01 1.06452250e+00
1.00140825e-01 5.76907098e-02 1.71174213e-01 -1.06111968e+00
7.70239353e-01 7.10385382e-01 -1.06375329e-01 -9.42734957e-01
-5.76868594e-01 -3.91085535e-01 -4.34491009e-01 -4.68992054e-01
3.26041281e-01 -3.35094690e-01 -3.01400632e-01 1.48487568e+00
2.80370295e-01 -6.01547770e-02 6.27838612e-01 4.72886354e-01
1.78484416e+00 5.98618805e-01 1.60088181e-01 9.75688398e-02
1.63818836e+00 -8.76690805e-01 -9.76151228e-01 -4.98358279e-01
4.73974138e-01 -3.04125100e-01 5.08433521e-01 2.30712369e-01
-9.60438251e-01 -2.12850973e-01 -8.83417487e-01 -7.97228158e-01
-1.00378585e+00 -1.62282124e-01 1.26600587e+00 3.05103719e-01
-7.92051792e-01 2.50951529e-01 -6.08519673e-01 -5.00393927e-01
5.56274533e-01 5.17644703e-01 -5.15940249e-01 -3.19869220e-02
-1.83740127e+00 1.14512932e+00 8.83528829e-01 5.77497929e-02
-3.76416624e-01 -7.18422771e-01 -9.45640445e-01 2.59292066e-01
1.07789063e+00 -9.29212093e-01 1.68359828e+00 -3.10568303e-01
-1.06141961e+00 1.17360866e+00 -1.13571271e-01 -9.92197573e-01
-2.88298845e-01 -6.33134544e-01 -6.31027639e-01 5.40720997e-03
1.51259154e-01 6.20087683e-01 -1.41096994e-01 -9.79432404e-01
-6.57798707e-01 -3.98886621e-01 6.83893204e-01 7.44372010e-02
-2.25451604e-01 3.38343114e-01 -2.60913521e-01 -3.94635379e-01
6.09389134e-02 -4.46161717e-01 -1.73636422e-01 -2.66660452e-01
-2.48381659e-01 -9.03566837e-01 3.61831725e-01 -4.51736003e-01
1.38598442e+00 -1.72072673e+00 -1.16156880e-02 -2.08369434e-01
7.32313156e-01 3.69272590e-01 1.40438959e-01 7.80647993e-01
-1.39557585e-01 1.46196559e-02 -1.31899063e-02 3.12981516e-01
4.35184650e-02 6.46878600e-01 -6.18090689e-01 -1.42926618e-01
1.22894615e-01 1.40918458e+00 -9.61996913e-01 -5.49820185e-01
-5.35198003e-02 1.71114892e-01 -4.59005535e-01 -1.50999185e-02
-9.52652156e-01 -6.04317859e-02 -7.51400173e-01 5.40910363e-01
3.40863436e-01 -5.19117296e-01 2.44723320e-01 -2.34201346e-02
-1.56203851e-01 9.84813869e-01 -1.01064086e+00 1.77968597e+00
-5.38929462e-01 8.65431726e-01 -6.16462231e-02 -1.13855624e+00
1.14731359e+00 6.67469561e-01 2.68255442e-01 -4.10033286e-01
2.12565467e-01 3.60790402e-01 5.50648309e-02 -7.76002526e-01
6.56751812e-01 -1.14520006e-01 -1.42705858e-01 3.64623576e-01
6.03145123e-01 1.48705840e-01 3.95136058e-01 3.06988686e-01
9.69407678e-01 -3.21685039e-02 8.68146598e-01 -1.43697351e-01
6.29874110e-01 3.91152641e-03 5.10470331e-01 7.68392384e-01
-1.86563417e-01 -2.63135377e-02 4.76912171e-01 -1.07402122e+00
-6.61655962e-01 -1.00079167e+00 -3.94501865e-01 1.13983750e+00
-1.25832468e-01 -5.02718985e-01 -5.78098774e-01 -7.96911538e-01
4.87502068e-02 6.43213749e-01 -5.62756181e-01 2.05667183e-01
-7.49766231e-01 -4.74608660e-01 7.50822484e-01 7.87874341e-01
6.73561931e-01 -1.53840709e+00 -3.11628193e-01 3.01614195e-01
-8.08988214e-01 -1.40370584e+00 5.10857761e-01 3.39714944e-01
-8.80266070e-01 -1.20383132e+00 -1.49552375e-01 -1.10253704e+00
-1.74130276e-02 1.80319864e-02 1.62758505e+00 -9.79332104e-02
-4.52966280e-02 1.74622968e-01 -4.80608433e-01 -5.99517703e-01
-8.92373472e-02 6.47096097e-01 -6.64340183e-02 -5.94205499e-01
1.28119946e+00 -4.37706530e-01 -1.96647868e-01 -1.06086187e-01
-5.29510140e-01 -1.91784993e-01 4.66415673e-01 7.67385781e-01
3.53634596e-01 1.31062195e-01 1.04465044e+00 -1.19102836e+00
7.63371944e-01 -8.11239541e-01 -3.94471020e-01 5.94379425e-01
-4.75610793e-01 2.36614347e-01 3.89427632e-01 1.13087393e-01
-1.60113382e+00 -6.19386375e-01 -5.90374470e-01 1.59936398e-01
-4.42177266e-01 9.53278840e-01 -1.83929652e-01 6.78754225e-02
9.35090721e-01 -2.19297677e-01 -3.83375794e-01 -6.65084362e-01
8.13157737e-01 8.30873787e-01 7.39517212e-01 -8.68746161e-01
3.04825664e-01 4.01235521e-01 -5.60123086e-01 -7.46256649e-01
-1.58740819e+00 -4.44901407e-01 -6.49420321e-01 3.16172332e-01
8.51179659e-01 -9.15377319e-01 -6.19298041e-01 1.37819797e-01
-1.41696143e+00 -6.23847768e-02 -4.46640015e-01 1.06352352e-01
-5.42571366e-01 1.39143154e-01 -1.03815293e+00 -4.71444368e-01
-6.76454961e-01 -7.01854587e-01 6.12593055e-01 3.75591129e-01
-4.92606431e-01 -1.33148324e+00 2.80811638e-01 7.86332190e-01
1.54959947e-01 4.96679544e-02 8.82054985e-01 -1.17130983e+00
-8.07197392e-01 -1.34644508e-01 -2.98537523e-01 1.08145669e-01
-1.95412293e-01 -2.87240863e-01 -1.04109359e+00 1.63214296e-01
-2.36449450e-01 -6.40189111e-01 8.33122134e-01 3.47500473e-01
9.04168785e-01 -2.48151183e-01 -6.10709071e-01 6.00896716e-01
1.39679325e+00 1.94640756e-01 6.14674151e-01 6.75406098e-01
5.35484135e-01 6.85094357e-01 5.71937680e-01 7.51810148e-02
9.36715066e-01 1.09930538e-01 2.22688898e-01 1.99906528e-01
-9.72536877e-02 -3.12571257e-01 -1.37256578e-01 7.29928136e-01
-1.12653799e-01 -2.23201886e-01 -1.16625321e+00 8.55345845e-01
-1.80129099e+00 -1.03104663e+00 -1.23587064e-01 1.49941492e+00
9.66244936e-01 2.10296407e-01 -3.32747310e-01 -6.56798556e-02
5.55501759e-01 3.56302381e-01 -6.81092918e-01 -6.87333047e-01
-2.53160268e-01 4.13796186e-01 1.29106119e-01 3.51576120e-01
-1.17381465e+00 1.30602896e+00 7.26496506e+00 5.29605150e-01
-4.29289401e-01 1.56335533e-01 3.31226885e-01 9.24522877e-02
-2.00355083e-01 1.49946749e-01 -1.16336954e+00 -3.34706545e-01
1.01019180e+00 -1.18034661e-01 2.54466951e-01 9.32282150e-01
-6.35999560e-01 1.02641344e-01 -1.32588255e+00 7.34194338e-01
7.76756406e-02 -1.92372537e+00 -1.24132954e-01 -4.69892584e-02
7.40966141e-01 7.21717715e-01 -4.94597495e-01 7.46131599e-01
7.91476190e-01 -7.07298756e-01 1.96718395e-01 3.76164705e-01
5.33485711e-01 -6.32047355e-01 8.63015115e-01 3.02196175e-01
-1.07312453e+00 -4.84441310e-01 -4.95810926e-01 -4.24886346e-01
7.39366561e-02 5.07305026e-01 -4.31972653e-01 7.64431536e-01
9.42595720e-01 7.41753519e-01 -2.38201022e-01 7.59921730e-01
-6.34908140e-01 3.85333389e-01 -2.91348070e-01 -9.06486139e-02
2.80215025e-01 2.37150282e-01 3.97723526e-01 1.05056572e+00
-3.77286196e-01 4.43281591e-01 9.69800130e-02 1.07512343e+00
-2.60480851e-01 9.57317725e-02 -8.51073503e-01 -1.36323854e-01
6.22546196e-01 9.04729128e-01 -1.97295368e-01 -3.04231733e-01
-1.13770473e+00 3.32534730e-01 1.07955837e+00 3.02473515e-01
-4.11283672e-01 -6.84219182e-01 6.29803538e-01 -2.33820692e-01
3.44017297e-01 -4.76145856e-02 -3.58229131e-01 -1.43636954e+00
3.15840691e-02 -8.76871228e-01 1.13067746e+00 -7.49165416e-01
-1.49394369e+00 7.44332910e-01 2.59316057e-01 -5.77689648e-01
-4.36088562e-01 -7.47843683e-01 -3.11021715e-01 5.07524610e-01
-1.65871155e+00 -1.02103674e+00 5.21102510e-02 4.14459854e-01
4.86122996e-01 -3.27881128e-01 1.15129435e+00 2.55367517e-01
-5.24472415e-01 4.07305866e-01 2.07407072e-01 7.30339944e-01
4.19071466e-01 -1.12866998e+00 8.89808357e-01 4.62028265e-01
4.07630473e-01 1.01764703e+00 4.51626867e-01 -6.15931988e-01
-1.21269667e+00 -6.33553803e-01 1.51347876e+00 -6.72605395e-01
1.09712458e+00 -2.93375969e-01 -1.09094560e+00 1.29311681e+00
3.93193752e-01 8.06689728e-03 9.49940264e-01 9.90312696e-01
-5.97562492e-01 -1.99322477e-01 -1.00176239e+00 3.86340320e-01
9.14374769e-01 -5.79497993e-01 -1.35657966e+00 1.74631819e-01
1.00965810e+00 -4.98241961e-01 -1.33901346e+00 3.00003856e-01
5.48099399e-01 -8.05561423e-01 1.06461620e+00 -8.85305524e-01
4.56958532e-01 1.46104053e-01 -2.92074904e-02 -1.00220156e+00
-3.89255434e-01 -5.30208468e-01 -6.96255445e-01 1.16928577e+00
4.96354342e-01 -7.53241599e-01 7.25327432e-01 7.86681235e-01
-1.04108721e-01 -1.01373649e+00 -8.85855079e-01 -5.56935608e-01
5.78212619e-01 -6.25274420e-01 7.54783511e-01 1.00855744e+00
7.64734745e-01 7.84987152e-01 -8.11751408e-04 -1.42613510e-02
2.94652998e-01 4.23817188e-01 5.85911989e-01 -1.56264055e+00
-1.29812971e-01 -2.05594614e-01 -1.90449312e-01 -1.30420744e+00
6.05253160e-01 -9.17366385e-01 -5.10685682e-01 -1.97032523e+00
1.13351911e-01 -5.33348203e-01 -2.41309568e-01 6.72600985e-01
-3.18626277e-02 -1.58017337e-01 -1.73570514e-02 -1.16573013e-01
-9.15408015e-01 3.36253941e-01 1.08368957e+00 -1.36791155e-01
-1.04317293e-01 -1.64778650e-01 -1.31718683e+00 8.02603483e-01
1.21823668e+00 -2.01778278e-01 -4.53347027e-01 -8.04213941e-01
8.13503265e-01 1.89270422e-01 3.31088938e-02 -7.46474624e-01
3.02852094e-01 -1.04727484e-01 -1.11900508e-01 -6.86059833e-01
1.53215095e-01 -7.23456383e-01 -4.77915913e-01 -1.42821455e-02
-4.52193022e-01 -1.83253855e-01 2.98757881e-01 3.72042507e-01
-4.60564554e-01 -3.79075140e-01 3.05098206e-01 -5.61101079e-01
-1.19600332e+00 2.16772601e-01 -4.29378867e-01 5.87884784e-01
8.30257714e-01 4.33376692e-02 -6.16890132e-01 -3.08542103e-01
-7.05174565e-01 6.47591174e-01 -4.32150155e-01 8.05523276e-01
5.24114311e-01 -1.00616872e+00 -6.78927362e-01 -3.87172513e-02
4.69995350e-01 3.67846102e-01 2.53216028e-01 3.41384470e-01
-3.18171114e-01 1.22246909e+00 5.48360348e-02 2.36522220e-02
-8.92148495e-01 9.08272326e-01 4.25287187e-01 -7.05891848e-01
-8.52092028e-01 9.47907209e-01 1.16865918e-01 -8.31972003e-01
3.86776656e-01 -3.71010274e-01 -5.92670798e-01 -2.60841429e-01
7.02204227e-01 2.22061470e-01 3.88651252e-01 -4.05266047e-01
-2.87558883e-01 2.06972942e-01 -5.35553515e-01 3.75373960e-01
1.34394670e+00 -1.72595024e-01 -4.79124337e-01 4.79972929e-01
1.03525829e+00 -3.88391495e-01 -5.10960579e-01 -7.32041180e-01
1.92575529e-01 2.12281439e-02 7.11067170e-02 -9.65225935e-01
-8.82458627e-01 8.17394257e-01 -5.44196293e-02 2.07126930e-01
8.31065416e-01 6.00566685e-01 1.11945200e+00 1.18006778e+00
6.30706608e-01 -1.03400040e+00 -4.77095544e-01 1.10006189e+00
8.38737011e-01 -1.31832325e+00 4.73396555e-02 -5.59525430e-01
-3.98587286e-01 1.06051159e+00 9.17795062e-01 -1.33420050e-01
9.78253126e-01 3.27168822e-01 1.36111051e-01 -7.98005939e-01
-1.18067873e+00 -3.22862178e-01 1.15454823e-01 7.97876000e-01
8.03847730e-01 -1.39422147e-02 -3.16619426e-01 8.97718132e-01
-5.37570417e-01 2.58368909e-01 2.51595974e-01 7.40014374e-01
-5.76306403e-01 -1.16272509e+00 -2.77337581e-01 5.97724795e-01
-4.65708882e-01 -3.57943684e-01 -4.53390032e-01 7.86691785e-01
2.70701081e-01 9.85622942e-01 2.15012372e-01 -5.22203371e-02
2.85033911e-01 3.79131258e-01 4.18802023e-01 -5.68878233e-01
-5.65988660e-01 -6.28260911e-01 6.60121441e-01 -3.63225698e-01
-4.13836420e-01 -5.54695487e-01 -1.46926010e+00 -4.07426566e-01
-6.30836487e-01 3.55505079e-01 2.70879120e-01 1.13586855e+00
4.21183914e-01 4.17274117e-01 -3.00977379e-01 2.74991132e-02
-2.85198450e-01 -8.15479815e-01 -3.62355143e-01 -8.47604498e-02
2.61974812e-01 -8.60347033e-01 2.12444037e-01 -2.34637469e-01] | [9.792695045471191, 8.369087219238281] |
639340eb-b096-489e-9dae-e022267d2bd7 | coherence-and-diversity-through-noise-self | 2302.02780 | null | https://arxiv.org/abs/2302.02780v1 | https://arxiv.org/pdf/2302.02780v1.pdf | Coherence and Diversity through Noise: Self-Supervised Paraphrase Generation via Structure-Aware Denoising | In this paper, we propose SCANING, an unsupervised framework for paraphrasing via controlled noise injection. We focus on the novel task of paraphrasing algebraic word problems having practical applications in online pedagogy as a means to reduce plagiarism as well as ensure understanding on the part of the student instead of rote memorization. This task is more complex than paraphrasing general-domain corpora due to the difficulty in preserving critical information for solution consistency of the paraphrased word problem, managing the increased length of the text and ensuring diversity in the generated paraphrase. Existing approaches fail to demonstrate adequate performance on at least one, if not all, of these facets, necessitating the need for a more comprehensive solution. To this end, we model the noising search space as a composition of contextual and syntactic aspects and sample noising functions consisting of either one or both aspects. This allows for learning a denoising function that operates over both aspects and produces semantically equivalent and syntactically diverse outputs through grounded noise injection. The denoising function serves as a foundation for learning a paraphrasing function which operates solely in the input-paraphrase space without carrying any direct dependency on noise. We demonstrate SCANING considerably improves performance in terms of both semantic preservation and producing diverse paraphrases through extensive automated and manual evaluation across 4 datasets. | ['Vikram Goyal', 'Mukesh Mohania', 'Venktesh V.', 'Rishabh Gupta'] | 2023-02-06 | null | null | null | null | ['paraphrase-generation', 'memorization', 'paraphrase-generation'] | ['computer-code', 'natural-language-processing', 'natural-language-processing'] | [ 4.56458569e-01 9.38397199e-02 2.90932469e-02 -9.34231430e-02
-8.32492948e-01 -1.03108370e+00 6.08747184e-01 4.36656713e-01
-3.58542740e-01 4.39084977e-01 5.66409528e-01 -5.22494197e-01
-3.49615037e-01 -8.49307716e-01 -8.37804973e-01 -5.15380263e-01
8.28361869e-01 3.15869987e-01 3.04875355e-02 -5.67953825e-01
5.59359372e-01 4.30090725e-01 -1.63389349e+00 4.18277830e-01
1.08450651e+00 3.02914411e-01 2.39677519e-01 5.75313210e-01
-5.22940278e-01 8.12211514e-01 -8.98530245e-01 -5.79649091e-01
8.56535360e-02 -8.40498626e-01 -1.15671587e+00 -2.52117421e-02
8.26232314e-01 1.73845403e-02 -1.62991285e-01 1.18404782e+00
4.36194390e-01 3.41479570e-01 4.17923480e-01 -9.02198553e-01
-6.78736210e-01 6.67879701e-01 -1.81329176e-01 2.32695207e-01
7.88215995e-01 1.25571132e-01 1.24161506e+00 -7.70360589e-01
5.94632030e-01 1.14960802e+00 7.10976303e-01 4.42271739e-01
-1.54221010e+00 -3.47859085e-01 -2.84184396e-01 2.58785069e-01
-8.78262758e-01 -2.93251187e-01 7.38097250e-01 -2.33737275e-01
9.50766742e-01 3.67423654e-01 8.55676234e-01 9.78854358e-01
2.38690168e-01 6.41063988e-01 9.85029578e-01 -6.90820158e-01
2.99718022e-01 3.41204017e-01 2.17642486e-01 5.01048148e-01
1.15707830e-01 -2.20075026e-01 -8.27025771e-01 -8.81844163e-02
4.53668296e-01 -2.62013078e-01 -3.82860959e-01 -5.06174743e-01
-8.18319261e-01 7.95060277e-01 1.88471347e-01 3.80608708e-01
2.33937167e-02 -5.95155470e-02 3.78807336e-01 9.30110216e-01
2.27900207e-01 1.12808597e+00 -1.57073036e-01 -4.04163986e-01
-1.02555835e+00 5.12715518e-01 1.02708554e+00 8.77096951e-01
6.33056462e-01 -1.30716830e-01 -9.98355523e-02 9.30274427e-01
-2.78584689e-01 2.01646984e-01 7.05390334e-01 -1.07043099e+00
4.16084021e-01 8.52078915e-01 -9.15733874e-02 -9.33018565e-01
-8.23965222e-02 -6.05960310e-01 -4.24079418e-01 1.49933890e-01
5.98182201e-01 3.44955683e-01 -6.99608922e-01 1.93518090e+00
2.90785134e-01 -4.39453274e-02 1.36986420e-01 6.60218716e-01
7.53116965e-01 5.66395223e-01 7.93374106e-02 3.52366194e-02
1.36725903e+00 -9.21089709e-01 -6.42429531e-01 -4.02922153e-01
8.55233967e-01 -1.07837701e+00 1.76208746e+00 4.89848197e-01
-1.66381240e+00 -2.75926679e-01 -1.06975913e+00 -5.56384921e-01
-3.56463581e-01 -2.94637471e-01 1.32312045e-01 6.63445592e-01
-1.03816235e+00 7.56888151e-01 -2.75599629e-01 -3.27234387e-01
2.24908963e-01 -1.09315878e-02 -3.66739988e-01 -3.06008607e-01
-1.04986930e+00 1.23688567e+00 5.53663075e-02 -4.18697953e-01
-2.42242754e-01 -1.21888530e+00 -1.05033612e+00 4.35894012e-01
2.96532601e-01 -8.68941367e-01 1.33785653e+00 -9.48658407e-01
-1.46565831e+00 9.31660593e-01 -5.44277914e-02 -3.44397068e-01
5.83705068e-01 -2.45687723e-01 1.69244498e-01 1.38857782e-01
2.28581041e-01 3.41465592e-01 7.96692848e-01 -6.72638953e-01
-1.10870138e-01 -3.46622795e-01 2.26742700e-01 5.68998456e-01
-5.46809316e-01 -9.94637422e-03 -5.46532199e-02 -9.65408921e-01
2.40595713e-01 -7.38948941e-01 -6.92887092e-03 -1.13805778e-01
-1.18740708e-01 -1.93488881e-01 6.05051041e-01 -7.53899813e-01
1.26441324e+00 -2.08034897e+00 5.95466852e-01 2.10359201e-01
1.45750731e-01 3.47484738e-01 -2.54921168e-01 6.15569472e-01
-2.62672573e-01 7.13618696e-02 -2.83455759e-01 -3.29473048e-01
-5.40845990e-02 8.12867433e-02 -5.96751750e-01 2.48906925e-01
8.80851299e-02 9.27300751e-01 -1.11346960e+00 -1.91997156e-01
1.80862889e-01 8.55135173e-02 -7.28204548e-01 3.03019136e-01
-6.26170337e-02 1.68829039e-01 -5.18806688e-02 1.65413409e-01
4.00503129e-01 7.10978732e-02 1.65310758e-03 1.74151778e-01
1.09759502e-01 8.02997231e-01 -1.17799449e+00 2.01910305e+00
-9.34041262e-01 5.06140292e-01 2.68116035e-02 -1.05699158e+00
7.84447670e-01 1.75041961e-03 9.86017436e-02 -8.05688560e-01
-8.39944184e-02 2.65512586e-01 -1.97485924e-01 -4.54612374e-01
7.25483239e-01 -5.18783987e-01 -1.92335486e-01 8.77677977e-01
1.59439713e-01 -8.45793545e-01 3.20059597e-01 6.56565368e-01
1.23655391e+00 3.63510824e-03 2.30441660e-01 -4.80169356e-01
5.99442124e-01 4.28385176e-02 2.67782006e-02 9.86971140e-01
1.06944241e-01 6.05503917e-01 6.39606774e-01 -2.39076614e-01
-1.14043856e+00 -1.14728093e+00 8.40277299e-02 1.18302131e+00
1.79368451e-01 -6.65741444e-01 -8.32670391e-01 -4.65108216e-01
-1.42950475e-01 1.13373470e+00 -3.15967709e-01 -6.29430711e-01
-6.48307979e-01 -2.99267354e-03 6.38834894e-01 3.88674587e-01
2.27984771e-01 -1.00926340e+00 -6.25880539e-01 6.40573576e-02
-2.39767939e-01 -8.03021133e-01 -5.77400565e-01 2.97149301e-01
-7.29372859e-01 -9.68539774e-01 -3.49919438e-01 -8.92042100e-01
4.35950845e-01 5.17195106e-01 1.29525614e+00 2.76514500e-01
-3.06391269e-01 4.13903058e-01 -1.96822777e-01 -2.65314639e-01
-9.54635680e-01 1.18983544e-01 -2.52442032e-01 -6.09147131e-01
3.53300005e-01 -9.48223352e-01 -2.53320813e-01 -2.63234433e-02
-1.28886461e+00 9.46690738e-02 2.95854002e-01 8.01497400e-01
1.56019017e-01 -7.35684261e-02 4.72721785e-01 -9.29356456e-01
1.13637722e+00 -3.53624642e-01 -2.71100789e-01 3.30011576e-01
-4.12235022e-01 2.92281657e-01 9.10394013e-01 -4.46863979e-01
-8.02932441e-01 -2.31004938e-01 -2.13742614e-01 -3.56078714e-01
2.78465375e-02 3.86166066e-01 -6.05161302e-02 3.08975228e-03
1.01515329e+00 4.51061159e-01 3.31047177e-01 -3.27911437e-01
6.21578872e-01 4.27924007e-01 6.41695976e-01 -8.52883041e-01
9.32428420e-01 2.81541664e-02 3.60787325e-02 -8.89555812e-01
-9.88225341e-01 -5.29802024e-01 -3.44285488e-01 3.57007757e-02
4.09565926e-01 -6.80147648e-01 -3.30127984e-01 2.62630641e-01
-1.14935124e+00 -1.02340639e-01 -8.85998726e-01 6.23357631e-02
-6.37963414e-01 6.91412270e-01 -6.59175515e-01 -1.96757317e-01
-2.03040630e-01 -1.26144898e+00 8.40858817e-01 1.28384247e-01
-7.99709797e-01 -1.00790882e+00 1.67632297e-01 6.73125207e-01
5.22215962e-01 -2.52330959e-01 1.33907044e+00 -8.59358549e-01
-3.73498410e-01 -1.99157491e-01 4.50955005e-04 4.90349472e-01
1.54071134e-02 -3.77142966e-01 -7.93066025e-01 -1.68640196e-01
4.87585574e-01 -6.83195531e-01 7.32813299e-01 -1.86070040e-01
9.88253891e-01 -5.30939162e-01 3.69169623e-01 5.10565102e-01
1.12936902e+00 -3.91229957e-01 6.28596961e-01 5.09597301e-01
4.19409662e-01 8.35664034e-01 1.41074106e-01 7.89464116e-02
-1.01164868e-02 5.67664027e-01 1.65026709e-02 2.46413678e-01
-4.35233742e-01 -5.90401947e-01 2.00438425e-01 7.67382681e-01
6.90961421e-01 -3.25455982e-03 -7.52506435e-01 7.85253048e-01
-1.61042261e+00 -1.03370547e+00 -4.48098183e-02 2.12856960e+00
1.24064827e+00 6.63207192e-03 4.64647561e-02 3.40297371e-01
4.21180129e-01 1.26758888e-01 -2.99993962e-01 -9.57437038e-01
5.02893701e-02 6.98084772e-01 -5.14427051e-02 8.18685770e-01
-5.39921761e-01 1.03753233e+00 5.62520504e+00 9.28979695e-01
-8.26328218e-01 -6.46815971e-02 1.64493859e-01 -1.94061711e-01
-8.96080732e-01 1.36629686e-01 -2.92428374e-01 3.09060693e-01
6.45084202e-01 -3.32105935e-01 6.97099447e-01 7.45345235e-01
9.07448009e-02 -1.36587545e-01 -1.09106779e+00 7.50440955e-01
2.08025962e-01 -1.33960772e+00 3.74154001e-01 -2.65954584e-01
6.67614341e-01 -7.05225766e-01 2.65837628e-02 2.17585087e-01
2.41510779e-01 -1.13921201e+00 7.20747352e-01 1.19610697e-01
4.55296457e-01 -7.83911467e-01 3.87741387e-01 5.01074553e-01
-7.66261041e-01 -4.20908583e-03 -2.44284257e-01 -4.60148692e-01
-2.35021368e-01 3.38779688e-01 -6.31490350e-01 3.27359408e-01
2.71231771e-01 5.32523453e-01 -5.25181174e-01 9.06646132e-01
-5.97070634e-01 5.36559284e-01 -2.65199363e-01 -5.71603402e-02
3.11132103e-01 -3.50279897e-01 8.05043519e-01 1.20009458e+00
1.12560146e-01 6.58429638e-02 -2.28544213e-02 1.01840425e+00
-9.03770104e-02 3.17629576e-01 -7.21749723e-01 5.55665605e-02
6.36207104e-01 9.78345692e-01 -5.10422587e-01 -1.39723286e-01
-2.09920093e-01 9.64266300e-01 5.55709422e-01 3.07845473e-01
-5.05160213e-01 -6.34884357e-01 6.43605411e-01 2.35502258e-01
3.41322757e-02 -1.39012396e-01 -9.00031030e-01 -1.31097281e+00
1.53425992e-01 -1.24381828e+00 3.83666664e-01 -8.04583967e-01
-1.11287510e+00 1.53000325e-01 -2.29924656e-02 -8.56570065e-01
-2.71140277e-01 -3.43358874e-01 -9.45238054e-01 1.14523065e+00
-1.21219456e+00 -9.12966490e-01 -2.93164462e-01 4.29413468e-01
7.87132025e-01 -4.09396775e-02 7.19404519e-01 -1.50527671e-01
-2.66417444e-01 7.64538288e-01 -1.75830781e-01 -4.42668766e-01
7.97472954e-01 -1.35529327e+00 3.77484977e-01 1.04417920e+00
2.73742437e-01 9.55248177e-01 1.17907488e+00 -4.87026274e-01
-1.35154235e+00 -6.75711572e-01 1.35437584e+00 -5.66602409e-01
9.03825104e-01 -2.79376358e-01 -1.22045267e+00 3.56957436e-01
2.06793249e-01 -6.93330765e-01 5.43439329e-01 7.55758723e-03
-6.94856405e-01 2.47181490e-01 -1.15535879e+00 7.76947856e-01
1.07350969e+00 -8.13042700e-01 -1.27076149e+00 3.48001957e-01
7.49321938e-01 -4.21599597e-01 -5.28745115e-01 4.69607674e-02
2.90125847e-01 -9.16276693e-01 1.04073584e+00 -9.02212918e-01
1.05382836e+00 4.76281159e-02 4.14835922e-02 -1.37861109e+00
-1.28520891e-01 -6.37054801e-01 1.37516350e-01 1.16065037e+00
1.22771420e-01 -3.42170000e-01 1.00951195e+00 7.43403733e-01
-2.83253878e-01 -6.79009557e-01 -8.72255683e-01 -6.14847183e-01
4.56175894e-01 -4.57291842e-01 3.01930368e-01 1.06606662e+00
3.81247878e-01 6.69444323e-01 7.58820102e-02 -2.43559852e-01
4.74958956e-01 2.35989496e-01 8.62656951e-01 -7.57184207e-01
-4.20939475e-01 -7.17247784e-01 -2.46097147e-01 -1.05069864e+00
4.65834945e-01 -1.18126178e+00 -8.67193639e-02 -1.47245002e+00
1.67049512e-01 -1.34387031e-01 2.39459068e-01 3.68390858e-01
-1.81362957e-01 1.44496500e-01 1.30079776e-01 2.26073682e-01
-2.48809561e-01 4.36126262e-01 1.14097917e+00 1.21991697e-03
-3.07612032e-01 -4.26210947e-02 -9.24454451e-01 6.87590420e-01
7.07818568e-01 -3.52899611e-01 -7.97401428e-01 -4.79827255e-01
4.55812752e-01 4.58896719e-03 5.08029521e-01 -7.70806432e-01
4.35391694e-01 -2.57416964e-01 -1.38798049e-02 -6.24442399e-02
1.15971774e-01 -6.69033706e-01 -1.75669760e-01 3.89600277e-01
-8.74013722e-01 2.29629934e-01 3.42566997e-01 3.99876148e-01
-1.60017043e-01 -7.69461691e-01 8.62192035e-01 -2.60262698e-01
-3.58073473e-01 -5.52878618e-01 -3.89496148e-01 4.96104181e-01
7.24280655e-01 -4.29525226e-01 -3.70071769e-01 -5.25235474e-01
-5.90281606e-01 7.49682635e-02 7.77083218e-01 3.88524532e-01
5.63846469e-01 -1.01500833e+00 -4.62896943e-01 4.66120988e-01
7.38307014e-02 4.48149396e-03 1.61212996e-01 2.57643729e-01
-5.58071613e-01 2.11537808e-01 -2.85932362e-01 -3.44734401e-01
-1.39226913e+00 3.46288353e-01 2.96312898e-01 -3.52701873e-01
-5.60919583e-01 9.04406011e-01 -2.66011115e-02 -5.92570603e-01
4.14810568e-01 -2.40288913e-01 6.97681308e-02 8.25390592e-02
5.81307769e-01 3.80402982e-01 4.52194273e-01 -2.55186617e-01
4.82817292e-02 4.83109713e-01 -1.33690372e-01 -1.81149542e-01
1.18748939e+00 -8.74004886e-02 -2.31711030e-01 1.84834167e-01
1.18751967e+00 2.29317844e-01 -6.29952669e-01 -3.03964883e-01
2.26391759e-02 -5.38521886e-01 -1.80654138e-01 -7.48196602e-01
-3.67805898e-01 7.59289384e-01 -1.15258448e-01 2.73922324e-01
1.03436768e+00 -4.80587035e-02 8.50118577e-01 5.53351820e-01
7.38815814e-02 -8.59199584e-01 4.28093165e-01 6.54353857e-01
8.43384802e-01 -7.07532346e-01 7.34734908e-02 -3.82473886e-01
-4.80286092e-01 1.09255779e+00 6.02429211e-01 -1.99028984e-01
1.16756797e-01 1.46305978e-01 -1.20627843e-01 -1.64709300e-01
-6.87213659e-01 1.46050602e-01 3.72319043e-01 1.88973129e-01
4.48624402e-01 -4.26972181e-01 -4.62042332e-01 4.47437882e-01
-7.71384060e-01 -2.55151331e-01 8.48657310e-01 1.15374219e+00
-6.31725192e-01 -1.20488441e+00 -2.47192889e-01 2.95575052e-01
-3.36182535e-01 -3.30681622e-01 -8.19134772e-01 5.54653823e-01
-2.65970558e-01 9.14014757e-01 -2.14364663e-01 -1.47087395e-01
5.61496317e-01 4.35692608e-01 6.14382267e-01 -9.29106057e-01
-1.05245614e+00 -4.30328369e-01 4.66287322e-02 -5.54091811e-01
5.33339977e-02 -3.53422493e-01 -9.53097224e-01 -5.73350191e-01
-1.23140670e-01 3.43520552e-01 7.26998210e-01 1.20753860e+00
2.64362514e-01 1.44380927e-01 3.54133368e-01 -4.57264960e-01
-1.28049302e+00 -6.91875756e-01 -2.57481933e-01 9.21334505e-01
2.18742847e-01 -2.67448574e-01 -3.18685710e-01 -2.00840712e-01] | [11.512219429016113, 9.203950881958008] |
2eebf657-702c-4776-bfcd-66f3ae87e1af | towards-demystifying-dimensions-of-source | 2008.13064 | null | https://arxiv.org/abs/2008.13064v3 | https://arxiv.org/pdf/2008.13064v3.pdf | Towards Demystifying Dimensions of Source Code Embeddings | Source code representations are key in applying machine learning techniques for processing and analyzing programs. A popular approach in representing source code is neural source code embeddings that represents programs with high-dimensional vectors computed by training deep neural networks on a large volume of programs. Although successful, there is little known about the contents of these vectors and their characteristics. In this paper, we present our preliminary results towards better understanding the contents of code2vec neural source code embeddings. In particular, in a small case study, we use the code2vec embeddings to create binary SVM classifiers and compare their performance with the handcrafted features. Our results suggest that the handcrafted features can perform very close to the highly-dimensional code2vec embeddings, and the information gains are more evenly distributed in the code2vec embeddings compared to the handcrafted features. We also find that the code2vec embeddings are more resilient to the removal of dimensions with low information gains than the handcrafted features. We hope our results serve a stepping stone toward principled analysis and evaluation of these code representations. | ['Md. Rafiqul Islam Rabin', 'Omprakash Gnawali', 'Mohammad Amin Alipour', 'Arjun Mukherjee'] | 2020-08-29 | null | null | null | null | ['method-name-prediction'] | ['natural-language-processing'] | [-4.68883604e-01 1.22330844e-01 -3.79161358e-01 -3.30711186e-01
-3.36051106e-01 -7.00256407e-01 4.93176699e-01 5.98808765e-01
-2.02756613e-01 5.23247942e-02 6.27456307e-01 -7.10448980e-01
5.35285845e-02 -8.03974628e-01 -7.72516549e-01 -2.64732808e-01
-3.81105214e-01 -8.96157995e-02 -7.10410625e-02 -3.36723000e-01
5.75910628e-01 2.79032558e-01 -1.61219418e+00 2.64972866e-01
5.99558890e-01 5.71793258e-01 -4.12939340e-02 8.91858518e-01
-4.82362747e-01 1.09413028e+00 -4.59586442e-01 -4.81817156e-01
2.02214241e-01 -1.69270992e-01 -7.49248147e-01 -4.02669519e-01
3.65035504e-01 -3.09511036e-01 -4.52287018e-01 1.25830841e+00
-2.95367315e-02 4.47657965e-02 7.70924330e-01 -1.30020392e+00
-1.26715481e+00 5.11526167e-01 -4.13879007e-01 1.79348394e-01
5.05006492e-01 -2.92372908e-02 1.22283435e+00 -9.39869523e-01
6.12443745e-01 9.41086888e-01 1.12810421e+00 3.93612653e-01
-1.38065851e+00 -1.98628053e-01 -3.21326613e-01 9.03082639e-02
-1.22075653e+00 -1.64296106e-01 7.74680018e-01 -1.16771126e+00
1.32226956e+00 9.92138535e-02 3.98191094e-01 7.43154705e-01
5.85217714e-01 7.03372657e-01 2.51407653e-01 -6.36043310e-01
1.21544980e-01 6.63800359e-01 5.74524999e-01 1.01766217e+00
5.34288704e-01 3.55229937e-02 -9.26154945e-03 -6.99269176e-01
4.79302973e-01 4.66583878e-01 -2.54692346e-01 -9.64240432e-01
-1.13141418e+00 1.51001227e+00 6.18697226e-01 4.78392780e-01
-1.55274570e-01 5.43114662e-01 8.55844021e-01 3.15529585e-01
2.03348413e-01 1.01210845e+00 -4.34412658e-01 -5.41444898e-01
-7.76580095e-01 2.69904852e-01 6.89463377e-01 1.00572872e+00
8.66921008e-01 2.75755614e-01 9.98525843e-02 9.02986050e-01
2.26096854e-01 1.63200676e-01 1.06840706e+00 -7.71604240e-01
4.45219338e-01 8.55442464e-01 -2.20658690e-01 -1.46861160e+00
-2.22319931e-01 1.54713541e-01 -2.55529970e-01 4.13942397e-01
1.63332954e-01 -1.61725760e-01 -5.72140038e-01 1.27920187e+00
-3.62436265e-01 -6.21552289e-01 9.59853977e-02 4.29340601e-01
5.92896342e-01 6.54416442e-01 -1.17566280e-01 6.53801203e-01
1.10171592e+00 -9.34004247e-01 -3.88162225e-01 -8.58529285e-02
1.21695864e+00 -7.21093059e-01 1.09146070e+00 -1.04229808e-01
-7.63833106e-01 -6.17643952e-01 -1.23781121e+00 -2.27279872e-01
-6.74210489e-01 7.63972476e-02 1.05248725e+00 9.81558800e-01
-1.15234983e+00 6.24492884e-01 -7.45570600e-01 -3.73939842e-01
4.49275672e-01 1.36322677e-01 -6.25186563e-01 -2.02483162e-01
-7.18698621e-01 8.27859461e-01 4.01331186e-01 -6.43592238e-01
-6.82832062e-01 -6.04055822e-01 -1.41350138e+00 5.02818704e-01
-3.56778294e-01 1.06698453e-01 9.65843499e-01 -8.98126066e-01
-8.16065550e-01 7.00074494e-01 6.00066874e-03 -2.49925792e-01
-2.28878140e-01 1.43946141e-01 -2.59795964e-01 -1.60577849e-01
-1.58981401e-02 2.29963943e-01 4.76701379e-01 -1.08343863e+00
-2.10533231e-01 -8.25237483e-02 7.01791719e-02 -3.61339450e-01
-1.03601885e+00 1.94144696e-02 -6.27457127e-02 -6.29741609e-01
-1.71137169e-01 -8.22733283e-01 -1.39400393e-01 1.85884356e-01
-3.66676338e-02 -1.93152532e-01 5.64312041e-01 -5.60700715e-01
1.29651690e+00 -2.54265618e+00 7.92450383e-02 1.45544827e-01
5.54079354e-01 3.43994766e-01 -3.26869816e-01 6.70624793e-01
-3.62045079e-01 3.95056486e-01 -4.13683578e-02 1.82661772e-01
1.84060454e-01 5.62776364e-02 -4.04306024e-01 5.41487038e-01
2.11255953e-01 8.44403923e-01 -9.78383541e-01 -3.81599933e-01
3.60054746e-02 4.58534032e-01 -1.14160264e+00 2.15970084e-01
1.30251467e-01 -7.94039547e-01 -4.43480015e-01 3.96323532e-01
3.67559403e-01 -1.30634308e-01 3.72302942e-02 1.28812581e-01
-2.78544668e-02 1.19492367e-01 -5.98171413e-01 1.51095533e+00
-7.52788186e-01 1.43661451e+00 -2.78835148e-01 -1.02958941e+00
1.01707399e+00 1.67130217e-01 1.87883228e-01 -3.08624715e-01
1.10468112e-01 3.23382206e-02 1.82371005e-01 -9.30280328e-01
9.41648543e-01 3.95866930e-02 -3.33285153e-01 6.83333695e-01
4.69660997e-01 -1.94484055e-01 5.15375473e-02 3.65250468e-01
1.27852166e+00 -2.05396622e-01 4.38739270e-01 -4.34429467e-01
8.96722451e-02 3.37817848e-01 1.27703071e-01 5.67757189e-01
-2.78440714e-01 6.70844853e-01 8.94105256e-01 -6.35661185e-01
-1.27508855e+00 -7.94572532e-01 -1.80944800e-01 1.07388330e+00
-2.27568626e-01 -8.58016253e-01 -6.27148092e-01 -8.18170369e-01
5.96024036e-01 7.17115223e-01 -1.04354525e+00 -4.88283843e-01
-3.00116837e-01 -3.84904504e-01 8.98602486e-01 1.00034499e+00
-3.17449182e-01 -6.47805393e-01 -6.06032431e-01 9.73565876e-02
3.86802435e-01 -3.71593207e-01 -4.72561002e-01 4.73199338e-01
-8.96614552e-01 -1.22969925e+00 -5.59449255e-01 -9.15359378e-01
8.53793681e-01 3.75801742e-01 1.04869294e+00 3.92482072e-01
-4.87659812e-01 4.39188391e-01 -5.71440876e-01 -4.74607617e-01
-6.73442721e-01 -2.22957119e-01 1.07106352e-02 -6.39942527e-01
9.31586683e-01 -2.80257404e-01 -2.52052933e-01 -1.89714193e-01
-9.88879919e-01 -5.99254549e-01 3.79011959e-01 1.04197049e+00
-1.20538607e-01 3.32276560e-02 1.07959569e-01 -9.04062748e-01
1.15234745e+00 -8.95529389e-01 -3.12479228e-01 4.77539748e-02
-7.74397612e-01 5.58655143e-01 9.53924954e-01 -3.89496714e-01
-6.19010091e-01 -9.51670632e-02 -2.77304232e-01 -4.31369334e-01
-4.76484448e-02 7.83459485e-01 4.91905034e-01 -3.28672618e-01
1.33150041e+00 1.17080532e-01 3.43527012e-02 -5.22336125e-01
6.32457554e-01 8.79070997e-01 4.72137658e-03 -7.87139058e-01
7.83792794e-01 1.71874896e-01 -5.23717940e-01 -7.21976697e-01
8.79801996e-03 -4.25649792e-01 -6.28396988e-01 2.98632145e-01
9.81889427e-01 -6.25655472e-01 -1.75942838e-01 -2.04290841e-02
-1.05340397e+00 -4.17061988e-03 -3.50198209e-01 5.17530918e-01
-4.50973123e-01 5.09062171e-01 -7.09541261e-01 -3.58059496e-01
1.45969018e-01 -1.46718550e+00 6.62993968e-01 2.93379892e-02
-6.48468733e-01 -1.37443721e+00 4.90046889e-01 -1.52819484e-01
6.58592045e-01 2.85123378e-01 1.60147440e+00 -9.85608518e-01
-7.63620660e-02 -8.13777328e-01 -2.00699523e-01 4.37648803e-01
2.02254742e-01 3.84154350e-01 -9.77775037e-01 -2.88508564e-01
-5.95091283e-02 -4.78241950e-01 7.98981011e-01 7.13926405e-02
1.30203867e+00 -3.46346945e-01 -2.59904057e-01 7.88773358e-01
1.82410991e+00 2.59096026e-01 5.21759868e-01 4.13365126e-01
7.96502054e-01 4.21021760e-01 7.36203343e-02 4.61863905e-01
2.17541903e-01 2.87447959e-01 3.59167457e-01 8.57498944e-02
2.05101326e-01 -4.46922421e-01 3.59533399e-01 1.07595277e+00
2.46371120e-01 2.73235172e-01 -1.30232203e+00 9.69076812e-01
-1.39285731e+00 -9.51339066e-01 -1.03334583e-01 1.92624378e+00
7.42130458e-01 -2.71010250e-02 -1.26771718e-01 6.44984618e-02
4.69231099e-01 2.55468726e-01 -1.95159137e-01 -1.20439327e+00
3.62537354e-01 2.63224036e-01 4.17632550e-01 2.44753838e-01
-8.74681890e-01 4.28726256e-01 7.30309105e+00 3.91159415e-01
-1.03267586e+00 7.09445104e-02 -6.30880287e-03 -2.31360495e-02
-5.58216453e-01 -1.71333570e-02 -2.79263228e-01 3.92623723e-01
1.13574433e+00 -6.20655835e-01 4.21899885e-01 1.71143758e+00
-4.68277037e-01 1.66504487e-01 -1.59902203e+00 8.53428423e-01
2.99879193e-01 -1.61459959e+00 -1.04813837e-01 7.38631412e-02
8.60626698e-01 2.17354447e-01 1.03373602e-01 6.81154609e-01
5.71198225e-01 -1.19308996e+00 5.65759838e-01 5.05117327e-02
8.74836445e-01 -7.05894709e-01 7.57109940e-01 5.54116331e-02
-9.34614658e-01 -4.59942997e-01 -9.07050014e-01 -2.08098099e-01
-5.97805619e-01 4.15047497e-01 -7.80131817e-01 -1.82207767e-02
5.17945766e-01 7.34435558e-01 -9.03977931e-01 9.61281538e-01
1.05441943e-01 4.27968919e-01 3.04274678e-01 -3.81057858e-01
2.15538338e-01 2.09605411e-01 7.99916759e-02 1.55824721e+00
4.02208328e-01 -4.27491874e-01 -2.79635668e-01 1.30339718e+00
-2.09600002e-01 4.45613079e-02 -1.24377894e+00 -6.93780005e-01
3.14203233e-01 1.09239900e+00 -4.38507825e-01 -4.15451199e-01
-8.47990930e-01 6.30062938e-01 5.36130607e-01 3.29788476e-01
-6.47262514e-01 -1.21924818e+00 1.24129081e+00 -2.23840207e-01
5.13630033e-01 -2.68340945e-01 -2.79253662e-01 -1.45526886e+00
-1.75017133e-01 -9.15807903e-01 9.46532339e-02 -7.31524110e-01
-1.19983053e+00 6.65384173e-01 -1.94593057e-01 -1.11039042e+00
-4.20096338e-01 -1.04910815e+00 -9.00471330e-01 8.90027881e-01
-1.01422608e+00 -4.18050706e-01 -8.94308686e-02 3.42608094e-01
3.44501674e-01 -5.48399806e-01 1.16378129e+00 6.19856939e-02
-3.27622592e-01 9.02646244e-01 9.12585437e-01 7.34852850e-01
4.63337272e-01 -1.30932605e+00 5.13340056e-01 7.53794014e-01
1.48458242e-01 1.31509948e+00 6.71097279e-01 -2.70397514e-01
-1.80600202e+00 -1.03880680e+00 5.79078376e-01 -7.62219787e-01
8.43392253e-01 -3.37080151e-01 -9.96374309e-01 1.02616525e+00
1.99439719e-01 1.16054542e-01 1.10479844e+00 4.03090417e-01
-9.33093309e-01 5.24798810e-01 -1.27320457e+00 3.17720115e-01
3.76848787e-01 -1.10295367e+00 -1.01418471e+00 1.41165093e-01
7.09322870e-01 -4.44389395e-02 -1.13006020e+00 -2.73820460e-01
6.19182646e-01 -9.00926411e-01 9.69967306e-01 -9.22891200e-01
1.24166322e+00 3.73913348e-02 -5.03611267e-01 -1.64455545e+00
-6.48633003e-01 2.63513744e-01 -1.76003277e-02 1.05821526e+00
3.76427025e-01 -5.25301576e-01 6.98012173e-01 8.71942759e-01
1.46983294e-02 -6.37261569e-01 -4.08365905e-01 -8.05616558e-01
5.81760943e-01 -3.09531510e-01 6.46796465e-01 1.47430766e+00
7.85700560e-01 -3.11866432e-01 8.66554976e-02 -2.56195635e-01
3.84646803e-01 1.71972498e-01 6.91818714e-01 -9.85005438e-01
-4.20584708e-01 -6.20345294e-01 -1.12585318e+00 -5.15510082e-01
3.26923698e-01 -1.48261929e+00 -5.86164445e-02 -1.38368416e+00
4.48455930e-01 -2.73808122e-01 -1.96098104e-01 4.85953510e-01
-1.80394113e-01 -1.54595360e-01 1.57957256e-01 9.82283205e-02
-7.72371963e-02 4.69514579e-01 5.80272853e-01 -2.84912914e-01
-7.23119900e-02 -5.43646216e-01 -9.54533994e-01 7.40166724e-01
7.50154793e-01 -6.57195628e-01 -4.20352906e-01 -8.17871213e-01
1.83196947e-01 -1.83768630e-01 4.18346841e-03 -8.81673634e-01
7.78531805e-02 -1.34416848e-01 4.90671456e-01 7.65069351e-02
-7.12410063e-02 -6.82122886e-01 -3.28134388e-01 5.25365353e-01
-6.15468800e-01 3.67779791e-01 4.77319121e-01 4.49234217e-01
-3.00202847e-01 -7.44142711e-01 4.72587049e-01 -3.16560358e-01
-8.17573071e-01 -1.82719752e-01 -7.42627919e-01 1.07452959e-01
9.51451540e-01 -2.91798830e-01 -4.36352223e-01 -1.63771048e-01
-1.73387960e-01 -1.97869048e-01 9.31134641e-01 7.51369059e-01
7.91967690e-01 -1.64821053e+00 -2.10718393e-01 5.84213376e-01
6.16731584e-01 -6.33544981e-01 -2.12308854e-01 1.65071249e-01
-8.62120748e-01 5.86636424e-01 -6.09312952e-01 -2.88207263e-01
-1.11904049e+00 1.07827926e+00 2.14759141e-01 2.19095945e-01
-4.05843288e-01 1.01442599e+00 1.04190618e-01 -7.03468621e-01
1.50871843e-01 -4.86374319e-01 -1.79691821e-01 -1.93637833e-02
6.83348358e-01 2.06292659e-01 -1.21295825e-01 -4.05479312e-01
-3.33888084e-01 4.32045162e-01 -1.21767320e-01 3.80167842e-01
1.62945116e+00 5.43386757e-01 -3.14061910e-01 4.73173290e-01
2.06216359e+00 1.85728088e-01 -8.27943027e-01 2.45730225e-02
1.14026189e-01 -9.18294907e-01 5.54688759e-02 -1.10971853e-01
-1.17131424e+00 1.27922893e+00 5.43923140e-01 4.15494680e-01
5.48129678e-01 -7.75007233e-02 5.33781648e-01 6.24783516e-01
3.14023316e-01 -8.44657779e-01 7.74061680e-02 5.32381058e-01
6.14726186e-01 -1.16450429e+00 -4.27027978e-02 1.65857881e-01
-5.43831170e-01 1.58402348e+00 5.99636793e-01 -4.58531082e-01
7.94277608e-01 3.89333367e-01 4.85501140e-02 -4.26697433e-01
-6.35061860e-01 2.70166904e-01 2.40314789e-02 8.93855929e-01
9.39749837e-01 5.47523275e-02 -1.04477722e-02 5.80704093e-01
-1.64267153e-01 -1.60128966e-01 1.12221289e+00 1.09867370e+00
-5.03340065e-01 -1.02767193e+00 -5.15116811e-01 8.51269364e-01
-3.40931296e-01 -3.70111167e-01 -2.58914083e-01 6.84911191e-01
-1.41429991e-01 6.68629587e-01 7.85726309e-02 -9.67646003e-01
2.82383323e-01 2.86125034e-01 1.74268737e-01 -9.09117937e-01
-5.65956712e-01 -8.80047381e-01 -2.60661721e-01 -6.03944898e-01
1.93891495e-01 -4.94236261e-01 -1.14121211e+00 -5.77892721e-01
-3.23700756e-01 2.57414341e-01 9.52641249e-01 3.38100791e-01
5.13712823e-01 5.86622059e-01 6.36439860e-01 -9.62254941e-01
-8.18667769e-01 -6.09811664e-01 -4.73726749e-01 8.05660963e-01
6.43971801e-01 -4.95347679e-01 -6.29567087e-01 1.26558259e-01] | [7.506735801696777, 7.861898899078369] |
c0b2fb24-7a6b-4a8c-b204-12dfd1729a23 | learning-to-infer | null | null | https://openreview.net/forum?id=B1Z3W-b0W | https://openreview.net/pdf?id=B1Z3W-b0W | Learning to Infer | Inference models, which replace an optimization-based inference procedure with a learned model, have been fundamental in advancing Bayesian deep learning, the most notable example being variational auto-encoders (VAEs). In this paper, we propose iterative inference models, which learn how to optimize a variational lower bound through repeatedly encoding gradients. Our approach generalizes VAEs under certain conditions, and by viewing VAEs in the context of iterative inference, we provide further insight into several recent empirical findings. We demonstrate the inference optimization capabilities of iterative inference models, explore unique aspects of these models, and show that they outperform standard inference models on typical benchmark data sets. | ['Yisong Yue', 'Joseph Marino', 'Stephan Mandt'] | 2018-01-01 | null | null | null | iclr-2018-1 | ['inference-optimization'] | ['audio'] | [ 3.43366601e-02 3.03456545e-01 -4.12154496e-01 -6.48975372e-01
-7.70369172e-01 -3.59083742e-01 8.93643379e-01 -3.71696770e-01
-3.51090223e-01 1.01745665e+00 3.00640136e-01 -4.24699038e-01
-3.83812845e-01 -6.96126342e-01 -1.12942004e+00 -6.84922218e-01
5.62080968e-05 6.39837682e-01 -1.48773402e-01 2.13055760e-01
4.06221539e-01 2.33983636e-01 -1.25978470e+00 -9.75040048e-02
6.49534225e-01 7.75901973e-01 5.87668866e-02 6.45575821e-01
-1.49348844e-02 8.10148001e-01 -2.60164589e-01 -7.98542023e-01
-2.68545628e-01 -3.33630234e-01 -6.82586014e-01 -3.60957801e-01
5.20663559e-01 -6.44555926e-01 -6.79602981e-01 1.12672222e+00
1.48248151e-01 3.74436319e-01 1.12913024e+00 -1.17000926e+00
-7.97920883e-01 9.30493653e-01 -3.04038525e-01 4.44238931e-01
-8.26749280e-02 -7.37901777e-02 1.47233987e+00 -1.01252532e+00
4.65965539e-01 1.54159307e+00 8.38332891e-01 6.12868547e-01
-1.54004169e+00 -4.88219261e-01 3.39529127e-01 4.28152829e-01
-1.42749441e+00 -7.19631195e-01 6.15831673e-01 -4.17139381e-01
9.80976880e-01 -3.62084433e-02 4.60623384e-01 1.52637136e+00
4.42933559e-01 1.22825122e+00 7.35723257e-01 -1.10475361e-01
4.63029921e-01 -1.17737107e-01 4.00758952e-01 9.23469901e-01
2.26300001e-01 4.29993629e-01 -6.77634895e-01 -1.86157450e-01
7.24132359e-01 -5.19560836e-03 -2.54338473e-01 -5.27188219e-02
-8.19057763e-01 1.18727005e+00 3.08945984e-01 -1.75049528e-01
-2.16833144e-01 9.63614881e-01 2.65495241e-01 -1.19050033e-01
5.47612429e-01 1.76900417e-01 -3.81382287e-01 -2.14960575e-01
-1.04378700e+00 5.52295029e-01 9.16771650e-01 8.22132587e-01
8.18248391e-01 1.47165880e-01 -5.77973604e-01 7.54706383e-01
8.62820745e-01 5.63204587e-01 -9.49848145e-02 -1.30287528e+00
1.49540767e-01 -2.67157942e-01 4.65969518e-02 -5.66804528e-01
-8.74961689e-02 -4.82052803e-01 -7.25189090e-01 -2.26480693e-01
1.82290077e-01 -2.69305468e-01 -7.95722783e-01 2.10609317e+00
2.46999809e-03 5.93846977e-01 -8.74459893e-02 5.91469824e-01
6.40114665e-01 6.60880566e-01 -4.27935980e-02 -2.62309790e-01
1.05710375e+00 -5.39793372e-01 -1.03899074e+00 -1.40969947e-01
-6.14992194e-02 -1.38495237e-01 7.96262324e-01 5.45218945e-01
-1.21735489e+00 -2.89746225e-01 -1.01967847e+00 -2.57714033e-01
-2.77315646e-01 -7.35239461e-02 9.00188148e-01 6.89703345e-01
-9.73031878e-01 7.24796176e-01 -1.36676645e+00 1.16177104e-01
7.06770658e-01 6.66562095e-02 2.98204869e-01 9.86829028e-02
-1.30482650e+00 8.14943969e-01 3.03964078e-01 4.35234904e-01
-1.62031293e+00 -9.11773801e-01 -8.59012783e-01 2.29543313e-01
4.89153743e-01 -9.21226025e-01 1.51989508e+00 -8.25680867e-02
-1.83211720e+00 3.05552334e-01 -6.85514450e-01 -7.49880254e-01
5.22645116e-01 -5.93605936e-01 -8.40076879e-02 -1.15147181e-01
-2.03143388e-01 5.28050721e-01 1.00728202e+00 -1.10597777e+00
-3.33513856e-01 -1.50274009e-01 8.51894915e-02 -2.90245652e-01
-6.12825491e-02 -3.93556267e-01 -5.36076605e-01 -4.49102700e-01
-7.40764588e-02 -6.43947184e-01 -2.05917835e-01 -9.48497802e-02
-5.18564284e-01 -5.94059885e-01 4.80266988e-01 -3.94264549e-01
1.24524891e+00 -1.80440855e+00 5.85778952e-01 1.65997013e-01
5.42594254e-01 -2.50042498e-01 2.23199725e-01 4.71906886e-02
3.53248775e-01 1.65342599e-01 -5.15958548e-01 -7.83389151e-01
5.56776047e-01 7.24957764e-01 -6.75233722e-01 4.10797924e-01
3.37228566e-01 1.25966465e+00 -1.09326458e+00 -4.76846129e-01
1.62403315e-01 6.56765342e-01 -9.94638860e-01 2.91086823e-01
-7.90705860e-01 3.06824476e-01 -4.23456401e-01 3.37529689e-01
4.35881108e-01 -4.01835650e-01 2.71133244e-01 -3.51861566e-01
1.91415790e-02 4.90175992e-01 -9.95324016e-01 1.67009413e+00
-4.21352386e-01 9.42701459e-01 -1.52830938e-02 -1.30275631e+00
4.31528538e-01 1.48459777e-01 1.11115851e-01 -1.89037234e-01
7.63669461e-02 -1.33724689e-01 -2.49403909e-01 -4.33734417e-01
4.18429881e-01 -1.28200516e-01 4.58669811e-02 4.72773433e-01
4.46203738e-01 -2.29882345e-01 3.44261110e-01 4.43431914e-01
7.34427691e-01 3.52581084e-01 7.43243173e-02 -3.16712946e-01
1.87125251e-01 -5.44544756e-01 6.02261603e-01 1.54793561e+00
4.93968576e-02 3.95851314e-01 6.40830874e-01 -1.91706464e-01
-8.40120614e-01 -1.81358826e+00 -6.84845448e-01 1.03290737e+00
-4.47133392e-01 -6.16165757e-01 -6.00733757e-01 -6.40141785e-01
1.45484403e-01 9.02081072e-01 -8.02092075e-01 -1.27147123e-01
-3.43263239e-01 -1.06997728e+00 7.41510153e-01 6.96903646e-01
4.26164418e-01 -5.56888163e-01 -3.54737878e-01 1.15714602e-01
-1.49709553e-01 -9.64037836e-01 -2.35910803e-01 1.76090911e-01
-9.47370887e-01 -8.12372386e-01 -4.00869250e-01 -8.42462480e-02
2.78234392e-01 -4.02030051e-01 1.45099711e+00 -1.69808984e-01
-1.99415624e-01 6.47211373e-01 1.27249375e-01 -4.67339545e-01
-4.04271156e-01 1.08369723e-01 2.82142580e-01 -2.46909916e-01
3.55510712e-01 -7.34759867e-01 -3.81903946e-01 -1.77727997e-01
-7.65359700e-01 -1.82107687e-01 5.21129906e-01 1.03132403e+00
6.21219456e-01 -2.02534467e-01 4.60241407e-01 -1.05460906e+00
6.26328766e-01 -8.17800820e-01 -9.37438548e-01 3.34384441e-01
-7.66929626e-01 6.82761788e-01 2.12826893e-01 -2.09708616e-01
-1.29972029e+00 -5.18603504e-01 -4.94888306e-01 -4.82670993e-01
2.03977644e-01 9.62131083e-01 4.72977124e-02 2.83254057e-01
2.71827579e-01 2.15605766e-01 -2.50166450e-02 -5.40109634e-01
6.11525297e-01 3.12015057e-01 7.08344340e-01 -1.09297800e+00
5.41349530e-01 5.17606556e-01 1.84187993e-01 -7.66321540e-01
-1.45402503e+00 2.86596715e-01 -3.94281805e-01 -1.00201368e-01
8.67827535e-01 -7.95411706e-01 -1.07351422e+00 1.03170104e-01
-1.13691974e+00 -4.48689193e-01 -1.26565516e-01 5.76977074e-01
-5.95416367e-01 2.73862332e-01 -7.16268003e-01 -1.16691577e+00
-9.42241699e-02 -1.13611615e+00 9.40427721e-01 3.33602399e-01
-1.55801505e-01 -1.38392258e+00 3.04681540e-01 -1.14079453e-02
3.23955148e-01 -1.75742984e-01 9.00339901e-01 -1.81804135e-01
-7.54087985e-01 3.75753880e-01 -2.15315163e-01 4.50753212e-01
-2.27596611e-01 3.22584569e-01 -9.95787501e-01 -1.12588279e-01
-2.47213915e-02 -3.87158722e-01 1.39249766e+00 8.75283420e-01
1.74919903e+00 -2.39432454e-01 -4.48471248e-01 1.02839327e+00
1.37403381e+00 -2.78883010e-01 5.69261014e-01 -3.49469900e-01
5.47840297e-01 -7.75552765e-02 2.45524640e-03 4.80598360e-01
4.55044389e-01 2.90249944e-01 4.19925630e-01 6.40082836e-01
7.45057687e-02 -5.19372582e-01 2.98793852e-01 9.37296271e-01
-3.55673909e-01 -1.72191173e-01 -5.70427775e-01 3.96257639e-01
-2.01329112e+00 -1.12656152e+00 1.85978666e-01 1.75914204e+00
1.11539578e+00 2.89903760e-01 -2.24187896e-01 -3.59841704e-01
2.56169528e-01 3.16814899e-01 -9.11806881e-01 -2.45469227e-01
4.12572995e-02 7.34040260e-01 3.23737383e-01 9.76258516e-01
-9.25982893e-01 8.71856332e-01 8.60338211e+00 7.13527679e-01
-4.58528131e-01 2.39196956e-01 4.82460499e-01 -3.73758703e-01
-6.49949670e-01 -5.63780181e-02 -1.21032476e+00 4.37320203e-01
1.08898664e+00 9.71426368e-02 8.23290288e-01 7.24473238e-01
6.49537798e-03 4.53942567e-02 -1.67270482e+00 8.01692486e-01
2.93986988e-03 -1.70875084e+00 5.93538582e-02 6.69509172e-02
8.53788078e-01 3.23042899e-01 3.81534517e-01 6.83675110e-01
9.79152501e-01 -1.33634114e+00 5.55911660e-01 1.03774512e+00
4.30381894e-01 -6.15604103e-01 5.02075076e-01 1.85400978e-01
-7.96873450e-01 8.58658329e-02 -5.42021394e-01 -1.65421888e-01
4.04552937e-01 8.03654730e-01 -3.54143620e-01 2.95071453e-01
7.01191306e-01 9.60983515e-01 -2.38860905e-01 6.31201804e-01
-7.02960908e-01 1.03476334e+00 -4.61121887e-01 -2.75842726e-01
2.09538773e-01 -4.19269234e-01 6.42101467e-01 1.35170805e+00
7.45598078e-02 -2.76009068e-02 -2.88454860e-01 1.84697151e+00
-2.82428980e-01 -8.35365772e-01 -2.94231474e-01 -1.64126858e-01
6.15274429e-01 7.34287202e-01 -8.23491141e-02 -4.82878238e-01
-3.69214267e-01 5.99213362e-01 7.46369481e-01 7.65420437e-01
-1.14979625e+00 -2.00364336e-01 1.07566643e+00 -5.57042778e-01
6.55164659e-01 -4.05520439e-01 -2.19048515e-01 -1.39380395e+00
-3.76720965e-01 -5.03829837e-01 2.78701395e-01 -5.90409279e-01
-1.45848978e+00 -1.30713299e-01 5.34864247e-01 -2.18614221e-01
-5.46926856e-01 -9.11190093e-01 -5.10586321e-01 6.62156105e-01
-1.40095305e+00 -5.98373890e-01 3.47176120e-02 3.23149055e-01
4.78960246e-01 -1.33092492e-03 6.03734672e-01 8.42079222e-02
-8.73444855e-01 8.03796768e-01 4.64679360e-01 2.27220058e-01
4.49573249e-02 -1.55448139e+00 3.30366939e-01 7.54549682e-01
6.47762716e-01 1.17726970e+00 9.68830407e-01 -3.54154885e-01
-1.68737566e+00 -7.26333261e-01 4.74494189e-01 -7.30420589e-01
6.83207452e-01 -5.91449916e-01 -8.64507675e-01 1.20006597e+00
2.29263574e-01 -1.16285160e-01 5.71912766e-01 7.44796455e-01
-5.25785804e-01 1.45445466e-01 -8.11221063e-01 7.41966605e-01
1.27200818e+00 -8.16043019e-01 -8.88606012e-01 1.98628336e-01
7.32140303e-01 -4.79993552e-01 -9.48937416e-01 3.85472119e-01
9.12142396e-01 -8.10764015e-01 1.22709715e+00 -8.44576478e-01
7.22161412e-01 -1.08251646e-02 -3.84464025e-01 -1.28126144e+00
-2.42954567e-01 -5.57850301e-01 -1.10074627e+00 8.96210194e-01
3.10428202e-01 -5.75503528e-01 6.99497044e-01 4.61100757e-01
-8.48203301e-02 -6.87932789e-01 -9.62725580e-01 -6.26859009e-01
4.59940702e-01 -9.22583342e-01 4.76227313e-01 5.44396579e-01
-1.87271044e-01 4.00709599e-01 -2.90991604e-01 2.44353563e-01
1.17400682e+00 -1.55648217e-01 5.88371456e-01 -1.15171218e+00
-7.92126536e-01 -8.19612861e-01 -2.53299847e-02 -1.71709168e+00
6.62938237e-01 -9.56149518e-01 2.19772115e-01 -1.30908179e+00
5.05375683e-01 6.54414622e-03 -4.22455013e-01 1.51237190e-01
-3.66864234e-01 4.06440608e-02 -2.67132580e-01 -2.28876263e-01
-5.65325916e-01 8.55870664e-01 7.28202164e-01 -2.37784132e-01
2.02787891e-01 6.88938722e-02 -5.86276174e-01 6.50163710e-01
4.00520205e-01 -4.27355796e-01 -5.42300880e-01 -6.63208127e-01
7.75165498e-01 -2.20687926e-01 6.19653225e-01 -5.31234205e-01
2.93172866e-01 -1.73736230e-01 4.38686371e-01 -7.83774078e-01
3.50881815e-01 -2.04233572e-01 -1.16272412e-01 3.20230365e-01
-5.76047361e-01 -3.41294676e-01 1.53657898e-01 9.44966495e-01
1.10089697e-01 -4.58892703e-01 5.67947149e-01 -4.12817225e-02
-5.67804337e-01 5.11244416e-01 -5.26166618e-01 3.64888281e-01
2.37529889e-01 4.70822930e-01 -1.16920054e-01 -4.14264053e-01
-9.69297051e-01 3.63713712e-01 -2.41311848e-01 2.60447692e-02
7.47354805e-01 -1.20602036e+00 -8.15845966e-01 3.14427540e-02
-1.55122340e-01 1.46723777e-01 1.47297466e-02 7.91622996e-01
-7.91179091e-02 4.93936867e-01 4.29817438e-01 -7.34545231e-01
-5.06095231e-01 3.28746378e-01 4.49579686e-01 -1.01104856e-01
-5.60773969e-01 1.23764741e+00 1.88779891e-01 -3.44019592e-01
5.74101150e-01 -5.18224716e-01 1.69873878e-01 -2.94520885e-01
5.97218990e-01 3.55090201e-01 -4.89432991e-01 1.03416607e-01
-2.31467053e-01 2.25901663e-01 -2.91012406e-01 -5.29158592e-01
1.34849942e+00 -2.18913600e-01 -5.39701097e-02 9.37349319e-01
1.27742815e+00 -3.54498237e-01 -1.63884354e+00 -5.70767522e-01
-5.93148731e-02 -3.47515225e-01 5.44921041e-01 -5.57104290e-01
-1.01524615e+00 1.07043922e+00 1.71317324e-01 -8.93106610e-02
6.37589335e-01 3.37179899e-01 4.05302256e-01 8.78046632e-01
1.70026615e-01 -1.12722898e+00 1.49053121e-02 7.42405951e-01
6.20325327e-01 -1.31198311e+00 2.42165491e-01 1.00457959e-01
-1.10128149e-01 9.65761244e-01 2.19892338e-01 -3.04192692e-01
1.16554296e+00 3.47849488e-01 -7.06276178e-01 -3.43564779e-01
-1.18167770e+00 -9.77687836e-02 3.40937108e-01 4.88479376e-01
4.60159689e-01 -5.88988066e-02 -1.28948629e-01 6.55963838e-01
-3.15003306e-01 2.29632840e-01 1.53922737e-01 5.86873889e-01
-2.90756643e-01 -7.40602553e-01 6.72147274e-02 6.19735837e-01
-4.93443996e-01 -3.84414017e-01 1.27500221e-01 5.77738523e-01
-3.13837647e-01 7.29935110e-01 4.15171742e-01 -1.09678634e-01
-2.76252598e-01 2.43990794e-01 1.10264397e+00 -3.60516310e-01
8.90121162e-02 -3.11203480e-01 -4.53276634e-02 -5.48241436e-01
-3.01693290e-01 -8.76603067e-01 -8.15758705e-01 -5.29734194e-01
-1.90451235e-01 1.85145736e-01 6.25524163e-01 1.47205329e+00
1.24479055e-01 7.28064060e-01 2.72848666e-01 -7.44636416e-01
-1.06010568e+00 -1.16438878e+00 -5.02499819e-01 -7.61987045e-02
6.39253318e-01 -9.92404521e-01 -5.88744760e-01 -4.13109735e-02] | [7.001053333282471, 3.9602222442626953] |
0e6fe21e-d11b-40c4-9bcd-fd4ec9cfdb87 | hyperbolic-manifold-regression | 2005.13885 | null | https://arxiv.org/abs/2005.13885v1 | https://arxiv.org/pdf/2005.13885v1.pdf | Hyperbolic Manifold Regression | Geometric representation learning has recently shown great promise in several machine learning settings, ranging from relational learning to language processing and generative models. In this work, we consider the problem of performing manifold-valued regression onto an hyperbolic space as an intermediate component for a number of relevant machine learning applications. In particular, by formulating the problem of predicting nodes of a tree as a manifold regression task in the hyperbolic space, we propose a novel perspective on two challenging tasks: 1) hierarchical classification via label embeddings and 2) taxonomy extension of hyperbolic representations. To address the regression problem we consider previous methods as well as proposing two novel approaches that are computationally more advantageous: a parametric deep learning model that is informed by the geodesics of the target space and a non-parametric kernel-method for which we also prove excess risk bounds. Our experiments show that the strategy of leveraging the hyperbolic geometry is promising. In particular, in the taxonomy expansion setting, we find that the hyperbolic-based estimators significantly outperform methods performing regression in the ambient Euclidean space. | ['Gian Maria Marconi', 'Carlo Ciliberto', 'Lorenzo Rosasco'] | 2020-05-28 | null | null | null | null | ['taxonomy-expansion'] | ['natural-language-processing'] | [ 4.23765704e-02 7.66365588e-01 -1.42959848e-01 -2.57076770e-01
-9.03386831e-01 -5.02260089e-01 5.76504707e-01 1.88881516e-01
-1.62215322e-01 1.71848685e-01 1.75338671e-01 -5.17068684e-01
-5.69938779e-01 -8.79765689e-01 -4.42971617e-01 -9.52951968e-01
-2.18971774e-01 4.93124545e-01 -3.36724780e-02 -7.16070607e-02
4.43241239e-01 6.40408814e-01 -1.17118549e+00 -5.05060017e-01
7.75324166e-01 1.09375286e+00 -5.71754217e-01 5.71208119e-01
2.76141793e-01 5.86426497e-01 1.25771705e-02 -7.07770050e-01
3.25206935e-01 -1.70603931e-01 -9.87996280e-01 -4.56689820e-02
2.86836892e-01 6.58800974e-02 -3.26015949e-01 9.24797714e-01
2.56445140e-01 5.14287293e-01 1.34565508e+00 -1.58149588e+00
-9.15617883e-01 5.72090507e-01 -7.12662339e-01 -5.78749478e-02
-8.22964981e-02 -5.42732656e-01 1.27217948e+00 -1.18570387e+00
3.15652132e-01 1.29549110e+00 9.58985627e-01 3.35191011e-01
-1.50401616e+00 -3.42534453e-01 -1.26349285e-01 4.57414687e-02
-1.64984965e+00 8.89490545e-02 8.67644966e-01 -9.41878080e-01
1.33802265e-01 1.49282351e-01 6.20220639e-02 1.04614055e+00
-7.34306546e-03 4.95885760e-01 8.86081159e-01 -4.93621618e-01
3.97011250e-01 2.28691563e-01 2.89201826e-01 9.36839879e-01
1.10047683e-01 -1.45695973e-02 -2.99588084e-01 -4.74520385e-01
6.83192432e-01 -4.13978565e-03 -2.98209153e-02 -9.55199897e-01
-9.30479348e-01 1.70815563e+00 7.00782835e-01 6.88711787e-03
-1.36975884e-01 3.86921972e-01 1.52313516e-01 -4.14973497e-02
8.93115461e-01 5.27293682e-01 -1.58211604e-01 3.75434369e-01
-5.86478889e-01 7.60565847e-02 8.10314655e-01 9.77967024e-01
6.37723804e-01 -3.36754620e-01 1.58671215e-01 6.80598259e-01
6.65135801e-01 -6.63983300e-02 5.27829155e-02 -9.36102808e-01
3.12899530e-01 4.24048781e-01 -2.51851678e-01 -9.85786378e-01
-6.35766625e-01 -3.86238247e-01 -7.78117180e-01 1.07342750e-01
7.05638111e-01 3.65439197e-03 -9.49712023e-02 1.91891980e+00
7.01032162e-01 3.72666597e-01 -3.64888795e-02 5.05911589e-01
2.01595157e-01 6.46520019e-01 1.45694718e-01 -2.96104718e-02
1.15212929e+00 -7.83713460e-01 -1.29451960e-01 4.97752845e-01
1.30799234e+00 -3.57176632e-01 9.73211288e-01 3.83522898e-01
-7.98529565e-01 -1.30948171e-01 -1.15887940e+00 -4.41688418e-01
-5.25279760e-01 -1.63664054e-02 6.64988160e-01 7.33043134e-01
-1.14457846e+00 8.16858411e-01 -7.20393181e-01 -6.21966302e-01
3.35290045e-01 2.38582522e-01 -1.17184609e-01 6.61129802e-02
-7.83196151e-01 8.10104132e-01 2.83000828e-03 -2.45070085e-01
-4.24259543e-01 -8.13600421e-01 -9.89955723e-01 -1.91699322e-02
2.39710212e-01 -5.54463089e-01 1.13315415e+00 -1.65817782e-01
-1.35464060e+00 8.81088793e-01 3.68308991e-01 -5.14136434e-01
6.43931091e-01 -1.45752057e-01 1.71184480e-01 1.61052689e-01
-8.64723921e-02 3.99423391e-01 9.44396138e-01 -7.65974462e-01
-3.66758138e-01 -1.01760507e+00 -4.51760255e-02 -1.29157268e-02
-5.42767048e-01 -1.75699621e-01 2.81982690e-01 -7.26795733e-01
3.29478621e-01 -1.24889457e+00 -3.34874779e-01 4.32245731e-01
-5.53905129e-01 -8.74639273e-01 6.34003639e-01 -5.07168174e-01
9.43598926e-01 -2.21703959e+00 6.03357494e-01 3.39375407e-01
4.68865842e-01 -3.31283003e-01 8.66215751e-02 3.79364967e-01
7.65461624e-02 4.08121645e-01 -5.32276273e-01 -5.30585408e-01
1.58581480e-01 -2.82910466e-01 -7.08858252e-01 1.00751817e+00
3.40791643e-01 8.00296962e-01 -9.26968098e-01 -4.41172212e-01
-7.35068247e-02 6.57192647e-01 -5.71686149e-01 2.12297827e-01
-8.24525356e-02 4.07422900e-01 -6.07474446e-01 2.75459945e-01
5.60377419e-01 -1.22970268e-01 -1.23100802e-01 1.23207696e-01
1.07790641e-01 2.65881032e-01 -9.35396135e-01 1.69249368e+00
-7.18399405e-01 6.15773380e-01 -1.50505915e-01 -1.16739488e+00
1.14768076e+00 8.17124620e-02 4.51981187e-01 1.82802632e-01
9.44119617e-02 1.09785341e-01 -4.05787200e-01 -2.72099942e-01
4.86803889e-01 -3.84321034e-01 -2.65556842e-01 5.84396720e-01
-5.64769609e-04 -1.23475231e-01 -5.39083362e-01 2.99158990e-01
1.02836812e+00 2.18151852e-01 3.27858835e-01 -4.64629799e-01
3.54110658e-01 -3.08485895e-01 1.71157852e-01 4.39671785e-01
-1.97173599e-02 7.58779407e-01 8.49184871e-01 -9.09244940e-02
-1.00636113e+00 -1.42219436e+00 -6.12362027e-01 1.37746906e+00
-9.57310870e-02 -4.96337295e-01 -7.64344215e-01 -8.68936956e-01
4.02366817e-02 9.56651926e-01 -1.08174026e+00 -6.11791790e-01
-4.13162261e-01 -7.54031718e-01 5.44721603e-01 7.45909631e-01
5.94737865e-02 -3.35075110e-01 -2.90338010e-01 -2.06120357e-01
2.69402772e-01 -9.67215240e-01 -4.95545030e-01 1.97549462e-01
-1.02914774e+00 -1.07338881e+00 -6.29918754e-01 -6.46590889e-01
3.18137646e-01 1.20114967e-01 6.38258576e-01 -2.09493428e-01
-4.09846395e-01 8.70391250e-01 -3.03950697e-01 -7.53125474e-02
-2.83783972e-01 5.66417396e-01 2.25771278e-01 2.85207927e-01
3.49908441e-01 -7.39419758e-01 -4.77414697e-01 3.28244925e-01
-7.80749559e-01 -2.53359586e-01 2.79935151e-01 4.96756852e-01
2.76346445e-01 -3.37357283e-01 7.34717607e-01 -8.26331735e-01
4.21441942e-01 -1.21886039e+00 -6.19626224e-01 2.23904118e-01
-7.67624795e-01 3.15906703e-01 5.89446545e-01 -4.83489245e-01
-6.45563304e-01 6.00183271e-02 2.63462991e-01 -2.60717958e-01
2.66191423e-01 4.01914418e-01 -9.35058445e-02 -7.97587819e-03
8.18706453e-01 -2.01991707e-01 9.31550488e-02 -5.84769785e-01
8.12381268e-01 6.49542511e-01 4.42019761e-01 -7.03941286e-01
1.20369089e+00 4.41981047e-01 7.01891124e-01 -9.72053885e-01
-9.67104375e-01 -4.98309195e-01 -1.11492872e+00 6.77392632e-02
1.20797896e+00 -6.38471067e-01 -5.66902578e-01 -1.91354245e-01
-9.54062879e-01 -1.78175554e-01 -4.17883933e-01 5.73198378e-01
-1.00450110e+00 2.13575959e-01 -5.55185318e-01 -1.11590087e+00
-3.78024280e-02 -8.13326955e-01 1.22553909e+00 -6.71723811e-03
-1.31823167e-01 -1.37497056e+00 4.72443283e-01 2.85898834e-01
1.14630386e-01 4.00105566e-01 1.33909142e+00 -1.06356514e+00
-6.18238151e-01 -2.86701202e-01 -2.70611316e-01 1.33056551e-01
-1.59779787e-01 -2.52716895e-02 -1.07716191e+00 -1.27870470e-01
2.02346653e-01 -3.19751859e-01 9.23044860e-01 1.78467840e-01
1.30439687e+00 -2.02828541e-01 -2.29482859e-01 9.63156402e-01
1.15086830e+00 -4.09065843e-01 3.75956833e-01 4.70163859e-02
8.42360020e-01 1.15580595e+00 4.75231916e-01 3.29516590e-01
5.34354568e-01 8.00994277e-01 4.90978986e-01 2.51764774e-01
2.63065159e-01 -5.50218999e-01 3.12285185e-01 5.26539445e-01
1.76923916e-01 2.23968521e-01 -7.89961278e-01 1.77956775e-01
-1.90458167e+00 -7.18631089e-01 -1.70274869e-01 2.58194208e+00
4.32375371e-01 -3.77901554e-01 5.85884690e-01 7.98249692e-02
7.49903977e-01 6.62179757e-03 -5.50378978e-01 -3.83655429e-01
2.72477090e-01 1.34184703e-01 2.97433048e-01 5.43814957e-01
-1.33969355e+00 7.18288660e-01 5.98352146e+00 5.21275043e-01
-6.49985909e-01 1.72484502e-01 6.96003497e-01 2.44024754e-01
-2.15153828e-01 -5.30106388e-03 -8.14725697e-01 -4.98670377e-02
1.17870128e+00 -3.21064025e-01 3.46880347e-01 1.27825844e+00
-2.44139001e-01 3.30763876e-01 -1.56722784e+00 8.44646871e-01
1.27314776e-01 -1.08511102e+00 -1.32475495e-01 6.43945336e-01
3.94074678e-01 -1.50622040e-01 7.18065441e-01 3.91897738e-01
3.90113890e-01 -1.35382080e+00 4.91669357e-01 2.96418816e-01
7.51695573e-01 -8.74786079e-01 1.52876452e-01 3.15858603e-01
-1.21179485e+00 -1.04304515e-01 -4.69743818e-01 1.79168776e-01
-2.85237551e-01 2.77226686e-01 -1.00801694e+00 4.14008498e-01
3.14744830e-01 8.00025105e-01 -7.17110872e-01 1.02663434e+00
-4.05395478e-02 7.36329675e-01 -3.19415301e-01 6.86575845e-02
8.74903202e-02 -7.17381001e-01 5.07753253e-01 9.95555937e-01
4.90170747e-01 -2.02526851e-03 -1.68638766e-01 1.26912272e+00
-1.90917224e-01 4.78556961e-01 -1.09312224e+00 1.96185932e-01
3.65399569e-01 1.45196116e+00 -8.47328126e-01 3.52117628e-01
-3.07266653e-01 7.48751640e-01 7.41024196e-01 2.54220098e-01
-8.17221165e-01 -2.66868621e-01 5.86457610e-01 2.35064462e-01
2.51272291e-01 -4.94160622e-01 -3.31833392e-01 -1.06197023e+00
-9.17729661e-02 -8.62526614e-03 5.25455177e-01 -4.61428225e-01
-1.38408446e+00 3.41572374e-01 1.60915896e-01 -9.71066535e-01
-5.01700878e-01 -8.76435578e-01 -6.86640441e-01 5.25187016e-01
-1.03186822e+00 -1.20427597e+00 1.16633121e-02 4.95564580e-01
2.22701311e-01 8.30934942e-03 7.34302938e-01 -9.74323973e-02
-6.94924653e-01 6.97146595e-01 3.91247213e-01 5.75290993e-02
3.63282293e-01 -1.70068502e+00 2.26693898e-01 5.78543186e-01
4.51789111e-01 4.62260872e-01 4.22890931e-01 -1.98438287e-01
-1.36310768e+00 -1.36380112e+00 6.23310447e-01 -9.02005374e-01
1.06737983e+00 -7.50385165e-01 -1.00738370e+00 9.73989964e-01
-6.45734787e-01 1.79975048e-01 1.06197119e+00 3.32697332e-01
-8.47744226e-01 1.70252964e-01 -1.17558491e+00 6.44247770e-01
1.17588687e+00 -7.00023711e-01 -3.19282711e-01 5.34986317e-01
9.74970996e-01 2.81292021e-01 -1.03090394e+00 1.67876229e-01
2.57100105e-01 -6.14668131e-01 1.04937434e+00 -9.96381879e-01
4.38857883e-01 6.58343956e-02 -7.23401487e-01 -1.18388104e+00
-2.09036559e-01 -8.39110017e-01 -3.67607415e-01 1.29970706e+00
2.86687970e-01 -8.00686717e-01 8.14197600e-01 6.40281558e-01
5.15832342e-02 -1.13329899e+00 -1.15230513e+00 -7.91025817e-01
9.19439793e-01 -3.62386525e-01 1.93592101e-01 8.81822467e-01
1.98818102e-01 5.99448562e-01 -1.15860470e-01 2.31543824e-01
9.38310266e-01 -1.71194464e-01 7.29276061e-01 -1.89511275e+00
-1.52334109e-01 -4.42559987e-01 -8.13092411e-01 -1.04625702e+00
6.52854383e-01 -1.39839220e+00 -1.22533873e-01 -9.67117667e-01
1.92274481e-01 -7.03737557e-01 3.79304364e-02 -2.61679173e-01
1.06302381e-01 -7.91174732e-03 -1.54994624e-02 1.74538448e-01
-2.76456088e-01 9.68968093e-01 5.33260882e-01 6.13438934e-02
-6.20290935e-02 5.39501250e-01 -8.65542650e-01 7.76891470e-01
6.96738183e-01 -5.02903700e-01 -5.15964329e-01 9.64190438e-02
3.36665362e-01 -1.02503769e-01 6.51062012e-01 -6.34429812e-01
2.78883636e-01 6.53167292e-02 -3.57308313e-02 -2.50909805e-01
5.62780440e-01 -5.60962200e-01 -4.64721948e-01 2.13002220e-01
-7.60739386e-01 8.38830881e-03 -4.98795301e-01 1.03939342e+00
3.71241242e-01 -5.07590711e-01 9.02996838e-01 5.64869046e-01
5.92339374e-02 5.34432292e-01 -1.51626185e-01 4.25301254e-01
1.22542000e+00 2.12767981e-02 -1.62227005e-01 -4.68677044e-01
-7.68518448e-01 -8.66339803e-02 4.10871416e-01 3.35489362e-01
5.61828852e-01 -1.37642276e+00 -7.36813307e-01 -1.26565561e-01
2.15747237e-01 -6.75076665e-03 -3.02791238e-01 1.08639622e+00
-1.67519122e-01 3.22969854e-01 4.66341913e-01 -5.91787815e-01
-7.59708166e-01 9.07275319e-01 4.33883071e-01 -1.38914853e-01
-8.10055256e-01 7.79207408e-01 6.08181417e-01 -5.69934368e-01
4.38370764e-01 -3.03933144e-01 -1.28410488e-01 1.74820885e-01
2.78730512e-01 7.90125012e-01 -9.46343541e-02 -7.55514622e-01
-1.39984176e-01 7.29992986e-01 5.24856634e-02 -4.60016340e-01
1.32825494e+00 -9.82284248e-02 -7.50305504e-02 8.33800435e-01
1.67082262e+00 -2.22273856e-01 -1.19614148e+00 -5.13571858e-01
6.32720947e-01 -1.51723951e-01 -1.36780776e-02 3.48956808e-02
-7.36277163e-01 1.12775230e+00 3.46609324e-01 5.73308170e-01
6.93479300e-01 5.35861731e-01 3.79610270e-01 3.98027599e-01
2.99802303e-01 -8.30989122e-01 3.64115328e-01 1.98339105e-01
9.41876471e-01 -1.08055830e+00 -1.74711064e-01 -7.20972359e-01
-3.48568678e-01 1.22442615e+00 1.77272335e-01 -4.56831783e-01
1.12852621e+00 -3.49326938e-01 -4.83438879e-01 -1.31034613e-01
-6.60495520e-01 -2.76753865e-02 4.59359109e-01 8.09869885e-01
3.22406024e-01 5.90849183e-02 1.48901120e-01 6.69411957e-01
-5.29035568e-01 -7.06998408e-01 6.74031556e-01 3.89352381e-01
-3.31014752e-01 -7.45613754e-01 -2.90044576e-01 2.82886177e-01
-2.18779624e-01 -4.74772081e-02 -6.48095071e-01 8.13975632e-01
-4.29944515e-01 8.71389568e-01 1.61035001e-01 -3.59210074e-01
-9.85923111e-02 3.25242937e-01 4.91285980e-01 -7.15068758e-01
-7.17001110e-02 -3.16843718e-01 -3.06466728e-01 -3.44277769e-01
1.06980197e-01 -9.05428886e-01 -1.08076286e+00 -2.10942969e-01
-3.43404680e-01 2.43899405e-01 8.99784744e-01 8.56158197e-01
1.17739595e-01 5.36695309e-02 1.04345727e+00 -7.51933873e-01
-1.43967271e+00 -9.50919688e-01 -9.30374265e-01 1.13412507e-01
3.92023653e-01 -9.62315857e-01 -8.27358305e-01 -3.90221715e-01] | [8.053925514221191, 4.119668483734131] |
865a5273-d93d-443f-b88d-96554469364f | automated-reasoning-in-non-classical-logics | 2202.09836 | null | https://arxiv.org/abs/2202.09836v1 | https://arxiv.org/pdf/2202.09836v1.pdf | Automated Reasoning in Non-classical Logics in the TPTP World | Non-classical logics are used in a wide spectrum of disciplines, including artificial intelligence, computer science, mathematics, and philosophy. The de-facto standard infrastructure for automated theorem proving, the TPTP World, currently supports only classical logics. Similar standards for non-classical logic reasoning do not exist (yet). This hampers practical development of reasoning systems, and limits their interoperability and application. This paper describes the latest extension of the TPTP World, which provides languages and infrastructure for reasoning in non-classical logics. The extensions integrate seamlessly with the existing TPTP World. | ['Christoph Benzmüller', 'Geoff Sutcliffe', 'Tobias Gleißner', 'David Fuenmayor', 'Alexander Steen'] | 2022-02-20 | null | null | null | null | ['automated-theorem-proving', 'automated-theorem-proving'] | ['miscellaneous', 'reasoning'] | [-2.09216446e-01 4.42551643e-01 -4.78552788e-01 -2.44607687e-01
-2.17992365e-01 -9.65503275e-01 8.07741940e-01 3.65593806e-02
-6.43718392e-02 1.26628304e+00 -1.96128145e-01 -1.09499061e+00
-4.18515086e-01 -1.35189712e+00 -2.48661518e-01 -1.16576537e-01
-8.50714669e-02 6.43071651e-01 8.10234070e-01 -4.85299438e-01
2.73486413e-02 4.48521256e-01 -1.52188182e+00 6.40097499e-01
6.20823026e-01 9.45551097e-01 -6.21865273e-01 4.48353857e-01
-4.22487587e-01 1.54607284e+00 -8.69092718e-03 -5.76211631e-01
1.09338507e-01 -2.67958969e-01 -1.56291902e+00 -6.03200436e-01
8.11253712e-02 -3.12629968e-01 -5.92724741e-01 1.22844994e+00
-3.33436519e-01 -2.56978661e-01 4.07446586e-02 -1.92313755e+00
-4.40708190e-01 1.00669408e+00 3.50828357e-02 -1.38773382e-01
9.09927011e-01 -4.81253490e-02 1.34471929e+00 -2.00628750e-02
9.61966753e-01 1.49600530e+00 6.74343467e-01 5.11364758e-01
-9.94471908e-01 -3.95397931e-01 -4.02912438e-01 9.80471969e-01
-1.66235828e+00 -3.62607419e-01 5.19993961e-01 -1.31705469e-02
1.12856317e+00 6.61874413e-01 4.95290875e-01 4.63239908e-01
5.49651146e-01 5.03635526e-01 1.22293055e+00 -9.19925213e-01
2.98345834e-01 5.40252626e-01 2.76844680e-01 3.39767516e-01
6.24915481e-01 2.14309450e-02 -3.32492739e-01 -4.05781657e-01
6.09833002e-01 -3.99623334e-01 1.68245658e-01 -4.70908999e-01
-1.21097124e+00 6.49352670e-01 -3.34703684e-01 8.65392923e-01
3.84128396e-03 8.13673139e-02 8.46637011e-01 7.73881376e-01
-3.19513351e-01 3.71643394e-01 -5.45741498e-01 -3.25428367e-01
-6.64335430e-01 8.68975997e-01 1.38957846e+00 1.31551635e+00
6.18826985e-01 -3.45427960e-01 1.42436936e-01 -1.65389836e-01
6.75053775e-01 5.18422484e-01 -3.32687348e-01 -1.80622625e+00
3.79379727e-02 8.22197735e-01 4.91675109e-01 -8.90046477e-01
-1.64678141e-01 2.96156377e-01 -2.25708395e-01 2.22667888e-01
5.40052474e-01 1.41499788e-01 5.66941239e-02 1.25637233e+00
3.03634524e-01 -2.18561172e-01 7.54325807e-01 4.66802448e-01
7.67112672e-01 5.17264962e-01 -2.14555319e-02 -5.64144492e-01
1.20651472e+00 -3.05629790e-01 -1.12798870e+00 1.51013926e-01
6.67319655e-01 -4.81706530e-01 4.19627547e-01 8.35544109e-01
-1.36267352e+00 3.43586862e-01 -8.52828205e-01 -2.53836811e-01
-7.84348190e-01 -8.46725106e-01 1.27748179e+00 4.94348407e-01
-1.19338334e+00 1.59318715e-01 -6.12457216e-01 -4.32275087e-01
2.73552269e-01 4.41342264e-01 -5.18785834e-01 -5.08426189e-01
-1.76259744e+00 1.44924343e+00 1.10437107e+00 -2.48173028e-01
-3.18917662e-01 -4.10094410e-01 -9.17271912e-01 -2.06757169e-02
8.98037314e-01 -3.40460837e-01 1.52533138e+00 -4.73747611e-01
-1.55862367e+00 7.31025517e-01 2.09923834e-02 -8.74396384e-01
4.26286757e-01 2.45625600e-01 -1.04637849e+00 2.44851917e-01
3.05743217e-01 1.26401737e-01 -2.46085778e-01 -8.83774042e-01
-8.74456704e-01 -1.99200824e-01 8.41400921e-01 -2.68195838e-01
4.03170675e-01 4.41887856e-01 -8.61619189e-02 4.49925005e-01
2.02256665e-01 -4.97218817e-01 2.56363656e-02 3.29657979e-02
6.92525208e-02 -4.47855115e-01 1.04980195e+00 1.95701554e-01
1.17299700e+00 -1.71640372e+00 -2.08924845e-01 4.87834156e-01
5.83451279e-02 5.34726307e-02 4.92786855e-01 9.89963830e-01
4.11639750e-01 2.03877628e-01 2.16312870e-01 8.14510465e-01
8.49460244e-01 7.91853070e-01 -5.27598560e-01 2.01689422e-01
-7.50459880e-02 7.25770354e-01 -1.13564289e+00 -1.23574030e+00
7.06943274e-01 -1.55158639e-01 -4.82512712e-01 -4.90230471e-01
-8.43440354e-01 -2.03451365e-01 -7.37943351e-01 8.28017294e-01
7.73573935e-01 -8.43857303e-02 9.43692148e-01 1.41108692e-01
-6.80208206e-01 4.28472042e-01 -1.47473419e+00 1.57600749e+00
-1.26455262e-01 4.89222586e-01 2.94846296e-01 -7.87908971e-01
4.13496643e-01 9.47116435e-01 2.44178161e-01 -4.50467587e-01
1.84755698e-01 5.30248642e-01 5.99445738e-02 -6.49529696e-01
3.45199823e-01 -6.41954243e-01 -2.58740038e-01 3.78515750e-01
-1.03997178e-02 -6.30405605e-01 9.97284591e-01 1.41939297e-01
1.08390594e+00 3.72251689e-01 7.51101971e-01 -4.77000445e-01
1.08437121e+00 8.09000969e-01 8.22421670e-01 5.89320362e-01
-3.05049211e-01 -4.51146007e-01 8.40827227e-01 -7.13385165e-01
-7.46344447e-01 -9.71268594e-01 -4.97251153e-01 7.00458169e-01
2.70755470e-01 -9.07923341e-01 -3.30673784e-01 -4.78357196e-01
1.54422317e-03 1.16775274e+00 3.53761166e-02 1.47918865e-01
-1.34001538e-01 9.66085494e-02 1.03305924e+00 2.94721663e-01
6.77678764e-01 -1.04969072e+00 -5.98802090e-01 1.25755116e-01
-3.65278363e-01 -1.39190578e+00 7.57171452e-01 -1.78046465e-01
-6.03453040e-01 -1.54864728e+00 8.04742098e-01 -3.45398992e-01
4.39364687e-02 -2.04061702e-01 9.70077574e-01 5.11449650e-02
1.29479110e-01 4.13452864e-01 -3.44390422e-01 -5.37782073e-01
-7.50077188e-01 -4.77646232e-01 -6.16637012e-03 -7.24230111e-01
8.54337335e-01 -2.32545450e-01 2.62107462e-01 1.56353191e-01
-1.03188825e+00 -1.06219269e-01 3.10446322e-01 2.79513657e-01
2.01631561e-01 9.80748951e-01 3.50863636e-01 -9.35796738e-01
4.37545449e-01 -1.94590688e-01 -9.00488317e-01 6.65805161e-01
-6.10937297e-01 -1.17875552e-02 8.79137337e-01 2.60880172e-01
-1.00839865e+00 -6.79437995e-01 1.26805544e-01 -9.24153998e-02
-4.82149720e-01 9.81452644e-01 -4.73197132e-01 -4.77013066e-02
3.20719779e-01 -5.51561303e-02 -5.32645956e-02 3.18043858e-01
2.38068566e-01 6.30328238e-01 5.87624490e-01 -1.04452932e+00
7.25106895e-01 5.07754147e-01 5.84058344e-01 -4.46221054e-01
-5.95679402e-01 -6.94310516e-02 -4.58275497e-01 -8.66686255e-02
4.09513384e-01 -3.63518149e-01 -1.22925973e+00 -6.56666830e-02
-1.32258129e+00 -2.32453957e-01 -5.47627866e-01 6.49536908e-01
-7.39149094e-01 4.01100636e-01 -5.40298522e-01 -1.13745487e+00
-8.51959586e-02 -7.69418180e-01 3.10034841e-01 6.08281977e-02
-4.97634947e-01 -1.14111686e+00 1.07673533e-01 3.26940149e-01
2.74322182e-01 2.51962721e-01 1.06031597e+00 -5.72548866e-01
-4.26012516e-01 -4.55138743e-01 -2.98471212e-01 3.27662438e-01
-8.98597911e-02 4.88904476e-01 -6.23192430e-01 2.36478820e-01
-2.88110822e-01 -3.88826311e-01 -4.16740894e-01 5.64277731e-02
4.06730264e-01 -3.24079573e-01 -3.14740568e-01 -3.24045569e-01
1.54979086e+00 4.33681488e-01 1.01450157e+00 8.05705607e-01
-2.37168759e-01 4.60980654e-01 9.09504414e-01 2.69793570e-01
7.69683897e-01 4.52591509e-01 1.50357813e-01 5.64770997e-01
6.05160773e-01 1.36016145e-01 1.55788541e-01 2.28614137e-01
-4.37381655e-01 4.07401174e-01 -1.45232534e+00 3.56035829e-01
-2.03589773e+00 -1.74332416e+00 -3.30293864e-01 1.76700962e+00
1.13216937e+00 3.61653835e-01 -2.54419684e-01 6.45873010e-01
3.62192303e-01 -4.47548032e-01 6.29318133e-03 -8.67307007e-01
-6.18135817e-02 3.34008157e-01 1.59562990e-01 5.80868185e-01
-9.17267740e-01 1.12845647e+00 7.39721775e+00 4.31064487e-01
-7.68978059e-01 2.00505834e-02 -5.51306069e-01 3.75833243e-01
-3.51566046e-01 6.99892163e-01 -3.71267825e-01 -1.94729656e-01
1.20925593e+00 -8.08889806e-01 6.47316039e-01 7.20337689e-01
1.28341869e-01 -2.46952400e-01 -1.22141647e+00 4.94865716e-01
-2.92753667e-01 -1.61900115e+00 -9.65327546e-02 -1.74252674e-01
3.02660882e-01 -2.17323273e-01 -6.30247295e-01 5.66999555e-01
4.37500119e-01 -7.33428299e-01 9.90126908e-01 5.27757108e-01
4.27508563e-01 -1.01861358e+00 1.28358471e+00 1.55594051e-01
-9.32399273e-01 -1.19391270e-01 -2.31640950e-01 -6.53796732e-01
8.34991597e-03 3.08203101e-01 -6.82885110e-01 1.19032049e+00
7.94530153e-01 4.57929015e-01 -9.45325792e-02 8.63692880e-01
-1.72336116e-01 1.97723836e-01 -3.35147381e-01 6.09972887e-02
2.02833384e-01 -2.63064474e-01 4.86927450e-01 1.06554198e+00
-1.01393759e-01 2.43945941e-01 2.13808641e-01 9.83244538e-01
3.68806154e-01 -3.68053555e-01 -7.69123852e-01 -3.18230391e-01
6.66098177e-01 8.41931283e-01 -5.54437399e-01 -5.67035615e-01
-8.94189119e-01 2.36376585e-03 -5.01254141e-01 1.61827177e-01
-8.22955072e-01 -6.23822272e-01 4.02111560e-01 2.04961374e-01
-1.24814168e-01 -1.94276452e-01 -3.05384725e-01 -1.05112672e+00
-8.81065130e-02 -1.23975241e+00 6.86051965e-01 -8.08479905e-01
-1.17041659e+00 3.15746665e-01 8.31701517e-01 -9.72977400e-01
-5.31313539e-01 -8.56181383e-01 -4.36339974e-01 5.44341981e-01
-1.45880926e+00 -1.46851075e+00 1.16718553e-01 7.02354252e-01
-5.59123695e-01 1.63585991e-01 1.38106334e+00 1.07118994e-01
-2.23658666e-01 -4.61773276e-02 -3.63099307e-01 -1.05688060e-02
6.04801595e-01 -1.28135073e+00 -2.76642531e-01 8.11710000e-01
-4.43510950e-01 8.64830196e-01 1.21978438e+00 -3.73702824e-01
-2.13894558e+00 -6.62400067e-01 1.21066809e+00 -1.19996063e-01
1.37141466e+00 4.17933092e-02 -5.31169713e-01 1.27315176e+00
2.96888173e-01 4.67274934e-02 5.44951618e-01 3.20069820e-01
-6.21961534e-01 -5.21768689e-01 -1.58565331e+00 7.79914498e-01
4.94617373e-01 -9.55882609e-01 -1.13041806e+00 3.04427505e-01
4.03341562e-01 -2.61597037e-01 -1.30829501e+00 4.59847659e-01
7.08145201e-01 -1.00767756e+00 5.84166646e-01 -5.54550648e-01
-2.63187904e-02 -9.13170040e-01 -4.74169672e-01 -2.38323346e-01
4.19552587e-02 -8.77815366e-01 -2.97055215e-01 1.06004727e+00
3.77908349e-01 -1.15345633e+00 4.96304601e-01 1.25573599e+00
-8.59926715e-02 -2.38282442e-01 -1.15154696e+00 -7.68889844e-01
3.83779824e-01 -1.04726958e+00 8.65481734e-01 1.08209550e+00
1.31348121e+00 1.37569621e-01 4.91440654e-01 9.61243436e-02
5.50042808e-01 5.50589681e-01 6.09323978e-01 -1.55353475e+00
-2.07495894e-02 -5.04914463e-01 -8.90315711e-01 -1.46446317e-01
3.96848351e-01 -8.23709130e-01 -1.10078596e-01 -2.01496744e+00
-9.19775590e-02 -2.44526058e-01 1.96803473e-02 1.04672432e+00
7.92652786e-01 -4.25934345e-02 -3.78151648e-02 -1.66000366e-01
-1.14741755e+00 -1.51980460e-01 1.08286273e+00 -7.72181600e-02
1.01828352e-01 -4.26252037e-01 -5.36136150e-01 8.30893159e-01
8.08944881e-01 -1.17935613e-01 -4.67640847e-01 1.13901712e-01
6.97187841e-01 1.05450079e-01 4.38831627e-01 -1.21828437e+00
3.57633203e-01 -1.06636167e+00 -4.02416915e-01 -4.71127540e-01
-1.62083596e-01 -1.09283650e+00 6.03876770e-01 3.09075981e-01
2.06447542e-02 -2.42737383e-01 4.08999443e-01 -3.39897603e-01
-5.15837014e-01 -4.20832932e-01 6.27830327e-01 -2.83254087e-01
-9.21909928e-01 -2.29996949e-01 -6.69127822e-01 -8.54092557e-03
1.63375652e+00 -1.32022738e-01 -8.76148403e-01 -2.79070549e-02
-5.51961303e-01 3.82553190e-01 8.03798616e-01 -2.46603981e-01
3.57634515e-01 -1.06295025e+00 -3.48559380e-01 -2.72500366e-01
1.17035219e-02 -3.42687650e-04 -2.67062914e-02 1.26206422e+00
-8.50478172e-01 1.08745515e+00 -3.55865479e-01 -7.49302581e-02
-9.63110209e-01 7.00843036e-01 3.58236879e-01 -2.37438619e-01
-6.71917856e-01 -2.30711803e-01 -5.47079146e-01 -4.31303978e-01
-7.47311786e-02 -4.42642063e-01 7.86382705e-02 -5.38882792e-01
9.28625822e-01 5.43698847e-01 1.97826996e-02 -4.67829406e-01
-7.71142662e-01 1.38163880e-01 7.31381029e-02 -2.76719093e-01
1.22100115e+00 -2.11124625e-02 -1.29842317e+00 9.82887030e-01
3.11124355e-01 -1.83614895e-01 1.14781849e-01 1.97712407e-02
4.95281100e-01 -1.19882151e-01 1.82672262e-01 -7.74672687e-01
-1.12461209e-01 2.76204467e-01 -2.45502919e-01 8.08271885e-01
8.28262627e-01 4.89248596e-02 3.71866196e-01 9.51631248e-01
1.20568812e+00 -1.28824353e+00 -1.07300222e+00 8.69346440e-01
6.83151424e-01 -6.24338865e-01 3.93251181e-01 -5.23672044e-01
-4.85492647e-01 1.39986289e+00 2.92702049e-01 2.96133012e-01
4.48566407e-01 8.36562872e-01 1.36654690e-01 -3.35810572e-01
-1.09454155e+00 -1.86476588e-01 -4.18177813e-01 4.90810752e-01
5.97334087e-01 1.79423988e-01 -5.31867802e-01 4.98964131e-01
-9.88901407e-02 8.84974182e-01 8.63933027e-01 1.76395237e+00
-3.46024960e-01 -1.61832023e+00 -8.47269595e-01 8.22059363e-02
-6.74799740e-01 1.56049803e-01 -4.67926919e-01 1.31933749e+00
4.40942258e-01 1.39022565e+00 -1.83160961e-01 -1.72855437e-01
2.25433901e-01 2.73420185e-01 8.65792811e-01 -3.86425614e-01
-2.83124566e-01 -5.22939742e-01 7.37397015e-01 -4.66301531e-01
-8.85156333e-01 -6.78027153e-01 -1.77187490e+00 -1.19310415e+00
-2.89194614e-01 8.73342872e-01 9.66536477e-02 1.10550117e+00
-1.39705285e-01 1.04523540e-01 -9.63713750e-02 -1.84305996e-01
-3.83852541e-01 -5.42210162e-01 -8.46662879e-01 -2.86705881e-01
-1.02642015e-01 -6.84925497e-01 -2.47572348e-01 5.83583303e-02] | [8.748467445373535, 6.808951377868652] |
ca471b4d-cc92-4ef4-b3d8-168a6fc479bf | a-generalization-of-vit-mlp-mixer-to-graphs | 2212.13350 | null | https://arxiv.org/abs/2212.13350v2 | https://arxiv.org/pdf/2212.13350v2.pdf | A Generalization of ViT/MLP-Mixer to Graphs | Graph Neural Networks (GNNs) have shown great potential in the field of graph representation learning. Standard GNNs define a local message-passing mechanism which propagates information over the whole graph domain by stacking multiple layers. This paradigm suffers from two major limitations, over-squashing and poor long-range dependencies, that can be solved using global attention but significantly increases the computational cost to quadratic complexity. In this work, we propose an alternative approach to overcome these structural limitations by leveraging the ViT/MLP-Mixer architectures introduced in computer vision. We introduce a new class of GNNs, called Graph ViT/MLP-Mixer, that holds three key properties. First, they capture long-range dependency and mitigate the issue of over-squashing as demonstrated on Long Range Graph Benchmark and TreeNeighbourMatch datasets. Second, they offer better speed and memory efficiency with a complexity linear to the number of nodes and edges, surpassing the related Graph Transformer and expressive GNN models. Third, they show high expressivity in terms of graph isomorphism as they can distinguish at least 3-WL non-isomorphic graphs. We test our architecture on 4 simulated datasets and 7 real-world benchmarks, and show highly competitive results on all of them. The source code is available for reproducibility at: \url{https://github.com/XiaoxinHe/Graph-ViT-MLPMixer}. | ['Xavier Bresson', 'Yann Lecun', 'Adam Perold', 'Thomas Laurent', 'Bryan Hooi', 'Xiaoxin He'] | 2022-12-27 | null | null | null | null | ['graph-regression'] | ['graphs'] | [ 9.47755203e-02 1.80188745e-01 -3.21787477e-01 -1.75514430e-01
-4.13251698e-01 -4.46577251e-01 7.40843892e-01 4.75637227e-01
-3.16771239e-01 5.26104271e-01 -1.27973169e-01 -5.35405040e-01
-3.81773829e-01 -1.19350052e+00 -9.57004070e-01 -6.21026337e-01
-4.27727729e-01 4.08177108e-01 3.43393326e-01 -3.11361760e-01
-2.73010470e-02 4.34331417e-01 -1.32584429e+00 1.42257705e-01
7.42072344e-01 8.95378828e-01 3.64236943e-02 8.65175366e-01
-2.43942797e-01 1.21879566e+00 -3.90910745e-01 -5.50444603e-01
6.80913925e-02 -4.03197259e-01 -8.93042445e-01 -2.86806673e-01
5.78801155e-01 -1.57256231e-01 -8.48753631e-01 1.13302100e+00
3.85185868e-01 -4.85282429e-02 2.92051405e-01 -1.54255831e+00
-8.87783289e-01 1.00371206e+00 -5.52433074e-01 2.63777912e-01
1.89468965e-01 -7.01116025e-02 1.39446163e+00 -3.85424525e-01
4.94789511e-01 1.24845207e+00 7.69179046e-01 3.95601809e-01
-1.19357109e+00 -6.75366580e-01 3.15362662e-01 3.64004970e-01
-1.42992246e+00 -8.76466408e-02 7.48688936e-01 -2.95459151e-01
1.12366879e+00 1.80804178e-01 6.37860119e-01 1.03879523e+00
2.81195134e-01 5.63198984e-01 8.22287142e-01 -2.19151467e-01
-1.84837449e-02 -5.24112642e-01 4.20751065e-01 1.13614762e+00
3.46941859e-01 5.11521138e-02 -3.67507607e-01 9.71607566e-02
7.77143180e-01 5.08607961e-02 -3.30610573e-01 -2.56571174e-01
-1.09854674e+00 8.87992561e-01 1.26638699e+00 4.45461512e-01
-7.67050311e-02 7.43009865e-01 4.88766968e-01 5.44654131e-01
2.78334737e-01 1.37792364e-01 -7.53951296e-02 2.45037943e-01
-5.01044452e-01 -1.26657948e-01 7.77449667e-01 9.39217150e-01
8.61850262e-01 5.23503646e-02 -5.35590167e-04 6.32708669e-01
2.87367105e-01 3.72413009e-01 3.05310816e-01 -3.55939537e-01
5.46970069e-01 9.08294976e-01 -7.52943277e-01 -1.41158211e+00
-6.60418272e-01 -6.92593634e-01 -1.38645089e+00 3.26828174e-02
3.37900490e-01 2.34120905e-01 -1.11440885e+00 1.82624936e+00
1.13840073e-01 3.82816851e-01 -6.61012977e-02 6.61494851e-01
1.27768242e+00 6.89178228e-01 8.65688175e-02 2.49683738e-01
1.23612404e+00 -1.18689847e+00 -3.89637768e-01 -4.23824549e-01
7.87656903e-01 -2.07447156e-01 9.15799022e-01 1.76586121e-01
-8.56777966e-01 -5.53050220e-01 -1.27944696e+00 -2.92360395e-01
-6.55262828e-01 -2.19466865e-01 9.06015515e-01 4.67685431e-01
-1.47791433e+00 8.14575851e-01 -8.41302514e-01 -3.81620765e-01
3.83415580e-01 4.82925564e-01 -5.10732949e-01 -1.67822123e-01
-1.19863844e+00 5.03208816e-01 5.87271810e-01 2.96534538e-01
-7.31934667e-01 -4.02373195e-01 -1.06371129e+00 3.95243585e-01
3.39955717e-01 -6.88588798e-01 8.79406929e-01 -8.24567080e-01
-1.19232655e+00 7.28305280e-01 2.40315437e-01 -5.14421701e-01
2.54938513e-01 7.80399656e-03 -4.24766153e-01 5.71869761e-02
-2.02166647e-01 5.50673544e-01 5.72414637e-01 -8.33215892e-01
-2.46678248e-01 -4.35123205e-01 4.66886461e-01 -7.20893294e-02
-4.06305790e-01 -2.25821346e-01 -7.08847284e-01 -6.27138734e-01
1.44332021e-01 -9.41699088e-01 -2.63386428e-01 -1.60619587e-01
-5.58880270e-01 -1.65841192e-01 7.02942252e-01 -2.93745190e-01
1.24419796e+00 -1.90681875e+00 1.59501165e-01 2.15151042e-01
8.12915623e-01 4.87449020e-01 -4.37973797e-01 7.33171284e-01
-1.87271550e-01 1.62003145e-01 -3.03792179e-01 -9.23930630e-02
3.86508927e-02 4.05363530e-01 -1.02133594e-01 5.09522557e-01
1.48171812e-01 1.41572690e+00 -9.79365826e-01 -3.29971313e-01
1.81224108e-01 6.32682204e-01 -3.66676092e-01 -3.24691422e-02
-2.63498247e-01 7.90952146e-02 -2.49788091e-01 3.48790586e-01
7.50369012e-01 -8.79929423e-01 5.30615926e-01 -6.55694902e-02
3.70284259e-01 3.69417906e-01 -1.14998138e+00 1.55802822e+00
-2.91078895e-01 6.64594173e-01 -5.73838428e-02 -1.31028330e+00
9.26942527e-01 -1.00873977e-01 1.55668825e-01 -8.89226973e-01
2.09392309e-01 1.07254982e-01 3.18704825e-03 -7.41389543e-02
2.46922195e-01 1.64758712e-01 -1.51671201e-01 3.16866249e-01
3.17024887e-01 2.77721584e-01 4.25696641e-01 5.30710638e-01
1.40856564e+00 -1.44556925e-01 3.29043686e-01 -3.33943069e-01
6.02840185e-01 -3.62540007e-01 3.07169437e-01 8.82430613e-01
1.04905151e-01 2.99248397e-01 9.33122396e-01 -4.83549953e-01
-5.98837674e-01 -1.02004206e+00 3.23444843e-01 1.20993614e+00
3.28886449e-01 -9.26179588e-01 -5.44787467e-01 -6.72700286e-01
-7.38655627e-02 2.01299638e-01 -5.44063509e-01 -3.43001008e-01
-7.08713949e-01 -6.67390943e-01 8.81691515e-01 6.83008134e-01
4.69977349e-01 -1.01284289e+00 -2.68204778e-01 2.05334485e-01
-1.73481666e-02 -1.33639300e+00 -2.93824732e-01 1.03117086e-01
-8.07595551e-01 -1.17808020e+00 -5.60901403e-01 -9.74613249e-01
7.01318562e-01 1.74725831e-01 1.32177866e+00 6.10960543e-01
-2.29905352e-01 1.75625384e-01 -2.06613660e-01 3.12252045e-02
-4.54324335e-01 3.99398118e-01 -3.84492546e-01 -9.17264894e-02
7.39588169e-03 -8.65542233e-01 -4.36632544e-01 2.01470152e-01
-8.11977506e-01 1.89299896e-01 6.49435997e-01 8.03906083e-01
5.33548772e-01 4.16912884e-02 4.13007945e-01 -1.10405791e+00
5.78175008e-01 -4.49048132e-01 -8.88261974e-01 2.69099712e-01
-6.55618608e-01 2.83346564e-01 9.24173236e-01 -1.41081125e-01
-4.46254194e-01 -2.58934647e-01 -2.96379745e-01 -2.27429062e-01
5.86284585e-02 7.83941984e-01 -4.96918112e-02 -4.00951654e-01
4.34981763e-01 1.19405322e-01 -1.39648393e-01 -3.91894519e-01
4.64133054e-01 9.55710933e-02 5.90807259e-01 -4.74619240e-01
8.28966916e-01 3.26738596e-01 4.71259952e-01 -8.20013881e-01
-5.32590330e-01 -2.24949881e-01 -4.23401564e-01 -9.79500860e-02
6.75614536e-01 -7.46213317e-01 -9.95545268e-01 7.90814638e-01
-1.03568590e+00 -5.24426162e-01 8.59246328e-02 3.30792725e-01
-2.57973373e-01 6.69234991e-01 -1.01376009e+00 -2.74929315e-01
-5.35758674e-01 -1.02864552e+00 7.50996292e-01 1.84482962e-01
2.42243856e-01 -1.24788344e+00 -6.88271374e-02 -1.98693164e-02
5.53104222e-01 5.86683869e-01 1.18084311e+00 -6.45095587e-01
-8.87838721e-01 -5.13811670e-02 -6.55687809e-01 1.95721567e-01
-1.30165905e-01 -7.29917958e-02 -6.43158615e-01 -6.28615320e-01
-6.13224745e-01 -2.21882954e-01 1.09900248e+00 2.06404269e-01
1.10591531e+00 -3.16445500e-01 -3.79905224e-01 9.29789960e-01
1.71030080e+00 -2.00933844e-01 5.55730999e-01 8.60866085e-02
1.23115695e+00 2.78574586e-01 -1.95356593e-01 -7.76697025e-02
6.74760938e-01 4.23803449e-01 7.30014622e-01 -2.19787136e-01
-4.71113563e-01 -5.07316291e-01 2.05168217e-01 1.18339479e+00
-1.39876857e-01 -8.26497793e-01 -1.09940445e+00 3.73266160e-01
-2.07726836e+00 -6.76697016e-01 -4.94822770e-01 1.97744250e+00
1.29504412e-01 3.34753960e-01 -3.14697176e-02 1.33883029e-01
7.61904776e-01 6.10513747e-01 -3.46987247e-01 -4.76023585e-01
-1.20098770e-01 3.50011200e-01 7.02310026e-01 5.67300200e-01
-9.67747629e-01 9.99104261e-01 5.47554302e+00 6.79134548e-01
-1.12334251e+00 -1.89026864e-03 4.68008846e-01 2.48366937e-01
-2.61334568e-01 2.83571854e-02 -4.64036375e-01 2.22844705e-01
1.04362321e+00 -4.89078984e-02 6.13072813e-01 5.66478968e-01
-5.95632911e-01 2.81960249e-01 -1.12266207e+00 1.05300128e+00
1.08848605e-02 -1.58193386e+00 1.39640510e-01 1.21176012e-01
3.30679178e-01 5.93361437e-01 -2.82452554e-01 4.25358504e-01
5.45010924e-01 -1.17529571e+00 4.03212935e-01 2.30536744e-01
7.74490297e-01 -7.10283816e-01 6.26532912e-01 2.32185870e-01
-1.66192126e+00 -3.48106585e-02 -2.92924643e-01 -1.34641796e-01
1.64417643e-02 4.83039349e-01 -5.32330215e-01 1.02773905e+00
6.81337476e-01 7.44231522e-01 -7.36066878e-01 8.66571069e-01
-5.47366679e-01 5.09010851e-01 -3.59252036e-01 -8.37303773e-02
5.11876881e-01 -1.89472556e-01 4.97324616e-01 1.20595968e+00
2.07699656e-01 -1.54794827e-01 1.69698447e-01 8.18104208e-01
-5.03476083e-01 -1.05658494e-01 -6.28883898e-01 -2.65339136e-01
2.58603305e-01 1.23723257e+00 -1.02117443e+00 -2.05614239e-01
-4.43247795e-01 8.83964121e-01 8.25102210e-01 3.08487356e-01
-9.63812351e-01 -6.77377343e-01 3.31069946e-01 7.87328370e-03
4.29522872e-01 -3.85954171e-01 1.93549812e-01 -1.02016068e+00
1.50932744e-01 -8.09467733e-01 7.97019362e-01 -4.63322699e-01
-1.19199407e+00 9.58937287e-01 -1.38989598e-01 -6.87000513e-01
-1.67365968e-02 -8.17383111e-01 -6.67752922e-01 4.45537269e-01
-1.60225105e+00 -1.39568770e+00 -6.14339292e-01 7.09682226e-01
1.85847264e-02 1.55740410e-01 7.84720242e-01 6.27086222e-01
-5.28237462e-01 8.75964582e-01 -9.47340205e-02 4.20411855e-01
1.98105142e-01 -1.31998467e+00 1.00905144e+00 9.11400855e-01
4.93536651e-01 4.27526504e-01 2.36083478e-01 -3.71083885e-01
-1.67959380e+00 -1.17889249e+00 8.09340537e-01 -1.03076130e-01
7.15742290e-01 -7.90952861e-01 -1.13153756e+00 9.32238638e-01
1.11450531e-01 3.34642053e-01 2.46364638e-01 3.01046759e-01
-6.80687189e-01 -2.40493298e-01 -7.30389416e-01 5.29695213e-01
1.45671821e+00 -6.02435350e-01 -1.08935185e-01 3.31106901e-01
8.64770353e-01 -4.30028439e-01 -8.72579932e-01 4.11033452e-01
3.16780001e-01 -1.10383880e+00 9.37331021e-01 -4.42845553e-01
1.61168084e-01 -1.73393577e-01 -1.60681698e-02 -1.16128552e+00
-5.52723348e-01 -8.62768471e-01 -2.15631321e-01 1.12716150e+00
4.24722493e-01 -1.14713180e+00 7.51752734e-01 -4.77745384e-02
-5.62237799e-02 -8.13349068e-01 -8.14721704e-01 -9.43129778e-01
-2.18252800e-02 -3.36533040e-01 7.99915314e-01 1.00651205e+00
-1.24826156e-01 7.59130418e-01 -2.80528873e-01 2.90875465e-01
6.98137224e-01 2.11750254e-01 8.20592165e-01 -1.29343390e+00
-3.28744411e-01 -6.64911449e-01 -8.61957192e-01 -1.22698712e+00
2.21454978e-01 -1.51086748e+00 -2.73214668e-01 -1.91956556e+00
4.99406084e-02 -3.69133323e-01 -5.29527426e-01 7.29685009e-01
7.43686706e-02 2.90133387e-01 2.63470083e-01 -6.53282702e-02
-6.99242294e-01 4.72705513e-01 1.12246406e+00 -3.43058735e-01
-2.03232914e-02 -2.21994951e-01 -6.62515104e-01 6.07720494e-01
9.41332817e-01 -5.60186565e-01 -5.64165711e-01 -7.75104821e-01
4.82361317e-01 8.12041983e-02 6.42601371e-01 -1.16059768e+00
4.72470790e-01 1.73535928e-01 4.16740254e-02 -4.15798753e-01
1.75149217e-01 -6.04947507e-01 3.33887458e-01 6.58574700e-01
-2.56316215e-01 4.24858361e-01 2.23332703e-01 7.78470576e-01
-2.31070563e-01 1.30898774e-01 7.12333798e-01 -5.44542000e-02
-7.97730148e-01 7.02056050e-01 -5.21704443e-02 1.68733820e-01
8.53621125e-01 -8.86813775e-02 -8.58935356e-01 -4.63122666e-01
-4.18094426e-01 3.89470607e-01 2.15160534e-01 6.15244091e-01
5.92542768e-01 -1.27851880e+00 -7.26855040e-01 3.40810150e-01
2.35220760e-01 -6.96925521e-02 2.83936203e-01 8.62057269e-01
-7.49444783e-01 3.78819257e-01 -1.43634409e-01 -6.53080463e-01
-1.10722423e+00 6.92435145e-01 4.24799830e-01 -5.14465451e-01
-1.05893183e+00 9.48281586e-01 3.42711985e-01 -5.91066360e-01
2.22309664e-01 -4.61550325e-01 1.39001366e-02 -2.51567632e-01
2.71127522e-01 4.25426185e-01 2.22596467e-01 -6.92073643e-01
-5.40187895e-01 6.36345804e-01 -1.20508909e-01 5.15103519e-01
1.20348740e+00 1.39615744e-01 -4.83254313e-01 6.18630387e-02
1.50986516e+00 -2.34724835e-01 -8.68267477e-01 -4.15357411e-01
2.35305950e-01 -1.68258205e-01 1.53071824e-02 -4.70678180e-01
-1.35653555e+00 8.67328346e-01 2.97139913e-01 5.55438340e-01
1.03841472e+00 2.46473834e-01 9.94317830e-01 4.84821081e-01
1.53306022e-01 -5.35451591e-01 -7.99860656e-02 6.86606407e-01
7.90959775e-01 -9.98466432e-01 -8.94191191e-02 -5.02407491e-01
-2.27218226e-01 1.05566561e+00 5.97659171e-01 -3.88936639e-01
6.78952575e-01 2.04097614e-01 -1.47395194e-01 -6.03934169e-01
-8.17210972e-01 -2.65110970e-01 2.88054526e-01 5.40625393e-01
2.00743854e-01 2.31827110e-01 -1.08520240e-01 3.23932499e-01
-1.14098087e-01 -2.39746332e-01 4.08629119e-01 6.60938561e-01
-2.08388776e-01 -1.08654296e+00 2.44063258e-01 3.56693149e-01
-3.50871831e-01 -1.67395651e-01 -4.84082669e-01 9.81452823e-01
-2.08076462e-01 8.29815805e-01 -1.45597914e-02 -4.78610277e-01
1.90255657e-01 -3.20776790e-01 5.11673391e-01 -2.68612951e-01
-5.32979310e-01 -2.49975637e-01 2.03983590e-01 -7.58295238e-01
-1.94205493e-01 1.59213051e-01 -1.37044251e+00 -6.47602618e-01
-3.47233087e-01 8.10555145e-02 4.41807210e-01 5.20622194e-01
6.69585407e-01 8.45244288e-01 2.67250359e-01 -6.33583486e-01
-3.61401469e-01 -7.09727466e-01 -4.73925769e-01 2.39841953e-01
5.32882452e-01 -3.86635303e-01 -4.28121090e-01 -4.45737898e-01] | [6.963223934173584, 6.283302307128906] |
c440a733-35f5-4164-999c-c70d4be3935d | multiplex-graph-neural-network-for-extractive | 2108.12870 | null | https://arxiv.org/abs/2108.12870v2 | https://arxiv.org/pdf/2108.12870v2.pdf | Multiplex Graph Neural Network for Extractive Text Summarization | Extractive text summarization aims at extracting the most representative sentences from a given document as its summary. To extract a good summary from a long text document, sentence embedding plays an important role. Recent studies have leveraged graph neural networks to capture the inter-sentential relationship (e.g., the discourse graph) to learn contextual sentence embedding. However, those approaches neither consider multiple types of inter-sentential relationships (e.g., semantic similarity & natural connection), nor model intra-sentential relationships (e.g, semantic & syntactic relationship among words). To address these problems, we propose a novel Multiplex Graph Convolutional Network (Multi-GCN) to jointly model different types of relationships among sentences and words. Based on Multi-GCN, we propose a Multiplex Graph Summarization (Multi-GraS) model for extractive text summarization. Finally, we evaluate the proposed models on the CNN/DailyMail benchmark dataset to demonstrate the effectiveness of our method. | ['Hanghang Tong', 'Wei Fan', 'Tao Yang', 'Zeyu You', 'Baoyu Jing'] | 2021-08-29 | null | https://aclanthology.org/2021.emnlp-main.11 | https://aclanthology.org/2021.emnlp-main.11.pdf | emnlp-2021-11 | ['extractive-document-summarization'] | ['natural-language-processing'] | [ 3.84491950e-01 3.34174961e-01 -2.37028122e-01 -3.73549759e-01
-5.04422247e-01 -3.91568720e-01 6.72697306e-01 7.08473802e-01
-1.31635547e-01 6.01643622e-01 1.16526258e+00 -1.53516665e-01
-2.01455727e-02 -8.41507554e-01 -7.26106942e-01 -1.46961123e-01
7.36748502e-02 3.12487707e-02 6.21534772e-02 -4.43108469e-01
6.25077367e-01 1.82164595e-01 -9.48680401e-01 4.39879775e-01
1.14167643e+00 5.06074309e-01 1.89372972e-01 8.51635754e-01
-5.52709341e-01 1.03939772e+00 -1.00108695e+00 -4.10604030e-01
-3.88825387e-01 -8.09691668e-01 -9.44454372e-01 1.82511136e-01
5.10558486e-01 -5.72447479e-01 -6.32703960e-01 1.19266903e+00
4.66717422e-01 2.77636558e-01 7.47386098e-01 -9.12154436e-01
-9.16577935e-01 1.09083080e+00 -5.43491066e-01 3.23711365e-01
4.70016390e-01 3.48614082e-02 1.62900424e+00 -6.19070888e-01
6.14324510e-01 1.42175937e+00 3.84121090e-01 2.59942412e-01
-9.11221862e-01 -3.11906904e-01 4.78049248e-01 1.07747443e-01
-7.03386664e-01 -4.20100331e-01 1.05908132e+00 7.80223683e-02
1.22606325e+00 5.31276524e-01 7.67375350e-01 9.72302496e-01
5.61194360e-01 1.19184673e+00 4.36583906e-01 -1.01784877e-01
-4.83141132e-02 -4.65927958e-01 5.98184347e-01 6.88766956e-01
5.16429603e-01 -7.58520961e-01 -5.07173836e-01 1.39622781e-02
2.74151146e-01 7.66038671e-02 -3.60456318e-01 2.57509977e-01
-1.23569679e+00 8.32753837e-01 7.17816591e-01 3.37610692e-01
-4.14675891e-01 3.04706961e-01 8.74881268e-01 1.17859915e-01
7.54901052e-01 6.19159341e-01 -8.58846307e-02 1.46468431e-01
-8.58056426e-01 1.79129303e-01 9.12370920e-01 1.10747576e+00
7.88214326e-01 2.00270578e-01 -6.34914994e-01 7.94381320e-01
2.78316170e-01 3.05889487e-01 5.43232203e-01 -5.58778703e-01
1.02642846e+00 1.03399312e+00 -4.30970281e-01 -1.49740064e+00
-3.99672002e-01 -4.84811544e-01 -1.26424336e+00 -6.95374727e-01
-5.24792731e-01 -3.03315550e-01 -6.66406572e-01 1.50274932e+00
-1.77171063e-02 1.02469116e-01 3.16652954e-01 6.09160602e-01
1.75349033e+00 9.74046350e-01 -1.98784962e-01 -4.51535732e-01
1.15345120e+00 -1.27310157e+00 -1.00642169e+00 -5.04523098e-01
6.55821979e-01 -4.62558031e-01 8.05046856e-01 -1.88579679e-01
-1.04921842e+00 -2.28348464e-01 -1.17101765e+00 -4.33765650e-01
-2.28449777e-01 1.92465950e-02 4.14961845e-01 -9.88879278e-02
-1.03910649e+00 7.24320829e-01 -5.78291416e-01 -4.89718258e-01
5.22789776e-01 2.57598132e-01 -2.71674752e-01 7.49401748e-02
-1.26776457e+00 5.50822079e-01 9.09434736e-01 1.83099553e-01
-4.65599567e-01 -2.43529543e-01 -1.11860323e+00 4.53644574e-01
4.74561960e-01 -1.02628982e+00 1.08646142e+00 -8.50343525e-01
-1.29685974e+00 3.74749541e-01 -3.07690173e-01 -5.54284811e-01
2.29026899e-02 -2.90934563e-01 -1.96987107e-01 4.60011065e-01
1.83774352e-01 5.17166078e-01 6.55636609e-01 -1.10253477e+00
-3.55590045e-01 -2.65749067e-01 3.26486707e-01 4.85943884e-01
-5.96436143e-01 1.51568130e-01 -2.30102345e-01 -8.02256644e-01
-1.40515091e-02 -4.38388526e-01 -1.00669838e-01 -7.24070489e-01
-1.16503191e+00 -6.20914102e-01 8.56297135e-01 -1.02556646e+00
1.63251948e+00 -1.95097530e+00 4.99945939e-01 -3.64233255e-01
6.74237013e-01 2.74589449e-01 -2.87111849e-01 9.51114118e-01
1.56349137e-01 3.78785163e-01 -5.21969140e-01 -4.48386103e-01
5.06631993e-02 1.15550578e-01 -3.46269846e-01 -7.51537681e-02
4.74139571e-01 1.36082280e+00 -1.17843044e+00 -7.74130225e-01
-8.46979246e-02 1.75752968e-01 -3.41870815e-01 1.69384658e-01
-4.29013520e-01 -2.68114079e-03 -8.90625477e-01 1.94861487e-01
3.90616506e-01 -3.21105808e-01 2.17697889e-01 -3.88813853e-01
1.72957882e-01 6.07043207e-01 -5.35856724e-01 1.75918412e+00
-3.70264560e-01 8.24545264e-01 -2.16126159e-01 -1.11173415e+00
9.87210691e-01 1.15639448e-01 2.74571061e-01 -4.59649116e-01
2.37186432e-01 9.24257115e-02 -2.00041160e-02 -6.63426280e-01
1.25292242e+00 1.25360429e-01 -1.04159452e-01 7.91361034e-01
1.63982391e-01 -2.10510343e-01 3.90592992e-01 8.75527501e-01
1.28343272e+00 -1.09390296e-01 6.16679609e-01 -1.70592263e-01
4.03091580e-01 -1.39843598e-01 4.42617923e-01 5.09302914e-01
-3.55282389e-02 7.38724768e-01 9.94853735e-01 -1.16376743e-01
-6.73112571e-01 -5.92305541e-01 7.24688888e-01 6.96978450e-01
2.89886206e-01 -9.21395421e-01 -7.20406771e-01 -9.30990160e-01
-1.18880071e-01 8.14266980e-01 -5.13942540e-01 -4.60516840e-01
-8.41352761e-01 -4.78051722e-01 4.61202770e-01 5.67066312e-01
7.01245129e-01 -1.33617938e+00 -3.20021600e-01 3.15346926e-01
-4.85600978e-01 -1.07984364e+00 -8.37979555e-01 -2.92174190e-01
-9.19845402e-01 -8.96114349e-01 -4.47716564e-01 -9.90617573e-01
5.88757336e-01 5.15282393e-01 1.23893118e+00 2.43943557e-01
2.35537648e-01 3.65795821e-01 -5.88078737e-01 -3.61267537e-01
-4.49533373e-01 5.63712955e-01 -3.87653440e-01 -1.21874483e-02
1.44153610e-01 -6.67162955e-01 -6.23954296e-01 -4.19772387e-01
-1.15118492e+00 3.59224916e-01 6.90072119e-01 7.00097740e-01
3.80094111e-01 -1.56663936e-02 1.06833196e+00 -1.07570624e+00
1.51263571e+00 -5.88900685e-01 2.74271160e-01 5.14655888e-01
-3.15741152e-01 1.02780096e-01 1.01127481e+00 -1.25938565e-01
-9.71758723e-01 -6.29906178e-01 -2.75173075e-02 -1.03059113e-01
2.07773268e-01 1.08701992e+00 -1.78898931e-01 6.55744195e-01
2.71467865e-01 5.63996911e-01 -1.86682656e-01 -1.90614045e-01
6.28086925e-01 7.88584650e-01 4.06715244e-01 -2.83707172e-01
4.79499668e-01 2.80662119e-01 -1.14220560e-01 -1.05555034e+00
-1.09928536e+00 -3.86039168e-01 -5.12550831e-01 -8.99982676e-02
1.00635862e+00 -6.07246339e-01 -3.03036332e-01 2.90452003e-01
-1.72739291e+00 3.01206056e-02 -2.80614763e-01 2.95299441e-01
-2.97261477e-01 8.66527200e-01 -5.89192033e-01 -4.49342191e-01
-1.14339960e+00 -8.24776471e-01 1.20780456e+00 4.65025187e-01
-3.87737691e-01 -1.28117299e+00 -1.34570166e-01 2.63788134e-01
2.61966705e-01 5.36064506e-01 9.87854302e-01 -9.31026638e-01
-4.01181877e-01 -2.00028285e-01 -4.63932633e-01 4.89619792e-01
6.12196028e-01 8.64704698e-02 -3.86215508e-01 -2.62066841e-01
-1.78126544e-01 -1.25128791e-01 1.40618575e+00 3.76766503e-01
9.26387310e-01 -8.94614577e-01 -2.40480378e-01 2.58763880e-01
9.62951839e-01 -2.32421339e-01 5.52067697e-01 7.51823783e-02
1.13418698e+00 6.39217913e-01 3.21271718e-01 3.99408221e-01
7.17905581e-01 2.03518867e-02 4.23156977e-01 1.26104549e-01
-3.16603214e-01 -4.05348957e-01 4.62010235e-01 1.58911085e+00
2.47764811e-01 -6.60849214e-01 -6.24968588e-01 4.93401647e-01
-2.07031035e+00 -9.78737354e-01 -2.75806963e-01 1.44222546e+00
7.62154996e-01 2.60433108e-01 -1.56548306e-01 -1.94898218e-01
9.20979381e-01 9.20743227e-01 -6.01834297e-01 -6.17256761e-01
-2.94937670e-01 -1.07665338e-01 -3.75848152e-02 4.65363503e-01
-8.96545589e-01 1.05198407e+00 4.73745871e+00 7.35360086e-01
-9.53227401e-01 -2.67206699e-01 4.37265128e-01 7.14966133e-02
-8.27862859e-01 7.50761703e-02 -5.59447467e-01 3.50028992e-01
5.41624546e-01 -6.91820502e-01 1.92603305e-01 3.36568534e-01
2.37831995e-01 2.20403727e-02 -9.27367389e-01 6.56402409e-01
6.82788253e-01 -1.66923344e+00 6.81178212e-01 -2.50324667e-01
7.81589031e-01 -2.01744899e-01 -3.40640336e-01 3.84890527e-01
9.35297087e-02 -8.71750355e-01 4.97178108e-01 4.28028822e-01
4.80014503e-01 -7.98817277e-01 8.09942007e-01 4.73860830e-01
-1.29851949e+00 2.23802269e-01 -3.09079736e-01 -1.38472226e-02
2.48726845e-01 5.97156286e-01 -6.25524759e-01 1.28232729e+00
2.08471835e-01 1.46384096e+00 -7.67463446e-01 4.82464433e-01
-5.95773578e-01 5.55172324e-01 1.38125941e-01 -5.46024144e-01
4.92745817e-01 -4.25258458e-01 9.05771315e-01 1.38813090e+00
1.45272106e-01 1.17550984e-01 -6.80253357e-02 8.66488814e-01
-6.24816537e-01 2.08780676e-01 -6.68680131e-01 -6.14080429e-01
3.11896205e-01 1.33397961e+00 -8.03272247e-01 -5.06405830e-01
-3.40184718e-01 9.68271315e-01 4.63542372e-01 3.90721500e-01
-4.62955892e-01 -8.13319564e-01 3.00541580e-01 -2.98643857e-01
1.09923288e-01 -3.22542757e-01 -2.19372660e-01 -1.36752212e+00
3.49117875e-01 -8.00929725e-01 3.79486620e-01 -8.29999089e-01
-1.19511831e+00 6.20374143e-01 -8.49419013e-02 -9.64908123e-01
-7.13543221e-02 -2.07093731e-02 -1.24771607e+00 6.43300951e-01
-1.50940311e+00 -1.23881054e+00 -2.44148389e-01 1.41027853e-01
9.25629258e-01 -1.17142141e-01 4.71672565e-01 -2.62300819e-01
-8.07727337e-01 2.13034496e-01 -1.22218475e-01 4.16043788e-01
4.41429734e-01 -1.36972117e+00 8.04610312e-01 1.04469299e+00
2.14994233e-02 8.95801544e-01 5.17944753e-01 -9.12329257e-01
-1.50629699e+00 -1.47843051e+00 1.14530480e+00 -1.05587758e-01
7.32194185e-01 -2.37837598e-01 -9.77309048e-01 8.72610986e-01
7.65155613e-01 -5.94524920e-01 5.29149950e-01 3.17999683e-02
-1.03560984e-01 -4.81595807e-02 -5.64240634e-01 8.74257326e-01
1.13567626e+00 -4.35252041e-01 -1.00294101e+00 4.14273262e-01
1.47660661e+00 -2.04674661e-01 -5.88495135e-01 2.31217772e-01
1.49587274e-01 -8.43031466e-01 5.38031995e-01 -8.26088965e-01
1.29548919e+00 -5.03181666e-02 9.50301737e-02 -1.75762141e+00
-1.15304232e-01 -7.81082213e-01 -4.98780340e-01 1.57829320e+00
2.46024266e-01 -6.93907320e-01 4.21677560e-01 -1.90476421e-02
-7.16229558e-01 -9.63583350e-01 -5.82346737e-01 -5.09073675e-01
5.50743332e-03 2.60644108e-02 7.39766359e-01 9.15974855e-01
1.96602121e-01 1.12743342e+00 -2.76856005e-01 -7.11261928e-02
3.44703496e-01 4.31125283e-01 7.74842083e-01 -1.05059326e+00
-2.98560765e-02 -7.50903368e-01 -2.80344546e-01 -1.15852559e+00
7.16094434e-01 -1.21145201e+00 -9.29368213e-02 -2.64375734e+00
5.33760548e-01 3.50339204e-01 -1.65848181e-01 1.56053349e-01
-7.15493262e-01 -5.19201756e-01 7.28168935e-02 1.07618896e-02
-9.67038751e-01 1.08918071e+00 1.54076433e+00 -5.52095890e-01
-1.36705622e-01 -3.50618452e-01 -1.09906352e+00 4.17317212e-01
9.49406266e-01 -2.48178050e-01 -5.62293768e-01 -6.85830832e-01
5.93112767e-01 2.52215832e-01 1.44142494e-01 -5.59260488e-01
3.77387762e-01 -1.92217976e-01 -5.14060259e-02 -8.81549299e-01
4.97620087e-04 -2.65918165e-01 -5.12646854e-01 3.52701724e-01
-6.74554467e-01 1.64874345e-01 -1.00432083e-01 8.61167669e-01
-6.15110099e-01 -1.69067115e-01 1.82926029e-01 -1.43896237e-01
-3.93129587e-01 3.47966075e-01 -8.23454633e-02 3.70251209e-01
6.23557568e-01 1.19866570e-02 -8.11252415e-01 -5.90963125e-01
-2.45501772e-01 6.81649327e-01 1.27953872e-01 5.31444192e-01
1.02221966e+00 -1.30176425e+00 -1.19573641e+00 -2.96374291e-01
1.29920349e-01 4.45385784e-01 3.16792846e-01 7.98339903e-01
-6.03390098e-01 3.35585415e-01 1.01158418e-01 -2.15592459e-01
-1.32188952e+00 2.92116612e-01 8.19398686e-02 -4.24385935e-01
-8.93742979e-01 6.04353130e-01 1.88686714e-01 -2.83784360e-01
-1.11171648e-01 -6.63519442e-01 -6.29697621e-01 2.78921694e-01
4.10383195e-01 4.07380402e-01 -1.46342829e-01 -5.83312631e-01
-1.33400351e-01 3.05897623e-01 -2.71639705e-01 2.16301009e-01
1.46748793e+00 -1.27805725e-01 -7.98961163e-01 4.65521723e-01
1.40099025e+00 -5.65941297e-02 -8.33728611e-01 -4.16376710e-01
1.65724292e-01 -9.73650906e-03 -6.19291253e-02 -1.80335805e-01
-8.35364819e-01 9.03143823e-01 -5.55168211e-01 5.38438618e-01
1.04372931e+00 -2.08245148e-03 1.26179147e+00 6.31585479e-01
-2.95007467e-01 -1.12871480e+00 4.83237773e-01 7.29897738e-01
1.26961589e+00 -1.11132503e+00 3.92864048e-01 -3.80842239e-01
-6.88624203e-01 1.41861296e+00 5.27948976e-01 -2.65547663e-01
2.60780990e-01 -3.22581977e-01 -4.60852891e-01 -5.22736669e-01
-7.57328510e-01 -1.47375360e-01 5.24264097e-01 1.23192519e-01
4.82095420e-01 2.24948227e-01 -5.41042566e-01 6.40060484e-01
-3.25028867e-01 -4.49052125e-01 9.25645888e-01 9.86702263e-01
-6.04151130e-01 -6.89193904e-01 2.26664394e-01 8.36619020e-01
-2.38195598e-01 -2.92971104e-01 -1.02387786e+00 4.72593218e-01
-4.90632325e-01 1.24818909e+00 -4.10711020e-02 -4.79430884e-01
4.05748606e-01 -1.40191555e-01 2.37690583e-01 -1.07929790e+00
-8.79036427e-01 -1.45376638e-01 3.36659998e-01 -1.21128671e-01
-4.62493330e-01 -4.09237564e-01 -1.44187343e+00 -3.62179041e-01
-1.65410608e-01 1.62882119e-01 3.86597991e-01 9.96198654e-01
5.19323170e-01 1.14181721e+00 6.61104202e-01 -5.20763040e-01
-5.01748860e-01 -1.27195811e+00 -5.35649955e-01 4.36402857e-01
4.72710371e-01 8.16443861e-02 -2.78124899e-01 -1.50294602e-01] | [12.640192985534668, 9.577865600585938] |
e086aca4-9950-41df-83cc-e0a92feb79b6 | multimodal-and-explainable-internet-meme | 2212.05612 | null | https://arxiv.org/abs/2212.05612v3 | https://arxiv.org/pdf/2212.05612v3.pdf | Multimodal and Explainable Internet Meme Classification | In the current context where online platforms have been effectively weaponized in a variety of geo-political events and social issues, Internet memes make fair content moderation at scale even more difficult. Existing work on meme classification and tracking has focused on black-box methods that do not explicitly consider the semantics of the memes or the context of their creation. In this paper, we pursue a modular and explainable architecture for Internet meme understanding. We design and implement multimodal classification methods that perform example- and prototype-based reasoning over training cases, while leveraging both textual and visual SOTA models to represent the individual cases. We study the relevance of our modular and explainable models in detecting harmful memes on two existing tasks: Hate Speech Detection and Misogyny Classification. We compare the performance between example- and prototype-based methods, and between text, vision, and multimodal models, across different categories of harmfulness (e.g., stereotype and objectification). We devise a user-friendly interface that facilitates the comparative analysis of examples retrieved by all of our models for any given meme, informing the community about the strengths and limitations of these explainable methods. | ['Luca Luceri', 'Zhivar Sourati', 'Riccardo Tommasini', 'Alain Mermoud', 'Hông-Ân Sandlin', 'Filip Ilievski', 'Abhinav Kumar Thakur'] | 2022-12-11 | null | null | null | null | ['explainable-models', 'hate-speech-detection', 'meme-classification'] | ['computer-vision', 'natural-language-processing', 'natural-language-processing'] | [-2.09522918e-02 -1.06079699e-02 -1.89063624e-01 1.39339948e-02
-2.46048883e-01 -8.28875959e-01 1.08933222e+00 6.35347962e-01
-1.09533682e-01 1.42640635e-01 5.96985221e-01 -4.14662153e-01
-1.42166078e-01 -4.78470504e-01 -1.39242604e-01 -8.17255527e-02
1.68319702e-01 1.67858332e-01 1.47329375e-01 -4.59132135e-01
8.33045125e-01 1.58081055e-01 -1.66531336e+00 9.28754866e-01
6.85629666e-01 5.92822731e-01 -1.01577066e-01 8.67656529e-01
-4.38057661e-01 1.03842580e+00 -8.60577583e-01 -8.54312181e-01
-2.64622986e-01 -3.36834341e-01 -8.07071626e-01 1.17238067e-01
8.52593064e-01 -2.71540284e-01 -4.93640661e-01 9.02765691e-01
3.18786532e-01 -2.59454787e-01 7.99599588e-01 -1.60337806e+00
-1.06609631e+00 6.53594255e-01 -5.17220795e-01 4.91537929e-01
6.99024141e-01 8.37381035e-02 9.11057889e-01 -8.13425779e-01
1.00091875e+00 1.53558636e+00 8.56915712e-01 6.73479676e-01
-1.19864190e+00 -5.09730577e-01 1.53270066e-01 3.68996590e-01
-1.03750873e+00 -4.18944687e-01 6.00981414e-01 -1.04392755e+00
7.86071002e-01 7.40286291e-01 6.76392019e-01 1.31637871e+00
-5.62100261e-02 7.39641905e-01 1.14605188e+00 -4.37199742e-01
-1.01093404e-01 6.94596112e-01 4.62903559e-01 9.36887383e-01
1.78014204e-01 -3.82528007e-01 -9.43941593e-01 -6.24151409e-01
1.17536530e-01 1.54247627e-01 -1.60986587e-01 8.20323005e-02
-1.02826858e+00 9.90892529e-01 2.27495417e-01 4.41593170e-01
-7.49847665e-02 1.38568506e-01 5.59647918e-01 2.19123408e-01
9.19248760e-01 8.24233413e-01 2.00207636e-01 2.15081498e-03
-1.07815731e+00 2.69219100e-01 1.00848210e+00 6.10560834e-01
6.87563062e-01 -5.01658261e-01 -2.72315681e-01 7.37601697e-01
1.95056379e-01 4.44821209e-01 1.47670850e-01 -6.10541284e-01
3.68897319e-01 8.37082028e-01 2.17802092e-01 -1.60809028e+00
-5.61175823e-01 -5.70147224e-02 -1.18009113e-01 7.42108971e-02
3.42900008e-01 3.03268190e-02 -6.08033359e-01 1.40576243e+00
3.67662221e-01 -2.34736115e-01 -7.01048315e-01 6.89714193e-01
9.69665170e-01 6.07671618e-01 4.11688477e-01 5.99283725e-02
1.54680634e+00 -6.58733189e-01 -9.66651142e-01 -2.92845041e-01
9.04722512e-01 -1.03741038e+00 1.22437608e+00 1.10915631e-01
-8.39077175e-01 8.90030637e-02 -9.18984890e-01 -3.81519377e-01
-8.96847486e-01 -8.61012191e-02 3.79337549e-01 8.45303595e-01
-8.70065749e-01 4.36118692e-01 -3.64773571e-01 -1.17506123e+00
3.61018002e-01 -2.42803797e-01 -3.70922148e-01 2.53819078e-01
-9.84115422e-01 1.28678465e+00 1.05190009e-01 -2.89207488e-01
-5.62611043e-01 -6.22907043e-01 -4.25086677e-01 -1.10275067e-01
2.96824217e-01 -5.54511607e-01 9.43522215e-01 -9.62145984e-01
-5.64278066e-01 1.25202155e+00 9.51593295e-02 2.28153411e-02
4.01364446e-01 -2.25042149e-01 -4.06817943e-01 2.15659484e-01
6.41500056e-02 5.19482136e-01 9.77116704e-01 -1.42877698e+00
-4.43126768e-01 -4.26652282e-01 3.81035060e-01 2.21369509e-02
-1.00314772e+00 6.27464175e-01 4.70285080e-02 -7.47582376e-01
-1.56284362e-01 -8.15769196e-01 3.00557435e-01 2.61087894e-01
-7.69239545e-01 -1.24800831e-01 1.23401678e+00 -1.07807076e+00
1.89507961e+00 -2.15392351e+00 7.49017969e-02 9.51388031e-02
7.94595301e-01 5.43861762e-02 7.52059296e-02 1.04908228e+00
3.01608235e-01 9.04315174e-01 8.92229378e-02 -4.30833280e-01
3.11049283e-01 -2.63221532e-01 -3.13242555e-01 5.68013608e-01
-1.57897249e-01 4.71935153e-01 -9.15000498e-01 -6.41553402e-01
-8.51109400e-02 5.28607368e-01 -5.53959608e-01 1.11378454e-01
-3.42874318e-01 8.90210867e-02 -4.18684989e-01 5.87197721e-01
2.26191536e-01 -2.78299481e-01 2.92223185e-01 -8.01702440e-02
-3.61521542e-01 3.93815279e-01 -7.68540323e-01 1.08733130e+00
-3.56164277e-01 1.44631231e+00 2.21946344e-01 7.91393816e-02
5.24111867e-01 1.74738213e-01 -3.69005278e-02 -5.58523834e-01
9.99775063e-03 2.40068957e-02 -2.85076320e-01 -1.14684558e+00
8.34033430e-01 -7.24778250e-02 1.11523224e-03 1.02968383e+00
-3.49447906e-01 2.53180355e-01 1.33094326e-01 6.95829928e-01
1.14970875e+00 2.36741528e-02 3.58582318e-01 -8.25063959e-02
1.78863525e-01 1.93030611e-01 -2.22362459e-01 1.03437817e+00
-3.16986591e-01 4.89078283e-01 8.25530648e-01 -6.65677786e-01
-1.35001087e+00 -4.83350098e-01 -6.25999039e-03 1.86394835e+00
3.33223939e-01 -8.65736544e-01 -9.25506949e-01 -7.38352299e-01
9.60400030e-02 8.73803914e-01 -8.88167739e-01 3.64130288e-02
-4.27753389e-01 -6.59470081e-01 6.21338069e-01 1.47052435e-02
8.42120796e-02 -9.11129773e-01 -7.10358262e-01 1.38514359e-02
-4.97928232e-01 -7.41780043e-01 -2.59800941e-01 -4.37199771e-01
-3.08920920e-01 -1.23483288e+00 -2.29965612e-01 -2.91136473e-01
4.95187044e-01 6.68342829e-01 1.09562290e+00 9.72966552e-01
-3.30348521e-01 8.43727171e-01 -5.01242876e-01 -4.71501261e-01
-5.76111674e-01 1.03124745e-01 -1.77223027e-01 3.96195538e-02
3.52320075e-01 -2.60673940e-01 -5.28149605e-01 2.57447153e-01
-1.12083840e+00 3.39206398e-01 -1.71940904e-02 4.72764939e-01
-3.67516577e-01 -4.09712374e-01 9.30770636e-02 -1.21248722e+00
9.64496672e-01 -1.19912326e+00 2.60837764e-01 5.56938231e-01
-3.56721461e-01 -3.20315182e-01 1.10526152e-01 -4.86819535e-01
-9.58021939e-01 -5.13864636e-01 1.90484747e-01 -9.00762454e-02
-5.87519631e-02 3.50787431e-01 3.02557021e-01 -6.65763617e-02
1.09169865e+00 -1.23784080e-01 -7.23079965e-02 -6.42123640e-01
5.74938893e-01 1.08763587e+00 9.03646555e-03 -3.33045781e-01
1.00989091e+00 6.40773952e-01 -4.94580239e-01 -1.00647044e+00
-8.39488029e-01 -5.24200797e-01 -2.67302334e-01 -7.34398067e-01
8.99677634e-01 -4.04377490e-01 -8.75219524e-01 2.60028958e-01
-1.46686745e+00 -9.05244425e-02 4.62931395e-01 -3.45217250e-02
-2.25087851e-01 5.20732880e-01 -7.96815813e-01 -1.09303272e+00
-1.78492442e-01 -6.38481915e-01 9.35351431e-01 -1.05363272e-01
-8.73999596e-01 -1.15632427e+00 7.12353066e-02 5.60747921e-01
5.73436201e-01 4.43628907e-01 1.30839705e+00 -1.10744286e+00
-4.65255767e-01 -2.02399001e-01 -3.54886502e-01 -5.39723754e-01
-3.27557683e-01 3.42425227e-01 -1.15692127e+00 -8.87803733e-02
-1.72567800e-01 -4.37177300e-01 6.78804874e-01 -3.75157535e-01
9.07887936e-01 -9.73010957e-01 -6.30374372e-01 1.84931327e-02
1.21718383e+00 -2.31557518e-01 4.16283160e-01 7.84574807e-01
8.14936280e-01 1.17684531e+00 2.76034266e-01 6.56647444e-01
4.12547499e-01 7.08128035e-01 6.93012714e-01 -6.03865869e-02
-1.50013775e-01 -3.68770301e-01 2.49192774e-01 3.56765658e-01
1.23916797e-01 -5.57356775e-01 -1.11944902e+00 3.74638766e-01
-2.01602125e+00 -1.41579127e+00 -3.77306402e-01 1.65690398e+00
4.82949525e-01 -1.50905073e-01 4.59475964e-01 -8.33881870e-02
1.16574407e+00 4.54910189e-01 -7.33164027e-02 -5.47086477e-01
1.16276115e-01 -5.83829463e-01 1.18546516e-01 6.69818342e-01
-1.02750063e+00 5.77007353e-01 6.60519981e+00 7.18134522e-01
-9.61489499e-01 4.65625048e-01 4.81068283e-01 -2.79087991e-01
-6.80550218e-01 -1.40371412e-01 -5.38050473e-01 6.48094654e-01
7.12828815e-01 2.19264962e-02 5.78760564e-01 6.56685412e-01
2.31967360e-01 -6.35221452e-02 -1.34504259e+00 7.67607212e-01
5.40992975e-01 -1.70755661e+00 2.72019655e-02 -2.92762276e-02
4.26274002e-01 -4.28167045e-01 3.22342366e-01 7.16680586e-02
-1.83703199e-01 -9.56890047e-01 1.30929518e+00 4.53628361e-01
5.37954032e-01 -1.00749478e-01 2.64346421e-01 3.60489070e-01
-4.94471431e-01 -2.90122271e-01 9.25171822e-02 -2.95832157e-01
-1.20021932e-01 5.04498705e-02 -8.69490325e-01 -2.35015437e-01
5.85495472e-01 5.72650492e-01 -1.06566155e+00 9.81093466e-01
-1.47945464e-01 4.67023253e-01 -4.14714254e-02 -4.64557648e-01
2.37665623e-02 3.14938843e-01 9.61046457e-01 1.77219176e+00
7.16034696e-02 -3.32847349e-02 -3.34610976e-02 8.62304568e-01
2.17496399e-02 1.44695058e-01 -7.08483100e-01 -3.72897387e-01
8.49621654e-01 1.44531024e+00 -7.75909364e-01 -2.29227602e-01
-3.65418732e-01 6.93922877e-01 5.63202798e-01 2.38576695e-01
-7.22744405e-01 -3.73614401e-01 4.20402974e-01 6.12708747e-01
-1.31009966e-01 -8.23316872e-02 -4.48086858e-01 -1.10478437e+00
-1.33805767e-01 -9.98278856e-01 5.73841810e-01 -1.05260277e+00
-1.55700672e+00 5.05581677e-01 7.89004788e-02 -7.85295308e-01
-1.36729199e-02 -4.87646729e-01 -6.99603498e-01 4.51268286e-01
-1.01539969e+00 -1.48555005e+00 -3.80944461e-01 1.45143747e-01
3.73837233e-01 2.15003118e-01 6.09264791e-01 2.85855889e-01
-7.64293730e-01 2.75259435e-01 -1.35159111e-02 8.85557607e-02
7.57466197e-01 -1.05062079e+00 2.41569802e-01 5.99264681e-01
1.71726912e-01 9.33473408e-01 1.25744629e+00 -8.51065516e-01
-1.21220696e+00 -6.86038136e-01 1.03036928e+00 -1.16286266e+00
1.27528787e+00 -7.67671287e-01 -9.72175956e-01 6.76671147e-01
6.27365351e-01 -6.68950558e-01 9.50103641e-01 5.67892849e-01
-9.32708621e-01 5.16356587e-01 -1.20288074e+00 7.83385873e-01
1.14107907e+00 -9.05907571e-01 -5.98666608e-01 8.88011813e-01
3.63184452e-01 -3.80282514e-02 -4.84523147e-01 -3.76458913e-01
7.35873461e-01 -1.16241252e+00 5.78559995e-01 -1.11997581e+00
8.54006886e-01 -9.32811201e-02 8.91386047e-02 -9.95340288e-01
-3.99929911e-01 -7.51315713e-01 -3.10153246e-01 1.48250735e+00
4.36955124e-01 -3.25753897e-01 4.66719121e-01 9.27009881e-01
-3.06868330e-02 -5.13996661e-01 -6.48610592e-01 -2.83915639e-01
-2.05207974e-01 -3.14925164e-01 2.85694629e-01 1.39301538e+00
5.94912231e-01 3.72260541e-01 -5.47624946e-01 8.00470188e-02
4.70703751e-01 -2.17390969e-01 7.06108212e-01 -1.12242997e+00
-3.35885137e-02 -7.34680712e-01 -4.52830613e-01 -3.32115740e-01
-2.18544118e-02 -7.93594599e-01 -5.00650764e-01 -1.67263377e+00
9.75056231e-01 -2.57728755e-01 2.49842063e-01 4.77623999e-01
-1.76244199e-01 5.83149672e-01 6.81239367e-01 7.46166348e-01
-8.58049989e-01 -1.38481945e-01 7.52683938e-01 -1.98890880e-01
4.36522402e-02 -6.78053439e-01 -8.17310452e-01 8.77019405e-01
5.47365963e-01 -7.08618522e-01 -1.63944989e-01 -3.93100858e-01
1.02311671e+00 -2.96622604e-01 7.64598310e-01 -7.35409319e-01
1.65993199e-01 -3.51598769e-01 2.09227651e-01 -2.22830877e-01
3.54365468e-01 -5.71770191e-01 9.10760239e-02 4.28120047e-01
-7.80239046e-01 4.53327559e-02 4.95669898e-03 7.31351733e-01
2.29876101e-01 -3.67697179e-01 5.41763425e-01 -1.74637556e-01
-5.94567597e-01 -8.62997100e-02 -7.49003530e-01 1.45288780e-01
9.90933061e-01 -4.46881592e-01 -1.27250123e+00 -7.05189586e-01
-8.35795820e-01 -4.38321158e-02 7.10299075e-01 5.62243700e-01
4.21576083e-01 -1.11840606e+00 -6.06417775e-01 -4.83979911e-01
4.28031683e-01 -1.25960279e+00 1.73132375e-01 8.26207399e-01
-4.57168370e-01 1.00141197e-01 -2.36103684e-01 -2.30186716e-01
-1.50130832e+00 6.84545100e-01 2.53509909e-01 2.60382086e-01
-3.86126578e-01 4.93615717e-01 2.15267241e-01 -2.47854024e-01
2.37040862e-01 5.37499964e-01 -5.22999942e-01 4.81091797e-01
8.61689985e-01 8.25058341e-01 -2.57412255e-01 -7.22903669e-01
-2.80877829e-01 1.60781205e-01 -8.95374790e-02 -5.85813411e-02
1.18670201e+00 -4.45022881e-01 -3.24623555e-01 5.71017087e-01
1.06199646e+00 2.48955935e-01 -6.31618977e-01 6.09552898e-02
2.68746674e-01 -7.41436720e-01 -1.94980949e-01 -1.04025781e+00
-3.86826932e-01 1.03106892e+00 4.18897271e-01 1.12531602e+00
4.24841911e-01 2.77394861e-01 6.53454781e-01 2.75546312e-01
-2.59754118e-02 -1.03390312e+00 5.11487544e-01 2.70189404e-01
1.02957737e+00 -1.23433220e+00 2.87239939e-01 -6.20322049e-01
-5.67995548e-01 1.27825761e+00 7.08818257e-01 4.60004151e-01
3.31786186e-01 -1.19917445e-01 1.60636008e-01 -7.53092766e-01
-8.18873763e-01 -1.10355049e-01 3.82935047e-01 4.71887082e-01
6.21560514e-01 -7.25276545e-02 -5.41150808e-01 2.70286143e-01
6.14766292e-02 -5.55984139e-01 7.55155325e-01 9.95404363e-01
-7.99077868e-01 -4.98449445e-01 -7.50001311e-01 4.14251596e-01
-7.03750253e-01 -1.36139065e-01 -1.17174041e+00 1.01341152e+00
1.48644254e-01 1.13555133e+00 1.67758822e-01 -5.79045057e-01
-5.83344437e-02 1.40275583e-01 2.35780761e-01 -7.14220881e-01
-1.06000876e+00 -3.44861507e-01 6.76498175e-01 -4.05381113e-01
-4.23639297e-01 -5.77839732e-01 -8.64532351e-01 -8.72148097e-01
-3.33608896e-01 -1.08055741e-01 9.93180037e-01 1.04415488e+00
5.21172881e-01 -1.77031711e-01 4.77614075e-01 -7.55092561e-01
-2.31683731e-01 -9.14903581e-01 -2.15770528e-01 8.70218754e-01
3.44441503e-01 -5.93916416e-01 -5.06064177e-01 -4.63694558e-02] | [8.508186340332031, 10.649076461791992] |
9f3017e1-61aa-4c7d-bcb0-fce798a90cae | 2305-14655 | 2305.14655 | null | https://arxiv.org/abs/2305.14655v1 | https://arxiv.org/pdf/2305.14655v1.pdf | Learning Survival Distribution with Implicit Survival Function | Survival analysis aims at modeling the relationship between covariates and event occurrence with some untracked (censored) samples. In implementation, existing methods model the survival distribution with strong assumptions or in a discrete time space for likelihood estimation with censorship, which leads to weak generalization. In this paper, we propose Implicit Survival Function (ISF) based on Implicit Neural Representation for survival distribution estimation without strong assumptions,and employ numerical integration to approximate the cumulative distribution function for prediction and optimization. Experimental results show that ISF outperforms the state-of-the-art methods in three public datasets and has robustness to the hyperparameter controlling estimation precision. | ['Bo Yan', 'Weimin Tan', 'Yu Ling'] | 2023-05-24 | null | null | null | null | ['numerical-integration', 'survival-analysis'] | ['miscellaneous', 'miscellaneous'] | [-2.83021092e-01 -2.74709553e-01 -7.38950253e-01 -8.12750041e-01
-9.05745983e-01 -1.22838564e-01 1.25510052e-01 1.84075400e-01
-3.39006186e-01 1.35067952e+00 3.70786726e-01 -5.79002082e-01
-2.86176413e-01 -8.28861654e-01 -3.81649107e-01 -6.74085855e-01
-4.94261980e-01 4.59696084e-01 -2.49358535e-01 2.96794564e-01
-2.16335207e-02 3.88165623e-01 -7.54969597e-01 -2.35936776e-01
8.76409590e-01 9.60863292e-01 -6.48638368e-01 6.18143976e-01
1.04852892e-01 9.12038624e-01 -6.13726020e-01 -3.36700469e-01
-3.86971712e-01 -1.72311470e-01 -1.79214761e-01 -7.33316898e-01
-3.09730172e-01 -6.92610145e-01 -7.88305879e-01 5.88810742e-01
4.97421086e-01 6.55881017e-02 1.37116671e+00 -1.61086297e+00
-1.02742136e+00 3.47651213e-01 -3.16833109e-01 9.24153551e-02
-3.86782624e-02 -1.28705844e-01 3.13275367e-01 -7.59097934e-01
-8.58483166e-02 1.16559565e+00 1.38559055e+00 5.55967510e-01
-1.02623963e+00 -6.17201030e-01 -1.53499603e-01 5.99519089e-02
-1.76249182e+00 -1.83430955e-01 5.59129357e-01 -4.34114277e-01
6.53182268e-01 2.37304062e-01 4.10073161e-01 1.19617343e+00
9.20981944e-01 5.65289378e-01 9.37852740e-01 -1.20114759e-01
3.57106239e-01 -1.28726348e-01 8.46472442e-01 1.99623376e-01
3.71995002e-01 1.08078623e+00 -9.21313688e-02 -8.12806010e-01
1.03254855e+00 5.49903870e-01 -2.24595055e-01 9.34867263e-02
-7.81303406e-01 1.18685961e+00 4.42487001e-01 -2.52382517e-01
-5.01385629e-01 3.79768163e-01 3.50873321e-01 3.03217441e-01
5.46996653e-01 -5.60858727e-01 -6.84121072e-01 1.58178076e-01
-1.06966293e+00 9.47712734e-02 1.00041544e+00 9.09201145e-01
1.62283238e-02 1.90541580e-01 -7.13688135e-01 4.75309044e-01
3.56570393e-01 4.02366966e-01 5.22346437e-01 -7.08494365e-01
-1.46823674e-01 4.27051395e-01 4.71590132e-01 -3.08036417e-01
-9.12881613e-01 -8.16859901e-01 -1.41615963e+00 3.14114988e-01
7.62480915e-01 -3.87330949e-01 -9.78298962e-01 1.76831818e+00
1.43743604e-01 4.75280523e-01 2.45860189e-01 6.96255922e-01
6.90421283e-01 6.45762265e-01 5.22348404e-01 -5.72443247e-01
1.14435387e+00 -6.04074836e-01 -9.19830561e-01 9.62411314e-02
4.67135817e-01 -6.12766221e-02 9.00932491e-01 3.98985982e-01
-8.14276278e-01 -2.86667366e-02 -7.32382596e-01 -6.57094568e-02
-3.00301939e-01 3.63504142e-01 8.44236970e-01 8.45910847e-01
-7.20710337e-01 8.07571828e-01 -8.10757995e-01 -1.22929804e-01
5.01476288e-01 5.58467448e-01 -1.17089227e-01 5.52953929e-02
-1.43873262e+00 6.69635594e-01 3.52160484e-01 1.65271580e-01
-1.04241216e+00 -5.55463195e-01 -6.16298616e-01 3.54531586e-01
-7.18351975e-02 -9.91722465e-01 9.61709917e-01 -9.61190820e-01
-1.50631404e+00 3.67859840e-01 -4.28598858e-02 -5.66447914e-01
6.50888741e-01 -2.68643588e-01 -5.27745426e-01 -4.05382097e-01
-4.09373015e-01 6.99901208e-02 4.84788239e-01 -5.55877805e-01
-2.45785415e-01 -5.42409718e-01 -5.71223676e-01 -1.93849578e-01
-4.36246037e-01 1.61247939e-01 -7.96874706e-03 -1.03889537e+00
-1.44626215e-01 -4.82113510e-01 -2.33527288e-01 2.38374665e-01
-7.87203014e-02 -4.01841581e-01 2.70493686e-01 -1.25102794e+00
1.64434695e+00 -2.27976251e+00 -1.96257100e-01 -1.40073761e-01
-1.68851361e-01 -4.14212316e-01 2.51373053e-01 4.44664299e-01
2.63715610e-02 -8.65660608e-02 -3.52705866e-01 -4.54084069e-01
-1.06470495e-01 9.71930325e-02 -3.37972730e-01 8.72920692e-01
-2.79470146e-01 9.14902329e-01 -6.46499813e-01 -7.16217339e-01
-2.06136510e-01 8.34925115e-01 -1.79305762e-01 4.92907971e-01
3.43492150e-01 3.59371036e-01 -4.90110993e-01 1.03881693e+00
7.30139673e-01 -4.25958574e-01 -1.27799511e-01 4.45952028e-01
1.12675823e-01 -2.09278643e-01 -7.17442513e-01 1.18491912e+00
-6.11164905e-02 1.36926338e-01 -3.80687863e-01 -9.52561677e-01
1.08674586e+00 6.22910976e-01 6.46101311e-02 -1.67968031e-02
5.48287272e-01 2.53046095e-01 -4.74085867e-01 -2.86461174e-01
-9.16712433e-02 -6.53919160e-01 -3.05523306e-01 -1.31013185e-01
-3.63140926e-02 5.77728748e-01 -6.67836487e-01 -2.53792167e-01
9.14819896e-01 1.23115323e-01 7.23232925e-01 -1.51397437e-01
3.37994248e-01 -2.40401432e-01 8.99391115e-01 1.03196442e+00
-5.71168780e-01 6.38978720e-01 7.56960154e-01 -5.62644541e-01
-6.40387535e-01 -1.40657830e+00 -6.80040419e-01 7.79590309e-01
-3.09678912e-01 4.20632005e-01 -5.52686155e-01 -7.00453222e-01
2.26386100e-01 1.18991649e+00 -1.14829814e+00 -5.49197972e-01
-1.57540441e-01 -1.23178101e+00 7.48694599e-01 1.05442131e+00
-7.94838592e-02 -9.81128216e-01 1.66903138e-02 4.16839302e-01
8.11110213e-02 -1.21596806e-01 -5.67769766e-01 4.46653098e-01
-1.43627036e+00 -9.34070408e-01 -1.20032108e+00 -7.17014194e-01
3.99182230e-01 -5.41307271e-01 1.06661093e+00 1.96778756e-02
4.77179736e-02 -1.58485115e-01 -1.77815426e-02 -1.97648868e-01
-7.02236071e-02 -1.26329929e-01 -1.50088936e-01 -2.56986171e-01
4.86655205e-01 -4.88646746e-01 -8.80459011e-01 3.20679873e-01
-7.88851976e-01 -3.73971283e-01 4.41712052e-01 1.35406733e+00
4.92462814e-01 -2.70693060e-02 1.46917534e+00 -6.53392971e-01
4.37152207e-01 -1.16013658e+00 -6.74549818e-01 4.48876292e-01
-7.89793372e-01 -1.56822056e-01 7.22058535e-01 -8.86783898e-01
-9.23826396e-01 7.74016662e-04 8.61928537e-02 -5.28361022e-01
-1.44888833e-01 7.95449793e-01 -1.38356179e-01 2.79743940e-01
4.03355002e-01 3.36621433e-01 -5.85710704e-02 -6.39283478e-01
-2.65573356e-02 8.57711673e-01 7.73395717e-01 -3.65654916e-01
9.37431678e-02 3.96976084e-01 2.51050651e-01 -1.69512704e-02
-7.32276618e-01 -1.03864156e-01 -3.94373417e-01 8.47643688e-02
5.73319316e-01 -9.73202825e-01 -1.01805651e+00 7.18116760e-01
-9.01478469e-01 -3.84867191e-01 -1.97894037e-01 9.83391583e-01
-7.78387308e-01 8.51559341e-02 -9.66694236e-01 -1.32232809e+00
-5.51420689e-01 -4.70270663e-01 7.60143757e-01 3.46034378e-01
-9.60583836e-02 -1.36529100e+00 1.25071570e-01 -1.80875391e-01
4.70544249e-01 5.70812583e-01 1.03375459e+00 -6.74805224e-01
-6.73549101e-02 -9.98262763e-01 -1.45815700e-01 2.04173386e-01
-1.07896686e-01 -1.81743309e-01 -5.91583788e-01 -5.41230798e-01
9.93943438e-02 -1.18410490e-01 7.80914545e-01 1.29049551e+00
1.58762562e+00 -4.24782217e-01 -5.21282017e-01 9.95208740e-01
1.36198509e+00 3.30812365e-01 7.76583076e-01 2.16698408e-01
-9.48437583e-03 6.98397875e-01 6.14496052e-01 7.34995008e-01
4.85976994e-01 1.81341812e-01 4.25876766e-01 -1.50007932e-02
3.02461982e-01 -3.40868115e-01 3.02803069e-01 2.14366525e-01
3.01127523e-01 -4.87989992e-01 -7.11264074e-01 5.60055554e-01
-2.00915480e+00 -6.93030417e-01 -4.42094147e-01 2.71012807e+00
8.96961093e-01 1.17092296e-01 3.38515997e-01 -1.43710345e-01
8.13866854e-01 -5.31699419e-01 -8.87419820e-01 -9.00027156e-03
-2.47855991e-01 -1.46653384e-01 7.41127551e-01 5.57702601e-01
-1.00303769e+00 4.89216954e-01 7.57879686e+00 8.74282956e-01
-7.53945291e-01 2.37621233e-01 1.15863907e+00 -8.41865689e-02
-7.30388686e-02 3.04015040e-01 -7.13250935e-01 7.75991976e-01
1.20537210e+00 -3.23227912e-01 2.13692024e-01 6.92593932e-01
2.49320686e-01 2.38519922e-01 -1.04989839e+00 7.05611765e-01
-2.86321163e-01 -7.61659980e-01 -3.80308717e-01 -1.97498333e-02
3.23427826e-01 -3.65735590e-01 1.42707258e-01 7.62837708e-01
3.18794757e-01 -1.56552327e+00 5.70425451e-01 1.31918073e+00
1.15175664e+00 -9.76067543e-01 1.20661342e+00 5.80877125e-01
-6.39742613e-01 -3.95749539e-01 -4.12261546e-01 -1.92922741e-01
4.92546856e-01 6.58314049e-01 -4.98510271e-01 3.71400207e-01
5.15068591e-01 4.14298773e-01 -3.51309597e-01 1.44944644e+00
2.42471993e-02 1.24372375e+00 -3.35300893e-01 -1.56819761e-01
-1.17958747e-01 3.37566971e-03 3.02113801e-01 9.38374698e-01
8.42402518e-01 2.01771110e-01 -1.32633865e-01 7.26736367e-01
1.83780223e-01 -1.70146264e-02 -1.47975385e-01 1.90415844e-01
5.55882752e-01 7.81595349e-01 -4.54220414e-01 -3.27792346e-01
-4.59439605e-01 7.17898965e-01 3.65693241e-01 6.81315720e-01
-1.06819332e+00 -3.02694499e-01 5.32808341e-02 1.87116042e-01
-8.16089585e-02 -8.64174888e-02 -7.41687775e-01 -9.67465222e-01
-4.50868696e-01 -9.30476412e-02 1.19582653e+00 -7.72656679e-01
-1.89953279e+00 3.20612818e-01 -7.03006983e-03 -1.09492338e+00
9.56635736e-03 -5.64955354e-01 -8.88225913e-01 1.18393302e+00
-1.41562152e+00 -1.18129516e+00 -1.01487324e-01 6.75598502e-01
4.92907129e-02 -6.96262717e-02 8.42894256e-01 3.00695270e-01
-7.91599274e-01 8.51941407e-01 7.14669049e-01 -3.45430970e-02
7.86799550e-01 -1.23430955e+00 1.90865740e-01 9.96374860e-02
-7.37240434e-01 5.37112474e-01 6.11034989e-01 -1.00334680e+00
-9.23552513e-01 -1.09921885e+00 9.37602580e-01 -4.04493600e-01
3.77967954e-01 1.82933882e-02 -1.25311995e+00 7.58886278e-01
-2.50264585e-01 1.68246195e-01 1.06349838e+00 2.25715563e-01
3.07545904e-02 1.00852229e-01 -1.50917459e+00 2.43527427e-01
5.23844659e-01 -5.44925313e-03 -4.42777753e-01 2.64661402e-01
7.58432448e-01 -1.38385534e-01 -1.04788589e+00 6.26728237e-01
8.45251620e-01 -4.83044595e-01 1.07212102e+00 -1.08476686e+00
1.66872248e-01 -8.87058210e-03 -2.16658279e-01 -8.56923342e-01
-7.41049230e-01 -4.29657876e-01 -5.92138052e-01 1.34111583e+00
9.57644433e-02 -9.16524768e-01 7.88867712e-01 6.00324988e-01
1.56314135e-01 -9.82883096e-01 -1.48641992e+00 -1.02639329e+00
8.27510834e-01 -8.03427473e-02 1.04007506e+00 7.49007225e-01
-1.88949630e-01 -5.13684005e-02 -8.40182781e-01 4.16380346e-01
9.98305202e-01 -1.48438737e-01 2.28315040e-01 -1.55868649e+00
-2.45976388e-01 -1.18760519e-01 -1.04376212e-01 -5.97148657e-01
3.08217674e-01 -5.70529580e-01 -2.29665235e-01 -1.28557849e+00
4.14641082e-01 -3.71790826e-01 -1.00080478e+00 5.19474745e-01
-4.95832562e-01 -2.24018753e-01 -6.53448761e-01 2.79440999e-01
-1.62514120e-01 1.04298258e+00 6.84410870e-01 2.03881577e-01
-2.22689331e-01 4.41935182e-01 -6.05977774e-01 5.87871134e-01
8.01576614e-01 -9.57739830e-01 -2.98154522e-02 9.13554281e-02
-5.26723415e-02 1.09831524e+00 8.21973562e-01 -7.62574553e-01
7.41790459e-02 -4.92992759e-01 8.75953317e-01 -8.95617485e-01
1.87402368e-01 -8.49855483e-01 4.77984935e-01 7.19415426e-01
-3.84140968e-01 -2.21675605e-01 -4.33413237e-02 9.33592737e-01
9.75095015e-03 -3.55323374e-01 6.90769315e-01 4.96833205e-01
1.55678108e-01 5.03655970e-01 -2.36181408e-01 -1.33276105e-01
8.42591286e-01 -1.65894121e-01 -3.29623610e-01 -6.73669219e-01
-9.60123003e-01 4.67224777e-01 4.24251109e-01 6.11562356e-02
7.32006431e-01 -1.66646492e+00 -8.93111169e-01 9.58003998e-02
-1.47695109e-01 -3.69868070e-01 3.99770766e-01 7.44215965e-01
-2.72896707e-01 2.81301230e-01 1.33149087e-01 -5.48920073e-02
-7.73082316e-01 8.76078308e-01 5.38382590e-01 -4.09734547e-01
-4.56169486e-01 5.71497381e-01 3.77957314e-01 -5.57631135e-01
4.40898031e-01 -8.77706241e-03 -3.35623235e-01 -2.60405213e-01
4.39321071e-01 7.84465551e-01 -4.93260771e-01 -2.95847028e-01
-3.65194380e-01 7.49841332e-02 2.13264838e-01 9.21611637e-02
1.17188382e+00 -1.11585028e-01 1.73300669e-01 7.68264055e-01
1.14196634e+00 -6.15038037e-01 -1.58569908e+00 -7.30610415e-02
-1.61356166e-01 -4.33148533e-01 2.64942259e-01 -1.34238851e+00
-7.47947037e-01 8.14114451e-01 9.00938272e-01 9.97921918e-03
1.21877837e+00 -2.44018465e-01 7.65940368e-01 -1.70656458e-01
2.84803867e-01 -6.05517387e-01 -4.64516908e-01 3.09800953e-01
9.61659431e-01 -1.17772162e+00 -6.38789684e-02 -2.09682763e-01
-4.45159018e-01 1.16459978e+00 3.72214139e-01 -3.20222765e-01
1.09536505e+00 1.92083210e-01 -8.49630535e-02 2.48844653e-01
-7.02782810e-01 6.62538528e-01 2.73811281e-01 6.33889496e-01
6.09860599e-01 3.42075050e-01 -7.39126861e-01 1.69513738e+00
4.05095667e-02 7.38500774e-01 2.44716957e-01 6.68951809e-01
-2.41798699e-01 -8.78790081e-01 -7.11660147e-01 6.84488356e-01
-7.40093112e-01 -1.75466001e-01 1.29580617e-01 6.68943465e-01
-3.31432849e-01 8.44163239e-01 -5.21314070e-02 2.91241139e-01
8.51263329e-02 2.46969074e-01 7.10089356e-02 3.42073590e-02
-3.70287865e-01 -3.13759409e-02 -2.75030673e-01 -5.59691638e-02
3.56228173e-01 -7.94789135e-01 -1.25562561e+00 -3.21588159e-01
-6.68716729e-01 1.27076060e-01 3.18652123e-01 5.76949894e-01
1.16172507e-01 6.43452823e-01 6.43952489e-01 -3.94257188e-01
-1.19064605e+00 -1.11899197e+00 -1.14609444e+00 -7.44578317e-02
6.33889854e-01 -7.89222479e-01 -7.32937098e-01 -2.19094217e-01] | [7.805413722991943, 5.583540916442871] |
eda6132e-6c91-4fa3-940e-ea5255be986f | an-intriguing-property-of-geophysics | 2204.13731 | null | https://arxiv.org/abs/2204.13731v2 | https://arxiv.org/pdf/2204.13731v2.pdf | An Intriguing Property of Geophysics Inversion | Inversion techniques are widely used to reconstruct subsurface physical properties (e.g., velocity, conductivity) from surface-based geophysical measurements (e.g., seismic, electric/magnetic (EM) data). The problems are governed by partial differential equations (PDEs) like the wave or Maxwell's equations. Solving geophysical inversion problems is challenging due to the ill-posedness and high computational cost. To alleviate those issues, recent studies leverage deep neural networks to learn the inversion mappings from measurements to the property directly. In this paper, we show that such a mapping can be well modeled by a very shallow (but not wide) network with only five layers. This is achieved based on our new finding of an intriguing property: a near-linear relationship between the input and output, after applying integral transform in high dimensional space. In particular, when dealing with the inversion from seismic data to subsurface velocity governed by a wave equation, the integral results of velocity with Gaussian kernels are linearly correlated to the integral of seismic data with sine kernels. Furthermore, this property can be easily turned into a light-weight encoder-decoder network for inversion. The encoder contains the integration of seismic data and the linear transformation without need for fine-tuning. The decoder only consists of a single transformer block to reverse the integral of velocity. Experiments show that this interesting property holds for two geophysics inversion problems over four different datasets. Compared to much deeper InversionNet, our method achieves comparable accuracy, but consumes significantly fewer parameters. | ['Youzuo Lin', 'Zicheng Liu', 'Peng Jin', 'Shihang Feng', 'Yinpeng Chen', 'Yinan Feng'] | 2022-04-28 | null | null | null | null | ['geophysics'] | ['miscellaneous'] | [ 2.39679396e-01 -7.32261762e-02 3.83163810e-01 -3.06531787e-01
-7.56506741e-01 -2.26116613e-01 5.28551936e-01 -2.85573572e-01
-5.21208644e-01 7.83467650e-01 -1.61286861e-01 -6.41763747e-01
-3.17263693e-01 -1.01200473e+00 -1.16274750e+00 -1.06372619e+00
-9.86244604e-02 4.12334681e-01 6.58824071e-02 -3.50009501e-01
4.07838494e-01 4.87259477e-01 -1.27125049e+00 -1.86967372e-03
9.30784225e-01 1.40418291e+00 7.48202056e-02 6.71951056e-01
6.62628561e-02 7.45503366e-01 -1.53354526e-01 6.26083389e-02
1.28596380e-01 -3.18442374e-01 -9.13408875e-01 -5.91824472e-01
2.55434185e-01 -7.42971957e-01 -4.42395836e-01 9.21676457e-01
3.76186460e-01 1.93110108e-01 9.48193073e-01 -1.03306735e+00
-6.93716347e-01 2.81669915e-01 -6.61047220e-01 8.42606649e-02
1.74265504e-01 -2.61307687e-01 9.25712764e-01 -9.47592854e-01
3.07878554e-01 8.92702103e-01 1.17330325e+00 1.77842826e-01
-1.35997844e+00 -5.70303679e-01 -4.34945911e-01 2.42166385e-01
-1.42761719e+00 -5.61754584e-01 7.22270668e-01 -3.88690174e-01
1.02574289e+00 1.57248899e-01 6.11998975e-01 1.02797651e+00
2.37268463e-01 4.07775104e-01 8.12764645e-01 -1.98613703e-01
3.00006300e-01 -6.10695779e-02 -7.66848251e-02 5.63630641e-01
3.17730486e-01 -4.30881977e-02 -4.44037795e-01 -9.90677774e-02
8.44919801e-01 -7.83125758e-02 -4.86279219e-01 -1.87896006e-02
-9.46917176e-01 8.51166248e-01 4.42649096e-01 1.07135378e-01
-4.38447297e-01 3.90970796e-01 2.76016265e-01 3.75057131e-01
4.19667363e-01 6.10888898e-01 -4.62002069e-01 -1.54664174e-01
-1.01947248e+00 3.54843587e-01 1.11480653e+00 4.78316158e-01
1.16974747e+00 4.20036286e-01 5.40982842e-01 6.09887421e-01
4.93281066e-01 1.09060812e+00 3.93609554e-01 -9.46660638e-01
3.01365018e-01 -1.18508495e-01 1.34875894e-01 -1.08250427e+00
-5.55332541e-01 -1.91488951e-01 -1.24588513e+00 5.62580209e-03
3.93819094e-01 -3.75280321e-01 -6.96861923e-01 1.73842800e+00
3.21781226e-02 7.15499580e-01 1.60173148e-01 1.03213441e+00
6.69553578e-01 9.23664093e-01 -3.36415917e-01 1.64994061e-01
1.23065567e+00 -2.80998796e-01 -6.99183941e-01 -2.98272848e-01
7.24893093e-01 -3.65893245e-01 7.22557425e-01 1.11955456e-01
-1.04628408e+00 -2.23904938e-01 -1.28367305e+00 -3.39289278e-01
-3.38274926e-01 -1.66205525e-01 7.40096152e-01 2.30596825e-01
-1.18212485e+00 8.08774054e-01 -1.08393931e+00 8.13454241e-02
1.23932905e-01 3.42119873e-01 -4.34847236e-01 3.30997646e-01
-1.67825401e+00 8.45763147e-01 -7.40460753e-02 5.77749193e-01
-6.63792491e-01 -1.19259465e+00 -1.09413862e+00 1.68912604e-01
-2.17658848e-01 -4.92771685e-01 1.10100162e+00 -6.67284250e-01
-1.62896955e+00 5.12806356e-01 -1.47594050e-01 -3.39788139e-01
3.23040485e-01 -1.85359806e-01 -2.94153064e-01 1.62132144e-01
2.35255614e-01 1.16541564e-01 7.30894089e-01 -1.02916694e+00
-2.45540395e-01 -2.23852322e-01 6.43133074e-02 -2.08147720e-01
-3.59137744e-01 -4.76920128e-01 -6.40887991e-02 -3.10033470e-01
5.59755981e-01 -7.48315275e-01 -2.97976993e-02 -2.39814445e-02
-3.58329207e-01 3.28097045e-01 6.10412896e-01 -1.05866766e+00
7.47719347e-01 -2.08092403e+00 1.28324345e-01 3.39266270e-01
1.37200341e-01 -3.26673081e-03 -6.99260756e-02 4.47248042e-01
-2.22378716e-01 1.38299063e-01 -7.64398813e-01 -1.69669256e-01
6.33931011e-02 1.89024359e-01 -7.17542410e-01 9.14958000e-01
3.49421620e-01 9.73691046e-01 -6.82515144e-01 -4.06340323e-02
-1.00451455e-01 8.20776820e-01 -6.90323055e-01 3.13957006e-01
3.13947201e-02 6.42728865e-01 -5.12584746e-01 2.60324121e-01
1.22027469e+00 -3.33186388e-01 1.05286181e-01 -4.28188413e-01
-2.62051523e-01 6.13336205e-01 -1.10881817e+00 1.64231181e+00
-9.84297216e-01 8.55779171e-01 3.25227559e-01 -1.79960585e+00
9.13038850e-01 3.79450977e-01 4.27244484e-01 -1.16895163e+00
5.60931414e-02 6.90024316e-01 -6.70506805e-02 -8.28015387e-01
2.78123051e-01 -5.96122980e-01 3.13115902e-02 6.30856454e-01
-4.32931446e-02 -4.14416283e-01 -1.33435577e-01 -1.17714360e-01
9.63064730e-01 1.14046007e-01 -1.25383258e-01 -3.40017945e-01
4.49918747e-01 -3.24813515e-01 3.53276730e-01 6.98245227e-01
5.88800311e-01 7.23165154e-01 7.58859754e-01 -3.07699531e-01
-1.08337915e+00 -1.08221531e+00 -5.26145756e-01 6.27728343e-01
5.07277623e-02 1.78815156e-01 -6.17214620e-01 2.63832361e-01
1.26915008e-01 5.25332510e-01 -6.94164932e-01 -3.76865536e-01
-8.56933892e-01 -8.27736974e-01 1.06296611e+00 6.08285427e-01
8.08432281e-01 -7.71331310e-01 -6.50681674e-01 1.71123832e-01
-3.21386725e-01 -1.33669448e+00 5.76543435e-02 4.54793453e-01
-9.36425924e-01 -6.13893747e-01 -7.51684070e-01 -5.84904730e-01
4.54541951e-01 -8.65119770e-02 8.68145466e-01 -3.17152892e-03
-4.01038537e-03 2.94769052e-02 -6.88372254e-02 -1.62824288e-01
-1.46508873e-01 1.51073173e-01 5.26431762e-02 3.21426123e-01
4.55891974e-02 -1.09963274e+00 -5.33754587e-01 1.10808037e-01
-1.12106800e+00 5.79069033e-02 4.25186515e-01 8.59684646e-01
3.00753266e-01 -2.81107366e-01 5.09157419e-01 -6.43867671e-01
3.77765507e-01 -7.35506594e-01 -7.77382314e-01 -9.22635384e-03
-1.94858432e-01 3.94100100e-01 7.53001988e-01 -3.16265613e-01
-1.18949378e+00 -3.86457771e-01 -5.83071589e-01 -4.92814071e-02
2.10162967e-01 1.03686452e+00 1.05141096e-01 -4.33147252e-01
4.75585520e-01 3.73691350e-01 1.04057975e-01 -6.63936257e-01
-1.34487405e-01 6.86604381e-01 7.09797323e-01 -7.01660693e-01
7.23684072e-01 8.00210536e-01 3.44571471e-01 -1.12095857e+00
-7.30652690e-01 -1.60213932e-01 -4.87275571e-01 1.64792210e-01
7.08964765e-01 -8.17424119e-01 -1.10424125e+00 6.56390369e-01
-1.35714602e+00 -4.15157557e-01 -1.53882697e-01 7.76012897e-01
-4.61273998e-01 2.92329639e-01 -7.48289526e-01 -6.37935817e-01
-1.73305050e-01 -1.05169678e+00 1.10062015e+00 2.65554208e-02
-9.51903313e-02 -1.36756563e+00 9.32570174e-02 -1.20105214e-01
7.24412441e-01 1.82066098e-01 9.06739712e-01 -1.37207344e-01
-5.21135569e-01 -1.86413497e-01 -6.03478491e-01 4.12173033e-01
-1.15926847e-01 -2.95202583e-01 -1.18513489e+00 -3.99483442e-02
5.75944483e-01 -1.97019041e-01 1.14025247e+00 6.48072958e-01
1.20356071e+00 -1.15467921e-01 1.49590420e-02 1.57265544e+00
1.46658707e+00 -1.81901902e-02 8.22272599e-01 2.78124928e-01
8.12646449e-01 2.51872838e-01 -5.01679666e-02 6.10402167e-01
3.82187903e-01 3.96210045e-01 3.77031326e-01 -2.64936984e-01
2.52312422e-01 3.89785320e-02 2.80722111e-01 1.02581227e+00
-5.35410523e-01 -8.90930742e-02 -1.14910614e+00 3.72608662e-01
-1.48643219e+00 -7.89746404e-01 -3.67710859e-01 2.09952641e+00
8.75022054e-01 -1.74477562e-01 -7.19324946e-01 3.02004188e-01
-2.55709770e-03 2.50636488e-01 -6.46759510e-01 -4.71546084e-01
-2.29397148e-01 6.38189375e-01 8.28722894e-01 7.55227506e-01
-9.28380072e-01 3.91725659e-01 6.10330820e+00 3.98173809e-01
-1.70725715e+00 1.62482157e-01 2.76658028e-01 3.06432366e-01
-7.81909466e-01 -8.87198150e-02 -4.41404670e-01 3.10225338e-01
1.19040442e+00 2.77616326e-02 4.77463275e-01 4.29253839e-02
1.38619795e-01 -1.48851305e-01 -1.12858415e+00 8.95143449e-01
-1.21713407e-01 -1.48348236e+00 -1.00731388e-01 -8.19056109e-02
4.86950338e-01 2.37203136e-01 1.63415387e-01 1.17913164e-01
-1.40961722e-01 -1.23867369e+00 7.02651143e-01 7.73108602e-01
1.22126293e+00 -3.84629637e-01 6.99979186e-01 2.77691722e-01
-1.01377106e+00 2.19015539e-01 -4.15520668e-01 -4.21075881e-01
5.38722157e-01 8.58732104e-01 -3.98116261e-01 4.12863612e-01
7.90067434e-01 8.39390934e-01 2.10730150e-01 6.12617850e-01
-1.97544977e-01 7.03596115e-01 -5.32620728e-01 2.18793154e-01
3.84439677e-01 -5.98944068e-01 2.90149719e-01 1.13974202e+00
7.66435981e-01 2.19930604e-01 -5.59155524e-01 1.14585805e+00
3.67719564e-03 -3.88519943e-01 -6.06588900e-01 1.08558960e-01
-4.71061282e-03 9.13417816e-01 -3.34492028e-01 -2.22043023e-01
-4.23197210e-01 7.22426772e-01 1.68654561e-01 9.34792936e-01
-9.82114673e-01 -5.75187206e-01 8.43684196e-01 2.95626193e-01
4.68809873e-01 -4.25792426e-01 -5.18631399e-01 -1.08893442e+00
1.76725373e-01 -5.28495312e-01 -2.08381280e-01 -7.88927495e-01
-9.90680814e-01 3.95541400e-01 -8.20280537e-02 -1.00384867e+00
-2.47287631e-01 -8.17737699e-01 -5.55844367e-01 1.29623675e+00
-1.96063530e+00 -7.59854913e-01 -4.21367079e-01 4.59994078e-01
-2.49769747e-01 2.66492337e-01 8.58219862e-01 6.41267836e-01
-3.13438743e-01 3.73801321e-01 4.01098073e-01 2.59317100e-01
3.86140794e-01 -1.02565289e+00 6.26733840e-01 4.15525168e-01
-4.18347269e-01 6.16298020e-01 7.83986747e-01 -4.25698727e-01
-1.92502439e+00 -6.54798388e-01 8.60693932e-01 2.50914376e-02
9.98836815e-01 -4.02336866e-01 -1.48886585e+00 6.48426056e-01
-4.33792286e-02 3.77908200e-01 3.20802063e-01 -1.82425573e-01
-4.16657478e-01 -1.56400189e-01 -8.57437253e-01 1.48029849e-01
6.94640636e-01 -9.05352235e-01 -3.15774381e-01 1.34116977e-01
4.81665045e-01 -6.12738013e-01 -8.76974225e-01 4.94896293e-01
8.18782568e-01 -9.21792924e-01 1.20120537e+00 -5.10361552e-01
9.57813203e-01 -2.18628850e-02 -2.95462400e-01 -1.30409777e+00
-2.27663480e-02 -4.63174790e-01 -3.65924537e-02 5.18122077e-01
4.91626173e-01 -1.20427167e+00 4.32550639e-01 4.59973305e-01
-2.35791415e-01 -7.40998447e-01 -1.02144337e+00 -7.06152320e-01
3.06424111e-01 -5.77580094e-01 5.73894680e-01 1.03043103e+00
-1.65159017e-01 1.94189489e-01 -4.56638098e-01 5.26422143e-01
6.12129271e-01 7.53094032e-02 3.39773476e-01 -1.30925679e+00
-2.48563305e-01 -2.46358708e-01 -2.52541035e-01 -1.46263373e+00
4.21901435e-01 -9.71820176e-01 3.68809938e-01 -1.40301836e+00
-2.07427233e-01 -6.02862597e-01 -1.91780478e-01 2.87629396e-01
2.24927515e-01 4.11586255e-01 -4.22075689e-01 -1.10977935e-02
1.22987531e-01 6.68617487e-01 1.24727547e+00 -1.74099401e-01
1.21215634e-01 -1.16000675e-01 -2.96666086e-01 8.44645500e-01
6.28775179e-01 -4.77973878e-01 -1.49261087e-01 -9.97826159e-01
8.74978602e-01 4.67905045e-01 6.59450948e-01 -8.81186187e-01
5.18807769e-01 2.23042890e-02 1.01752736e-01 -2.37548277e-01
5.13345778e-01 -5.47453463e-01 9.75778624e-02 3.43596011e-01
-1.49609447e-02 -2.71010660e-02 3.12549919e-01 1.72782719e-01
-5.89704394e-01 -3.08180809e-01 6.27684712e-01 1.73318073e-01
-6.12138271e-01 3.08512032e-01 -3.77821714e-01 2.75402367e-01
2.99301773e-01 -1.89110130e-01 -1.72629561e-02 -4.90402937e-01
-3.25894535e-01 -1.01207308e-01 -1.19310338e-02 -1.95953697e-01
6.93317831e-01 -1.13963377e+00 -9.00682449e-01 7.19411790e-01
-4.02431548e-01 2.52326310e-01 3.79583865e-01 1.40181744e+00
-9.67849851e-01 2.54876763e-01 -1.93411395e-01 -6.69905961e-01
-3.20028752e-01 -1.76117152e-01 7.14329243e-01 -6.30489259e-04
-7.63744295e-01 9.98091340e-01 3.88673425e-01 -6.54665649e-01
-1.64447740e-01 -5.30734599e-01 4.68242653e-02 2.02654481e-01
4.56409633e-01 4.75255191e-01 2.47212440e-01 -5.13410747e-01
-4.43220437e-01 9.82263625e-01 3.68012458e-01 -3.08179706e-01
1.71270227e+00 1.50098518e-01 -5.00763655e-01 5.51739395e-01
1.90193570e+00 -2.30239719e-01 -1.17635667e+00 -2.67519772e-01
-3.43326628e-01 -7.14988932e-02 2.74035454e-01 -1.59470156e-01
-1.48651826e+00 1.30183542e+00 9.48190466e-02 3.76048982e-01
1.00856113e+00 -2.52145976e-01 1.01085913e+00 6.84770525e-01
3.32708210e-02 -8.11496258e-01 -3.26243877e-01 1.09693873e+00
8.92536700e-01 -1.13608813e+00 -1.26849748e-02 -1.06751010e-01
1.90817062e-02 1.18512261e+00 1.16444483e-01 -1.90538928e-01
1.14490712e+00 9.17781115e-01 1.03668623e-01 -2.27711424e-01
-3.40627551e-01 3.55779171e-01 9.02954638e-02 3.37496817e-01
4.94498044e-01 -5.64714260e-02 7.50742182e-02 4.45584744e-01
-3.74820322e-01 -2.44188588e-02 5.10881186e-01 6.41301036e-01
-1.72348529e-01 -4.98077393e-01 -3.35442603e-01 4.44529146e-01
-2.80644506e-01 -2.95171469e-01 3.50826591e-01 5.42573333e-01
-3.11736614e-01 4.58155006e-01 5.78023672e-01 -1.00305282e-01
2.10773438e-01 -1.47713095e-01 3.12336057e-01 -1.35034844e-01
-1.64301111e-03 -3.39064479e-01 -1.89190507e-01 -5.34269333e-01
-2.75964290e-01 -5.89907229e-01 -1.65672433e+00 -5.32540858e-01
-9.72903147e-02 2.01865971e-01 8.29917789e-01 1.10463631e+00
6.39706701e-02 7.80671299e-01 4.89272475e-01 -1.18916523e+00
-6.98058784e-01 -9.65015054e-01 -9.45541680e-01 2.91906506e-01
9.68352437e-01 -6.62637889e-01 -7.57124066e-01 -5.50226122e-02] | [6.852907180786133, 2.5395312309265137] |
7e4772dc-1a67-41fb-9638-3a3504be34bd | spatio-temporal-contrastive-learning-enhanced | 2209.11461 | null | https://arxiv.org/abs/2209.11461v2 | https://arxiv.org/pdf/2209.11461v2.pdf | Spatio-Temporal Contrastive Learning Enhanced GNNs for Session-based Recommendation | Session-based recommendation (SBR) systems aim to utilize the user's short-term behavior sequence to predict the next item without the detailed user profile. Most recent works try to model the user preference by treating the sessions as between-item transition graphs and utilize various graph neural networks (GNNs) to encode the representations of pair-wise relations among items and their neighbors. Some of the existing GNN-based models mainly focus on aggregating information from the view of spatial graph structure, which ignores the temporal relations within neighbors of an item during message passing and the information loss results in a sub-optimal problem. Other works embrace this challenge by incorporating additional temporal information but lack sufficient interaction between the spatial and temporal patterns. To address this issue, inspired by the uniformity and alignment properties of contrastive learning techniques, we propose a novel framework called Session-based Recommendation with Spatio-Temporal Contrastive Learning Enhanced GNNs (RESTC). The idea is to supplement the GNN-based main supervised recommendation task with the temporal representation via an auxiliary cross-view contrastive learning mechanism. Furthermore, a novel global collaborative filtering graph (CFG) embedding is leveraged to enhance the spatial view in the main task. Extensive experiments demonstrate the significant performance of RESTC compared with the state-of-the-art baselines e.g., with an improvement as much as 27.08% gain on HR@20 and 20.10% gain on MRR@20. | ['Yang Wang', 'Guangyong Chen', 'Ting Guo', 'Boyu Li', 'Jiezhong Qiu', 'Xin Liu', 'Benyou Wang', 'Zhongwei Wan'] | 2022-09-23 | null | null | null | null | ['session-based-recommendations'] | ['miscellaneous'] | [-1.23123839e-01 -4.68030035e-01 -4.88969475e-01 -4.04939920e-01
-2.62922913e-01 -4.39361155e-01 5.52045047e-01 3.08691114e-01
-2.63582468e-01 1.14260152e-01 5.44214785e-01 -4.32014257e-01
-6.86631858e-01 -7.71474004e-01 -4.93193865e-01 -6.32775486e-01
-5.63053191e-01 -7.10103586e-02 3.14300954e-01 -6.83979273e-01
1.55714244e-01 2.18433022e-01 -1.00295985e+00 4.39234555e-01
9.90027845e-01 1.07621992e+00 3.50820541e-01 3.91978562e-01
-7.24691227e-02 5.61455488e-01 -1.32156452e-02 -3.36849660e-01
2.49423683e-01 -5.58422327e-01 -5.35960257e-01 1.31013379e-01
3.24081033e-01 -1.54917985e-01 -9.91135359e-01 8.22887957e-01
3.83892864e-01 9.23709929e-01 4.54401702e-01 -9.66965020e-01
-1.34958363e+00 6.64557219e-01 -8.54699016e-01 4.89623934e-01
5.03403783e-01 -7.53820688e-02 1.47580874e+00 -8.60591352e-01
3.49309832e-01 1.01206636e+00 5.59775114e-01 3.14435124e-01
-1.18548179e+00 -2.95576483e-01 1.03405881e+00 6.39538467e-01
-1.27987289e+00 1.53705418e-01 1.08940840e+00 -3.51398662e-02
8.60801876e-01 4.31906074e-01 6.96226835e-01 1.22395015e+00
1.04751311e-01 7.64986753e-01 8.28008771e-01 -1.04803860e-01
-8.01033527e-02 -1.24083176e-01 6.03455126e-01 7.28314340e-01
9.39839557e-02 1.14806965e-01 -6.02611601e-01 -5.39538544e-03
6.95956290e-01 6.68469429e-01 -4.67456043e-01 -3.92732620e-01
-1.00250971e+00 8.35043848e-01 1.10440445e+00 3.99831057e-01
-3.03738862e-01 -2.00754389e-01 3.41180474e-01 5.11908054e-01
7.35475183e-01 2.19864368e-01 -4.63699937e-01 3.90885919e-01
-6.51834846e-01 -5.20972535e-02 5.27430773e-01 8.73589694e-01
5.90273321e-01 3.80817205e-02 -4.40065205e-01 9.50696528e-01
6.60495698e-01 2.52864417e-02 5.72352111e-01 -2.17853695e-01
6.39703572e-01 6.86992884e-01 -2.06519350e-01 -1.38675940e+00
-4.51695353e-01 -1.03432572e+00 -9.26466346e-01 -4.68347281e-01
2.21418336e-01 6.26573637e-02 -1.02775121e+00 1.54715574e+00
2.26216659e-01 4.52181339e-01 -3.85568440e-01 9.97589946e-01
9.72891033e-01 7.93500483e-01 5.34983203e-02 -3.10342133e-01
1.10703123e+00 -1.38972998e+00 -7.19053447e-01 -7.22430572e-02
7.05005884e-01 -4.91546720e-01 1.21451771e+00 1.22979581e-01
-8.64434421e-01 -7.25358248e-01 -1.00153601e+00 -9.18278191e-03
-5.89762270e-01 -1.26489073e-01 7.14257598e-01 3.48481745e-01
-1.09257400e+00 7.71935225e-01 -6.23115361e-01 -2.95351475e-01
1.46379694e-01 3.67790669e-01 -1.94151431e-01 -4.72623199e-01
-1.43170965e+00 4.54454809e-01 1.32646471e-01 2.91894943e-01
-3.31983298e-01 -8.06153595e-01 -8.50549519e-01 2.27534637e-01
5.43200731e-01 -6.57803357e-01 7.11616099e-01 -8.21215808e-01
-1.35371935e+00 1.25021994e-01 -1.86528236e-01 -1.97059691e-01
2.40511522e-01 -4.47738655e-02 -1.00282466e+00 -1.87358603e-01
-2.69794017e-01 -3.23485881e-02 4.51365560e-01 -1.21177518e+00
-7.15160310e-01 -5.83862543e-01 4.77546334e-01 5.24123192e-01
-6.25298560e-01 -4.14023280e-01 -9.47830379e-01 -1.11741602e+00
2.88135231e-01 -1.08237755e+00 -4.45251316e-01 -3.61605644e-01
-1.77417800e-01 -3.97754699e-01 8.88336003e-01 -6.62799656e-01
1.81388712e+00 -2.06476307e+00 2.59164035e-01 4.02710050e-01
2.11967081e-01 2.67156273e-01 -6.08592391e-01 6.81817532e-01
-2.70025190e-02 9.84362885e-03 2.86803901e-01 -2.91386217e-01
-6.25952557e-02 1.70659974e-01 -9.24470052e-02 4.74658817e-01
-5.09763718e-01 1.16863585e+00 -1.11322248e+00 -2.19913805e-03
2.16304824e-01 7.34350860e-01 -6.68436587e-01 1.99345440e-01
-8.24195445e-02 4.79645908e-01 -4.63057876e-01 1.68376118e-01
6.91116631e-01 -7.57018328e-01 6.13519311e-01 -4.54715073e-01
1.79952189e-01 7.07445502e-01 -1.02677691e+00 1.96177733e+00
-5.76295972e-01 4.31184173e-02 -2.58001238e-01 -9.78724957e-01
6.41064584e-01 1.56081319e-01 5.24837971e-01 -1.10653722e+00
-2.89839923e-01 -1.83939427e-01 9.62116718e-02 -2.26026207e-01
6.32167935e-01 3.33655775e-01 1.02640986e-01 4.26136732e-01
1.53414771e-01 9.76650774e-01 -1.51912151e-02 6.21246934e-01
1.05686009e+00 2.37417579e-01 1.15063883e-01 -2.01763317e-01
5.56320608e-01 -5.41721523e-01 4.38573301e-01 7.40974605e-01
6.75157756e-02 4.63923275e-01 2.51306277e-02 -3.26331049e-01
-4.16446388e-01 -9.68390524e-01 3.43986422e-01 1.48564684e+00
5.76020181e-01 -7.72607565e-01 6.97493227e-03 -1.22840023e+00
-5.39020225e-02 6.18000925e-01 -8.46045256e-01 -3.39573830e-01
-7.57610857e-01 -7.68776834e-01 -3.66520137e-01 5.70975125e-01
4.45211381e-01 -7.21844912e-01 4.77821141e-01 1.87494248e-01
-2.52231568e-01 -8.77834260e-01 -1.21257603e+00 -1.82102770e-01
-9.74902749e-01 -7.18548179e-01 -8.07602942e-01 -7.81695843e-01
6.82280660e-01 1.16093898e+00 1.03366959e+00 2.29353383e-01
1.90219745e-01 4.90395606e-01 -8.95443320e-01 3.00825387e-01
4.01829362e-01 1.56035528e-01 -3.05622024e-03 4.82272506e-01
2.03668639e-01 -8.87033820e-01 -1.25939143e+00 5.91590047e-01
-7.35093474e-01 -1.04870327e-01 4.13007826e-01 7.95011461e-01
5.66391051e-01 4.58970182e-02 6.78956568e-01 -1.28584349e+00
6.65924251e-01 -9.30868506e-01 -2.91133840e-02 4.36199635e-01
-1.04101455e+00 -1.19795710e-01 9.49488640e-01 -6.28739953e-01
-1.22722578e+00 -4.15801764e-01 -1.01949863e-01 -4.24785703e-01
1.51405409e-01 8.81777525e-01 -9.86573249e-02 -1.74765121e-02
4.00992543e-01 2.80176014e-01 -4.83653873e-01 -7.35396504e-01
5.91941118e-01 2.90448040e-01 -1.89882237e-02 -1.80285826e-01
7.20917046e-01 3.75125229e-01 -1.45658717e-01 -4.20233577e-01
-1.14039743e+00 -1.07807946e+00 -4.13012803e-01 -1.73797131e-01
4.53690082e-01 -8.63191724e-01 -5.98968387e-01 4.19955775e-02
-5.00807822e-01 -3.07673007e-01 -2.59123128e-02 7.13367105e-01
-1.35853589e-01 6.71647489e-01 -8.09204161e-01 -5.18275321e-01
-2.91798800e-01 -6.66984141e-01 5.63282251e-01 1.58282727e-01
2.44587988e-01 -1.38992548e+00 1.36068137e-02 4.03187782e-01
5.52295089e-01 -2.41968483e-01 9.22098398e-01 -8.30936670e-01
-4.85536098e-01 -1.02918319e-01 -3.19488376e-01 -2.28529032e-02
5.61011136e-01 -6.09504163e-01 -4.54407245e-01 -6.13211274e-01
-1.27800465e-01 2.42748827e-01 1.12063479e+00 4.02540058e-01
1.38605678e+00 -4.99619037e-01 -4.25901681e-01 5.76592803e-01
1.50813079e+00 2.00120106e-01 5.34918845e-01 -3.24283801e-02
1.29448295e+00 5.29343903e-01 6.13238335e-01 3.00019920e-01
6.40427470e-01 9.84193742e-01 3.38854820e-01 4.49323319e-02
-3.39421570e-01 -5.34843326e-01 3.37059826e-01 1.24420547e+00
-3.48116964e-01 -7.32955635e-01 -2.52192020e-01 4.65708733e-01
-2.39197159e+00 -1.00719273e+00 -2.99237549e-01 2.25188804e+00
3.46212924e-01 -1.23910280e-02 2.85144657e-01 -2.19891310e-01
7.21975088e-01 6.57849848e-01 -4.89482164e-01 -1.07586682e-01
1.01661325e-01 -1.64845333e-01 5.31971216e-01 4.57916975e-01
-1.03984845e+00 6.03959501e-01 5.20727682e+00 9.93017972e-01
-9.79609370e-01 1.73332348e-01 4.09673929e-01 -1.52868912e-01
-5.99206448e-01 -2.18359202e-01 -5.85013151e-01 5.82471728e-01
9.07220662e-01 1.54233083e-01 7.65134573e-01 2.90092379e-01
3.96220654e-01 3.99312079e-01 -8.81546617e-01 9.34106231e-01
3.22080851e-01 -1.21117675e+00 1.15702108e-01 1.09508581e-01
8.76030564e-01 -3.38470712e-02 4.07108575e-01 6.02379322e-01
3.14355284e-01 -6.73767924e-01 1.80863127e-01 5.25417507e-01
4.69206482e-01 -7.49637067e-01 5.56725919e-01 2.46727854e-01
-1.75577152e+00 -1.46606416e-01 -3.17966193e-01 -5.14778979e-02
2.94877142e-01 4.08044159e-01 -3.12319309e-01 1.33872259e+00
7.22397447e-01 1.31574965e+00 -7.47374356e-01 1.13534784e+00
-1.22572124e-01 8.96129966e-01 -1.15016177e-02 -2.87412805e-03
4.45914805e-01 -6.54936969e-01 6.20140672e-01 1.13184321e+00
3.38454723e-01 1.92413285e-01 3.51459056e-01 3.11865985e-01
-2.41211444e-01 3.80774409e-01 -4.83924657e-01 1.99112102e-01
4.17472422e-02 1.41129291e+00 -5.85738003e-01 3.59414667e-02
-9.96233344e-01 1.25963235e+00 5.13170779e-01 6.68530583e-01
-8.63057673e-01 -1.37256995e-01 3.91859502e-01 3.06829393e-01
7.98721910e-01 -2.37114862e-01 -3.20276730e-02 -1.31124699e+00
1.29809082e-01 -6.26779854e-01 8.39024663e-01 -3.66227120e-01
-1.86119628e+00 4.76767898e-01 -1.75957918e-01 -1.49966848e+00
1.80308387e-01 -3.62556249e-01 -7.64935434e-01 7.93109655e-01
-1.64322901e+00 -1.51353252e+00 -3.78250331e-02 8.03246021e-01
6.73645139e-01 -3.49012651e-02 6.24525547e-01 6.27421975e-01
-5.59222937e-01 9.79118764e-01 2.17838287e-01 -2.63478849e-02
5.86910069e-01 -1.32742119e+00 2.62390673e-01 7.24163294e-01
4.70909923e-01 9.44006562e-01 3.40982884e-01 -5.71503997e-01
-1.57302058e+00 -1.52707231e+00 8.25021267e-01 -2.54489332e-01
6.87195361e-01 -3.66853446e-01 -1.10387838e+00 8.11201811e-01
1.92165092e-01 4.06243145e-01 8.03772867e-01 7.85282910e-01
-7.10882485e-01 -3.15312594e-01 -7.13871479e-01 8.47058356e-01
1.66371047e+00 -5.36172688e-01 -3.14851373e-01 2.92754143e-01
9.11473393e-01 -8.56417120e-02 -8.80230129e-01 3.99059922e-01
5.17371058e-01 -7.78283060e-01 1.12471318e+00 -8.58502984e-01
6.56168237e-02 -3.31287444e-01 -1.19496599e-01 -1.54079330e+00
-1.06239688e+00 -6.27391696e-01 -5.85616767e-01 1.15634322e+00
5.39140821e-01 -5.55538952e-01 6.95820153e-01 2.63701648e-01
-2.75466472e-01 -9.64987993e-01 -4.09519136e-01 -7.72596240e-01
-2.94615060e-01 -1.42919794e-01 3.88189405e-01 1.14967406e+00
1.04685634e-01 7.24651635e-01 -7.84640670e-01 3.71580780e-01
3.47531945e-01 3.88566643e-01 1.38061658e-01 -1.06296849e+00
-5.64680398e-01 -2.40863338e-01 -1.58783436e-01 -1.60055721e+00
-7.33769611e-02 -1.10774207e+00 -2.58611798e-01 -1.92594278e+00
2.37011924e-01 -3.78604025e-01 -1.18578041e+00 1.62655473e-01
-4.05384630e-01 1.95533484e-01 1.96428686e-01 2.23393753e-01
-1.21178961e+00 5.61651230e-01 1.46515179e+00 -1.36994094e-01
-5.32147348e-01 2.61901826e-01 -9.42342103e-01 3.84331971e-01
6.11969531e-01 -2.08785340e-01 -9.16466892e-01 -4.29747939e-01
2.75267541e-01 4.41785380e-02 5.81556112e-02 -5.81422210e-01
4.40608114e-01 -1.91645637e-01 2.76562333e-01 -5.33764541e-01
2.40301207e-01 -9.49525774e-01 9.26925540e-02 1.31724505e-02
-5.57507217e-01 1.76623344e-01 -1.42667338e-01 1.43467021e+00
-9.75036025e-02 2.26789400e-01 3.95596981e-01 2.00944748e-02
-9.20114338e-01 9.33511972e-01 1.35247419e-02 -3.10659498e-01
7.36578524e-01 -9.06401966e-03 -3.19462001e-01 -5.51099062e-01
-9.66505647e-01 3.43953311e-01 6.19021542e-02 7.98346043e-01
6.96857512e-01 -1.65788937e+00 -4.44786787e-01 -1.15776695e-01
3.87934059e-01 -5.50213039e-01 8.74531686e-01 9.66523647e-01
2.37234086e-01 4.90253597e-01 1.23133898e-01 -2.01855332e-01
-1.13005471e+00 9.62852657e-01 1.07959077e-01 -7.99533308e-01
-6.81479692e-01 7.76419461e-01 4.57844734e-01 -3.54616702e-01
2.00583562e-01 8.65601227e-02 -7.07488120e-01 4.31041010e-02
5.24993658e-01 4.08882052e-01 2.27228731e-01 -6.00832462e-01
-2.76917011e-01 4.11154062e-01 -5.09396434e-01 2.47582421e-01
1.43272293e+00 -5.70645928e-01 1.75662100e-01 4.72894430e-01
1.39760280e+00 1.83051929e-01 -1.02789426e+00 -8.66027951e-01
-2.14803010e-01 -5.60433865e-01 2.57707834e-01 -8.63824069e-01
-1.36951125e+00 6.02749586e-01 5.97171307e-01 5.51186800e-01
1.12249386e+00 -1.27115473e-01 9.45431530e-01 1.26580000e-01
3.48866463e-01 -7.69986033e-01 2.55193979e-01 5.04950285e-01
8.42181623e-01 -1.27218020e+00 -2.33197748e-03 -4.66102034e-01
-7.16521502e-01 8.01727414e-01 6.10018075e-01 -3.03551942e-01
1.24266696e+00 -5.53987503e-01 -2.79965550e-01 -2.66850829e-01
-7.48595774e-01 -4.23118234e-01 1.08495593e+00 3.52008820e-01
6.16126359e-01 1.49141133e-01 -6.57062709e-01 7.87927151e-01
5.49379349e-01 -3.73944789e-01 -2.52770688e-02 6.39392078e-01
-1.92305986e-02 -1.22387779e+00 4.45739031e-01 7.44645774e-01
-4.20742989e-01 -2.54111707e-01 3.38228159e-02 5.15183330e-01
-1.71540767e-01 1.20227063e+00 -1.53995976e-01 -8.37012768e-01
4.24737930e-01 -3.19126427e-01 3.15605074e-01 -7.34511137e-01
-7.02257812e-01 2.79693663e-01 3.01438048e-02 -7.12968647e-01
-4.16871637e-01 -3.41868699e-01 -9.64626074e-01 -2.73426026e-01
-4.71544027e-01 1.94536254e-01 1.48478597e-01 8.31815004e-01
7.04426825e-01 7.90604770e-01 1.02568233e+00 -7.61513948e-01
-3.63801450e-01 -8.58684003e-01 -9.14595127e-01 7.22617626e-01
3.20843399e-01 -5.25792360e-01 -3.59387606e-01 -3.12926382e-01] | [10.203490257263184, 5.609766960144043] |
fce62aaf-aa69-4058-8d73-ae1ab06a9859 | feature-selection-approaches-for-optimising | 2212.13369 | null | https://arxiv.org/abs/2212.13369v1 | https://arxiv.org/pdf/2212.13369v1.pdf | Feature Selection Approaches for Optimising Music Emotion Recognition Methods | The high feature dimensionality is a challenge in music emotion recognition. There is no common consensus on a relation between audio features and emotion. The MER system uses all available features to recognize emotion; however, this is not an optimal solution since it contains irrelevant data acting as noise. In this paper, we introduce a feature selection approach to eliminate redundant features for MER. We created a Selected Feature Set (SFS) based on the feature selection algorithm (FSA) and benchmarked it by training with two models, Support Vector Regression (SVR) and Random Forest (RF) and comparing them against with using the Complete Feature Set (CFS). The result indicates that the performance of MER has improved for both Random Forest (RF) and Support Vector Regression (SVR) models by using SFS. We found using FSA can improve performance in all scenarios, and it has potential benefits for model efficiency and stability for MER task. | ['Gengfa Fang', 'Haiyan Lu', 'Sam Ferguson', 'Le Cai'] | 2022-12-27 | null | null | null | null | ['music-emotion-recognition'] | ['music'] | [ 3.21952254e-01 -4.82675940e-01 1.00453787e-01 -4.27677065e-01
-6.32928729e-01 -4.59662348e-01 1.58897176e-01 -3.18809092e-01
-2.97096550e-01 5.59172571e-01 3.72848630e-01 1.46616027e-01
-5.89319408e-01 -5.77040434e-01 4.30582911e-02 -6.54842734e-01
9.65312868e-02 7.91454464e-02 -1.54081419e-01 -3.78783733e-01
4.48696971e-01 6.51310623e-01 -2.22185516e+00 5.69855452e-01
6.22022212e-01 1.24751806e+00 5.52009158e-02 5.41050315e-01
1.73495505e-02 6.16401434e-01 -7.69428015e-01 6.28567487e-02
3.41408610e-01 -5.05308628e-01 -8.43355179e-01 -2.84483880e-01
-1.73109367e-01 2.77782708e-01 1.33557051e-01 4.66425389e-01
8.21360171e-01 3.53276938e-01 5.58915436e-01 -1.50238597e+00
-5.73567562e-02 5.45033872e-01 -3.75885904e-01 -1.03770219e-01
7.24741042e-01 -3.86183619e-01 1.00320208e+00 -1.03059995e+00
5.43910921e-01 1.09733868e+00 8.86861503e-01 4.60261405e-01
-1.14668667e+00 -1.01877522e+00 -5.42942919e-02 2.95938700e-01
-1.50394702e+00 -5.55439174e-01 9.29068983e-01 -4.55167621e-01
1.26647627e+00 9.90571380e-01 1.04036272e+00 9.86725509e-01
9.20565352e-02 4.86263841e-01 1.46079731e+00 -8.39335024e-01
2.98360050e-01 3.39881122e-01 4.31854337e-01 1.28193974e-01
7.42184818e-02 2.92849630e-01 -1.07188451e+00 -6.06533706e-01
3.28166038e-01 -2.91279376e-01 -2.62624204e-01 -6.82022870e-02
-8.62306178e-01 9.97896612e-01 -1.44361272e-01 4.13231224e-01
-5.68113148e-01 -3.14159036e-01 5.12005091e-01 8.82094860e-01
1.92242801e-01 9.31921721e-01 -7.52155542e-01 -3.14577222e-01
-1.13623250e+00 2.51580387e-01 9.37997282e-01 2.65499890e-01
4.72334325e-01 4.03096259e-01 -1.95294559e-01 1.21629477e+00
1.42473474e-01 2.35386953e-01 8.28203261e-01 -1.00943625e+00
-1.84067249e-01 7.06179738e-01 4.66445182e-03 -1.25008821e+00
-6.55919254e-01 -5.00447273e-01 -7.27623284e-01 3.31807435e-01
3.43094543e-02 -1.01959147e-01 -6.35154784e-01 1.57931995e+00
1.68230250e-01 -3.17744762e-02 1.16643064e-01 8.87092650e-01
9.78019238e-01 3.74551296e-01 -1.91752493e-01 -6.57526374e-01
9.00662005e-01 -7.15700567e-01 -8.16754818e-01 3.04275230e-02
5.08027136e-01 -1.09084404e+00 1.01716375e+00 1.18858802e+00
-4.56922084e-01 -6.77312791e-01 -9.16485012e-01 5.72018921e-01
-2.64494091e-01 3.07812542e-01 1.04781771e+00 9.48435783e-01
-9.08109426e-01 6.26700997e-01 -3.63722324e-01 -3.09656739e-01
-7.55618960e-02 7.95561731e-01 -5.80484509e-01 3.07138652e-01
-1.10678601e+00 6.72926366e-01 5.15536219e-02 7.55496994e-02
-2.44091243e-01 -4.67472076e-02 -5.77746153e-01 -8.25764984e-02
3.27981226e-02 -5.35540342e-01 9.03905034e-01 -1.28010809e+00
-1.75982583e+00 3.49016070e-01 -3.57747823e-01 -1.19655922e-01
1.53033122e-01 -4.01254028e-01 -7.48342454e-01 -1.28779784e-01
-1.89187959e-01 1.82508647e-01 1.11284876e+00 -1.17678750e+00
-5.34364879e-01 -3.58154714e-01 -5.71626008e-01 1.91164687e-01
-2.09831625e-01 5.50326526e-01 2.75022775e-01 -7.17313945e-01
4.91837084e-01 -9.58491802e-01 -9.03789103e-02 -6.54805422e-01
-2.55099952e-01 -9.98333097e-02 6.65784001e-01 -5.90031743e-01
1.82601893e+00 -2.24079394e+00 -9.43432152e-02 7.64671922e-01
-3.21323216e-01 1.09257340e-01 -1.17240965e-01 5.63361645e-01
-5.18792331e-01 1.60380691e-01 5.09625003e-02 2.54046410e-01
-3.25277507e-01 1.17601424e-01 -4.00536209e-01 1.87107936e-01
6.45022914e-02 4.65665162e-01 -3.91289204e-01 -3.66746813e-01
3.80731374e-02 5.00193357e-01 -5.51128983e-01 1.13437094e-01
4.17826146e-01 2.43139192e-01 -3.89962494e-01 8.22469175e-01
4.66589034e-01 2.75609225e-01 7.76297525e-02 -2.36097142e-01
-1.84399411e-01 1.74131706e-01 -1.90122199e+00 1.15111995e+00
-1.88266918e-01 5.11510253e-01 -4.94386479e-02 -8.21111560e-01
1.57120383e+00 4.90070313e-01 7.60523438e-01 -3.47003013e-01
2.54733376e-02 1.98986113e-01 2.05583930e-01 -5.86136281e-01
4.51997668e-01 -2.92482048e-01 -4.01089564e-02 4.69510227e-01
1.60744548e-01 1.57625359e-02 -9.11915451e-02 -2.03385189e-01
1.05382586e+00 1.04343794e-01 5.84770918e-01 1.42443748e-02
4.63907599e-01 1.16122685e-01 8.51620078e-01 6.85870886e-01
-2.51884051e-02 6.20010436e-01 1.76589280e-01 -3.03474754e-01
-3.56069446e-01 -6.36086106e-01 -1.18108556e-01 1.09146154e+00
-3.61503154e-01 -8.41214955e-01 -1.76468655e-01 -5.95457435e-01
-8.01421236e-03 6.57060027e-01 -4.25575286e-01 -3.24438691e-01
-2.33792067e-01 -7.85188258e-01 4.61692452e-01 1.42058015e-01
5.80500998e-02 -1.30575228e+00 -8.03963184e-01 2.41611868e-01
-2.31066197e-01 -3.44922453e-01 5.51231764e-03 6.77759469e-01
-7.62638867e-01 -9.16271985e-01 -1.80249184e-01 -3.80466193e-01
6.68893605e-02 3.29986542e-01 9.17544007e-01 -4.63078432e-02
-2.90987879e-01 3.59169006e-01 -9.42217946e-01 -6.48985982e-01
-1.90466374e-01 -2.42485981e-02 3.18643212e-01 1.83763817e-01
4.08330917e-01 -5.09793758e-01 -3.41175109e-01 4.84526753e-01
-5.94915748e-01 -3.29429746e-01 5.29761851e-01 9.77596641e-01
5.69226503e-01 4.72645044e-01 1.03683507e+00 -7.86917925e-01
8.74000847e-01 -1.87678531e-01 1.43871635e-01 3.85334753e-02
-1.06832576e+00 -2.11577415e-01 2.53562361e-01 -5.06099284e-01
-8.23354602e-01 6.15321696e-01 -2.83625901e-01 -3.83359879e-01
-1.38953775e-01 6.51957154e-01 -4.73367907e-02 -3.46712023e-01
8.08720946e-01 4.85251620e-02 7.07797557e-02 -8.26973736e-01
-1.75671607e-01 1.26874149e+00 1.10142762e-02 -2.67612845e-01
4.07620400e-01 9.01359469e-02 -1.15238400e-02 -9.24571216e-01
-7.06780374e-01 -7.69541383e-01 -6.19012773e-01 -4.77142662e-01
3.54558617e-01 -7.47534275e-01 -4.78053808e-01 2.62225002e-01
-5.50955772e-01 3.44125122e-01 -4.85944599e-01 9.19005096e-01
-5.18319190e-01 2.79442295e-02 -1.08592905e-01 -1.33681059e+00
-5.11826456e-01 -8.61122847e-01 8.83127749e-01 8.72722361e-03
-9.78042960e-01 -1.96080148e-01 2.61431962e-01 1.92968130e-01
3.62435699e-01 4.21077162e-01 7.85291612e-01 -9.29100215e-01
3.17399323e-01 -3.64753097e-01 3.74125391e-01 4.27943468e-01
2.18655765e-01 3.81870985e-01 -1.36415398e+00 -1.26109496e-01
1.96813390e-01 -3.54465127e-01 8.77128065e-01 3.19638461e-01
8.43004704e-01 -1.44666925e-01 -6.73761442e-02 4.63789165e-01
1.09166610e+00 3.60009432e-01 6.02808833e-01 6.38374805e-01
2.31857747e-01 7.37437248e-01 1.13580859e+00 7.28741765e-01
2.66430881e-02 7.50544369e-01 7.78118819e-02 3.97086516e-03
1.48025259e-01 -7.52465203e-02 6.23339236e-01 9.11819696e-01
-2.14340955e-01 2.16432646e-01 -6.25675142e-01 7.93244615e-02
-1.85586691e+00 -1.13735473e+00 -3.60624850e-01 2.25445890e+00
5.83929121e-01 -2.39554077e-01 5.59637427e-01 1.14633262e+00
4.03469682e-01 -1.59753159e-01 -2.75262386e-01 -9.07858193e-01
-3.72004032e-01 5.57624936e-01 -5.33713996e-02 -6.63758628e-03
-1.08876371e+00 6.29936755e-01 7.19432259e+00 7.11807191e-01
-1.35348904e+00 -2.39979222e-01 1.75994873e-01 -2.00326145e-01
-5.63045107e-02 4.20852974e-02 -5.64464569e-01 1.09335154e-01
9.44988370e-01 -1.45061612e-01 5.34978509e-01 9.61233795e-01
2.88222760e-01 -6.20240383e-02 -5.13765633e-01 1.31823993e+00
2.23008431e-02 -7.25050807e-01 9.02888328e-02 -2.22406745e-01
3.15685868e-01 -4.09961849e-01 -4.00467753e-01 4.45009559e-01
-6.80440888e-02 -1.17598724e+00 6.14419401e-01 7.25976229e-01
6.47005379e-01 -9.04219866e-01 9.11772192e-01 2.06495687e-01
-1.09876597e+00 -3.28131914e-01 -1.62059218e-01 -4.40756798e-01
-2.57637739e-01 5.30220211e-01 -7.62118876e-01 6.85381770e-01
9.89995658e-01 6.13791287e-01 -7.66078830e-01 1.13709688e+00
2.36380935e-01 8.15597534e-01 -4.41690981e-01 -2.11354680e-02
-4.59053278e-01 -1.44202501e-01 6.87684536e-01 1.17819691e+00
5.68757892e-01 -1.30971387e-01 -8.96579865e-03 1.28284007e-01
7.36552298e-01 6.61522806e-01 -4.55952257e-01 -8.14280510e-02
4.44381684e-01 1.38394046e+00 -8.73941481e-01 1.67109862e-01
-2.01857284e-01 7.04243600e-01 -1.15289778e-01 4.24281731e-02
-3.67787093e-01 -5.23387432e-01 5.68413496e-01 6.44399747e-02
2.32127175e-01 1.11895919e-01 -4.48105425e-01 -1.01468146e+00
-2.32130066e-01 -1.25917614e+00 7.15027750e-01 -9.02779877e-01
-1.20745218e+00 9.33930576e-01 -2.85613477e-01 -1.61595297e+00
-3.82155210e-01 -3.41568083e-01 -3.80394161e-01 8.41624081e-01
-8.97371888e-01 -7.92573512e-01 -1.98700428e-01 8.00502062e-01
3.64870101e-01 -3.97928268e-01 1.44721949e+00 9.82177705e-02
-3.55245113e-01 5.89481354e-01 -8.33870843e-02 -4.07739639e-01
9.92028654e-01 -9.69300449e-01 -4.63146687e-01 2.83788323e-01
5.40961742e-01 5.47905982e-01 8.25997531e-01 -6.29413009e-01
-1.43599355e+00 -8.02501678e-01 9.88640547e-01 -1.46527588e-01
1.60282910e-01 -6.88145757e-02 -6.96351171e-01 2.69324511e-01
-2.38790810e-01 -4.55866218e-01 1.37358963e+00 5.15164971e-01
-2.52492309e-01 -1.93691567e-01 -1.31613493e+00 1.17005274e-01
7.47563183e-01 -3.10946763e-01 -5.58073163e-01 -1.33669630e-01
2.84962773e-01 7.48795867e-02 -9.84155953e-01 6.91560507e-01
1.16595507e+00 -1.11200368e+00 7.74300158e-01 -5.39297581e-01
1.47960056e-02 -4.11560476e-01 -5.54711640e-01 -1.34960043e+00
-5.40125608e-01 -5.54132938e-01 -1.75145697e-02 1.47242796e+00
4.95882422e-01 -6.74015045e-01 4.55151767e-01 4.27519977e-01
2.77570397e-01 -8.33121300e-01 -8.96259546e-01 -8.78486514e-01
-4.84598756e-01 -8.06627035e-01 6.60466552e-01 1.06194174e+00
4.01039608e-02 6.74086034e-01 -5.58817506e-01 -2.74855226e-01
-2.47371029e-02 5.39806426e-01 7.87631154e-01 -1.86784518e+00
-3.58269572e-01 -3.03031504e-01 -5.72348595e-01 6.36923537e-02
5.45700416e-02 -7.53761649e-01 -3.17261755e-01 -1.27961659e+00
1.91428810e-02 -3.82851750e-01 -6.38181806e-01 6.52445674e-01
4.73585837e-02 3.33525449e-01 2.69996136e-01 3.37996691e-01
-9.70895216e-02 3.45509410e-01 7.82323301e-01 2.53776938e-01
-7.46091604e-01 5.22611380e-01 -9.81103241e-01 7.11559236e-01
9.69505012e-01 -7.35994577e-01 -2.61705726e-01 5.51502764e-01
3.30940336e-01 -3.55218947e-02 3.29923853e-02 -8.82233560e-01
-1.47682101e-01 -2.03806713e-01 7.58860588e-01 -5.17481148e-01
5.10219157e-01 -8.49803686e-01 7.19221950e-01 2.51029849e-01
-2.48729527e-01 5.34572825e-02 9.36259329e-02 2.35000595e-01
-4.50114906e-01 -2.66641438e-01 5.18115044e-01 5.67402728e-02
-6.53964460e-01 -4.19887275e-01 -4.77490246e-01 -5.01016498e-01
7.81340718e-01 -5.95597327e-01 2.92957455e-01 -5.61198413e-01
-1.09069145e+00 -3.21406752e-01 2.13191360e-01 5.44983625e-01
8.22846353e-01 -1.27109146e+00 -5.62581956e-01 6.49658918e-01
4.56482507e-02 -7.50422776e-01 8.33549574e-02 7.92174637e-01
2.71089803e-02 1.81847945e-01 -3.26706409e-01 -3.31733406e-01
-1.95456696e+00 3.23598891e-01 2.00622268e-02 -2.03715190e-01
-3.10363472e-01 7.94623137e-01 -5.26890635e-01 -5.73036313e-01
2.19294876e-01 7.36875609e-02 -8.70639920e-01 4.56863165e-01
4.74151194e-01 6.22945368e-01 2.64480382e-01 -9.02130783e-01
-5.14305592e-01 6.73028827e-01 2.32321098e-01 -1.94937006e-01
1.61156929e+00 -3.40058431e-02 -1.60596669e-01 9.18352127e-01
7.50667810e-01 3.28656137e-01 -3.40582222e-01 2.26668820e-01
1.96013495e-01 -6.35075569e-01 2.30579108e-01 -1.02552998e+00
-8.52913618e-01 3.35039258e-01 8.91847134e-01 3.32876682e-01
1.65993416e+00 -4.02708709e-01 1.81871429e-01 3.45436215e-01
4.53093648e-01 -1.17898893e+00 -3.86947453e-01 5.11975110e-01
1.18545926e+00 -8.34425330e-01 1.68877542e-01 -4.91750032e-01
-1.05606294e+00 1.24501133e+00 3.02974403e-01 -2.08164304e-01
1.00425243e+00 2.77912050e-01 2.26340130e-01 -1.07481569e-01
-1.04594851e+00 -3.27054948e-01 5.07867754e-01 6.59396887e-01
6.71872675e-01 5.73207438e-02 -6.02036417e-01 1.44847047e+00
-8.78421903e-01 2.51522660e-01 2.74639815e-01 9.25547063e-01
-4.03912812e-01 -1.26744533e+00 -5.19683361e-01 1.07405019e+00
-5.03724694e-01 6.66245222e-02 -9.15344596e-01 7.75178671e-01
2.51799315e-01 1.43985128e+00 -4.99615341e-01 -1.06688559e+00
5.28652489e-01 4.56622809e-01 2.66271204e-01 -4.97209132e-01
-1.12984180e+00 4.31419164e-01 3.62832636e-01 -6.57174826e-01
-4.92181271e-01 -9.17774439e-01 -8.94018710e-01 1.33479238e-01
-8.85829508e-01 4.41793382e-01 7.67988563e-01 6.20017052e-01
4.53917563e-01 4.70263958e-01 1.15368521e+00 -6.21234655e-01
-5.24537444e-01 -9.52259302e-01 -1.03769743e+00 2.51463056e-01
4.92267907e-02 -8.82512450e-01 -6.05681479e-01 -8.24088007e-02] | [15.846609115600586, 5.194525241851807] |
c634dc18-0e9a-4f78-a806-5ee704b650b3 | construction-d-un-systeme-de-recommandation | 2306.03247 | null | https://arxiv.org/abs/2306.03247v1 | https://arxiv.org/pdf/2306.03247v1.pdf | Construction d'un système de recommandation basé sur des contraintes via des graphes de connaissances | Knowledge graphs in RDF model entities and their relations using ontologies, and have gained popularity for information modeling. In recommender systems, knowledge graphs help represent more links and relationships between users and items. Constraint-based recommender systems leverage deep recommendation knowledge to identify relevant suggestions. When combined with knowledge graphs, they offer benefits in constraint sets. This paper explores a constraint-based recommender system using RDF knowledge graphs for the vehicle purchase/sale domain. Our experiments demonstrate that the proposed approach efficiently identifies recommendations based on user preferences. | ['Philippe Gouspillou', 'Marie-Hélène Abel', 'Ngoc Luyen Le'] | 2023-06-05 | null | null | null | null | ['knowledge-graphs'] | ['knowledge-base'] | [-7.81518698e-01 4.55538094e-01 -1.15581930e+00 -6.84045672e-01
2.77212530e-01 -5.54019272e-01 2.76964337e-01 5.52830100e-01
-5.49066029e-02 5.79892397e-01 6.70843840e-01 -5.21290712e-02
-8.41732860e-01 -1.38557708e+00 -2.94900864e-01 3.21789265e-01
-2.51307398e-01 7.54597247e-01 4.91357356e-01 -7.24201441e-01
3.55176598e-01 3.70550722e-01 -1.72495520e+00 5.76382875e-01
1.02441442e+00 8.72432470e-01 -4.17315215e-02 -1.24894835e-01
-6.79365337e-01 9.04569864e-01 -6.12811670e-02 -8.94788682e-01
8.48395973e-02 1.19929217e-01 -6.12481296e-01 -4.32426065e-01
4.50789124e-01 -1.48136213e-01 -5.50307214e-01 7.92347252e-01
1.00369990e-01 7.39665985e-01 5.60979903e-01 -1.30539548e+00
-1.34458089e+00 1.20013356e+00 2.06817016e-01 6.50015995e-02
9.27580357e-01 -9.26452160e-01 1.56096160e+00 -1.00527990e+00
1.33953953e+00 1.04355204e+00 3.89389008e-01 4.53353137e-01
-5.05017459e-01 -5.32615662e-01 4.68458176e-01 8.56233656e-01
-1.52279270e+00 -3.91964875e-02 4.38784361e-01 -3.82066995e-01
1.67629373e+00 4.58486080e-01 1.26647198e+00 2.12849498e-01
-1.76545933e-01 5.48065960e-01 1.17959470e-01 -2.71173000e-01
3.72598231e-01 4.22973096e-01 5.16206682e-01 3.58756781e-01
8.12492192e-01 -1.75426483e-01 -7.36013651e-01 -1.10055231e-01
3.87033880e-01 4.65189248e-01 -9.11146700e-02 -4.03198987e-01
-2.47984737e-01 9.25759614e-01 5.71454704e-01 2.16535151e-01
-4.27137554e-01 -1.31236717e-01 1.27080232e-01 2.87196785e-01
3.40865791e-01 7.50404954e-01 -3.45181286e-01 3.34556162e-01
-3.81291121e-01 3.52362573e-01 1.24716377e+00 1.95272553e+00
6.87888265e-01 -2.77740777e-01 2.68381298e-01 9.09319580e-01
9.88275766e-01 4.76155967e-01 9.92896631e-02 -6.63632274e-01
2.02131495e-01 1.07769287e+00 1.81754172e-01 -1.35489368e+00
-4.49201316e-01 -3.38934034e-01 9.17955786e-02 -4.56647307e-01
-3.23184997e-01 1.10355042e-01 -5.03254712e-01 8.57407093e-01
4.39474136e-01 1.76316842e-01 2.64490962e-01 9.99968410e-01
1.46365952e+00 3.42346728e-01 3.45816851e-01 -2.31664598e-01
1.18006504e+00 -8.94139886e-01 -9.68951941e-01 1.56460181e-01
8.54727030e-01 -5.88568211e-01 4.94877249e-01 2.35765040e-01
-7.12757587e-01 -1.60907224e-01 -8.28518867e-01 1.41523346e-01
-1.02186525e+00 -3.41312110e-01 1.16307843e+00 8.72197628e-01
-7.47612476e-01 6.19597137e-01 -2.14470848e-01 -7.46439815e-01
3.45575958e-01 4.11925822e-01 -6.56650821e-03 -3.22286457e-01
-1.52799356e+00 1.10529590e+00 5.63482165e-01 -1.43758595e-01
-1.63643286e-02 -8.34563613e-01 -7.40103006e-01 3.25734913e-01
6.90317869e-01 -7.33605981e-01 5.50090849e-01 -4.21282440e-01
-1.22407198e+00 1.36810258e-01 1.89313278e-01 -3.21883112e-01
-3.99722964e-01 -2.83054322e-01 -1.56412458e+00 -2.35959277e-01
-3.67152929e-01 -4.20922711e-02 1.63979620e-01 -1.07251811e+00
-1.08457136e+00 -7.90098906e-02 6.72510028e-01 3.46990675e-01
-3.41207325e-01 2.54899144e-01 -7.89679408e-01 -2.68236458e-01
2.24826727e-02 -6.15277886e-01 -1.79565877e-01 -6.01453006e-01
-1.83209121e-01 -5.06001890e-01 5.10044992e-01 -1.00663297e-01
1.84907722e+00 -1.49300194e+00 -1.57106519e-01 1.13119078e+00
2.00696707e-01 1.99997216e-01 7.28219002e-02 9.70238507e-01
4.05202478e-01 4.56223071e-01 8.76965046e-01 7.15465486e-01
4.45565850e-01 6.01543605e-01 -1.59875378e-01 -1.36713237e-01
-5.77408493e-01 9.77975309e-01 -1.00940752e+00 -4.11751479e-01
8.53032246e-02 3.72887731e-01 -1.14548361e+00 -1.43292680e-01
-6.68798566e-01 -4.12631869e-01 -9.80365098e-01 7.73541570e-01
3.53683382e-01 -2.94437498e-01 1.26531410e+00 -5.78338146e-01
9.51168761e-02 3.84899527e-01 -1.29065156e+00 1.41088235e+00
-4.22234148e-01 2.46401921e-01 -4.45865750e-01 -5.73235095e-01
1.03536177e+00 4.51863781e-02 6.84515238e-01 -9.99893486e-01
-5.68681508e-02 3.45876068e-01 -6.17812015e-02 -7.78614044e-01
1.06369948e+00 3.67126763e-01 1.16793498e-01 2.54734218e-01
6.93913549e-02 2.13180676e-01 5.54444790e-01 6.51550114e-01
9.40924346e-01 2.93547213e-01 6.45673275e-01 -3.45186859e-01
4.82970953e-01 2.27182433e-01 6.18439555e-01 4.09880847e-01
7.40508437e-01 -3.69556427e-01 -2.51178503e-01 -5.51001608e-01
-4.79779869e-01 -6.55617297e-01 -1.60032198e-01 1.28197896e+00
4.74305928e-01 -1.43539798e+00 7.09486827e-02 -8.97870958e-01
7.18852341e-01 8.19692492e-01 -2.71749228e-01 1.07399551e-02
-5.05288482e-01 -2.19271369e-02 -8.11889693e-02 7.70442605e-01
-3.61326486e-01 -8.24749351e-01 3.55354697e-02 2.92691380e-01
3.03454906e-01 -7.92696059e-01 -2.99498141e-01 -2.79087156e-01
-7.77755916e-01 -1.53195286e+00 1.68977246e-01 -6.61451876e-01
7.29641616e-01 3.95119697e-01 1.41418791e+00 8.66925538e-01
3.74012589e-01 8.56712639e-01 -9.93347466e-01 -3.47825855e-01
8.94148573e-02 4.06510532e-02 2.69521475e-01 -5.22756338e-01
8.30577791e-01 -4.59345192e-01 -6.56433403e-01 9.64559197e-01
-5.82008421e-01 -4.48496640e-01 -1.35963961e-01 2.56741047e-01
7.29033947e-01 9.98186618e-02 9.96339738e-01 -1.66162705e+00
6.79362059e-01 -1.07973790e+00 -6.59451902e-01 8.00574780e-01
-1.48985720e+00 -9.42127183e-02 3.28302205e-01 9.17976350e-02
-1.20579350e+00 -3.65674853e-01 4.37738858e-02 -9.67728868e-02
2.77756363e-01 1.43284881e+00 -3.13653469e-01 -3.78132313e-01
7.58939385e-01 -7.74662375e-01 -8.51156652e-01 -6.78487062e-01
1.03118622e+00 4.45865512e-01 -3.23497914e-02 -7.47647464e-01
4.35318440e-01 1.45955443e-01 -9.53654498e-02 -4.39127862e-01
-7.08761096e-01 -9.17953491e-01 -5.49590409e-01 -5.14759541e-01
3.94030005e-01 -8.32620740e-01 -8.90994251e-01 -8.72994900e-01
-6.40445232e-01 2.08995700e-01 -6.18870318e-01 8.01105499e-01
-2.14760050e-01 1.45937085e-01 -1.57093614e-01 -6.09179914e-01
-7.07639605e-02 -2.98896968e-01 -8.93832594e-02 6.26510322e-01
-1.07890606e-01 -1.28537047e+00 1.15839779e-01 4.24698114e-01
6.05462015e-01 -1.32105321e-01 8.81835163e-01 -1.41049480e+00
-8.14392567e-01 -4.15439278e-01 -5.62375113e-02 -3.87190342e-01
1.63249448e-01 8.49446580e-02 -1.73164174e-01 6.68611974e-02
-1.05571902e+00 2.13301703e-01 3.83478642e-01 7.01806843e-02
6.15914643e-01 -5.40084600e-01 -7.12519705e-01 4.33191419e-01
1.61172616e+00 5.53575039e-01 7.60272741e-01 2.90237755e-01
8.55333686e-01 5.81676006e-01 1.14426470e+00 6.79474652e-01
9.06520247e-01 7.81741440e-01 3.30937922e-01 5.89229345e-01
-1.06378369e-01 -3.28841120e-01 -2.62732774e-01 1.17679119e+00
-6.17304265e-01 -6.52482212e-01 -7.09281743e-01 3.23386908e-01
-2.30525136e+00 -1.16123068e+00 -3.03088635e-01 1.99859977e+00
2.90343851e-01 -3.24998021e-01 1.43016055e-01 -5.00562310e-01
5.85315943e-01 -2.89050937e-01 -6.20306432e-01 -4.69624013e-01
-1.51289105e-01 2.17550695e-01 6.54769063e-01 4.28638697e-01
-3.84671897e-01 1.19594550e+00 6.55319262e+00 5.29383421e-01
-3.09463918e-01 -4.05107625e-02 -5.90995312e-01 -2.52365410e-01
-1.16008341e+00 2.51650423e-01 -1.04444873e+00 1.82124469e-02
1.01410544e+00 -1.04136527e+00 4.97645050e-01 1.05406725e+00
-1.09329246e-01 2.25184724e-01 -1.04106486e+00 6.66118562e-01
1.61035925e-01 -2.09047484e+00 2.95567453e-01 2.30381772e-01
9.37172711e-01 -6.81488439e-02 -3.86405051e-01 5.07885814e-01
9.18179095e-01 -8.25490177e-01 3.93845826e-01 1.12753713e+00
3.20194155e-01 -9.06293094e-01 7.05228329e-01 -1.05458036e-01
-1.60602403e+00 -1.77297860e-01 -9.15797710e-01 3.83131355e-02
3.63709688e-01 4.05467004e-01 -8.71940792e-01 1.07317090e+00
5.64468086e-01 1.25190616e+00 -1.26899526e-01 1.27662027e+00
-3.59389663e-01 3.86557102e-01 -2.63192475e-01 -3.32090914e-01
-3.97690058e-01 -5.23045778e-01 1.51253834e-01 1.16976440e+00
4.95089412e-01 7.06398487e-01 2.61615455e-01 4.03052688e-01
-4.72776532e-01 9.10734236e-01 -5.55442333e-01 -1.66411892e-01
1.06222045e+00 1.12211931e+00 -4.54314172e-01 -3.90161753e-01
-8.53113949e-01 -1.15548395e-01 2.55322993e-01 3.38948280e-01
-5.67476451e-01 -2.04273179e-01 9.37880814e-01 3.34847331e-01
5.76547265e-01 3.60794105e-02 2.78407127e-01 -1.23557341e+00
-3.16652745e-01 -4.98252720e-01 1.04849946e+00 -6.09989643e-01
-1.49659133e+00 2.16222450e-01 2.07753330e-01 -1.39949191e+00
-3.90716255e-01 -3.56128931e-01 -2.05439344e-01 4.27925229e-01
-1.59660697e+00 -1.12807703e+00 -1.39178827e-01 7.02549219e-01
-5.34139201e-02 -3.87243181e-01 9.86965835e-01 8.13271523e-01
-2.51511514e-01 4.05436784e-01 2.87571192e-01 -3.38229418e-01
5.03992856e-01 -1.03025770e+00 -4.09923348e-04 3.81257534e-01
4.44922328e-01 1.23906875e+00 3.55210662e-01 -1.26373041e+00
-1.82776392e+00 -9.95219707e-01 8.20360065e-01 -2.73854792e-01
6.18783355e-01 2.22625077e-01 -7.31307924e-01 8.95095289e-01
2.73671627e-01 2.79167920e-01 1.57093334e+00 9.57048476e-01
-5.56134820e-01 -1.41484171e-01 -1.21463847e+00 2.50034600e-01
1.54913723e+00 -4.41634387e-01 -5.04504740e-01 5.42118907e-01
5.82500935e-01 -3.16011548e-01 -1.85420585e+00 2.06481561e-01
9.68755603e-01 -6.54821813e-01 1.19125938e+00 -1.20123088e+00
-9.73405018e-02 -5.55773854e-01 -5.20889342e-01 -1.40999532e+00
-7.39040554e-01 -2.14270294e-01 -8.22150230e-01 1.17198634e+00
8.96957338e-01 -4.94960368e-01 8.71621907e-01 1.06629407e+00
-3.58215630e-01 -5.00546098e-01 -3.64080817e-01 -7.66837239e-01
-4.22414213e-01 -4.26127046e-01 1.33783853e+00 1.37744832e+00
9.81625140e-01 1.05708428e-01 -3.59970629e-01 2.99206346e-01
2.91872770e-01 4.67991114e-01 4.66894716e-01 -1.83136559e+00
-1.23701803e-01 -1.45356014e-01 -5.95236182e-01 -7.67099798e-01
6.60079345e-02 -1.46552169e+00 -6.32616520e-01 -2.17151523e+00
-1.52766034e-01 -1.06626689e+00 -6.35278881e-01 3.52761596e-01
3.56666028e-01 -2.21703537e-02 2.53440320e-01 3.61919403e-01
-1.21724916e+00 1.41094383e-02 9.60255206e-01 -5.37398458e-02
-2.34322309e-01 -6.99694827e-02 -9.38919306e-01 6.30327761e-01
6.29114926e-01 -4.38149542e-01 -8.81880939e-01 -3.86718422e-01
1.57500935e+00 -5.00092879e-02 -5.55504978e-01 -5.30257106e-01
7.94277370e-01 -7.02477694e-01 3.11543029e-02 -4.97813314e-01
6.60265386e-02 -1.29435968e+00 9.50849891e-01 -2.55635172e-01
-4.82398748e-01 -1.96284711e-01 -1.34542689e-01 8.08323383e-01
-1.47746116e-01 -4.92715210e-01 1.12292618e-02 3.30326855e-02
-1.21739721e+00 3.83044004e-01 -2.43473560e-01 8.32216628e-03
7.87701905e-01 -1.22888267e-01 -7.30404854e-01 -3.23091179e-01
-1.18252492e+00 5.28588712e-01 4.25449401e-01 7.18238294e-01
9.60666597e-01 -1.60980856e+00 -2.23984227e-01 -2.98057377e-01
5.54571867e-01 -5.82050920e-01 1.27680510e-01 2.53237009e-01
-2.52894670e-01 7.06076622e-01 -3.31356347e-01 2.29885414e-01
-1.09364367e+00 5.05845964e-01 7.40273343e-03 -1.50205325e-02
-5.61948657e-01 8.10343802e-01 -6.28681719e-01 -4.59748507e-01
1.22263461e-01 -2.19598889e-01 -1.09504271e+00 4.42406982e-01
5.79214811e-01 6.69532478e-01 1.46726936e-01 -5.19005001e-01
-6.02726340e-01 4.28482801e-01 -1.28614441e-01 4.42630589e-01
1.59002507e+00 -3.02473694e-01 1.00407161e-01 -1.61047205e-01
5.33292115e-01 7.23200083e-01 -2.61432886e-01 -5.61351240e-01
5.02421200e-01 -8.69284809e-01 1.01720527e-01 -1.06173277e+00
-1.46866179e+00 -1.74419716e-01 1.09054260e-02 5.76796055e-01
7.75325477e-01 1.56607822e-01 7.26675868e-01 8.72738838e-01
7.75451183e-01 -1.39585781e+00 -5.23535311e-01 5.18535554e-01
7.30513573e-01 -8.63118947e-01 6.85185850e-01 -1.02428925e+00
-7.65934646e-01 1.42511141e+00 7.10334837e-01 -1.16274528e-01
1.18614388e+00 5.50383739e-02 -1.71626002e-01 -7.86820233e-01
-1.01686108e+00 -6.63574994e-01 8.06203187e-01 9.83558416e-01
7.81657040e-01 2.33577684e-01 -7.26736188e-01 1.07505548e+00
-1.07652764e-03 3.23823243e-01 3.80996704e-01 9.62631464e-01
-7.08517492e-01 -1.64590728e+00 4.47500408e-01 9.81022954e-01
5.50227873e-02 -2.30778381e-01 -3.01176518e-01 6.45154774e-01
2.57217973e-01 1.12625599e+00 -1.01749420e-01 -7.33436227e-01
9.01951611e-01 -8.53670090e-02 3.10079098e-01 -1.02209187e+00
-9.61704493e-01 -2.24367648e-01 1.05726528e+00 -8.57085586e-01
-7.64645457e-01 -4.09543991e-01 -1.86976099e+00 -3.48144203e-01
-9.44189608e-01 7.01136768e-01 7.05558002e-01 7.28285611e-01
7.77554035e-01 4.56178427e-01 2.95099795e-01 3.46770231e-03
2.49922410e-01 -4.38078970e-01 -8.74078274e-01 3.70449513e-01
-6.70421362e-01 -1.03799736e+00 3.03438365e-01 -1.86527967e-01] | [9.99104118347168, 5.84326171875] |
4878d44b-fffa-420d-aef1-31edec877ee6 | doodlenet-double-deeplab-enhanced-feature | 2204.10266 | null | https://arxiv.org/abs/2204.10266v1 | https://arxiv.org/pdf/2204.10266v1.pdf | DooDLeNet: Double DeepLab Enhanced Feature Fusion for Thermal-color Semantic Segmentation | In this paper we present a new approach for feature fusion between RGB and LWIR Thermal images for the task of semantic segmentation for driving perception. We propose DooDLeNet, a double DeepLab architecture with specialized encoder-decoders for thermal and color modalities and a shared decoder for final segmentation. We combine two strategies for feature fusion: confidence weighting and correlation weighting. We report state-of-the-art mean IoU results on the MF dataset. | ['Catherine Wacongne', 'Lucien Martin-Gaffé', 'Oriel Frigo'] | 2022-04-21 | null | null | null | null | ['thermal-image-segmentation'] | ['computer-vision'] | [ 6.54748827e-02 -4.57913309e-01 2.85701137e-02 -1.05026507e+00
-1.00626707e+00 -3.49381655e-01 3.79813641e-01 -3.35606605e-01
-8.10727179e-01 2.24374175e-01 -3.81450087e-01 -3.43767971e-01
4.06213582e-01 -5.47367156e-01 -4.95270073e-01 -7.24491715e-01
4.07343626e-01 3.11583746e-02 7.10855246e-01 -1.66652411e-01
1.77111879e-01 2.17431381e-01 -2.04999185e+00 5.86079419e-01
6.42726362e-01 2.03207469e+00 1.23001046e-01 1.45452261e+00
2.14054063e-01 7.15759635e-01 -1.66437164e-01 -3.25266123e-01
5.39264798e-01 -2.29285374e-01 -7.77715385e-01 -2.35063881e-01
6.72018170e-01 -3.71995032e-01 -4.75967079e-01 7.88539410e-01
8.60279679e-01 4.73269969e-01 4.46293503e-01 -1.39924502e+00
-1.73411071e-01 1.99820042e-01 -6.21168435e-01 4.47647274e-01
1.61172777e-01 3.94302458e-01 8.56074989e-01 -6.43220544e-01
1.84575111e-01 1.04958963e+00 5.43469727e-01 5.13219833e-01
-7.27492809e-01 -4.76972014e-01 -2.36297980e-01 7.99052060e-01
-1.34976709e+00 -4.69274282e-01 4.38411921e-01 -2.64827907e-01
1.48370874e+00 4.64777380e-01 6.38013482e-01 9.73021805e-01
6.69019163e-01 9.57003713e-01 1.73070467e+00 -1.21976145e-01
3.53955090e-01 -1.11468043e-02 5.36909163e-01 8.05285215e-01
-9.95007008e-02 5.57125032e-01 -1.05815673e+00 3.48850280e-01
-6.49356544e-02 -7.22240508e-01 2.40758300e-01 3.20162736e-02
-8.24561715e-01 7.28206515e-01 6.16162300e-01 -1.31412268e-01
4.02949899e-02 6.78561568e-01 4.24066037e-01 9.79519933e-02
7.37455070e-01 -5.72883971e-02 -6.29240513e-01 -2.74517357e-01
-1.14849365e+00 -6.57683983e-02 4.63863164e-01 9.55763459e-01
1.05323005e+00 -2.54001200e-01 -2.78581887e-01 4.36786532e-01
7.95263648e-01 8.91378403e-01 3.19587499e-01 -1.35477722e+00
-1.51522849e-02 7.90552702e-03 -1.50548801e-01 -2.39394978e-01
-4.55390304e-01 1.52755290e-01 -3.19130421e-01 4.94196236e-01
-1.04760550e-01 -2.27259517e-01 -1.53684163e+00 9.16443288e-01
2.31812056e-02 3.19119453e-01 1.22583680e-01 1.09027922e+00
1.11571586e+00 3.58430028e-01 4.86359186e-02 4.85515356e-01
1.22155547e+00 -1.25157869e+00 -6.40665770e-01 -6.24326825e-01
2.48572052e-01 -7.53109276e-01 4.34691817e-01 4.32344735e-01
-7.32023180e-01 -8.58430624e-01 -1.37007535e+00 -4.98097569e-01
-6.77309811e-01 1.31399557e-01 7.77710617e-01 1.06931317e+00
-1.28603327e+00 6.17520571e-01 -1.21317959e+00 -2.19564378e-01
4.26762074e-01 2.18707174e-01 -7.90600106e-02 -8.87157768e-02
-1.11780488e+00 1.29953396e+00 6.02336109e-01 3.54581535e-01
-8.26951206e-01 -3.23052078e-01 -8.22543979e-01 -6.55407131e-01
8.08408260e-02 -6.96035266e-01 1.47413683e+00 -6.34166360e-01
-1.83704495e+00 1.21916115e+00 -3.42919618e-01 -5.61156750e-01
4.48137432e-01 -6.13275647e-01 -5.16126752e-01 8.12210143e-02
-1.32160690e-02 8.96944940e-01 1.01871431e+00 -1.13827002e+00
-9.63591039e-01 -3.50234896e-01 -4.14295971e-01 2.02303678e-01
3.74706089e-01 -3.95769961e-02 -6.47953868e-01 1.37861654e-01
7.55390450e-02 -8.21303725e-01 -4.85672235e-01 -2.13471487e-01
-3.81685883e-01 -2.27465574e-02 1.04865491e+00 -6.68608487e-01
7.05735981e-01 -2.22381878e+00 -3.45792621e-02 2.85350114e-01
5.81805073e-02 8.63977894e-02 7.34511912e-02 -4.35102165e-01
2.86832094e-01 -6.83427453e-02 -2.42776200e-01 -6.94909036e-01
1.75947681e-01 3.12740654e-01 3.74394879e-02 5.79255879e-01
1.85038894e-02 1.05931580e+00 -7.22434163e-01 -5.24689972e-01
8.86578441e-01 4.24383432e-01 -9.01211947e-02 3.10711920e-01
-1.01735637e-01 3.41066450e-01 -3.54100883e-01 6.68166161e-01
8.18699360e-01 4.40296143e-01 -4.86418277e-01 -5.86198688e-01
-3.90129596e-01 1.14257440e-01 -9.06597674e-01 2.18608451e+00
-5.08164287e-01 9.64907885e-01 2.77208298e-01 -2.58135319e-01
7.78668404e-01 -3.47097293e-02 3.32127005e-01 -9.41307425e-01
8.87964904e-01 3.21245223e-01 -3.46701086e-01 -2.97636062e-01
1.04693222e+00 -3.60360257e-02 -2.38976598e-01 3.82218957e-01
3.18194389e-01 -8.42454135e-01 7.03785494e-02 -1.08309008e-01
8.12369406e-01 6.02067292e-01 -2.95810252e-01 -7.63327554e-02
2.69362509e-01 1.00521736e-01 2.32559532e-01 6.21092558e-01
-7.32976794e-01 8.59282970e-01 -2.99474727e-02 -2.88672715e-01
-8.13956380e-01 -1.16408527e+00 -2.63817132e-01 1.37237620e+00
3.69492322e-01 -3.00150007e-01 -6.70048237e-01 -3.56921196e-01
-1.15423359e-01 9.30970371e-01 -7.25093544e-01 -3.61749798e-01
6.36181794e-03 -6.68546557e-01 7.24380016e-01 8.57408524e-01
1.11668885e+00 -5.91192126e-01 -1.49661982e+00 -1.76704034e-01
-1.38109028e-01 -1.44782746e+00 -7.21301958e-02 8.28125358e-01
-4.25055474e-01 -1.03012514e+00 -1.81069180e-01 -1.81218922e-01
8.37402232e-03 2.51399219e-01 1.08372319e+00 -3.38863313e-01
-7.42626488e-01 4.73101228e-01 -4.39476997e-01 -4.56907034e-01
-9.11139101e-02 -2.70770073e-01 -3.82950395e-01 -1.07008636e-01
6.68399334e-01 3.24114799e-01 -8.46542120e-01 2.74490476e-01
-6.81014657e-01 1.09902240e-01 3.05343300e-01 3.66628826e-01
6.15846872e-01 -1.57677233e-01 -5.56181729e-01 -2.24136025e-01
2.91564167e-01 1.24265049e-02 -5.56068778e-01 2.49787197e-01
-4.32122350e-01 2.17109069e-01 -2.44517699e-01 6.04521811e-01
-1.23597491e+00 3.97139847e-01 -4.11755949e-01 -4.81345892e-01
-3.78033012e-01 -4.04492998e-03 2.22444057e-01 -4.30693388e-01
6.08446836e-01 -1.25927448e-01 -4.17459942e-02 -2.03693509e-01
9.52423096e-01 9.27969992e-01 7.50794768e-01 -3.30483496e-01
2.51769453e-01 5.19264340e-01 6.82171136e-02 -5.91588795e-01
-8.62647831e-01 -8.02863657e-01 -1.04212224e+00 -5.35713911e-01
1.57352829e+00 -1.08242714e+00 -5.39133728e-01 9.55550492e-01
-8.30941975e-01 -4.79229599e-01 -2.73289382e-01 4.60703582e-01
-8.85186136e-01 4.48823236e-02 -4.93350953e-01 -5.70317149e-01
-4.42856371e-01 -1.41768682e+00 1.22078156e+00 4.89389688e-01
2.54684448e-01 -6.84101641e-01 2.65420109e-01 6.92839921e-01
6.72862411e-01 2.85887003e-01 -3.30574736e-02 -4.55270670e-02
-3.61976385e-01 -2.50407428e-01 -5.24721265e-01 7.92871118e-01
-3.45972151e-01 3.66945654e-01 -1.88039303e+00 3.49582024e-02
-1.75073788e-01 -7.52639771e-01 1.70486784e+00 5.67472160e-01
1.20139456e+00 8.12063098e-01 -3.14398378e-01 8.99774194e-01
1.31081486e+00 -8.50970596e-02 6.35109007e-01 2.75507927e-01
8.63471091e-01 3.49952281e-01 7.49234438e-01 4.79640990e-01
5.72423875e-01 3.35853755e-01 4.25448924e-01 -1.51792809e-01
-2.86786407e-01 4.72955465e-01 3.33709508e-01 3.15494716e-01
-2.25880638e-01 -1.44276440e-01 -1.13707316e+00 3.95355105e-01
-1.88249695e+00 -4.59902525e-01 -2.88332462e-01 1.74160755e+00
4.80012745e-01 2.03666557e-02 -4.61390764e-02 -1.57914013e-02
3.62290710e-01 3.19630146e-01 -5.89959025e-01 -1.02293754e+00
-1.12180881e-01 5.67121327e-01 1.28029549e+00 5.26115358e-01
-1.39545012e+00 1.12741995e+00 7.81459427e+00 8.78910661e-01
-1.06623900e+00 5.06169319e-01 9.31460559e-01 -2.26427227e-01
-3.07173878e-02 -2.21910357e-01 -6.13129199e-01 6.33202121e-02
1.55834126e+00 2.80959398e-01 3.85345548e-01 7.12934077e-01
-3.08687955e-01 -1.02736354e+00 -8.36400330e-01 1.01385713e+00
1.25696555e-01 -8.56477797e-01 -8.48138213e-01 -4.06993240e-01
5.86654186e-01 1.06190562e+00 3.17152619e-01 -9.59574878e-02
5.14583230e-01 -1.01423359e+00 8.91851366e-01 6.84968829e-01
1.01051533e+00 -8.79432678e-01 8.44208121e-01 -2.65762538e-01
-1.34489644e+00 3.40443283e-01 -3.10437441e-01 2.81021118e-01
3.00628040e-02 6.00754917e-01 -2.48054013e-01 8.59504461e-01
1.14082122e+00 7.40178585e-01 -9.91737783e-01 8.40932429e-01
-3.60843688e-01 4.14186746e-01 -5.35445273e-01 1.92337513e-01
4.18920100e-01 1.18314207e-01 1.32455677e-01 1.49216747e+00
2.12754592e-01 -7.72193447e-02 4.07352895e-02 6.08326733e-01
4.45628285e-01 -5.86149871e-01 -4.82473999e-01 2.43404537e-01
-9.18995962e-02 1.60024905e+00 -8.76104534e-01 -5.09003997e-01
-4.48199332e-01 1.27285898e+00 -1.72556594e-01 1.29964381e-01
-1.05482101e+00 -4.31645721e-01 9.87570822e-01 -7.71505237e-01
5.38659930e-01 -5.60079575e-01 -7.45632529e-01 -9.09306288e-01
-4.67877597e-01 5.71058656e-04 1.93557739e-01 -1.07051802e+00
-1.09006798e+00 8.71130645e-01 3.62310894e-02 -8.85930538e-01
-6.10328577e-02 -1.06907582e+00 -2.38602057e-01 9.41193938e-01
-1.88240182e+00 -9.81835425e-01 -6.62629843e-01 7.76821017e-01
1.95499077e-01 1.68676287e-01 7.19422698e-01 -2.09412817e-02
-5.38581192e-01 2.42777795e-01 6.82756007e-02 -1.93691999e-01
8.25025320e-01 -1.37871647e+00 6.61157846e-01 1.00410628e+00
-2.35112473e-01 -2.05515012e-01 6.65033877e-01 -3.25105518e-01
-1.66412842e+00 -1.18077743e+00 2.46599793e-01 -6.30707204e-01
5.59486687e-01 -3.19929570e-01 -2.26717442e-01 4.87759173e-01
9.40896809e-01 3.71762753e-01 8.91093612e-01 -9.93378535e-02
-3.98010343e-01 -1.56040072e-01 -1.34570193e+00 -1.59702804e-02
6.32506847e-01 -7.56827056e-01 -4.02320117e-01 6.09159768e-02
6.27512872e-01 -8.19344640e-01 -6.92630172e-01 3.53815228e-01
6.51498795e-01 -1.38316917e+00 6.91297472e-01 1.68235496e-01
2.10651219e-01 -5.08223236e-01 -5.58019400e-01 -1.37527633e+00
-8.15690234e-02 -2.57827997e-01 1.04016632e-01 4.31412697e-01
3.97969723e-01 -4.48356628e-01 6.14634097e-01 9.37110066e-01
-6.20084226e-01 -3.91479582e-01 -1.48641634e+00 -5.38445234e-01
-2.45577455e-01 -1.37904501e+00 1.03501067e-01 1.07170548e-02
-2.24790722e-01 1.49053261e-01 -1.24058187e-01 -1.77148152e-02
7.65994012e-01 1.74062792e-02 1.89987525e-01 -5.69180548e-01
7.97505025e-04 -4.99438614e-01 -5.71626902e-01 -5.97679496e-01
1.14864081e-01 -6.24329686e-01 8.46602321e-01 -1.66641879e+00
3.28449383e-02 -1.32266104e-01 -6.11973524e-01 5.12258887e-01
-7.21411854e-02 8.52676213e-01 1.37638181e-01 -3.95220220e-01
-1.10474193e+00 4.48291987e-01 9.11997318e-01 -2.49766976e-01
1.90432236e-01 -4.27182645e-01 -1.91762283e-01 4.64203715e-01
8.85148585e-01 -3.21605313e-03 -6.79911673e-02 -6.25486672e-01
1.78526044e-01 -3.62721354e-01 4.62086171e-01 -1.56079566e+00
2.65932649e-01 4.63087447e-02 4.93612885e-01 -1.09199083e+00
5.81152618e-01 -8.06978464e-01 -3.05732697e-01 3.32077682e-01
-8.67784470e-02 4.14190516e-02 6.29528642e-01 4.47418749e-01
-1.92038819e-01 2.24182710e-01 8.94852102e-01 -7.66182095e-02
-1.58242202e+00 1.25288099e-01 -8.07577491e-01 -1.69717610e-01
1.18546021e+00 -2.39956483e-01 -5.25057316e-01 1.12821989e-01
-8.16841066e-01 4.94744837e-01 3.12081516e-01 6.43737555e-01
9.37668204e-01 -1.11784923e+00 -5.63915908e-01 3.45419496e-01
2.58435816e-01 -2.15572461e-01 2.50761330e-01 1.03377342e+00
-6.38228059e-01 5.11782408e-01 -4.97282118e-01 -8.36192489e-01
-1.25097060e+00 1.26348153e-01 5.93971074e-01 3.25314999e-01
-1.77275538e-01 1.41476369e+00 -4.61671561e-01 -1.66133702e-01
-2.53705233e-02 -6.54494643e-01 1.51158512e-01 7.12620243e-02
3.97403061e-01 5.93460441e-01 5.67657888e-01 -8.56374502e-01
-8.33978057e-01 7.05641687e-01 2.26647258e-01 -5.16985416e-01
8.03490698e-01 -4.37044054e-01 -2.05994979e-01 9.11365390e-01
1.23885632e+00 -8.77945125e-01 -1.53471255e+00 5.66186709e-03
-3.62697244e-01 -4.28199947e-01 1.06045389e+00 -1.21764183e+00
-1.29969573e+00 1.04156005e+00 1.36770833e+00 -1.43559605e-01
1.40064669e+00 -5.75687103e-02 8.95826757e-01 1.73426121e-01
4.30281907e-01 -1.64668369e+00 -2.19721764e-01 8.43195260e-01
3.24027807e-01 -1.58947480e+00 1.07566476e-01 -1.12919509e-01
-1.00080228e+00 1.07667613e+00 6.03939533e-01 -1.45525917e-01
7.94594824e-01 4.37452793e-01 4.37083185e-01 -2.62525737e-01
-8.69559824e-01 -9.70903933e-01 6.87031388e-01 4.92194772e-01
5.95271289e-01 4.71947789e-01 8.55181832e-03 2.09504321e-01
-1.71515256e-01 4.16362174e-02 1.80739328e-01 1.11469042e+00
-6.70789182e-01 -8.21140587e-01 -2.77396828e-01 4.82547939e-01
-2.24500477e-01 -8.71571600e-02 -3.81077826e-01 1.30167052e-01
5.12324750e-01 1.40367830e+00 2.39272147e-01 -1.21584439e+00
3.85057144e-02 2.33882621e-01 8.79236221e-01 -3.00820589e-01
-6.40335500e-01 8.66562352e-02 3.46020639e-01 -1.33195400e+00
-5.95517874e-01 -6.02989852e-01 -1.22895777e+00 -2.53096193e-01
-4.22318131e-01 -2.76750118e-01 1.34127545e+00 9.59061265e-01
3.02822351e-01 9.55566585e-01 6.60159886e-01 -1.10498583e+00
1.05071202e-01 -8.71970296e-01 -6.43483996e-01 -3.03695202e-01
4.11552966e-01 -5.10735691e-01 -2.43031532e-01 -1.89707249e-01] | [9.12325668334961, -1.5716617107391357] |
6f2baca9-a730-4c08-a071-b1a789818e78 | dive-into-the-power-of-neuronal-heterogeneity | 2305.11484 | null | https://arxiv.org/abs/2305.11484v1 | https://arxiv.org/pdf/2305.11484v1.pdf | Dive into the Power of Neuronal Heterogeneity | The biological neural network is a vast and diverse structure with high neural heterogeneity. Conventional Artificial Neural Networks (ANNs) primarily focus on modifying the weights of connections through training while modeling neurons as highly homogenized entities and lacking exploration of neural heterogeneity. Only a few studies have addressed neural heterogeneity by optimizing neuronal properties and connection weights to ensure network performance. However, this strategy impact the specific contribution of neuronal heterogeneity. In this paper, we first demonstrate the challenges faced by backpropagation-based methods in optimizing Spiking Neural Networks (SNNs) and achieve more robust optimization of heterogeneous neurons in random networks using an Evolutionary Strategy (ES). Experiments on tasks such as working memory, continuous control, and image recognition show that neuronal heterogeneity can improve performance, particularly in long sequence tasks. Moreover, we find that membrane time constants play a crucial role in neural heterogeneity, and their distribution is similar to that observed in biological experiments. Therefore, we believe that the neglected neuronal heterogeneity plays an essential role, providing new approaches for exploring neural heterogeneity in biology and new ways for designing more biologically plausible neural networks. | ['Yi Zeng', 'Yang Li', 'Yiting Dong', 'Dongcheng Zhao', 'Guobin Shen'] | 2023-05-19 | null | null | null | null | ['continuous-control'] | ['playing-games'] | [ 3.13651592e-01 -1.00942291e-01 1.38344929e-01 1.09699890e-01
2.37363175e-01 -2.47295067e-01 4.54581618e-01 -2.93058425e-01
-7.22942412e-01 1.10597563e+00 -2.47547776e-01 6.74803555e-02
-2.76649922e-01 -6.19760215e-01 -9.64581847e-01 -1.38483036e+00
-2.25486327e-02 3.28445226e-01 4.76904213e-01 -3.34280938e-01
4.65053469e-01 7.25398719e-01 -1.80790818e+00 7.19010606e-02
7.76575863e-01 8.20038140e-01 3.86565238e-01 5.54380894e-01
9.18990895e-02 6.01846814e-01 -6.31643772e-01 -1.05051389e-02
2.50098377e-01 -6.70536697e-01 -2.97102422e-01 -3.99879009e-01
-6.59454018e-02 5.76432467e-01 -3.06468338e-01 1.12906718e+00
8.75093699e-01 1.06074885e-01 1.03527141e+00 -1.01692867e+00
-5.92489064e-01 8.53253543e-01 -1.92313015e-01 4.41509843e-01
-7.93524623e-01 3.77111614e-01 5.51572680e-01 -3.23685199e-01
5.39334238e-01 9.71647322e-01 8.89209151e-01 1.05962348e+00
-1.56433582e+00 -6.96590960e-01 1.87758580e-01 1.00319140e-01
-1.55238020e+00 -5.13527095e-01 5.92932463e-01 -4.81797963e-01
1.16339982e+00 1.22258492e-01 9.84867692e-01 1.10725605e+00
7.55790770e-01 1.91803619e-01 1.01649666e+00 -2.75709003e-01
6.44505501e-01 -1.53795257e-01 -3.48348655e-02 3.87810856e-01
6.77835226e-01 1.07903056e-01 -4.82940346e-01 6.71788491e-03
9.17464435e-01 -2.24372715e-01 -3.92796576e-01 6.77031931e-03
-1.37282574e+00 5.65295517e-01 1.97385401e-01 5.72387993e-01
-3.76895279e-01 5.23619890e-01 2.62356281e-01 2.10051499e-02
-6.71868622e-02 9.79789972e-01 -4.24888939e-01 1.57630771e-01
-7.42646337e-01 2.86724329e-01 7.14128852e-01 3.26258808e-01
5.78503430e-01 4.30943161e-01 6.25641569e-02 1.07941735e+00
-2.13779137e-02 3.74426425e-01 8.13313603e-01 -1.42060292e+00
-7.03309402e-02 4.56525594e-01 -3.27856451e-01 -8.17445219e-01
-7.14807987e-01 -6.16153657e-01 -1.38772631e+00 6.50120020e-01
5.22819221e-01 -2.36789480e-01 -7.57144213e-01 2.01754856e+00
-4.79556434e-02 1.30053952e-01 1.42187655e-01 8.23261321e-01
5.48824131e-01 4.89372939e-01 2.03660969e-02 -2.57630825e-01
9.64105368e-01 -7.15868354e-01 -5.50386250e-01 -1.40085429e-01
3.48775685e-01 -2.01986760e-01 6.31089032e-01 3.32811266e-01
-1.37163913e+00 -1.94226965e-01 -1.14636624e+00 4.15259153e-01
-4.70307738e-01 -3.27692717e-01 5.05977750e-01 7.23187268e-01
-1.11715412e+00 9.63347554e-01 -8.59387219e-01 -3.66844416e-01
4.41418469e-01 7.48425364e-01 -2.42300145e-02 4.92245704e-01
-1.11543655e+00 1.17367268e+00 4.46770400e-01 2.37307191e-01
-4.85367179e-01 -7.56285489e-01 -2.92134136e-01 1.07877418e-01
-2.48702273e-01 -1.15705955e+00 6.25396013e-01 -8.93637240e-01
-1.83174968e+00 6.61804318e-01 -1.10613659e-01 -5.95342398e-01
1.99855894e-01 8.49182963e-01 2.90277507e-02 -1.38166174e-01
-5.41534424e-01 1.12131643e+00 3.48021477e-01 -1.40743482e+00
-5.95430210e-02 -4.04776961e-01 -4.72223043e-01 -9.28100124e-02
-3.08889985e-01 -1.94227383e-01 1.38893560e-01 -7.11662948e-01
3.28658134e-01 -1.15303993e+00 -3.90587896e-01 8.97997022e-02
-1.67094380e-01 1.31999508e-01 2.11367518e-01 -1.45258754e-01
8.86468351e-01 -1.81385553e+00 6.17881715e-01 2.93441147e-01
4.13956404e-01 1.97754905e-01 -2.63721585e-01 1.96223065e-01
9.29374695e-02 3.88891608e-01 -4.18803662e-01 6.16878606e-02
-2.15513676e-01 3.09083015e-01 1.11347146e-01 4.04056638e-01
3.61003429e-01 1.04424584e+00 -3.48087698e-01 -2.41958126e-01
-2.04094470e-01 7.71994114e-01 -5.58661401e-01 -2.06239834e-01
-2.04584762e-01 5.52597225e-01 -1.20133296e-01 5.39082110e-01
4.64088976e-01 -2.92773873e-01 1.23006791e-01 -1.79507196e-01
-3.62348258e-01 -2.00733498e-01 -8.81355226e-01 1.05056584e+00
-3.22520912e-01 8.73482287e-01 1.37716725e-01 -1.37424242e+00
9.20260966e-01 1.07877158e-01 4.51105505e-01 -1.06086326e+00
4.49005932e-01 3.94732386e-01 7.76573777e-01 -4.54521507e-01
-6.85182065e-02 -1.74780056e-01 5.43871284e-01 3.79154295e-01
8.22749436e-02 -2.37290800e-01 3.01728398e-01 -3.81841838e-01
1.06233966e+00 -2.28149012e-01 -1.44305408e-01 -7.97347665e-01
1.34388685e-01 -1.65597200e-01 6.67326272e-01 8.75250936e-01
-2.00770393e-01 8.86198401e-01 6.01445258e-01 -2.13935599e-01
-1.34814143e+00 -7.69537330e-01 -4.01012391e-01 6.38277829e-01
1.51167169e-01 1.73485726e-01 -9.21345294e-01 5.72614729e-01
-2.14009434e-01 3.42482775e-01 -9.05924380e-01 -5.14085591e-01
-5.64754963e-01 -1.45695889e+00 8.39403570e-01 2.15855092e-01
3.83618414e-01 -1.19687283e+00 -8.00588608e-01 4.65317756e-01
6.23968802e-02 -9.03412342e-01 9.30223167e-02 7.90293694e-01
-1.13495028e+00 -6.62542701e-01 -9.36906815e-01 -7.66531169e-01
5.81879735e-01 -1.56698436e-01 8.60624015e-01 3.00095499e-01
-6.38497174e-01 -2.18536839e-01 1.86755121e-01 -4.55688894e-01
-3.60732466e-01 3.45883667e-01 1.49142191e-01 -3.95770878e-01
-8.18905681e-02 -1.08088148e+00 -6.55203760e-01 4.92556065e-01
-9.41785336e-01 1.35281220e-01 5.84875822e-01 9.48383868e-01
6.10581338e-01 1.78690150e-01 7.23118603e-01 -6.57538295e-01
6.84946597e-01 -4.14352208e-01 -4.31383431e-01 2.60155678e-01
-7.03139186e-01 4.56910133e-01 6.14561737e-01 -9.22839344e-01
-6.83457911e-01 -2.57247567e-01 -9.89190713e-02 -3.78842279e-02
2.16404982e-02 2.96864957e-01 7.73601606e-02 -5.55128455e-01
9.39787269e-01 3.86550277e-01 3.76065582e-01 1.21501766e-01
-3.84154201e-01 2.85609774e-02 1.61832690e-01 -7.19329953e-01
1.24504425e-01 4.76789176e-01 3.59128922e-01 -9.43280518e-01
-1.00154795e-01 1.32244661e-01 -3.83739799e-01 -3.16777438e-01
6.03889585e-01 -4.42081392e-01 -1.01688766e+00 8.57467711e-01
-1.10924947e+00 -8.12908351e-01 -1.96501687e-01 4.90234762e-01
-7.03921199e-01 -1.58373982e-01 -7.15122223e-01 -7.76344419e-01
-2.11008891e-01 -1.27016664e+00 5.80424845e-01 5.02071381e-01
-1.77276865e-01 -1.08226168e+00 3.24357271e-01 3.98528799e-02
9.58542228e-01 7.20616207e-02 1.22368145e+00 -2.53870219e-01
-4.93237019e-01 4.08841789e-01 -1.45802140e-01 -5.15345447e-02
-1.89455464e-01 5.41686535e-01 -1.01243901e+00 1.11957356e-01
-1.19860545e-01 -1.99678496e-01 1.17486584e+00 9.26697195e-01
1.14710605e+00 -1.99982256e-01 -4.24744785e-01 7.57101536e-01
1.48219430e+00 3.34560305e-01 9.64461863e-01 6.29145443e-01
3.53211641e-01 9.82614338e-01 -5.82485378e-01 2.93951243e-01
-1.80484802e-02 3.74335855e-01 5.16599417e-01 9.94008034e-02
-2.46728167e-01 4.40490305e-01 7.84938261e-02 1.00804651e+00
-5.52563071e-01 -2.49905720e-01 -9.08419073e-01 2.35406563e-01
-1.99149394e+00 -1.15219474e+00 -1.76465549e-02 1.97605515e+00
9.95003521e-01 1.02798194e-01 -1.48311304e-02 1.12886578e-01
9.69785213e-01 -3.23470205e-01 -1.03278470e+00 -2.56112397e-01
-8.81859779e-01 8.80412683e-02 6.50857627e-01 3.02106351e-01
-2.31027827e-01 6.58970535e-01 7.60013914e+00 6.34619236e-01
-1.55137050e+00 -1.67965010e-01 7.58699000e-01 -5.59270322e-01
-4.49294835e-01 -4.19121742e-01 -9.12229836e-01 7.43450940e-01
1.16199100e+00 -8.83210450e-02 6.90317631e-01 4.73786630e-02
3.13485086e-01 -1.27988368e-01 -5.26592195e-01 9.24209297e-01
-2.27878973e-01 -1.76230764e+00 -4.30626906e-02 1.45431414e-01
9.70826626e-01 3.91927958e-01 1.78005889e-01 6.24036565e-02
4.66136672e-02 -1.20160258e+00 7.20190883e-01 9.00643349e-01
2.18201354e-01 -5.51720083e-01 6.45202518e-01 2.99487919e-01
-7.32574821e-01 5.16170934e-02 -9.74071920e-01 -5.88836633e-02
-8.73426646e-02 8.74426544e-01 -1.44177288e-01 -4.78814632e-01
8.60934496e-01 3.97206515e-01 -5.15340328e-01 1.34844398e+00
5.64389586e-01 4.35182512e-01 -5.58525026e-01 -5.93563974e-01
-1.09291874e-01 -3.39812338e-01 3.36426288e-01 1.06343400e+00
3.78622890e-01 -8.63303095e-02 -7.17799008e-01 1.38283491e+00
-1.19552806e-01 -8.16285759e-02 -5.81977606e-01 -3.09602380e-01
4.94722992e-01 1.10807920e+00 -1.26635802e+00 5.69124185e-02
4.40119728e-02 5.06925642e-01 4.50132787e-01 5.81352472e-01
-7.71680832e-01 -3.49837393e-01 6.48598909e-01 -2.72601601e-02
3.37053925e-01 -3.36073160e-01 -7.94608474e-01 -1.02005482e+00
-2.40847960e-01 -5.51192582e-01 -3.74915957e-01 -6.94813371e-01
-1.11352062e+00 6.20660901e-01 -3.60405326e-01 -6.45499110e-01
6.09467737e-02 -7.70242393e-01 -6.97499394e-01 6.87143564e-01
-1.31144536e+00 -4.58223641e-01 -1.68086007e-01 3.93916935e-01
2.14104369e-01 -2.21621349e-01 6.39024198e-01 1.46428928e-01
-9.11910415e-01 3.55149865e-01 4.90944624e-01 -4.98308122e-01
3.54745209e-01 -8.00348461e-01 1.66130930e-01 3.90018761e-01
-2.43006513e-01 6.83253229e-01 8.99710417e-01 -4.20754582e-01
-1.20865643e+00 -8.58344018e-01 5.74079514e-01 2.44709570e-02
6.62448347e-01 -4.21835393e-01 -1.13810098e+00 -4.80906479e-02
2.57511437e-01 -8.06305259e-02 6.51857078e-01 -1.93942115e-01
-1.84612826e-01 1.06789224e-01 -1.03720260e+00 1.08464372e+00
1.23836601e+00 -1.44584507e-01 -1.63083032e-01 8.66120309e-02
3.73265564e-01 1.89898118e-01 -8.07604313e-01 5.89057028e-01
9.78050888e-01 -1.01500392e+00 7.72704482e-01 -5.21437049e-01
2.99205869e-01 -1.22368000e-01 -4.19970229e-02 -1.47145879e+00
-4.32802975e-01 -3.33354354e-01 7.99218193e-02 8.91193628e-01
7.50051618e-01 -1.09388149e+00 9.93460417e-01 6.87429428e-01
-4.13171276e-02 -9.10703838e-01 -9.34533536e-01 -8.89598310e-01
5.16485333e-01 1.34598613e-01 2.02410892e-01 7.27266192e-01
-2.40545586e-01 -3.81248258e-02 1.66468754e-01 -1.31430596e-01
6.63192987e-01 -3.65968972e-01 7.88428262e-02 -1.35314739e+00
-1.26956001e-01 -1.25559425e+00 -5.22113562e-01 -3.84649575e-01
4.91933167e-01 -6.73813641e-01 2.68562704e-01 -1.23277247e+00
3.38247955e-01 -5.26400983e-01 -3.52361411e-01 9.56899226e-02
1.53434463e-02 4.37278628e-01 -1.39327645e-01 4.11617398e-01
-3.02766740e-01 5.06901085e-01 1.24634528e+00 -1.52440593e-01
-5.16765229e-02 -2.17217803e-01 -5.40220737e-01 5.27787745e-01
1.23309672e+00 -6.27368450e-01 -2.52303839e-01 -4.71093774e-01
4.86441463e-01 -3.95874441e-01 2.16211021e-01 -1.37669861e+00
6.67001128e-01 -3.77681345e-01 3.72442275e-01 4.73103486e-02
3.13302338e-01 -6.05574012e-01 4.50314552e-01 8.03800106e-01
-5.94719112e-01 1.34843439e-01 2.84318388e-01 2.87414491e-01
-9.28427354e-02 -5.60128272e-01 1.13203788e+00 -2.82156587e-01
-2.45296314e-01 2.38193989e-01 -1.21095741e+00 2.06663638e-01
7.86562443e-01 -5.96687615e-01 -6.44817293e-01 1.05158255e-01
-7.06354499e-01 -1.13309480e-01 6.17945313e-01 -2.85716057e-02
3.09061229e-01 -1.04992545e+00 -4.22406852e-01 1.09325163e-01
-2.59244204e-01 -1.72433525e-01 3.06883723e-01 9.61736858e-01
-6.92436278e-01 3.33115608e-01 -7.75325775e-01 -6.28961325e-01
-7.83779979e-01 2.37550959e-01 9.23979223e-01 1.61106199e-01
-1.58401027e-01 9.85783756e-01 1.54299855e-01 -3.28022301e-01
2.25829408e-01 -3.47977102e-01 -2.78855264e-01 -1.02091923e-01
2.95518458e-01 3.77207875e-01 1.51251152e-01 -2.68483400e-01
-1.56366542e-01 8.90638351e-01 1.98289946e-01 -8.43095481e-02
1.41898596e+00 -2.26691552e-03 -3.82794380e-01 7.65657783e-01
9.71142590e-01 -5.08464992e-01 -1.39388144e+00 1.65941536e-01
-1.10999919e-01 2.94019759e-01 2.76828054e-02 -5.46976149e-01
-1.44059348e+00 9.26227212e-01 5.04305601e-01 2.49593481e-01
7.72820890e-01 -5.41696027e-02 4.60104793e-01 7.62124777e-01
3.74724776e-01 -1.11792254e+00 9.18921530e-02 6.79650545e-01
8.15201581e-01 -9.26946461e-01 -2.91838259e-01 -4.50964011e-02
-3.20902824e-01 1.31838739e+00 8.41348171e-01 -3.39419842e-01
8.25667858e-01 7.45322585e-01 -6.53972328e-02 -1.45970434e-02
-1.18843913e+00 -2.55100429e-02 -7.32055828e-02 7.26545751e-01
4.70262825e-01 -2.87299961e-01 -2.91472465e-01 3.47755373e-01
-8.79777968e-02 -3.06398124e-02 5.68743229e-01 7.91311145e-01
-6.91000938e-01 -8.83048832e-01 -1.15357719e-01 4.94477242e-01
-3.78463894e-01 -1.82468936e-01 -2.99253464e-01 5.80437601e-01
2.69921236e-02 4.18394864e-01 1.43991262e-01 -3.11294675e-01
8.24927688e-02 1.01175956e-01 9.55628335e-01 -3.55380744e-01
-7.71312773e-01 -2.81731755e-01 -3.80673081e-01 -1.56680658e-01
-4.82148200e-01 -4.41964328e-01 -1.42618811e+00 -4.88293886e-01
-5.02144635e-01 4.73064184e-02 1.15426540e+00 9.26560879e-01
4.74575669e-01 8.42861772e-01 1.44872770e-01 -1.22422326e+00
-1.61966860e-01 -6.95421100e-01 -7.09323287e-01 -4.54852078e-03
2.06040852e-02 -8.65928054e-01 -8.05075288e-01 9.04354304e-02] | [8.055468559265137, 2.8377957344055176] |
3cbe6eeb-6507-43bd-a1ea-f81bf87e243f | knowledge-integration-networks-for-action | 2002.07471 | null | https://arxiv.org/abs/2002.07471v1 | https://arxiv.org/pdf/2002.07471v1.pdf | Knowledge Integration Networks for Action Recognition | In this work, we propose Knowledge Integration Networks (referred as KINet) for video action recognition. KINet is capable of aggregating meaningful context features which are of great importance to identifying an action, such as human information and scene context. We design a three-branch architecture consisting of a main branch for action recognition, and two auxiliary branches for human parsing and scene recognition which allow the model to encode the knowledge of human and scene for action recognition. We explore two pre-trained models as teacher networks to distill the knowledge of human and scene for training the auxiliary tasks of KINet. Furthermore, we propose a two-level knowledge encoding mechanism which contains a Cross Branch Integration (CBI) module for encoding the auxiliary knowledge into medium-level convolutional features, and an Action Knowledge Graph (AKG) for effectively fusing high-level context information. This results in an end-to-end trainable framework where the three tasks can be trained collaboratively, allowing the model to compute strong context knowledge efficiently. The proposed KINet achieves the state-of-the-art performance on a large-scale action recognition benchmark Kinetics-400, with a top-1 accuracy of 77.8%. We further demonstrate that our KINet has strong capability by transferring the Kinetics-trained model to UCF-101, where it obtains 97.8% top-1 accuracy. | ['Li-Min Wang', 'Matthew R. Scott', 'Shiwen Zhang', 'Weilin Huang', 'Sheng Guo'] | 2020-02-18 | null | null | null | null | ['scene-recognition', 'human-parsing'] | ['computer-vision', 'computer-vision'] | [ 3.29819560e-01 -2.32295945e-01 -4.44611460e-01 -2.95847237e-01
-7.13953733e-01 -3.00596029e-01 4.65495139e-01 -1.14813410e-01
-6.22126877e-01 4.54257071e-01 3.59523386e-01 -1.78068146e-01
1.41438156e-01 -6.93668842e-01 -8.16086769e-01 -6.60876274e-01
-7.17544407e-02 1.34021863e-01 6.59676135e-01 1.62927315e-01
-4.90144528e-02 3.57692927e-01 -1.50783479e+00 8.95917118e-01
5.45514584e-01 1.29673576e+00 1.21210709e-01 9.92850304e-01
2.01146957e-02 1.76303959e+00 -6.14195704e-01 -2.39697024e-01
-1.37768477e-01 -3.91586483e-01 -1.21848440e+00 2.06315756e-01
3.58419627e-01 -6.67785645e-01 -6.73625231e-01 4.55750912e-01
3.80264521e-01 2.69199312e-01 4.40481395e-01 -1.18249965e+00
-5.54604471e-01 3.28074425e-01 -3.60388786e-01 4.74525660e-01
4.39143181e-01 6.04150891e-01 7.46001422e-01 -6.20521963e-01
5.19902170e-01 1.18762672e+00 3.30861807e-01 5.84253550e-01
-6.23028338e-01 -4.80444163e-01 3.95537496e-01 7.90500939e-01
-1.25080502e+00 -3.94577652e-01 4.87734288e-01 -4.40089077e-01
1.40023696e+00 5.24357297e-02 7.61988819e-01 1.25840843e+00
9.73769873e-02 1.43510437e+00 7.33048022e-01 -9.64471623e-02
1.44542843e-01 -5.14432847e-01 2.19497830e-01 7.73405552e-01
-1.44182459e-01 -2.21722722e-01 -7.31386423e-01 1.95647880e-01
1.06616843e+00 2.47532278e-02 -1.91096857e-01 -9.74893272e-02
-1.23404098e+00 3.57836068e-01 5.93281925e-01 1.36704087e-01
-3.97998065e-01 6.00708783e-01 6.61125541e-01 -1.10463239e-01
6.19329922e-02 2.46007144e-02 -5.88664234e-01 -5.51684678e-01
-2.84508228e-01 -3.48760858e-02 6.05254471e-01 9.83123422e-01
5.28268337e-01 -1.09811120e-01 -6.55197203e-01 6.13033950e-01
1.49199232e-01 3.57459724e-01 4.65288132e-01 -1.03303123e+00
5.61574221e-01 1.03268576e+00 -9.91085893e-04 -5.91143250e-01
-4.02603835e-01 -2.48651281e-01 -6.99658513e-01 -2.40345880e-01
3.13556492e-01 1.23069622e-01 -1.02917445e+00 1.78492439e+00
5.95633924e-01 9.42248762e-01 3.81075740e-01 8.86956990e-01
1.05004776e+00 6.21493101e-01 5.76746821e-01 -1.89482328e-02
1.56051624e+00 -1.45936382e+00 -5.19965410e-01 -2.38061473e-01
9.25666511e-01 -2.08474621e-01 9.78937268e-01 3.29786628e-01
-9.31947410e-01 -8.14687848e-01 -7.66482472e-01 -3.74439955e-01
-3.72113883e-01 3.96810859e-01 8.79237115e-01 3.54621187e-02
-8.66866410e-01 4.64420795e-01 -1.12526226e+00 -2.91669250e-01
8.51355612e-01 2.67494053e-01 -4.61339951e-01 -2.74128288e-01
-1.11106455e+00 6.11311376e-01 7.51161933e-01 8.96029025e-02
-1.35092998e+00 -4.85249937e-01 -9.73066628e-01 6.80044591e-02
7.36388206e-01 -7.99575210e-01 1.34911120e+00 -9.60701644e-01
-1.62057841e+00 5.84934711e-01 -4.17053923e-02 -2.95449048e-01
2.16110736e-01 -4.16588634e-01 -4.80933100e-01 5.76476514e-01
-4.44951467e-02 6.94115818e-01 5.94145715e-01 -6.13925159e-01
-1.03832579e+00 -2.74085760e-01 4.56817716e-01 2.72329181e-01
-2.90212989e-01 1.76606789e-01 -1.18674099e+00 -5.77326953e-01
-2.64018327e-01 -6.76732242e-01 -2.06465069e-02 -1.10864215e-01
-7.09350631e-02 -4.84665632e-01 8.49922657e-01 -6.09155238e-01
1.36196923e+00 -1.99733329e+00 2.66054839e-01 -3.40866655e-01
1.92492589e-01 6.44462883e-01 -3.99128675e-01 2.42039502e-01
-1.25788199e-03 -2.25344032e-01 -9.28465426e-02 -1.77976638e-01
-1.39573529e-01 5.37229776e-01 -2.58600563e-02 1.72789723e-01
6.76366031e-01 1.42867219e+00 -1.14917302e+00 -3.83877933e-01
2.28245005e-01 5.41278124e-01 -5.31451881e-01 5.04834652e-01
-4.14386779e-01 5.10506392e-01 -7.36939311e-01 7.15195000e-01
1.44086286e-01 -6.13012075e-01 3.86711240e-01 -2.24137515e-01
2.95512557e-01 1.74269080e-01 -1.02882898e+00 1.85681570e+00
-2.49258861e-01 2.41733760e-01 -2.93269485e-01 -9.37697947e-01
4.34013784e-01 2.21301451e-01 5.09021997e-01 -7.34912932e-01
1.73575327e-01 -2.62971878e-01 -1.88184068e-01 -6.49321258e-01
8.66358951e-02 4.92892981e-01 -1.98694244e-01 2.75487483e-01
3.67757440e-01 4.88523543e-01 2.85870552e-01 3.11324894e-01
1.63702166e+00 5.78418195e-01 2.49918431e-01 2.00863704e-01
8.11195910e-01 -1.55944243e-01 7.91586280e-01 4.64788228e-01
-4.09600884e-01 5.04030995e-02 6.40382230e-01 -6.47127926e-01
-4.83005911e-01 -8.71594071e-01 4.12641019e-01 1.29542565e+00
1.60938203e-01 -7.65346706e-01 -7.66407609e-01 -1.00285995e+00
-1.43803209e-01 3.31546664e-01 -7.61015415e-01 -4.79028553e-01
-5.69987535e-01 -2.82392502e-01 7.17839301e-01 1.25414312e+00
1.01122952e+00 -1.21635640e+00 -7.47212529e-01 2.19221637e-01
-3.71913522e-01 -1.52523994e+00 -4.33886081e-01 1.57871500e-01
-7.56867707e-01 -1.54939401e+00 -2.70259738e-01 -6.10678256e-01
4.53492343e-01 2.71425307e-01 1.11521494e+00 3.09482813e-01
-2.51420647e-01 6.81342065e-01 -6.12354159e-01 -1.13202855e-02
-8.74642059e-02 -1.98126107e-01 -1.55301213e-01 1.70969173e-01
5.12105048e-01 -4.99831706e-01 -6.74801886e-01 5.63922405e-01
-1.00830591e+00 3.89367640e-01 7.30397880e-01 7.93589413e-01
8.81066263e-01 -1.24681629e-02 4.11457837e-01 -5.18388331e-01
-1.38362776e-03 -4.12297398e-01 -3.05884510e-01 5.56178451e-01
-3.25901434e-03 -4.24622931e-02 6.11803532e-01 -5.17110646e-01
-1.05921495e+00 4.14877892e-01 -5.62984347e-02 -5.89233160e-01
-4.28695172e-01 5.77344537e-01 -5.05187273e-01 1.05030343e-01
4.35853839e-01 4.26384389e-01 -3.51918757e-01 -5.05842507e-01
5.62928140e-01 6.23472691e-01 9.75140274e-01 -7.95646906e-01
1.53959453e-01 2.93952346e-01 -7.34137967e-02 -5.77789068e-01
-1.19209290e+00 -7.07747459e-01 -9.73350883e-01 -2.74186999e-01
1.41786921e+00 -1.16798937e+00 -9.94730175e-01 1.04993033e+00
-9.81351495e-01 -8.84099543e-01 -2.81127810e-01 3.27580720e-01
-8.53258669e-01 4.06843305e-01 -8.41617703e-01 -4.94720787e-01
-1.67338088e-01 -1.24738002e+00 1.32485592e+00 3.69289458e-01
2.33052745e-01 -7.45186567e-01 -1.32033467e-01 8.80951822e-01
-1.27897402e-02 1.47785082e-01 6.29183888e-01 -5.72949290e-01
-7.52406359e-01 9.00582671e-02 -3.50593418e-01 4.53779757e-01
8.06686357e-02 -2.18193188e-01 -1.01011872e+00 -1.66002765e-01
-4.38811749e-01 -9.17009890e-01 1.10244870e+00 -4.94367676e-04
1.44289744e+00 -1.14796042e-01 -3.65502119e-01 6.48257554e-01
1.03731489e+00 3.44474584e-01 1.04130244e+00 1.14386171e-01
1.05512834e+00 2.31809262e-02 6.48653805e-01 4.83947426e-01
7.17134476e-01 8.43921542e-01 3.49171579e-01 4.34870161e-02
-2.86135703e-01 -2.96776265e-01 5.95858276e-01 7.46791959e-01
-2.32248098e-01 -1.83027834e-01 -9.28191781e-01 3.66024077e-01
-2.33908772e+00 -1.09615886e+00 -3.83174210e-03 1.77432978e+00
9.59958255e-01 1.06017496e-02 3.06610554e-01 -4.33306135e-02
3.71152431e-01 1.05450265e-01 -7.70050228e-01 -5.03800996e-02
2.61645198e-01 1.26531556e-01 2.08160758e-01 1.98064089e-01
-1.58723843e+00 1.38294220e+00 5.92576981e+00 1.07586801e+00
-8.79205585e-01 -1.06907319e-02 5.50316989e-01 1.77318770e-02
3.51585656e-01 -9.65413973e-02 -7.83713102e-01 4.00908202e-01
1.10525668e+00 3.04756165e-01 2.01758415e-01 9.89088774e-01
9.80151817e-02 -2.99628228e-01 -1.14655566e+00 1.01219940e+00
4.25251052e-02 -1.32896996e+00 2.28621736e-01 -2.35651061e-01
5.39323568e-01 -4.45392951e-02 -4.78082299e-01 7.71977127e-01
4.08344656e-01 -9.67707276e-01 4.54243809e-01 7.23148882e-01
8.46075475e-01 -6.50461137e-01 5.50346255e-01 3.97716969e-01
-1.72680020e+00 -3.21933568e-01 -1.81162655e-01 -1.62157074e-01
5.91083691e-02 5.68295941e-02 -4.93665874e-01 8.68452013e-01
6.38670385e-01 1.19318116e+00 -6.99483097e-01 8.38400960e-01
-6.50233448e-01 7.64728725e-01 -6.38969317e-02 2.22921848e-01
3.70847046e-01 1.78451031e-01 -1.11182563e-01 1.25875473e+00
-2.27050990e-01 6.47063553e-01 6.51393533e-01 4.51896638e-01
-1.09824464e-01 -1.47135094e-01 -1.88194394e-01 -2.89595097e-01
1.25966191e-01 1.18730152e+00 -7.12336838e-01 -7.97722399e-01
-6.71408951e-01 1.33814526e+00 7.17756152e-01 3.41298282e-01
-1.15095592e+00 -1.68315247e-01 8.12993288e-01 -4.24287468e-01
5.93159616e-01 -2.58875966e-01 3.40212166e-01 -1.34225953e+00
9.39160362e-02 -9.46889818e-01 7.97492862e-01 -9.59507763e-01
-9.71083105e-01 4.10309613e-01 4.58241813e-02 -9.24251914e-01
-6.21621264e-03 -9.28555787e-01 -6.27738297e-01 5.83071589e-01
-1.46547055e+00 -1.58357751e+00 -6.27870321e-01 9.84398901e-01
6.11613631e-01 -6.91907108e-02 6.55982077e-01 4.19671297e-01
-1.08693731e+00 5.26629210e-01 -3.77242297e-01 6.23029232e-01
4.55845326e-01 -1.19592547e+00 4.07820582e-01 8.93594921e-01
2.32615322e-01 3.72674733e-01 -2.65953034e-01 -7.77409911e-01
-1.65162516e+00 -1.50305355e+00 5.46148717e-01 -7.42414773e-01
6.96769953e-01 -1.84516937e-01 -8.53008330e-01 8.25355411e-01
-2.64036626e-01 4.39464629e-01 8.08992743e-01 -2.67150998e-02
-6.43256724e-01 -2.82447785e-02 -5.99905491e-01 4.50736612e-01
1.59583032e+00 -7.47249067e-01 -6.34702146e-01 3.71514529e-01
8.96240354e-01 -5.83126426e-01 -1.01314116e+00 4.03660148e-01
5.42595029e-01 -7.02789962e-01 9.66871679e-01 -1.21443903e+00
6.91576183e-01 -4.71156925e-01 -1.23111896e-01 -7.75685549e-01
-5.83481133e-01 -3.00017089e-01 -6.46769404e-01 1.06236339e+00
7.67996907e-03 -1.65570438e-01 7.59200335e-01 5.09019017e-01
-3.42568547e-01 -1.13790953e+00 -8.09509158e-01 -8.97810340e-01
-2.92169034e-01 -7.72597194e-01 4.56893891e-01 7.44164944e-01
4.10046428e-02 5.87745607e-01 -3.64448190e-01 1.29357949e-01
1.39562637e-01 5.41115627e-02 9.00621116e-01 -6.96350515e-01
-5.11697948e-01 -2.83130288e-01 -7.83854485e-01 -1.56403291e+00
2.35673338e-01 -8.70556176e-01 -2.74435394e-02 -1.64475572e+00
5.90017080e-01 -1.46602139e-01 -7.17176735e-01 1.20012450e+00
-4.58892643e-01 5.78468516e-02 1.46395221e-01 1.90278158e-01
-1.42981219e+00 6.24437273e-01 1.39262533e+00 -2.31792226e-01
-6.24091737e-02 -2.96179980e-01 -5.50661206e-01 6.96329772e-01
4.63759243e-01 -1.34872556e-01 -7.52156556e-01 -6.54886425e-01
-1.92901701e-01 1.08138792e-01 6.79136515e-01 -1.35846436e+00
3.01415861e-01 -4.18921292e-01 4.73333508e-01 -4.69473362e-01
3.38594526e-01 -7.94596672e-01 -3.83070260e-02 2.62201667e-01
-2.92157769e-01 -3.55311185e-01 2.83395231e-01 8.36450160e-01
-3.00702572e-01 2.77653396e-01 5.26914299e-01 -2.28583682e-02
-1.43864381e+00 6.58951879e-01 -1.40075892e-01 4.62927148e-02
1.25009692e+00 -9.88163650e-02 -5.74783504e-01 -2.26346731e-01
-7.10738838e-01 5.81753969e-01 2.06314102e-01 4.65994298e-01
5.57271302e-01 -1.43459260e+00 -4.38167691e-01 3.26239690e-02
3.92797202e-01 -6.52336003e-03 5.47756851e-01 1.01642227e+00
-3.57341141e-01 3.74357343e-01 -2.85951376e-01 -6.44360662e-01
-1.26363921e+00 6.66965365e-01 2.83063859e-01 -5.69399178e-01
-6.43946290e-01 9.38303292e-01 4.00551051e-01 -6.86498806e-02
4.64664936e-01 -5.43541133e-01 -2.69857287e-01 -1.89355403e-01
9.87036407e-01 2.93626010e-01 -1.41615480e-01 -7.33215809e-01
-6.62218451e-01 5.09254813e-01 3.94296832e-03 3.46571803e-01
1.17170727e+00 2.55985528e-01 5.99168129e-02 5.31196147e-02
1.09895682e+00 -6.26296401e-01 -1.67159057e+00 -3.10751230e-01
-5.44191115e-02 -3.01854849e-01 -7.05511123e-02 -1.19851303e+00
-1.15910721e+00 9.18460250e-01 3.84103507e-01 -4.85729963e-01
1.51868558e+00 9.52825695e-02 8.03685546e-01 7.03135729e-01
5.30730724e-01 -1.14415228e+00 5.38127542e-01 8.13346088e-01
7.34003961e-01 -1.11473894e+00 -9.24515277e-02 -4.65690583e-01
-9.30225074e-01 9.89509106e-01 9.90723133e-01 1.12103872e-01
2.99353421e-01 1.04686044e-01 -7.06483349e-02 -3.03352326e-01
-1.10354352e+00 -5.31604052e-01 4.33233500e-01 6.12200618e-01
1.55592754e-01 -3.36556584e-02 -4.68953103e-02 1.10076082e+00
6.15230143e-01 5.48140049e-01 8.51252824e-02 1.26086056e+00
-4.02122825e-01 -1.00315452e+00 1.67088479e-01 2.94326693e-01
-2.58343488e-01 2.47477256e-02 -5.30296385e-01 6.40229166e-01
3.09771597e-01 8.65582287e-01 -6.34851260e-03 -9.18162704e-01
4.58598137e-01 2.64773756e-01 4.60604280e-01 -6.76018119e-01
-6.25049293e-01 -1.99442238e-01 3.31378639e-01 -1.34829974e+00
-5.18047392e-01 -2.75862634e-01 -1.46663547e+00 -1.89669263e-02
1.99967735e-02 -1.46919310e-01 5.87926246e-02 1.17938817e+00
6.97145462e-01 7.86490321e-01 2.73645699e-01 -6.61276519e-01
-2.24785998e-01 -7.47765720e-01 -3.28222036e-01 4.80771184e-01
-1.50610566e-01 -8.87463450e-01 2.42523357e-01 4.17089015e-01] | [8.5090970993042, 0.609455406665802] |
d3036f88-99b2-47cc-a870-722ba2a4517c | a-comparative-evaluation-of-heart-rate | 2005.11101 | null | https://arxiv.org/abs/2005.11101v1 | https://arxiv.org/pdf/2005.11101v1.pdf | A Comparative Evaluation of Heart Rate Estimation Methods using Face Videos | This paper presents a comparative evaluation of methods for remote heart rate estimation using face videos, i.e., given a video sequence of the face as input, methods to process it to obtain a robust estimation of the subjects heart rate at each moment. Four alternatives from the literature are tested, three based in hand crafted approaches and one based on deep learning. The methods are compared using RGB videos from the COHFACE database. Experiments show that the learning-based method achieves much better accuracy than the hand crafted ones. The low error rate achieved by the learning based model makes possible its application in real scenarios, e.g. in medical or sports environments. | ['Javier Hernandez-Ortega', 'Julian Fierrez', 'Aythami Morales', 'David Diaz'] | 2020-05-22 | null | null | null | null | ['heart-rate-estimation'] | ['medical'] | [-1.65058747e-02 1.74670681e-01 -9.02225971e-02 -4.41261441e-01
-2.71371514e-01 -4.07478139e-02 2.20245123e-01 -3.05791169e-01
-7.06044674e-01 6.33865654e-01 -2.03344822e-01 -1.22441828e-01
7.58513957e-02 -4.34367746e-01 -2.46538952e-01 -1.00512493e+00
-1.82254568e-01 2.64910191e-01 -2.21013069e-01 1.01266317e-01
3.17522943e-01 6.78068459e-01 -1.63640952e+00 1.17543396e-02
2.61508524e-01 1.15508115e+00 -1.98898464e-01 9.40499127e-01
8.31455737e-02 9.35555995e-01 -7.72388935e-01 -1.55658498e-01
3.80105734e-01 -7.00267136e-01 -6.48103595e-01 1.36464253e-01
5.08207858e-01 -5.54682791e-01 -2.05615848e-01 5.86501539e-01
8.50313187e-01 -2.13402649e-03 4.57418710e-01 -1.12977064e+00
1.42877147e-01 1.80007309e-01 -4.07516599e-01 4.08080369e-01
7.89837539e-01 1.90104157e-01 2.31785521e-01 -7.38925934e-01
4.66746390e-01 1.19690144e+00 7.65942514e-01 8.18613470e-01
-1.18861544e+00 -4.70805496e-01 -4.60892439e-01 3.28004867e-01
-1.56729794e+00 -7.77031720e-01 7.65537560e-01 -4.60091352e-01
7.71675825e-01 2.77204126e-01 8.65453362e-01 1.06850410e+00
1.57485023e-01 3.82127762e-01 1.49577081e+00 -6.04489326e-01
4.33293909e-01 3.83136928e-01 -2.80270904e-01 8.21501374e-01
2.37114400e-01 1.52916715e-01 -6.57858372e-01 -1.87784936e-02
1.18675375e+00 -2.57185578e-01 -3.67627382e-01 -2.38083616e-01
-1.09194803e+00 7.27325320e-01 7.50768036e-02 5.45788705e-01
-5.94045877e-01 2.09002778e-01 5.31575918e-01 3.06817174e-01
4.03126031e-01 -2.11425480e-02 -4.18341964e-01 -2.09031850e-01
-1.43907332e+00 1.67290375e-01 1.10240579e+00 5.00157595e-01
5.88040292e-01 2.59893984e-01 -7.15358555e-02 3.42504561e-01
6.73253536e-01 5.26366830e-01 5.85310757e-01 -8.83815527e-01
-2.72432733e-02 7.86025524e-02 1.08197466e-01 -9.06316757e-01
-5.58144093e-01 1.08182587e-01 -5.69071889e-01 4.42498177e-01
7.24391460e-01 -3.29172194e-01 -8.17623138e-01 1.32066858e+00
5.55483639e-01 3.79111767e-01 8.66857823e-03 1.00136423e+00
1.18512416e+00 4.83069301e-01 1.59332529e-02 -7.52558410e-01
1.20206904e+00 -5.83093286e-01 -9.04554069e-01 1.90497801e-01
7.45713785e-02 -7.42457807e-01 5.56707025e-01 9.53139722e-01
-8.83111775e-01 -8.28845799e-01 -9.39167976e-01 2.16339499e-01
-1.20071270e-01 2.54546493e-01 5.26071012e-01 1.29326451e+00
-1.16984975e+00 8.45712960e-01 -7.88648367e-01 -5.35925567e-01
1.94581598e-01 4.97034639e-01 -2.55622178e-01 3.82710904e-01
-9.44783807e-01 9.87373233e-01 3.72792304e-01 5.14805496e-01
-9.54873025e-01 -3.11603814e-01 -6.64495945e-01 -2.39807010e-01
2.25886822e-01 -2.46513784e-01 1.38354945e+00 -1.27981293e+00
-2.12082243e+00 1.01322150e+00 -2.76063792e-02 -3.81119251e-01
9.41095471e-01 -1.16347134e-01 -1.89817309e-01 6.42379642e-01
-6.77604437e-01 4.66295898e-01 1.35169113e+00 -7.99765289e-01
-1.71311498e-01 -2.24243820e-01 -1.70089856e-01 2.87824869e-03
-1.30523816e-01 5.03238030e-02 -2.51600504e-01 -1.61143780e-01
-1.09027304e-01 -8.96173120e-01 7.13214697e-03 2.54181862e-01
-2.58805812e-03 -2.51199603e-01 7.28374004e-01 -8.88152719e-01
7.96013236e-01 -1.77826178e+00 1.20523565e-01 3.22866470e-01
1.28338009e-01 4.94907320e-01 1.40895322e-01 2.23541483e-01
-2.58125275e-01 -1.68771148e-01 3.17694485e-01 -1.50551543e-01
-2.68824428e-01 1.65919885e-01 2.45102599e-01 8.59811962e-01
-1.87101483e-01 4.52817321e-01 -6.72840953e-01 -8.11940849e-01
5.67219019e-01 1.13679183e+00 -2.90265143e-01 4.66650784e-01
3.41302067e-01 8.08621109e-01 -3.57894897e-01 6.31156683e-01
4.85994577e-01 2.43750408e-01 2.73481518e-01 -5.18565953e-01
-7.02601746e-02 -1.77370787e-01 -1.26614857e+00 1.54248583e+00
-6.44389272e-01 7.72967875e-01 -2.44647376e-02 -9.59328949e-01
1.16231740e+00 9.28218424e-01 7.32718110e-01 -4.01412815e-01
5.54044187e-01 2.09625065e-01 -7.09354086e-03 -9.26205993e-01
-9.27773938e-02 -4.49602067e-01 6.28770471e-01 3.77638370e-01
4.51841086e-01 -5.47675118e-02 1.57780528e-01 -3.05562079e-01
7.80333459e-01 5.53827763e-01 6.62294507e-01 -3.53870839e-01
8.47092628e-01 -6.50972009e-01 3.54987621e-01 4.54906136e-01
-5.83362460e-01 4.77006137e-01 4.47177500e-01 -1.03154278e+00
-7.58719742e-01 -5.04035830e-01 -1.48019403e-01 4.96310502e-01
-2.36633182e-01 -2.59804487e-01 -1.21762729e+00 -6.22970462e-01
-1.78803727e-01 -1.25934444e-02 -8.72837305e-01 1.25623271e-01
-5.88121533e-01 -6.54851913e-01 2.99052536e-01 4.28761810e-01
5.78094602e-01 -1.34362614e+00 -1.56465697e+00 1.23753294e-01
-3.61624777e-01 -1.02121484e+00 6.48559332e-02 -7.31827542e-02
-1.15941572e+00 -1.06331408e+00 -7.61349738e-01 -1.46330953e-01
4.02799428e-01 -1.74111113e-01 1.13520324e+00 3.53246361e-01
-8.12240124e-01 8.34953189e-01 -3.73653799e-01 -5.62671363e-01
-3.41034383e-01 -2.40619749e-01 2.07924277e-01 3.79041493e-01
6.07484579e-01 -3.29690576e-01 -7.05271900e-01 1.76456571e-01
-4.88055289e-01 -4.52298760e-01 5.44060946e-01 4.42641944e-01
4.52756763e-01 -2.19721243e-01 2.40514144e-01 -6.86950326e-01
1.56718194e-01 -5.68421185e-02 -6.61519408e-01 1.47773504e-01
-5.78249991e-01 3.64326574e-02 2.29189470e-01 -4.65639383e-01
-8.31080914e-01 5.40849686e-01 -2.66258359e-01 -5.23674846e-01
-3.84106994e-01 -3.19383554e-02 -2.95777712e-02 -4.51266497e-01
8.48822773e-01 2.23824531e-02 4.14680243e-01 -3.76120448e-01
-6.15976192e-02 5.46940982e-01 5.36228418e-01 -2.78811216e-01
6.00141108e-01 3.01830739e-01 8.85546729e-02 -1.31447148e+00
-3.96363378e-01 -3.64172280e-01 -9.91963446e-01 -1.06683111e+00
1.06782818e+00 -8.01444530e-01 -1.14247966e+00 4.86030847e-01
-1.08096409e+00 -3.60847712e-01 -1.95958674e-01 8.02205920e-01
-1.00732708e+00 2.23502591e-01 -4.73424673e-01 -1.04513812e+00
-7.07385361e-01 -9.75164175e-01 1.00028288e+00 6.28845572e-01
-1.42606094e-01 -1.23112738e+00 1.60822034e-01 3.77139509e-01
5.06994367e-01 5.86118102e-01 1.08292662e-01 -4.72035587e-01
-2.54105240e-01 -2.89419383e-01 3.57175142e-01 4.00115013e-01
8.32538158e-02 2.66472697e-01 -1.47428858e+00 -3.96618277e-01
3.51874679e-01 -3.30600947e-01 4.74574357e-01 4.60329831e-01
1.10006702e+00 -2.15009525e-01 -3.58109139e-02 4.85407203e-01
1.57524216e+00 1.05178729e-01 9.85750556e-01 8.91504139e-02
2.42065012e-01 4.22717035e-01 5.95516801e-01 6.66071713e-01
-1.58960879e-01 6.04146481e-01 4.44790542e-01 -2.78697163e-01
-6.48837537e-02 2.81831443e-01 4.37451601e-01 3.02002996e-01
-9.25907791e-01 1.92672538e-03 -6.82054996e-01 -7.73339644e-02
-1.41440856e+00 -1.16457999e+00 9.02765989e-03 2.34401631e+00
5.24612665e-01 -1.08576484e-01 5.83736360e-01 5.51724732e-01
6.78928733e-01 -1.33914314e-02 -3.02622557e-01 -5.59754670e-01
5.65451562e-01 6.23699546e-01 2.33404756e-01 3.55604887e-01
-1.06331694e+00 3.27671409e-01 7.10954285e+00 1.84923872e-01
-1.42391860e+00 2.52763242e-01 6.02162004e-01 -5.18486872e-02
4.60696161e-01 -3.65441054e-01 -7.28631854e-01 2.84250200e-01
1.38389695e+00 4.59573679e-02 4.74451393e-01 8.13762188e-01
4.46683705e-01 -5.01802266e-01 -1.10712028e+00 1.49608743e+00
4.53780949e-01 -7.58453727e-01 -5.11985421e-01 1.53622711e-02
1.53209656e-01 -3.23299766e-01 -2.69605398e-01 5.48075177e-02
-5.71218073e-01 -9.29602921e-01 6.42662883e-01 8.58906686e-01
8.42093050e-01 -6.90024674e-01 9.25799549e-01 -1.45257851e-02
-9.16571140e-01 6.31126240e-02 -3.35966945e-01 1.87278842e-03
-1.65909350e-01 6.42564595e-01 -8.15476120e-01 3.69345427e-01
7.20584273e-01 3.97867352e-01 -5.35333514e-01 1.13138402e+00
-3.15053999e-01 6.85696781e-01 -3.08737516e-01 -4.77712341e-02
-2.33168513e-01 -1.39874652e-01 2.45641142e-01 1.27701318e+00
3.78356934e-01 1.66715756e-01 -3.86190414e-02 5.44387937e-01
3.02999079e-01 3.10923219e-01 -5.40199816e-01 9.05135274e-02
-6.16800785e-03 1.79383576e+00 -1.03777981e+00 -3.91994357e-01
-4.46199179e-01 9.90907431e-01 -1.08968124e-01 1.11325160e-01
-9.95418310e-01 -3.99328232e-01 2.75363803e-01 2.19285160e-01
2.15882361e-01 -7.56796822e-02 2.85016000e-01 -1.04679585e+00
-1.44487366e-01 -8.17501485e-01 3.99525076e-01 -7.07792640e-01
-6.05129659e-01 8.47742617e-01 1.81204796e-01 -1.05100155e+00
-5.27612865e-01 -8.28813910e-01 -6.04383528e-01 8.08346391e-01
-1.39361882e+00 -6.83721304e-01 -7.95898438e-01 8.15737963e-01
5.81060052e-01 -1.63379997e-01 7.91873872e-01 3.83595407e-01
-6.62793159e-01 3.49901319e-01 -4.57161009e-01 7.49696121e-02
7.67529666e-01 -1.18918121e+00 -2.74398535e-01 6.57171786e-01
2.38873616e-01 2.90091604e-01 8.87528241e-01 -1.97220072e-01
-1.54080784e+00 -5.60804605e-01 7.44034946e-01 -4.10496980e-01
1.64293370e-03 -2.02504754e-01 -7.02752650e-01 5.47151566e-01
4.73900914e-01 5.47038674e-01 6.82399690e-01 -1.53072953e-01
5.18156849e-02 -4.68971968e-01 -1.48039699e+00 -1.17534369e-01
2.93808728e-01 -2.92868108e-01 -5.24408460e-01 3.97608876e-01
-2.64040709e-01 -5.14379501e-01 -1.14642119e+00 1.97837189e-01
9.82874930e-01 -1.35317779e+00 8.55252743e-01 -2.31370121e-01
-1.50849875e-02 -1.78526208e-01 3.41172129e-01 -1.05181801e+00
1.01152264e-01 -8.14486027e-01 -2.95655698e-01 9.34358120e-01
5.30051839e-05 -5.23302555e-01 7.70100772e-01 4.62003022e-01
5.91084957e-01 -4.57937688e-01 -9.93812799e-01 -6.71759009e-01
-4.53079104e-01 -2.36338284e-03 9.08288583e-02 6.26763821e-01
-1.17585659e-01 1.14256449e-01 -5.67005157e-01 -8.36125314e-02
7.37944841e-01 -5.84595613e-02 8.62831771e-01 -1.39243352e+00
-1.85922265e-01 4.28790227e-02 -8.64328861e-01 -1.69347063e-01
2.21816860e-02 -2.93945670e-01 1.53698176e-01 -1.06243956e+00
2.90222727e-02 -1.02796331e-01 -2.10377827e-01 6.32761359e-01
-6.32196292e-03 6.87972903e-01 1.21753328e-01 -8.12498257e-02
-2.38737196e-01 -5.47725633e-02 8.54509413e-01 3.53443682e-01
-2.84414411e-01 1.57495037e-01 -6.39215717e-03 7.43951917e-01
1.04340839e+00 -4.07932997e-01 -3.12140077e-01 9.41097736e-02
-6.73476830e-02 3.61706167e-01 4.31692511e-01 -1.52163565e+00
-5.30008785e-02 7.60924816e-02 8.58448863e-01 -1.58091635e-01
2.77297676e-01 -8.96103859e-01 3.00916702e-01 9.60932791e-01
-1.31921098e-01 -2.70792693e-02 -8.67838692e-03 3.17395210e-01
-4.88056867e-05 -4.53833133e-01 1.22594607e+00 -4.12773907e-01
-4.55825537e-01 1.35181069e-01 -4.35773224e-01 -3.27327132e-01
9.82494593e-01 -4.51475918e-01 2.22653478e-01 -4.63015646e-01
-1.13860440e+00 -5.60673356e-01 -4.13415246e-02 6.15028925e-02
4.72202748e-01 -1.08157718e+00 -6.88298702e-01 4.25603539e-01
-2.83074707e-01 -6.47504508e-01 6.08191825e-02 1.33335900e+00
-8.95102203e-01 2.45537445e-01 -6.49874866e-01 -7.00577021e-01
-1.64723051e+00 6.40586853e-01 7.17006147e-01 9.95965451e-02
-6.15904093e-01 5.45689046e-01 -5.40381670e-01 6.91239238e-02
3.28879356e-01 -3.39907646e-01 -3.70057374e-01 2.33263880e-01
9.21333671e-01 5.65822601e-01 3.29781294e-01 -6.18829250e-01
-4.75913614e-01 8.21066141e-01 4.98357058e-01 -1.12266861e-01
1.15364742e+00 1.96722783e-02 -7.18259662e-02 6.08428836e-01
9.68568683e-01 -1.21394411e-01 -1.18991971e+00 9.20836329e-02
-1.74925223e-01 -6.15984440e-01 2.48535991e-01 -6.79176807e-01
-1.52983236e+00 1.07826364e+00 1.20288420e+00 4.83743288e-02
1.52859247e+00 -3.95460844e-01 1.70418933e-01 4.79195416e-01
3.68234515e-01 -1.17548859e+00 1.49667799e-01 -1.97197169e-01
7.44847834e-01 -1.17129374e+00 3.60469699e-01 -2.46470451e-01
-5.10502994e-01 1.94848943e+00 4.31821972e-01 -2.96219792e-02
6.67976975e-01 2.09068075e-01 3.30396980e-01 -4.19973060e-02
-4.76201743e-01 -2.09842369e-01 8.17543045e-02 6.61474586e-01
6.27257884e-01 -2.42631242e-01 -5.48240602e-01 -1.86389849e-01
1.79697990e-01 4.41346049e-01 7.01119602e-01 9.69976664e-01
-3.57751012e-01 -7.96711326e-01 -7.60961056e-01 1.51607662e-01
-5.83895147e-01 4.11521643e-01 -2.57047683e-01 1.02708960e+00
2.28729516e-01 8.82471025e-01 -7.04821274e-02 -9.79399905e-02
2.32376948e-01 3.63704681e-01 1.13758242e+00 -1.73239395e-01
-6.77366376e-01 2.32778400e-01 -9.36543494e-02 -9.78975832e-01
-1.05055094e+00 -7.89774835e-01 -9.99625266e-01 -1.84945554e-01
-2.88302720e-01 1.12317830e-01 1.05483115e+00 8.22172105e-01
-2.08264798e-01 1.70611307e-01 8.61921787e-01 -1.30298507e+00
-2.45419249e-01 -1.04427493e+00 -7.68921793e-01 1.05157308e-01
4.37785476e-01 -6.17092073e-01 -4.07603055e-01 4.07607615e-01] | [13.858908653259277, 2.666558265686035] |
d7bb7467-6f64-4bcd-819f-93b960cbb882 | gender-bias-evaluation-in-luganda-english | null | null | https://aclanthology.org/2022.amta-research.21 | https://aclanthology.org/2022.amta-research.21.pdf | Gender bias Evaluation in Luganda-English Machine Translation | We have seen significant growth in the area of building Natural Language Processing (NLP) tools for African languages. However, the evaluation of gender bias in the machine translation systems for African languages is not yet thoroughly investigated. This is due to the unavailability of explicit text data available for addressing the issue of gender bias in machine translation. In this paper, we use transfer learning techniques based on a pre-trained Marian MT model for building machine translation models for English-Luganda and Luganda-English. Our work attempts to evaluate and quantify the gender bias within a Luganda-English machine translation system using Word Embeddings Fairness Evaluation Framework (WEFE). Luganda is one of the languages with gender-neutral pronouns in the world, therefore we use a small set of trusted gendered examples as the test set to evaluate gender bias by biasing word embeddings. This approach allows us to focus on Luganda-Engish translations with gender-specific pronouns, and the results of the gender bias evaluation are confirmed by human evaluation. To compare and contrast the results of the word embeddings evaluation metric, we used a modified version of the existing Translation Gender Bias Index (TGBI) based on the grammatical consideration for Luganda. | ['Eric Peter Wairagala'] | null | null | null | null | amta-2022-9 | ['embeddings-evaluation'] | ['natural-language-processing'] | [-3.08792800e-01 2.10562080e-01 -4.90489691e-01 -6.48617208e-01
-2.34061450e-01 -5.82550228e-01 1.22632766e+00 2.57879615e-01
-8.72541130e-01 9.54717398e-01 3.39190096e-01 -6.60714686e-01
1.47274107e-01 -7.90150940e-01 -2.10783288e-01 -6.37231469e-01
3.69908810e-01 1.09100795e+00 -5.01095951e-01 -7.21119106e-01
6.27030551e-01 4.18327838e-01 -1.14673829e+00 -1.19947009e-01
8.99811506e-01 1.78260118e-01 -1.19643174e-01 4.30875629e-01
-1.04000725e-01 -4.74904701e-02 -6.05321944e-01 -8.88585329e-01
2.43978947e-01 -5.78495502e-01 -9.33374822e-01 -5.64113259e-01
5.13986349e-01 -1.03353932e-01 2.20312014e-01 1.11782622e+00
8.62082183e-01 -2.37584829e-01 9.77292418e-01 -1.24962485e+00
-7.79991984e-01 6.68460250e-01 -6.35902822e-01 3.99562150e-01
1.68006361e-01 -2.45847940e-01 9.45159435e-01 -1.11834288e+00
9.52057421e-01 1.93279386e+00 4.06948149e-01 8.43473494e-01
-1.06223798e+00 -9.19754446e-01 -3.72276366e-01 1.04840860e-01
-1.14927816e+00 -4.51237470e-01 4.55721855e-01 -4.59583342e-01
7.92538643e-01 2.44046077e-01 4.04524982e-01 1.32616365e+00
6.70221806e-01 8.24568197e-02 1.80889189e+00 -8.78022313e-01
-7.54326284e-02 5.63339889e-01 -3.54203433e-02 3.33633661e-01
6.01029336e-01 2.63688236e-01 -6.15277112e-01 -2.03986168e-01
3.64942372e-01 -6.01686835e-01 4.00191426e-01 -1.04814753e-01
-1.26752615e+00 1.15681720e+00 -1.29381746e-01 6.56577647e-01
-8.82826895e-02 5.63760735e-02 7.62128294e-01 5.26745141e-01
9.17717576e-01 5.53618014e-01 -4.35317993e-01 -3.14882576e-01
-1.02930117e+00 6.27275586e-01 7.23560154e-01 4.98425722e-01
6.75521016e-01 -5.63870408e-02 -2.64155060e-01 9.33127284e-01
5.33405840e-01 9.79985118e-01 5.68790078e-01 -4.51652497e-01
6.22531593e-01 5.56657076e-01 -1.32532297e-02 -1.10008705e+00
8.05637762e-02 1.23037055e-01 -3.89300227e-01 2.23375663e-01
6.02357805e-01 -6.33348972e-02 -8.22958231e-01 1.76034009e+00
1.90688625e-01 -1.06042361e+00 1.32025212e-01 9.16209400e-01
6.35258675e-01 6.59477711e-01 5.38469493e-01 -5.28746098e-02
1.49551094e+00 -4.22702461e-01 -7.39500642e-01 -1.12856470e-01
7.34862566e-01 -1.23636830e+00 1.25167418e+00 -6.40280098e-02
-7.33387351e-01 -2.74006963e-01 -9.54633474e-01 -1.52278230e-01
-7.94411242e-01 6.53639883e-02 6.24792099e-01 1.31009626e+00
-7.82498479e-01 4.43129122e-01 -5.92804670e-01 -9.82921422e-01
2.24800974e-01 4.83479828e-01 -5.80365479e-01 -9.04227570e-02
-1.36636651e+00 1.54015017e+00 2.03197137e-01 -2.08156049e-01
-4.29642379e-01 -4.87830967e-01 -7.66071320e-01 -5.01210749e-01
-4.34930056e-01 -2.52367496e-01 8.91899049e-01 -1.24490499e+00
-1.07336879e+00 1.64626050e+00 -3.48200649e-01 2.98101306e-02
6.12405598e-01 5.25785498e-02 -5.34007847e-01 -3.17080379e-01
5.90698719e-01 9.00115073e-01 5.79873264e-01 -1.14366889e+00
-5.88727057e-01 -7.57810116e-01 -1.51734471e-01 1.24104194e-01
-3.65788311e-01 9.52914894e-01 4.71562117e-01 -7.13444769e-01
-2.64899731e-01 -9.38270628e-01 2.36644909e-01 -6.27970874e-01
9.63936001e-02 -6.19644642e-01 7.41964042e-01 -1.07302094e+00
1.25492775e+00 -1.87085652e+00 7.95069113e-02 2.87594736e-01
-1.26140624e-01 2.40641817e-01 3.22040319e-02 6.78897440e-01
-8.94807130e-02 4.40713257e-01 -1.25358015e-01 -1.69422328e-01
3.54497403e-01 5.85812449e-01 -1.96291283e-01 5.37578285e-01
2.92101681e-01 7.21768200e-01 -8.35468829e-01 -9.15916681e-01
-9.16104391e-02 2.80838251e-01 -4.27812874e-01 1.48560042e-02
4.22645211e-01 4.07013237e-01 4.66785580e-02 8.18371236e-01
9.74693239e-01 1.01704299e+00 2.76511282e-01 -3.04981247e-02
-4.61513817e-01 3.88772219e-01 -6.54273927e-01 1.29491472e+00
-5.21713197e-01 6.96271837e-01 -1.91533700e-01 -7.37555265e-01
1.34709060e+00 1.02510087e-01 -1.53127506e-01 -8.91037464e-01
5.74609816e-01 9.69783068e-01 6.58202767e-01 -1.96444228e-01
9.08303320e-01 -5.32657862e-01 -3.61064225e-01 5.58584154e-01
3.94926876e-01 -2.85048187e-01 6.42042696e-01 -2.63362169e-01
3.15775126e-01 2.52259970e-01 2.34251261e-01 -1.09538782e+00
5.40585995e-01 1.55662060e-01 6.79800391e-01 5.99390306e-02
-4.68064725e-01 4.84264612e-01 7.20530689e-01 -5.66626370e-01
-1.47518384e+00 -8.58713686e-01 -4.35388684e-01 1.48482931e+00
-2.29628384e-01 -2.31215805e-01 -8.81389737e-01 -7.57997870e-01
-9.69343260e-02 1.02081239e+00 -8.13916266e-01 1.54813811e-01
-8.84434342e-01 -1.15958703e+00 8.99512053e-01 1.32909015e-01
3.95515889e-01 -8.99070263e-01 -6.01299345e-01 -1.71502680e-01
-7.03163445e-02 -6.64001703e-01 -2.05634385e-01 6.18892573e-02
-8.24339747e-01 -8.39743316e-01 -7.06411004e-01 -9.16228771e-01
4.97321993e-01 -5.28875470e-01 1.40082908e+00 -2.22782224e-01
-9.57460850e-02 -1.04581743e-01 -2.74069756e-01 -1.08964002e+00
-8.33273888e-01 3.74955654e-01 3.34320486e-01 -4.76274639e-01
1.23163831e+00 -3.18349957e-01 -3.01030189e-01 4.05393481e-01
-6.75905585e-01 -3.52502674e-01 4.10615146e-01 7.23941445e-01
-1.59169719e-01 -6.88951075e-01 4.46535885e-01 -8.88355315e-01
8.63949120e-01 -4.18928981e-01 -2.03901142e-01 4.00019176e-02
-1.03673899e+00 -3.66884023e-02 1.07023261e-01 -3.05049241e-01
-9.13541794e-01 -8.35883737e-01 -1.24162689e-01 3.08393240e-01
-1.05300023e-04 1.33293957e-01 -2.38388151e-01 5.40123284e-02
6.76155269e-01 -3.73090446e-01 1.13290980e-01 -1.34532154e-01
1.52233332e-01 1.11560297e+00 1.15718067e-01 -1.11508834e+00
7.03055203e-01 -5.88960834e-02 1.04893139e-03 -6.72329187e-01
5.37884124e-02 -2.64681783e-02 -6.62777781e-01 -1.32866949e-01
9.13348079e-01 -7.32839346e-01 -1.40750691e-01 2.48047948e-01
-1.25708342e+00 -1.11075975e-02 4.80947532e-02 6.72015667e-01
-3.00673932e-01 6.73614666e-02 -3.87873054e-01 -7.52785265e-01
-4.99643624e-01 -9.94134665e-01 9.30599391e-01 -5.57371452e-02
-9.79251444e-01 -1.04655218e+00 5.00958979e-01 1.66462541e-01
4.90117401e-01 1.24544889e-01 1.44185865e+00 -8.51981461e-01
4.87846136e-01 1.75211597e-02 -3.68176192e-01 3.46933812e-01
2.24420950e-02 2.12745711e-01 -7.43326247e-01 -1.82274818e-01
-6.12600893e-02 -2.32562885e-01 3.71986359e-01 2.56766751e-02
1.71603724e-01 -1.00847639e-01 2.24912073e-02 9.99903232e-02
1.45174825e+00 1.40665784e-01 6.75617337e-01 7.04816997e-01
5.62981963e-01 1.07485366e+00 1.00338697e+00 -9.87486728e-03
6.11624658e-01 4.89046335e-01 1.23791337e-01 -4.95131947e-02
-6.69380426e-02 -2.42024228e-01 4.91219550e-01 8.42684984e-01
-6.22653008e-01 4.40178672e-03 -1.34074903e+00 9.20720994e-01
-1.51151776e+00 -5.66282511e-01 -1.32688493e-01 1.98535323e+00
8.13375533e-01 -1.99611023e-01 1.14479214e-01 1.67627215e-01
8.68281484e-01 2.13741168e-01 3.30675781e-01 -1.68927312e+00
-6.66573197e-02 7.52786219e-01 7.14936018e-01 4.81428325e-01
-9.49729741e-01 1.09729099e+00 6.13957930e+00 9.15890455e-01
-1.11502993e+00 3.06477547e-01 7.52179801e-01 3.40272278e-01
-4.34136599e-01 4.69440967e-03 -7.38593757e-01 3.43421191e-01
1.10589933e+00 -2.00516224e-01 2.33470753e-01 5.16723752e-01
3.81912291e-01 -5.77873364e-02 -1.25908494e+00 1.01852131e+00
2.50367701e-01 -5.87279260e-01 2.28231609e-01 3.57848287e-01
6.49182498e-01 -1.47301525e-01 1.32848203e-01 4.73009557e-01
1.25286309e-02 -1.28702986e+00 6.71485662e-01 -1.38500556e-01
9.93133426e-01 -1.16884887e+00 1.24960005e+00 3.67130749e-02
-2.16888934e-01 9.94748473e-02 -7.09384203e-01 -3.69238526e-01
-2.68062204e-01 3.40808988e-01 -9.23326194e-01 4.63193655e-01
7.38305926e-01 1.35964736e-01 -6.16955221e-01 1.51270166e-01
-2.35372752e-01 6.95784390e-01 2.00311448e-02 -3.31417561e-01
2.95620054e-01 -5.60192466e-01 5.38687527e-01 1.51208806e+00
5.68207264e-01 -4.52659965e-01 -3.68106574e-01 7.42794991e-01
3.36801820e-02 9.01236415e-01 -7.82015324e-01 -3.93380761e-01
2.93817967e-01 1.32426393e+00 -6.57084346e-01 -9.53317955e-02
-2.24280894e-01 7.27066875e-01 -5.32251596e-02 -8.54839683e-02
-5.34650505e-01 -4.44806159e-01 9.48295116e-01 2.21938491e-01
-3.25761497e-01 -2.92548567e-01 -6.64173484e-01 -8.34775746e-01
-2.72648066e-01 -1.20061970e+00 5.28283753e-02 -2.21148178e-01
-1.16184843e+00 3.70505542e-01 3.76623243e-01 -8.10173154e-01
-3.99844646e-01 -9.07642066e-01 -5.02216876e-01 1.41130757e+00
-1.13580751e+00 -1.61955583e+00 3.08538407e-01 -1.30234659e-02
2.49988154e-01 -7.23901749e-01 1.15176952e+00 5.52125156e-01
-1.88015029e-01 8.50617886e-01 -3.18771064e-01 1.50936067e-01
1.40607762e+00 -1.25735366e+00 4.01631564e-01 7.57132471e-01
-3.49310189e-01 8.66435111e-01 1.16491008e+00 -6.34130239e-01
-9.83875453e-01 -6.99463129e-01 2.11848330e+00 -7.34358013e-01
5.77562034e-01 -5.77017546e-01 -3.16855609e-01 4.41091150e-01
7.80096710e-01 -4.82338518e-01 9.08542395e-01 3.52696806e-01
-4.63665068e-01 5.89670055e-03 -1.55414104e+00 6.69077098e-01
8.68668973e-01 -3.58604997e-01 -1.07567859e+00 -1.91236846e-03
1.87849134e-01 -6.87070787e-02 -7.31618524e-01 3.21633071e-01
6.80159748e-01 -7.12246895e-01 5.06599307e-01 -7.83853173e-01
1.12165642e+00 -7.75665194e-02 -4.75526005e-01 -1.50462234e+00
-7.32147247e-02 -1.74150079e-01 6.67814732e-01 1.67363906e+00
4.82391417e-01 -8.07200193e-01 5.20779133e-01 4.10419255e-01
2.88988203e-01 -4.53514516e-01 -1.15451694e+00 -5.94518065e-01
1.01362729e+00 2.90431958e-02 8.24023902e-01 1.19589925e+00
-1.42008051e-01 3.53471428e-01 -1.80466995e-01 -2.85803527e-01
4.17833954e-01 -1.06304236e-01 8.69617462e-01 -1.20636332e+00
1.90831110e-01 -4.56156582e-01 -6.52456284e-01 1.85751319e-01
4.03313428e-01 -1.06286860e+00 -2.22697362e-01 -1.27746701e+00
3.51993263e-01 -3.13178450e-01 6.51841685e-02 2.44080141e-01
2.18029018e-03 6.88849151e-01 1.67484269e-01 2.96707470e-02
2.31572196e-01 2.82188088e-01 1.01798594e+00 -1.04577422e-01
1.97311789e-01 -4.22454596e-01 -9.17941749e-01 6.24391854e-01
9.06240404e-01 -7.99393177e-01 -1.18626967e-01 -4.21190143e-01
5.89710712e-01 -7.19810784e-01 2.34105922e-02 -4.50274378e-01
-5.67866087e-01 -4.00134027e-01 4.12730753e-01 -3.10475558e-01
-2.86155969e-01 -5.65350473e-01 -2.43875384e-01 5.10164738e-01
3.61571945e-02 6.67001426e-01 3.82681079e-02 -3.45852613e-01
-3.86684239e-01 -3.92265767e-01 5.93380153e-01 -1.37751296e-01
-4.70557630e-01 5.96849211e-02 -5.27321875e-01 2.03426592e-02
7.14097321e-01 -3.36416543e-01 -2.87806541e-01 7.69803077e-02
-6.69389963e-02 -3.47483084e-02 7.27596700e-01 7.43580759e-01
7.96099082e-02 -1.62511456e+00 -1.13439453e+00 3.73742320e-02
4.25398588e-01 -8.70433748e-01 -4.30149794e-01 7.58451521e-01
-9.09816682e-01 3.08075190e-01 -8.94108057e-01 -2.24251181e-01
-1.43363369e+00 9.27123874e-02 -8.32090825e-02 -2.53423333e-01
2.72417456e-01 5.54974914e-01 -1.09569162e-01 -1.17463934e+00
-4.66706991e-01 -1.27758294e-01 -4.36289966e-01 4.55630183e-01
1.13060996e-01 7.00151742e-01 1.54863939e-01 -1.13374853e+00
-6.58852458e-01 4.59260881e-01 1.26947105e-01 -6.32320523e-01
1.24567413e+00 -5.15952259e-02 -9.26652849e-01 5.14820635e-01
1.07105720e+00 5.44018209e-01 1.82321817e-02 5.20567417e-01
1.04421616e-01 -7.80040026e-01 -3.64912927e-01 -8.48869860e-01
-3.57148439e-01 9.88633275e-01 9.96860802e-01 -3.78071278e-01
5.66663682e-01 -4.70889568e-01 4.10473764e-01 6.05430827e-02
4.18948412e-01 -1.63688946e+00 -6.25090539e-01 8.81141305e-01
7.67659485e-01 -1.25480592e+00 -7.96051100e-02 -1.54505327e-01
-4.95525062e-01 1.15891218e+00 4.32897419e-01 -1.56786770e-01
3.47775996e-01 -1.57801107e-01 6.73508346e-01 -6.34389967e-02
-2.87314564e-01 1.76016569e-01 1.21426769e-01 8.66071165e-01
1.15298188e+00 4.92230713e-01 -1.78848660e+00 5.31238437e-01
-8.45150709e-01 -2.99964875e-01 3.54650110e-01 6.26926899e-01
-1.14048302e-01 -1.91937733e+00 -6.60281241e-01 1.51174232e-01
-9.56958652e-01 -1.42685816e-01 -7.28587568e-01 1.13832557e+00
5.51418543e-01 8.90099049e-01 2.40012541e-01 -3.15173417e-01
1.41954303e-01 3.89828056e-01 8.39950860e-01 -5.31339884e-01
-8.91023397e-01 -2.24607825e-01 5.75753748e-01 2.04441976e-02
-5.61032772e-01 -7.67058969e-01 -8.00690889e-01 -8.23242128e-01
1.53219521e-01 3.12166333e-01 1.12145698e+00 8.37331533e-01
-5.80662452e-02 -3.63813080e-02 3.67412686e-01 -8.23540032e-01
-4.46703613e-01 -1.40033138e+00 -4.98647422e-01 5.49674273e-01
-4.03431088e-01 -6.77051723e-01 -1.94321916e-01 -3.75265568e-01] | [9.407405853271484, 10.254837036132812] |
a44e8ed4-4674-4c9d-a8d3-8875466d4f55 | taming-diffusion-models-for-audio-driven-co | 2303.09119 | null | https://arxiv.org/abs/2303.09119v2 | https://arxiv.org/pdf/2303.09119v2.pdf | Taming Diffusion Models for Audio-Driven Co-Speech Gesture Generation | Animating virtual avatars to make co-speech gestures facilitates various applications in human-machine interaction. The existing methods mainly rely on generative adversarial networks (GANs), which typically suffer from notorious mode collapse and unstable training, thus making it difficult to learn accurate audio-gesture joint distributions. In this work, we propose a novel diffusion-based framework, named Diffusion Co-Speech Gesture (DiffGesture), to effectively capture the cross-modal audio-to-gesture associations and preserve temporal coherence for high-fidelity audio-driven co-speech gesture generation. Specifically, we first establish the diffusion-conditional generation process on clips of skeleton sequences and audio to enable the whole framework. Then, a novel Diffusion Audio-Gesture Transformer is devised to better attend to the information from multiple modalities and model the long-term temporal dependency. Moreover, to eliminate temporal inconsistency, we propose an effective Diffusion Gesture Stabilizer with an annealed noise sampling strategy. Benefiting from the architectural advantages of diffusion models, we further incorporate implicit classifier-free guidance to trade off between diversity and gesture quality. Extensive experiments demonstrate that DiffGesture achieves state-of-theart performance, which renders coherent gestures with better mode coverage and stronger audio correlations. Code is available at https://github.com/Advocate99/DiffGesture. | ['Lequan Yu', 'Ziwei Liu', 'Rui Qian', 'Xuanyu Liu', 'Xian Liu', 'Lingting Zhu'] | 2023-03-16 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Zhu_Taming_Diffusion_Models_for_Audio-Driven_Co-Speech_Gesture_Generation_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Zhu_Taming_Diffusion_Models_for_Audio-Driven_Co-Speech_Gesture_Generation_CVPR_2023_paper.pdf | cvpr-2023-1 | ['gesture-generation'] | ['robots'] | [ 4.64200117e-02 -3.74684483e-02 -7.38129243e-02 -5.32199182e-02
-1.03254771e+00 -4.11067039e-01 7.68374503e-01 -6.98686123e-01
9.10982117e-02 3.88326287e-01 5.80361068e-01 6.76939711e-02
2.02984009e-02 -5.74160278e-01 -4.56101537e-01 -1.13295472e+00
-7.65714096e-03 1.41565412e-01 2.63011120e-02 -2.67076969e-01
-4.34519917e-01 1.37128159e-01 -1.28015828e+00 6.55604824e-02
8.66394281e-01 7.50555336e-01 2.40178499e-02 9.53975677e-01
1.98185425e-02 7.80353725e-01 -6.75135612e-01 -4.08357829e-01
1.04823880e-01 -1.01318276e+00 -2.01590031e-01 -4.94278707e-02
-4.87516038e-02 -4.75393593e-01 -6.75520897e-01 8.95211816e-01
9.36712086e-01 1.92886293e-01 7.39368796e-01 -1.40816176e+00
-5.07602930e-01 7.21702754e-01 -5.20743966e-01 -1.48346618e-01
4.54410225e-01 7.32859910e-01 9.67594147e-01 -7.44782805e-01
5.58035314e-01 1.32924330e+00 5.63779712e-01 9.28325653e-01
-1.05179501e+00 -1.00347269e+00 1.93040535e-01 8.41425508e-02
-1.46939802e+00 -6.45340502e-01 1.18235779e+00 -2.05872297e-01
3.66071463e-01 5.38232327e-01 9.48016524e-01 1.97408688e+00
-1.70644447e-01 9.99747992e-01 7.53901184e-01 -1.82053909e-01
1.09843493e-01 -4.21024501e-01 -5.44716716e-01 4.72950220e-01
-6.51733041e-01 4.32863116e-01 -7.56686389e-01 -8.57032165e-02
1.09619808e+00 7.55760372e-02 -3.71819437e-01 -6.54829964e-02
-1.13219690e+00 5.99605441e-01 3.10122043e-01 5.08637249e-01
-4.85857517e-01 3.32689077e-01 3.40483010e-01 2.44198501e-01
2.35341132e-01 -1.59526691e-01 1.44362077e-01 -5.91375470e-01
-8.11283588e-01 3.40770006e-01 5.87153792e-01 8.18635702e-01
1.22551821e-01 4.94862914e-01 -3.65196019e-01 8.86327147e-01
5.53466558e-01 6.83686852e-01 5.69477797e-01 -9.15433824e-01
4.72786933e-01 2.02677101e-02 -2.93785244e-01 -9.54885364e-01
-3.96359451e-02 -3.78111601e-01 -1.36144543e+00 1.77855760e-01
2.54708827e-01 -3.45377445e-01 -7.92676210e-01 2.07971525e+00
5.31697869e-01 5.86928010e-01 4.68189530e-02 1.08655906e+00
7.12499857e-01 6.97703123e-01 1.63120300e-01 -3.59778076e-01
1.01960647e+00 -7.60568619e-01 -8.88630748e-01 2.29441240e-01
3.98574546e-02 -7.21772015e-01 1.17317665e+00 3.21066916e-01
-9.72649932e-01 -5.81407309e-01 -1.02397847e+00 2.83057600e-01
4.68529195e-01 -1.46879286e-01 4.47651416e-01 5.13994813e-01
-7.45524526e-01 3.96799028e-01 -1.30202150e+00 -2.56180242e-02
2.92610705e-01 1.38678208e-01 -1.83754005e-02 1.95329919e-01
-1.19808829e+00 2.02346474e-01 8.09696168e-02 2.48978764e-01
-1.11358011e+00 -4.53842998e-01 -7.45068073e-01 -2.75408328e-01
3.73104900e-01 -7.43376672e-01 1.28180873e+00 -1.08665633e+00
-2.20750308e+00 2.06582293e-01 -2.75220908e-02 -1.18425265e-01
1.01612163e+00 -2.03805700e-01 -5.37197590e-01 1.74253538e-01
-2.90807009e-01 5.07376373e-01 1.27784169e+00 -1.30472159e+00
-3.03324223e-01 -1.89841628e-01 -2.23198265e-01 4.44854677e-01
-4.68706608e-01 -1.55692145e-01 -7.15215504e-01 -1.27073205e+00
8.71583670e-02 -1.05588078e+00 -1.33067593e-01 -8.88131335e-02
-6.10328913e-01 6.59227185e-03 7.41031945e-01 -7.22189367e-01
1.34042645e+00 -2.24749327e+00 5.18887103e-01 2.67823458e-01
2.79716522e-01 1.28624216e-01 -2.68516511e-01 5.04780352e-01
1.76438361e-01 -8.06526989e-02 -1.30396619e-01 -4.68762279e-01
1.50334939e-01 1.97041675e-01 -2.56375372e-01 3.45742434e-01
1.54152051e-01 9.35898066e-01 -8.07489097e-01 -6.32110715e-01
2.43348375e-01 8.96365225e-01 -6.38301492e-01 4.52990294e-01
-1.95321709e-01 1.12418163e+00 -7.32848763e-01 6.59040689e-01
4.17844057e-01 1.37961879e-01 2.31873065e-01 -9.94965136e-02
2.15666592e-01 3.48352343e-02 -1.12953699e+00 1.97961724e+00
-5.35475254e-01 3.57501596e-01 3.30474705e-01 -6.63716197e-01
8.99419665e-01 6.77595437e-01 5.98105788e-01 -6.20915353e-01
3.26330096e-01 1.69442192e-01 -7.55483061e-02 -5.79174101e-01
1.11174524e-01 -3.47545147e-01 -6.95629716e-02 3.82875979e-01
8.85323733e-02 -2.17692941e-01 -3.97927552e-01 7.09745809e-02
1.02872276e+00 3.16631258e-01 -1.02621354e-01 3.95401448e-01
2.64848441e-01 -5.71543872e-01 5.30388892e-01 3.27245533e-01
-1.08714484e-01 8.59340847e-01 2.35600665e-01 1.03303395e-01
-8.49175930e-01 -1.20100224e+00 3.65427494e-01 9.12618518e-01
2.42853150e-01 -3.23638529e-01 -8.40483546e-01 -4.09599304e-01
-3.90901357e-01 4.00080621e-01 -4.98791516e-01 -2.59749234e-01
-6.80520117e-01 -5.77758908e-01 9.59780097e-01 5.11454344e-01
4.76900667e-01 -1.06395495e+00 -1.92257702e-01 3.37844342e-01
-4.41188782e-01 -7.10352957e-01 -9.04302061e-01 -1.97130337e-01
-7.08314478e-01 -6.97298765e-01 -1.20599568e+00 -5.88870943e-01
5.29505312e-02 -8.47922638e-02 5.99891007e-01 -3.49985547e-02
-2.96243820e-02 2.78826326e-01 -5.84914744e-01 -4.22560126e-02
-6.80419147e-01 9.90621820e-02 1.12307049e-01 2.36380205e-01
-3.44203934e-02 -1.36717308e+00 -6.16400242e-01 3.35222989e-01
-9.81405437e-01 1.27564520e-01 7.18609214e-01 1.16211331e+00
4.56946284e-01 -1.34620413e-01 5.94190776e-01 -2.45973989e-01
7.69782186e-01 -4.36670721e-01 -6.16184808e-02 -5.08045740e-02
-2.78614372e-01 -5.66840433e-02 4.76765931e-01 -1.07332802e+00
-1.12094581e+00 4.41999957e-02 -5.77114463e-01 -9.96575058e-01
-8.30892548e-02 2.76000202e-01 -5.57072699e-01 3.86284292e-01
5.41105926e-01 4.97654080e-01 2.70716935e-01 -3.16923887e-01
5.88628650e-01 6.92675054e-01 9.03748035e-01 -6.44393802e-01
1.01957440e+00 2.51083881e-01 -3.76035094e-01 -7.61147380e-01
-1.20884910e-01 -1.11392274e-01 -2.60200858e-01 -4.58943605e-01
8.58333170e-01 -9.58231032e-01 -7.23249912e-01 8.03998888e-01
-1.06669247e+00 -6.97821796e-01 -2.40662009e-01 5.55127680e-01
-6.77970707e-01 3.02827835e-01 -9.04017925e-01 -1.06222641e+00
-3.96316797e-01 -9.87910986e-01 1.07258797e+00 3.23509693e-01
-2.45661542e-01 -6.43918753e-01 2.45357245e-01 1.91389933e-01
2.74753302e-01 6.02679074e-01 3.48182917e-01 -3.38164657e-01
-5.71941078e-01 4.04774360e-02 2.91839331e-01 2.25170910e-01
1.98782608e-01 1.04486346e-01 -9.66889918e-01 -1.41593963e-01
-9.15231854e-02 -2.91967809e-01 5.74598134e-01 3.53357881e-01
6.66752636e-01 -5.63455164e-01 -1.59959257e-01 7.26727664e-01
8.66251111e-01 2.84488350e-01 7.74354458e-01 -1.45423323e-01
9.83585954e-01 4.04034287e-01 5.39337695e-01 6.78043723e-01
2.49244809e-01 9.71534669e-01 3.37047964e-01 -1.01713398e-02
-4.53491151e-01 -8.15061688e-01 5.77908337e-01 1.32973754e+00
-3.31871241e-01 -3.16839397e-01 -5.08506656e-01 3.42690200e-01
-1.75909162e+00 -1.06814158e+00 1.85293168e-01 1.76630425e+00
1.02905333e+00 3.79535444e-02 4.97631103e-01 2.69029021e-01
6.74674213e-01 3.95735919e-01 -6.89894319e-01 1.96848065e-01
-2.13028714e-02 2.36948803e-01 -6.49634823e-02 4.03619468e-01
-8.96698356e-01 7.58038759e-01 5.05134487e+00 1.38377380e+00
-1.22012269e+00 3.37231189e-01 3.90049487e-01 -3.02368730e-01
-5.76354563e-01 -3.98736984e-01 -3.18368554e-01 5.70381939e-01
6.28396928e-01 -1.06998570e-02 6.16183996e-01 5.07955313e-01
3.95183563e-01 4.27047968e-01 -8.71861756e-01 1.07910132e+00
-1.64364547e-01 -9.48811769e-01 -4.95265536e-02 7.05210567e-02
5.62008679e-01 -3.27734739e-01 2.79588491e-01 2.06296667e-01
4.96827513e-01 -1.02612734e+00 1.09117460e+00 5.49934626e-01
1.18354213e+00 -8.31330419e-01 3.39453042e-01 4.27018672e-01
-1.44053030e+00 8.83670375e-02 3.87250274e-01 2.07487985e-01
4.52651203e-01 2.91943967e-01 -4.60846782e-01 6.25335932e-01
4.58745688e-01 6.38661861e-01 8.50861222e-02 5.63171864e-01
-4.79610592e-01 7.75495112e-01 -3.65253001e-01 -7.24068359e-02
3.38009298e-02 -1.04472287e-01 9.95051324e-01 1.00952363e+00
4.62607205e-01 3.77320111e-01 8.15968812e-02 8.49962294e-01
1.13323294e-01 -2.85537150e-02 -4.50788468e-01 -1.61562428e-01
8.14672172e-01 8.54114950e-01 -5.09568453e-01 -9.19893291e-03
-3.80300509e-04 1.29906523e+00 -1.52287886e-01 4.55527395e-01
-1.07150996e+00 -2.21270114e-01 8.51622999e-01 -1.13409638e-01
2.52461582e-01 -3.41394007e-01 -1.42177284e-01 -1.25470805e+00
1.14805944e-01 -1.24780786e+00 1.10459864e-01 -4.45153803e-01
-1.21997809e+00 7.20794797e-01 -1.66704729e-01 -1.50088358e+00
-6.48463428e-01 1.21044733e-01 -8.07455063e-01 7.01635897e-01
-8.89884710e-01 -1.45196867e+00 -3.54736596e-01 8.43244374e-01
7.39441991e-01 -5.68255745e-02 7.45391309e-01 4.24177408e-01
-5.92145741e-01 9.89771783e-01 -7.44684637e-02 1.32116035e-01
6.17933989e-01 -8.07314932e-01 3.52511525e-01 7.45755553e-01
3.19696516e-01 4.53633517e-01 7.95599163e-01 -5.71976781e-01
-1.42334223e+00 -9.29250419e-01 1.69177264e-01 -3.50760549e-01
5.70304513e-01 -3.22745055e-01 -8.10674548e-01 4.29725796e-01
9.29788351e-02 -2.82006472e-01 5.08330643e-01 -1.61876097e-01
-3.39972943e-01 -2.18771454e-02 -8.02548707e-01 9.21040118e-01
1.37980044e+00 -7.20623732e-01 -3.38327527e-01 -1.29249975e-01
8.78370225e-01 -5.66692114e-01 -7.18127370e-01 2.33482748e-01
8.51695001e-01 -9.34085846e-01 9.72787559e-01 -1.06882811e-01
4.53088284e-01 -1.91043898e-01 -4.43740226e-02 -1.31691372e+00
-9.01499018e-03 -1.29100084e+00 -3.56935471e-01 1.62467766e+00
1.33080542e-01 -4.49210703e-01 7.42064893e-01 2.00909749e-01
-3.40637118e-02 -6.30192637e-01 -9.47669983e-01 -7.91869581e-01
-1.16531484e-01 -6.48210227e-01 7.24465251e-01 9.26872015e-01
-1.38881141e-02 3.51767868e-01 -8.60177934e-01 1.63400963e-01
5.91349781e-01 -1.52529433e-01 9.90000725e-01 -8.05385828e-01
-8.28012943e-01 -5.63884139e-01 -2.38840312e-01 -1.41908085e+00
-3.98420915e-03 -5.60429454e-01 3.50176364e-01 -1.00763547e+00
-7.85964355e-03 -5.58434546e-01 -1.20546140e-01 3.38456601e-01
-2.71733224e-01 3.10263216e-01 3.44829202e-01 4.37027335e-01
-3.89237195e-01 1.20987332e+00 1.52307916e+00 -1.61259666e-01
-5.33302069e-01 1.53217375e-01 -4.38229710e-01 5.24173975e-01
6.04888141e-01 -4.43656832e-01 -5.20984769e-01 -2.90355712e-01
-3.57103795e-01 6.04520142e-01 4.35587376e-01 -1.00362945e+00
7.52139241e-02 -1.67640552e-01 1.17643923e-01 -1.66222990e-01
6.15082264e-01 -5.81662893e-01 6.57435238e-01 4.47204590e-01
-3.61676008e-01 -2.27873266e-01 -1.17838010e-01 8.40651512e-01
-3.85041058e-01 4.16149229e-01 5.64231932e-01 1.53032809e-01
-2.09321171e-01 5.15652597e-01 -2.74684221e-01 -5.23351831e-03
8.47212374e-01 4.79382724e-02 1.44010812e-01 -9.34039712e-01
-8.40040565e-01 8.86947289e-02 3.73627484e-01 5.55544317e-01
5.99957228e-01 -1.84088719e+00 -7.70530522e-01 2.17404112e-01
-1.14328079e-01 6.88599646e-02 5.84250033e-01 9.35443759e-01
-2.26186395e-01 -1.03503577e-01 -8.18559751e-02 -6.47683322e-01
-1.27436340e+00 2.11416855e-01 1.28971100e-01 -2.02294633e-01
-8.31656039e-01 1.00038815e+00 1.08497471e-01 -1.99531853e-01
4.32880461e-01 -1.56059265e-01 1.09700160e-02 -2.93792635e-02
4.95018214e-01 2.59151965e-01 -3.82374436e-01 -7.07213521e-01
-1.83113754e-01 4.23411846e-01 4.41711307e-01 -7.15215266e-01
1.19171190e+00 -1.53746188e-01 3.72171789e-01 4.79266852e-01
9.91921067e-01 2.94381112e-01 -1.77943075e+00 -1.44559041e-01
-5.45872450e-01 -3.18936944e-01 -1.69015244e-01 -5.11431396e-01
-1.31333780e+00 8.40068042e-01 5.06989181e-01 4.99185398e-02
1.38503826e+00 -4.20902371e-02 1.18557870e+00 -2.27461383e-01
2.97854722e-01 -6.60560787e-01 4.79951978e-01 2.01261640e-01
9.96512890e-01 -7.31199265e-01 -5.09581447e-01 -2.87529320e-01
-8.34086359e-01 8.36775541e-01 4.44845557e-01 7.34849041e-03
5.67392468e-01 5.41307032e-01 3.19771975e-01 7.62868375e-02
-6.06164038e-01 -7.24756271e-02 1.71157613e-01 8.81378651e-01
1.94580212e-01 1.98371232e-01 -1.82712525e-01 9.68244255e-01
-3.44919682e-01 -8.71250778e-03 -1.13915987e-02 6.96506917e-01
5.81387505e-02 -1.26698279e+00 -4.07481343e-01 -4.07827944e-02
-2.82485664e-01 -8.37793425e-02 -3.61464053e-01 6.49783850e-01
-4.62122001e-02 9.03966784e-01 -1.90753728e-01 -9.70985353e-01
2.17218980e-01 3.30183357e-02 5.03983915e-01 -1.03125572e-01
-5.04951775e-01 7.23240018e-01 -3.35537605e-02 -5.59921920e-01
-5.16836464e-01 -7.93786407e-01 -1.23999524e+00 -4.47719455e-01
-2.84760296e-01 -6.07285388e-02 3.56929153e-01 8.05766761e-01
3.37350547e-01 7.92613924e-01 7.74401486e-01 -1.22169340e+00
-6.27461255e-01 -1.22234952e+00 -5.82192421e-01 3.83361995e-01
4.70688134e-01 -5.37722647e-01 -3.29508603e-01 8.68140981e-02] | [5.738527297973633, -0.19869668781757355] |
23df0298-f293-4a53-aee8-bb9b1542a2c0 | cross-task-transfer-for-multimodal-aerial | 2005.08449 | null | https://arxiv.org/abs/2005.08449v2 | https://arxiv.org/pdf/2005.08449v2.pdf | Cross-Task Transfer for Geotagged Audiovisual Aerial Scene Recognition | Aerial scene recognition is a fundamental task in remote sensing and has recently received increased interest. While the visual information from overhead images with powerful models and efficient algorithms yields considerable performance on scene recognition, it still suffers from the variation of ground objects, lighting conditions etc. Inspired by the multi-channel perception theory in cognition science, in this paper, for improving the performance on the aerial scene recognition, we explore a novel audiovisual aerial scene recognition task using both images and sounds as input. Based on an observation that some specific sound events are more likely to be heard at a given geographic location, we propose to exploit the knowledge from the sound events to improve the performance on the aerial scene recognition. For this purpose, we have constructed a new dataset named AuDio Visual Aerial sceNe reCognition datasEt (ADVANCE). With the help of this dataset, we evaluate three proposed approaches for transferring the sound event knowledge to the aerial scene recognition task in a multimodal learning framework, and show the benefit of exploiting the audio information for the aerial scene recognition. The source code is publicly available for reproducibility purposes. | ['Xuhong LI', 'Xiaoxiang Zhu', 'Liping Jing', 'Lichao Mou', 'Dong Chen', 'Pu Jin', 'Di Hu', 'Dejing Dou'] | 2020-05-18 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/4513_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123690069.pdf | eccv-2020-8 | ['scene-recognition'] | ['computer-vision'] | [ 7.52272248e-01 -6.57180071e-01 2.56428719e-01 -3.68478209e-01
-5.33660471e-01 -6.21568918e-01 5.80047488e-01 3.68167996e-01
-3.14856708e-01 4.46239114e-01 2.66098708e-01 4.40750495e-02
-2.96867341e-01 -8.69930804e-01 -5.17842650e-01 -8.61828983e-01
2.09399208e-01 -4.86261487e-01 1.01054788e-01 -2.04600524e-02
3.48924011e-01 4.71435606e-01 -2.13149858e+00 5.33752084e-01
6.70633972e-01 1.06352031e+00 6.60055578e-01 8.54651630e-01
-2.21405461e-01 8.46151352e-01 -4.78929102e-01 2.86751062e-01
1.70353875e-01 -4.98488456e-01 -6.64080679e-01 3.62811804e-01
3.35745543e-01 -1.21550426e-01 -1.99602664e-01 1.09238148e+00
5.36299646e-01 2.98204452e-01 6.37275219e-01 -1.18902659e+00
-6.54675663e-01 4.93251473e-01 -3.31698686e-01 3.43389481e-01
5.10068715e-01 -7.71229938e-02 9.16190386e-01 -9.70958114e-01
4.46503498e-02 1.12912226e+00 2.31647298e-01 -3.04877907e-02
-6.69498563e-01 -4.53891367e-01 5.06954610e-01 4.56892937e-01
-1.74066472e+00 -5.12149572e-01 9.58957911e-01 -4.99737740e-01
4.02904987e-01 4.56572622e-01 5.85096836e-01 7.37183213e-01
-1.66742310e-01 8.29092562e-01 1.08234859e+00 -7.05904365e-01
1.29047975e-01 2.15121567e-01 8.04273039e-02 4.64964151e-01
-6.23754337e-02 -1.35513246e-01 -5.42399883e-01 -6.45006523e-02
3.91701043e-01 4.91160750e-01 -5.97650945e-01 -9.18881968e-02
-1.05133104e+00 6.45834923e-01 6.51724279e-01 5.08640468e-01
-4.54334378e-01 -1.21735111e-01 2.04616114e-01 1.08538508e-01
1.11813463e-01 5.59069999e-02 -1.44388273e-01 2.54158080e-01
-8.32277536e-01 -1.75734594e-01 4.81916547e-01 5.37705481e-01
8.63219619e-01 1.18065841e-01 7.78795173e-03 1.00062716e+00
4.76925850e-01 1.01024210e+00 4.15599257e-01 -3.76939565e-01
2.89788753e-01 6.73200071e-01 5.60634546e-02 -1.51835072e+00
-2.77513623e-01 -3.41601580e-01 -9.52691913e-01 -5.60724875e-04
-5.76170720e-02 3.09625510e-02 -8.91637921e-01 1.36281490e+00
2.99504936e-01 3.64398807e-01 4.45693403e-01 1.04560673e+00
1.25615799e+00 1.20891249e+00 2.21889287e-01 -1.50631413e-01
1.23583102e+00 -6.27514005e-01 -5.05066872e-01 -1.22544974e-01
9.89333093e-02 -1.00188041e+00 9.34987843e-01 4.40337628e-01
-3.16021860e-01 -9.79355156e-01 -9.96256053e-01 3.41592193e-01
-5.89442194e-01 7.04852521e-01 5.44467390e-01 4.80702609e-01
-8.60149205e-01 3.88025120e-03 -4.63962346e-01 -7.90698528e-01
2.15291560e-01 -4.23592106e-02 -2.64582336e-01 -3.63794357e-01
-9.99936283e-01 4.27647948e-01 5.33941746e-01 4.75061119e-01
-1.13837337e+00 -1.88010812e-01 -7.97661781e-01 5.96389696e-02
4.10980642e-01 -3.19774956e-01 7.88440824e-01 -1.16444027e+00
-1.37724221e+00 5.36878824e-01 -9.38429609e-02 2.63188267e-03
2.79610395e-03 -7.63196647e-02 -6.74346089e-01 3.44241619e-01
3.63330469e-02 3.74605030e-01 1.00816882e+00 -1.46126747e+00
-7.03212142e-01 -4.15528119e-01 6.88529760e-02 4.61266994e-01
-3.06695670e-01 7.51014799e-02 -1.60505041e-01 -4.55648810e-01
8.22633877e-02 -7.77414680e-01 6.19930476e-02 -1.49038732e-01
-2.78002918e-01 1.19511701e-01 9.20704484e-01 -5.89273989e-01
8.57577860e-01 -2.47270656e+00 1.11205839e-02 2.80718625e-01
-3.89728069e-01 3.03294182e-01 -2.79749423e-01 5.67729175e-01
1.05064817e-01 -1.16950199e-01 -5.67396343e-01 1.66505396e-01
-2.88619965e-01 1.11096464e-01 -5.61305225e-01 2.06923515e-01
2.04364717e-01 2.61086732e-01 -7.91168928e-01 -5.27130425e-01
4.63765353e-01 4.67463195e-01 -2.86033422e-01 3.43084365e-01
3.34727988e-02 5.04911363e-01 -7.54980028e-01 9.37698364e-01
6.94576681e-01 -1.93850864e-02 -5.79834618e-02 -2.43112653e-01
-3.34441423e-01 -4.23345327e-01 -1.45744252e+00 1.64472377e+00
-4.65702385e-01 4.80079442e-01 3.06041501e-02 -9.58436072e-01
1.22143698e+00 4.13854927e-01 3.68821770e-01 -3.69726747e-01
1.32540494e-01 -2.22641286e-02 -2.92626798e-01 -8.67236853e-01
6.21275902e-01 -1.43935159e-01 9.56229120e-02 -2.84025874e-02
1.01664603e-01 -3.52035463e-01 -1.01862505e-01 -3.80895846e-02
6.22164190e-01 -1.05196819e-01 4.93863195e-01 1.63576648e-01
8.77546370e-01 1.15958095e-01 1.55306116e-01 6.22809291e-01
1.27802556e-02 4.46332514e-01 -2.34597459e-01 -3.44540000e-01
-4.54902649e-01 -8.89579356e-01 -2.37172276e-01 1.15335572e+00
2.81005710e-01 -3.53784978e-01 -2.82359898e-01 -2.82330871e-01
-2.04808369e-01 4.64596272e-01 -3.74649823e-01 -1.45125240e-02
2.05191001e-01 -9.52316344e-01 6.85441732e-01 2.73893088e-01
9.37681437e-01 -1.08527613e+00 -7.14229047e-01 2.09619179e-02
-3.65086198e-01 -1.01231885e+00 8.77902657e-02 -7.43473545e-02
-4.42755997e-01 -1.11855805e+00 -8.05420518e-01 -6.16040945e-01
4.30725753e-01 8.24451506e-01 5.75905323e-01 -2.44932598e-03
-6.04145348e-01 1.01249850e+00 -6.68765545e-01 -6.43368304e-01
-1.67414650e-01 -2.32947662e-01 -6.83530197e-02 7.58547723e-01
1.97788611e-01 -5.10635793e-01 -4.13641185e-01 2.56973267e-01
-1.18091762e+00 -5.98149784e-02 6.75836980e-01 5.19006968e-01
5.57993650e-01 5.22153795e-01 2.65724450e-01 -2.72440553e-01
3.20377499e-01 -5.63341677e-01 -4.62335467e-01 3.75201106e-01
1.44156352e-01 -4.52860177e-01 5.19967318e-01 -1.83087811e-01
-1.20439243e+00 4.15368348e-01 1.09007239e-01 -3.20028096e-01
-8.04813147e-01 8.60243201e-01 -4.15882528e-01 -2.65222371e-01
5.66885233e-01 5.70554435e-01 -3.60064566e-01 -4.08047259e-01
2.49416605e-01 9.99353588e-01 3.39620262e-01 -2.44890645e-01
5.22064686e-01 4.40338075e-01 7.77313039e-02 -1.54141676e+00
-9.63400722e-01 -7.80702949e-01 -5.94736278e-01 -6.19265258e-01
1.02540743e+00 -1.25509584e+00 -5.96875012e-01 4.52718675e-01
-1.08697605e+00 1.49273068e-01 1.55064881e-01 7.75074422e-01
-1.64894775e-01 5.41276634e-01 1.25021279e-01 -1.23110116e+00
1.06961103e-02 -9.86650109e-01 1.16468716e+00 3.36729825e-01
3.36723030e-01 -6.66110158e-01 1.80530623e-01 7.02296570e-02
2.02844054e-01 1.91517442e-01 7.38867342e-01 -3.32649380e-01
-7.10225523e-01 -6.37502521e-02 -3.67102921e-01 4.67438817e-01
4.87008572e-01 2.06020430e-01 -1.38794661e+00 -2.06829514e-02
5.39785847e-02 -2.95658588e-01 9.73458350e-01 3.20664525e-01
9.89507496e-01 -3.22097689e-02 -1.20084040e-01 4.02016610e-01
1.71974254e+00 4.12972540e-01 4.68718499e-01 8.06569755e-02
7.80748308e-01 6.85055435e-01 8.20142865e-01 7.76481628e-01
1.43798023e-01 4.78129238e-01 5.67419589e-01 -1.07495882e-01
8.80435184e-02 -2.67291784e-01 3.92565936e-01 8.04258287e-01
-2.85950303e-01 -4.63912010e-01 -9.72075999e-01 3.89382690e-01
-1.73095477e+00 -1.08225274e+00 8.88918247e-03 2.01737237e+00
2.27430969e-01 -5.46393573e-01 -2.14258105e-01 3.65141898e-01
8.08771729e-01 5.15904844e-01 -1.87585011e-01 1.19602166e-01
-4.10999060e-01 -1.72160998e-01 2.97059029e-01 3.50049227e-01
-1.35520840e+00 8.35712194e-01 5.97634506e+00 6.26726806e-01
-1.36576247e+00 -1.65442705e-01 5.29492274e-02 2.72257894e-01
-5.69904922e-03 -8.30798075e-02 -5.58745921e-01 2.13822931e-01
4.63968515e-01 -1.13535589e-02 4.50661093e-01 6.27589166e-01
2.19990447e-01 -3.26792359e-01 -6.38973534e-01 1.40572596e+00
5.21205306e-01 -8.23864222e-01 6.71170652e-01 -3.57750237e-01
4.06086028e-01 -1.63922563e-01 2.79542077e-02 -3.51097435e-02
-8.87864083e-03 -8.09528828e-01 6.31141901e-01 8.02973390e-01
4.94232237e-01 -5.87393224e-01 6.89956665e-01 3.37275207e-01
-1.68150604e+00 -2.96065062e-01 -6.17219925e-01 -1.38187647e-01
-1.15986370e-01 3.81826103e-01 -7.62899637e-01 1.13377035e+00
8.19156826e-01 1.10849071e+00 -8.76329839e-01 1.36377525e+00
-1.56925946e-01 6.20116472e-01 -2.61873662e-01 -8.86864588e-02
2.71452159e-01 -1.89307243e-01 5.89565158e-01 1.11717486e+00
7.46099830e-01 3.56732279e-01 5.21375597e-01 6.35795534e-01
2.80505240e-01 5.00455141e-01 -1.18741882e+00 -3.20852548e-01
1.49778605e-01 1.28442764e+00 -6.58809483e-01 -1.52373552e-01
-3.26826423e-01 8.70648921e-01 -3.75638634e-01 4.16688532e-01
-5.95939219e-01 -4.87314671e-01 1.39304549e-01 -5.05813956e-01
2.57410109e-01 -3.06399316e-01 2.43555978e-01 -1.01975238e+00
-3.05414535e-02 -6.97581947e-01 4.57747757e-01 -1.29171908e+00
-9.92455184e-01 7.07390130e-01 -1.23710781e-02 -1.52653527e+00
3.14532638e-01 -6.56719804e-01 -5.51780999e-01 7.08557725e-01
-1.79341245e+00 -1.34267664e+00 -8.43041837e-01 9.88581598e-01
5.37041903e-01 -2.96035558e-01 9.38165188e-01 2.72442073e-01
-2.68474400e-01 -3.45728882e-02 -1.67343318e-02 5.22683710e-02
7.25358784e-01 -7.47006953e-01 -4.58681881e-01 1.04220521e+00
3.98957372e-01 1.14378877e-01 5.82356453e-01 -4.26777214e-01
-1.45982194e+00 -1.32524920e+00 6.16920829e-01 -6.70350790e-02
4.09937561e-01 1.59802228e-01 -7.52334952e-01 4.46649283e-01
1.84075311e-01 -1.74065620e-01 1.05209410e+00 -1.13318950e-01
-1.64494678e-01 -2.92306453e-01 -8.36257875e-01 3.02140862e-01
7.37391651e-01 -7.98221052e-01 -6.53876603e-01 2.71875799e-01
4.36104685e-01 1.49695471e-01 -4.66178894e-01 3.73641908e-01
2.29605973e-01 -7.96194077e-01 1.01036513e+00 -2.59469926e-01
2.89701134e-01 -8.14973235e-01 -9.17897463e-01 -1.32325757e+00
-3.11697483e-01 -1.91774629e-02 7.09833860e-01 1.40053833e+00
1.69858143e-01 -3.47945392e-01 7.38960579e-02 -2.50789877e-02
-8.76492858e-02 2.53579080e-01 -6.85600936e-01 -4.74431574e-01
-6.04244173e-01 -5.18439233e-01 6.09587848e-01 8.50622237e-01
-1.52077019e-01 3.07672650e-01 -6.20501339e-01 9.04531896e-01
3.60279441e-01 4.83806431e-01 7.69687235e-01 -1.47602892e+00
-1.72674373e-01 6.77555725e-02 -6.79419100e-01 -8.93062174e-01
2.79698372e-02 -9.57266986e-01 3.07103038e-01 -1.55681419e+00
2.56563842e-01 -9.12075192e-02 -5.49258053e-01 5.17681777e-01
-2.11947151e-02 3.20375055e-01 4.93152976e-01 1.98840201e-01
-7.57904291e-01 8.43369007e-01 1.01590228e+00 -4.63783920e-01
-2.68575430e-01 2.39141770e-02 -5.41642785e-01 7.44056106e-01
6.50050819e-01 -2.82311320e-01 -4.37481254e-01 -4.91814405e-01
7.16455132e-02 1.79244995e-01 8.40709925e-01 -1.27311754e+00
3.47818702e-01 -1.80974379e-01 3.90353471e-01 -7.08963454e-01
4.55128133e-01 -1.15193534e+00 2.27608621e-01 2.24236250e-01
-2.38806292e-01 -1.63233414e-01 4.20068711e-01 8.71940970e-01
-6.63723767e-01 -2.15025127e-01 5.56811094e-01 -2.62363493e-01
-1.05656219e+00 1.25981256e-01 -7.26441443e-01 -5.02409756e-01
8.67369056e-01 -1.08091570e-01 -1.32151991e-01 -3.64457309e-01
-6.56719685e-01 -6.31638989e-02 2.47861873e-02 4.43132907e-01
7.86122859e-01 -1.26223743e+00 -6.22967720e-01 3.39005500e-01
4.85037893e-01 -3.47990811e-01 7.95248747e-01 6.29069269e-01
-3.01283479e-01 4.34301764e-01 -4.33405071e-01 -7.38682091e-01
-1.63936913e+00 5.15477896e-01 1.88217327e-01 4.94702041e-01
-3.06921929e-01 5.53345382e-01 2.98566222e-01 -3.07584465e-01
1.90714955e-01 -4.06371951e-01 -8.25856566e-01 1.29786044e-01
6.66757405e-01 3.33842993e-01 -1.87057316e-01 -8.67467999e-01
-3.52922380e-01 9.86731350e-01 4.96939868e-01 -9.33027342e-02
1.18138576e+00 -4.40899402e-01 -9.51440781e-02 9.81341064e-01
8.34852278e-01 1.78436756e-01 -6.47015393e-01 -4.57049221e-01
-3.11048478e-01 -6.01837993e-01 2.48735890e-01 -8.76882911e-01
-8.18044126e-01 1.09256721e+00 8.22077215e-01 4.39882249e-01
1.40988505e+00 -2.52230495e-01 1.89134434e-01 7.15180457e-01
3.94668549e-01 -7.74828970e-01 -7.97601864e-02 4.74916756e-01
1.00924659e+00 -1.35805309e+00 -1.02883838e-01 -4.97350603e-01
-8.01886022e-01 1.30856776e+00 1.36863425e-01 1.72771171e-01
8.03872585e-01 -1.80584714e-01 3.21700126e-01 -1.33700967e-01
-3.38823199e-01 -5.91940284e-01 3.05713177e-01 6.10414565e-01
1.56840310e-01 1.33322686e-01 3.39991093e-01 4.81653720e-01
1.91670265e-02 9.04468670e-02 3.73705059e-01 8.36833894e-01
-8.79610062e-01 -7.79781640e-01 -6.29716098e-01 -2.60392912e-02
-2.50418037e-01 6.68688584e-03 -6.88798964e-01 3.70599955e-01
2.76048034e-01 1.42790091e+00 -2.37012610e-01 -4.01777506e-01
4.02621657e-01 1.76791936e-01 2.81633526e-01 -5.69301963e-01
-1.88832402e-01 3.22841369e-02 -1.73385337e-01 -2.32660949e-01
-1.12193561e+00 -5.47806382e-01 -9.15806472e-01 1.20096289e-01
-1.99909225e-01 2.20095947e-01 6.67437315e-01 8.64609361e-01
2.47125626e-01 6.93569422e-01 7.76740372e-01 -1.01212740e+00
-3.52161564e-02 -8.45375180e-01 -9.17462349e-01 9.51152220e-02
3.76864314e-01 -6.80955291e-01 -5.23048937e-01 3.33577096e-01] | [9.796740531921387, 1.7832187414169312] |
d2469126-0801-4739-80bf-625f437d81a2 | cell-nuclei-classification-in | 2202.10177 | null | https://arxiv.org/abs/2202.10177v1 | https://arxiv.org/pdf/2202.10177v1.pdf | Cell nuclei classification in histopathological images using hybrid OLConvNet | Computer-aided histopathological image analysis for cancer detection is a major research challenge in the medical domain. Automatic detection and classification of nuclei for cancer diagnosis impose a lot of challenges in developing state of the art algorithms due to the heterogeneity of cell nuclei and data set variability. Recently, a multitude of classification algorithms has used complex deep learning models for their dataset. However, most of these methods are rigid and their architectural arrangement suffers from inflexibility and non-interpretability. In this research article, we have proposed a hybrid and flexible deep learning architecture OLConvNet that integrates the interpretability of traditional object-level features and generalization of deep learning features by using a shallower Convolutional Neural Network (CNN) named as $CNN_{3L}$. $CNN_{3L}$ reduces the training time by training fewer parameters and hence eliminating space constraints imposed by deeper algorithms. We used F1-score and multiclass Area Under the Curve (AUC) performance parameters to compare the results. To further strengthen the viability of our architectural approach, we tested our proposed methodology with state of the art deep learning architectures AlexNet, VGG16, VGG19, ResNet50, InceptionV3, and DenseNet121 as backbone networks. After a comprehensive analysis of classification results from all four architectures, we observed that our proposed model works well and perform better than contemporary complex algorithms. | ['Satish Kumar Singh', 'Suvidha Tripathi'] | 2022-02-21 | null | null | null | null | ['nuclei-classification'] | ['medical'] | [-1.00133851e-01 2.42827669e-01 -1.48919418e-01 -2.28684992e-01
-2.86486477e-01 -3.78925443e-01 3.43938917e-01 3.91310036e-01
-7.77199149e-01 8.57572615e-01 -1.76604196e-01 -6.45130455e-01
-3.27733606e-01 -8.46001923e-01 -3.48780274e-01 -8.82580340e-01
-9.69069079e-02 3.12652141e-01 2.94489503e-01 -2.63151288e-01
1.30690426e-01 7.98875690e-01 -1.31765592e+00 4.67122883e-01
7.49195099e-01 1.20119798e+00 7.41551965e-02 7.07302570e-01
-1.09152861e-01 6.16025865e-01 -5.37643373e-01 -3.57601792e-01
-4.10595536e-02 -1.18321411e-01 -8.85594964e-01 -7.61682615e-02
2.26726055e-01 -4.74525243e-02 -3.65236968e-01 1.08168578e+00
6.58338308e-01 -3.45748037e-01 7.19646454e-01 -1.03049767e+00
-4.87067461e-01 4.13782328e-01 -3.48820686e-01 5.17203987e-01
-2.71153450e-01 3.71680558e-01 7.83200681e-01 -7.29849279e-01
6.05591774e-01 8.15472782e-01 1.06510067e+00 5.45542240e-01
-9.03761864e-01 -6.95988238e-01 -1.30765498e-01 3.23279589e-01
-1.51611149e+00 -1.37514099e-01 4.08619463e-01 -3.49706084e-01
9.98680592e-01 4.54154819e-01 9.00560975e-01 8.67391706e-01
6.35544717e-01 6.22290909e-01 9.98988271e-01 -2.62282997e-01
1.60887882e-01 2.10010037e-01 3.53647798e-01 1.08351743e+00
6.95299864e-01 -1.22186914e-01 1.54029429e-01 9.68676582e-02
7.75645435e-01 3.56102884e-01 -1.82967976e-01 7.17521012e-02
-8.54326189e-01 9.84426081e-01 8.97728264e-01 6.86753809e-01
-9.75824967e-02 2.52925128e-01 7.26841688e-01 1.30075499e-01
2.75154054e-01 4.08502609e-01 -3.43821853e-01 3.79763275e-01
-7.73756325e-01 3.51505727e-02 4.26488578e-01 6.14843965e-01
4.97273177e-01 1.10674482e-02 -1.58184499e-01 6.45722568e-01
3.23152840e-01 2.93054916e-02 9.85687494e-01 -4.44400460e-01
-1.18529730e-01 1.17608547e+00 -5.17954350e-01 -9.00404453e-01
-9.92072761e-01 -9.63129759e-01 -1.43328881e+00 2.93213636e-01
4.43960249e-01 4.08146307e-02 -1.30279195e+00 1.34069133e+00
-4.01968583e-02 -9.23562273e-02 9.26016197e-02 6.49917185e-01
1.26245689e+00 2.24624261e-01 2.97160804e-01 2.36869141e-01
1.43340945e+00 -9.58102167e-01 -5.19852579e-01 5.71800545e-02
1.21446335e+00 -4.87383634e-01 8.58997524e-01 2.23301083e-01
-7.75024056e-01 -4.22210842e-01 -1.19643664e+00 -2.53510684e-01
-7.81063557e-01 3.36462528e-01 9.65070784e-01 5.85371554e-01
-1.33630705e+00 4.90100443e-01 -9.28689420e-01 -5.96340299e-01
8.26221585e-01 8.96886110e-01 -5.87226152e-01 5.90904132e-02
-1.01306665e+00 8.22569728e-01 7.44355738e-01 1.95729956e-01
-8.01850975e-01 -5.48158228e-01 -6.83304489e-01 2.25731775e-01
-1.01089239e-01 -6.62495196e-01 9.64478612e-01 -1.02536845e+00
-1.27832389e+00 1.09670722e+00 4.77723300e-01 -4.66510206e-01
4.20561552e-01 3.35814953e-01 -2.46998996e-01 1.77252159e-01
-3.21916044e-01 8.70790064e-01 1.12410948e-01 -7.65756726e-01
-5.11397481e-01 -4.47915912e-01 1.35214984e-01 -7.15450943e-02
-4.52115685e-01 -4.50431675e-01 -3.52652907e-01 -5.26779175e-01
2.02764586e-01 -9.21029627e-01 -4.54414964e-01 2.36836284e-01
-4.32332724e-01 -2.77717739e-01 9.51989770e-01 -2.69933999e-01
1.19813335e+00 -2.20856881e+00 -1.77727103e-01 2.93448985e-01
5.21196842e-01 6.87639475e-01 -6.93518221e-02 2.87011149e-04
-9.39766988e-02 4.53406096e-01 -9.64579731e-02 -3.41172591e-02
-3.97815317e-01 3.07506472e-01 5.39954901e-01 7.03330398e-01
1.22788139e-01 1.02390754e+00 -6.41889334e-01 -7.86224663e-01
3.46823245e-01 5.13169229e-01 -6.79239392e-01 -7.92365819e-02
-9.96827148e-03 7.80814746e-03 -3.20934057e-01 9.00967360e-01
6.35477722e-01 -6.14433110e-01 2.07420334e-01 -3.67464215e-01
2.83640355e-01 -3.63531888e-01 -7.92829812e-01 1.50604117e+00
-2.35694349e-01 7.10885227e-01 3.50880176e-02 -1.11709893e+00
7.69078016e-01 3.24975044e-01 3.74863893e-01 -6.44805014e-01
5.93430519e-01 4.11142349e-01 4.99444723e-01 -5.11796713e-01
1.85438365e-01 -1.30868152e-01 1.16273761e-01 -9.25053656e-02
3.57009321e-01 1.85034648e-01 8.26939791e-02 -1.19740345e-01
1.22686541e+00 -4.70145196e-01 6.57014072e-01 -5.93941033e-01
8.57459009e-01 2.25634943e-03 3.66201043e-01 5.54649651e-01
-4.43752736e-01 5.67814052e-01 4.92161840e-01 -9.44509089e-01
-9.48620319e-01 -7.89417624e-01 -4.34869468e-01 7.63402939e-01
-9.82090384e-02 -1.62958115e-01 -7.41731346e-01 -7.24217713e-01
-7.79116750e-02 1.81074396e-01 -1.10872626e+00 -1.72638312e-01
-4.40532744e-01 -1.12540472e+00 1.07476556e+00 6.51017725e-01
7.02259183e-01 -1.13362885e+00 -6.43207431e-01 1.33712605e-01
2.17425764e-01 -9.43109632e-01 1.34844169e-01 4.61946964e-01
-1.03778183e+00 -1.37416172e+00 -4.60020453e-01 -8.95340204e-01
1.00618625e+00 -1.01087935e-01 1.10039723e+00 8.13466191e-01
-7.01881945e-01 -2.74119973e-01 -2.66390115e-01 -7.10388303e-01
-4.23688263e-01 6.17664933e-01 -2.76321560e-01 -1.60939425e-01
4.56597388e-01 -2.14386463e-01 -9.19670999e-01 6.61289543e-02
-1.17006791e+00 3.59804593e-02 9.82466877e-01 1.05537820e+00
6.39335155e-01 1.85540579e-02 5.04800081e-01 -1.12335098e+00
4.79571939e-01 -4.77910042e-01 -2.93245018e-01 -3.90120707e-02
-5.28953910e-01 -6.23651817e-02 8.51923823e-01 -9.45801437e-02
-5.12869298e-01 -3.27234454e-02 -4.78792787e-01 -1.53813884e-01
-1.97033226e-01 6.52895689e-01 1.30941302e-01 -3.09446990e-01
7.84897327e-01 8.23980048e-02 3.24075490e-01 -8.82916972e-02
-1.50770798e-01 7.56101251e-01 4.60548460e-01 -2.02726126e-02
2.98343271e-01 5.72023690e-01 3.27884048e-01 -8.30014944e-01
-7.41841555e-01 -3.61438602e-01 -5.05154967e-01 -1.50816485e-01
8.83410513e-01 -7.98844814e-01 -9.02531266e-01 7.35721350e-01
-9.16502595e-01 -1.44520760e-01 -1.16610155e-01 3.56102705e-01
-1.39893681e-01 9.32937190e-02 -7.68755555e-01 -2.13467762e-01
-7.43503094e-01 -1.34373081e+00 7.05887258e-01 4.55915302e-01
-2.80301154e-01 -1.33804131e+00 -1.26826674e-01 2.44490117e-01
7.42763162e-01 7.29269683e-01 1.14971197e+00 -8.93465698e-01
-4.13302600e-01 -4.92019594e-01 -4.31168586e-01 3.15264225e-01
1.65333629e-01 2.14360774e-01 -1.07156575e+00 -5.44940233e-01
-2.31161937e-01 -3.81181091e-01 1.02450335e+00 4.74315137e-01
1.71667027e+00 -2.77327687e-01 -7.13156283e-01 1.06198514e+00
1.76116121e+00 1.99171290e-01 7.86508739e-01 6.52152717e-01
6.13753438e-01 2.05882221e-01 -3.87425572e-02 9.88309458e-02
1.53648153e-01 1.29988179e-01 7.87736833e-01 -6.59921885e-01
-9.32004526e-02 3.08426350e-01 -2.03150913e-01 5.82243681e-01
-2.46528924e-01 -4.97302115e-01 -1.10632443e+00 5.34638405e-01
-1.52472937e+00 -8.11142623e-01 -1.01882331e-01 1.55254197e+00
6.71315968e-01 3.26752722e-01 -3.53864998e-01 5.09279251e-01
4.69564557e-01 -7.96912909e-02 -4.94038373e-01 -5.24730921e-01
-1.15431331e-01 4.19318378e-01 5.92823625e-01 1.06517941e-01
-1.08243597e+00 6.51164711e-01 5.97546721e+00 9.70739186e-01
-1.53550625e+00 7.92160034e-02 1.10072553e+00 3.76478434e-02
6.42996877e-02 -6.31448925e-01 -6.73595428e-01 2.77300090e-01
8.85679662e-01 1.87710986e-01 -2.83269465e-01 7.21288681e-01
-7.55406395e-02 -2.43541468e-02 -1.12996924e+00 8.67130697e-01
-6.73542991e-02 -1.84480226e+00 3.00758243e-01 2.29982853e-01
5.30402303e-01 2.51580358e-01 1.28899157e-01 3.45827520e-01
1.43946216e-01 -1.64047110e+00 1.34029314e-01 3.19438964e-01
8.98476183e-01 -6.93501055e-01 1.48740005e+00 1.88594162e-01
-9.84405279e-01 -2.59235531e-01 -3.01063836e-01 4.90323529e-02
-5.90398610e-01 2.21204191e-01 -1.05069792e+00 4.06073391e-01
7.37391472e-01 4.71468776e-01 -9.03698623e-01 1.02701318e+00
2.68659323e-01 4.83736724e-01 -2.77351588e-01 -3.69872332e-01
6.45346701e-01 3.91732007e-01 6.62636664e-03 1.54272461e+00
2.48878583e-01 1.06766954e-01 7.03679747e-04 5.20132363e-01
-2.14887038e-01 4.69063781e-02 -4.14091706e-01 3.67022236e-03
1.25020027e-01 1.60884881e+00 -1.31440747e+00 -2.68998384e-01
-3.29510450e-01 3.86359453e-01 3.88772637e-01 -3.93065736e-02
-7.96221435e-01 -4.91217911e-01 4.42402601e-01 4.08066124e-01
9.95209962e-02 1.12048529e-01 -5.57178915e-01 -7.84428120e-01
-4.50086802e-01 -7.76246369e-01 6.60820305e-01 -3.65655035e-01
-1.08724451e+00 8.98377001e-01 -2.60238230e-01 -1.19095218e+00
1.55495226e-01 -1.04548764e+00 -6.13638937e-01 5.21033347e-01
-1.40708220e+00 -1.19200003e+00 -8.09522808e-01 5.99400759e-01
5.40568769e-01 -3.98393631e-01 9.27515984e-01 3.25993359e-01
-5.99600911e-01 1.01949310e+00 1.04475833e-01 4.50671405e-01
2.45721698e-01 -1.18994558e+00 -9.70173851e-02 4.35716212e-01
-3.84352505e-01 3.88184935e-01 3.49336505e-01 -1.25714540e-01
-1.07921433e+00 -1.35292888e+00 4.97584552e-01 -1.28386647e-01
5.01290560e-01 -2.49869540e-01 -8.15166235e-01 5.50212085e-01
2.81798571e-01 4.31650907e-01 9.51801360e-01 -2.51150995e-01
8.29517692e-02 -2.89417028e-01 -1.40778565e+00 4.37605828e-01
6.66250348e-01 -1.89688191e-01 -1.15624234e-01 2.41733417e-01
5.00918269e-01 -4.51434314e-01 -9.09556091e-01 6.00713670e-01
4.21666801e-01 -1.05654752e+00 7.90278554e-01 -4.81341302e-01
3.64190847e-01 -2.23909378e-01 -5.07519692e-02 -1.02836657e+00
-6.21091723e-01 7.46654496e-02 1.91061974e-01 6.43707037e-01
5.38830578e-01 -6.87438548e-01 1.12004268e+00 2.03015059e-01
-6.36011362e-01 -1.48847640e+00 -8.90582800e-01 -4.04036611e-01
2.93624461e-01 2.51726285e-02 4.78107154e-01 8.27684700e-01
-2.08232835e-01 2.50227731e-02 1.80012122e-01 1.33116191e-04
3.28444153e-01 -3.04214627e-01 3.50807637e-01 -1.29483414e+00
4.67504896e-02 -7.36682296e-01 -1.05099571e+00 -1.42773226e-01
-1.00119628e-01 -1.12792468e+00 -3.76594126e-01 -1.59368253e+00
3.54981601e-01 -4.50946361e-01 -8.21958423e-01 7.44787037e-01
2.68398132e-02 7.01546371e-01 -7.87497461e-02 3.82290669e-02
-4.20422912e-01 1.14576161e-01 1.37145233e+00 -4.14980203e-01
3.31926756e-02 -2.33494535e-01 -7.01955080e-01 1.01184773e+00
8.41093481e-01 -3.45733494e-01 -4.72952813e-01 -4.20743555e-01
5.02287671e-02 -2.35950768e-01 3.37056935e-01 -1.36869371e+00
3.72493446e-01 1.93287507e-02 9.13307786e-01 -5.70216119e-01
5.71751371e-02 -7.78097928e-01 2.28039846e-01 9.48792398e-01
-1.35671571e-01 1.32066652e-01 4.18544531e-01 1.99101403e-01
-3.84140998e-01 -2.32305482e-01 1.10848284e+00 -3.66398841e-01
-5.89000523e-01 6.09263182e-01 -4.92971480e-01 -2.45465711e-01
1.23001420e+00 -6.64286494e-01 -5.39466739e-01 3.10034901e-01
-7.22252309e-01 1.06381848e-01 3.87262762e-01 1.21414416e-01
6.14920318e-01 -1.10666955e+00 -5.85229754e-01 1.87385783e-01
2.29335561e-01 3.26301843e-01 3.59545916e-01 8.59877050e-01
-1.34868193e+00 6.23093069e-01 -6.65968001e-01 -8.38839829e-01
-1.24420249e+00 2.71245390e-01 8.09904039e-01 -6.92676365e-01
-3.97200733e-01 1.04724574e+00 3.15380841e-01 -3.64655614e-01
1.76738888e-01 -6.74033642e-01 -5.63671887e-01 -1.17271937e-01
2.26249859e-01 2.76871145e-01 3.81935865e-01 -3.83430779e-01
-5.24213970e-01 2.87118465e-01 -4.96045709e-01 6.02026761e-01
1.31501305e+00 3.13078701e-01 -1.48525268e-01 -3.28104496e-02
1.30288947e+00 -6.18793309e-01 -7.85426497e-01 1.18443526e-01
-1.11015283e-01 6.61341250e-02 2.01668411e-01 -9.03328776e-01
-1.44399428e+00 8.59618783e-01 1.06267834e+00 9.94367450e-02
1.31216276e+00 -1.68457180e-01 5.76868057e-01 5.11565268e-01
3.45708840e-02 -9.17494774e-01 9.09424126e-02 5.22266507e-01
4.94205058e-01 -1.28925300e+00 1.65951744e-01 -2.20485657e-01
8.15131515e-03 1.41166198e+00 9.85793769e-01 -1.94017515e-01
7.75746346e-01 3.59626025e-01 2.50555933e-01 -5.71664155e-01
-6.47502661e-01 7.56235719e-02 3.98841426e-02 3.08655888e-01
6.74295068e-01 1.16108336e-01 -4.66897994e-01 7.54963875e-01
-3.20443422e-01 2.92261183e-01 5.81981540e-01 8.94683003e-01
-4.17484909e-01 -7.38663495e-01 6.28353357e-02 6.72266960e-01
-9.38687801e-01 1.68470610e-02 -2.27325812e-01 1.27795732e+00
4.49600160e-01 4.87520218e-01 1.76884517e-01 -3.97907674e-01
1.90983668e-01 -3.16883236e-01 2.53835976e-01 -4.27581578e-01
-8.34370255e-01 -2.80573428e-01 -2.01091602e-01 -2.17609391e-01
-3.65706146e-01 -1.40380159e-01 -1.48942459e+00 -3.99886101e-01
-3.44810605e-01 7.84846768e-02 5.66199422e-01 9.24191535e-01
2.87193179e-01 8.66906345e-01 2.83392727e-01 -5.26083112e-01
-3.64977777e-01 -1.06608939e+00 -5.34605503e-01 2.85941064e-01
3.41904074e-01 -6.12558663e-01 -1.42004803e-01 -4.00664359e-02] | [15.056147575378418, -2.7994613647460938] |
8d84a33b-3231-4e67-89c1-7b4801575451 | from-perception-to-programs-regularize | 2206.05922 | null | https://arxiv.org/abs/2206.05922v2 | https://arxiv.org/pdf/2206.05922v2.pdf | From Perception to Programs: Regularize, Overparameterize, and Amortize | Toward combining inductive reasoning with perception abilities, we develop techniques for neurosymbolic program synthesis where perceptual input is first parsed by neural nets into a low-dimensional interpretable representation, which is then processed by a synthesized program. We explore several techniques for relaxing the problem and jointly learning all modules end-to-end with gradient descent: multitask learning; amortized inference; overparameterization; and a differentiable strategy for penalizing lengthy programs. Collectedly this toolbox improves the stability of gradient-guided program search, and suggests ways of learning both how to perceive input as discrete abstractions, and how to symbolically process those abstractions as programs. | ['Kevin Ellis', 'Hao Tang'] | 2022-06-13 | null | null | null | null | ['program-synthesis'] | ['computer-code'] | [ 4.49920356e-01 6.46390736e-01 -3.94397378e-01 -5.07393360e-01
-6.88182414e-01 -8.05639446e-01 5.89420080e-01 1.14535183e-01
-4.99038130e-01 5.51406085e-01 2.15221524e-01 -7.37892091e-01
-7.51805678e-02 -9.65184152e-01 -1.07205451e+00 -1.64860934e-01
-1.79727420e-01 6.47582650e-01 -2.19027519e-01 1.37548432e-01
2.79371232e-01 4.12024260e-01 -1.59706819e+00 6.45809770e-01
1.08767843e+00 5.84310591e-01 1.01140715e-01 1.16350698e+00
-4.23333585e-01 1.37130558e+00 -4.85037893e-01 -3.84304821e-01
2.27575060e-02 -3.04207921e-01 -1.04527116e+00 -2.21936971e-01
4.65003788e-01 -3.90516907e-01 1.16910502e-01 1.06031144e+00
-1.45528272e-01 2.52312422e-01 7.29912996e-01 -1.27203262e+00
-9.24249291e-01 1.16535723e+00 -9.60716307e-02 -1.73956156e-01
4.25730616e-01 5.71164787e-01 1.07834291e+00 -8.89723539e-01
4.34486091e-01 1.79514873e+00 4.26210493e-01 6.96574867e-01
-1.87771153e+00 -5.04427254e-01 2.33224407e-01 -2.32280478e-01
-7.59297967e-01 -2.72154927e-01 5.78355491e-01 -8.69224846e-01
1.39627826e+00 2.09681943e-01 9.03080821e-01 1.00350261e+00
1.96083918e-01 9.05742228e-01 8.42631578e-01 -5.25667667e-01
3.71611923e-01 4.76840138e-02 3.77667516e-01 1.53533888e+00
-7.23268837e-02 2.86868453e-01 -4.67594862e-01 -5.60506880e-01
9.83185947e-01 -1.39485151e-01 3.08115929e-01 -4.34446514e-01
-1.22364581e+00 8.18556249e-01 3.67067516e-01 -3.17805111e-01
-4.12403829e-02 8.21012735e-01 6.44406438e-01 6.25695705e-01
6.55935425e-03 1.06746745e+00 -3.75417411e-01 -8.34003836e-02
-6.30235374e-01 5.90535343e-01 9.88380611e-01 9.41693306e-01
9.18941319e-01 4.48222578e-01 -1.61127463e-01 6.13650322e-01
1.31658688e-01 3.90878946e-01 2.67621845e-01 -1.49387670e+00
5.66506505e-01 6.56462610e-01 -1.58453614e-01 -6.01397395e-01
-3.54978621e-01 2.40586866e-02 -5.13416290e-01 6.56902790e-01
4.28657144e-01 -5.80826223e-01 -1.00121260e+00 1.97112167e+00
-1.81075349e-01 -5.22390604e-02 2.37398863e-01 4.78685528e-01
6.76452875e-01 7.78667927e-01 4.95745093e-01 3.14138293e-01
1.15107024e+00 -1.17434895e+00 -1.13957122e-01 -5.42126000e-01
8.35162163e-01 -1.07046269e-01 1.52521825e+00 5.93880475e-01
-1.65054095e+00 -7.40826845e-01 -1.00439835e+00 -3.59415948e-01
-3.83505195e-01 2.98767030e-01 1.11456656e+00 6.19327188e-01
-1.20380688e+00 6.19466245e-01 -8.73939991e-01 2.75871515e-01
6.44947708e-01 5.78889310e-01 4.48236689e-02 3.50262165e-01
-8.22722435e-01 9.78595018e-01 7.82195926e-01 -3.11195195e-01
-1.38817418e+00 -9.62622166e-01 -1.00664270e+00 3.62966567e-01
1.51395470e-01 -1.29818666e+00 1.44755208e+00 -1.27601397e+00
-1.73051667e+00 9.54982638e-01 -1.95682645e-01 -5.94622552e-01
1.53722972e-01 -1.23309672e-01 7.75495842e-02 3.22883110e-03
-1.78703874e-01 8.38476241e-01 9.61075068e-01 -1.11536229e+00
-4.79106218e-01 -1.32707909e-01 7.03071415e-01 3.95775467e-01
-2.77845442e-01 -4.13215831e-02 -3.50865833e-02 -5.22673786e-01
3.88745926e-02 -1.00145912e+00 -4.78998572e-01 2.75526285e-01
-5.30445099e-01 -4.27315354e-01 1.44641191e-01 -5.18651307e-01
8.22956860e-01 -2.10099196e+00 8.27198565e-01 2.88737953e-01
5.12026608e-01 -1.75417334e-01 -3.04416567e-01 -6.12175651e-02
-1.99243993e-01 1.98038831e-01 -2.53320515e-01 -2.19814599e-01
3.34903002e-01 1.71408311e-01 -6.37925029e-01 -4.67546917e-02
3.45694900e-01 1.03270447e+00 -1.14188254e+00 -4.62444097e-01
6.47558039e-03 -1.68530196e-01 -1.07918274e+00 5.43023169e-01
-8.14945698e-01 1.42466009e-01 -4.04203475e-01 6.33486509e-01
1.46703511e-01 -8.86682495e-02 1.78826898e-01 -1.14492709e-02
5.07590398e-02 5.08627295e-01 -1.01200998e+00 1.86025560e+00
-9.19443011e-01 6.63897097e-01 1.88885391e-01 -1.09411526e+00
8.03301454e-01 -2.33544055e-02 -8.77114460e-02 -1.83457255e-01
-1.11865073e-01 -9.33547616e-02 -9.96079668e-02 -6.46687508e-01
3.27223390e-01 -2.14003950e-01 -3.52085948e-01 6.06513798e-01
2.62053937e-01 -8.49729180e-01 1.59566194e-01 6.35679439e-02
1.06594825e+00 4.95167792e-01 2.40561683e-02 -2.49797165e-01
2.43424669e-01 3.62142861e-01 2.15597406e-01 1.10119045e+00
2.74149805e-01 -6.04400001e-02 9.78002906e-01 -4.97292548e-01
-1.15167093e+00 -1.53426564e+00 3.02094609e-01 1.94343972e+00
-4.13500339e-01 -2.74482489e-01 -8.86531055e-01 -4.08057123e-01
3.45183343e-01 1.13179076e+00 -5.68065166e-01 -4.33295339e-01
-6.28412724e-01 -2.93529630e-01 9.49928999e-01 8.76911879e-01
2.62030046e-02 -1.37178290e+00 -8.73272955e-01 5.81187382e-02
3.77159625e-01 -4.32561308e-01 -7.36750960e-02 8.39717925e-01
-1.11396301e+00 -8.45347345e-01 -1.96346030e-01 -1.15545905e+00
1.03246772e+00 -3.93165708e-01 1.47575665e+00 1.66501269e-01
-1.18342765e-01 3.32502425e-01 4.25197601e-01 -4.03770834e-01
-7.34053314e-01 -4.60807495e-02 -5.04422337e-02 -7.23635972e-01
1.07398845e-01 -6.91802442e-01 -2.47892719e-02 -4.91062135e-01
-7.35117137e-01 4.09958452e-01 6.89353228e-01 9.16984975e-01
3.27602476e-01 -1.93104595e-01 2.79962689e-01 -1.29929936e+00
1.10307622e+00 -3.45970392e-01 -8.56713176e-01 2.71085501e-01
-2.91233838e-01 5.76176345e-01 9.96200800e-01 -5.35429835e-01
-1.11809015e+00 3.79487932e-01 5.40514216e-02 -3.95306945e-01
-1.36922628e-01 5.77558696e-01 1.65356085e-01 -7.90963769e-02
1.14270294e+00 2.12476090e-01 4.51702364e-02 -1.78215560e-02
9.98627186e-01 8.44633579e-02 8.03665638e-01 -1.53895569e+00
8.83361042e-01 2.73105688e-02 -1.39701933e-01 -5.43813944e-01
-6.14868224e-01 3.97648871e-01 -4.97721970e-01 2.28791580e-01
7.68206000e-01 -9.61307943e-01 -1.14077485e+00 7.12005002e-03
-1.31186581e+00 -9.92558777e-01 -4.51741725e-01 3.96396331e-02
-1.01785064e+00 -3.37294601e-02 -7.71775544e-01 -6.38327718e-01
-3.00148696e-01 -1.51717794e+00 7.62567878e-01 6.57702833e-02
-8.53695631e-01 -1.04701793e+00 -2.94022728e-03 7.87290409e-02
1.79940313e-01 8.86421204e-02 1.85080206e+00 -4.29912359e-01
-5.63319266e-01 2.12483749e-01 -2.50999838e-01 2.21079722e-01
-3.41300964e-01 2.31963560e-01 -7.71755934e-01 1.25299683e-02
-3.71714860e-01 -9.67020392e-01 8.32586288e-01 3.24592471e-01
1.84703481e+00 -6.00300729e-01 -3.21842909e-01 1.02658570e+00
1.18564534e+00 1.96373090e-01 3.31620514e-01 2.34305382e-01
8.85901451e-01 5.35958111e-01 -1.77544169e-02 2.26763755e-01
3.73007119e-01 1.21250562e-02 3.15145433e-01 -3.67257781e-02
5.94821982e-02 -4.64450479e-01 4.44435745e-01 3.27192158e-01
1.48650929e-01 4.59590405e-01 -1.07967818e+00 4.57206219e-01
-1.80507004e+00 -9.03356552e-01 3.91807318e-01 1.78338170e+00
1.28099990e+00 4.83629107e-01 3.01003695e-01 -3.06855530e-01
1.89527497e-01 -7.73889199e-02 -8.84743035e-01 -1.12502503e+00
3.25674713e-01 6.11790478e-01 2.90469646e-01 7.37629950e-01
-8.34620655e-01 1.25379765e+00 8.19856739e+00 3.93819720e-01
-8.75726879e-01 -2.53770053e-01 6.80421829e-01 -1.65667102e-01
-7.68153310e-01 1.76686898e-01 -5.08532405e-01 7.22142980e-02
9.55964804e-01 -5.20946503e-01 1.01066959e+00 1.43391657e+00
-1.15737945e-01 9.49729532e-02 -1.87147307e+00 7.00048685e-01
-2.68918723e-01 -1.64605021e+00 2.31678575e-01 -5.42475522e-01
7.64614820e-01 -2.50945657e-01 2.20383674e-01 8.88595104e-01
1.21009648e+00 -1.26160741e+00 9.29497182e-01 5.34768462e-01
6.71808302e-01 -8.35384011e-01 -3.36545110e-01 3.16854298e-01
-8.59654129e-01 -5.76628745e-01 -3.77099752e-01 -3.62337440e-01
-5.08383930e-01 3.02646250e-01 -6.42338693e-01 -3.60909700e-01
3.76654387e-01 3.62661272e-01 -5.93277037e-01 6.10241175e-01
-6.97607875e-01 3.43189776e-01 -9.56919566e-02 -3.99203181e-01
2.36607030e-01 -1.46620557e-01 4.27830786e-01 1.39600182e+00
3.50398459e-02 3.65553051e-02 4.77507323e-01 1.77944529e+00
-1.58362567e-01 -4.43691254e-01 -8.54125321e-01 -2.65733242e-01
4.84665364e-01 1.09005737e+00 -4.02967304e-01 -8.00241590e-01
-2.30456784e-01 6.89511299e-01 8.38708162e-01 6.58908129e-01
-8.18433762e-01 -6.05904698e-01 7.61268556e-01 -3.21789265e-01
8.06388631e-02 -4.35433000e-01 -1.03538394e+00 -1.21375620e+00
-3.21362495e-01 -1.22245383e+00 2.91844875e-01 -1.04295945e+00
-8.52720916e-01 3.87037665e-01 1.14829041e-01 -4.43867743e-01
-6.32121146e-01 -8.16215098e-01 -9.14686382e-01 1.05028522e+00
-8.81116033e-01 -8.29560459e-01 2.71686800e-02 4.22471821e-01
5.43418467e-01 -4.19555724e-01 1.03425658e+00 -4.02500570e-01
-5.17406702e-01 7.14847088e-01 -3.39715093e-01 5.58171794e-02
1.45159900e-01 -1.54680943e+00 5.66127002e-01 6.55322790e-01
-2.30030656e-01 1.08554113e+00 7.25047827e-01 -4.36688632e-01
-1.94429326e+00 -9.36861396e-01 1.71932071e-01 -4.91098046e-01
1.03142691e+00 -4.79738474e-01 -7.74187922e-01 1.03567553e+00
1.22096471e-01 -4.65662479e-01 3.34149003e-01 5.94185889e-01
-6.96168125e-01 4.68437932e-02 -1.03808391e+00 1.10967112e+00
9.37833011e-01 -9.81252432e-01 -1.06768656e+00 4.41865206e-01
1.00576687e+00 -6.77459478e-01 -7.47808576e-01 -1.10060208e-01
4.80791360e-01 -5.06874204e-01 1.13498247e+00 -9.33783829e-01
9.84605014e-01 2.47638784e-02 1.57299191e-01 -1.40502226e+00
-3.65932912e-01 -6.19625092e-01 -3.95480335e-01 7.35148013e-01
6.48146451e-01 -4.68182683e-01 1.03271413e+00 1.02644062e+00
-2.99790412e-01 -8.96937311e-01 -3.78984541e-01 -4.21070814e-01
5.30497253e-01 -3.96625221e-01 4.88379061e-01 6.79336905e-01
4.61873621e-01 4.36498374e-01 1.29350945e-01 5.08832652e-03
6.70323968e-01 4.29649651e-01 7.13190496e-01 -1.00141430e+00
-8.23997259e-01 -9.93631721e-01 2.14173540e-01 -1.10808921e+00
7.84267783e-01 -1.19865203e+00 2.17981145e-01 -1.17704427e+00
2.44396120e-01 -4.35361475e-01 9.62473825e-02 8.80792260e-01
-5.30397817e-02 -4.95630294e-01 9.70402434e-02 -1.13337219e-01
-2.74257779e-01 1.74356475e-01 1.09052265e+00 -4.78289425e-01
-4.64796066e-01 -2.01712653e-01 -8.10479105e-01 9.93503869e-01
6.13064289e-01 -3.96054447e-01 -6.51124597e-01 -7.93037355e-01
6.21357977e-01 4.58051294e-01 6.57927573e-01 -1.00132596e+00
3.06822598e-01 -5.85837722e-01 5.39569199e-01 -3.42040777e-01
2.51023740e-01 -3.86220932e-01 -1.72456831e-01 6.48856938e-01
-1.20537567e+00 1.73583925e-01 4.90188599e-01 2.48442277e-01
1.36957660e-01 -6.19004428e-01 7.92355001e-01 -6.57671332e-01
-7.78769612e-01 -7.73849562e-02 -8.05329978e-01 5.60516082e-02
6.02093875e-01 -8.92008692e-02 -2.96332300e-01 -1.78511620e-01
-1.07338154e+00 4.50960189e-01 3.78656000e-01 2.09105790e-01
5.97770751e-01 -1.06309271e+00 -2.84631193e-01 2.25772038e-01
-1.85701072e-01 -5.06579764e-02 -2.84770489e-01 1.74557135e-01
-6.06805086e-01 2.00489730e-01 -4.35926259e-01 -3.86500865e-01
-1.05986774e+00 5.96411645e-01 4.25912440e-01 -9.71125960e-02
-3.44683200e-01 1.08924818e+00 2.56151199e-01 -9.68816757e-01
6.59637272e-01 -8.55332315e-01 1.73452899e-01 -3.06504071e-01
3.42642754e-01 1.24449648e-01 -4.39578563e-01 4.13844973e-01
-5.35856299e-02 1.81360826e-01 -1.61004737e-02 -2.45438769e-01
1.30610144e+00 4.96518821e-01 -5.25074720e-01 5.73968530e-01
1.01691377e+00 -3.93550277e-01 -1.36875534e+00 -1.03472829e-01
-8.64369143e-03 -2.64386889e-02 -1.01545811e-01 -8.67123306e-01
-4.87322122e-01 9.60635781e-01 9.46580768e-02 2.00925902e-01
9.81719077e-01 -1.32593885e-01 3.30998629e-01 1.30371404e+00
9.46659297e-02 -9.73996162e-01 5.10986686e-01 8.47165704e-01
8.19348633e-01 -7.66839266e-01 2.91964561e-02 -1.64553255e-01
-4.22288358e-01 1.53350294e+00 8.74920607e-01 -5.87445498e-01
3.50789070e-01 5.70425391e-01 -4.07219291e-01 -2.03162715e-01
-9.40486610e-01 1.33252159e-01 1.67875797e-01 6.95020258e-01
5.49385011e-01 6.79488108e-02 2.87867874e-01 4.57616866e-01
-5.81374764e-01 6.28017783e-02 4.71473783e-01 8.63590419e-01
-5.00945926e-01 -9.13972676e-01 -2.89592594e-01 6.46032691e-01
-1.03901075e-02 -4.06555742e-01 -2.04322532e-01 4.93785828e-01
4.19569351e-02 2.72296518e-01 9.37612578e-02 -2.65213668e-01
2.17119381e-01 2.81242758e-01 8.10332477e-01 -1.04019940e+00
-8.77844930e-01 -4.50918972e-01 2.16937214e-01 -5.98496914e-01
1.80144027e-01 -3.46771359e-01 -1.57665598e+00 -2.73860902e-01
4.04914707e-01 -1.21387020e-01 4.55504447e-01 9.64401364e-01
2.92884946e-01 8.45573008e-01 2.77747303e-01 -1.01316941e+00
-8.24078619e-01 -3.31969261e-01 1.76094286e-02 2.43434906e-01
4.53878492e-01 -3.62318039e-01 -2.78049499e-01 2.89520681e-01] | [8.75537395477295, 7.16160249710083] |
297aae60-2ac7-4910-8ce9-5c8b02fc2b1c | semi-supervised-learning-made-simple-with-1 | 2306.07483 | null | https://arxiv.org/abs/2306.07483v1 | https://arxiv.org/pdf/2306.07483v1.pdf | Semi-supervised learning made simple with self-supervised clustering | Self-supervised learning models have been shown to learn rich visual representations without requiring human annotations. However, in many real-world scenarios, labels are partially available, motivating a recent line of work on semi-supervised methods inspired by self-supervised principles. In this paper, we propose a conceptually simple yet empirically powerful approach to turn clustering-based self-supervised methods such as SwAV or DINO into semi-supervised learners. More precisely, we introduce a multi-task framework merging a supervised objective using ground-truth labels and a self-supervised objective relying on clustering assignments with a single cross-entropy loss. This approach may be interpreted as imposing the cluster centroids to be class prototypes. Despite its simplicity, we provide empirical evidence that our approach is highly effective and achieves state-of-the-art performance on CIFAR100 and ImageNet. | ['Elisa Ricci', 'Moin Nabi', 'Julien Mairal', 'Xavier Alameda-Pineda', 'Karteek Alahari', 'Pietro Astolfi', 'Enrico Fini'] | 2023-06-13 | semi-supervised-learning-made-simple-with | http://openaccess.thecvf.com//content/CVPR2023/html/Fini_Semi-Supervised_Learning_Made_Simple_With_Self-Supervised_Clustering_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Fini_Semi-Supervised_Learning_Made_Simple_With_Self-Supervised_Clustering_CVPR_2023_paper.pdf | cvpr-2023-1 | ['clustering'] | ['methodology'] | [ 1.66610613e-01 4.84744221e-01 -5.54444313e-01 -7.53771782e-01
-7.52821207e-01 -6.03947580e-01 8.95794332e-01 3.89809102e-01
-6.34013534e-01 7.94971049e-01 3.73681188e-02 -4.78295051e-02
-2.36241687e-02 -3.08148891e-01 -6.30349696e-01 -7.79438436e-01
-2.06967726e-01 6.71386182e-01 1.04889147e-01 2.91983724e-01
7.59813702e-04 1.89743862e-01 -1.61885881e+00 3.16978544e-01
8.83737266e-01 8.69110823e-01 6.30681440e-02 3.68601233e-01
-4.38201688e-02 1.01956189e+00 -2.70494848e-01 -3.59688044e-01
1.87522441e-01 -5.63199341e-01 -1.22167277e+00 4.85188156e-01
5.64519644e-01 2.20655024e-01 1.73422813e-01 1.05208921e+00
2.04662681e-01 2.13616520e-01 1.11707449e+00 -1.35617661e+00
-5.17307520e-01 7.54350543e-01 -6.18230462e-01 -3.00286543e-02
-5.09006418e-02 2.27159858e-02 1.29110241e+00 -7.34261572e-01
6.59757257e-01 1.09810758e+00 6.02383077e-01 4.25226003e-01
-1.81975341e+00 -4.51415449e-01 1.61618933e-01 1.70825720e-01
-1.48287237e+00 -4.23370481e-01 8.85340691e-01 -6.46010816e-01
5.66961169e-01 -1.98003665e-01 2.36734524e-01 9.56907153e-01
-3.30608159e-01 1.06287181e+00 1.47623765e+00 -7.95949519e-01
5.83711207e-01 6.91618204e-01 2.98059255e-01 7.50973880e-01
2.04563498e-01 1.47859022e-01 -3.66188526e-01 -4.13010605e-02
3.80208939e-01 -1.54896840e-01 6.88383058e-02 -1.20458257e+00
-1.16500556e+00 1.12441921e+00 7.49100447e-01 4.88224298e-01
-2.26708762e-02 5.14877178e-02 5.15963852e-01 1.08718097e-01
7.46440172e-01 5.30086577e-01 -2.50656366e-01 3.22872221e-01
-1.23014092e+00 -2.24577680e-01 6.16793513e-01 8.78686070e-01
1.12943220e+00 1.57020055e-02 5.82814664e-02 7.77290523e-01
3.82431149e-01 1.45902827e-01 5.53573608e-01 -9.12434399e-01
-5.57242818e-02 4.81204778e-01 -8.22395645e-03 -5.40862978e-01
-5.74955583e-01 -6.84851885e-01 -8.01844239e-01 4.84401077e-01
4.70694333e-01 -1.12471335e-01 -1.00011325e+00 1.76179552e+00
8.09553266e-02 1.09975316e-01 4.12761122e-02 8.43740940e-01
5.77247143e-01 3.19638729e-01 4.36353862e-01 -3.85495245e-01
9.11842883e-01 -1.28655994e+00 -4.88029480e-01 -1.78007528e-01
7.55440116e-01 -4.04391885e-01 1.06396925e+00 4.16875660e-01
-7.84573376e-01 -7.78036773e-01 -1.03399229e+00 1.97828054e-01
-4.53662395e-01 4.50729758e-01 7.15897322e-01 6.89857543e-01
-1.01332700e+00 7.55284309e-01 -8.53477657e-01 -4.58766967e-01
8.50161612e-01 1.35051489e-01 -4.44537759e-01 1.13614373e-01
-6.77491724e-01 8.58021557e-01 7.62260795e-01 -4.08719480e-01
-1.04779923e+00 -3.55468273e-01 -8.66120815e-01 -9.24166888e-02
3.41337979e-01 -3.72418344e-01 1.13294482e+00 -1.59598660e+00
-1.28003395e+00 1.45069313e+00 -1.21163823e-01 -7.74209678e-01
5.28443754e-01 -6.05034605e-02 -1.42468870e-01 3.61404091e-01
1.89445272e-01 1.00023937e+00 1.10845792e+00 -1.95214701e+00
-5.11686563e-01 -2.74590403e-01 -9.18085054e-02 2.18350306e-01
-4.69377905e-01 -1.49687752e-01 -5.43093793e-02 -6.06655598e-01
-3.54510918e-02 -8.80884945e-01 -3.04696143e-01 -4.79346234e-03
-5.81824541e-01 -5.38207829e-01 6.84804201e-01 -1.06301084e-02
8.34439456e-01 -2.18576980e+00 7.10687339e-02 2.08772883e-01
3.45128655e-01 3.63886863e-01 -4.02897857e-02 1.90325141e-01
-3.06975722e-01 1.18727997e-01 -6.06642425e-01 -8.02582860e-01
4.96605486e-02 2.82374918e-01 -2.82227486e-01 5.83643734e-01
2.53877789e-01 8.84626269e-01 -1.18724632e+00 -8.88933241e-01
4.55969512e-01 4.11764532e-01 -2.48435020e-01 1.90798774e-01
-2.14942575e-01 5.19153595e-01 -1.44716367e-01 3.61232430e-01
3.38778466e-01 -5.22840261e-01 2.57267982e-01 -1.42255677e-02
5.85755408e-02 1.04004107e-01 -1.07832730e+00 1.75068557e+00
-3.53337079e-01 7.03782380e-01 -1.12408794e-01 -1.51597416e+00
8.13604832e-01 1.53498232e-01 3.32365870e-01 -4.30842906e-01
5.32103814e-02 -7.22146109e-02 -3.50399494e-01 -1.93268836e-01
1.23030260e-01 -4.99982536e-01 2.13561296e-01 6.94601715e-01
7.20338166e-01 -9.31013152e-02 2.31904045e-01 2.59931922e-01
6.06310844e-01 4.46766973e-01 6.36480749e-01 -4.74757254e-01
4.23936635e-01 1.53119579e-01 2.93366134e-01 7.77800798e-01
-3.47450376e-01 7.34521866e-01 2.80353159e-01 -3.45021635e-01
-9.13764775e-01 -1.25098276e+00 -2.67367393e-01 1.25414062e+00
-1.32914990e-01 -4.21719342e-01 -7.98397005e-01 -1.13028407e+00
-1.55558631e-01 6.26483142e-01 -7.79656768e-01 -7.89904520e-02
-2.25955516e-01 -5.93956053e-01 3.99870932e-01 5.30909896e-01
4.00966257e-01 -1.06440747e+00 -5.15503824e-01 2.45628809e-03
-1.25536472e-02 -1.04689252e+00 -5.69394082e-02 7.92287409e-01
-1.02514899e+00 -1.05251861e+00 -7.55800486e-01 -1.04132581e+00
1.05580819e+00 2.77020305e-01 1.36989295e+00 -1.10392936e-01
-2.50713289e-01 4.40507501e-01 -4.81049389e-01 -3.41119975e-01
-3.11947435e-01 1.35208711e-01 1.22543655e-01 3.14367950e-01
3.42896074e-01 -5.45988560e-01 -2.49251604e-01 2.50465006e-01
-8.40941131e-01 -5.71186375e-03 5.58265507e-01 8.39566588e-01
8.13298404e-01 4.00814116e-02 6.83422327e-01 -1.29860210e+00
1.21464953e-01 -4.14011121e-01 -4.24627692e-01 3.03923666e-01
-8.39853883e-01 2.88148671e-01 8.69392395e-01 -4.34634775e-01
-1.09083557e+00 7.21117854e-01 3.36420655e-01 -5.93698204e-01
-6.49843812e-01 3.31462413e-01 -4.05515023e-02 -8.41860026e-02
9.92196620e-01 1.43601641e-01 -1.60536483e-01 -3.63125592e-01
7.73761153e-01 6.27414465e-01 5.08791625e-01 -4.92441297e-01
1.03143215e+00 8.69360626e-01 -2.13501468e-01 -7.86385477e-01
-1.43710947e+00 -8.25980186e-01 -1.16223121e+00 -1.35385305e-01
9.89228010e-01 -1.04483414e+00 -3.50655556e-01 1.03362761e-01
-6.71890199e-01 -5.90753436e-01 -5.86706638e-01 5.90641260e-01
-9.35270905e-01 5.16543746e-01 -1.97347552e-01 -9.26645577e-01
5.62597662e-02 -6.92561507e-01 8.31026137e-01 1.79072648e-01
-2.06069946e-01 -1.40388191e+00 1.74425617e-01 4.84277517e-01
1.52159631e-01 1.36052728e-01 6.33603036e-01 -9.99588549e-01
-3.06362331e-01 6.10140823e-02 -3.35233182e-01 6.63104773e-01
1.42421871e-01 -2.50269771e-01 -1.37631440e+00 -4.54739153e-01
-1.90256476e-01 -1.02914310e+00 1.32444894e+00 3.24289024e-01
1.09212685e+00 -9.37840194e-02 -3.45821500e-01 6.14478350e-01
1.59614062e+00 -3.06859761e-01 2.24012464e-01 9.30459648e-02
7.79497981e-01 7.27981150e-01 4.79229063e-01 2.56730288e-01
3.55200887e-01 5.18685102e-01 2.63765633e-01 -3.70633513e-01
-2.00665250e-01 -4.32151705e-01 4.06991541e-02 5.10146379e-01
6.03733286e-02 8.94351229e-02 -9.69298005e-01 6.87322617e-01
-1.96221805e+00 -9.54780161e-01 1.04311943e-01 2.30559874e+00
9.47525203e-01 1.99613124e-01 3.26834828e-01 1.86102927e-01
7.10628331e-01 2.45074511e-01 -4.94788408e-01 6.55939430e-02
-3.26699287e-01 1.31043375e-01 4.26996201e-01 4.22103167e-01
-1.77264953e+00 1.07093561e+00 6.53988028e+00 8.06864738e-01
-9.23199654e-01 2.46420264e-01 7.67785907e-01 1.29275948e-01
5.09883873e-02 1.32770151e-01 -4.82035697e-01 2.78949589e-01
7.18632281e-01 -8.01843777e-03 1.11062460e-01 1.07185757e+00
-6.55029789e-02 -1.97589695e-01 -1.32725561e+00 1.02296484e+00
4.59064573e-01 -1.24115038e+00 -4.35657986e-02 -7.60806575e-02
1.03654814e+00 1.38701618e-01 1.64643279e-03 2.11472407e-01
6.85160637e-01 -1.01014805e+00 5.80901027e-01 2.86551058e-01
8.03154886e-01 -5.72983086e-01 5.86557686e-01 3.69424909e-01
-9.55404222e-01 -3.76469642e-02 -2.77644306e-01 -1.25627518e-01
-8.17191899e-02 6.68030381e-01 -7.88961530e-01 4.33031797e-01
6.24111712e-01 1.01231980e+00 -8.41072619e-01 1.13849592e+00
-4.67266440e-01 1.02014875e+00 -1.73327208e-01 2.19417080e-01
4.24894631e-01 -5.67864925e-02 2.16409966e-01 1.42806399e+00
-5.07399559e-01 -2.70886391e-01 5.37647486e-01 8.23893189e-01
-1.17740318e-01 9.76435766e-02 -6.37462020e-01 9.41634104e-02
1.47859812e-01 1.38404441e+00 -1.15519929e+00 -5.20475090e-01
-3.16685796e-01 9.09729958e-01 7.08676279e-01 4.55094397e-01
-4.86391425e-01 -2.21471027e-01 1.26022875e-01 7.29163438e-02
3.88661504e-01 -1.24892622e-01 -4.65952307e-01 -1.28872418e+00
-2.31102899e-01 -4.98415291e-01 5.09792030e-01 -7.11622298e-01
-1.56632698e+00 5.40645361e-01 1.21081971e-01 -1.39909875e+00
-2.87389308e-01 -5.02257586e-01 -5.79378009e-01 2.23888040e-01
-1.77134633e+00 -1.24289763e+00 -9.53811258e-02 6.09803021e-01
4.16409016e-01 -2.38510132e-01 9.45819080e-01 -7.46082002e-03
-3.00290495e-01 5.56533873e-01 4.45771009e-01 3.78153414e-01
8.84721816e-01 -1.69905090e+00 -8.46525580e-02 7.20313847e-01
8.82007957e-01 4.40893143e-01 6.87764883e-01 -2.90057540e-01
-5.95476568e-01 -1.14453781e+00 8.66577685e-01 -3.51829052e-01
5.96495032e-01 -4.72252697e-01 -8.61794949e-01 7.44276524e-01
2.82341480e-01 4.83564585e-01 8.52371037e-01 2.40394995e-01
-7.32341945e-01 -5.57724051e-02 -1.05723929e+00 1.83974683e-01
9.74834383e-01 -6.29230440e-01 -8.84660602e-01 6.58311546e-01
3.49558651e-01 1.89765930e-01 -5.28182924e-01 2.02671334e-01
8.38448107e-03 -9.66359437e-01 9.11460876e-01 -6.18188262e-01
2.77155817e-01 -4.58698630e-01 1.75286401e-02 -1.47073448e+00
-2.14620367e-01 -5.27147353e-01 5.45733571e-02 1.16238904e+00
3.31450760e-01 -3.03798586e-01 9.19963062e-01 7.48024955e-02
2.64202058e-02 -4.29426581e-01 -8.47700238e-01 -9.59733903e-01
1.55376792e-01 -1.85574517e-01 -3.00273418e-01 1.42779803e+00
3.06638986e-01 6.18487358e-01 -3.74969751e-01 -1.54335693e-01
1.15441954e+00 1.94441110e-01 5.77908456e-01 -1.52539706e+00
-2.42303714e-01 -3.61769319e-01 -5.43182969e-01 -8.28695595e-01
6.37543678e-01 -1.12791765e+00 1.92262113e-01 -1.37865222e+00
5.09211898e-01 -5.85509658e-01 -4.52083498e-01 6.95974052e-01
-4.47054505e-02 6.08743370e-01 1.63169548e-01 4.29151922e-01
-1.31846428e+00 5.73464930e-01 7.00314462e-01 -2.59187907e-01
-1.85572132e-01 1.47259829e-03 -5.94879568e-01 9.52364564e-01
8.35923910e-01 -6.21360004e-01 -5.28459609e-01 -1.34433597e-01
-7.96663016e-02 -4.73102570e-01 4.44324464e-01 -1.07353020e+00
2.90828586e-01 1.43046500e-02 4.02303874e-01 -3.02949488e-01
1.40979275e-01 -7.46886313e-01 -4.97931004e-01 2.17705920e-01
-7.16892481e-01 -5.38808048e-01 -1.67818889e-01 6.67498410e-01
-5.01630187e-01 -2.91160077e-01 1.10886979e+00 -1.77649558e-01
-7.34121203e-01 -4.54229712e-02 -2.17346460e-01 2.91012019e-01
1.12329900e+00 -1.62758246e-01 -1.59529716e-01 -4.40384984e-01
-9.63250816e-01 2.83038735e-01 5.15310585e-01 2.46407971e-01
3.81328583e-01 -1.13296807e+00 -6.55162632e-01 1.64325107e-02
4.92688388e-01 -1.77997306e-01 3.83919105e-02 5.40630937e-01
-1.14392325e-01 4.50341702e-01 -2.15624124e-01 -8.88907671e-01
-1.15048230e+00 8.56172800e-01 2.40511999e-01 -3.02907556e-01
-3.96116525e-01 7.34814584e-01 3.50776553e-01 -5.81773281e-01
5.99880159e-01 1.41710900e-02 -2.91354388e-01 2.20152214e-01
4.14089561e-01 1.26046926e-01 -1.82129398e-01 -7.48063207e-01
-2.63312489e-01 5.64098537e-01 6.70068292e-03 -1.02334410e-01
1.28366935e+00 -1.52731121e-01 2.14814126e-01 8.67765009e-01
1.42905045e+00 -3.65819246e-01 -1.48216116e+00 -6.43651664e-01
3.41214925e-01 -5.43394424e-02 7.10055232e-02 -8.44231009e-01
-8.70320261e-01 1.08009398e+00 6.81895971e-01 5.83928525e-02
9.51934814e-01 5.24202049e-01 5.86158410e-02 6.38394475e-01
3.17038059e-01 -1.34341109e+00 3.21297050e-01 2.31065735e-01
4.48567778e-01 -1.89036012e+00 1.79501548e-01 -2.52133965e-01
-9.14226472e-01 7.36347377e-01 4.99834985e-01 -3.34409148e-01
6.96369767e-01 -1.43536747e-01 2.62969941e-01 -3.07366159e-02
-6.15882754e-01 -7.21826255e-01 4.45119202e-01 9.50322092e-01
4.44781780e-01 1.61159411e-01 -1.14535086e-01 3.18587750e-01
1.07460357e-01 -9.60313305e-02 2.80235410e-01 8.29178810e-01
-5.46482086e-01 -9.31561768e-01 -8.98981020e-02 3.54654223e-01
-2.74844557e-01 -6.35737628e-02 -5.25188982e-01 7.16189802e-01
7.53067508e-02 9.07974780e-01 1.60619780e-01 -1.41638294e-01
-1.38630241e-01 3.38168353e-01 3.21586341e-01 -8.86545241e-01
-4.22079653e-01 4.93409485e-02 -1.12040743e-01 -2.62596995e-01
-1.03225672e+00 -6.11927450e-01 -1.21540797e+00 3.78909975e-01
-3.55778575e-01 2.76006877e-01 4.92535204e-01 1.06878245e+00
6.88238442e-02 1.07351094e-01 7.74331748e-01 -9.39496577e-01
-4.70489025e-01 -9.32662606e-01 -5.73018014e-01 7.56979465e-01
3.19218427e-01 -7.68124223e-01 -5.61351418e-01 4.41160083e-01] | [9.438148498535156, 2.8539044857025146] |
4093db2a-a187-4a88-b6ae-4a1783865896 | cross-lingual-and-cross-domain-transfer | null | null | https://aclanthology.org/2022.lrec-1.68 | https://aclanthology.org/2022.lrec-1.68.pdf | Cross-lingual and Cross-domain Transfer Learning for Automatic Term Extraction from Low Resource Data | Automatic Term Extraction (ATE) is a key component for domain knowledge understanding and an important basis for further natural language processing applications. Even with persistent improvements, ATE still exhibits weak results exacerbated by small training data inherent to specialized domain corpora. Recently, transformers-based deep neural models, such as BERT, have proven to be efficient in many downstream NLP tasks. However, no systematic evaluation of ATE has been conducted so far. In this paper, we run an extensive study on fine-tuning pre-trained BERT models for ATE. We propose strategies that empirically show BERT’s effectiveness using cross-lingual and cross-domain transfer learning to extract single and multi-word terms. Experiments have been conducted on four specialized domains in three languages. The obtained results suggest that BERT can capture cross-domain and cross-lingual terminologically-marked contexts shared by terms, opening a new design-pattern for ATE. | ['Beatrice Daille', 'Florian Boudin', 'Merieme Bouhandi', 'Amir Hazem'] | null | null | null | null | lrec-2022-6 | ['term-extraction'] | ['natural-language-processing'] | [ 3.35805677e-02 -1.04397923e-01 -6.25682890e-01 -4.64020163e-01
-1.12672246e+00 -6.71089530e-01 7.67962337e-01 9.32529047e-02
-6.70479298e-01 9.17522132e-01 1.70018539e-01 -5.59080303e-01
-1.57381073e-01 -6.68102682e-01 -6.67960346e-01 -3.48238349e-01
-2.47796208e-01 7.23497927e-01 -7.23158345e-02 -4.95083153e-01
-3.80299725e-02 2.76676416e-01 -1.18750179e+00 5.94093382e-01
1.07576799e+00 6.80716753e-01 3.43274623e-01 6.57787398e-02
-6.85620189e-01 4.79315430e-01 -7.23273635e-01 -5.21995246e-01
1.72204107e-01 -1.70458660e-01 -9.84597802e-01 -3.70475978e-01
2.15522394e-01 -1.17829226e-01 3.49560790e-02 9.91912484e-01
4.19869184e-01 -1.17085350e-03 6.25577569e-01 -8.04798663e-01
-8.68152082e-01 1.17909694e+00 -4.40903932e-01 2.93879211e-01
-5.38307475e-04 -1.69254377e-01 1.25637686e+00 -7.39062548e-01
7.46681571e-01 1.06410420e+00 8.40531826e-01 6.25643611e-01
-1.18735039e+00 -8.03096771e-01 1.13157734e-01 3.30278277e-01
-1.32361519e+00 -4.41975713e-01 8.86172116e-01 -2.21991256e-01
1.55553830e+00 -1.63708366e-02 2.93836534e-01 1.25921547e+00
1.49142802e-01 9.44627106e-01 1.07484746e+00 -1.05019879e+00
9.65739600e-03 3.26680273e-01 2.09518224e-01 3.21671665e-01
3.46885920e-01 5.63914180e-02 -4.89358038e-01 -1.02730341e-01
6.13403797e-01 -5.47032773e-01 -7.29915351e-02 -5.38523123e-02
-8.55914414e-01 1.05953467e+00 2.17047721e-01 1.02132750e+00
-3.92366618e-01 -4.48924713e-02 8.82070065e-01 5.32636881e-01
9.08329844e-01 6.75273299e-01 -1.12845337e+00 -3.84067148e-02
-8.96923184e-01 2.82618076e-01 7.96739757e-01 1.09846210e+00
6.97385430e-01 -1.02548599e-02 -1.28302395e-01 1.34792244e+00
4.89151012e-03 2.87547588e-01 8.62042129e-01 -5.17046273e-01
6.42316699e-01 6.10897839e-01 -1.97983742e-01 -6.58570111e-01
-3.31717223e-01 -5.61272681e-01 -6.19857132e-01 -3.33195746e-01
2.58782297e-01 -2.94858724e-01 -9.90041614e-01 1.89270020e+00
1.30458608e-01 -2.27164343e-01 1.19066993e-02 4.81525928e-01
7.55035341e-01 7.20359623e-01 3.85054946e-01 -1.90361422e-02
1.66896963e+00 -7.86736012e-01 -8.05559814e-01 -6.45821989e-01
9.24707651e-01 -6.56387389e-01 1.06839073e+00 4.39530253e-01
-9.99322355e-01 -4.15242761e-01 -8.17493737e-01 -4.09956038e-01
-8.76039743e-01 3.47681820e-01 7.85616159e-01 6.28410876e-01
-9.01422381e-01 3.95602286e-01 -5.67355037e-01 -4.76531237e-01
3.53953123e-01 2.16444969e-01 -3.37905705e-01 -2.84253299e-01
-1.79501998e+00 1.08969378e+00 8.08661640e-01 2.77191792e-02
-7.17817605e-01 -6.86400235e-01 -8.89166832e-01 1.06133021e-01
4.13730860e-01 -5.96687257e-01 1.49954903e+00 -9.79450762e-01
-1.23825228e+00 1.39178920e+00 -2.76925087e-01 -7.77813017e-01
3.02530229e-02 -5.25888681e-01 -7.75447488e-01 -1.58460706e-01
3.01995695e-01 5.14208198e-01 5.83820105e-01 -1.01494408e+00
-7.21615493e-01 -4.44605082e-01 1.80204615e-01 1.28852710e-01
-6.96417451e-01 5.22261560e-01 -2.64094114e-01 -9.13161635e-01
-4.32556778e-01 -6.73243105e-01 -8.79138801e-03 -4.73275095e-01
-6.14469498e-02 -7.25396812e-01 3.43785316e-01 -7.91754246e-01
1.48422420e+00 -1.82566977e+00 -7.83050209e-02 -1.05515406e-01
-3.46830845e-01 7.67606795e-01 -8.50306824e-02 7.68714309e-01
-3.22617800e-03 1.98052138e-01 -2.49883890e-01 -2.97831059e-01
1.30449712e-01 3.85062039e-01 -3.72700512e-01 7.61942798e-03
4.26911235e-01 9.48060274e-01 -8.40258181e-01 -4.94962603e-01
4.41704839e-02 3.95661473e-01 -3.63570333e-01 -7.19394861e-03
-5.12785137e-01 8.93554240e-02 -4.94990200e-01 4.64083910e-01
5.77086508e-01 -5.34205213e-02 3.74866009e-01 1.55607283e-01
-1.97581813e-01 1.01677489e+00 -6.47843778e-01 2.00701070e+00
-9.31098163e-01 6.54202938e-01 1.49798319e-01 -1.28122294e+00
7.85350859e-01 5.61309874e-01 2.17560232e-01 -9.12052691e-01
1.51557073e-01 5.68579376e-01 -7.61701763e-02 -5.19406259e-01
6.64940596e-01 -4.25288498e-01 -2.60196179e-01 1.57978803e-01
4.34360445e-01 1.06480725e-01 3.78461957e-01 -2.13844389e-01
8.97309005e-01 2.20637038e-01 5.23796499e-01 -4.68095154e-01
7.00744033e-01 4.63323861e-01 4.85212862e-01 4.40488875e-01
-1.42919838e-01 2.68395215e-01 2.04273030e-01 -3.99064243e-01
-9.71813738e-01 -6.61642849e-01 -3.57230425e-01 1.46279252e+00
-1.79974079e-01 -3.16381693e-01 -7.28460193e-01 -7.54023314e-01
-4.82909158e-02 8.95234823e-01 -3.74065995e-01 -2.41570082e-02
-9.35984612e-01 -7.03721941e-01 9.43440318e-01 3.55638921e-01
5.04623413e-01 -1.36441696e+00 -1.10560589e-01 6.07405424e-01
-4.28443879e-01 -1.25602126e+00 -2.62378752e-01 6.44278586e-01
-7.62431979e-01 -6.25598609e-01 -8.10606897e-01 -1.29917431e+00
2.00613942e-02 8.70297402e-02 1.51169443e+00 -1.92967027e-01
-1.38050690e-01 -3.03905457e-01 -5.03825426e-01 -6.03753448e-01
-3.28027815e-01 7.15531230e-01 -9.25336778e-02 -3.92626941e-01
1.22485471e+00 -4.94602799e-01 -2.89295882e-01 6.91588297e-02
-9.47145224e-01 -4.04782653e-01 7.94697642e-01 9.28211331e-01
3.66398841e-01 7.06368312e-03 1.01021636e+00 -1.07662225e+00
9.84494746e-01 -5.45511782e-01 -5.35076857e-01 3.22027922e-01
-5.41409492e-01 3.19154829e-01 7.27414310e-01 -3.02959919e-01
-1.40891039e+00 -3.58398646e-01 -4.97370899e-01 8.07249099e-02
-5.07865429e-01 9.09426391e-01 -1.12883188e-01 3.11836332e-01
9.13266301e-01 1.25624046e-01 -4.87699956e-01 -9.31980491e-01
4.65021640e-01 9.66638088e-01 2.15096414e-01 -9.21019673e-01
3.43040645e-01 1.90499336e-01 -6.30027592e-01 -8.81559610e-01
-1.07868528e+00 -8.13304186e-01 -5.89197695e-01 3.92340571e-01
7.96086729e-01 -1.06708515e+00 -1.04067037e-02 2.75527924e-01
-1.37693751e+00 -3.02998185e-01 -6.97481632e-02 3.37893337e-01
-4.16818261e-02 2.50784606e-01 -7.53067791e-01 -4.76876944e-01
-6.58982337e-01 -7.01720357e-01 1.08797753e+00 -3.35217938e-02
-3.75627905e-01 -1.25945306e+00 3.91373932e-01 2.70305783e-01
4.71180260e-01 -1.89345181e-01 1.16941905e+00 -9.38009560e-01
-1.78272679e-01 -1.91519484e-02 -1.90050587e-01 5.40461659e-01
8.67059603e-02 -3.41977954e-01 -1.17181993e+00 -1.16263971e-01
4.15114779e-03 -4.29955661e-01 8.68712068e-01 4.24109548e-01
7.64736295e-01 -1.33519530e-01 -4.91555333e-01 5.19734800e-01
1.53979325e+00 3.34766775e-01 4.62912053e-01 7.66183317e-01
4.09959376e-01 8.07783604e-01 5.13113499e-01 8.49406980e-03
2.28158623e-01 5.63232899e-01 -1.68347195e-01 7.05497563e-02
-2.37182394e-01 -1.02403693e-01 5.01628101e-01 9.75031972e-01
9.67806429e-02 -2.53540367e-01 -1.21972609e+00 1.08228946e+00
-1.46999156e+00 -7.84252465e-01 -1.35237589e-01 1.86616457e+00
1.30638266e+00 3.42229217e-01 -1.40603140e-01 -1.52155504e-01
6.50219679e-01 -3.42205912e-02 -2.25115016e-01 -7.70289004e-01
-1.08402066e-01 7.80682325e-01 3.78620207e-01 3.07665914e-01
-1.23488903e+00 1.42460144e+00 6.30091143e+00 1.30385733e+00
-1.13818765e+00 2.41497084e-01 2.83712804e-01 1.96616203e-01
-3.72964829e-01 -6.85193539e-02 -1.12519109e+00 1.60651594e-01
1.09132731e+00 -3.35427493e-01 2.97411859e-01 1.11976552e+00
-1.94237188e-01 2.22725168e-01 -1.10867691e+00 6.99622095e-01
-7.76586607e-02 -1.03817713e+00 9.27537307e-02 6.03614934e-02
7.47484565e-01 4.52237070e-01 -1.66950211e-01 8.75152409e-01
4.37240899e-01 -1.00341165e+00 3.56692255e-01 -4.25955683e-01
9.11423981e-01 -9.29616988e-01 8.69517863e-01 4.71137047e-01
-1.09503198e+00 1.34550393e-01 -3.89066726e-01 -4.44348995e-03
1.11791372e-01 7.65212655e-01 -1.05649555e+00 8.19715500e-01
5.42476654e-01 4.24717724e-01 -3.09492260e-01 8.09144557e-01
-4.55787212e-01 8.60664248e-01 -3.70472878e-01 -1.64245754e-01
6.40417814e-01 -5.53318411e-02 4.79099780e-01 1.69697380e+00
1.82716146e-01 -3.31849813e-01 -9.74994302e-02 9.47993219e-01
-2.24214047e-01 3.35245907e-01 -7.54693925e-01 -2.26244062e-01
4.13883269e-01 1.10169947e+00 -4.40614641e-01 -4.97917682e-01
-5.69608927e-01 8.62332344e-01 7.23136723e-01 3.75517964e-01
-6.72103643e-01 -5.95833540e-01 8.01959097e-01 -9.54810977e-02
4.79953945e-01 -3.93941879e-01 -4.73181933e-01 -1.18965638e+00
1.27999619e-01 -9.86485124e-01 5.38715780e-01 -3.43850553e-01
-1.55553818e+00 8.94343793e-01 2.93812364e-01 -9.39961672e-01
-6.16190314e-01 -8.83871794e-01 -2.36127049e-01 1.05403876e+00
-1.99001789e+00 -1.35441494e+00 3.16716760e-01 5.50781608e-01
7.90665567e-01 -2.32469156e-01 1.13768351e+00 6.85500085e-01
-4.14274842e-01 7.00273752e-01 4.30384666e-01 3.49432856e-01
8.67353082e-01 -1.31851542e+00 6.89666152e-01 7.30132818e-01
3.81515741e-01 9.77556407e-01 5.41002393e-01 -5.29139102e-01
-9.59478438e-01 -9.02339756e-01 1.55340564e+00 -2.67734677e-01
8.33980381e-01 -5.58838010e-01 -1.12216258e+00 8.05551469e-01
6.19785845e-01 -4.63471681e-01 7.29557395e-01 7.49273896e-01
-4.47633684e-01 -9.24283713e-02 -9.71702397e-01 4.70412612e-01
1.08623362e+00 -7.07286477e-01 -1.12715495e+00 2.94395685e-01
8.65896106e-01 -9.81712565e-02 -7.48752475e-01 3.89789522e-01
4.19799715e-01 -6.46910548e-01 9.28070962e-01 -6.34838820e-01
4.95983064e-01 2.11106345e-01 1.29850343e-01 -1.55099201e+00
-3.38862717e-01 -4.43721294e-01 4.85584587e-02 1.54514921e+00
5.51513612e-01 -5.64540029e-01 5.61624289e-01 2.34730899e-01
-3.30597997e-01 -4.26450968e-01 -9.16049123e-01 -1.16512787e+00
7.00931132e-01 -6.10111058e-01 6.14744425e-01 1.31224084e+00
9.02885869e-02 7.28728592e-01 -5.41811027e-02 -1.54932216e-01
3.02703053e-01 1.53234586e-01 3.76272649e-01 -1.10315454e+00
-2.43312210e-01 -5.78112602e-01 -2.19669240e-03 -1.07914257e+00
4.99590904e-01 -1.13822067e+00 1.41569883e-01 -1.46480906e+00
-1.10705175e-01 -6.38364792e-01 -5.25784433e-01 5.69026232e-01
-7.65255392e-02 -3.03255022e-02 -4.75907296e-01 -8.74987096e-02
-1.71687961e-01 4.74912882e-01 8.92023146e-01 -1.15543343e-01
-1.71036124e-01 -3.11181068e-01 -7.38060176e-01 6.80002987e-01
8.99160266e-01 -6.95008099e-01 -2.88770586e-01 -1.04053831e+00
2.13551760e-01 -4.64104503e-01 -1.91531405e-01 -5.79673469e-01
-4.68326956e-02 -4.68741134e-02 -1.62710976e-02 -4.80465233e-01
1.15232982e-01 -8.32291424e-01 -3.04271311e-01 1.66636452e-01
-3.88973087e-01 -1.38252005e-01 5.45914888e-01 1.96467414e-01
-5.72275817e-01 -4.79616106e-01 4.67644542e-01 -3.10981721e-01
-1.09469378e+00 7.92649165e-02 -3.56693774e-01 4.43984061e-01
7.73784339e-01 1.14099927e-01 -1.41359657e-01 4.43126336e-02
-3.83340359e-01 2.11535096e-01 7.01678991e-02 6.29411101e-01
1.23420671e-01 -1.30914545e+00 -8.39052320e-01 1.85696065e-01
4.25873220e-01 9.57392082e-02 7.03699607e-03 2.85134703e-01
-4.08273280e-01 9.05134201e-01 -1.02348998e-01 -3.74278843e-01
-1.07569849e+00 6.74027801e-01 2.32664883e-01 -8.21396708e-01
-4.46788222e-01 1.06244612e+00 1.78822041e-01 -7.45308459e-01
2.09240749e-01 -5.58194280e-01 -2.51734167e-01 1.06120363e-01
4.38256949e-01 -3.55561194e-03 4.89079624e-01 -4.85778213e-01
-4.55692738e-01 2.96189785e-01 -4.56386536e-01 -2.77493864e-01
1.40958905e+00 -1.55664638e-01 -1.18963003e-01 4.16922897e-01
1.30368114e+00 -1.32154420e-01 -4.07337129e-01 -6.16254032e-01
6.76351666e-01 -8.81190449e-02 1.00380301e-01 -9.18340445e-01
-6.49817228e-01 1.00459254e+00 1.34758770e-01 3.48927110e-01
1.16936004e+00 -7.85994008e-02 1.13265610e+00 5.02208531e-01
5.89205980e-01 -1.38932264e+00 -4.10392046e-01 9.59242463e-01
8.28394711e-01 -1.33203709e+00 -3.24020028e-01 -3.21794093e-01
-4.67390329e-01 8.64540875e-01 4.57630664e-01 -1.94022879e-01
6.44854724e-01 1.88664928e-01 1.76636785e-01 -2.59405553e-01
-7.09859133e-01 -3.65461349e-01 4.53466952e-01 5.25507867e-01
8.87346029e-01 9.41518545e-02 -6.58445656e-01 6.29918158e-01
-2.48788387e-01 2.32112899e-01 -1.33855492e-01 9.45870996e-01
-3.02928478e-01 -1.66307139e+00 -1.92297056e-01 1.55298576e-01
-9.54734623e-01 -6.61935329e-01 -5.46804368e-01 1.04966307e+00
3.07527393e-01 8.78345370e-01 -2.25234240e-01 -3.96941975e-02
2.63331234e-01 6.22514665e-01 5.58177531e-01 -8.51879179e-01
-8.93586099e-01 2.78959330e-02 6.17635787e-01 -2.03675032e-01
-4.78887141e-01 -4.28368121e-01 -1.09466970e+00 -7.99927860e-02
-4.80696678e-01 2.81271309e-01 7.46534348e-01 1.04727876e+00
4.56155509e-01 4.88066912e-01 1.66307911e-01 -5.09111106e-01
-6.52233660e-01 -1.34734809e+00 -5.14912963e-01 4.55978900e-01
2.22198069e-01 -5.64309895e-01 -1.28658146e-01 8.97328332e-02] | [10.363790512084961, 9.182446479797363] |
fa75cbb4-dfcb-4976-9d73-49c844abbe74 | normalized-compression-distance-of-multisets | 1212.5711 | null | http://arxiv.org/abs/1212.5711v4 | http://arxiv.org/pdf/1212.5711v4.pdf | Normalized Compression Distance of Multisets with Applications | Normalized compression distance (NCD) is a parameter-free, feature-free,
alignment-free, similarity measure between a pair of finite objects based on
compression. However, it is not sufficient for all applications. We propose an
NCD of finite multisets (a.k.a. multiples) of finite objects that is also a
metric. Previously, attempts to obtain such an NCD failed. We cover the entire
trajectory from theoretical underpinning to feasible practice. The new NCD for
multisets is applied to retinal progenitor cell classification questions and to
related synthetically generated data that were earlier treated with the
pairwise NCD. With the new method we achieved significantly better results.
Similarly for questions about axonal organelle transport. We also applied the
new NCD to handwritten digit recognition and improved classification accuracy
significantly over that of pairwise NCD by incorporating both the pairwise and
NCD for multisets. In the analysis we use the incomputable Kolmogorov
complexity that for practical purposes is approximated from above by the length
of the compressed version of the file involved, using a real-world compression
program.
Index Terms--- Normalized compression distance, multisets or multiples,
pattern recognition, data mining, similarity, classification, Kolmogorov
complexity, retinal progenitor cells, synthetic data, organelle transport,
handwritten character recognition | ['Andrew R. Cohen', 'Paul M. B. Vitanyi'] | 2012-12-22 | null | null | null | null | ['handwritten-digit-recognition'] | ['computer-vision'] | [ 6.69766128e-01 -4.71485615e-01 1.28359005e-01 -3.33860159e-01
-6.16180301e-01 -6.25192523e-01 6.04879022e-01 5.72429836e-01
-7.03676462e-01 9.59263861e-01 -8.38707089e-02 -2.09524602e-01
-8.31342578e-01 -7.27199495e-01 -4.82125849e-01 -8.81748855e-01
-1.93058938e-01 6.15824342e-01 3.47753465e-01 -1.18714638e-01
1.03605366e+00 8.75205994e-01 -2.09763241e+00 2.77403176e-01
7.46586442e-01 9.53806460e-01 3.32142711e-01 1.01800311e+00
-1.44764751e-01 3.53748024e-01 -5.14697015e-01 -3.50107402e-01
5.90757608e-01 -4.90693569e-01 -1.02259576e+00 2.17542201e-01
4.87618893e-01 -6.27623312e-03 -1.51027173e-01 1.00189090e+00
4.35481668e-01 -7.50041902e-02 1.08878708e+00 -1.37356627e+00
-5.85281909e-01 2.44201258e-01 -1.93903089e-01 3.46289873e-01
2.93155313e-01 -1.34658724e-01 6.52616799e-01 -6.70446575e-01
7.35539734e-01 1.15256536e+00 6.64587975e-01 3.29863578e-01
-1.18517935e+00 -1.38456956e-01 -7.93682754e-01 1.66809306e-01
-1.50214839e+00 -1.96392536e-01 -1.02382816e-01 -4.56272334e-01
9.97587621e-01 8.61911476e-01 7.10004210e-01 3.89208972e-01
2.64842510e-01 5.84015287e-02 1.17167664e+00 -8.41678619e-01
3.90646607e-01 -8.82200450e-02 3.19583982e-01 4.31031257e-01
6.01051211e-01 -9.99749228e-02 -8.68173502e-03 -2.91755140e-01
7.97220528e-01 -2.00199261e-01 -1.48978278e-01 6.78251684e-02
-1.32777143e+00 7.42637157e-01 -3.35946590e-01 5.41217744e-01
3.57581489e-02 -1.06288046e-01 4.57193673e-01 7.29488194e-01
-1.01067601e-02 4.89218652e-01 -3.43293995e-01 -3.86624157e-01
-9.03573215e-01 4.36803758e-01 9.91837502e-01 1.26720810e+00
3.86814624e-01 -4.53070402e-01 1.08083664e-02 8.62593889e-01
-3.40074487e-02 2.34249473e-01 8.88870418e-01 -1.18648684e+00
1.19759642e-01 6.03038013e-01 -2.74464488e-01 -8.25292051e-01
-3.71049911e-01 1.00573100e-01 -1.05145097e+00 2.87014663e-01
6.03180230e-01 4.79057789e-01 -6.48019493e-01 1.47864211e+00
1.26333088e-01 4.89737419e-03 2.56594390e-01 5.58164120e-01
4.12390977e-01 4.10106659e-01 -4.36060101e-01 -5.72612822e-01
1.27431679e+00 -4.91208911e-01 -3.57576042e-01 6.09987319e-01
8.57159555e-01 -9.88242447e-01 7.52710879e-01 5.75853765e-01
-1.09632289e+00 -2.30900168e-01 -1.06700623e+00 -4.37069796e-02
-4.43981171e-01 -1.51618853e-01 3.82656783e-01 8.19365978e-01
-1.18513095e+00 9.68808949e-01 -4.88267094e-01 -7.42943704e-01
4.88843858e-01 5.91042161e-01 -5.35007298e-01 -2.13436201e-01
-5.96986711e-01 1.00712395e+00 4.63098884e-01 -5.97197056e-01
-9.05280337e-02 -3.77363533e-01 -3.59783590e-01 -1.78669676e-01
-2.10858032e-01 -5.63833833e-01 7.72952795e-01 -4.57713902e-01
-1.19900632e+00 9.32327509e-01 5.06818928e-02 -7.09266961e-01
4.03268188e-01 5.12718439e-01 -2.36131370e-01 4.30591285e-01
-2.09975228e-01 5.50387740e-01 3.65456641e-01 -6.60546482e-01
-4.31574851e-01 -2.96512932e-01 -3.32971215e-01 -1.17519088e-01
-4.64244157e-01 1.64510220e-01 1.51873529e-01 -7.23684728e-01
2.55306780e-01 -9.40183461e-01 -1.74578056e-01 4.23969477e-01
-1.42416790e-01 -2.92570502e-01 6.11615658e-01 -2.89151192e-01
1.12455416e+00 -2.00253916e+00 2.21166566e-01 4.07144845e-01
7.52685666e-02 3.89291018e-01 -4.00962263e-01 6.16523981e-01
-7.06320032e-02 4.54330027e-01 -7.00573266e-01 -8.60897005e-02
-3.85104954e-01 5.47733366e-01 -7.76010007e-03 6.94877565e-01
1.82532087e-01 3.51134628e-01 -6.44772530e-01 -6.02675259e-01
-1.05691016e-01 8.59360695e-02 -4.30871397e-01 -2.66703486e-01
3.72527167e-02 -1.85584158e-01 -2.90802643e-02 5.91192842e-01
8.97862554e-01 -1.55009523e-01 -2.16504216e-01 -5.62027209e-02
-3.04547966e-01 -2.28872135e-01 -1.30481160e+00 1.54911089e+00
1.07019588e-01 7.60752678e-01 -4.94462162e-01 -1.04595578e+00
1.21067834e+00 5.62555194e-02 5.87608039e-01 -4.93924171e-01
9.75761041e-02 5.28356373e-01 3.44748437e-01 -5.07957518e-01
5.54336905e-01 -1.21825628e-01 3.00764591e-01 5.70192397e-01
1.18803255e-01 -1.97862595e-01 5.07543445e-01 1.79029793e-01
1.41889799e+00 -4.91807759e-01 6.28238261e-01 -5.76424360e-01
7.14976668e-01 -1.23102382e-01 3.30434620e-01 6.07845306e-01
-7.09204152e-02 1.07793128e+00 4.04426515e-01 -2.17813849e-01
-1.78956580e+00 -8.42488408e-01 -8.06788921e-01 3.04595858e-01
9.18991566e-02 -5.04119039e-01 -8.62261355e-01 -7.37572610e-02
1.50339961e-01 3.44816327e-01 -3.35694939e-01 -8.28312412e-02
-3.08551788e-01 -7.18880773e-01 1.07866275e+00 1.42645657e-01
3.38694841e-01 -7.29311824e-01 -7.00477540e-01 1.97131395e-01
3.00924312e-02 -1.08077896e+00 -3.37802649e-01 -7.67843612e-03
-1.22799826e+00 -1.12815714e+00 -7.19247043e-01 -8.57935190e-01
4.67627794e-01 9.54342261e-02 6.42033041e-01 2.91432858e-01
-7.29043961e-01 2.61919022e-01 -4.54057813e-01 -4.90490019e-01
-4.85279590e-01 -5.32227993e-01 5.13419211e-01 -3.71962190e-01
3.27819526e-01 -7.69183993e-01 -2.51594961e-01 5.36579967e-01
-1.31985164e+00 -2.56181031e-01 5.80971301e-01 8.41107428e-01
8.75894606e-01 6.23276755e-02 6.35513127e-01 -2.64973253e-01
5.96244812e-01 -3.29935819e-01 -3.19112837e-01 3.25730652e-01
-5.17639995e-01 3.40685666e-01 7.06181347e-01 -7.39164770e-01
-1.43965647e-01 -2.42588148e-01 -3.86692174e-02 -2.89646655e-01
-2.48538151e-01 2.83251017e-01 -5.77541664e-02 -5.23051023e-01
9.21270907e-01 5.67224145e-01 5.20514131e-01 -3.99594575e-01
1.93933457e-01 1.13815641e+00 6.40587568e-01 -4.86434996e-01
2.69114316e-01 5.45714438e-01 6.04151070e-01 -1.10436440e+00
-2.26722984e-03 -5.94372451e-01 -8.13963592e-01 1.03141390e-01
4.69604850e-01 -2.36265436e-01 -7.30736732e-01 7.53172100e-01
-1.39091563e+00 2.22132757e-01 -5.58987021e-01 8.01482797e-01
-1.13542891e+00 8.77958596e-01 -5.00429451e-01 -6.14468098e-01
-4.02664155e-01 -8.40914071e-01 5.56735873e-01 -1.17964618e-01
-2.91210301e-02 -6.37053192e-01 1.94785461e-01 1.53863773e-01
2.46665776e-01 4.04190183e-01 1.28279293e+00 -1.08572876e+00
-3.51133466e-01 -2.52400130e-01 -1.81185082e-01 6.59972310e-01
1.75950617e-01 2.34353259e-01 -4.12660182e-01 -1.87325493e-01
3.33106279e-01 -3.67138684e-01 6.29030287e-01 2.34921128e-01
1.43285871e+00 -6.68920279e-01 -9.17846113e-02 6.35385931e-01
1.61844349e+00 2.79539853e-01 1.19600880e+00 5.19909322e-01
4.52544447e-03 5.21669686e-01 5.84211588e-01 8.64306629e-01
1.09576676e-02 4.07356083e-01 2.73944676e-01 5.62142491e-01
-7.11776465e-02 4.78487074e-01 2.89261248e-02 1.15938604e+00
-4.74561155e-01 -5.83726704e-01 -7.45612979e-01 5.18118382e-01
-1.54620242e+00 -1.15792227e+00 -4.27931905e-01 2.51395202e+00
8.26538861e-01 7.69278333e-02 2.67816037e-01 7.28035569e-01
8.96257639e-01 -6.84097588e-01 -5.03345609e-01 -7.31572688e-01
-6.72750056e-01 3.45502645e-01 6.99744105e-01 2.50978470e-01
-5.18855631e-01 1.66113421e-01 5.88619184e+00 1.32796359e+00
-7.29857922e-01 -1.95707992e-01 3.78152937e-01 -1.14880748e-01
-1.21697120e-01 8.67709070e-02 -6.74568653e-01 6.06368721e-01
1.25764692e+00 -5.76898098e-01 4.85579401e-01 3.42706949e-01
-2.50654250e-01 -2.69672006e-01 -1.34117568e+00 1.38542676e+00
1.87092498e-01 -1.49395657e+00 5.22797167e-01 4.90638256e-01
4.98900741e-01 -8.89090449e-02 -9.45304185e-02 -3.91941398e-01
-2.98846781e-01 -1.22756851e+00 5.83664238e-01 7.49133706e-01
9.11740720e-01 -7.90322781e-01 7.74133861e-01 5.54232717e-01
-8.75739694e-01 -1.28661886e-01 -9.51522052e-01 -5.97743429e-02
-1.13664962e-01 6.42795324e-01 -8.43487024e-01 5.75356305e-01
2.70207524e-01 5.29037297e-01 -6.18492246e-01 1.64045286e+00
1.12680435e+00 2.12082654e-01 -6.77748978e-01 -3.05526435e-01
2.06401005e-01 -4.93966162e-01 6.49432242e-01 1.33735812e+00
8.92566144e-01 4.51800644e-01 -4.03634608e-01 4.48728770e-01
1.06838457e-01 3.14368993e-01 -8.23106766e-01 -1.07846066e-01
6.96146488e-01 9.94509161e-01 -9.11969900e-01 -2.55666494e-01
-1.72683567e-01 8.35149765e-01 1.09808840e-01 -2.73776770e-01
-4.73066390e-01 -8.61392200e-01 6.81658745e-01 2.24746227e-01
2.35005856e-01 -2.55243659e-01 -5.16055226e-01 -6.80321932e-01
-1.39700565e-02 -7.64813006e-01 3.48673016e-01 -5.51365674e-01
-1.44522333e+00 4.49804753e-01 8.08932185e-02 -1.67943501e+00
8.60854052e-03 -8.16431344e-01 -2.89148688e-01 7.36854315e-01
-9.28532362e-01 -4.63770837e-01 1.64703559e-02 5.46567082e-01
3.50745767e-01 -4.00329143e-01 1.02445388e+00 4.06545639e-01
5.41507117e-02 6.38388932e-01 6.22817457e-01 -2.00189084e-01
5.57208598e-01 -1.03022373e+00 7.84088224e-02 3.51983011e-01
-6.92376867e-02 6.33913696e-01 5.72521031e-01 -4.93826479e-01
-1.23833764e+00 -1.02824795e+00 9.33845580e-01 -1.48256332e-01
3.93711716e-01 -2.48581227e-02 -1.04491234e+00 -1.09204324e-02
-9.13920477e-02 9.77139026e-02 9.91140902e-01 -7.05590189e-01
-2.12393120e-01 4.22258750e-02 -1.72121072e+00 3.64615440e-01
1.22499752e+00 -2.92153686e-01 -5.82403302e-01 7.59137630e-01
5.56105375e-01 2.51741279e-02 -1.38582051e+00 3.44945788e-01
7.13240504e-01 -9.72533703e-01 1.10177827e+00 -8.07771146e-01
5.07798791e-01 -2.13500172e-01 -7.23555326e-01 -8.03815365e-01
-3.43726635e-01 -4.56498176e-01 1.12409718e-01 1.06412625e+00
1.48531929e-01 -4.92922813e-01 4.97405767e-01 3.89602333e-01
-2.11259082e-01 -1.03242826e+00 -1.35396886e+00 -1.44892418e+00
3.16336453e-01 -2.47407049e-01 6.57230377e-01 9.28334594e-01
2.95787007e-01 -1.08737275e-01 -1.80188790e-01 -1.97307721e-01
8.15683484e-01 -1.15896292e-01 3.47387999e-01 -1.57070589e+00
-3.02355438e-01 -5.20413280e-01 -1.49312878e+00 -7.96472073e-01
-3.37228984e-01 -1.24118257e+00 -3.12009692e-01 -1.29550946e+00
2.59863853e-01 -7.23931313e-01 -2.25897934e-02 6.19794540e-02
6.43054545e-01 1.73437148e-01 2.23937690e-01 3.71641487e-01
-3.51068914e-01 2.16632977e-01 1.21900499e+00 8.31363276e-02
-4.65685204e-02 -1.07041582e-01 -4.00308043e-01 6.04755461e-01
9.19083893e-01 -6.72409296e-01 -1.61268622e-01 -6.82305619e-02
-6.13855310e-02 1.17272958e-01 3.22643995e-01 -1.29978907e+00
6.22801721e-01 -2.06991300e-01 -6.80402014e-03 -5.77265799e-01
3.95151675e-01 -6.54857397e-01 3.45094562e-01 6.54330909e-01
-4.51017886e-01 2.14200795e-01 -2.59992201e-02 6.48308992e-01
-6.29000366e-02 -9.18259442e-01 9.42234457e-01 -1.92305431e-01
-2.67057151e-01 2.45681211e-01 -6.51382208e-01 -6.15204163e-02
1.25361097e+00 -8.12419355e-01 -5.44286788e-01 -3.30119990e-02
-4.70575243e-01 -3.55064869e-01 6.25201881e-01 -5.43591827e-02
1.01221657e+00 -1.33042812e+00 -8.98958623e-01 2.98329830e-01
1.04805075e-01 -1.98462516e-01 -1.53388143e-01 8.76230299e-01
-9.49469566e-01 4.52015013e-01 -6.99993610e-01 -6.44471169e-01
-1.84752703e+00 5.60268164e-01 1.23246303e-02 2.51816243e-01
-3.06521952e-01 5.71011424e-01 -5.48565507e-01 -3.17295432e-01
1.07513614e-01 -7.78881431e-01 -1.57599568e-01 -2.06139348e-02
6.29051387e-01 8.69882524e-01 4.13764268e-01 -6.38251722e-01
-1.11063890e-01 9.28478301e-01 -2.81617772e-02 -3.49598303e-02
1.41930437e+00 -1.29564703e-01 -5.79157889e-01 4.58685517e-01
1.61165011e+00 -4.09042060e-01 -4.73864675e-01 9.73841846e-02
1.12124257e-01 -7.13994682e-01 -5.52610278e-01 -4.20972973e-01
-4.52474236e-01 5.46092033e-01 8.40855360e-01 3.58872592e-01
1.13285232e+00 -3.20498683e-02 6.11548126e-01 8.75308454e-01
4.95768070e-01 -9.91578400e-01 -1.02000497e-01 6.07359827e-01
9.14038360e-01 -8.13248515e-01 8.45808163e-02 -4.53153104e-01
-2.81567961e-01 1.62611103e+00 1.51651517e-01 -1.27773508e-01
7.40134180e-01 5.10789514e-01 -6.17870152e-01 2.08522663e-01
-8.68534327e-01 7.12701902e-02 -1.15590066e-01 7.91945994e-01
4.33804765e-02 -2.33326927e-01 -9.59591866e-01 2.44449139e-01
-3.20086509e-01 3.56519341e-01 9.63562906e-01 1.14279211e+00
-8.07381809e-01 -1.33608305e+00 -4.23762918e-01 1.16675782e+00
-3.54564279e-01 -2.39388242e-01 -3.51541877e-01 5.09236813e-01
2.43421704e-01 7.41938174e-01 3.67988169e-01 -6.05490029e-01
1.24936260e-01 -1.91296950e-01 9.13717687e-01 -2.35539943e-01
-3.05830061e-01 -5.07753193e-01 -1.07961975e-01 -1.63948968e-01
-5.37711620e-01 -9.14806426e-01 -1.38004076e+00 -6.99983776e-01
-4.66240942e-01 7.84323364e-02 8.41041565e-01 8.03881705e-01
3.16532820e-01 -2.21046731e-01 5.87800205e-01 -4.80572104e-01
-8.40398133e-01 -8.92447948e-01 -8.95254493e-01 2.74709344e-01
2.26213545e-01 -3.58796299e-01 -5.10009468e-01 2.83961445e-01] | [7.258962154388428, 3.812659740447998] |
82ad8a97-c28f-4223-81da-05e88d1d9b7d | uv-gan-adversarial-facial-uv-map-completion | 1712.04695 | null | http://arxiv.org/abs/1712.04695v1 | http://arxiv.org/pdf/1712.04695v1.pdf | UV-GAN: Adversarial Facial UV Map Completion for Pose-invariant Face Recognition | Recently proposed robust 3D face alignment methods establish either dense or
sparse correspondence between a 3D face model and a 2D facial image. The use of
these methods presents new challenges as well as opportunities for facial
texture analysis. In particular, by sampling the image using the fitted model,
a facial UV can be created. Unfortunately, due to self-occlusion, such a UV map
is always incomplete. In this paper, we propose a framework for training Deep
Convolutional Neural Network (DCNN) to complete the facial UV map extracted
from in-the-wild images. To this end, we first gather complete UV maps by
fitting a 3D Morphable Model (3DMM) to various multiview image and video
datasets, as well as leveraging on a new 3D dataset with over 3,000 identities.
Second, we devise a meticulously designed architecture that combines local and
global adversarial DCNNs to learn an identity-preserving facial UV completion
model. We demonstrate that by attaching the completed UV to the fitted mesh and
generating instances of arbitrary poses, we can increase pose variations for
training deep face recognition/verification models, and minimise pose
discrepancy during testing, which lead to better performance. Experiments on
both controlled and in-the-wild UV datasets prove the effectiveness of our
adversarial UV completion model. We achieve state-of-the-art verification
accuracy, $94.05\%$, under the CFP frontal-profile protocol only by combining
pose augmentation during training and pose discrepancy reduction during
testing. We will release the first in-the-wild UV dataset (we refer as WildUV)
that comprises of complete facial UV maps from 1,892 identities for research
purposes. | ['Yuxiang Zhou', 'Shiyang Cheng', 'Niannan Xue', 'Stefanos Zafeiriou', 'Jiankang Deng'] | 2017-12-13 | uv-gan-adversarial-facial-uv-map-completion-1 | http://openaccess.thecvf.com/content_cvpr_2018/html/Deng_UV-GAN_Adversarial_Facial_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Deng_UV-GAN_Adversarial_Facial_CVPR_2018_paper.pdf | cvpr-2018-6 | ['robust-face-recognition'] | ['computer-vision'] | [ 3.49566847e-01 2.30275750e-01 1.86492592e-01 -6.40813529e-01
-9.07841504e-01 -5.17964184e-01 4.58028823e-01 -7.53885210e-01
5.96913062e-02 3.18051368e-01 -3.34594816e-01 1.23022676e-01
2.08524078e-01 -7.71065712e-01 -1.15768695e+00 -5.75846076e-01
1.36945173e-01 5.20323694e-01 -4.33926344e-01 -2.71966726e-01
-8.58556181e-02 8.64939153e-01 -1.83920658e+00 1.15655832e-01
4.12609905e-01 1.43276203e+00 -4.09110665e-01 3.41522247e-01
1.42963216e-01 5.32610379e-02 -4.62730765e-01 -9.08887029e-01
8.34367454e-01 -2.47934654e-01 -6.27258897e-01 3.98554862e-01
1.50000978e+00 -7.18779624e-01 -1.85773984e-01 8.84739876e-01
6.65916860e-01 -1.73214912e-01 5.86395383e-01 -1.40243304e+00
-5.71423709e-01 -1.95335209e-01 -7.25725770e-01 -5.92156351e-01
5.54317772e-01 5.95592968e-02 5.84351540e-01 -1.21283424e+00
1.11628222e+00 1.30961490e+00 9.80118155e-01 1.00748038e+00
-1.47518384e+00 -9.49746132e-01 -9.16097537e-02 -2.87013263e-01
-1.61322868e+00 -8.77681971e-01 1.08665860e+00 -2.83006012e-01
5.67666352e-01 2.02269539e-01 5.58501363e-01 1.48131311e+00
1.00119300e-01 3.13129425e-01 1.17963111e+00 -4.07699585e-01
-2.46524196e-02 -1.92271382e-01 -4.77482021e-01 1.05989623e+00
-1.05088025e-01 2.20527604e-01 -6.39673531e-01 -3.88072520e-01
9.04667258e-01 -8.00512359e-02 -2.24830002e-01 -6.11177623e-01
-5.49386203e-01 4.69279170e-01 2.46417716e-01 -3.63224268e-01
-1.18633926e-01 1.05095707e-01 2.33684987e-01 4.08058971e-01
7.39689350e-01 1.57648325e-01 -3.86650115e-01 3.26029986e-01
-8.41416895e-01 5.27170420e-01 4.18605536e-01 1.08547091e+00
1.11419940e+00 1.47264734e-01 1.86521262e-01 1.04493713e+00
2.67373830e-01 8.77348781e-01 -7.81713054e-02 -1.13505840e+00
2.92454153e-01 3.94149572e-01 -2.16445386e-01 -1.02370489e+00
-8.62004235e-02 -5.65111823e-03 -7.96301544e-01 5.72958112e-01
2.85834700e-01 4.86120917e-02 -1.22153437e+00 2.15812659e+00
7.08947003e-01 2.45955288e-01 -2.38888010e-01 8.42019558e-01
8.68347168e-01 7.42142946e-02 -3.18810314e-01 -1.26086250e-01
1.20817637e+00 -4.91070062e-01 -4.64354277e-01 -5.97889908e-02
2.35300884e-01 -9.06286478e-01 8.96858573e-01 3.28609556e-01
-1.17064822e+00 -5.43191671e-01 -9.97115314e-01 -1.08186468e-01
-1.55820414e-01 1.29719693e-02 4.17364150e-01 7.44984090e-01
-1.31622648e+00 6.84192061e-01 -7.27752268e-01 -1.61797434e-01
8.95544887e-01 6.32960677e-01 -1.06746316e+00 -3.04703027e-01
-7.72350550e-01 5.13715982e-01 -3.81957918e-01 3.36305648e-01
-1.12733793e+00 -9.12220776e-01 -1.17490757e+00 -4.91655529e-01
2.16414034e-01 -6.86684906e-01 9.06423867e-01 -1.10748887e+00
-1.48464561e+00 1.42116737e+00 -2.90775537e-01 2.35220104e-01
7.84687281e-01 -3.28979269e-02 -3.62974465e-01 9.53774899e-02
-1.79104973e-03 6.77727878e-01 1.38930845e+00 -1.44545388e+00
-1.71280857e-02 -7.68984258e-01 -1.98600106e-02 -1.74751267e-01
-2.06366718e-01 6.14202619e-02 -7.02682257e-01 -6.27633810e-01
2.62576848e-01 -9.72830832e-01 2.64709353e-01 6.25149131e-01
-2.58453578e-01 1.90887630e-01 9.54631388e-01 -8.95803034e-01
4.30953443e-01 -2.16213083e+00 1.23778746e-01 5.58511674e-01
2.35520959e-01 1.63790300e-01 -5.22633493e-01 5.25357574e-02
-2.63558447e-01 5.63772395e-02 -1.82937726e-01 -1.08338559e+00
-3.52777541e-02 2.82721996e-01 -1.77223802e-01 7.56841600e-01
5.24843097e-01 8.69958937e-01 -3.75026494e-01 -3.11969638e-01
-2.72549165e-04 8.21095645e-01 -8.46506357e-01 2.79581845e-01
-1.15055293e-01 5.39037228e-01 -1.40648991e-01 1.25245690e+00
1.24123847e+00 2.03785747e-01 1.53840512e-01 -4.45957720e-01
4.32956696e-01 -2.32653990e-01 -1.14834917e+00 1.92705202e+00
-4.52678323e-01 3.45076412e-01 6.18314445e-01 -7.07735658e-01
1.12914670e+00 2.25756884e-01 5.47815025e-01 -6.83972180e-01
3.52019459e-01 3.82929593e-01 -6.63002253e-01 -1.77319840e-01
1.88792467e-01 -5.09927757e-02 1.94793880e-01 3.93028080e-01
4.07723278e-01 -2.22482279e-01 -3.51140797e-01 -3.97183895e-01
6.88748121e-01 5.27632177e-01 -2.43910030e-01 -1.80918589e-01
4.59300071e-01 -5.88063002e-01 5.67838609e-01 2.29484618e-01
-1.13468446e-01 9.92833674e-01 4.60179895e-01 -6.89407110e-01
-1.17712212e+00 -1.19115961e+00 -4.88529861e-01 5.97701430e-01
-1.16425738e-01 -2.45918274e-01 -1.09828901e+00 -7.61636674e-01
2.14765653e-01 -7.39436150e-02 -1.04094410e+00 -9.56300795e-02
-7.56248295e-01 -3.61537457e-01 6.88545167e-01 2.75997698e-01
5.67635000e-01 -6.78317249e-01 5.65398894e-02 -2.69644558e-01
1.06684089e-01 -1.31579876e+00 -6.93153739e-01 -3.93064111e-01
-4.88954067e-01 -1.34136450e+00 -5.42181611e-01 -8.38854074e-01
1.07180941e+00 -4.27187122e-02 1.02511644e+00 2.49955371e-01
-4.12319541e-01 4.96582121e-01 -3.68083119e-02 -2.59040922e-01
-5.50769329e-01 -2.16358632e-01 4.52792406e-01 4.89474803e-01
-1.17707765e-02 -8.48925769e-01 -5.64305484e-01 5.80569983e-01
-8.05896819e-01 -8.53915513e-02 1.50267363e-01 9.93800163e-01
8.19545329e-01 -5.33242345e-01 3.03326339e-01 -7.57635951e-01
1.41881481e-01 2.55608689e-02 -7.57642150e-01 1.35366589e-01
-3.31375867e-01 -2.28775218e-01 3.90225798e-01 -3.61675471e-01
-8.55158627e-01 2.06891164e-01 -5.08375466e-01 -1.12048626e+00
-9.35620368e-02 -4.63314056e-02 -5.96840620e-01 -8.36087763e-01
7.29941726e-01 -9.08860937e-02 7.46163309e-01 -4.48871881e-01
2.33089760e-01 5.20357907e-01 6.20386243e-01 -8.84248257e-01
1.18324649e+00 6.72648787e-01 3.76490474e-01 -7.08679497e-01
-5.54723799e-01 1.95576027e-01 -7.37275779e-01 -3.74084353e-01
6.13687932e-01 -8.78308415e-01 -8.17910850e-01 7.42536306e-01
-1.08083713e+00 -3.00678551e-01 -4.86957021e-02 -1.82818878e-03
-6.54866934e-01 2.54318535e-01 -4.38561827e-01 -5.50668299e-01
-5.00921726e-01 -1.36317515e+00 1.64118075e+00 -1.08375326e-01
-7.79149607e-02 -6.48075342e-01 -3.98912840e-02 5.15539587e-01
2.03055188e-01 8.08223128e-01 5.30474842e-01 -6.81343228e-02
-4.85624522e-01 -3.93903702e-01 -2.10051574e-02 6.08765423e-01
2.69352168e-01 1.03409521e-01 -1.39376092e+00 -4.08626676e-01
-1.59981355e-01 -6.93590224e-01 4.58175868e-01 -6.53515710e-03
1.23762739e+00 -2.70205706e-01 -1.24033550e-02 1.27026355e+00
1.27231944e+00 -3.16150755e-01 6.78445101e-01 2.24076193e-02
8.73503208e-01 7.19606400e-01 4.11660314e-01 3.15936983e-01
9.66485590e-02 9.79703307e-01 7.32234597e-01 -1.48113057e-01
-2.49391556e-01 -3.39066982e-01 2.73326516e-01 2.99322188e-01
-4.16976392e-01 4.59620893e-01 -6.70914233e-01 1.50437444e-01
-1.20352352e+00 -7.67800987e-01 3.83668154e-01 2.40164709e+00
9.27978873e-01 -2.47130021e-01 -8.37461129e-02 -9.36376303e-02
5.11418819e-01 1.75693378e-01 -3.42694849e-01 -3.27388555e-01
-1.23456500e-01 7.83420205e-01 1.74490690e-01 5.95488787e-01
-1.11837959e+00 9.73372340e-01 5.72224903e+00 7.82658756e-01
-1.22703469e+00 -2.76876893e-03 7.22139716e-01 -2.15658545e-01
-3.09272468e-01 -4.10328001e-01 -8.52661490e-01 7.65318573e-02
5.49555480e-01 3.71049136e-01 6.25523448e-01 9.25897717e-01
-4.45099249e-02 3.86216193e-01 -1.17420316e+00 1.21537662e+00
4.29902852e-01 -1.37876463e+00 2.28940561e-01 4.12095010e-01
9.03822899e-01 -3.10490429e-01 2.35963836e-01 3.38028632e-02
-1.78828880e-01 -1.22968650e+00 8.37360442e-01 4.56083328e-01
1.61572707e+00 -8.52131069e-01 4.88401115e-01 -1.03058286e-01
-1.13390827e+00 3.93691093e-01 -3.32632601e-01 4.00147557e-01
-1.22516409e-01 2.54720956e-01 -5.80079615e-01 7.33292758e-01
7.74994671e-01 5.40836930e-01 -4.92707580e-01 4.08115119e-01
-4.01856303e-02 1.99687947e-02 -3.28921854e-01 5.93381524e-01
-3.07008833e-01 -3.52208018e-01 3.86779338e-01 6.25611484e-01
3.44881177e-01 -1.09613508e-01 -1.02809183e-01 1.03979266e+00
-5.60094595e-01 5.04849777e-02 -7.77975023e-01 3.60436469e-01
3.60821933e-01 1.25681555e+00 -1.22559793e-01 1.07025445e-01
-2.66022831e-01 1.13847756e+00 4.77921873e-01 2.85368413e-01
-7.74603724e-01 1.91864789e-01 1.34118152e+00 2.08474398e-01
5.72456829e-02 -1.50210366e-01 -4.53204550e-02 -1.07491803e+00
4.10289139e-01 -1.08737874e+00 -1.07891858e-01 -6.66709840e-01
-1.35565829e+00 6.86984420e-01 -1.52814999e-01 -1.12609494e+00
-3.26785952e-01 -6.95593953e-01 -3.34256202e-01 1.14520180e+00
-1.40250325e+00 -1.72083402e+00 -6.31485581e-01 8.09189200e-01
1.30857378e-01 -1.95256233e-01 1.17491686e+00 4.72846985e-01
-5.08691072e-01 1.27114260e+00 -2.87491113e-01 3.51734638e-01
9.53762412e-01 -6.78506076e-01 5.96589029e-01 5.17035484e-01
7.03050569e-02 5.82546830e-01 3.21583927e-01 -4.66392726e-01
-1.89809752e+00 -1.27862537e+00 4.86817926e-01 -7.85399497e-01
1.48983434e-01 -8.30413580e-01 -7.73061633e-01 7.89187849e-01
-2.04210058e-01 5.89091122e-01 5.12946308e-01 -1.57522216e-01
-8.17082942e-01 -3.42278361e-01 -1.57762945e+00 7.50507116e-01
1.52550137e+00 -7.72456884e-01 -7.73405954e-02 2.45990485e-01
4.37103242e-01 -8.95465493e-01 -1.13640916e+00 7.36583650e-01
8.62715065e-01 -8.70690465e-01 1.13458753e+00 -5.86173236e-01
4.72671509e-01 -8.50193799e-02 -3.32979083e-01 -1.10203934e+00
1.88792765e-01 -8.82366359e-01 -5.79053536e-02 1.32353282e+00
2.01954752e-01 -6.83815241e-01 1.08786309e+00 6.87742233e-01
-1.49660036e-01 -8.93120527e-01 -1.19454169e+00 -6.74443245e-01
5.35592586e-02 -4.89040762e-01 9.02321100e-01 1.00541437e+00
-7.31337130e-01 -2.28524566e-01 -4.09948707e-01 3.38672578e-01
8.91585410e-01 -9.88118798e-02 1.31936657e+00 -1.04307604e+00
-5.30244745e-02 -2.89202601e-01 -6.54487193e-01 -6.25791848e-01
6.30779922e-01 -7.91947484e-01 -1.43043533e-01 -8.68819058e-01
-7.71452114e-02 -3.96608979e-01 3.59071076e-01 6.76215768e-01
9.92529243e-02 8.00258040e-01 -7.62277395e-02 2.18391176e-02
2.31424734e-01 6.44405425e-01 1.42687142e+00 -1.47711694e-01
7.48986974e-02 -2.03099251e-01 -3.42366338e-01 5.32553077e-01
4.44322467e-01 3.49489762e-03 -3.18032086e-01 -4.20371711e-01
-3.51560079e-02 -9.72477868e-02 6.71987832e-01 -7.08008409e-01
-2.52709955e-01 -1.12746805e-02 6.45645738e-01 -1.66537225e-01
7.84780741e-01 -8.65613699e-01 4.01639938e-01 -1.09819405e-01
8.43262672e-02 -7.11541921e-02 4.51886207e-01 3.04628581e-01
-4.21283627e-03 2.03269914e-01 9.31181431e-01 1.43112332e-01
-4.36478734e-01 9.14709926e-01 4.54385191e-01 -1.49736861e-02
9.63411212e-01 -3.44039083e-01 -2.34600693e-01 -2.19763294e-01
-6.30569696e-01 -1.60933137e-01 9.27039742e-01 4.82397735e-01
7.44801879e-01 -1.66189969e+00 -8.05628955e-01 1.07274139e+00
3.13360661e-01 2.07875445e-01 3.04672182e-01 6.21150196e-01
-7.40358412e-01 -2.58250922e-01 -3.55594337e-01 -7.50321388e-01
-1.55487216e+00 2.82795876e-01 7.29680836e-01 3.19746137e-01
-5.16055822e-01 9.64468122e-01 1.22013271e-01 -9.48157132e-01
9.14538056e-02 2.25151554e-01 3.05474550e-01 -7.52966627e-02
5.13544202e-01 -7.30115622e-02 5.20352900e-01 -9.61338460e-01
-3.30482841e-01 1.07884109e+00 1.66959539e-01 -6.43743053e-02
1.25216174e+00 1.77279055e-01 -2.98133492e-01 -3.22633862e-01
1.73120666e+00 2.07453400e-01 -1.49647236e+00 -1.29041746e-01
-7.24520922e-01 -7.99407005e-01 -1.87203601e-01 -3.33276182e-01
-1.46521032e+00 6.95879519e-01 6.39487028e-01 -4.55057621e-01
1.10341692e+00 -8.66753682e-02 6.47459805e-01 -3.01923417e-02
6.24449134e-01 -8.05788815e-01 2.42091808e-02 2.86740094e-01
1.06960273e+00 -1.15894604e+00 -1.56810626e-01 -7.11168051e-01
-1.31036818e-01 1.08230913e+00 6.82567000e-01 -1.64898396e-01
7.73767531e-01 2.67473191e-01 2.18076929e-01 -2.32192591e-01
-3.50362033e-01 2.27145746e-01 4.32251871e-01 8.81015360e-01
9.07952711e-02 -1.87732428e-01 3.17496240e-01 3.10560435e-01
-4.70169157e-01 -1.50669187e-01 1.02799900e-01 8.90633702e-01
2.18857914e-01 -1.31235349e+00 -6.08543336e-01 1.55059293e-01
-3.24944556e-01 1.01281688e-01 -5.14807343e-01 9.44565415e-01
1.47253439e-01 5.23108900e-01 1.54884934e-01 -5.95210373e-01
5.25016546e-01 2.13247791e-01 1.02219784e+00 -3.31440806e-01
-3.06456238e-01 -2.08689496e-02 1.19243778e-01 -1.04367018e+00
-2.36999780e-01 -7.51391232e-01 -7.30558455e-01 -6.36554897e-01
1.34155462e-02 -4.81774569e-01 7.66427755e-01 7.28681207e-01
5.39954901e-01 6.89047575e-02 9.83690321e-01 -1.32055295e+00
-5.58642566e-01 -7.71755815e-01 -5.26130080e-01 6.58058763e-01
3.69540691e-01 -9.56086516e-01 -4.69915360e-01 -5.17694838e-03] | [13.081192970275879, 0.04916268214583397] |
7d0bedd1-ab14-4f79-88c6-9dfc1b4bb549 | an-energy-based-model-for-neuro-symbolic | 2110.01639 | null | https://arxiv.org/abs/2110.01639v1 | https://arxiv.org/pdf/2110.01639v1.pdf | An energy-based model for neuro-symbolic reasoning on knowledge graphs | Machine learning on graph-structured data has recently become a major topic in industry and research, finding many exciting applications such as recommender systems and automated theorem proving. We propose an energy-based graph embedding algorithm to characterize industrial automation systems, integrating knowledge from different domains like industrial automation, communications and cybersecurity. By combining knowledge from multiple domains, the learned model is capable of making context-aware predictions regarding novel system events and can be used to evaluate the severity of anomalies that might be indicative of, e.g., cybersecurity breaches. The presented model is mappable to a biologically-inspired neural architecture, serving as a first bridge between graph embedding methods and neuromorphic computing - uncovering a promising edge application for this upcoming technology. | ['Josep Soler Garrido', 'Dominik Dold'] | 2021-10-04 | null | null | null | null | ['automated-theorem-proving', 'automated-theorem-proving'] | ['miscellaneous', 'reasoning'] | [ 4.83228445e-01 4.40582335e-01 -2.62302935e-01 1.02593973e-01
5.22468537e-02 -5.46099961e-01 6.69025540e-01 1.03201544e+00
2.20081881e-01 5.16797245e-01 -1.94678470e-01 -8.21981847e-01
-6.03765070e-01 -1.13401890e+00 -5.56382596e-01 -6.85238898e-01
-5.53409278e-01 2.06837550e-01 1.44973904e-01 -2.75435388e-01
3.39285553e-01 7.54389226e-01 -1.20352113e+00 -8.59765485e-02
3.98193747e-01 1.14441156e+00 -1.41135320e-01 2.83617109e-01
2.54968733e-01 5.45025170e-01 -5.95949888e-01 -1.14385381e-01
2.44327588e-03 -4.26132709e-01 -5.12863934e-01 -3.32357049e-01
-2.08757922e-01 4.22517061e-01 -5.62838793e-01 1.40510094e+00
6.10364303e-02 -3.32331285e-02 3.08699220e-01 -1.55272210e+00
-3.69380474e-01 6.48308516e-01 -1.78829417e-01 2.90028483e-01
3.39379877e-01 2.97955602e-01 9.17461514e-01 -1.83075711e-01
6.17818415e-01 1.01419449e+00 4.34103578e-01 4.11851645e-01
-1.33147609e+00 -4.91746604e-01 2.12542579e-01 7.71843433e-01
-7.78057098e-01 2.19725534e-01 1.19441509e+00 -4.14525121e-01
9.83368039e-01 1.58033565e-01 9.75088418e-01 1.02642405e+00
9.34948742e-01 2.75316000e-01 9.38073337e-01 -3.51496428e-01
8.79323065e-01 2.26673074e-02 2.37690121e-01 7.91764796e-01
7.14085102e-01 3.61706495e-01 -2.21909881e-01 -3.68284971e-01
3.92515838e-01 2.74614722e-01 -2.19216093e-01 -3.50844294e-01
-1.24650836e+00 5.94124317e-01 6.81110084e-01 6.06589377e-01
-3.99440736e-01 5.67794263e-01 6.63852394e-01 4.60964441e-01
1.66048333e-01 9.40921307e-01 -4.21999872e-01 -1.51740725e-03
-2.73372650e-01 -1.04660146e-01 7.92809129e-01 6.29208565e-01
8.45557570e-01 4.16620493e-01 4.08314914e-01 8.21662135e-03
2.90107816e-01 2.43780315e-01 3.04016560e-01 -4.25443411e-01
-3.25169750e-02 1.13722837e+00 -5.72497189e-01 -1.08819020e+00
-5.99233925e-01 -3.08024913e-01 -8.72504413e-01 4.52225626e-01
-1.03049003e-01 -5.88009655e-02 -2.82909691e-01 1.43157375e+00
2.66889840e-01 7.08258808e-01 -8.49954560e-02 6.17575467e-01
2.66861692e-02 5.61387122e-01 -2.02509403e-01 -1.79701686e-01
1.37670207e+00 -2.43661880e-01 -4.87098306e-01 -9.13285837e-02
5.75168133e-01 5.74895404e-02 3.75076383e-01 7.48823762e-01
-3.89562279e-01 -2.86561489e-01 -1.53571999e+00 6.32846713e-01
-8.22229922e-01 -6.42782092e-01 7.62441039e-01 6.94594741e-01
-8.26447368e-01 1.06083179e+00 -8.48913789e-01 -4.77350950e-01
3.19516420e-01 5.15311301e-01 -2.01147839e-01 -7.28653893e-02
-1.24447060e+00 9.26748574e-01 6.19638145e-01 -7.36552328e-02
-1.01874804e+00 -4.49154466e-01 -8.33748758e-01 1.92814827e-01
5.50905168e-01 -6.06720746e-01 4.69121337e-01 -3.07368666e-01
-1.25706017e+00 1.96292207e-01 4.84514534e-01 -1.00570703e+00
-2.04703808e-01 3.38823169e-01 -1.02321088e+00 1.09325293e-02
-6.18794203e-01 -1.99214026e-01 9.25912976e-01 -6.63405061e-01
-2.66628057e-01 -8.02433193e-01 2.64589220e-01 -6.14401877e-01
-6.21412098e-01 -4.70089853e-01 4.85737145e-01 -3.92133951e-01
-5.35740778e-02 -9.20406163e-01 -4.63913321e-01 -9.81132761e-02
-5.05298436e-01 -1.89152211e-01 1.11781955e+00 -2.51287699e-01
1.20699084e+00 -1.90901470e+00 4.53525096e-01 5.26318491e-01
3.63439143e-01 2.89352149e-01 -1.31163523e-01 9.06379521e-01
-2.33443558e-01 1.17689811e-01 -3.73676270e-02 6.87818229e-01
-4.64544147e-02 -1.20950788e-01 -4.08449262e-01 6.16246402e-01
6.00418985e-01 8.74447048e-01 -1.06527519e+00 2.09453866e-01
6.54012978e-01 3.17035735e-01 -1.09123014e-01 -1.60503328e-01
-4.30294752e-01 3.55244070e-01 -7.42798567e-01 5.57805121e-01
5.57496175e-02 -1.53067619e-01 5.36649823e-01 -2.86367118e-01
2.13938266e-01 9.84011441e-02 -7.23339677e-01 1.62534106e+00
-6.58974171e-01 8.33673000e-01 -2.10406497e-01 -1.36669540e+00
1.14424670e+00 1.23707511e-01 7.17196226e-01 -7.10604846e-01
5.08520126e-01 -1.82391644e-01 2.11835414e-01 -2.22956374e-01
-5.32851033e-02 6.33356795e-02 -3.13460261e-01 4.03056234e-01
1.30238548e-01 -2.21406743e-01 5.76668531e-02 6.79635778e-02
1.87802303e+00 -4.13113296e-01 4.76045668e-01 -3.73954505e-01
6.71205223e-01 -1.47099122e-01 3.61194044e-01 9.05291066e-02
4.64616111e-03 -3.78338456e-01 6.51105106e-01 -5.47943115e-01
-8.14444363e-01 -1.02630651e+00 7.14589506e-02 1.81083128e-01
3.87020707e-01 -5.64829946e-01 -6.95644259e-01 -8.82038713e-01
4.09600079e-01 9.04279828e-01 -4.64384794e-01 -1.20245159e+00
-1.39797643e-01 -3.42125952e-01 3.43450159e-01 4.63818848e-01
-2.22752616e-01 -8.53693187e-01 -6.71756029e-01 4.41190928e-01
6.27652228e-01 -9.52585518e-01 2.20992684e-01 6.85304463e-01
-1.12797248e+00 -1.40031517e+00 1.78745672e-01 -4.60356563e-01
6.06471598e-01 -1.51952595e-01 9.68146086e-01 9.38966721e-02
-9.29450393e-01 5.04814863e-01 -2.79475152e-01 -5.67127943e-01
-8.64389420e-01 -2.49225423e-01 5.46817064e-01 -3.32341641e-02
2.73642331e-01 -8.98031533e-01 -4.06373203e-01 2.58417785e-01
-9.35970783e-01 -5.07426620e-01 6.00469768e-01 7.05391407e-01
5.80700040e-01 5.24956286e-01 1.09065855e+00 -7.68506587e-01
9.05875742e-01 -7.61689365e-01 -1.14249706e+00 1.92442670e-01
-1.21297514e+00 3.58810186e-01 1.13299882e+00 -4.13904995e-01
-3.49005610e-01 -4.45020385e-02 1.50083780e-01 -5.27132869e-01
-2.69666553e-01 6.00908637e-01 -4.96858299e-01 -2.62806565e-01
5.28990865e-01 3.02432403e-02 2.30675675e-02 -1.44476935e-01
3.55374932e-01 2.84299135e-01 2.84875274e-01 -4.18197930e-01
9.60329056e-01 1.42037645e-01 6.86054766e-01 -8.58857393e-01
-2.92212646e-02 -3.59070867e-01 -1.14724211e-01 -5.15887856e-01
6.40564859e-01 -2.99666256e-01 -1.24092579e+00 -5.60952611e-02
-9.34291422e-01 1.17165864e-01 -5.87613583e-01 3.77804905e-01
-5.20540357e-01 2.79008925e-01 -4.04204756e-01 -7.04181731e-01
-2.95946836e-01 -8.44189703e-01 6.47384346e-01 1.27540082e-01
-2.37913966e-01 -1.25383496e+00 4.26537871e-01 -1.33567199e-01
1.87466502e-01 5.90273321e-01 1.50137532e+00 -1.02371693e+00
-6.28446758e-01 -6.55496776e-01 2.33608916e-01 3.94228548e-01
3.15067440e-01 -2.37471610e-01 -7.67796576e-01 -5.18781483e-01
5.72781451e-02 -5.89518920e-02 5.17329276e-01 -2.61413511e-02
1.09472787e+00 -2.37429470e-01 -6.81005120e-01 8.71959552e-02
1.49623847e+00 4.05432850e-01 6.01207912e-01 1.90349072e-02
3.46330732e-01 3.90365034e-01 4.53281313e-01 4.85512078e-01
-2.48618782e-01 5.97978473e-01 1.12145770e+00 3.24862570e-01
2.13809013e-01 -2.00710893e-01 6.18035674e-01 6.15864277e-01
9.62864421e-03 -3.02886456e-01 -7.93763697e-01 2.11973622e-01
-1.74230754e+00 -9.63538110e-01 4.31387722e-02 2.33468771e+00
6.77286685e-02 4.75859940e-01 5.12294956e-02 5.16982317e-01
6.85911357e-01 -4.79867123e-02 -1.08609486e+00 -6.74936175e-01
3.01331550e-01 3.13292563e-01 4.55089867e-01 -1.16003841e-01
-6.32812142e-01 3.88609767e-01 5.63625193e+00 3.98689568e-01
-1.05784953e+00 -1.24062613e-01 3.65957856e-01 3.18315268e-01
-3.57494116e-01 1.08610153e-01 -1.33926317e-01 3.68252933e-01
1.43720841e+00 -7.45387018e-01 6.83565855e-01 8.35564137e-01
-1.18712284e-01 1.46463916e-01 -1.56606829e+00 7.79731989e-01
-6.36186376e-02 -1.33804321e+00 -6.75715655e-02 4.52481925e-01
3.30574691e-01 -1.50350397e-02 -1.58894937e-02 6.24139048e-02
1.58126935e-01 -7.69041836e-01 3.04583639e-01 4.01900530e-01
4.43033010e-01 -9.62335765e-01 6.70195401e-01 2.11706206e-01
-1.14490926e+00 -4.50985998e-01 -2.35026717e-01 -1.13961793e-01
1.90475792e-01 9.29645360e-01 -1.21732855e+00 8.35137427e-01
4.03953910e-01 8.01950276e-01 -3.52966160e-01 8.58205497e-01
-1.39078379e-01 5.96197128e-01 -1.36264756e-01 -4.24405485e-01
-1.36460006e-01 -2.78607100e-01 8.60239625e-01 6.97086692e-01
5.50892174e-01 -2.75502205e-01 2.34926697e-02 1.12779081e+00
-1.04581155e-01 -4.68207300e-01 -1.23098707e+00 -7.87044823e-01
3.82588327e-01 1.44205415e+00 -1.10900617e+00 9.13320482e-02
-2.86221057e-01 7.43740976e-01 -1.34572148e-01 1.79344326e-01
-8.35273266e-01 -5.79373121e-01 9.39977944e-01 4.04341780e-02
9.80172232e-02 -2.51837075e-01 -1.92237318e-01 -8.36986601e-01
-2.67560512e-01 -6.21207535e-01 1.45759523e-01 -3.54206294e-01
-1.14741683e+00 5.82982779e-01 -4.24607873e-01 -1.44923639e+00
-3.87334347e-01 -1.10709643e+00 -7.78848529e-01 2.72457272e-01
-1.41934931e+00 -5.52113950e-01 -1.91118624e-02 4.16172624e-01
1.47909716e-01 -4.61971998e-01 9.98968720e-01 6.95938338e-03
-5.83185136e-01 1.54030547e-01 -1.52842933e-03 -2.76679993e-01
1.53210238e-01 -1.23751009e+00 6.51928067e-01 6.85363948e-01
6.64032102e-01 3.39999914e-01 8.72994900e-01 -6.66357815e-01
-2.24433661e+00 -1.30704904e+00 -2.14347169e-02 -3.56870353e-01
1.23943031e+00 -3.49889338e-01 -6.43488407e-01 3.53498518e-01
1.13203272e-01 1.01105362e-01 4.82042760e-01 5.66075109e-02
-2.53821820e-01 -4.06730950e-01 -1.21413267e+00 4.94568676e-01
9.01305974e-01 -6.51291907e-01 -3.74907643e-01 4.06830072e-01
7.93117940e-01 2.43540466e-01 -1.22010326e+00 2.96290010e-01
2.32116997e-01 -6.51161432e-01 9.22325730e-01 -7.20233142e-01
7.99064860e-02 -2.41830945e-01 1.24587774e-01 -1.55050504e+00
-4.44123268e-01 -7.27557361e-01 -7.06492305e-01 8.90087008e-01
2.74388760e-01 -9.91230786e-01 6.77105725e-01 1.43581346e-01
-2.21667215e-01 -7.94674814e-01 -1.08274162e+00 -1.15359735e+00
-2.18570858e-01 -4.36273664e-01 6.49374843e-01 1.02095604e+00
7.43566632e-01 3.48712564e-01 2.31457084e-01 3.18299472e-01
4.67519909e-01 1.09209038e-01 2.34528184e-01 -1.75822711e+00
-3.56955349e-01 -6.40397906e-01 -1.30393517e+00 -1.62513450e-01
3.84447396e-01 -1.18345463e+00 -2.85356790e-01 -1.25216615e+00
-2.84609646e-01 -7.03886449e-02 -1.01422274e+00 3.94797117e-01
5.29188812e-01 -6.13506474e-02 1.59511883e-02 -6.03406072e-01
-3.77046883e-01 4.79581535e-01 6.51762128e-01 -6.79449499e-01
8.80609527e-02 1.51742429e-01 -3.99357408e-01 5.24999321e-01
8.40938687e-01 -4.26365137e-01 -5.11025488e-01 2.82219142e-01
5.12118757e-01 2.14527428e-01 5.20040691e-01 -1.37926865e+00
2.38441169e-01 -8.23027343e-02 -1.17226854e-01 -4.10319380e-02
5.32396026e-02 -1.50349820e+00 3.10576797e-01 9.33185399e-01
-1.63912579e-01 1.77595332e-01 4.47755575e-01 1.27546668e+00
-2.03195393e-01 -4.34841104e-02 2.88654596e-01 3.43264192e-01
-9.54873741e-01 4.21045840e-01 -5.12301087e-01 -3.30548972e-01
1.51387048e+00 -1.50397643e-02 -4.66365218e-01 -1.98604077e-01
-5.26849627e-01 1.07500955e-01 4.80856448e-01 9.28234279e-01
8.23418140e-01 -1.19899571e+00 -2.52156973e-01 1.32032081e-01
3.79001647e-01 -6.91495419e-01 6.12036176e-02 5.32787919e-01
-1.27558783e-01 3.14956367e-01 -3.60846162e-01 -5.24694562e-01
-9.03573513e-01 1.04382861e+00 -5.83394393e-02 -2.63810992e-01
-6.22254848e-01 1.96041122e-01 -3.07987124e-01 6.50767284e-03
2.66471552e-03 -5.61061323e-01 9.98255610e-02 -3.02077204e-01
3.57776195e-01 6.50817454e-01 4.13953215e-01 9.49643552e-02
-7.73061097e-01 2.60401994e-01 2.90407479e-01 4.08145756e-01
1.39380145e+00 3.09873432e-01 -3.39610785e-01 6.84737146e-01
9.07634914e-01 -4.11175728e-01 -8.01249325e-01 4.78591621e-02
4.91748899e-01 7.78804049e-02 3.38059723e-01 -9.97828424e-01
-9.98863995e-01 1.06854796e+00 7.87288249e-01 8.22294831e-01
1.28696442e+00 7.25457445e-02 6.50127172e-01 1.00634575e+00
1.11178744e+00 -1.01452935e+00 7.91198239e-02 2.59478182e-01
6.19647980e-01 -7.69864857e-01 6.35361671e-02 -1.16112836e-01
-9.67719555e-02 1.41285861e+00 2.69859046e-01 -2.14423373e-01
8.62691760e-01 5.38948119e-01 -7.31057942e-01 -3.35633636e-01
-8.69097650e-01 1.50393769e-01 2.55019754e-01 7.45571613e-01
-6.04169741e-02 2.25441977e-01 2.40038782e-02 5.88866770e-01
3.38244945e-01 -7.22030029e-02 5.63104630e-01 8.81238818e-01
-3.60118628e-01 -1.30181384e+00 5.93455322e-02 6.16460800e-01
3.64821516e-02 2.32142299e-01 -6.69280469e-01 5.77668428e-01
-6.54453859e-02 7.43724704e-01 -1.15482457e-01 -1.11127877e+00
3.36873084e-01 4.97922599e-02 5.12252569e-01 -6.86016560e-01
-4.29091752e-01 -5.77435434e-01 -1.93917993e-02 -9.17198658e-01
3.82960401e-02 -4.83773798e-01 -1.30854154e+00 -1.68483630e-02
-6.70025229e-01 4.46446687e-02 9.67756987e-01 8.26348841e-01
8.35190594e-01 1.10495412e+00 8.31990659e-01 -6.99194133e-01
-3.53001624e-01 -4.62771416e-01 -7.13048935e-01 1.56353533e-01
1.53287381e-01 -7.38240302e-01 -2.10510612e-01 -2.53408074e-01] | [7.288451194763184, 2.977374315261841] |
1c14415c-b0b4-48f1-a8bf-38401f7c8aa6 | class-distribution-aware-pseudo-labeling-for | 2305.02795 | null | https://arxiv.org/abs/2305.02795v2 | https://arxiv.org/pdf/2305.02795v2.pdf | Class-Distribution-Aware Pseudo Labeling for Semi-Supervised Multi-Label Learning | Pseudo-labeling has emerged as a popular and effective approach for utilizing unlabeled data. However, in the context of semi-supervised multi-label learning (SSMLL), conventional pseudo-labeling methods encounter difficulties when dealing with instances associated with multiple labels and an unknown label count. These limitations often result in the introduction of false positive labels or the neglect of true positive ones. To overcome these challenges, this paper proposes a novel solution called Class-Aware Pseudo-Labeling (CAP) that performs pseudo-labeling in a class-aware manner. The proposed approach introduces a regularized learning framework incorporating class-aware thresholds, which effectively control the assignment of positive and negative pseudo-labels for each class. Notably, even with a small proportion of labeled examples, our observations demonstrate that the estimated class distribution serves as a reliable approximation. Motivated by this finding, we develop a class-distribution-aware thresholding strategy to ensure the alignment of pseudo-label distribution with the true distribution. The correctness of the estimated class distribution is theoretically verified, and a generalization error bound is provided for our proposed method. Extensive experiments on multiple benchmark datasets confirm the efficacy of CAP in addressing the challenges of SSMLL problems. | ['Hao-Zhe Liu', 'Sheng-Jun Huang', 'Masashi Sugiyama', 'Gang Niu', 'Jia-Hao Xiao', 'Ming-Kun Xie'] | 2023-05-04 | null | null | null | null | ['multi-label-learning', 'pseudo-label'] | ['methodology', 'miscellaneous'] | [ 5.79812467e-01 1.61063656e-01 -6.45079315e-01 -7.28606105e-01
-1.20998979e+00 -6.80346727e-01 3.34074199e-01 5.88016808e-01
-3.16844642e-01 1.04122901e+00 -4.71684784e-01 -9.65809524e-02
-6.59864172e-02 -4.97513831e-01 -6.04216754e-01 -1.01995182e+00
4.61819679e-01 4.79387343e-01 7.23778307e-02 5.26469827e-01
2.62271732e-01 1.73764527e-01 -1.71147251e+00 2.22822338e-01
1.02549601e+00 1.00340044e+00 -3.94067587e-03 2.10148677e-01
-4.19045001e-01 6.76499367e-01 -6.34699047e-01 -2.53637612e-01
3.59128714e-01 -4.59113210e-01 -7.54422426e-01 6.61993504e-01
4.27010953e-01 -1.14727423e-01 5.87608993e-01 1.29399574e+00
1.16280250e-01 3.94764682e-03 8.68540168e-01 -1.49858034e+00
-1.74156100e-01 5.81245720e-01 -1.06230843e+00 -1.97432429e-01
5.91135547e-02 -2.22309634e-01 1.10587049e+00 -9.02390540e-01
3.07690293e-01 9.45996284e-01 6.37070835e-01 3.46197665e-01
-1.16290331e+00 -7.83967853e-01 2.75326371e-01 -1.25103295e-01
-1.77715230e+00 -1.82693794e-01 7.60523081e-01 -3.11528474e-01
1.95466697e-01 2.32603282e-01 2.29662701e-01 5.64733744e-01
-1.93864763e-01 8.48109663e-01 1.55954278e+00 -7.23073900e-01
4.72661495e-01 6.56087458e-01 3.70313793e-01 5.81789255e-01
4.96609271e-01 -2.13394016e-01 -3.31495136e-01 -5.18472373e-01
3.98324996e-01 2.49639377e-02 -1.84507385e-01 -5.48866808e-01
-9.77549255e-01 6.64596200e-01 -1.29700959e-01 2.56441951e-01
-2.26628944e-01 -2.58625180e-01 4.36286300e-01 -7.46025816e-02
6.88992321e-01 9.07293707e-03 -4.45075661e-01 2.34548882e-01
-9.91472423e-01 -8.00880715e-02 6.38004243e-01 1.09147596e+00
1.05310643e+00 -2.12165415e-01 -2.06074238e-01 1.00814056e+00
2.83383578e-01 4.70738113e-01 3.99480402e-01 -6.99937105e-01
3.25459957e-01 8.97680044e-01 4.82996672e-01 -7.80280471e-01
-2.43915856e-01 -6.46528304e-01 -6.35199010e-01 -5.23140877e-02
7.16601312e-01 3.99776101e-02 -7.61665404e-01 1.86468434e+00
8.01956952e-01 2.32239455e-01 1.02208167e-01 7.38702059e-01
2.77340233e-01 3.96838903e-01 4.39443946e-01 -8.58381510e-01
1.21383965e+00 -7.85546184e-01 -9.01579618e-01 -1.07017919e-01
8.52902293e-01 -6.22848332e-01 1.04727149e+00 3.40465814e-01
-5.52506387e-01 -4.09786642e-01 -1.08722365e+00 4.60680544e-01
-1.47958741e-01 5.41779339e-01 4.88088131e-01 8.54397595e-01
-4.95708108e-01 2.25522131e-01 -4.98490542e-01 -2.06125453e-01
2.44979799e-01 3.20214927e-01 -3.15392524e-01 -6.47379011e-02
-8.92753839e-01 5.54371595e-01 8.00027013e-01 1.69360742e-01
-4.92806226e-01 -3.48287404e-01 -4.84506756e-01 -1.15187660e-01
6.77335203e-01 8.63599256e-02 1.18881953e+00 -1.14958560e+00
-1.01302481e+00 9.37714517e-01 -2.25817934e-01 -1.35369733e-01
4.93444592e-01 1.50095984e-01 -3.99244368e-01 2.81744808e-01
4.35925394e-01 3.98932904e-01 8.99030745e-01 -1.77451110e+00
-1.15019155e+00 -3.25146705e-01 -1.40559629e-01 2.31553435e-01
-5.55436254e-01 -3.80336404e-01 -1.40423477e-01 -3.52520615e-01
4.96316195e-01 -8.34741592e-01 -2.38820866e-01 -2.47120976e-01
-5.67817748e-01 -4.93388981e-01 6.08758330e-01 -8.98045450e-02
1.29696882e+00 -2.25920296e+00 -4.94253993e-01 4.17630047e-01
2.49935195e-01 2.09357992e-01 1.43993437e-01 1.91569373e-01
-1.24735691e-01 5.30307889e-02 -5.31646490e-01 -2.97805518e-01
-1.30432397e-01 2.60380715e-01 -2.80493587e-01 6.22982621e-01
2.10634604e-01 2.90504187e-01 -1.14158583e+00 -9.52213705e-01
1.97270349e-01 8.11847299e-02 4.14851159e-02 2.56671637e-01
-1.53175667e-01 3.52094233e-01 -5.80627322e-01 8.26250374e-01
8.88501406e-01 -5.69914877e-01 5.69923162e-01 -3.18889469e-01
1.12789713e-01 -2.97840923e-01 -1.52629876e+00 1.00459373e+00
-3.53754789e-01 -1.56862304e-01 1.81562826e-02 -1.13429940e+00
1.10026991e+00 3.84441495e-01 6.42002761e-01 -8.12112689e-02
3.09865713e-01 5.53718328e-01 -3.89528215e-01 -3.38719398e-01
3.39000046e-01 -4.09760505e-01 1.71501562e-02 7.21496284e-01
-4.05102447e-02 4.91095670e-02 4.31049734e-01 6.69821948e-02
6.30300879e-01 1.73830792e-01 7.38531947e-01 -1.60617426e-01
7.67974734e-01 1.53220549e-01 7.89051235e-01 7.07365692e-01
-5.60743928e-01 4.75473106e-01 2.68891990e-01 -1.86526492e-01
-7.89626718e-01 -8.17996562e-01 -3.67566556e-01 1.05993044e+00
3.29268366e-01 -1.26270771e-01 -6.88068748e-01 -1.09333122e+00
-7.09716454e-02 6.39029503e-01 -5.06277084e-01 -6.16607862e-03
-1.94090188e-01 -1.20957959e+00 3.46328110e-01 3.20546240e-01
4.08944339e-01 -6.72818363e-01 -3.18697244e-01 2.87567526e-01
-3.10050756e-01 -1.24693167e+00 -3.61504316e-01 4.88489211e-01
-6.75124288e-01 -1.48888338e+00 -4.87292647e-01 -9.10867095e-01
1.35046363e+00 3.98486793e-01 7.13765144e-01 2.67936401e-02
9.14931819e-02 3.17684144e-01 -4.41293061e-01 -3.45643871e-02
-8.06336701e-01 -1.03911487e-02 1.17771529e-01 5.14094710e-01
6.01290762e-01 -3.20283622e-01 -2.89445698e-01 5.11381030e-01
-9.58380342e-01 -7.90154561e-02 5.72070122e-01 8.72926056e-01
9.95560348e-01 3.65977168e-01 1.07489073e+00 -1.34478426e+00
4.06087548e-01 -5.26257813e-01 -6.91505313e-01 6.88685477e-01
-9.74277794e-01 -3.16349603e-02 8.15320134e-01 -7.25018382e-01
-1.03803051e+00 4.95211244e-01 2.08554417e-01 -2.66693026e-01
-3.32997769e-01 3.56679410e-01 -2.37265870e-01 -1.04256488e-01
5.01456499e-01 1.77961469e-01 -7.55998939e-02 -3.04286897e-01
3.22200745e-01 1.04663336e+00 4.72286522e-01 -6.53729677e-01
7.41442680e-01 5.29308021e-01 1.14164390e-01 -3.65701139e-01
-1.45930862e+00 -7.37060130e-01 -7.15282261e-01 -3.34657997e-01
3.36495191e-01 -9.08578098e-01 -5.42527318e-01 3.20235252e-01
-6.35039508e-01 8.09792727e-02 -2.35096455e-01 4.25174594e-01
-2.82447845e-01 6.49598122e-01 -3.44255328e-01 -1.28594124e+00
-2.00288534e-01 -1.04010653e+00 1.01966250e+00 2.21782893e-01
-1.51665494e-01 -9.80317652e-01 -1.85500637e-01 4.24964756e-01
-1.22514047e-01 2.49038905e-01 8.78084958e-01 -1.07917821e+00
-2.02616006e-01 -6.40300632e-01 -3.65779787e-01 4.72476006e-01
4.84101892e-01 -1.10818796e-01 -1.07236314e+00 -4.01889056e-01
7.47026578e-02 -6.68174207e-01 4.64078069e-01 1.63752094e-01
1.10475314e+00 -9.10579711e-02 -3.61863524e-01 2.70219427e-03
1.57976985e+00 3.94763723e-02 -2.08062440e-01 1.24188811e-01
5.83163619e-01 6.39250100e-01 1.22540236e+00 8.17079186e-01
3.31961155e-01 3.55235368e-01 3.80125105e-01 4.36132997e-02
1.65722132e-01 -2.46777207e-01 -8.40546489e-02 6.70455992e-01
5.30601680e-01 -3.58239174e-01 -7.21622825e-01 4.85579580e-01
-1.82514119e+00 -4.27719921e-01 -2.91929752e-01 2.43760467e+00
1.23093879e+00 1.58332616e-01 2.30935756e-02 5.96578240e-01
1.18386102e+00 -1.44406721e-01 -6.27927661e-01 1.04948856e-01
-2.58814692e-01 -1.36441708e-01 5.55378675e-01 4.30957109e-01
-1.27981460e+00 6.54862583e-01 6.20651293e+00 1.24249601e+00
-9.65981007e-01 2.13246241e-01 7.74682462e-01 3.97697985e-01
-2.35997915e-01 5.93086034e-02 -1.11042941e+00 4.28817421e-01
6.93020642e-01 -1.13596216e-01 1.56823453e-02 1.05016589e+00
1.10546939e-01 -4.34493244e-01 -9.34256315e-01 9.01771903e-01
6.48375601e-02 -7.12632716e-01 -2.68853568e-02 -1.32182658e-01
8.11090291e-01 -5.80213070e-01 -1.49477810e-01 2.14435935e-01
1.29266277e-01 -3.85054916e-01 7.93523550e-01 -3.08013265e-03
1.02084029e+00 -7.72556484e-01 8.08847785e-01 5.82381487e-01
-1.21910882e+00 -1.47291109e-01 -2.74779290e-01 1.01766691e-01
9.28650517e-03 1.05600369e+00 -1.04408205e+00 5.33692658e-01
1.91853195e-01 4.76640761e-01 -5.35773575e-01 8.59522760e-01
-3.19603115e-01 7.76322901e-01 -3.42396915e-01 9.49865133e-02
3.41055281e-02 -1.33662477e-01 1.04815803e-01 9.69165981e-01
5.80530763e-02 1.19182719e-02 6.82830632e-01 5.57200968e-01
-2.63402648e-02 5.26052117e-01 -2.16619670e-01 -9.27915722e-02
8.00185502e-01 1.52790534e+00 -1.26993024e+00 -5.57593822e-01
-3.66698354e-01 5.27353704e-01 4.16836321e-01 3.07689637e-01
-8.09478879e-01 -1.11642428e-01 -1.54315352e-01 -1.84187442e-01
-2.56241467e-02 2.14254037e-01 -4.45751905e-01 -9.22427893e-01
1.11210763e-01 -5.63195527e-01 5.32866955e-01 -2.86362797e-01
-1.38930392e+00 3.71924698e-01 1.72959715e-01 -1.55701745e+00
-1.89471051e-01 -2.65078425e-01 -8.81597251e-02 4.16221142e-01
-1.43120313e+00 -1.10983860e+00 -2.10795283e-01 1.27894491e-01
4.07748878e-01 6.15630671e-02 7.69328594e-01 4.34419185e-01
-7.01094925e-01 7.87035525e-01 1.85150504e-01 -1.40172407e-01
8.64792109e-01 -1.31881392e+00 -4.14269805e-01 7.52702951e-01
-1.26039371e-01 4.79951859e-01 7.88175762e-01 -6.60283744e-01
-8.91898334e-01 -1.29425359e+00 6.86147690e-01 -8.48570168e-02
4.31458235e-01 -2.00470731e-01 -1.01499987e+00 5.55742145e-01
-3.74526441e-01 3.16969156e-01 1.05784309e+00 -1.27816468e-01
-2.93920457e-01 -2.19467357e-01 -1.47896540e+00 2.75590152e-01
5.98396599e-01 -4.21432316e-01 -2.29570225e-01 6.86659038e-01
3.81003737e-01 -2.34273195e-01 -8.47032309e-01 5.80147266e-01
4.81458277e-01 -7.36110151e-01 5.35562515e-01 -4.53217700e-02
1.35313869e-02 -6.42903388e-01 -1.86233804e-01 -9.71890926e-01
3.36906090e-02 -1.92347109e-01 -4.79452051e-02 1.60510731e+00
3.48135889e-01 -6.32944822e-01 8.24432969e-01 6.25887573e-01
1.24708585e-01 -7.89753139e-01 -8.31302285e-01 -8.20251703e-01
-3.02784801e-01 -2.18244538e-01 4.30411816e-01 1.19132912e+00
2.96475347e-02 6.56260177e-02 -4.26552087e-01 3.70404929e-01
9.13851202e-01 2.80384988e-01 5.04258871e-01 -1.17725396e+00
-1.42254934e-01 -3.08730938e-02 -1.01587929e-01 -7.76119828e-01
3.72285157e-01 -7.65189648e-01 3.53569657e-01 -1.10015357e+00
5.14934957e-01 -8.34143102e-01 -6.11512363e-01 6.50186181e-01
-5.47323763e-01 5.38570106e-01 -3.12227577e-01 3.82512689e-01
-8.90779138e-01 3.07045132e-01 1.05672848e+00 2.92105023e-02
-8.52325857e-02 1.69832796e-01 -6.25961065e-01 7.26353526e-01
8.08258593e-01 -7.04037070e-01 -5.22668421e-01 1.76377550e-01
1.09698378e-01 -1.02649696e-01 -5.28494827e-02 -7.31489122e-01
1.77914125e-03 -3.29416007e-01 2.16774810e-02 -7.66813397e-01
1.11528765e-02 -9.63524342e-01 -1.20883016e-02 2.32247740e-01
-5.75967014e-01 -4.68029082e-01 -1.52063429e-01 7.49673247e-01
-2.20088050e-01 -4.96708333e-01 8.71785939e-01 7.07184244e-03
-5.75253904e-01 1.71514489e-02 -1.24750629e-01 2.79411376e-02
1.32025373e+00 -2.47817397e-01 -1.04098640e-01 -5.82896844e-02
-5.52135766e-01 3.82359028e-01 3.91272992e-01 5.12160473e-02
3.21156979e-01 -1.26220083e+00 -4.92638409e-01 1.62016481e-01
5.06729901e-01 2.43972111e-02 5.59627637e-02 7.45025098e-01
-1.14903249e-01 2.13918447e-01 2.62415171e-01 -7.56347537e-01
-1.14281023e+00 5.67137122e-01 1.07105590e-01 -3.39812964e-01
-1.94697365e-01 5.49410522e-01 2.06981108e-01 -5.65929592e-01
2.92352587e-01 -6.93452358e-02 -1.95726231e-01 7.46214092e-02
4.36810493e-01 2.64425546e-01 -3.19059752e-02 -6.71590984e-01
-2.78394282e-01 4.95140225e-01 -1.66286841e-01 7.84913003e-02
9.09578562e-01 -4.69334722e-01 -1.37407616e-01 8.35308552e-01
1.02416122e+00 -2.56577916e-02 -1.23609555e+00 -4.49111670e-01
2.36730829e-01 -4.68575627e-01 -1.95676059e-01 -7.13073611e-01
-8.90256345e-01 5.72888374e-01 5.31497478e-01 2.23140791e-01
1.12237656e+00 -1.41334355e-01 5.08107424e-01 1.59877151e-01
5.94549716e-01 -1.18352222e+00 -4.97683957e-02 -4.32910025e-02
1.43263161e-01 -1.57031584e+00 1.02976970e-01 -8.71359110e-01
-4.98901308e-01 9.18705165e-01 8.73624861e-01 3.00285697e-01
4.60693181e-01 1.27535373e-01 2.31293410e-01 1.50810376e-01
-5.64389050e-01 -1.25073358e-01 -6.27863929e-02 2.46879309e-01
4.22680080e-01 1.69019848e-01 -7.18412519e-01 3.62938076e-01
3.51870030e-01 -3.56026576e-03 5.19991040e-01 1.22368670e+00
-5.71922004e-01 -1.18576694e+00 -6.30308449e-01 5.15316129e-01
-4.97988045e-01 2.44101554e-01 -9.89370272e-02 6.04493201e-01
2.21498728e-01 1.14608550e+00 -2.63978004e-01 -2.25823194e-01
-4.31593098e-02 3.25568646e-01 2.94690132e-01 -8.22908938e-01
-2.02697799e-01 3.09735149e-01 -1.47615761e-01 5.86369485e-02
-8.20477962e-01 -4.59627420e-01 -1.50194669e+00 2.76350081e-01
-9.65003729e-01 5.24167120e-01 5.02329826e-01 1.13010776e+00
-1.07453752e-03 1.38801545e-01 9.96969163e-01 -2.76090026e-01
-9.53312874e-01 -8.76726806e-01 -1.02130103e+00 5.62352479e-01
1.80154726e-01 -8.51149440e-01 -7.28382587e-01 1.41664684e-01] | [9.362951278686523, 4.01597261428833] |
de725bef-0c1e-4f0d-aae1-d21b77cd8417 | mderank-a-masked-document-embedding-rank | 2110.06651 | null | https://arxiv.org/abs/2110.06651v3 | https://arxiv.org/pdf/2110.06651v3.pdf | MDERank: A Masked Document Embedding Rank Approach for Unsupervised Keyphrase Extraction | Keyphrase extraction (KPE) automatically extracts phrases in a document that provide a concise summary of the core content, which benefits downstream information retrieval and NLP tasks. Previous state-of-the-art (SOTA) methods select candidate keyphrases based on the similarity between learned representations of the candidates and the document. They suffer performance degradation on long documents due to discrepancy between sequence lengths which causes mismatch between representations of keyphrase candidates and the document. In this work, we propose a novel unsupervised embedding-based KPE approach, Masked Document Embedding Rank (MDERank), to address this problem by leveraging a mask strategy and ranking candidates by the similarity between embeddings of the source document and the masked document. We further develop a KPE-oriented BERT (KPEBERT) model by proposing a novel self-supervised contrastive learning method, which is more compatible to MDERank than vanilla BERT. Comprehensive evaluations on six KPE benchmarks demonstrate that the proposed MDERank outperforms state-of-the-art unsupervised KPE approach by average 1.80 $F1@15$ improvement. MDERank further benefits from KPEBERT and overall achieves average 3.53 $F1@15$ improvement over the SOTA SIFRank. Our code is available at \url{https://github.com/LinhanZ/mderank}. | ['Xin Cao', 'Wei Wang', 'Bing Li', 'Shiliang Zhang', 'Chong Deng', 'Wen Wang', 'Qian Chen', 'Linhan Zhang'] | 2021-10-13 | null | https://aclanthology.org/2022.findings-acl.34 | https://aclanthology.org/2022.findings-acl.34.pdf | findings-acl-2022-5 | ['document-embedding'] | ['methodology'] | [-9.22361538e-02 -6.38493150e-03 -5.87912500e-01 1.80556089e-01
-1.19712698e+00 -6.10491455e-01 9.16075289e-01 6.78412199e-01
-5.63282430e-01 3.87750149e-01 7.96365917e-01 -2.29165107e-01
-4.54125732e-01 -6.93706274e-01 -5.74544072e-01 -5.50390065e-01
-3.25332791e-01 3.37827116e-01 3.95797402e-01 -2.15359524e-01
5.83304286e-01 3.40965003e-01 -1.26777506e+00 4.80491310e-01
6.98661625e-01 7.58002996e-01 1.65717587e-01 8.72260571e-01
-4.53987479e-01 5.03801107e-01 -4.62905139e-01 -2.63418198e-01
3.47302765e-01 -1.00447074e-01 -7.01925695e-01 -3.64902973e-01
4.23433065e-01 -4.35969293e-01 -8.45638752e-01 8.62254441e-01
4.30576444e-01 -3.12278662e-02 7.20743597e-01 -1.36980546e+00
-7.35041380e-01 9.70270574e-01 -5.53178310e-01 4.20787841e-01
1.35042459e-01 -3.56589675e-01 1.93501365e+00 -1.68423533e+00
5.46779871e-01 8.90126109e-01 3.08136255e-01 3.35271239e-01
-1.17614686e+00 -8.16268325e-01 8.98605809e-02 3.64520878e-01
-1.48490703e+00 -2.84249157e-01 7.97799766e-01 -1.68003380e-01
1.04677331e+00 4.05049890e-01 4.41513002e-01 8.05570722e-01
1.91810727e-01 1.28192079e+00 9.07705724e-01 -4.77886319e-01
6.84299767e-02 3.46903533e-01 5.05138934e-01 6.70794487e-01
4.65238720e-01 -1.09818708e-02 -8.91319573e-01 -5.43193996e-01
2.04128489e-01 1.01185434e-01 -3.08485657e-01 -2.31216699e-01
-1.39544797e+00 8.85762155e-01 2.73059636e-01 3.30504060e-01
-3.77155632e-01 1.54574960e-01 3.77788663e-01 1.81867525e-01
3.27281088e-01 7.62441754e-01 -5.55554092e-01 -1.87647313e-01
-1.21291554e+00 4.81092036e-01 8.54977071e-01 8.55509639e-01
6.70866251e-01 -2.45918348e-01 -5.28057516e-01 7.65243769e-01
4.03777063e-01 3.47740322e-01 5.74247360e-01 -4.20462787e-01
7.21914828e-01 7.47003794e-01 1.51708946e-01 -1.15725815e+00
-1.73507273e-01 -5.22764444e-01 -6.07024610e-01 -4.16676998e-01
-1.56974167e-01 2.21229836e-01 -7.08238363e-01 1.17559361e+00
8.93195495e-02 5.23078963e-02 1.43961325e-01 5.48752487e-01
9.10707235e-01 1.38664138e+00 -2.71741658e-01 -1.19715385e-01
1.52790642e+00 -1.26476336e+00 -7.03852654e-01 -2.51444399e-01
6.93925500e-01 -8.97261500e-01 9.06814158e-01 3.92085463e-01
-6.87977314e-01 -2.30644062e-01 -1.30519211e+00 -8.82887542e-02
-6.08885407e-01 4.53539848e-01 3.17715198e-01 3.36689472e-01
-8.04597199e-01 5.38487017e-01 -4.61774975e-01 2.73956521e-03
3.18997741e-01 3.12125415e-01 -2.69305348e-01 -1.83353722e-01
-1.40470374e+00 6.01252854e-01 6.99504793e-01 -1.79480478e-01
-7.05458760e-01 -9.44378734e-01 -6.39525414e-01 2.16251910e-01
6.46992803e-01 -2.39647254e-01 9.62553740e-01 -5.55423908e-02
-1.10817444e+00 4.10108089e-01 -2.92159200e-01 -5.67306221e-01
3.90736200e-02 -7.89621115e-01 -6.10046983e-01 5.42688131e-01
1.90701887e-01 6.80356443e-01 1.07445765e+00 -1.17397416e+00
-7.58669972e-01 8.15041512e-02 -1.80305421e-01 1.09494418e-01
-9.55725014e-01 4.51611504e-02 -6.37871861e-01 -1.09420013e+00
-1.01712951e-02 -8.33916903e-01 1.52021600e-02 -1.40951961e-01
-5.80896497e-01 -5.35580039e-01 9.38796163e-01 -6.97397530e-01
2.11081648e+00 -2.19979048e+00 2.91035138e-02 2.95332462e-01
5.62911272e-01 5.16176224e-01 -2.99170315e-01 1.11523223e+00
-5.12222107e-03 3.15687388e-01 -9.82947089e-03 -1.85631931e-01
3.28924716e-01 -6.54661143e-03 -9.40374136e-01 2.79352605e-01
2.95365214e-01 9.04521108e-01 -1.15832376e+00 -6.01282060e-01
-7.69972950e-02 2.23734200e-01 -3.88039351e-01 2.33046323e-01
-1.47755146e-01 -3.68357599e-01 -5.14723241e-01 6.79172516e-01
5.16115308e-01 -2.32671052e-01 -9.62058008e-02 -4.40583557e-01
-1.73462808e-01 6.30524516e-01 -1.25374043e+00 1.21508265e+00
-3.35411578e-01 8.27514589e-01 -3.16937596e-01 -7.43265033e-01
8.77441645e-01 3.69420797e-01 6.20938599e-01 -3.02943975e-01
-1.37886375e-01 4.41701502e-01 -1.68022588e-01 -2.90261377e-02
1.06176376e+00 3.27837467e-01 -1.60691172e-01 6.92609012e-01
1.18021220e-01 -4.22403552e-02 5.02067566e-01 8.78001630e-01
1.35850394e+00 -3.07864457e-01 3.95790875e-01 -3.16357404e-01
4.41959262e-01 3.43017988e-02 3.93580914e-01 8.98379028e-01
-9.10922661e-02 3.98560256e-01 4.44274694e-01 -1.46334603e-01
-1.02284694e+00 -1.06978381e+00 -2.34320071e-02 1.02028430e+00
1.19387835e-01 -1.31086588e+00 -3.59472811e-01 -1.00167227e+00
2.31653661e-01 7.84902811e-01 -4.65616018e-01 -2.37648904e-01
-5.99008322e-01 -6.08311892e-01 5.87315023e-01 4.51025009e-01
1.63564816e-01 -6.71343684e-01 -1.63366348e-01 2.98800796e-01
-7.59049132e-02 -1.05829394e+00 -8.09362352e-01 1.15178876e-01
-6.94669604e-01 -6.96580589e-01 -8.29194307e-01 -6.45881712e-01
6.10553503e-01 5.05982697e-01 9.01197195e-01 -6.39770627e-02
-2.06731930e-01 5.39164662e-01 -7.62943387e-01 -3.52773309e-01
-2.59599715e-01 3.26396853e-01 1.88566267e-01 -1.98437069e-02
7.21326828e-01 -3.85156214e-01 -7.73174167e-01 3.90648097e-02
-1.18123293e+00 -7.28842840e-02 9.80360866e-01 9.85311389e-01
5.63129842e-01 3.29988927e-01 4.46283877e-01 -6.69114649e-01
9.96735871e-01 -5.69835424e-01 -4.85514998e-01 2.37458497e-01
-1.39178360e+00 4.88694191e-01 5.94681621e-01 -4.51817840e-01
-4.62486476e-01 -4.31366712e-01 1.20361298e-01 -3.33221585e-01
4.13732946e-01 6.54856026e-01 1.96442693e-01 4.36380506e-01
4.36677307e-01 5.10555804e-01 -4.63833481e-01 -5.27269185e-01
5.28554082e-01 9.91913915e-01 2.65189320e-01 -5.57501376e-01
1.29305589e+00 3.75937730e-01 -2.39862546e-01 -8.46851349e-01
-7.46756852e-01 -1.05832624e+00 -3.79248619e-01 8.29404965e-02
2.33790427e-01 -1.00176001e+00 -1.08137481e-01 -5.60362749e-02
-1.08529556e+00 2.34422073e-01 -3.20820540e-01 6.11331642e-01
-8.36794749e-02 5.50631225e-01 -5.97290039e-01 -5.98727882e-01
-8.01077843e-01 -8.19959402e-01 1.07055449e+00 -4.22715135e-02
-4.44363952e-01 -6.76098168e-01 2.24432305e-01 1.58497050e-01
1.68103665e-01 -2.48765707e-01 1.07654464e+00 -1.03914642e+00
-6.95646167e-01 -5.26638448e-01 -4.72422540e-01 3.47918630e-01
3.06154281e-01 6.45262673e-02 -6.55631065e-01 -4.51914936e-01
-4.40239072e-01 -1.19419061e-01 1.36132669e+00 -1.40477018e-02
7.40842819e-01 -7.10757732e-01 -4.08633947e-01 1.88013658e-01
1.52121270e+00 -5.44372946e-02 5.33796072e-01 5.80359042e-01
6.57370627e-01 5.37262380e-01 8.52498591e-01 6.36362791e-01
2.47842208e-01 5.65782368e-01 1.40853435e-01 2.34298766e-01
-1.87902138e-01 -7.31129408e-01 7.55582273e-01 1.25393915e+00
4.19698954e-01 -2.53363460e-01 -8.09826732e-01 9.98588324e-01
-1.70659304e+00 -7.64000773e-01 1.19399853e-01 1.95260954e+00
1.28332663e+00 5.19264460e-01 -1.10143330e-02 4.54264522e-01
4.50621068e-01 6.02400780e-01 -2.41375446e-01 -3.54637295e-01
1.34418011e-01 4.01729643e-01 7.48955667e-01 3.83327156e-01
-1.09415185e+00 9.83022928e-01 4.68702888e+00 1.30209863e+00
-7.63927698e-01 -2.57201456e-02 1.05091548e-02 -4.64039296e-02
-5.68586588e-01 2.77726650e-01 -1.46385717e+00 4.04558063e-01
1.01901078e+00 -5.40592194e-01 1.23783506e-01 8.24942231e-01
6.50714263e-02 1.07143573e-01 -1.06819892e+00 8.38119388e-01
1.55947447e-01 -1.61931205e+00 3.49593818e-01 8.22552368e-02
6.32986069e-01 -1.46038299e-02 2.63415873e-01 3.01248163e-01
2.57558286e-01 -6.24796450e-01 5.45644462e-01 1.05824649e-01
5.20019948e-01 -6.39441550e-01 6.42641664e-01 1.88007638e-01
-1.47739232e+00 -2.65515208e-01 -2.83715189e-01 3.51090282e-01
1.62993290e-03 8.84428144e-01 -1.03150964e+00 6.23635054e-01
6.92917168e-01 6.58243299e-01 -6.54639661e-01 7.81075716e-01
-4.14242685e-01 8.21246028e-01 -2.10658893e-01 -4.42178369e-01
5.57667911e-01 1.04361929e-01 9.97341812e-01 1.65495908e+00
2.16364563e-01 -1.61303788e-01 1.32436715e-02 6.44026995e-01
-2.70268559e-01 4.46513772e-01 -3.77293438e-01 -7.63913870e-01
9.05984640e-01 1.29461670e+00 -6.43218040e-01 -4.86753821e-01
-3.44673991e-01 9.51797783e-01 4.38562371e-02 1.66415572e-01
-5.04842520e-01 -9.41663325e-01 6.66079402e-01 1.78487767e-02
7.34225631e-01 -3.67742747e-01 1.92860559e-01 -1.13494253e+00
2.88523257e-01 -9.50605154e-01 3.83235455e-01 -2.77956486e-01
-1.20872474e+00 4.34155881e-01 2.78210759e-01 -1.67441165e+00
-4.55821306e-02 -6.88027799e-01 -3.92476290e-01 5.38095176e-01
-1.78479719e+00 -9.07032013e-01 2.19093710e-01 1.24449693e-01
7.13648558e-01 -3.13976228e-01 8.69312763e-01 1.56880543e-01
-5.14961958e-01 6.46829486e-01 5.91855645e-01 3.79061460e-01
8.13139915e-01 -1.37650502e+00 4.80280995e-01 8.90018880e-01
7.18099892e-01 9.90465701e-01 6.13436162e-01 -5.45008898e-01
-1.72166705e+00 -1.07942307e+00 1.26374388e+00 -3.71008039e-01
1.06912613e+00 -4.15911078e-01 -7.41242647e-01 2.97834277e-01
3.24129462e-01 7.89173171e-02 7.87550747e-01 -1.19417161e-01
-8.14836025e-01 -3.30836535e-01 -5.63198745e-01 8.71604621e-01
6.12205267e-01 -8.35584044e-01 -9.37017679e-01 1.89418629e-01
1.11686826e+00 -5.73267266e-02 -8.48488331e-01 3.33340168e-01
5.87372243e-01 -6.04268074e-01 1.11699700e+00 -2.77375609e-01
4.97698665e-01 -3.67270678e-01 -3.17642897e-01 -1.12596810e+00
-2.65550762e-01 -8.50340843e-01 -8.70589554e-01 1.29497969e+00
6.16776943e-01 -4.51342463e-01 5.99594414e-01 1.26901999e-01
1.78366065e-01 -1.12614882e+00 -7.40566671e-01 -1.10604358e+00
1.04705885e-01 -3.63905609e-01 4.64443624e-01 6.78159773e-01
2.00169116e-01 4.95367378e-01 -3.19473147e-01 3.60991180e-01
5.92227995e-01 1.60590589e-01 6.85830534e-01 -9.18337286e-01
-3.35182786e-01 -4.99630123e-01 -1.45660400e-01 -1.38310480e+00
-1.34816393e-02 -1.10440874e+00 -7.92757198e-02 -1.70331049e+00
2.02652305e-01 -2.08690748e-01 -7.00725734e-01 3.03196371e-01
-3.34973961e-01 -9.41503569e-02 2.40112543e-01 4.70506310e-01
-5.50754428e-01 8.13107789e-01 7.19939947e-01 -4.91659224e-01
-3.59448761e-01 -2.53728926e-01 -7.77855515e-01 3.54706526e-01
8.53444517e-01 -8.73588443e-01 -2.93730855e-01 -1.79208696e-01
3.60205710e-01 -4.45209503e-01 7.85185322e-02 -6.48505807e-01
4.53470021e-01 7.84471035e-02 2.31786445e-02 -9.66789544e-01
3.07331413e-01 -6.73306644e-01 -5.23520052e-01 4.67239231e-01
-5.27727544e-01 2.13296235e-01 3.21963072e-01 7.79234111e-01
-4.14884329e-01 -5.57139337e-01 1.43490180e-01 1.79672524e-01
-7.59964406e-01 3.31665546e-01 -4.10138100e-01 8.55164006e-02
5.26172936e-01 8.11025426e-02 -5.15545726e-01 -2.62317568e-01
-6.45727441e-02 2.40177467e-01 -1.11989513e-01 5.04009664e-01
1.18088675e+00 -1.40101719e+00 -8.59622538e-01 3.64399776e-02
5.70307493e-01 -3.00882936e-01 -2.46287957e-01 7.50754118e-01
-3.33408147e-01 7.61745036e-01 5.29147387e-01 -6.88656420e-02
-1.28586900e+00 4.66003835e-01 -3.83528352e-01 -7.09705114e-01
-7.36401141e-01 8.81115615e-01 -1.26582012e-01 -2.84316510e-01
2.47558057e-01 -3.98006231e-01 -2.51380235e-01 3.39593261e-01
7.98594475e-01 4.87220556e-01 1.40828744e-01 -4.01669621e-01
-2.32668191e-01 4.46170807e-01 -7.69977331e-01 -3.71664524e-01
1.26760805e+00 1.26532882e-01 -2.88541019e-01 2.20523387e-01
1.58355463e+00 5.01470804e-01 -7.84057856e-01 -8.04861546e-01
5.42584121e-01 -5.75044096e-01 4.25432801e-01 -4.39411849e-01
-4.94283259e-01 5.69915354e-01 3.71790260e-01 1.07833870e-01
9.88635778e-01 1.33049577e-01 1.21614897e+00 6.43244863e-01
4.20009159e-02 -1.23235512e+00 5.15131235e-01 4.25622821e-01
9.27070200e-01 -1.07695627e+00 5.75237811e-01 -1.86646953e-01
-4.55199987e-01 1.16103470e+00 1.48423746e-01 -1.59609929e-01
9.09496546e-01 2.93420292e-02 -1.66799486e-01 -2.70548224e-01
-1.00238085e+00 -1.45871431e-01 6.62871301e-01 9.79268700e-02
1.11829430e-01 -1.39527589e-01 -5.22303462e-01 7.59340525e-01
-1.81795835e-01 -3.31530213e-01 3.12809825e-01 1.22644663e+00
-6.05026722e-01 -1.41941786e+00 -2.21896112e-01 6.26297832e-01
-4.62635398e-01 -5.88245451e-01 -6.26460195e-01 6.36473835e-01
-1.64670587e-01 8.00966680e-01 -2.81473637e-01 -6.08055353e-01
9.01652202e-02 9.14763138e-02 5.27064763e-02 -8.21620464e-01
-3.37552637e-01 1.77889764e-01 -3.92088573e-03 -2.94610918e-01
-1.64730415e-01 -3.87884498e-01 -1.25432348e+00 -1.53067052e-01
-6.21348619e-01 5.35410047e-01 4.70313609e-01 6.24985039e-01
6.08762503e-01 2.53605515e-01 9.80765045e-01 -5.08809090e-01
-8.00135016e-01 -8.73555064e-01 -5.22892296e-01 -8.12461004e-02
5.23974061e-01 -3.93135488e-01 -4.97953534e-01 -1.83871001e-01] | [12.222155570983887, 8.85249137878418] |
934d9f95-43dc-47a0-8d3c-935c2febe69a | throttling-poisson-processes | null | null | http://papers.nips.cc/paper/4025-throttling-poisson-processes | http://papers.nips.cc/paper/4025-throttling-poisson-processes.pdf | Throttling Poisson Processes | We study a setting in which Poisson processes generate sequences of decision-making events. The optimization goal is allowed to depend on the rate of decision outcomes; the rate may depend on a potentially long backlog of events and decisions. We model the problem as a Poisson process with a throttling policy that enforces a data-dependent rate limit and reduce the learning problem to a convex optimization problem that can be solved efficiently. This problem setting matches applications in which damage caused by an attacker grows as a function of the rate of unsuppressed hostile events. We report on experiments on abuse detection for an email service. | ['Michael Brückner', 'Peter Haider', 'Uwe Dick', 'Tobias Scheffer', 'Thomas Vanck'] | 2010-12-01 | null | null | null | neurips-2010-12 | ['abuse-detection'] | ['natural-language-processing'] | [ 1.57081753e-01 -3.03802192e-01 -3.25662792e-01 -2.53930271e-01
-5.53498924e-01 -6.01588964e-01 4.82559055e-01 5.89056373e-01
-1.02991068e+00 5.90085924e-01 -1.45306617e-01 -6.36497378e-01
1.98328555e-01 -8.70759010e-01 -5.75218379e-01 -6.35111570e-01
-6.34366393e-01 9.00375843e-01 3.21745545e-01 3.27247679e-01
3.45812082e-01 6.17282450e-01 -4.05748963e-01 3.91572639e-02
4.30175588e-02 5.67313910e-01 -3.03308725e-01 1.20561397e+00
-2.35263444e-03 1.07317519e+00 -1.14915025e+00 -3.71625721e-01
3.23487848e-01 8.78814012e-02 -6.66135371e-01 1.38241068e-01
-2.19785929e-01 -6.38650477e-01 -5.38132787e-01 6.61068261e-01
3.97938341e-01 -4.18073200e-02 7.30388224e-01 -1.67615771e+00
-1.52164057e-01 7.22512722e-01 -9.97928083e-01 1.16471088e+00
1.93396211e-01 3.21099609e-01 8.62224281e-01 -7.40519688e-02
1.89021915e-01 1.34373510e+00 1.83587581e-01 4.42509711e-01
-1.49587905e+00 -6.14687502e-01 1.86940372e-01 8.37446973e-02
-7.22991467e-01 -2.90553600e-01 2.69719273e-01 -3.43412340e-01
9.51668441e-01 -1.47541717e-01 1.33265108e-01 1.47147512e+00
5.53587914e-01 6.00562453e-01 7.85310209e-01 -3.07934284e-01
9.17980373e-01 -4.67052609e-01 3.84618282e-01 1.80926546e-01
4.20051157e-01 1.03370115e-01 -4.55752462e-01 -1.06145334e+00
9.59005296e-01 3.00960511e-01 -1.34786904e-01 2.00949371e-01
-4.60228771e-01 1.30367076e+00 -2.28010431e-01 -2.30327606e-01
-6.67424083e-01 5.85047781e-01 5.86534679e-01 6.66887522e-01
2.33143404e-01 2.10165873e-01 -2.72922128e-01 -5.06665587e-01
-6.45710468e-01 2.36302763e-01 1.59640992e+00 5.90214372e-01
2.13731661e-01 1.97289199e-01 -1.54192820e-01 3.67539763e-01
1.57386333e-01 5.68705618e-01 7.74161369e-02 -9.69543517e-01
6.30541742e-01 -3.45059454e-01 4.06554580e-01 -4.67814088e-01
-8.59076679e-02 -7.88542554e-02 -4.37813610e-01 2.87779033e-01
5.65525174e-01 -6.59511745e-01 -6.66111588e-01 1.64623177e+00
7.29737431e-02 4.45028216e-01 -5.41775584e-01 4.32746649e-01
-6.19797468e-01 8.56927276e-01 2.52669543e-01 -7.65479982e-01
9.66366887e-01 -1.85419843e-01 -6.15042090e-01 2.26190072e-02
2.47967497e-01 -7.03006506e-01 6.55705810e-01 4.18663919e-01
-1.27531910e+00 4.68758672e-01 -4.36377674e-01 5.63107073e-01
4.16031688e-01 -9.58808959e-01 4.22293693e-01 7.58921266e-01
-6.80187643e-01 5.06797493e-01 -9.81464207e-01 -6.76614866e-02
5.30235529e-01 3.14352453e-01 1.58301085e-01 2.30153993e-01
-7.99834669e-01 6.63512647e-01 2.28340939e-01 -4.44576412e-01
-1.56697607e+00 -7.33434021e-01 -3.38115960e-01 2.03925997e-01
8.29182506e-01 -6.42723560e-01 1.63771760e+00 -1.63845927e-01
-9.57999527e-01 7.54990220e-01 8.88480842e-02 -7.95888543e-01
7.63485312e-01 -1.68138921e-01 2.89833415e-02 2.58500904e-01
2.48473331e-01 -5.38167536e-01 1.16578567e+00 -9.93244350e-01
-6.02269769e-01 -2.99195737e-01 2.66906142e-01 1.40788913e-01
-2.91512638e-01 6.01413310e-01 6.10700399e-02 -5.31663835e-01
-7.86080956e-01 -7.90542364e-01 -5.94142973e-01 -1.41213283e-01
-3.33501250e-01 -2.23610386e-01 9.86322999e-01 -3.34690839e-01
1.38460219e+00 -1.84396839e+00 -2.01473877e-01 1.30966842e-01
1.20460413e-01 -8.26353282e-02 4.29762378e-02 8.35646808e-01
2.38921881e-01 6.70877397e-01 -2.56676316e-01 -6.95183575e-01
-2.74268895e-01 5.22738457e-01 -7.92698503e-01 7.76267827e-01
-9.85941850e-03 3.53698060e-02 -8.56422961e-01 -2.14734748e-01
-1.91704601e-01 -1.89444393e-01 -4.72308099e-01 7.30964839e-01
-1.75871789e-01 -3.45532931e-02 -7.98322141e-01 5.56110263e-01
3.23512256e-01 -3.13272387e-01 -1.12326682e-01 7.83556163e-01
2.96077251e-01 2.81037420e-01 -1.02986121e+00 6.44677341e-01
-6.80176735e-01 1.95998475e-01 3.15779120e-01 -6.10690951e-01
3.89942199e-01 3.64058375e-01 7.15634406e-01 9.19198543e-02
2.51330405e-01 -1.92626581e-01 -2.21483678e-01 -3.27306777e-01
-7.46452138e-02 -7.37272561e-01 -3.51812214e-01 1.22796237e+00
-2.19722569e-01 1.00449488e-01 -1.43088344e-02 4.73163933e-01
2.07099986e+00 -7.73073137e-01 2.76594549e-01 2.43049979e-01
-2.54172474e-01 -2.79038548e-01 7.48703599e-01 1.42757428e+00
-1.79032236e-01 -1.12636657e-02 1.21087456e+00 -4.31117237e-01
-1.39932072e+00 -1.46254194e+00 3.53851095e-02 1.02431679e+00
2.68907356e-03 -9.65960845e-02 -5.21143198e-01 -8.00697565e-01
2.69021064e-01 9.30629015e-01 -3.99795830e-01 1.36493951e-01
-1.00880361e+00 -1.16387892e+00 4.91732746e-01 3.94642651e-01
1.96707800e-01 -1.16626680e+00 -5.80423772e-01 5.58159947e-01
2.92860687e-01 -1.49079514e+00 -7.27749825e-01 2.60980427e-01
-6.32835567e-01 -1.17107356e+00 -6.33207634e-02 -1.34794250e-01
4.75891292e-01 1.31351396e-01 1.03252268e+00 -8.57834984e-03
-5.56565166e-01 7.96768308e-01 -2.59946827e-02 -7.86585927e-01
-6.97846532e-01 -3.00784931e-02 1.79894626e-01 7.06681237e-02
3.29690099e-01 -9.28971708e-01 -4.44300771e-01 -1.17334120e-01
-1.34430146e+00 -8.19597304e-01 -1.28112420e-01 5.37448049e-01
2.49619365e-01 1.03574358e-01 7.54703462e-01 -1.04525411e+00
1.21022248e+00 -1.07162035e+00 -6.55390382e-01 -7.49511570e-02
-2.45075345e-01 -1.02044940e-01 6.32141948e-01 -8.30831826e-01
-8.75623584e-01 -3.71728748e-01 3.65798473e-01 -6.15230322e-01
-1.77452117e-01 -3.67203914e-02 1.95934266e-01 2.33581573e-01
5.52864611e-01 -8.74821469e-02 -2.73971051e-01 -2.09093645e-01
-1.44782998e-02 4.77356493e-01 2.59044379e-01 -8.68965387e-01
8.38632226e-01 3.32882911e-01 7.98011646e-02 -7.72202671e-01
-5.41192532e-01 -5.36771476e-01 -1.39354467e-02 -6.48854971e-02
4.66852963e-01 -5.97262919e-01 -1.20298922e+00 5.39419532e-01
-1.33930850e+00 -6.25183642e-01 -3.52675110e-01 9.95860770e-02
-7.84554422e-01 4.42460865e-01 -1.32975042e+00 -1.23256767e+00
-2.89092600e-01 -4.52430606e-01 6.64955139e-01 3.92701738e-02
-2.83034891e-01 -1.35959029e+00 2.01019078e-01 5.53521141e-03
4.31672961e-01 4.22972471e-01 1.09332669e+00 -1.11264718e+00
-6.96376860e-01 -4.29549307e-01 7.58030415e-02 2.68997163e-01
-4.58349138e-02 -1.13407902e-01 -3.72881830e-01 -4.23603833e-01
6.95862889e-01 -1.54649109e-01 4.71481860e-01 1.40823349e-01
1.55477715e+00 -1.07345295e+00 -1.78822339e-01 3.65090549e-01
1.49967694e+00 2.61525929e-01 9.07730088e-02 3.42587084e-01
4.14869070e-01 1.57133475e-01 1.93151727e-01 1.12642789e+00
4.78791967e-02 3.18932116e-01 7.26347685e-01 3.51335645e-01
7.03595579e-01 -2.05131501e-01 4.00577813e-01 -1.09025434e-01
7.20524266e-02 -6.36029065e-01 -1.19069469e+00 5.20538509e-01
-1.60189629e+00 -1.29389286e+00 1.34411409e-01 2.67697453e+00
9.15937245e-01 6.23107076e-01 5.84455192e-01 -2.77382527e-02
8.20401251e-01 3.54552656e-01 -6.80923343e-01 -1.02272570e+00
2.17430085e-01 1.79423288e-01 9.41845119e-01 8.02975118e-01
-9.42934453e-01 5.20672023e-01 7.42597532e+00 6.84250712e-01
-8.74075353e-01 3.04488420e-01 9.36020494e-01 -7.04110384e-01
-3.51825333e-03 4.08952571e-02 -8.76291156e-01 9.44008589e-01
1.49711263e+00 -7.61387587e-01 5.93983412e-01 7.71187842e-01
4.95142221e-01 -8.36697295e-02 -1.36515999e+00 8.75882566e-01
-3.73055071e-01 -9.11180377e-01 -1.47784427e-01 5.92410326e-01
3.57310742e-01 4.46073897e-02 -2.43871156e-02 1.52001798e-01
1.20813596e+00 -8.81799936e-01 1.38288558e-01 1.08821318e-02
3.49914342e-01 -9.05488610e-01 3.87291551e-01 8.32073271e-01
-2.46837914e-01 -7.13814378e-01 -1.86199531e-01 -1.79944187e-01
8.73164713e-01 9.39325750e-01 -1.08625770e+00 -4.78079408e-01
3.91122133e-01 7.34400377e-02 2.73971856e-01 1.15705514e+00
-2.80916095e-01 1.41931093e+00 -7.39259243e-01 2.36099780e-01
1.50595933e-01 4.61682267e-02 9.45418954e-01 1.44113135e+00
-4.51509766e-02 3.77986640e-01 6.98785841e-01 4.27651584e-01
-4.43394035e-01 -1.25916049e-01 -6.43435061e-01 2.72412062e-01
8.19299757e-01 8.67301702e-01 -4.21836495e-01 -1.98894218e-01
-9.80583057e-02 8.43330920e-01 3.09148133e-01 5.91503978e-01
-8.49938810e-01 1.01665735e-01 7.54125893e-01 5.66870630e-01
1.09290168e-01 -5.95938623e-01 -3.59931052e-01 -1.15962684e+00
-9.87400860e-03 -5.94301999e-01 6.31746233e-01 -1.08580306e-01
-1.81805182e+00 2.43167415e-01 1.33886725e-01 -4.84000057e-01
-4.14616078e-01 2.05165986e-02 -1.45558298e+00 8.18656147e-01
-1.21740627e+00 -3.04131866e-01 5.25128722e-01 4.28070992e-01
6.84192002e-01 -5.14699519e-02 4.47272301e-01 -1.51675502e-02
-8.58449936e-01 1.52481943e-01 -1.51730835e-01 5.93217686e-02
4.23967630e-01 -1.50794482e+00 5.58919370e-01 8.13531637e-01
-3.45165461e-01 3.11900616e-01 8.86201322e-01 -8.17232907e-01
-9.35406864e-01 -8.30407202e-01 8.20531905e-01 -5.61325252e-01
1.20724320e+00 -3.05401981e-01 -1.13422859e+00 7.87237525e-01
-4.40634675e-02 -2.25816388e-02 8.08892310e-01 -2.66434010e-02
-4.44933981e-01 1.58722147e-01 -1.31629050e+00 4.72817272e-01
7.43054509e-01 -4.89577442e-01 -2.78802991e-01 4.87872273e-01
5.58997869e-01 -4.35455628e-02 -7.10303962e-01 1.78565215e-02
-1.55659512e-01 -3.71354222e-01 5.63528180e-01 -1.20827496e+00
1.27574518e-01 3.49444926e-01 2.04454586e-01 -1.03894401e+00
-1.45946681e-01 -1.26858985e+00 -8.58213723e-01 1.11097980e+00
7.53780678e-02 -4.40412730e-01 6.37293935e-01 9.14447010e-01
8.47027779e-01 -7.98189700e-01 -1.38719642e+00 -7.82194972e-01
2.24416628e-01 -1.50131240e-01 2.22410947e-01 7.02938676e-01
-1.94745407e-01 3.15663606e-01 -5.83689749e-01 3.39419305e-01
1.17562020e+00 -4.50199485e-01 4.02697593e-01 -7.68027782e-01
-7.05784142e-01 -2.46013477e-01 -1.52979344e-01 -8.39676261e-01
9.93055627e-02 -3.47765982e-01 -1.18989654e-01 -6.68096781e-01
4.48878407e-01 -5.81175268e-01 -3.19193065e-01 4.23706889e-01
-3.79113495e-01 -5.48211515e-01 3.41037840e-01 5.51579356e-01
-6.38354361e-01 2.76340872e-01 3.57351869e-01 2.67288685e-01
-2.36411586e-01 6.76800251e-01 -3.19019228e-01 6.75535619e-01
1.08936334e+00 -1.07201231e+00 -2.48853788e-01 -3.00760448e-01
9.46467593e-02 7.41613030e-01 2.59038568e-01 -4.42721426e-01
2.09363312e-01 -7.18411803e-01 -3.07529531e-02 -2.09625289e-01
1.60653368e-01 -4.59277213e-01 -2.64976919e-01 4.88655090e-01
-7.53328979e-01 5.21148920e-01 -2.96519965e-01 1.21537483e+00
2.69619703e-01 -3.27320307e-01 8.76829982e-01 -1.52347624e-01
2.88461179e-01 8.51824343e-01 -9.38853920e-01 3.89293581e-01
1.41416156e+00 1.42399505e-01 -3.00874770e-01 -8.77846122e-01
-7.98549414e-01 5.40984154e-01 6.11131191e-01 3.50069463e-01
5.06502032e-01 -7.35243142e-01 -8.08822989e-01 -3.89778674e-01
-6.23314798e-01 -1.63116738e-01 -1.75922647e-01 1.11658432e-01
-2.57372439e-01 -2.32395947e-01 4.14772555e-02 -2.09431276e-01
-1.38723075e+00 7.82343090e-01 3.09381992e-01 -7.62547255e-01
-4.07806575e-01 5.44402361e-01 -1.64227679e-01 4.27628666e-01
5.02321601e-01 5.27943313e-01 1.88788995e-01 -7.46802688e-02
8.01933110e-01 6.95226610e-01 -3.66932511e-01 2.19449643e-02
1.20648900e-02 -4.19158787e-01 -5.21403909e-01 -4.81219918e-01
1.18570507e+00 -1.44556090e-01 -5.56520410e-02 5.24509728e-01
1.01523805e+00 -2.56417722e-01 -1.31925082e+00 -6.61321044e-01
1.73201874e-01 -7.48545647e-01 -1.74195409e-01 -3.29566628e-01
-7.09874988e-01 6.78398609e-01 -4.45696078e-02 5.19767582e-01
8.28611672e-01 3.01542152e-02 1.11663139e+00 2.81593084e-01
7.87977576e-01 -8.80208611e-01 5.97345114e-01 5.80669701e-01
5.66730022e-01 -8.24251473e-01 -2.70601928e-01 -3.81447524e-01
-4.22338933e-01 1.05415988e+00 4.92089540e-01 -3.97174776e-01
8.66394937e-01 9.17061031e-01 -3.07796925e-01 1.90169215e-01
-1.20458281e+00 4.94293064e-01 -8.50364208e-01 4.58143681e-01
-1.34204403e-01 3.31584275e-01 -4.30981010e-01 2.37790838e-01
1.22304596e-01 5.61884530e-02 1.21495235e+00 1.35529029e+00
-7.61915267e-01 -1.39498091e+00 -5.76708376e-01 6.72238111e-01
-1.29739428e+00 -4.91825119e-03 4.78327721e-02 1.07996389e-01
-3.22328478e-01 9.63358998e-01 2.20194951e-01 3.76081645e-01
2.02070192e-01 -8.00697058e-02 9.34149325e-02 -1.10593367e+00
-7.09411740e-01 -2.54004717e-01 1.36153758e-01 -3.65117133e-01
4.61452454e-01 -8.48536551e-01 -1.06191111e+00 -9.10727620e-01
-2.05013491e-02 -7.67988572e-03 4.66888458e-01 9.19440866e-01
-9.18577239e-03 5.85305393e-02 1.25625527e+00 -2.05922306e-01
-1.42589772e+00 -5.10122716e-01 -7.09017456e-01 4.41253901e-01
4.40141410e-01 -8.80899653e-02 -7.45914936e-01 -4.23016064e-02] | [4.4743242263793945, 3.178588390350342] |
abe77bcc-20e4-4055-8232-8c137e77530f | macsaar-at-semeval-2016-task-11-zipfian-and | null | null | https://aclanthology.org/S16-1155 | https://aclanthology.org/S16-1155.pdf | MacSaar at SemEval-2016 Task 11: Zipfian and Character Features for ComplexWord Identification | null | ['Josef van Genabith', 'Marcos Zampieri', 'Liling Tan'] | 2016-06-01 | null | null | null | semeval-2016-6 | ['complex-word-identification'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.40109395980835, 3.6928813457489014] |
25f1b871-fe2d-4df1-955c-fac0ed0ed082 | understanding-the-capabilities-of-large | 2305.16151 | null | https://arxiv.org/abs/2305.16151v1 | https://arxiv.org/pdf/2305.16151v1.pdf | Understanding the Capabilities of Large Language Models for Automated Planning | Automated planning is concerned with developing efficient algorithms to generate plans or sequences of actions to achieve a specific goal in a given environment. Emerging Large Language Models (LLMs) can answer questions, write high-quality programming code, and predict protein folding, showcasing their versatility in solving various tasks beyond language-based problems. In this paper, we aim to explore how LLMs can also be used for automated planning. To do so, we seek to answer four key questions. Firstly, we want to understand the extent to which LLMs can be used for plan generation. Secondly, we aim to identify which pre-training data is most effective in facilitating plan generation. Thirdly, we investigate whether fine-tuning or prompting is a more effective approach for plan generation. Finally, we explore whether LLMs are capable of plan generalization. By answering these questions, the study seeks to shed light on the capabilities of LLMs in solving complex planning problems and provide insights into the most effective approaches for using LLMs in this context. | ['Andrea Loreggia', 'Francesco Fabiano', 'Lior Horesh', 'Biplav Srivastava', 'Francesca Rossi', 'Keerthiram Murugesan', 'Bharath Muppasani', 'Vishal Pallagani'] | 2023-05-25 | null | null | null | null | ['protein-folding'] | ['natural-language-processing'] | [ 4.91533220e-01 6.00213170e-01 -2.67814666e-01 -2.04443783e-01
-6.15775347e-01 -5.23081899e-01 6.72401011e-01 4.09920007e-01
-1.73790351e-01 8.61426532e-01 5.68427980e-01 -5.09022534e-01
-2.66052991e-01 -1.00934207e+00 -6.39315665e-01 -1.90680087e-01
-1.71759859e-01 6.95158303e-01 5.16338311e-02 -4.21472490e-01
7.57524848e-01 7.02869296e-01 -1.57017803e+00 5.40822029e-01
1.04456937e+00 2.78910697e-01 6.77443802e-01 5.13967335e-01
-3.11703593e-01 1.13947105e+00 -5.65089524e-01 1.12021029e-01
5.47051840e-02 -6.64835334e-01 -1.48319316e+00 8.13430473e-02
-1.90534249e-01 -1.26104876e-01 1.99662521e-01 7.22776413e-01
1.74035043e-01 2.80814260e-01 3.50882351e-01 -9.75029647e-01
-3.96483168e-02 7.33303487e-01 3.24493766e-01 1.04282066e-01
1.12706316e+00 6.29829288e-01 8.41830254e-01 -2.17986032e-01
8.35929155e-01 1.25384545e+00 2.50112981e-01 6.79694891e-01
-1.15743482e+00 -5.00969440e-02 2.61406638e-02 1.34881914e-01
-1.18673646e+00 -5.57583630e-01 6.06572986e-01 -4.35374171e-01
1.40850031e+00 3.50675583e-01 7.14338124e-01 7.67859101e-01
4.33215231e-01 6.93781435e-01 1.20576322e+00 -6.90495133e-01
3.22956473e-01 3.97934802e-02 -1.32911548e-01 8.29748333e-01
9.82172340e-02 3.16372901e-01 -7.65705109e-01 -2.69479513e-01
5.70794821e-01 -5.38549244e-01 -2.65073538e-01 -3.96456048e-02
-1.34448934e+00 1.00131834e+00 1.39014795e-01 5.69001794e-01
-4.92531657e-01 6.77047595e-02 3.22897881e-01 2.66475230e-01
1.17356693e-02 1.52814925e+00 -5.54124713e-01 -4.26387101e-01
-6.27636909e-01 6.35254443e-01 1.02036691e+00 9.89846587e-01
8.00618768e-01 -1.12295255e-01 -2.74188191e-01 2.89080590e-01
2.78599113e-01 -3.74328047e-02 4.93325233e-01 -1.09254944e+00
6.07640624e-01 8.26517463e-01 3.06357592e-01 -5.76404095e-01
-7.46835470e-01 1.90124050e-01 -1.18528098e-01 2.71657892e-02
3.23674887e-01 -2.31682584e-01 -5.53577900e-01 1.57926214e+00
8.54250416e-02 -2.69863337e-01 3.13892305e-01 5.46211898e-01
4.37931299e-01 8.61397326e-01 2.92117774e-01 -5.59073448e-01
1.26672614e+00 -1.07461715e+00 -3.43742937e-01 -6.05526149e-01
1.28097808e+00 -5.95230222e-01 1.22974074e+00 2.28470489e-01
-1.06036639e+00 -6.07401609e-01 -6.05611563e-01 1.45278454e-01
-1.65571898e-01 -1.50594413e-01 9.13902402e-01 4.52711910e-01
-1.02233624e+00 9.12404537e-01 -8.70951772e-01 -5.89428246e-01
6.87925518e-02 3.35613221e-01 -7.08355308e-02 -2.29262322e-01
-1.08009100e+00 1.19543445e+00 8.46131504e-01 -2.87597835e-01
-9.32938516e-01 -3.84092987e-01 -9.25189555e-01 -1.21290684e-01
6.06243253e-01 -6.72051609e-01 1.41709983e+00 -6.99808478e-01
-1.52313423e+00 5.09935379e-01 -3.63457918e-01 -5.60043752e-01
-5.30400500e-02 2.41443768e-01 -6.63987920e-02 5.14886938e-02
2.74593353e-01 7.94216454e-01 2.05111742e-01 -9.21137273e-01
-5.84329486e-01 -2.66940266e-01 5.44996798e-01 4.13235545e-01
7.97387883e-02 3.18390191e-01 3.18357311e-02 -1.51931822e-01
7.11501315e-02 -1.09402323e+00 -8.22370946e-01 -7.40544736e-01
-3.45648438e-01 -5.50725877e-01 8.87546465e-02 -5.48469603e-01
1.36524439e+00 -1.64122140e+00 4.54309881e-01 9.63976681e-02
-8.50031003e-02 2.33073980e-01 -2.59397537e-01 1.13903558e+00
3.15438449e-01 2.42444307e-01 -2.69307256e-01 3.58770609e-01
-2.62095156e-04 4.09793496e-01 -2.15348065e-01 -3.47292244e-01
4.62655932e-01 1.16632891e+00 -8.69836330e-01 -5.04290640e-01
3.44477817e-02 -9.12525356e-02 -7.39166558e-01 3.67996335e-01
-9.06082511e-01 9.03318405e-01 -9.85329330e-01 6.59843028e-01
-8.37662816e-02 -1.60796106e-01 4.75693345e-01 4.06215400e-01
-4.14426208e-01 7.55377889e-01 -8.38240623e-01 1.45953310e+00
-4.90527302e-01 3.73880684e-01 -3.45903665e-01 -9.45025384e-01
9.79081452e-01 1.25071302e-01 2.94535965e-01 -7.25748479e-01
-1.25467122e-01 2.46203989e-01 1.80182606e-01 -8.65648270e-01
4.31902677e-01 -4.30950850e-01 -2.37885952e-01 6.44399583e-01
-3.39626819e-01 -6.88076735e-01 6.04577363e-01 -3.02845389e-01
1.18112159e+00 3.73561800e-01 7.07078874e-01 -2.94783324e-01
5.16296685e-01 6.85129702e-01 4.73059356e-01 9.14820492e-01
-1.00837061e-02 2.46686544e-02 5.76039016e-01 -7.69393086e-01
-7.90553153e-01 -2.21621349e-01 3.31277221e-01 1.08739126e+00
-1.58393994e-01 -7.38940537e-01 -9.01350558e-01 -4.46278900e-01
-4.52146262e-01 1.13309312e+00 -1.89946800e-01 4.59544137e-02
-8.25135171e-01 -6.57071888e-01 4.23104852e-01 4.59595233e-01
3.75171721e-01 -1.82341611e+00 -1.32993865e+00 4.00329292e-01
-5.78545928e-01 -7.15631664e-01 -1.11867137e-01 3.05253297e-01
-1.20501173e+00 -1.16497135e+00 -2.64521599e-01 -7.68558681e-01
6.47153974e-01 1.12284914e-01 1.27114141e+00 3.67551744e-01
1.06899276e-01 3.67002308e-01 -5.87930620e-01 -4.97061074e-01
-9.29646432e-01 3.80272597e-01 -3.06551103e-02 -8.04412484e-01
4.02423054e-01 -4.22811538e-01 3.52230109e-02 2.09790573e-01
-1.02437496e+00 4.44000453e-01 6.02187991e-01 4.64858353e-01
4.49464440e-01 2.96394408e-01 4.74317700e-01 -1.11789691e+00
1.26548910e+00 -2.23205402e-01 -4.16746646e-01 4.34617341e-01
-7.92553723e-01 6.09916449e-01 7.03998208e-01 -1.10551685e-01
-1.03848875e+00 2.12856129e-01 -5.48847020e-01 5.74410379e-01
-6.21579647e-01 9.68963742e-01 -1.26526698e-01 -1.00375399e-01
9.60022926e-01 3.74810994e-01 -1.54368430e-01 -1.67415649e-01
2.31828779e-01 2.08078504e-01 -1.73899129e-01 -1.07408273e+00
4.59839612e-01 -2.04181243e-02 1.29007944e-03 -7.55773544e-01
-7.54615903e-01 -2.68905312e-01 -6.90301418e-01 3.27339284e-02
8.89961481e-01 -3.55795622e-01 -5.45362890e-01 -1.43603459e-02
-9.79523659e-01 -9.14164364e-01 -2.95001179e-01 9.50694531e-02
-1.26346052e+00 2.75996208e-01 -4.37519640e-01 -7.44486928e-01
-1.77341238e-01 -1.48194790e+00 9.86340642e-01 4.71283555e-01
-8.74245584e-01 -8.93466830e-01 2.29595020e-01 5.09319305e-01
3.06439489e-01 1.37248605e-01 1.16062415e+00 -6.75009787e-01
-7.83933640e-01 1.85588002e-01 2.82676846e-01 -1.77617148e-01
-3.29806507e-02 -3.56908858e-01 -3.86731625e-01 -8.88864323e-02
4.97785918e-02 -5.44939101e-01 1.81577042e-01 3.16200107e-01
9.65471566e-01 -6.25399649e-01 -5.05173326e-01 2.62073427e-01
1.09205258e+00 6.75595403e-01 7.97438264e-01 6.71358049e-01
1.80988029e-01 1.15538096e+00 1.17714906e+00 2.78082579e-01
3.91561270e-01 7.11581469e-01 1.41716674e-01 5.00606894e-01
2.34981090e-01 -4.77285326e-01 2.51165956e-01 3.81604314e-01
-3.67705315e-01 8.10562745e-02 -1.42635536e+00 3.35967273e-01
-1.73571515e+00 -1.07982552e+00 -4.48081940e-02 1.84878588e+00
8.92886162e-01 1.32887498e-01 1.65123791e-01 -1.19777672e-01
1.03702885e-03 1.88198552e-01 -3.87932628e-01 -5.68284750e-01
2.59946585e-01 3.39096963e-01 1.38525471e-01 5.97437203e-01
-7.65680373e-01 1.22190630e+00 6.73009920e+00 4.25478578e-01
-7.98728883e-01 -2.92132050e-01 5.41078269e-01 4.45360422e-01
-3.15191150e-01 4.38408256e-01 -7.34072685e-01 5.20258546e-02
1.28382611e+00 -1.88159868e-01 8.10440958e-01 8.55110943e-01
5.49788713e-01 -3.35783690e-01 -1.16205835e+00 1.49351701e-01
-1.48802176e-01 -1.61070514e+00 -2.71241814e-02 1.30851954e-01
5.98140299e-01 -3.47054958e-01 -5.20422161e-01 6.17866993e-01
3.59455049e-01 -1.39052582e+00 6.04639053e-01 6.42878890e-01
2.40208149e-01 -6.77221775e-01 6.90602899e-01 1.08051610e+00
-1.00149953e+00 -2.70890117e-01 -2.71017313e-01 -6.42111063e-01
4.25227314e-01 -1.69992715e-01 -1.28038478e+00 5.51746309e-01
1.76368341e-01 1.23715162e-01 -5.61410487e-01 8.83658469e-01
-4.92018282e-01 3.58892739e-01 4.45106532e-03 -4.58341509e-01
4.61716741e-01 -4.49511856e-02 3.48772585e-01 1.06233776e+00
2.37906456e-01 4.50578332e-01 5.63749433e-01 7.75538802e-01
6.66723251e-01 1.87627628e-01 -8.98806393e-01 -6.35097682e-01
2.27248505e-01 6.05066299e-01 -8.20880413e-01 -1.74496382e-01
-1.71199933e-01 7.14941204e-01 4.64829296e-01 1.70214698e-01
-4.28365648e-01 -8.76165405e-02 4.06729877e-01 3.27967912e-01
-2.69584656e-01 -4.62093323e-01 -3.38586807e-01 -8.15720856e-01
-3.44149590e-01 -1.50267816e+00 2.60801166e-01 -9.29166913e-01
-5.26448846e-01 6.42206371e-01 2.91330904e-01 -7.30736315e-01
-6.14859223e-01 -4.20423359e-01 -5.95757842e-01 7.26869285e-01
-1.29328811e+00 -9.63816524e-01 -1.38620306e-02 1.28197476e-01
8.61980200e-01 -1.17548019e-01 1.09448266e+00 -3.63175243e-01
-2.95523643e-01 -1.46621183e-01 -4.44852889e-01 -3.06014061e-01
1.88427657e-01 -8.77659142e-01 4.36343819e-01 7.80775428e-01
1.92786798e-01 8.95951271e-01 9.55248117e-01 -8.82371664e-01
-1.50252283e+00 -9.20988977e-01 1.29359949e+00 -5.85532367e-01
2.59461254e-01 1.11890584e-01 -7.41995633e-01 7.31052756e-01
-6.71787560e-02 -8.40935886e-01 5.66299379e-01 3.18252057e-01
2.67998964e-01 4.80417252e-01 -1.07422423e+00 7.08946645e-01
1.07952845e+00 -2.95585603e-01 -7.55132973e-01 6.53913558e-01
7.86175132e-01 -6.34625554e-01 -7.47547328e-01 3.21536422e-01
9.64041427e-02 -1.07572150e+00 8.10462773e-01 -1.01747489e+00
7.77023613e-01 -2.97003001e-01 -1.18436627e-01 -1.26213729e+00
-3.46720606e-01 -7.07528830e-01 2.27085903e-01 8.20558310e-01
6.17391586e-01 -5.85396171e-01 7.95619428e-01 8.56388688e-01
-3.87389481e-01 -1.14416146e+00 -6.07747078e-01 -5.54122031e-01
7.38567784e-02 -6.64223611e-01 7.94927776e-01 6.57021224e-01
5.14832616e-01 3.31164509e-01 -2.73836970e-01 8.02538618e-02
5.64005598e-03 3.19817275e-01 8.27684343e-01 -1.04741502e+00
-4.72860068e-01 -3.64115328e-01 2.24019145e-03 -1.02547705e+00
3.47633243e-01 -8.40939462e-01 2.33059555e-01 -1.86860764e+00
8.40601549e-02 -5.51695108e-01 3.52585793e-01 6.13959014e-01
-1.35723520e-02 -4.43887621e-01 4.26158518e-01 1.93029135e-01
-4.91653293e-01 2.07548335e-01 1.34972537e+00 7.21233860e-02
-6.89137697e-01 3.28129143e-01 -1.11931741e+00 6.78817391e-01
1.15720844e+00 -3.72769594e-01 -6.17515564e-01 -1.61893710e-01
4.84802395e-01 6.09035611e-01 -1.06149241e-01 -7.98341453e-01
1.83901727e-01 -1.15672147e+00 -1.61314040e-01 -1.45729929e-01
3.68287176e-04 -4.42272007e-01 2.75829196e-01 6.95130110e-01
-5.29995322e-01 2.23087907e-01 3.49235415e-01 2.97205865e-01
-1.55199811e-01 -9.78215337e-01 3.96032423e-01 -7.50274003e-01
-9.19860184e-01 -9.95520279e-02 -9.00722206e-01 8.21795091e-02
1.21862650e+00 -1.64258033e-01 -7.52386823e-02 -3.45831156e-01
-6.07861578e-01 4.75825936e-01 6.46079004e-01 2.43036523e-01
5.50363064e-01 -9.81020033e-01 -3.44901919e-01 5.22147156e-02
2.11978376e-01 -4.10338081e-02 -3.23867127e-02 6.37108386e-01
-8.16992819e-01 1.10453498e+00 -2.71926284e-01 -1.40197963e-01
-1.19639599e+00 6.06344521e-01 2.39788070e-01 -7.48433411e-01
-4.62386936e-01 9.51489747e-01 -3.99087071e-02 -6.74883664e-01
-5.31707518e-02 -3.41857642e-01 -2.39972904e-01 -2.90170461e-01
5.64192176e-01 9.77186114e-02 -2.26042375e-01 -3.67999822e-01
-3.70031029e-01 2.48285562e-01 2.19359726e-01 -1.24058247e-01
1.48314512e+00 1.17972866e-01 -2.57927001e-01 2.00810462e-01
4.19880301e-01 -1.28523320e-01 -1.02797282e+00 1.25226468e-01
5.79453826e-01 -1.46574214e-01 -4.31551576e-01 -7.31746972e-01
-1.27343222e-01 6.19935691e-01 -3.72270606e-02 4.00320381e-01
1.19229007e+00 2.63665766e-01 4.99815285e-01 7.08039641e-01
1.00626123e+00 -1.06292772e+00 2.58565933e-01 8.29675198e-01
9.95408773e-01 -1.11918032e+00 1.44075872e-02 -3.15089136e-01
-1.08876896e+00 1.25900078e+00 5.65277219e-01 3.29926193e-01
-2.82079983e-03 1.35453582e-01 -2.68532991e-01 -4.58104938e-01
-8.84632945e-01 -4.17778939e-01 1.44830808e-01 7.07933366e-01
6.43871367e-01 1.89720571e-01 -5.60491085e-01 1.13938853e-01
-6.02654517e-01 8.96888748e-02 6.53620005e-01 1.26215744e+00
-8.91489208e-01 -1.63294578e+00 -5.37523210e-01 6.42807841e-01
3.12610413e-03 -1.95620768e-02 -7.28730619e-01 5.19883335e-01
1.41023532e-01 1.23214793e+00 -5.22927165e-01 -3.63671601e-01
3.31996948e-01 5.47986805e-01 6.36801898e-01 -1.19092941e+00
-5.55789948e-01 -2.97861099e-01 6.84467196e-01 -7.28253782e-01
-3.96047503e-01 -7.83660471e-01 -1.32770920e+00 -5.44103347e-02
-1.07783169e-01 4.23620880e-01 1.61912963e-01 1.25366688e+00
2.38776326e-01 4.89568293e-01 5.34563661e-02 -6.52737021e-01
-5.94843268e-01 -7.26025701e-01 -4.89652157e-02 4.27439250e-03
-1.77670762e-01 -3.19667220e-01 2.38204896e-01 5.81883229e-02] | [4.302672863006592, 1.1118580102920532] |
4e106037-ba94-41fd-bc4c-664ec533519b | improving-chinese-grammatical-error-detection | null | null | https://aclanthology.org/2022.findings-acl.233 | https://aclanthology.org/2022.findings-acl.233.pdf | Improving Chinese Grammatical Error Detection via Data augmentation by Conditional Error Generation | Chinese Grammatical Error Detection(CGED) aims at detecting grammatical errors in Chinese texts. One of the main challenges for CGED is the lack of annotated data. To alleviate this problem, previous studies proposed various methods to automatically generate more training samples, which can be roughly categorized into rule-based methods and model-based methods. The rule-based methods construct erroneous sentences by directly introducing noises into original sentences. However, the introduced noises are usually context-independent, which are quite different from those made by humans. The model-based methods utilize generative models to imitate human errors. The generative model may bring too many changes to the original sentences and generate semantically ambiguous sentences, so it is difficult to detect grammatical errors in these generated sentences. In addition, generated sentences may be error-free and thus become noisy data. To handle these problems, we propose CNEG, a novel Conditional Non-Autoregressive Error Generation model for generating Chinese grammatical errors. Specifically, in order to generate a context-dependent error, we first mask a span in a correct text, then predict an erroneous span conditioned on both the masked text and the correct span. Furthermore, we filter out error-free spans by measuring their perplexities in the original sentences. Experimental results show that our proposed method achieves better performance than all compared data augmentation methods on the CGED-2018 and CGED-2020 benchmarks. | ['TingHao Yu', 'Shengkang Song', 'Tao Yang', 'Huihui Cai', 'Shulin Liu', 'Tianchi Yue'] | null | null | null | null | findings-acl-2022-5 | ['grammatical-error-detection'] | ['natural-language-processing'] | [ 2.29853362e-01 -4.62090820e-02 6.61319673e-01 -4.20093089e-01
-6.60930574e-01 -1.39607146e-01 6.92908317e-02 1.63684309e-01
-4.07636732e-01 8.06781530e-01 2.99476206e-01 -3.37043285e-01
5.29106855e-01 -9.19587195e-01 -6.51765108e-01 -4.15126801e-01
5.59292436e-01 2.82700241e-01 8.65699444e-03 -3.95184487e-01
2.21268788e-01 -2.38992229e-01 -1.32134974e+00 1.99418455e-01
1.81656790e+00 5.59025168e-01 5.71001172e-01 6.74424648e-01
-4.35879409e-01 7.88333118e-01 -1.19229519e+00 -5.14386833e-01
-3.15173626e-01 -9.77074206e-01 -5.80534041e-01 1.78595349e-01
-1.28992692e-01 -2.49778166e-01 -2.73149282e-01 1.51933169e+00
5.46327055e-01 8.90013874e-02 2.78729141e-01 -8.48822236e-01
-1.03059137e+00 8.41048241e-01 -5.45249358e-02 2.30302542e-01
3.49812359e-01 4.82557043e-02 6.30778134e-01 -1.19100916e+00
3.42216462e-01 1.52347398e+00 5.57431757e-01 9.23343897e-01
-9.17297125e-01 -7.21503317e-01 3.01825792e-01 1.27158433e-01
-1.14987576e+00 -2.92937934e-01 8.16963851e-01 -2.03715086e-01
7.98036039e-01 3.70602340e-01 3.66934180e-01 1.23018813e+00
2.13745087e-01 8.16995680e-01 9.81572628e-01 -7.04673767e-01
7.78792426e-02 -8.97759199e-02 2.17900336e-01 4.25494999e-01
3.49386126e-01 -3.18888575e-02 -2.21382409e-01 9.43444520e-02
2.18762353e-01 7.36784041e-02 -4.37563151e-01 8.50749016e-01
-1.04857206e+00 7.59484053e-01 1.51334167e-01 5.01681447e-01
-1.88997433e-01 -1.39516249e-01 5.49888313e-01 2.59662807e-01
6.54625356e-01 2.29170412e-01 -4.45771307e-01 -3.20536405e-01
-7.12104559e-01 2.97637582e-01 4.05415297e-01 1.13541782e+00
6.80594802e-01 2.66190559e-01 -3.84765357e-01 9.12207484e-01
3.96668047e-01 6.47796750e-01 9.55356419e-01 -2.31492952e-01
1.03570771e+00 6.71501160e-01 1.51309133e-01 -1.25612223e+00
-2.19186679e-01 -5.32870412e-01 -1.08563876e+00 -4.45493817e-01
1.93281516e-01 -4.15393084e-01 -8.88758600e-01 1.76825655e+00
1.12018667e-01 -8.91934037e-02 6.18461482e-02 6.69031024e-01
9.79916573e-01 7.59057701e-01 2.45406181e-01 -5.18269479e-01
9.54707444e-01 -8.99741530e-01 -1.19939744e+00 -4.87278104e-01
9.18143570e-01 -1.01365280e+00 1.51274455e+00 2.47348636e-01
-8.22096288e-01 -6.88966691e-01 -9.39317942e-01 4.03071232e-02
-1.16887331e-01 5.34133196e-01 -5.58321103e-02 5.74995220e-01
-4.78150845e-01 4.00004655e-01 -7.07381368e-01 2.11711645e-01
1.10926308e-01 -3.43879700e-01 3.56097221e-02 -1.57494336e-01
-1.56587267e+00 5.92711627e-01 6.66764677e-01 3.72898221e-01
-4.38486457e-01 -2.70855933e-01 -1.05329990e+00 3.37204500e-03
2.41667256e-01 -2.68884540e-01 1.30682898e+00 -1.07365656e+00
-1.13069248e+00 4.82184887e-01 -5.01558721e-01 -2.77026057e-01
6.11855268e-01 -2.36753404e-01 -8.86649251e-01 -4.45364714e-01
1.57870904e-01 2.33447924e-02 5.69244385e-01 -9.87795949e-01
-4.98045623e-01 -2.93992996e-01 -4.51716781e-01 1.59981236e-01
-3.33271682e-01 2.41393879e-01 -2.72157252e-01 -1.16665196e+00
2.82996237e-01 -6.99104428e-01 -2.53266037e-01 -7.51834631e-01
-7.94527173e-01 -3.37011784e-01 7.05120146e-01 -1.18553317e+00
1.92970240e+00 -2.07959557e+00 -1.30245462e-01 -1.29898205e-01
8.74650031e-02 5.22163808e-01 -2.42591113e-01 2.25338623e-01
1.27922162e-01 5.02692759e-01 -4.51475292e-01 -5.91283143e-01
-1.52949348e-01 3.46965879e-01 -3.90095651e-01 -1.32472619e-01
3.21384043e-01 6.66751742e-01 -1.16679728e+00 -4.34897304e-01
-6.85366690e-02 8.46423395e-03 -4.69751507e-01 5.06980538e-01
-4.62805361e-01 6.67177558e-01 -4.05618876e-01 4.04340744e-01
9.66078043e-01 1.77957490e-01 -3.81527469e-02 2.42800593e-01
-2.74710241e-03 7.34328568e-01 -1.15686166e+00 1.22106385e+00
-3.77688378e-01 2.85714895e-01 -4.01593596e-01 -7.58259118e-01
1.20679462e+00 2.46045575e-01 -3.47048134e-01 -6.01305783e-01
2.42862388e-01 4.29587752e-01 9.78246853e-02 -7.70086646e-01
7.42000639e-01 -2.30451748e-01 -2.79874831e-01 2.42306590e-01
-2.12284237e-01 1.53493369e-02 4.51669753e-01 1.16496749e-01
1.02249074e+00 -1.62325367e-01 -1.42036229e-01 2.25884244e-01
6.69352531e-01 -2.41794989e-01 1.34318411e+00 6.17735922e-01
-1.01571374e-01 8.15947056e-01 4.14983988e-01 -1.26402140e-01
-9.97792006e-01 -6.49592221e-01 1.01340644e-01 6.33226991e-01
-6.52087759e-03 -7.55444467e-01 -1.01508117e+00 -9.73714828e-01
-3.96147847e-01 1.06109285e+00 -3.68206769e-01 -4.37488347e-01
-7.32925594e-01 -8.94767284e-01 7.88397670e-01 4.85463232e-01
8.13539267e-01 -1.34300101e+00 3.53563800e-02 6.42362177e-01
-7.82918751e-01 -1.01849234e+00 -7.95588851e-01 -3.62829030e-01
-7.33782172e-01 -1.11850822e+00 -3.70633692e-01 -8.61518383e-01
1.05814123e+00 -8.11523274e-02 1.03914869e+00 6.12222135e-01
-1.83756032e-03 -4.07204360e-01 -8.57477307e-01 -7.06553638e-01
-1.06719565e+00 -1.32572353e-01 7.90114924e-02 -1.14162959e-01
5.42992294e-01 -8.59097913e-02 -2.45160207e-01 1.43703043e-01
-1.01530957e+00 8.48088264e-02 4.40258622e-01 1.14115703e+00
5.36402225e-01 2.20054299e-01 7.78509080e-01 -1.02838480e+00
1.06138098e+00 -3.94940555e-01 -3.74405056e-01 3.14781755e-01
-7.01070487e-01 1.63751133e-02 9.92593765e-01 -4.32535946e-01
-1.26597297e+00 -3.24940175e-01 -6.40933573e-01 -2.11810499e-01
-1.00941904e-01 7.57263720e-01 -4.51023012e-01 5.07743835e-01
4.99906600e-01 6.32622421e-01 -2.47807980e-01 -6.88181877e-01
-1.03930473e-01 1.02149832e+00 4.39502567e-01 -4.99586582e-01
5.90812325e-01 -4.06598330e-01 -4.75358427e-01 -5.41693568e-01
-8.07611227e-01 8.90720934e-02 -2.73005605e-01 -6.42543063e-02
6.46859884e-01 -7.30281472e-01 -9.78395492e-02 1.00354409e+00
-1.65199566e+00 7.18836114e-02 1.01514235e-02 3.90185505e-01
-5.54721570e-03 8.43233585e-01 -8.35226655e-01 -9.51887429e-01
-6.35123968e-01 -1.05397487e+00 7.72083282e-01 5.94853222e-01
-1.22219741e-01 -7.79370725e-01 -9.19629559e-02 1.14795715e-01
3.75833541e-01 3.45239602e-02 9.86117661e-01 -6.99671686e-01
-1.91649005e-01 -2.00042486e-01 5.18985949e-02 9.60533559e-01
2.30205774e-01 1.31017864e-01 -5.09301126e-01 -4.20483425e-02
1.31334394e-01 -1.56887040e-01 5.53899229e-01 -1.69796899e-01
1.37810051e+00 -6.44682109e-01 1.63581334e-02 2.50335515e-01
1.13198578e+00 3.07423472e-01 8.70831370e-01 9.40805674e-03
6.44006848e-01 2.42187694e-01 9.30959642e-01 5.36274374e-01
3.47640693e-01 3.00740510e-01 1.63631722e-01 9.31823850e-02
7.39141628e-02 -8.03325713e-01 3.11583549e-01 1.42933226e+00
1.77860856e-01 -5.48337102e-01 -8.39186490e-01 6.58028007e-01
-1.65982389e+00 -7.65738785e-01 -7.64257908e-01 2.14110851e+00
1.27351665e+00 2.44169995e-01 -3.89482886e-01 4.44003046e-01
1.03391433e+00 -1.75216496e-01 -2.21552342e-01 -3.96504343e-01
-4.19816703e-01 2.47557774e-01 -8.29593167e-02 3.72967690e-01
-7.48677433e-01 1.01975787e+00 5.42726517e+00 8.11855495e-01
-9.08672810e-01 1.80794775e-01 7.71397710e-01 2.11426854e-01
-7.00295269e-01 1.07326031e-01 -1.03512204e+00 1.40036690e+00
7.71950424e-01 -1.60497636e-01 1.61802500e-01 6.16928160e-01
5.86178064e-01 -3.50400619e-03 -6.76771462e-01 8.81778836e-01
2.27162004e-01 -8.55623603e-01 6.23270683e-02 -3.97920817e-01
7.47176647e-01 -3.29131037e-01 -2.57168472e-01 6.58677161e-01
1.93400726e-01 -9.02727008e-01 7.27626741e-01 6.84658885e-01
6.07155383e-01 -7.59769440e-01 1.06677628e+00 7.78511167e-01
-8.08890820e-01 6.94404542e-02 -7.88661122e-01 -3.63088131e-01
1.95291415e-01 1.24046123e+00 -4.23917204e-01 4.57533568e-01
6.11513555e-01 4.76526618e-01 -7.65691280e-01 9.14768577e-01
-7.32433677e-01 1.06552899e+00 -7.70619661e-02 -3.20247918e-01
-1.17395587e-01 -2.79983282e-01 5.50325572e-01 1.21370578e+00
7.66866505e-01 5.27594276e-02 4.51849541e-03 1.11642528e+00
-4.04405504e-01 3.64883155e-01 -2.53152877e-01 -1.23323612e-01
8.41114044e-01 8.28241587e-01 -1.10423520e-01 -4.26088810e-01
-2.88310587e-01 1.21514714e+00 3.98422778e-01 2.77849853e-01
-8.99307907e-01 -8.92990053e-01 2.84849703e-01 -1.90063775e-01
8.09154194e-03 -9.87010524e-02 -4.06113982e-01 -1.52163005e+00
4.77621436e-01 -1.19824600e+00 1.17154248e-01 -8.48963678e-01
-1.31086564e+00 6.53618813e-01 -6.35849893e-01 -1.24411988e+00
-1.01296686e-01 -2.08820924e-01 -8.26163590e-01 1.23794663e+00
-1.18788612e+00 -6.66928411e-01 -5.69318235e-01 1.60492957e-01
8.99487793e-01 2.19338182e-02 7.77709961e-01 4.85733956e-01
-9.70416009e-01 9.54468012e-01 1.07915580e-01 4.78941232e-01
6.87435567e-01 -1.11148620e+00 4.70598072e-01 1.41143048e+00
-1.31934360e-01 6.52047753e-01 6.71470940e-01 -1.18542135e+00
-7.17618167e-01 -1.45824063e+00 1.60107660e+00 -2.43793011e-01
2.02181116e-01 -3.11707050e-01 -1.30236375e+00 7.28548229e-01
-1.67546839e-01 -5.74104637e-02 3.34694862e-01 -1.69448793e-01
6.38007894e-02 9.18329582e-02 -9.67546999e-01 5.31908453e-01
9.91410017e-01 -2.62282461e-01 -8.75180662e-01 2.57549226e-01
9.58059430e-01 -7.34579027e-01 -3.67144525e-01 5.31805634e-01
-1.88770205e-01 -8.88808787e-01 1.82732642e-01 -6.08566701e-01
7.47912526e-01 -5.10930240e-01 1.69467032e-01 -1.80170643e+00
-1.15525633e-01 -3.79896104e-01 3.51225259e-03 1.69621074e+00
4.10529345e-01 -5.87374330e-01 4.47771549e-01 4.80869621e-01
-6.23478055e-01 -6.71884060e-01 -7.40390599e-01 -7.82049537e-01
2.11581022e-01 -5.04666030e-01 9.27303612e-01 7.59575963e-01
-1.28587246e-01 1.93564579e-01 -4.69588459e-01 7.73274675e-02
1.38871908e-01 -2.79401243e-01 5.32250464e-01 -7.85578787e-01
-1.91316038e-01 -2.76908666e-01 -1.64750472e-01 -1.20538998e+00
1.57034583e-02 -5.45577884e-01 6.08542085e-01 -1.35852849e+00
4.75626849e-02 -6.33711398e-01 1.74601361e-01 2.90243506e-01
-1.15654886e+00 -2.44257137e-01 -8.79463404e-02 4.13862318e-02
-3.16624969e-01 9.95075703e-01 1.17203617e+00 -5.04289493e-02
-1.05926067e-01 8.63511637e-02 -6.47709906e-01 6.50741518e-01
8.58314455e-01 -5.88571548e-01 -1.88130364e-01 -6.78877711e-01
3.83369952e-01 -1.34756982e-01 4.94066030e-02 -9.28626180e-01
-5.46747819e-02 -9.25988480e-02 2.22926736e-01 -6.78460419e-01
-5.13870537e-01 -3.94568563e-01 -6.01238050e-02 5.52165806e-01
-1.56641781e-01 3.14479768e-01 3.23582031e-02 6.41541541e-01
-5.11401832e-01 -5.54727733e-01 6.48334563e-01 -1.24971971e-01
-2.98212200e-01 3.54853123e-02 -4.84068185e-01 4.12223309e-01
6.73972070e-01 2.89620191e-01 -3.88052762e-01 -3.12621385e-01
-5.99655867e-01 3.00942212e-01 2.50553250e-01 5.16288638e-01
8.10710013e-01 -1.37938201e+00 -1.04844224e+00 3.96042049e-01
1.89147398e-01 3.25424969e-01 4.91367787e-01 5.98820031e-01
-5.36362350e-01 1.44539624e-01 4.57504511e-01 -2.90051430e-01
-1.32488596e+00 3.79679829e-01 4.12828505e-01 -4.30956244e-01
-1.86456278e-01 8.65590870e-01 -1.67258814e-01 -6.83820486e-01
9.38322842e-02 -4.14079159e-01 -5.42900860e-02 -4.79095995e-01
6.70861661e-01 3.07691902e-01 3.41803223e-01 -4.94100094e-01
4.21043150e-02 5.60019091e-02 -1.15778059e-01 2.84628421e-01
8.64393890e-01 -2.55242229e-01 -3.05985093e-01 4.16734278e-01
8.26461792e-01 2.46904209e-01 -7.74576187e-01 -3.65637600e-01
-5.18329255e-03 -4.84025627e-01 -2.36268625e-01 -8.56595933e-01
-7.77758360e-01 9.08245146e-01 2.65687048e-01 1.32834643e-01
1.20309997e+00 -3.33837211e-01 1.27731884e+00 1.63760602e-01
2.40908384e-01 -1.37822151e+00 6.77559972e-02 9.84527469e-01
9.17881250e-01 -1.28255594e+00 -5.46834886e-01 -6.32770061e-01
-4.50905830e-01 1.03165817e+00 1.09474301e+00 1.00318886e-01
3.37992132e-01 -1.37472972e-02 1.30789187e-02 3.43693942e-01
-4.61559117e-01 3.08145918e-02 1.05348051e-01 3.44010502e-01
5.94018757e-01 1.19280525e-01 -9.73750055e-01 1.31452811e+00
-4.06785101e-01 -7.69402683e-02 6.66870892e-01 8.05252671e-01
-4.80478853e-01 -1.36845219e+00 -4.99763846e-01 5.84526479e-01
-4.24798608e-01 -4.05080199e-01 -2.59375453e-01 3.08605939e-01
4.25853848e-01 1.35795414e+00 2.18776278e-02 -5.46144068e-01
4.67873096e-01 2.63595521e-01 2.01872364e-01 -8.05362940e-01
-3.99294734e-01 -1.67224973e-01 1.50657058e-01 2.55496055e-02
1.74705654e-01 -5.52672327e-01 -1.54914296e+00 -2.82525063e-01
-7.09346533e-01 3.32070857e-01 4.65669870e-01 1.22252643e+00
4.34852391e-01 7.33589053e-01 7.51792431e-01 2.46789362e-02
-7.18592584e-01 -1.64600813e+00 -3.77454907e-01 8.77335608e-01
-3.35880741e-02 -2.17006281e-01 -5.42382658e-01 -7.56585076e-02] | [10.9917631149292, 10.779138565063477] |
c19c67c4-3024-40a8-8cfb-831f294ce628 | toward-multi-agent-reinforcement-learning-for | 2305.08723 | null | https://arxiv.org/abs/2305.08723v1 | https://arxiv.org/pdf/2305.08723v1.pdf | Toward Multi-Agent Reinforcement Learning for Distributed Event-Triggered Control | Event-triggered communication and control provide high control performance in networked control systems without overloading the communication network. However, most approaches require precise mathematical models of the system dynamics, which may not always be available. Model-free learning of communication and control policies provides an alternative. Nevertheless, existing methods typically consider single-agent settings. This paper proposes a model-free reinforcement learning algorithm that jointly learns resource-aware communication and control policies for distributed multi-agent systems from data. We evaluate the algorithm in a high-dimensional and nonlinear simulation example and discuss promising avenues for further research. | ['Dominik Baumann', 'Sebastian Trimpe', 'Lukas Kesper'] | 2023-05-15 | null | null | null | null | ['multi-agent-reinforcement-learning'] | ['methodology'] | [-1.77679598e-01 -1.76626295e-02 -3.12981904e-01 2.24895850e-02
-4.05978322e-01 -4.21799600e-01 5.68539917e-01 4.65038091e-01
-5.55549622e-01 1.56641448e+00 -2.94831038e-01 -3.91712129e-01
-5.24541318e-01 -8.33797574e-01 -3.92319173e-01 -8.44525397e-01
-7.22553194e-01 6.34419322e-01 1.23693980e-01 -1.90404743e-01
-1.77349765e-02 4.26514328e-01 -9.64649796e-01 -5.47963560e-01
5.98538518e-01 7.38219798e-01 1.84945211e-01 1.18469703e+00
3.21970463e-01 9.08473670e-01 -1.02327394e+00 5.65609217e-01
3.93124104e-01 -5.18232524e-01 -3.63090724e-01 4.70564812e-01
-4.23854709e-01 -6.29666448e-01 -3.28084797e-01 7.56766856e-01
5.97692430e-01 3.86127025e-01 4.25289005e-01 -1.57923448e+00
-1.77626923e-01 5.84047496e-01 -4.41104233e-01 2.53603846e-01
1.80238307e-01 3.97052258e-01 5.60769141e-01 -3.67687270e-02
4.34888214e-01 1.27172303e+00 2.93583006e-01 5.74479163e-01
-1.35148680e+00 -8.01663995e-01 4.83381361e-01 3.55170742e-02
-1.15364599e+00 -2.14622006e-01 5.25168538e-01 -1.51619211e-01
9.68691170e-01 -8.27202126e-02 9.84710157e-01 8.40501845e-01
4.78727371e-01 5.52545309e-01 1.27403784e+00 -4.34445441e-01
6.17890120e-01 -2.00241536e-01 -4.14101034e-01 5.52421868e-01
4.07239676e-01 6.91945851e-01 -1.83921039e-01 -4.97057259e-01
1.23794866e+00 3.65396664e-02 -9.47662145e-02 -6.03115916e-01
-1.22969568e+00 9.23667371e-01 1.54221743e-01 -1.10406056e-01
-9.52635407e-01 4.76719290e-01 2.35657156e-01 9.95740235e-01
2.99625367e-01 5.59264958e-01 -5.54626346e-01 -6.86677024e-02
-4.32620347e-01 5.20161033e-01 1.18152428e+00 8.40390742e-01
5.70828438e-01 5.43224931e-01 2.64899880e-01 3.58052731e-01
3.53982568e-01 6.47577822e-01 6.00580201e-02 -1.36594248e+00
2.25213662e-01 1.32408351e-01 6.21579885e-01 -7.09862769e-01
-6.23087764e-01 -1.97970971e-01 -8.63952696e-01 6.76915884e-01
3.58027816e-01 -1.32233799e+00 -3.42263699e-01 1.50875521e+00
5.64261198e-01 3.02598029e-01 5.63758612e-01 9.29246306e-01
-2.73176223e-01 8.91290188e-01 -2.25491539e-01 -9.36708748e-01
6.00529015e-01 -9.73151743e-01 -8.53925049e-01 -5.60716949e-02
3.34143728e-01 -5.93089163e-01 3.48894238e-01 2.12070107e-01
-1.24932945e+00 5.17561734e-02 -1.01069093e+00 1.07589126e+00
-5.56129403e-02 -3.43034714e-01 5.06536603e-01 2.55886585e-01
-1.12862599e+00 4.81707156e-01 -1.03306079e+00 -5.17831266e-01
-1.63182244e-01 6.87789202e-01 3.35711062e-01 2.25063100e-01
-1.10878313e+00 9.77186799e-01 1.60184801e-01 -6.82464316e-02
-1.47102201e+00 -5.15095055e-01 -5.34130335e-01 -9.75084156e-02
1.10758686e+00 -7.46126711e-01 1.86108363e+00 -8.94379318e-01
-2.14067531e+00 -3.68835866e-01 3.20363462e-01 -6.12413704e-01
6.06060624e-01 2.72009313e-01 -2.91903287e-01 2.56243467e-01
-1.97228357e-01 1.83690667e-01 9.15470541e-01 -1.45930731e+00
-9.84577179e-01 3.68062049e-01 5.20811975e-01 5.56263804e-01
-3.58190447e-01 -2.96294570e-01 4.41314697e-01 -5.24566770e-01
-6.98162556e-01 -9.71144915e-01 -8.39422464e-01 8.39800984e-02
-1.18904717e-01 -9.28668454e-02 1.00692844e+00 1.57280073e-01
9.55999732e-01 -1.74465656e+00 1.91195905e-01 4.88610923e-01
2.03266591e-01 1.22979939e-01 -2.45289028e-01 1.12428093e+00
4.06968266e-01 -1.16259754e-01 2.84030765e-01 2.42322907e-02
7.64365420e-02 5.13244927e-01 -9.70066041e-02 5.59169829e-01
-2.30995864e-01 2.79594749e-01 -9.15954471e-01 -4.33100939e-01
3.57528925e-01 2.09039301e-01 -3.57599884e-01 5.38048387e-01
-5.29880106e-01 6.31209433e-01 -1.04205155e+00 1.69058353e-01
1.51240095e-01 -3.07295978e-01 6.57454908e-01 5.33899188e-01
-2.60338724e-01 -1.29033685e-01 -1.41254950e+00 8.77241075e-01
-6.56645417e-01 3.04531991e-01 1.10171950e+00 -1.30914485e+00
5.67484081e-01 6.98410869e-01 9.81134295e-01 -6.04154170e-01
2.14274451e-01 -6.10248260e-02 1.97243348e-01 -4.51227725e-01
1.20249182e-01 -2.67516643e-01 -5.54328971e-02 9.29885209e-01
-4.03060138e-01 -3.26561958e-01 4.31341857e-01 6.74221590e-02
1.34122610e+00 -5.29868543e-01 5.31281412e-01 -3.84053290e-01
3.62901807e-01 1.80534333e-01 5.15984118e-01 8.92729759e-01
-4.25346971e-01 -6.62249148e-01 3.95814657e-01 -2.19773173e-01
-9.70503747e-01 -9.10389006e-01 3.86916250e-01 1.04858446e+00
3.66703957e-01 -1.40681803e-01 -4.91522491e-01 -3.11913997e-01
2.99778700e-01 3.09330791e-01 -4.14468616e-01 7.32232183e-02
-5.13574183e-01 -6.45693302e-01 -9.21787992e-02 1.57031462e-01
1.81170389e-01 -9.08660233e-01 -7.02325284e-01 8.40120018e-01
5.41581929e-01 -9.50201631e-01 -5.74146748e-01 1.60413042e-01
-5.67086518e-01 -1.41286480e+00 -4.62840259e-01 -7.10484028e-01
7.82777071e-01 2.82618791e-01 6.34021163e-01 3.14960599e-01
-6.41085804e-02 1.12530947e+00 4.42764424e-02 -5.01907945e-01
-6.72076464e-01 6.27206033e-03 4.65813845e-01 -1.77438378e-01
-3.86067867e-01 -4.46005613e-01 -5.50997853e-01 3.52035910e-01
-6.93710327e-01 -1.73813313e-01 4.84597504e-01 9.61112201e-01
4.20710474e-01 4.46597248e-01 1.24201393e+00 -5.04319489e-01
1.24700308e+00 -5.08756757e-01 -1.27747786e+00 3.63633819e-02
-7.28574634e-01 -1.47022426e-01 1.21286130e+00 -7.75034904e-01
-9.97632563e-01 3.62092815e-02 4.90142316e-01 -8.55552256e-02
-1.39226019e-01 4.29414272e-01 4.11403060e-01 -3.25112104e-01
2.81433105e-01 8.85939375e-02 5.92867076e-01 9.43671465e-02
2.17954908e-02 4.70134139e-01 -2.26022899e-01 -8.13357830e-01
7.36150801e-01 1.24409221e-01 1.19110547e-01 -9.09302533e-01
-3.15065354e-01 4.89994958e-02 -5.23182899e-02 -4.14102316e-01
2.63088256e-01 -8.17780972e-01 -1.43262744e+00 2.11510241e-01
-7.80774415e-01 -1.01288557e+00 -2.22661227e-01 7.96263933e-01
-9.44867790e-01 -7.88803250e-02 -7.64808953e-01 -1.11642611e+00
8.82079080e-02 -8.66368294e-01 3.76652002e-01 3.43483478e-01
-3.72861028e-02 -1.42537034e+00 4.46400285e-01 -3.63884687e-01
8.36827695e-01 2.98119307e-01 5.12269497e-01 -4.70377594e-01
-6.98192120e-01 -1.37162432e-01 2.80288190e-01 -1.64396558e-02
2.69630849e-01 -4.60138954e-02 -7.65761808e-02 -9.76174653e-01
-2.95660883e-01 -6.52401567e-01 1.22075655e-01 6.79575503e-01
7.83935547e-01 -8.30841184e-01 -5.66757619e-01 -9.61098373e-02
1.52880502e+00 5.51381409e-01 -3.43841612e-01 1.84602767e-01
2.09751874e-01 4.06747729e-01 4.98489112e-01 1.09469151e+00
6.74223542e-01 4.10122901e-01 6.84240520e-01 -4.63762023e-02
4.06993151e-01 9.16571245e-02 5.18559396e-01 6.27011359e-01
1.88024625e-01 -5.20299852e-01 -7.38746762e-01 4.50579762e-01
-2.07262206e+00 -7.69362152e-01 5.15212834e-01 2.02753878e+00
7.61552572e-01 7.55253211e-02 5.07365763e-01 -4.73721325e-02
8.48029137e-01 -1.00471132e-01 -9.18669641e-01 -4.66890723e-01
1.23950958e-01 -3.11172634e-01 5.31498432e-01 8.23903918e-01
-9.15185034e-01 5.20690262e-01 7.42566347e+00 4.25390095e-01
-9.80116069e-01 -4.04081084e-02 2.28166327e-01 -3.56779903e-01
3.59818339e-01 -2.68398821e-01 -5.45004487e-01 2.82242715e-01
1.36207783e+00 -8.60895216e-01 1.00355494e+00 5.56330204e-01
9.75870430e-01 -3.34910870e-01 -9.70929027e-01 6.60945594e-01
-5.34190655e-01 -1.20136404e+00 -4.69938964e-01 2.28711158e-01
1.01260364e+00 5.28384522e-02 -4.17578042e-01 3.02702814e-01
1.04279852e+00 -6.77424312e-01 1.83963284e-01 3.90981078e-01
1.64349541e-01 -9.55304980e-01 5.27579963e-01 5.82553148e-01
-1.12142277e+00 -6.10360742e-01 -1.86355054e-01 -7.49454677e-01
3.70460629e-01 7.96190426e-02 -6.64143443e-01 1.29874051e-01
2.05220267e-01 6.27074957e-01 8.41411427e-02 1.07680750e+00
1.49760857e-01 5.86962938e-01 -6.20385885e-01 -6.63868546e-01
4.43832666e-01 3.11395880e-02 6.47370338e-01 7.51527488e-01
1.34078696e-01 3.47359598e-01 1.33158231e+00 2.29206920e-01
1.36854127e-01 -1.89410970e-01 -7.39816844e-01 -1.74869791e-01
8.37918043e-01 1.19786358e+00 -7.22323835e-01 -4.63460296e-01
-5.36725342e-01 2.71792829e-01 2.87981629e-01 7.54494727e-01
-5.78254580e-01 -1.99037030e-01 1.01794708e+00 -1.00241557e-01
3.14649194e-01 -7.89610863e-01 3.56497943e-01 -7.96760559e-01
-4.42879379e-01 -9.03045058e-01 3.86906445e-01 -1.15780979e-01
-1.49917889e+00 1.15528010e-01 7.20375478e-02 -1.13690400e+00
-9.05281007e-01 -2.87855297e-01 -5.68023205e-01 2.07559243e-01
-1.52030122e+00 -5.66380739e-01 1.98695332e-01 8.36166680e-01
5.53763688e-01 -3.41786921e-01 9.91184175e-01 2.59853210e-02
-7.74044752e-01 2.20607780e-02 6.19764864e-01 -1.07804134e-01
5.82867503e-01 -1.29356301e+00 -5.19689739e-01 4.07110363e-01
-5.43817580e-01 7.09269643e-02 9.52841818e-01 -3.26274663e-01
-1.93547344e+00 -1.09267902e+00 7.11934194e-02 4.19849515e-01
1.21614110e+00 -1.32304467e-02 -6.32992506e-01 6.28379285e-01
8.31897020e-01 8.00002292e-02 3.75763685e-01 -3.25082660e-01
4.60685670e-01 -3.01226079e-01 -1.06413686e+00 6.24024332e-01
4.00451273e-01 1.05754547e-02 5.80896949e-03 5.36224663e-01
5.00505269e-01 -3.00066888e-01 -1.04075480e+00 -1.34572238e-01
1.24799065e-01 -1.80281371e-01 6.21569753e-01 -5.61956942e-01
-3.49037230e-01 -2.38474652e-01 -2.99589466e-02 -2.02772474e+00
-1.19236812e-01 -1.21877074e+00 -4.92085516e-01 7.90611386e-01
3.97189200e-01 -9.36280608e-01 6.10821247e-01 4.46776122e-01
2.09940180e-01 -5.92538416e-01 -9.77274477e-01 -1.09602070e+00
1.99470416e-01 1.28214642e-01 2.84754932e-01 8.53621125e-01
4.22697663e-01 4.82956916e-01 -3.79979461e-01 4.26665932e-01
8.96016717e-01 3.57909724e-02 6.20414674e-01 -7.38632917e-01
-2.96854734e-01 -4.43532974e-01 1.74363256e-01 -9.10286188e-01
6.13135278e-01 -1.31984293e-01 1.01196632e-01 -1.52656674e+00
-2.89773107e-01 -3.99385154e-01 -8.09646249e-02 2.96823531e-01
2.82157630e-01 -3.11210603e-01 2.53064692e-01 1.12248557e-02
-1.09766340e+00 8.40012789e-01 1.27129161e+00 -1.82632193e-01
-3.66248101e-01 1.67571485e-01 -1.59133852e-01 5.28606832e-01
1.49399567e+00 -2.78119385e-01 -8.11109900e-01 -4.13895011e-01
-2.38862813e-01 8.32741380e-01 2.92054266e-01 -7.59843707e-01
6.74953520e-01 -1.09229791e+00 -1.83424875e-02 6.09534867e-02
2.40581512e-01 -1.19535935e+00 -1.01029605e-01 1.07344937e+00
-5.38711548e-01 5.14771760e-01 6.97979629e-02 1.12425745e+00
2.51394808e-02 1.70717850e-01 8.54950845e-01 -1.29230514e-01
-4.97670710e-01 3.95183414e-01 -1.25262809e+00 -3.85205150e-02
1.45988882e+00 3.46289426e-01 -2.96282291e-01 -8.98787916e-01
-8.58892560e-01 1.04899037e+00 1.77751794e-01 2.11073756e-01
5.72654843e-01 -9.33790028e-01 -6.07186377e-01 2.52426043e-02
-5.66682935e-01 -3.67871761e-01 -7.79436454e-02 4.55146044e-01
-8.18267614e-02 2.48812109e-01 -2.38834739e-01 -2.62880087e-01
-9.82661903e-01 7.12383032e-01 7.25434542e-01 -2.17999026e-01
-3.23477000e-01 -7.17623383e-02 -4.05728430e-01 -3.56346458e-01
3.26758116e-01 -3.29040177e-02 5.24457805e-02 -1.84158877e-01
4.84475672e-01 5.18975437e-01 -5.50231338e-01 3.10086808e-03
-1.24020495e-01 2.10332870e-01 1.55464426e-01 -3.77699137e-01
1.35440326e+00 -6.33268118e-01 4.84686159e-02 4.03649420e-01
5.67906678e-01 -6.96675658e-01 -1.75408649e+00 -3.83289009e-01
-7.01401383e-02 -2.93136001e-01 2.75579304e-01 -6.60762966e-01
-1.13172185e+00 2.69402564e-01 2.63177812e-01 8.95252407e-01
1.03958011e+00 -4.02999610e-01 4.69201535e-01 8.69143724e-01
7.25458205e-01 -1.38021493e+00 3.42115104e-01 5.89524269e-01
7.42157221e-01 -1.14924276e+00 -6.88015390e-03 -8.36201385e-02
-3.63395274e-01 1.17664218e+00 9.65494037e-01 -5.77366948e-01
1.06126320e+00 8.04057240e-01 6.00650571e-02 1.56880528e-01
-1.53478861e+00 -1.50468960e-01 -6.29614890e-01 8.60739887e-01
1.04813568e-01 9.86576453e-02 -4.46159989e-01 -3.75268161e-02
3.12586427e-01 -3.00020445e-02 1.05428267e+00 1.35171795e+00
-6.64211333e-01 -1.12559712e+00 -4.60541815e-01 4.79258895e-01
-2.95960516e-01 4.91642922e-01 -1.48510933e-01 1.02621174e+00
-5.74308336e-01 1.32435155e+00 1.68073878e-01 2.57616818e-01
3.48856807e-01 -4.61198509e-01 3.99733841e-01 -4.96973336e-01
-4.94761586e-01 4.65274870e-01 2.40248784e-01 -4.70172614e-01
-6.19391084e-01 -5.47897935e-01 -1.48994493e+00 -8.05579603e-01
-2.08182007e-01 4.13272947e-01 4.09200430e-01 7.73270667e-01
5.29562652e-01 8.09543908e-01 9.40549433e-01 -9.84531462e-01
-1.09925961e+00 -6.39111280e-01 -6.68794930e-01 -2.66306579e-01
9.21090484e-01 -8.52892458e-01 -3.31152052e-01 -2.39164233e-01] | [4.001095771789551, 2.2108500003814697] |
11fde43e-b067-4cac-bcf6-415d6516215c | mer-gcn-micro-expression-recognition-based-on | 2004.08915 | null | https://arxiv.org/abs/2004.08915v1 | https://arxiv.org/pdf/2004.08915v1.pdf | MER-GCN: Micro Expression Recognition Based on Relation Modeling with Graph Convolutional Network | Micro-Expression (ME) is the spontaneous, involuntary movement of a face that can reveal the true feeling. Recently, increasing researches have paid attention to this field combing deep learning techniques. Action units (AUs) are the fundamental actions reflecting the facial muscle movements and AU detection has been adopted by many researches to classify facial expressions. However, the time-consuming annotation process makes it difficult to correlate the combinations of AUs to specific emotion classes. Inspired by the nodes relationship building Graph Convolutional Networks (GCN), we propose an end-to-end AU-oriented graph classification network, namely MER-GCN, which uses 3D ConvNets to extract AU features and applies GCN layers to discover the dependency laying between AU nodes for ME categorization. To our best knowledge, this work is the first end-to-end architecture for Micro-Expression Recognition (MER) using AUs based GCN. The experimental results show that our approach outperforms CNN-based MER networks. | ['Wen-Huang Cheng', 'Hong-Han Shuai', 'Hong-Xia Xie', 'Ling Lo'] | 2020-04-19 | null | null | null | null | ['micro-expression-recognition'] | ['computer-vision'] | [ 5.9352990e-02 2.8235134e-01 -1.7879961e-01 -6.4764053e-01
2.4559048e-01 2.5158726e-02 3.9082149e-01 -3.7397444e-01
-7.8128114e-02 2.0611800e-01 -2.5662687e-02 1.8009916e-01
8.7391764e-02 -7.2650182e-01 -5.2705675e-01 -7.0314330e-01
-2.0577955e-01 -1.8061031e-01 -3.2999715e-01 -5.9212160e-01
-4.5286559e-02 7.2394782e-01 -1.4033968e+00 3.8060480e-01
2.8396726e-01 1.5717114e+00 -4.0205973e-01 3.8574204e-01
-1.6474190e-01 1.3176017e+00 -3.7662411e-01 -6.3903946e-01
-1.2410405e-01 -8.8412261e-01 -7.1522558e-01 1.9542703e-01
2.6116201e-01 -1.4400542e-01 -3.1170017e-01 1.1773317e+00
4.8829246e-01 -1.4185274e-01 6.1284298e-01 -1.6475842e+00
-7.3431742e-01 3.0859658e-01 -8.7671024e-01 2.3476267e-01
3.4745356e-01 -2.5306693e-01 9.7530699e-01 -5.7410693e-01
7.8858751e-01 1.2933527e+00 5.6467724e-01 7.2813624e-01
-6.7091858e-01 -8.4249252e-01 6.6398047e-02 3.5846016e-01
-1.2575816e+00 -4.0930989e-01 1.2745479e+00 -2.1471712e-01
1.0500281e+00 3.1113945e-02 9.4032097e-01 1.3679681e+00
4.0661517e-01 9.8792458e-01 1.0398421e+00 -2.1795471e-01
-4.5088112e-02 -2.3580337e-01 -8.2729183e-02 1.1579617e+00
-2.7748561e-01 -2.5164154e-01 -5.3917265e-01 1.5309823e-01
6.8133670e-01 -9.3207285e-02 -2.1546258e-01 -7.2555900e-02
-4.9974221e-01 5.6669569e-01 8.9721608e-01 5.8957088e-01
-5.4230773e-01 5.6886590e-01 7.3860925e-01 5.6606674e-01
7.1414858e-01 7.2118267e-02 -2.2477430e-01 -3.1528005e-01
-5.0883085e-01 -2.1022429e-01 5.0899208e-01 6.8095541e-01
7.8303236e-01 3.2607180e-01 1.0520133e-01 8.4236389e-01
2.1824765e-01 1.4863440e-02 5.4868978e-01 -5.2406853e-01
3.3944708e-03 1.2268109e+00 -6.3111442e-01 -1.5688027e+00
-7.6083177e-01 -3.5806251e-01 -1.0210963e+00 2.0089456e-01
-2.6779061e-03 -4.3086010e-01 -5.5011952e-01 1.9403425e+00
1.9131972e-01 2.3061079e-01 -1.9836856e-01 1.1240064e+00
1.2982911e+00 4.0096915e-01 2.0323518e-01 -2.2856198e-01
1.1501251e+00 -7.7097756e-01 -9.1154379e-01 9.6549846e-02
1.0418055e+00 -3.0490163e-01 7.0862770e-01 2.3921306e-01
-6.4869612e-01 -5.6173730e-01 -9.1040969e-01 1.4599822e-01
-3.8554475e-01 2.9160601e-01 1.0043854e+00 4.5708624e-01
-1.1110158e+00 6.8185794e-01 -8.6248428e-01 -5.3741181e-01
9.3842721e-01 6.1020720e-01 -9.0925753e-01 4.4710365e-01
-1.0966140e+00 7.0600939e-01 6.9908552e-02 7.3676074e-01
-6.5515280e-01 -2.0881841e-01 -8.4894210e-01 -1.2905481e-01
2.9690573e-01 -1.4844608e-01 8.3605802e-01 -1.8497486e+00
-1.7497261e+00 1.2044345e+00 4.1928221e-02 -2.6761677e-02
1.5718286e-01 7.2053269e-02 -6.6899478e-01 2.5110260e-01
-4.9981064e-01 6.2694585e-01 9.4912159e-01 -7.9948479e-01
7.8108661e-02 -7.6640767e-01 -2.1939317e-02 -7.4578926e-02
-5.5283684e-01 4.6841556e-01 -9.9746615e-02 -2.6522622e-01
2.0360672e-01 -8.3998621e-01 2.3509645e-01 9.6026562e-02
-5.0217271e-01 -7.1649528e-01 1.1014204e+00 -3.5514921e-01
1.2208627e+00 -2.1387358e+00 3.8116059e-01 2.0315985e-01
6.4903283e-01 2.9007810e-01 -3.2437050e-01 2.6839116e-01
-3.8151428e-01 4.1836448e-02 1.8912406e-01 -3.1984559e-01
-3.7222791e-02 1.6756174e-01 1.4772071e-01 7.0472711e-01
4.1893259e-01 1.3455359e+00 -7.4102354e-01 -4.9174219e-01
1.9807875e-02 5.3785294e-01 -2.6407111e-01 4.2506558e-01
-1.1641239e-01 3.6139086e-01 -6.4792299e-01 9.2383432e-01
6.0664803e-01 -1.6228279e-01 3.4339735e-01 -5.6537956e-01
2.8413361e-01 -2.8392488e-01 -5.0511819e-01 1.6437987e+00
-4.1173741e-01 9.1813022e-01 3.3541915e-01 -1.3970610e+00
1.3438401e+00 3.1731230e-01 5.8413959e-01 -7.1474117e-01
9.5516843e-01 1.2782423e-01 1.4265218e-01 -7.4594605e-01
-7.7273630e-02 -1.0878334e-01 1.7374154e-02 3.2654655e-01
4.2239618e-01 2.5734970e-01 -1.9826715e-01 3.0415164e-02
9.1740209e-01 3.0311063e-01 3.9730316e-01 -7.7031627e-02
6.1072552e-01 -3.8110504e-01 5.9870285e-01 -7.1695417e-02
-4.0495753e-01 1.7700358e-01 1.0544469e+00 -8.6135483e-01
-3.2879949e-01 -3.3945730e-01 1.1028993e-01 1.1387190e+00
5.7091575e-02 -5.7809335e-01 -8.8883936e-01 -8.8242936e-01
-1.8211974e-01 -1.8085499e-03 -1.2525181e+00 -3.2719359e-01
-3.0420160e-01 -6.5115702e-01 8.1439668e-01 6.9263768e-01
8.6262506e-01 -1.4971433e+00 -5.8323753e-01 2.4885567e-01
2.2422630e-02 -1.2199383e+00 -3.8248073e-02 8.2648881e-03
-5.8862513e-01 -1.1591991e+00 -4.0934607e-01 -8.2056862e-01
6.9558692e-01 -8.4777236e-02 8.8775831e-01 4.1152033e-01
-4.7434554e-01 2.7123061e-01 -5.6194532e-01 -4.6636173e-01
-2.4575660e-01 -3.0186580e-02 -3.1747043e-02 5.9312510e-01
8.1870240e-01 -9.7323424e-01 -5.3892231e-01 1.9946769e-01
-6.4330590e-01 -3.1578206e-02 6.9138908e-01 4.8372385e-01
4.4846481e-01 -3.6482525e-01 4.5339131e-01 -8.0017686e-01
7.8848988e-01 -3.3849111e-01 -1.9537017e-01 2.4011023e-01
-1.3344909e-01 -2.2766489e-01 5.2773064e-01 -3.5134411e-01
-7.0736724e-01 8.2671516e-02 -3.4083781e-01 -7.4734282e-01
-4.0280917e-01 5.0410885e-01 -2.3125197e-01 -5.3390604e-01
4.9232754e-01 4.5946930e-02 3.0225435e-01 -8.9682870e-02
3.2023329e-01 8.2066882e-01 2.7952805e-01 -1.9835421e-01
1.3461091e-01 4.7456762e-01 3.5837165e-01 -1.0921861e+00
-8.8598293e-01 -1.4963239e-01 -7.3026508e-01 -7.5361872e-01
1.3923976e+00 -6.7589504e-01 -1.1211469e+00 6.7156923e-01
-1.1909097e+00 -3.2882369e-01 4.1376722e-01 1.0177284e-01
-5.3205091e-01 3.6998457e-01 -8.1513894e-01 -8.2988667e-01
-6.5206009e-01 -1.0463362e+00 1.1428057e+00 2.5924399e-01
-4.2077059e-01 -9.5359582e-01 -1.3659756e-01 1.2912555e-01
3.2518849e-01 8.4816188e-01 6.4798427e-01 -2.9406837e-01
-8.7203598e-03 -4.0571567e-01 -2.9699752e-01 3.7763348e-01
1.0724313e-01 2.8932622e-01 -1.0422503e+00 -3.6259506e-02
-1.9057941e-01 -9.4211000e-01 5.7401508e-01 3.1254137e-01
1.4733391e+00 -2.7676821e-01 -2.2614430e-01 8.8070232e-01
1.1376169e+00 1.7243508e-01 8.7880933e-01 1.4465377e-01
9.3022776e-01 6.0962969e-01 3.6182955e-01 4.3370929e-01
2.4045801e-02 7.1070224e-01 8.2415980e-01 -4.6096319e-01
-6.2558590e-03 1.1692397e-02 4.1642240e-01 6.7226654e-01
-7.3798120e-01 -2.1203119e-01 -4.4815615e-01 4.0298935e-02
-1.7870307e+00 -9.3561655e-01 -1.9368891e-01 1.3296125e+00
3.2084793e-01 -1.1074005e-01 1.3980679e-01 -6.1035424e-02
4.2556754e-01 5.8964378e-01 -4.3728209e-01 -9.1300249e-01
-5.6262281e-02 3.7480339e-01 -1.4246403e-01 -4.5276932e-02
-1.0987961e+00 1.1015110e+00 4.8191509e+00 6.1186725e-01
-1.6590382e+00 6.9267578e-02 7.7770919e-01 1.5781920e-01
1.9450022e-01 -5.3496265e-01 -3.4412679e-01 9.5752478e-02
7.3028350e-01 3.0967313e-01 2.0065349e-01 1.0405779e+00
8.8158220e-02 3.9557260e-01 -8.6439061e-01 1.2910886e+00
2.7790505e-01 -1.0142616e+00 -5.7991412e-02 8.7946244e-03
4.5896110e-01 -6.5361366e-02 -1.8501318e-01 1.9833201e-01
-9.2391342e-02 -1.1956594e+00 3.1786692e-01 3.9089200e-01
9.3957430e-01 -7.9559737e-01 9.7279042e-01 3.6345370e-02
-1.3579898e+00 4.6913553e-02 -5.6659889e-01 -2.0503162e-01
-3.5120431e-02 9.8053515e-02 -4.6035182e-01 4.6795884e-01
6.7419481e-01 1.1251720e+00 -3.7233415e-01 2.1245804e-01
-5.0526392e-01 6.6033238e-01 -2.4332406e-02 -4.9161848e-01
4.5179096e-01 -4.9196553e-01 2.1297233e-01 1.1833749e+00
2.5724742e-01 1.9217120e-01 -2.4550563e-01 1.0267812e+00
-4.8208004e-01 3.8529444e-01 -6.9949055e-01 -5.8158398e-01
-2.7054623e-01 1.8322428e+00 -6.0943115e-01 6.2313117e-02
-4.2165893e-01 1.3781775e+00 6.8684071e-01 1.6111827e-01
-8.7416869e-01 -4.1573131e-01 7.4677628e-01 -7.4269749e-02
2.2990342e-02 -4.2197272e-02 3.0724460e-01 -9.0986234e-01
-5.1438060e-02 -6.4673418e-01 3.0218115e-01 -9.6250093e-01
-1.1586357e+00 1.0866427e+00 -5.1020581e-01 -1.1730963e+00
-9.8826870e-02 -8.9689761e-01 -7.9925704e-01 5.4195952e-01
-1.2807637e+00 -1.3900898e+00 -8.4480965e-01 7.9751402e-01
1.7328632e-01 -1.7403124e-01 1.0285174e+00 2.1157663e-02
-8.8084394e-01 7.9461020e-01 -6.1502588e-01 7.1007639e-01
2.2450626e-01 -8.4642774e-01 -1.9740254e-01 4.3426391e-01
2.4123576e-01 3.5433125e-01 3.0108100e-01 -3.2116961e-01
-1.4029835e+00 -9.6881002e-01 7.2447133e-01 -1.0436673e-01
8.4483677e-01 -5.4959208e-01 -7.7179420e-01 7.5719756e-01
3.0495483e-01 5.2307421e-01 8.2699180e-01 6.2265828e-02
-3.6134779e-01 -2.8092825e-01 -8.9111292e-01 4.4167277e-01
1.3916804e+00 -5.4475546e-01 -1.7855965e-01 2.4778968e-01
2.5195888e-01 -3.0494043e-01 -9.8090047e-01 4.5066115e-01
6.2476903e-01 -1.1494879e+00 4.3752533e-01 -9.8940372e-01
9.0471506e-01 7.7824973e-02 1.1279021e-01 -1.2796232e+00
-1.6059594e-01 -5.0133801e-01 -2.5829734e-02 9.9650383e-01
1.8137479e-01 -4.4738600e-01 8.9720577e-01 3.1372035e-01
-1.7168014e-01 -1.1439644e+00 -8.8324207e-01 -5.0183457e-01
-3.2995874e-01 -5.0571984e-01 5.9206533e-01 1.2356964e+00
2.7202460e-01 6.2108701e-01 -4.2850658e-01 -1.8270518e-01
2.3718739e-01 2.7047300e-01 7.6283455e-01 -1.2273064e+00
-4.4447877e-02 -7.1848160e-01 -1.2279899e+00 -7.5749761e-01
5.9111100e-01 -1.0953732e+00 -3.0368713e-01 -1.3815912e+00
-6.7938631e-04 -6.1593551e-02 -2.8255245e-01 7.9948854e-01
1.7596677e-01 4.4631681e-01 -8.3091203e-03 -1.1285108e-01
-7.5756317e-01 8.7487596e-01 1.4676311e+00 -9.0651371e-02
-1.3062902e-01 -2.8004181e-01 -5.6003529e-01 7.4173379e-01
8.3781028e-01 -7.6184228e-02 -4.1650894e-01 -2.0436598e-01
4.2808318e-01 -5.5701070e-02 2.6202580e-01 -8.5517889e-01
9.4774831e-03 7.0005260e-02 4.2725655e-01 -1.4030428e-01
2.8867027e-01 -9.2898542e-01 -1.0529227e-01 2.1984527e-01
-2.2597024e-01 -7.4085998e-03 1.9127582e-01 3.6775714e-01
-5.5064535e-01 2.4960147e-01 7.4320453e-01 -7.8935027e-02
-1.0969620e+00 7.3176837e-01 -4.0042183e-01 -3.7891644e-01
1.2629663e+00 -2.8589633e-01 -8.7862447e-02 -6.4268124e-01
-7.9865730e-01 -1.1494156e-01 -5.7715386e-02 4.8479980e-01
6.7135960e-01 -1.4201875e+00 -5.4255277e-01 7.8512698e-02
3.8444397e-01 -2.7685705e-01 3.4407726e-01 1.1432247e+00
-5.1020414e-01 -4.1670680e-02 -5.2476037e-01 -5.1281762e-01
-1.6912676e+00 2.7235344e-01 7.5372159e-01 1.4522363e-01
-4.1475037e-01 1.2498728e+00 -1.2158937e-02 -3.4445253e-01
6.3677609e-02 -1.6981045e-03 -6.6252971e-01 2.0235904e-01
4.2133844e-01 1.1963644e-01 -1.5494860e-02 -9.4750887e-01
-4.5394498e-01 6.2640685e-01 7.0283934e-02 5.5288947e-01
1.3854585e+00 1.5261330e-01 -6.8378252e-01 2.1923308e-01
1.6606326e+00 -5.0676930e-01 -8.2991481e-01 -1.6088702e-02
-1.3381702e-01 -1.9158722e-01 1.4338489e-01 -4.0258291e-01
-1.8207355e+00 1.3070115e+00 4.9635914e-01 3.5818178e-02
1.4013827e+00 1.8033816e-01 7.2606575e-01 4.0251097e-01
3.6009416e-01 -1.0782087e+00 3.0603158e-01 1.5666479e-01
1.1543751e+00 -1.2287340e+00 -3.3347967e-01 -5.0064808e-01
-7.6278287e-01 1.6059036e+00 1.1226377e+00 -1.2844071e-01
8.0148512e-01 7.9301722e-02 3.2800815e-01 -8.9683610e-01
-4.9534449e-01 -2.4947342e-01 2.6112378e-01 4.2001250e-01
7.5682980e-01 1.2572023e-01 -3.3869252e-01 5.9151542e-01
-1.2669218e-01 2.5870925e-01 5.3478491e-02 6.5068716e-01
-1.9260627e-01 -8.3891988e-01 3.6360487e-01 5.0335282e-01
-5.1127034e-01 2.9144850e-01 -9.9748731e-01 9.3354559e-01
1.9326641e-01 7.7174157e-01 2.2021246e-01 -8.7949920e-01
3.9411569e-01 3.4380011e-02 6.2622619e-01 -2.1780027e-01
-5.1261294e-01 -6.5695904e-02 1.8455584e-01 -1.0808069e+00
-8.9679980e-01 -2.1166100e-01 -1.5366769e+00 -3.7043071e-01
-8.0906510e-02 -1.6704294e-01 4.3656316e-01 8.8551891e-01
4.9224862e-01 5.0670522e-01 7.8323334e-01 -7.3693144e-01
-3.7004821e-02 -1.2086432e+00 -8.4087431e-01 8.2720286e-01
7.8021702e-03 -8.0709499e-01 -2.6210251e-01 -2.9367551e-01] | [13.647160530090332, 1.6395379304885864] |
1ac70edd-96f8-4e6e-93e3-56e62a407cd1 | dna-inspired-online-behavioral-modeling-and | 1602.00110 | null | http://arxiv.org/abs/1602.00110v1 | http://arxiv.org/pdf/1602.00110v1.pdf | DNA-inspired online behavioral modeling and its application to spambot detection | We propose a strikingly novel, simple, and effective approach to model online
user behavior: we extract and analyze digital DNA sequences from user online
actions and we use Twitter as a benchmark to test our proposal. We obtain an
incisive and compact DNA-inspired characterization of user actions. Then, we
apply standard DNA analysis techniques to discriminate between genuine and
spambot accounts on Twitter. An experimental campaign supports our proposal,
showing its effectiveness and viability. To the best of our knowledge, we are
the first ones to identify and adapt DNA-inspired techniques to online user
behavioral modeling. While Twitter spambot detection is a specific use case on
a specific social media, our proposed methodology is platform and technology
agnostic, hence paving the way for diverse behavioral characterization tasks. | ['Maurizio Tesconi', 'Angelo Spognardi', 'Marinella Petrocchi', 'Stefano Cresci', 'Roberto Di Pietro'] | 2016-01-30 | null | null | null | null | ['dna-analysis'] | ['medical'] | [ 1.74441203e-01 -2.40314752e-01 -4.64658767e-01 -7.37318099e-02
-2.88834184e-01 -9.57493067e-01 9.72057760e-01 5.32394469e-01
-6.37113094e-01 6.14924788e-01 8.97448603e-03 -5.60854137e-01
5.05579971e-02 -9.98776138e-01 -3.88649225e-01 -5.77584982e-01
1.17211767e-01 6.54170454e-01 5.34288049e-01 -1.67809486e-01
5.19819915e-01 6.57187998e-01 -1.52525902e+00 1.35882631e-01
7.44626105e-01 4.58103240e-01 -2.79914171e-01 8.39695811e-01
-3.03829797e-02 4.66805249e-01 -4.30730402e-01 -8.88035655e-01
1.45560414e-01 -4.34831440e-01 -7.94465005e-01 8.33240971e-02
1.95012540e-01 -2.55540580e-01 -4.32723165e-01 1.06018603e+00
3.25195551e-01 -8.81227106e-02 7.32165933e-01 -9.95215297e-01
-4.17808235e-01 6.21703506e-01 -2.15068027e-01 1.73083276e-01
4.13454264e-01 4.23776805e-01 8.78273010e-01 -2.46430680e-01
7.95935988e-01 9.46838379e-01 8.17124486e-01 5.71770132e-01
-1.23245215e+00 -2.71230191e-01 -1.70067504e-01 8.99220407e-02
-1.01649690e+00 -3.94577533e-01 4.26552802e-01 -5.77265561e-01
4.96565789e-01 4.82256591e-01 5.93847573e-01 1.50721455e+00
-8.62785801e-02 9.57301140e-01 1.12564826e+00 -3.15672964e-01
4.07415360e-01 3.65136147e-01 8.08083355e-01 6.92865431e-01
7.91711509e-01 -3.47660333e-01 -2.06921175e-01 -6.89503968e-01
8.81483555e-02 9.01280642e-02 7.67881870e-02 -1.89158261e-01
-7.34867334e-01 6.83462858e-01 -2.57051200e-01 6.90459013e-01
-1.48562983e-01 9.46271196e-02 6.94288194e-01 1.43851861e-01
2.90577143e-01 2.27029383e-01 -8.64395946e-02 -4.83425736e-01
-7.66552925e-01 3.38015318e-01 1.11512315e+00 5.31815648e-01
7.41101444e-01 -3.04316968e-01 -4.84729446e-02 6.21740639e-01
1.70501664e-01 4.23167884e-01 7.69893706e-01 -6.97993696e-01
-1.27637982e-01 6.91673338e-01 2.93888658e-01 -1.10945702e+00
-4.60231841e-01 -2.42646620e-01 -3.77687663e-01 -4.57448691e-01
9.24575686e-01 -2.01047095e-03 -2.83220083e-01 1.43611920e+00
1.94306076e-01 3.19844246e-01 -4.22585726e-01 3.21987063e-01
3.36439610e-01 2.47783765e-01 2.04939947e-01 -2.17084512e-01
1.49230874e+00 -3.88622433e-01 -2.06247166e-01 -4.62321341e-02
9.71308947e-01 -3.12157363e-01 1.11377168e+00 3.16045105e-01
-7.36636996e-01 7.55999237e-02 -7.36817181e-01 1.97912619e-01
-6.71057284e-01 -1.23816133e-01 8.04479063e-01 1.48257720e+00
-5.32973588e-01 4.77008134e-01 -7.38923728e-01 -8.71023178e-01
3.81403774e-01 2.42054656e-01 -1.99138910e-01 2.14447137e-02
-9.99215484e-01 5.97016156e-01 8.35101902e-02 -3.22317868e-01
-5.87485373e-01 -3.41151029e-01 -2.28509799e-01 -8.10831636e-02
2.80841202e-01 -4.74551409e-01 1.33659065e+00 -6.43375456e-01
-1.57823992e+00 1.15953267e+00 -1.31108537e-01 -7.33905435e-01
6.53788507e-01 3.35655719e-01 -2.36741751e-01 5.70948832e-02
-1.29748851e-01 -2.55563945e-01 4.48255181e-01 -9.55579877e-01
-2.02040330e-01 -4.41961080e-01 -3.15233096e-02 -7.70333588e-01
-8.02665353e-01 1.55250221e-01 -2.39386663e-01 -4.39719945e-01
-3.49421799e-01 -1.17795706e+00 -3.20740649e-03 -4.52396244e-01
-3.42237115e-01 -1.21319212e-01 4.67504025e-01 -4.62208480e-01
1.23780370e+00 -1.82353067e+00 -1.90204471e-01 5.01253128e-01
3.22666556e-01 7.55474985e-01 6.44885097e-03 8.14857185e-01
5.26786149e-01 4.38941807e-01 -3.31162989e-01 -3.24334294e-01
2.83465236e-01 -2.14739025e-01 -2.91587830e-01 9.04341280e-01
-2.30320916e-01 9.77833152e-01 -1.08988762e+00 -3.12795907e-01
-3.09389066e-02 2.06315577e-01 -6.92812085e-01 -9.12575573e-02
-3.96254718e-01 -3.09527013e-02 -6.77710295e-01 6.87430859e-01
6.87120616e-01 -1.42947093e-01 7.44583607e-01 2.85908341e-01
-2.07789525e-01 3.76344413e-01 -7.15780735e-01 9.35563028e-01
-2.54032463e-01 5.36101401e-01 -7.16603175e-02 -8.89587283e-01
8.39008451e-01 -2.04426363e-01 3.20916712e-01 -5.83444715e-01
6.36913180e-01 3.34065109e-01 8.02655146e-02 -6.19596422e-01
5.15358746e-01 -6.26203716e-02 -1.66061938e-01 9.37273383e-01
-2.78454483e-03 4.57657009e-01 5.29343128e-01 2.92874128e-01
1.38497770e+00 -2.10798636e-01 4.07225907e-01 -2.47330055e-01
7.70125091e-01 -1.16395161e-01 3.44246961e-02 1.06829965e+00
-6.48739338e-01 1.30403250e-01 8.88087273e-01 -1.37144580e-01
-9.15088117e-01 -8.54448318e-01 -7.13239685e-02 1.29921985e+00
-1.39222756e-01 -6.02825880e-01 -1.11424518e+00 -1.00252795e+00
3.16674918e-01 4.31196243e-01 -6.75918996e-01 -2.06081178e-02
-6.80521607e-01 -1.23194063e+00 1.12944913e+00 -5.44626564e-02
3.21413219e-01 -7.13308454e-01 -8.03633332e-02 2.08932176e-01
-4.55999896e-02 -1.13432336e+00 -2.52140701e-01 -3.35981905e-01
-5.70339441e-01 -1.34481728e+00 -4.34347957e-01 -3.48948121e-01
2.78061211e-01 4.23515528e-01 6.66350842e-01 3.21163028e-01
-3.32220882e-01 5.40166199e-01 -3.23188037e-01 -3.04242522e-01
-8.33839893e-01 4.69751567e-01 1.64105728e-01 4.76894230e-01
5.68906724e-01 -7.23120928e-01 -5.33103168e-01 3.34055156e-01
-9.00072575e-01 -7.28317738e-01 4.56924140e-01 3.14291775e-01
-3.17604423e-01 -5.24744332e-01 6.48482919e-01 -1.38399303e+00
8.36060762e-01 -8.26999009e-01 -6.58380866e-01 1.22175045e-01
-5.01275301e-01 -7.35503137e-02 9.23568904e-01 -3.90729398e-01
-8.80481064e-01 4.77162888e-03 -4.62769359e-01 4.97203022e-01
-2.77861983e-01 2.75617391e-01 -3.91296893e-02 -2.41480932e-01
9.59591925e-01 5.90926409e-01 2.81707555e-01 -5.68855166e-01
3.31363916e-01 9.96037245e-01 2.72383451e-01 -3.76817644e-01
6.73980236e-01 8.80586386e-01 -8.47481862e-02 -1.32414520e+00
-2.37406969e-01 -7.72877216e-01 -4.30015951e-01 -1.74980283e-01
5.35048962e-01 -2.31033176e-01 -1.66152692e+00 7.45835066e-01
-7.86052644e-01 -1.56374454e-01 3.56042147e-01 4.64486219e-02
-5.16595364e-01 1.00435436e+00 -8.65404606e-01 -1.10602546e+00
-1.18767172e-01 -6.15975082e-01 8.50246549e-01 -1.59401610e-01
-3.70538592e-01 -1.07679927e+00 4.12893414e-01 4.29951668e-01
3.14471453e-01 8.85653421e-02 7.05197036e-01 -1.13995993e+00
-6.11202776e-01 -7.72680402e-01 -1.87258884e-01 -1.78786379e-03
-7.56238820e-03 2.37562254e-01 -9.74488676e-01 -3.62659544e-02
2.13646200e-02 -1.06368117e-01 9.09246683e-01 -1.97882444e-01
9.59481001e-01 -3.85107905e-01 -5.46462655e-01 4.27616835e-02
1.13604605e+00 -2.81118959e-01 7.90790439e-01 5.01484334e-01
2.69985855e-01 6.49883687e-01 1.96851909e-01 5.46162307e-01
3.09221685e-01 7.24304616e-01 2.47869164e-01 4.78906304e-01
1.78229168e-01 -2.09297076e-01 5.95493913e-01 5.04791141e-01
5.74658066e-02 -5.38543463e-01 -1.05089295e+00 4.09897327e-01
-1.94672024e+00 -1.34764695e+00 -5.70481598e-01 2.11335731e+00
6.78138912e-01 1.09437302e-01 1.04797876e+00 1.41362056e-01
7.35838771e-01 5.08335652e-03 -2.73844488e-02 -3.35448712e-01
-6.48718476e-02 2.10515276e-01 7.15976954e-01 4.97086257e-01
-9.14300025e-01 7.93749034e-01 6.88455296e+00 6.76854491e-01
-9.84454274e-01 2.25450188e-01 3.35027337e-01 -2.99838465e-02
-2.00743467e-01 -1.23350814e-01 -8.85715246e-01 7.63919771e-01
1.30122554e+00 -2.73586333e-01 3.86273682e-01 6.13661408e-01
2.59802431e-01 -1.13051564e-01 -8.41935992e-01 5.66052198e-01
7.72829950e-02 -1.65091515e+00 -3.87512259e-02 3.48793119e-01
3.71897966e-01 1.30564556e-01 -6.64684102e-02 2.88915318e-02
2.11990505e-01 -5.83660960e-01 4.38124627e-01 5.81434608e-01
-1.49572000e-01 -5.41251600e-01 6.12529755e-01 5.43773413e-01
-5.50424635e-01 -3.10372144e-01 -7.66414925e-02 -1.73887506e-01
-6.39165984e-03 5.32550812e-01 -1.08829689e+00 2.07146779e-01
1.07362561e-01 4.75766063e-01 -8.96902502e-01 1.06410038e+00
2.30862379e-01 9.45935309e-01 -1.39982119e-01 -8.22931588e-01
2.34490946e-01 -3.26141417e-01 3.83876055e-01 1.77105391e+00
4.08932231e-02 -3.86972368e-01 -1.58040151e-01 6.41420782e-01
-2.93460637e-01 2.69873381e-01 -6.69185221e-01 -4.70742077e-01
3.39100391e-01 1.28981435e+00 -9.96272624e-01 -4.01131690e-01
-3.51946831e-01 7.76747167e-01 3.82788867e-01 -1.31005589e-02
-6.49883628e-01 -3.66263092e-01 7.81306267e-01 4.20522332e-01
5.55511713e-02 -4.45047200e-01 -2.57456869e-01 -1.29413342e+00
-2.13878378e-01 -9.59406674e-01 3.04881722e-01 1.36164695e-01
-1.43655682e+00 -1.11153491e-01 -1.60440639e-01 -9.28250492e-01
-1.97986796e-01 -8.86658609e-01 -5.00032783e-01 3.62632960e-01
-1.02568412e+00 -9.02230203e-01 -1.12118408e-01 1.56068549e-01
-3.14209200e-02 -1.32061705e-01 5.12135863e-01 4.05715734e-01
-7.92386293e-01 5.41876137e-01 6.22688830e-01 1.46643817e-01
4.28603977e-01 -8.54021966e-01 7.26348460e-01 6.49679840e-01
-8.17784816e-02 1.06029034e+00 8.41713071e-01 -6.77926123e-01
-1.79984474e+00 -9.37726676e-01 9.90402400e-01 -9.08668995e-01
1.25551379e+00 -1.09727418e+00 -8.67818236e-01 6.28007948e-01
-2.15684965e-01 -5.21264613e-01 9.07616436e-01 8.17471445e-02
-5.63471437e-01 1.09178394e-01 -1.17645895e+00 9.17470634e-01
1.14966667e+00 -8.27638507e-01 -1.72590062e-01 7.20007837e-01
1.29920736e-01 3.59697968e-01 -5.32357872e-01 -2.39976510e-01
9.08774257e-01 -1.30237448e+00 8.33679020e-01 -5.48637509e-01
4.65494916e-02 -7.26442710e-02 -9.61242616e-03 -6.56301916e-01
-6.69362918e-02 -7.93751299e-01 6.99593825e-03 1.42968953e+00
2.64506340e-01 -9.16724265e-01 1.12451839e+00 2.70267874e-01
2.61795580e-01 -3.33496243e-01 -5.66554666e-01 -9.83869553e-01
1.31421953e-01 -4.65701461e-01 5.14586389e-01 8.47704530e-01
5.12227118e-01 -3.37505452e-02 -3.96781772e-01 -2.54816562e-01
6.13095284e-01 -8.27350244e-02 1.15646660e+00 -1.19916511e+00
-4.99449790e-01 -7.19011307e-01 -4.46663052e-01 -8.13354015e-01
4.19171035e-01 -8.88561666e-01 -2.30617031e-01 -8.64268124e-01
4.37484741e-01 -1.83359608e-01 1.70082241e-01 -1.73387304e-02
2.15653017e-01 6.58671141e-01 6.75164834e-02 1.15952924e-01
-5.37968159e-01 1.15956157e-01 5.77190936e-01 -1.32268116e-01
-6.29265755e-02 3.42949361e-01 -8.10834110e-01 5.92562914e-01
1.00808656e+00 -4.86825973e-01 -7.80669004e-02 3.93296123e-01
4.59872991e-01 -5.09341836e-01 4.51972663e-01 -6.44209802e-01
6.68382570e-02 -3.66926134e-01 -2.80841887e-01 5.36511764e-02
6.69680238e-02 -3.58642370e-01 -1.31388977e-01 8.77704561e-01
-2.49677390e-01 -4.47485298e-01 -9.51243043e-02 7.69142151e-01
3.81738454e-01 -5.39526403e-01 8.61002147e-01 -2.86562592e-01
-2.62294143e-01 1.37599051e-01 -1.13422084e+00 -1.75911769e-01
9.72159445e-01 -1.48592144e-01 -8.69003892e-01 -3.16097617e-01
-3.90889674e-01 -2.74236858e-01 8.38966072e-01 1.65604189e-01
8.96968506e-03 -7.94935286e-01 -4.58874881e-01 -3.19209397e-02
2.94444978e-01 -1.35595012e+00 8.89742449e-02 1.02027440e+00
-5.46282947e-01 5.93040884e-01 -7.74964914e-02 -3.91127557e-01
-1.35151589e+00 7.41302371e-01 1.85605898e-01 6.84235012e-03
-6.20774478e-02 1.95146963e-01 -3.20595145e-01 -5.00371099e-01
-8.79575163e-02 9.00631100e-02 4.23185565e-02 3.29071671e-01
5.75801492e-01 8.30034256e-01 2.73152649e-01 -5.89295328e-01
-3.56133014e-01 2.78041642e-02 5.53934909e-02 2.73434594e-02
1.03713751e+00 -1.03108421e-01 -3.88878644e-01 4.57094520e-01
1.33274150e+00 5.31616688e-01 -4.36576158e-01 5.25593273e-02
4.24848437e-01 -5.73683918e-01 -5.97956955e-01 -3.05846155e-01
-1.86140448e-01 5.73605239e-01 1.67187944e-01 1.03148043e+00
5.31506062e-01 -5.32366894e-02 8.16819549e-01 6.74817741e-01
3.16406369e-01 -9.26059365e-01 1.08364329e-01 5.52060902e-01
2.51238495e-01 -8.34941447e-01 -1.65079936e-01 -5.18541694e-01
-8.28324109e-02 8.78569067e-01 2.84876943e-01 -8.70830715e-02
5.45588493e-01 2.19407063e-02 -3.04068685e-01 4.18450609e-02
-5.67557395e-01 -4.60620284e-01 -2.13739589e-01 7.16370463e-01
5.07627010e-01 -1.08282872e-01 -9.21492815e-01 3.44761729e-01
-5.19166663e-02 4.28779349e-02 9.33196425e-01 9.15575683e-01
-7.29308784e-01 -1.60529470e+00 -1.89984560e-01 4.19632316e-01
-5.00142574e-01 1.30948812e-01 -1.14014232e+00 8.23726296e-01
-3.44065070e-01 8.33605886e-01 -3.15171480e-01 -4.06262070e-01
3.22586484e-02 4.83604938e-01 5.45122623e-01 -4.81138647e-01
-9.78111863e-01 -4.56704021e-01 4.42751348e-01 -1.92700282e-01
-3.36719334e-01 -8.82647812e-01 -8.63889992e-01 -9.81085360e-01
-7.30842277e-02 4.04161960e-02 7.28216290e-01 9.45740938e-01
4.71418887e-01 -2.68666118e-01 7.80473948e-01 -5.14156580e-01
-4.84068036e-01 -6.82012796e-01 -5.59956014e-01 6.01284623e-01
2.84747869e-01 -3.60061407e-01 -3.57570529e-01 -2.73492724e-01] | [8.188316345214844, 10.200197219848633] |
5d7222d3-e852-4f07-95fe-5ae1071518f2 | open-world-continual-learning-unifying | 2304.10038 | null | https://arxiv.org/abs/2304.10038v1 | https://arxiv.org/pdf/2304.10038v1.pdf | Open-World Continual Learning: Unifying Novelty Detection and Continual Learning | As AI agents are increasingly used in the real open world with unknowns or novelties, they need the ability to (1) recognize objects that (i) they have learned and (ii) detect items that they have not seen or learned before, and (2) learn the new items incrementally to become more and more knowledgeable and powerful. (1) is called novelty detection or out-of-distribution (OOD) detection and (2) is called class incremental learning (CIL), which is a setting of continual learning (CL). In existing research, OOD detection and CIL are regarded as two completely different problems. This paper theoretically proves that OOD detection actually is necessary for CIL. We first show that CIL can be decomposed into two sub-problems: within-task prediction (WP) and task-id prediction (TP). We then prove that TP is correlated with OOD detection. The key theoretical result is that regardless of whether WP and OOD detection (or TP) are defined explicitly or implicitly by a CIL algorithm, good WP and good OOD detection are necessary and sufficient conditions for good CIL, which unifies novelty or OOD detection and continual learning (CIL, in particular). A good CIL algorithm based on our theory can naturally be used in open world learning, which is able to perform both novelty/OOD detection and continual learning. Based on the theoretical result, new CIL methods are also designed, which outperform strong baselines in terms of CIL accuracy and its continual OOD detection by a large margin. | ['Bing Liu', 'Zixuan Ke', 'Tatsuya Konishi', 'Changnan Xiao', 'Gyuhak Kim'] | 2023-04-20 | null | null | null | null | ['class-incremental-learning'] | ['computer-vision'] | [ 5.92015423e-02 9.63714719e-02 -4.28375304e-01 -6.18101005e-03
-4.87973958e-01 -5.28853536e-01 7.07158864e-01 2.73796737e-01
-3.80081326e-01 8.44113708e-01 -1.56865641e-01 -8.93364400e-02
-3.75454873e-01 -7.31647551e-01 -1.04153180e+00 -6.06147051e-01
-4.30629760e-01 7.12645292e-01 4.58050966e-01 -3.61802876e-02
1.27728358e-01 3.01749200e-01 -1.87001634e+00 8.01664367e-02
1.12732971e+00 1.17260373e+00 3.59858960e-01 5.77805281e-01
-2.64491409e-01 8.47666383e-01 -6.70380950e-01 -1.44802064e-01
6.93695486e-01 -2.82717615e-01 -7.53946364e-01 -2.57430773e-04
4.52726841e-01 -1.18577078e-01 3.66544798e-02 1.16585088e+00
4.42626566e-01 2.89193273e-01 8.76604736e-01 -1.71037078e+00
-9.62584019e-01 5.67102969e-01 -5.94064236e-01 3.28355789e-01
2.99329907e-01 1.58387691e-01 1.04175770e+00 -1.18530011e+00
5.80641448e-01 9.59707141e-01 5.72096109e-01 4.26133394e-01
-1.06799448e+00 -6.58433020e-01 4.72210526e-01 3.10324907e-01
-1.20287776e+00 -1.50267571e-01 7.28469729e-01 -5.48039854e-01
9.50484216e-01 2.93047369e-01 7.98558116e-01 9.25621867e-01
1.94472209e-01 1.28580666e+00 1.07077944e+00 -6.66149855e-01
6.07684433e-01 4.13803518e-01 2.40178406e-01 5.88385403e-01
4.32851136e-01 3.89057070e-01 -5.31261742e-01 5.05847037e-02
6.29605353e-01 2.91431159e-01 -7.41860569e-02 -4.93084282e-01
-1.39515793e+00 7.04997778e-01 4.22215819e-01 4.85081494e-01
-5.01188099e-01 -9.38506201e-02 4.16108698e-01 9.76394236e-01
5.35186827e-01 8.70447814e-01 -6.50179982e-01 -7.34045431e-02
-5.99563479e-01 2.80436784e-01 9.06058550e-01 9.47798371e-01
1.01105142e+00 5.45196347e-02 6.27360912e-03 7.97078550e-01
-1.14895545e-01 4.42599654e-01 1.01690066e+00 -5.75329125e-01
2.49773562e-01 7.44833708e-01 2.72869825e-01 -8.85384679e-01
-5.19587576e-01 -6.34458125e-01 -7.75681615e-01 2.27704585e-01
2.93760478e-01 -4.79041487e-02 -8.27740610e-01 1.99489832e+00
4.14540440e-01 5.24487793e-01 2.66459942e-01 5.04176140e-01
6.69942081e-01 7.26324618e-01 -1.21423848e-01 -7.80277908e-01
9.83523846e-01 -1.09786284e+00 -5.78836977e-01 -4.17834580e-01
7.87079990e-01 -3.70812654e-01 1.19602191e+00 5.68080306e-01
-7.04986572e-01 -6.89964533e-01 -1.06466925e+00 2.30096012e-01
-5.19667029e-01 -3.41722459e-01 8.25627446e-01 3.92514616e-01
-8.95315528e-01 1.73297495e-01 -4.56629127e-01 -2.58683503e-01
3.98011178e-01 1.71593666e-01 -3.17635626e-01 -1.26740877e-02
-1.19515944e+00 8.95217955e-01 7.68763244e-01 -2.33480185e-01
-1.06561351e+00 -8.23765874e-01 -6.85838044e-01 -1.92175284e-01
9.20338392e-01 -6.64613545e-01 1.24517763e+00 -1.43740797e+00
-1.17309427e+00 9.45422113e-01 -7.14842603e-02 -6.58210814e-01
6.91185713e-01 -4.07054275e-01 -4.55680877e-01 -1.64058268e-01
4.13833916e-01 5.06085396e-01 1.17058241e+00 -1.23184955e+00
-1.17573273e+00 -3.61104757e-01 2.23248884e-01 4.51601028e-01
-4.91600960e-01 -4.17978168e-01 -1.28224909e-01 -8.34565282e-01
2.03860149e-01 -7.31822133e-01 1.15323320e-01 1.20867277e-02
-2.49961734e-01 -8.20262730e-01 8.63663793e-01 -2.70359874e-01
1.13254404e+00 -2.04627728e+00 6.83559179e-02 7.75645301e-02
5.13162374e-01 3.24876040e-01 -2.85881549e-01 2.06683859e-01
-1.29487991e-01 4.99803051e-02 3.29255918e-03 -2.40673609e-02
9.72959921e-02 1.86539933e-01 -4.92088765e-01 1.48629904e-01
4.57561053e-02 1.20232391e+00 -1.09095716e+00 -2.13359445e-01
2.92876381e-02 -3.97560239e-01 -3.40581328e-01 1.90076753e-01
-5.94578683e-01 8.01569894e-02 -2.93711841e-01 8.05354178e-01
4.31547076e-01 -2.60628819e-01 -2.37346441e-01 1.39050826e-01
-7.57619441e-02 3.65264267e-02 -1.46398270e+00 1.23304033e+00
-3.36146235e-01 4.11576331e-01 -3.96638721e-01 -1.29549038e+00
8.68857265e-01 1.94215938e-01 3.29173654e-01 -8.24220061e-01
-2.61014730e-01 5.02686799e-01 -3.30782160e-02 -5.45640945e-01
3.00985545e-01 -3.05338234e-01 -6.09936602e-02 5.43899536e-01
5.39530627e-02 2.06564426e-01 3.30115020e-01 1.13437831e-01
1.08095443e+00 -2.93345928e-01 6.89822197e-01 -3.45141828e-01
3.84878725e-01 -1.15980923e-01 7.46532500e-01 1.25279927e+00
-3.39854360e-01 -7.05747083e-02 5.53844810e-01 -7.81624436e-01
-8.39571416e-01 -1.22609484e+00 -1.77959532e-01 1.26802671e+00
3.26215714e-01 6.53092787e-02 -1.07672974e-01 -9.99927282e-01
2.34850168e-01 5.41723728e-01 -7.97488868e-01 -2.71634042e-01
-3.07964593e-01 -4.62546319e-01 1.35133252e-01 5.50931036e-01
4.79708493e-01 -1.31834328e+00 -5.20753503e-01 3.29927832e-01
1.02091998e-01 -5.98581731e-01 -3.81889403e-01 5.63047707e-01
-7.86820054e-01 -1.13861203e+00 -9.12425280e-01 -1.02878141e+00
4.66412485e-01 4.93310630e-01 1.06220722e+00 -3.03010680e-02
-7.01203793e-02 4.65804338e-01 -5.54906309e-01 -8.11413050e-01
-3.38215172e-01 3.66591699e-02 6.49552643e-01 -1.74720455e-02
4.34459299e-01 -6.78288817e-01 -2.00647235e-01 4.82500225e-01
-9.81723428e-01 -9.22599584e-02 9.53593552e-01 1.00438571e+00
7.61599064e-01 4.71114874e-01 1.12163162e+00 -9.68313694e-01
5.78045845e-01 -6.44216299e-01 -5.69969654e-01 4.82342899e-01
-8.70246232e-01 4.87272814e-02 5.91951787e-01 -9.93809700e-01
-7.57383943e-01 -2.08830595e-01 1.29735753e-01 -6.50086284e-01
-2.50351816e-01 8.05099249e-01 -1.39430791e-01 6.99584410e-02
8.55140209e-01 4.81759906e-01 -2.57523626e-01 -2.92870224e-01
2.61700362e-01 6.24151051e-01 3.83109599e-01 -6.26676142e-01
9.78628874e-01 2.70550489e-01 -2.70326644e-01 -8.45113993e-01
-1.15044117e+00 -4.84321654e-01 -6.38768017e-01 -1.17616788e-01
4.09798205e-01 -9.31862652e-01 -5.14918387e-01 6.52174413e-01
-7.30203152e-01 -4.67172325e-01 -8.18069041e-01 4.50351208e-01
-4.55823332e-01 3.12933177e-01 -1.69006929e-01 -9.50867832e-01
-1.52679279e-01 -7.44596601e-01 6.59665942e-01 2.36635640e-01
6.44781440e-02 -1.05014956e+00 9.49532986e-02 -4.52245809e-02
2.99496174e-01 1.07802294e-01 9.33766901e-01 -1.14669776e+00
-4.00870353e-01 -3.07462394e-01 -4.76266518e-02 5.54750741e-01
1.52699396e-01 -5.36878705e-01 -7.27992535e-01 -3.54566187e-01
2.11301073e-01 -6.03507996e-01 8.98548305e-01 3.11815500e-01
8.29239488e-01 -5.68868577e-01 -2.26236805e-01 2.41670564e-01
1.29448891e+00 4.83204573e-01 4.14944917e-01 5.14352024e-01
5.39896727e-01 2.99336106e-01 6.87548637e-01 3.20573628e-01
2.34262735e-01 6.32529140e-01 2.83621013e-01 9.37567055e-02
-6.36948943e-02 -3.61311466e-01 5.66577077e-01 7.12008893e-01
-1.86564494e-02 -5.02516478e-02 -9.23869550e-01 6.29342318e-01
-2.07598209e+00 -1.10273206e+00 1.27542555e-01 2.34490728e+00
1.03386748e+00 4.48210090e-01 3.98991197e-01 1.22199990e-01
7.48083591e-01 -9.20238271e-02 -1.01730478e+00 -2.82946229e-01
-3.03525925e-01 -5.85270822e-02 -1.82049535e-03 2.81698048e-01
-1.15174139e+00 9.34170604e-01 6.10971785e+00 1.17438614e+00
-9.26820397e-01 2.36119524e-01 5.28432131e-01 2.04063386e-01
-2.09441900e-01 -1.35607734e-01 -9.15397525e-01 3.78246099e-01
3.08315456e-01 -4.53991175e-01 3.15824181e-01 1.40062177e+00
-4.56904203e-01 -2.52123952e-01 -1.44481182e+00 1.00865459e+00
2.13546604e-01 -1.11034107e+00 1.48252338e-01 -6.95860237e-02
1.25337505e+00 -1.60311550e-01 1.24302685e-01 9.84048784e-01
2.62292266e-01 -6.64500475e-01 6.55234575e-01 4.80431348e-01
5.54307282e-01 -6.12235308e-01 7.48804986e-01 9.36564088e-01
-1.17894173e+00 -3.96428883e-01 -6.04556620e-01 -3.69373679e-01
-2.73990005e-01 8.45687151e-01 -6.40700221e-01 5.68250179e-01
6.35851741e-01 9.49153483e-01 -5.27178109e-01 1.16711354e+00
-4.89494771e-01 4.04099554e-01 -3.87678534e-01 -1.02079831e-01
1.65091410e-01 1.71653584e-01 8.63234401e-01 6.62905395e-01
7.60713071e-02 -2.64476061e-01 5.30981779e-01 8.25262547e-01
-2.37661377e-01 4.94603813e-02 -5.43534398e-01 1.07860669e-01
6.87482774e-01 8.29901457e-01 -5.55496097e-01 -5.78378439e-01
-4.12250161e-01 8.17817330e-01 3.53524476e-01 2.05879882e-01
-6.32059872e-01 -3.53053898e-01 4.76216614e-01 -1.10991642e-01
3.54194909e-01 2.83439010e-01 -2.49181278e-02 -1.43320251e+00
3.25164795e-01 -9.30921793e-01 5.65428555e-01 -3.92352581e-01
-1.96981204e+00 3.52905124e-01 -3.04548591e-02 -1.33405125e+00
-1.14255687e-02 -4.89450336e-01 -6.16785347e-01 2.68829316e-01
-1.66509151e+00 -8.95637155e-01 -7.38728195e-02 7.20165730e-01
8.67315114e-01 -3.47659320e-01 6.86191916e-01 7.73520023e-02
-4.46360201e-01 7.00977504e-01 2.14605093e-01 -2.12248370e-01
5.66045582e-01 -1.43211770e+00 -1.12119829e-02 8.40628207e-01
4.82491523e-01 4.38461155e-01 3.62676531e-01 -8.69006813e-01
-1.21000159e+00 -1.14757693e+00 8.66130173e-01 -4.45593804e-01
6.94628954e-01 -3.36542875e-01 -9.33190703e-01 8.47021222e-01
-4.72671747e-01 1.94745976e-02 3.51641148e-01 4.22221661e-01
-5.10091186e-01 -2.23966569e-01 -9.33082521e-01 3.99158388e-01
1.08459306e+00 -3.40330482e-01 -1.24801493e+00 6.72237158e-01
1.03062248e+00 -2.44547352e-01 -5.11940479e-01 5.54429412e-01
5.31107545e-01 -8.42331052e-01 9.10959303e-01 -7.70460248e-01
9.15355533e-02 -2.37562895e-01 -1.19078226e-01 -1.10503292e+00
-4.28998470e-01 -4.98633593e-01 -6.12425923e-01 1.00977170e+00
4.62594360e-01 -9.16285396e-01 4.02525425e-01 2.44951278e-01
-2.21690163e-01 -8.27302575e-01 -1.02215314e+00 -1.64396298e+00
1.18769981e-01 -7.60006309e-01 3.90082240e-01 1.30389965e+00
7.51008168e-02 3.40349674e-01 -5.65368056e-01 1.20618097e-01
3.58681947e-01 3.18256199e-01 8.73305559e-01 -1.60235167e+00
-4.00171399e-01 -5.67551911e-01 -5.30145407e-01 -1.25071728e+00
1.98181439e-02 -1.00789428e+00 5.27153127e-02 -1.32771957e+00
3.58750433e-01 -6.26517057e-01 -6.45408988e-01 7.36327291e-01
-3.45456272e-01 -2.82714888e-02 8.54016766e-02 7.23711014e-01
-1.09385967e+00 5.49697697e-01 1.05770087e+00 -8.88535455e-02
-6.34583235e-01 3.09990615e-01 -6.51574910e-01 9.39748347e-01
4.52705801e-01 -4.30429012e-01 -2.96423465e-01 -1.01027943e-01
5.67820191e-01 -2.38793522e-01 4.05658692e-01 -1.07701111e+00
2.75525957e-01 -1.99469075e-01 2.89559960e-01 -4.09799427e-01
-5.70998341e-02 -7.70589948e-01 -2.89417624e-01 8.20410192e-01
-4.87073660e-01 -2.09721819e-01 5.05009294e-02 8.78592074e-01
-6.87188283e-02 -2.97744572e-01 5.20133078e-01 -1.45343989e-01
-1.23179340e+00 4.53206122e-01 -2.79898942e-01 5.30696869e-01
1.40618932e+00 -3.42341453e-01 -3.09802055e-01 -2.61722594e-01
-7.32059419e-01 3.88684362e-01 2.38282979e-01 6.25460446e-01
6.54488981e-01 -1.37509096e+00 -6.20104253e-01 2.70650238e-01
4.24945235e-01 4.21791747e-02 1.78582266e-01 9.61092889e-01
1.05329165e-02 2.53547519e-01 1.30417824e-01 -6.51994884e-01
-7.55730391e-01 9.97950852e-01 3.03839091e-02 -5.07103205e-01
-7.42273629e-01 1.05939913e+00 4.67830062e-01 -5.30377209e-01
6.31565750e-01 -1.85954794e-01 -3.73068713e-02 1.73624739e-01
6.73763096e-01 3.00561398e-01 -1.09148465e-01 -1.23105794e-01
-1.81561455e-01 3.96020144e-01 -3.81581277e-01 3.20141256e-01
1.43616855e+00 -4.77736145e-02 5.99105395e-02 9.46852148e-01
9.02045012e-01 -1.83805391e-01 -1.15027869e+00 -7.62383461e-01
1.99884057e-01 -4.40004438e-01 -1.39696151e-01 -9.73760843e-01
-6.28898323e-01 5.63151419e-01 7.19955385e-01 5.60664713e-01
1.11254907e+00 3.18835109e-01 7.84401357e-01 8.24224472e-01
5.91309607e-01 -1.18126523e+00 6.38561964e-01 6.49227858e-01
9.63864625e-01 -1.37459111e+00 -1.84341714e-01 -2.28984132e-02
-6.33181870e-01 9.33956921e-01 7.19260454e-01 -1.08758762e-01
6.16542816e-01 9.09690410e-02 -4.24768955e-01 -2.11321533e-01
-9.13956523e-01 -3.73980045e-01 5.41653275e-01 6.71098113e-01
-2.08752856e-01 -4.56093960e-02 -1.76824510e-01 7.29227424e-01
-8.66736174e-02 9.96318366e-03 7.99966678e-02 9.00100172e-01
-9.42032456e-01 -7.46586502e-01 -1.29714534e-01 6.74006462e-01
6.18163757e-02 -6.63826093e-02 -3.32049787e-01 8.46467376e-01
5.17035902e-01 6.32519782e-01 -4.34337109e-02 -4.35696602e-01
2.51592398e-01 1.52259544e-01 2.87273914e-01 -8.74862552e-01
-1.93010390e-01 -2.01123655e-01 -3.44313681e-01 -2.64729261e-01
-2.91437179e-01 -4.30216670e-01 -9.94233251e-01 -4.22709994e-02
-7.05930054e-01 7.97555149e-02 2.34471917e-01 1.19710970e+00
2.31381223e-01 2.84902006e-01 7.58796096e-01 -5.13378203e-01
-7.48689055e-01 -8.98524761e-01 -8.41659904e-01 3.71721089e-01
2.73164034e-01 -8.88081968e-01 -7.83138812e-01 -3.37832496e-02] | [9.855354309082031, 3.3109638690948486] |
5a26b0ba-61f7-4f7b-bc24-d449bcc1cd75 | lemmatization-and-morphological-tagging-in | null | null | https://aclanthology.org/L16-1239 | https://aclanthology.org/L16-1239.pdf | Lemmatization and Morphological Tagging in German and Latin: A Comparison and a Survey of the State-of-the-art | This paper relates to the challenge of morphological tagging and lemmatization in morphologically rich languages by example of German and Latin. We focus on the question what a practitioner can expect when using state-of-the-art solutions out of the box. Moreover, we contrast these with old(er) methods and implementations for POS tagging. We examine to what degree recent efforts in tagger development are reflected by improved accuracies ― and at what cost, in terms of training and processing time. We also conduct in-domain vs. out-domain evaluation. Out-domain evaluations are particularly insightful because the distribution of the data which is being tagged by a user will typically differ from the distribution on which the tagger has been trained. Furthermore, two lemmatization techniques are evaluated. Finally, we compare pipeline tagging vs. a tagging approach that acknowledges dependencies between inflectional categories. | ['er', 'R{\\"u}diger Gleim', 'Alex Mehler', 'Steffen Eger'] | 2016-05-01 | lemmatization-and-morphological-tagging-in-1 | https://aclanthology.org/L16-1239 | https://aclanthology.org/L16-1239.pdf | lrec-2016-5 | ['morphological-tagging'] | ['natural-language-processing'] | [-9.47610661e-02 3.08079690e-01 -3.40799652e-02 -4.47244942e-01
-9.74173248e-01 -1.19365036e+00 5.21160066e-01 6.73623800e-01
-8.92477453e-01 6.20349288e-01 5.03535926e-01 -5.79020560e-01
9.35207084e-02 -7.22998559e-01 -2.52911747e-01 -3.84267181e-01
1.38590902e-01 8.51870775e-01 6.04496859e-02 -2.12548465e-01
1.36990622e-01 5.14771819e-01 -8.92965019e-01 3.79021585e-01
7.43542314e-01 2.77451456e-01 2.12348074e-01 3.11926663e-01
-5.22281349e-01 3.88024807e-01 -8.09943736e-01 -1.06457353e+00
5.38411178e-02 -1.28566563e-01 -1.25221801e+00 -3.82771827e-02
4.28126931e-01 4.08416480e-01 2.30279475e-01 1.23600698e+00
4.52119201e-01 -8.56281910e-03 6.85855329e-01 -3.09723258e-01
-7.12166727e-01 1.10438442e+00 -2.31595099e-01 4.89497721e-01
3.76157522e-01 -5.24405427e-02 1.01186466e+00 -6.28063202e-01
9.86054838e-01 1.17552602e+00 8.23925495e-01 6.13597572e-01
-1.19860446e+00 -4.45472062e-01 3.20953280e-01 -2.87881136e-01
-1.08185232e+00 -4.05544996e-01 4.06047106e-01 -6.36326432e-01
1.04229236e+00 -1.56350344e-01 3.06609452e-01 7.27954268e-01
4.78954911e-02 6.12915099e-01 1.32287824e+00 -6.80083334e-01
7.55631849e-02 3.68814945e-01 1.47110879e-01 3.72517049e-01
4.40927804e-01 -7.25049376e-02 -1.99709535e-01 4.13866760e-03
5.53258955e-01 -7.43093371e-01 1.40916929e-01 4.20341454e-02
-1.12928450e+00 7.40644395e-01 -8.07129070e-02 9.20062900e-01
-2.17038274e-01 -3.07945818e-01 6.16668880e-01 2.37643868e-01
6.07608795e-01 6.96390748e-01 -1.02296209e+00 -2.30518833e-01
-1.18365455e+00 -6.18809089e-02 8.91094565e-01 9.63280439e-01
5.06542027e-01 1.84331343e-01 1.96075439e-01 9.09107983e-01
1.96599856e-01 3.53217781e-01 4.69749719e-01 -6.43665195e-01
6.08912408e-01 4.25203234e-01 8.83329138e-02 -1.88207328e-01
-3.95253241e-01 -2.33385295e-01 -1.50849119e-01 7.03667328e-02
7.51132250e-01 -4.47598934e-01 -1.03343642e+00 1.74640846e+00
2.22460687e-01 -4.64366257e-01 2.32864872e-01 3.86988312e-01
6.33766830e-01 3.71717691e-01 8.04258168e-01 -1.36660114e-01
1.83748233e+00 -5.47610462e-01 -6.65009856e-01 -6.18547499e-01
1.00916636e+00 -1.19228661e+00 8.59090924e-01 2.58096695e-01
-1.34205258e+00 -5.25762856e-01 -5.80655217e-01 -1.51478097e-01
-7.31204629e-01 1.99971452e-01 8.01948488e-01 8.89833808e-01
-1.06405497e+00 7.86474347e-01 -9.21394110e-01 -7.99113631e-01
9.16119516e-02 4.31363434e-01 -5.84054828e-01 2.13708982e-01
-1.03340042e+00 1.20315814e+00 7.73958564e-01 -2.97077715e-01
-2.52040982e-01 -5.78153431e-01 -9.70979512e-01 -1.29183337e-01
-1.22833252e-01 -2.39832476e-01 1.56433582e+00 -9.12367821e-01
-1.14727724e+00 1.51111162e+00 -7.66876340e-02 -1.61310181e-01
7.76816458e-02 -1.87229693e-01 -7.67337084e-01 -5.83537370e-02
3.74415547e-01 6.03025079e-01 1.45598069e-01 -1.08354449e+00
-1.14863133e+00 -2.12494820e-01 2.01010704e-02 6.53555691e-02
-9.50426236e-03 9.55260396e-01 -1.77624114e-02 -7.86288619e-01
-2.39722468e-02 -6.79947376e-01 -1.34485304e-01 -8.20916831e-01
-6.22638082e-03 -5.27322531e-01 3.50406051e-01 -1.12062621e+00
1.28402698e+00 -2.22140932e+00 -2.80795485e-01 -6.13612048e-02
-2.62083322e-01 5.60853660e-01 8.73081852e-03 6.57249868e-01
-3.60818833e-01 6.04202688e-01 -6.69106022e-02 -3.22401315e-01
9.15355459e-02 4.62640762e-01 -2.18327761e-01 3.86707038e-01
5.38096666e-01 6.81680083e-01 -1.12614727e+00 -7.18846202e-01
-1.19612418e-01 2.48352647e-01 -3.51531118e-01 1.10546470e-01
1.34708107e-01 3.99615258e-01 -1.09258875e-01 8.32071662e-01
5.61818242e-01 4.30341393e-01 6.95706666e-01 9.44154933e-02
-7.60622025e-01 1.09745061e+00 -9.46693361e-01 1.66174710e+00
-5.57371974e-01 2.99945474e-01 1.30774617e-01 -7.39214540e-01
8.93570364e-01 6.93788469e-01 1.11800358e-01 -1.96117461e-01
2.64334381e-01 5.90016246e-01 3.20023000e-01 -2.15886772e-01
6.73678994e-01 -5.45995951e-01 -4.95567530e-01 2.09089160e-01
4.16717559e-01 -6.63756207e-02 5.47427118e-01 -2.10108221e-01
1.06838322e+00 5.03620327e-01 6.71326816e-01 -7.23858356e-01
1.21633522e-01 5.49844086e-01 8.14695299e-01 4.38219547e-01
-2.11907819e-01 2.85238653e-01 5.13430238e-01 -1.41979232e-01
-1.06153977e+00 -1.14797199e+00 -4.02700394e-01 1.15985799e+00
-5.00193775e-01 -1.81127727e-01 -8.85462880e-01 -1.02934170e+00
-3.71318489e-01 9.17956829e-01 -4.76292014e-01 3.51424873e-01
-1.00420952e+00 -6.45890594e-01 8.35671663e-01 6.66909456e-01
5.54964645e-03 -1.61197782e+00 -3.50869566e-01 5.18475115e-01
-1.46295913e-02 -1.19163728e+00 -2.55969286e-01 5.97483337e-01
-1.02111447e+00 -6.93459392e-01 -3.78243387e-01 -1.49444377e+00
5.42149425e-01 -3.20026726e-01 1.56127131e+00 -3.08309346e-01
3.67646396e-01 1.10195495e-01 -7.36932099e-01 -4.92110699e-01
-6.62189782e-01 4.45522040e-01 -8.35396647e-02 -6.58471167e-01
7.99743831e-01 -4.97752488e-01 -1.95459738e-01 -8.19557533e-02
-7.90966213e-01 -5.13941646e-01 8.09030831e-01 4.15024221e-01
4.31220055e-01 -7.83390477e-02 2.79711723e-01 -1.50162172e+00
3.83067936e-01 -5.01084745e-01 -3.31610382e-01 3.31534058e-01
-4.61643100e-01 -1.20102633e-02 5.33953071e-01 -4.63111609e-01
-1.23726797e+00 1.02705985e-01 -6.04171336e-01 2.79873848e-01
-7.37816393e-01 5.30359328e-01 -3.58698010e-01 2.50693917e-01
4.72687155e-01 -5.10264397e-01 -4.56720382e-01 -9.12857890e-01
3.80784065e-01 6.55482173e-01 7.08665907e-01 -8.93296421e-01
7.60022283e-01 9.99944732e-02 -4.30732995e-01 -5.71209490e-01
-7.93138385e-01 -7.86127269e-01 -1.10533130e+00 3.20547640e-01
1.04844296e+00 -8.29626977e-01 -4.80034128e-02 3.58286619e-01
-1.24980044e+00 -4.81609553e-01 -5.98942637e-01 4.05358225e-01
-2.72833943e-01 3.08511406e-01 -8.94195914e-01 -6.78177118e-01
-2.72764117e-01 -7.32908428e-01 8.28472078e-01 2.48096615e-01
-5.24911463e-01 -1.53756905e+00 2.49234989e-01 5.70230111e-02
-4.89415638e-02 1.10183477e-01 1.18317747e+00 -1.18286550e+00
2.41605654e-01 -1.00009143e-01 -5.83927557e-02 3.72082174e-01
8.09728950e-02 -8.55870619e-02 -8.43712270e-01 -2.35968679e-01
-1.57191515e-01 -1.81796728e-03 3.36745620e-01 1.78697288e-01
2.25720599e-01 -1.54063225e-01 -1.65194660e-01 3.74844640e-01
1.43316483e+00 2.83104599e-01 5.59486806e-01 5.34788489e-01
4.76296276e-01 1.07070231e+00 9.42143142e-01 -2.10936055e-01
4.35403764e-01 3.88652563e-01 -2.48819858e-01 -1.06642179e-01
-2.60339588e-01 -4.31279242e-01 6.35006487e-01 1.18306077e+00
1.97713837e-01 -2.15952948e-01 -1.29081857e+00 1.09095776e+00
-1.51734638e+00 -6.54981017e-01 -2.66279459e-01 2.01306701e+00
1.16669238e+00 2.72236973e-01 1.99583992e-01 1.59381062e-01
8.41331184e-01 -9.87362191e-02 2.80239433e-01 -9.88863170e-01
-7.25483894e-02 9.42953169e-01 6.39029920e-01 6.29437923e-01
-1.28036153e+00 1.58050501e+00 6.66774273e+00 8.63714159e-01
-1.07968223e+00 3.50693256e-01 3.73025507e-01 5.96959233e-01
-1.44176677e-01 4.17245001e-01 -1.14458013e+00 2.71539062e-01
1.21695721e+00 2.28007346e-01 -8.93457085e-02 7.63122916e-01
6.28299359e-03 -1.04666024e-01 -9.86693501e-01 3.87714148e-01
-2.33329639e-01 -6.26358092e-01 -2.39274666e-01 2.08824024e-01
5.12985468e-01 7.92330578e-02 -1.52786851e-01 4.89321947e-01
8.66316140e-01 -6.72881067e-01 1.03036189e+00 1.03320837e-01
9.52293336e-01 -6.70813262e-01 1.07317579e+00 1.21090084e-01
-1.33370590e+00 3.18840235e-01 -4.71478939e-01 -1.13643259e-01
4.45294261e-01 5.58733046e-01 -9.02760804e-01 5.09025276e-01
4.79160160e-01 2.29852796e-01 -6.68764949e-01 1.07713115e+00
-9.05304313e-01 1.09184408e+00 -2.28707045e-01 9.66432616e-02
4.26708639e-01 -2.97717839e-01 3.50123227e-01 2.07513189e+00
2.66866863e-01 7.24853277e-02 2.32350886e-01 2.64121413e-01
1.65505618e-01 4.85286981e-01 -4.61842507e-01 -2.95652658e-01
5.79808414e-01 1.48239124e+00 -1.09958696e+00 -3.81834120e-01
-4.44357753e-01 7.45547891e-01 5.54999471e-01 9.24260169e-03
-3.24403584e-01 -4.76537406e-01 7.21804440e-01 3.43923777e-01
5.35424471e-01 -3.19708109e-01 -4.25546408e-01 -9.04547751e-01
-2.55813718e-01 -7.15194881e-01 7.27725983e-01 -4.43533510e-01
-1.40007031e+00 7.17364669e-01 1.03314199e-01 -7.50924647e-01
-2.93695629e-01 -9.53943014e-01 -4.88469064e-01 1.02416945e+00
-1.49247229e+00 -1.29692817e+00 4.98418510e-01 -5.27292304e-02
2.65302807e-01 1.08461887e-01 1.09847796e+00 4.97852296e-01
-3.80991757e-01 5.67617178e-01 -1.26274854e-01 7.91886687e-01
8.25056493e-01 -1.73134458e+00 8.82274330e-01 1.07718658e+00
4.36811298e-01 7.01999962e-01 6.05556190e-01 -7.56121516e-01
-4.34156865e-01 -1.01644158e+00 1.89293861e+00 -7.25982904e-01
1.00061929e+00 -3.51300508e-01 -7.97848284e-01 8.42063427e-01
3.66431206e-01 -2.33805552e-01 8.98899913e-01 5.16455472e-01
-2.57435113e-01 3.06508929e-01 -1.31020963e+00 3.61882776e-01
1.07020056e+00 -5.16446710e-01 -1.13295090e+00 7.91641399e-02
3.44101638e-01 -1.87365681e-01 -1.18310869e+00 2.25699380e-01
2.24016756e-01 -5.66735923e-01 5.21118402e-01 -8.94095182e-01
3.03531051e-01 -1.25899643e-01 6.25018403e-02 -1.50256276e+00
-6.02999151e-01 -4.99089807e-01 7.17558801e-01 2.08215618e+00
7.45279968e-01 -4.50326174e-01 6.67073190e-01 3.05212736e-01
-4.86495227e-01 -3.93585503e-01 -7.67181933e-01 -8.34097803e-01
7.89028883e-01 -3.34328115e-01 4.77972329e-01 1.18507445e+00
3.11050832e-01 3.52513343e-01 3.03401232e-01 1.19525909e-01
2.67011136e-01 -1.53711975e-01 1.97684795e-01 -1.23657215e+00
-3.35682392e-01 -4.07671213e-01 -4.55777436e-01 -6.06119215e-01
2.51441151e-01 -1.08341837e+00 1.64689958e-01 -1.57491529e+00
-1.60020500e-01 -6.85388863e-01 -2.01719224e-01 7.29305267e-01
-2.58016437e-01 3.11916649e-01 1.50811344e-01 1.07250988e-01
-3.15835655e-01 -3.63223433e-01 6.83356524e-01 4.35223281e-01
-1.84929773e-01 -2.02242196e-01 -8.72426510e-01 9.53205884e-01
1.11139739e+00 -8.66763115e-01 2.62904406e-01 -7.71602631e-01
2.46526852e-01 -2.61013418e-01 -2.68685728e-01 -9.10441697e-01
-1.88912436e-01 -1.34493321e-01 1.16997927e-01 -2.46267617e-01
-1.61860734e-01 -5.52845120e-01 7.45820329e-02 1.67884141e-01
6.03412353e-02 5.28994560e-01 3.39723825e-01 -1.77603066e-01
-2.17954472e-01 -9.01535451e-01 1.05068552e+00 -5.59056878e-01
-6.71743214e-01 8.05617720e-02 -6.07183099e-01 5.86888194e-01
5.24613917e-01 -2.28929251e-01 -1.00920327e-01 2.96538025e-01
-8.85534406e-01 -2.36998171e-01 6.41364157e-01 3.20527181e-02
-2.40949556e-01 -1.02003968e+00 -6.64163291e-01 -2.42420092e-01
2.80252949e-04 -2.45806843e-01 -3.75600606e-01 6.36783600e-01
-6.14347398e-01 2.63210505e-01 -1.72565714e-01 1.34551629e-01
-1.26820016e+00 4.97533947e-01 7.52562955e-02 -7.96625197e-01
-2.87268341e-01 9.70851600e-01 -2.91694328e-02 -6.51349843e-01
-1.38374135e-01 -2.26902261e-01 -4.23746258e-01 4.16139722e-01
6.57854378e-02 2.01829642e-01 3.13011289e-01 -8.85544419e-01
-5.60873389e-01 3.28924119e-01 -1.01076268e-01 -4.88157064e-01
1.30406857e+00 2.07790911e-01 -1.72735810e-01 4.68626738e-01
7.18769491e-01 7.89882779e-01 -7.62517571e-01 -7.53996670e-02
7.10051477e-01 -3.52552757e-02 -1.96903944e-01 -1.00989032e+00
-6.82243347e-01 9.46641326e-01 3.04082274e-01 3.15637380e-01
7.59753227e-01 2.56552219e-01 7.94164360e-01 4.06194665e-03
2.57684648e-01 -1.43627787e+00 -6.23381674e-01 7.93074608e-01
2.32587516e-01 -7.68340349e-01 -6.02920353e-02 -7.39959717e-01
-6.70261383e-01 8.80918443e-01 2.05378950e-01 -2.95418739e-01
6.97965086e-01 6.19869769e-01 5.80676258e-01 -2.24443987e-01
-2.60970145e-01 -7.74294138e-01 4.24837656e-02 9.79455471e-01
1.12846947e+00 2.65995950e-01 -8.72291625e-01 6.01054728e-01
-8.29093039e-01 -4.39118713e-01 2.03139260e-01 1.01805103e+00
-4.36501145e-01 -1.83445561e+00 -3.63335401e-01 1.21753275e-01
-1.22510862e+00 -5.38654089e-01 -6.28874898e-01 1.14793003e+00
6.70913637e-01 1.00450957e+00 1.13101035e-01 -1.35010540e-01
4.74291265e-01 5.48601747e-01 5.29457331e-01 -1.37482822e+00
-1.20588171e+00 1.24435320e-01 7.25014210e-01 1.84507489e-01
-4.15529341e-01 -1.36051774e+00 -1.27616966e+00 -1.22562476e-01
-3.76317948e-01 6.48608863e-01 6.01974726e-01 1.02600980e+00
-1.28715113e-01 2.10405052e-01 6.11835420e-02 -4.82194602e-01
-4.77276176e-01 -1.21135080e+00 -7.16434300e-01 4.16759521e-01
-4.42802072e-01 -3.79585654e-01 -2.84824044e-01 2.01972976e-01] | [10.38170337677002, 10.02133560180664] |
8acecbdd-cb69-4d21-99dc-0f93d5f25f59 | global-relation-modeling-and-refinement-for | 2303.14888 | null | https://arxiv.org/abs/2303.14888v1 | https://arxiv.org/pdf/2303.14888v1.pdf | Global Relation Modeling and Refinement for Bottom-Up Human Pose Estimation | In this paper, we concern on the bottom-up paradigm in multi-person pose estimation (MPPE). Most previous bottom-up methods try to consider the relation of instances to identify different body parts during the post processing, while ignoring to model the relation among instances or environment in the feature learning process. In addition, most existing works adopt the operations of upsampling and downsampling. During the sampling process, there will be a problem of misalignment with the source features, resulting in deviations in the keypoint features learned by the model. To overcome the above limitations, we propose a convolutional neural network for bottom-up human pose estimation. It invovles two basic modules: (i) Global Relation Modeling (GRM) module globally learns relation (e.g., environment context, instance interactive information) among region of image by fusing multiple stages features in the feature learning process. It combines with the spatial-channel attention mechanism, which focuses on achieving adaptability in spatial and channel dimensions. (ii) Multi-branch Feature Align (MFA) module aggregates features from multiple branches to align fused feature and obtain refined local keypoint representation. Our model has the ability to focus on different granularity from local to global regions, which significantly boosts the performance of the multi-person pose estimation. Our results on the COCO and CrowdPose datasets demonstrate that it is an efficient framework for multi-person pose estimation. | ['Jianqin Yin', 'Ruoqi Yin'] | 2023-03-27 | null | null | null | null | ['multi-person-pose-estimation'] | ['computer-vision'] | [-6.85119182e-02 -3.35190415e-01 9.64428410e-02 -3.55659306e-01
-6.22489393e-01 -1.71142310e-01 3.91557425e-01 4.36453484e-02
-5.18182755e-01 3.63876075e-01 5.03877640e-01 6.35925412e-01
-6.82345554e-02 -8.09953570e-01 -7.76771903e-01 -4.14750993e-01
1.66793689e-02 5.39760709e-01 3.33479613e-01 -3.75390738e-01
-1.15642678e-02 4.51114088e-01 -1.77438092e+00 3.75123680e-01
6.30692542e-01 1.00248861e+00 1.22444443e-01 6.12064719e-01
9.76463482e-02 1.99942216e-01 -7.10063756e-01 -3.82341981e-01
2.19570979e-01 -1.15038320e-01 -7.13612497e-01 1.08432613e-01
5.67772985e-01 -3.80629569e-01 -2.69817501e-01 8.50932062e-01
9.12348330e-01 8.89408067e-02 4.94467378e-01 -1.22633743e+00
-1.58392504e-01 3.22050124e-01 -7.99020767e-01 9.91379991e-02
7.01499939e-01 1.18198380e-01 6.96304142e-01 -1.03050566e+00
4.65080768e-01 1.69012260e+00 8.46892595e-01 3.27358335e-01
-7.72834241e-01 -6.67140841e-01 5.50353646e-01 3.45470399e-01
-1.56881070e+00 -1.62275687e-01 8.16047668e-01 -4.57587808e-01
6.74739242e-01 4.12366241e-01 1.11203253e+00 1.01876950e+00
3.79195303e-01 9.18738186e-01 9.49249029e-01 -2.92635560e-01
-3.22109938e-01 -7.46093914e-02 8.85218009e-02 5.80542266e-01
2.81808674e-01 -5.08596227e-02 -8.00103009e-01 -8.23154151e-02
9.75585520e-01 1.62418917e-01 -1.24142036e-01 -3.43790978e-01
-1.19258320e+00 5.04010260e-01 7.47379124e-01 2.48988450e-01
-3.54244798e-01 1.52107790e-01 3.48080724e-01 -7.44114891e-02
2.24957988e-01 1.39398068e-01 -5.92723072e-01 8.44580159e-02
-8.00880909e-01 6.45053685e-01 4.37731475e-01 1.05353463e+00
8.38854730e-01 -4.52996761e-01 -4.82742459e-01 6.24196887e-01
5.45451403e-01 4.85605478e-01 5.65853536e-01 -4.34955746e-01
8.05317223e-01 7.95062244e-01 2.08780691e-01 -1.29355299e+00
-9.31816697e-01 -5.73241472e-01 -6.37452960e-01 -1.27076596e-01
3.57107431e-01 -3.00873071e-01 -8.30581605e-01 1.67815125e+00
7.13660061e-01 -1.29348999e-02 -4.60903198e-01 1.14576447e+00
1.03346777e+00 2.07758352e-01 2.86991537e-01 7.23003596e-03
1.73151863e+00 -1.09525275e+00 -6.26461446e-01 -3.51036370e-01
3.60847354e-01 -5.68573713e-01 7.11120129e-01 2.81111807e-01
-9.24336612e-01 -1.22052348e+00 -1.01480651e+00 -2.01007575e-01
-4.45376933e-01 3.17075670e-01 6.72016740e-01 4.49032575e-01
-6.88719332e-01 4.87141371e-01 -6.00765347e-01 -4.49046761e-01
2.15161145e-01 6.50717616e-01 -6.10335946e-01 1.51661530e-01
-1.32974422e+00 7.49306202e-01 4.11458999e-01 5.20070016e-01
-5.27036548e-01 -4.14033949e-01 -9.02842879e-01 -3.73917557e-02
3.58299762e-01 -1.00484264e+00 8.38697076e-01 -9.56704617e-01
-1.29515100e+00 4.68169421e-01 -1.69579968e-01 1.00989090e-02
7.53716528e-01 -8.39573145e-01 -4.31038737e-01 -9.17812437e-02
2.17869565e-01 7.36055076e-01 9.33863819e-01 -1.13306546e+00
-9.41442192e-01 -7.55873382e-01 -8.83363783e-02 7.13942349e-01
-2.16467425e-01 5.79102933e-02 -9.92307842e-01 -5.73994696e-01
2.80564368e-01 -7.99189270e-01 -2.21553773e-01 -8.40604007e-02
-4.68909234e-01 -1.81436628e-01 5.71006238e-01 -7.73560226e-01
1.36293828e+00 -1.94011116e+00 4.13733810e-01 3.62055123e-01
1.82153329e-01 2.55600992e-03 9.89524499e-02 3.15337837e-01
-1.79862771e-02 -2.87170202e-01 2.46191770e-01 -4.22399104e-01
-9.65218470e-02 -5.62863238e-02 2.16474608e-01 5.49499631e-01
4.89455871e-02 1.06223071e+00 -5.97743630e-01 -9.17475581e-01
4.28739130e-01 6.57291591e-01 -6.74085975e-01 1.81034103e-01
2.16902554e-01 6.73289657e-01 -5.88923633e-01 6.61263049e-01
7.80132353e-01 -1.19446561e-01 4.54914384e-02 -6.42492056e-01
-2.19636299e-02 -2.39520594e-01 -1.72433913e+00 1.95337939e+00
-2.98719943e-01 -4.04497469e-03 -1.11342296e-01 -5.11796415e-01
7.14016140e-01 2.27474093e-01 6.44081652e-01 -2.86391973e-01
3.85064125e-01 -4.24954817e-02 -1.37664214e-01 -5.42456686e-01
6.33396566e-01 3.36883903e-01 -3.76977414e-01 -8.89311880e-02
3.08204830e-01 4.67901349e-01 -6.98956773e-02 -1.74496442e-01
5.22821724e-01 5.52684844e-01 5.24456322e-01 -8.39298144e-02
6.84279084e-01 -2.85836965e-01 5.86049139e-01 5.36960006e-01
-3.41555983e-01 5.90584159e-01 7.44128413e-03 -7.37803757e-01
-6.70835197e-01 -7.80804515e-01 6.71588164e-03 1.31915808e+00
5.14984250e-01 -6.39094949e-01 -8.20241690e-01 -6.83825314e-01
8.45236704e-02 -2.10799664e-01 -8.11439395e-01 -7.10255001e-03
-7.91268051e-01 -8.09054911e-01 3.36169809e-01 8.30230832e-01
9.24318552e-01 -9.84856784e-01 -7.03855872e-01 2.83524573e-01
-4.17744428e-01 -9.33382273e-01 -6.58282757e-01 -1.21518113e-01
-5.40834546e-01 -1.03944921e+00 -6.38041437e-01 -6.13590240e-01
6.17709816e-01 -1.24271741e-04 8.48414004e-01 3.33131803e-03
-2.47159421e-01 4.32507962e-01 -3.91831249e-01 -3.81454229e-01
4.43275779e-01 2.46987849e-01 1.99716657e-01 3.28113317e-01
5.41092932e-01 -4.03809398e-01 -8.07236135e-01 3.78454238e-01
-3.00454319e-01 1.12378001e-01 7.34934390e-01 6.99182212e-01
6.62020028e-01 -8.29723999e-02 2.54233301e-01 -4.72466439e-01
4.22123790e-01 -2.16681093e-01 -1.31191447e-01 3.06195676e-01
-1.12735361e-01 -2.01701969e-01 2.74264067e-01 -4.73529249e-01
-9.08000231e-01 5.07430553e-01 -2.70270765e-01 -2.12409541e-01
-4.45110172e-01 1.99421465e-01 -6.96421146e-01 -1.54029906e-01
5.09525180e-01 2.20035985e-01 -2.35120341e-01 -5.68846822e-01
2.58994251e-01 6.40441775e-01 4.59940851e-01 -6.37404084e-01
7.82870889e-01 4.28089947e-01 -3.70089225e-02 -6.34616256e-01
-7.43995488e-01 -4.81616139e-01 -1.13325012e+00 -6.21883333e-01
1.07000327e+00 -1.20878685e+00 -8.39291871e-01 6.36158586e-01
-1.03617001e+00 3.13532114e-01 -2.27225870e-02 5.50968409e-01
-4.74628836e-01 2.98307389e-01 -5.01132548e-01 -7.16622651e-01
-4.01552469e-01 -1.12430048e+00 1.46407068e+00 5.57987452e-01
-3.36220860e-01 -5.67749977e-01 -1.23747759e-01 3.57923418e-01
8.88667777e-02 3.47517550e-01 1.44075289e-01 -5.83409727e-01
-4.23417896e-01 -3.55458409e-01 -5.43450601e-02 -2.41077274e-01
-2.03138143e-02 -4.07392323e-01 -1.00397265e+00 -4.49699789e-01
-2.71448821e-01 -1.02004617e-01 6.72050416e-01 5.06797552e-01
1.04965413e+00 -1.38240844e-01 -6.29514575e-01 8.23614180e-01
1.20893383e+00 -2.21309483e-01 4.60740000e-01 4.31235731e-01
1.17711103e+00 7.29523242e-01 9.33219016e-01 6.42314613e-01
7.61141658e-01 1.05241203e+00 3.13008934e-01 -1.77174479e-01
-6.96127489e-02 -3.89810979e-01 1.80314198e-01 5.05997419e-01
-5.19449890e-01 1.06898554e-01 -6.11790717e-01 2.63904899e-01
-2.10524797e+00 -9.00934756e-01 1.01206310e-01 2.03137755e+00
5.30023575e-01 3.13895307e-02 5.35193920e-01 3.20649147e-02
9.04438496e-01 1.33744001e-01 -3.93444568e-01 -3.36130857e-02
1.13209262e-01 -1.30609304e-01 3.32885057e-01 3.14070851e-01
-1.54691458e+00 9.02097285e-01 5.62652063e+00 8.55921149e-01
-8.88101757e-01 8.69300663e-02 3.34344566e-01 -2.09567964e-01
1.46989301e-01 -4.59767193e-01 -1.21264291e+00 4.57413137e-01
2.65799016e-01 4.59142178e-01 1.84826836e-01 8.48491848e-01
-7.41386935e-02 -5.53508811e-02 -1.04756153e+00 1.27882731e+00
2.10200593e-01 -6.80699110e-01 2.09851980e-01 -2.05583908e-02
3.71205539e-01 -2.94124633e-01 -2.28379458e-01 3.92291337e-01
-1.74162626e-01 -1.01618576e+00 9.98751462e-01 8.58314872e-01
5.80910981e-01 -1.05224419e+00 8.09509337e-01 4.63875681e-01
-1.87128198e+00 -2.63266027e-01 -1.86673552e-01 -1.25586450e-01
2.22003654e-01 2.88350046e-01 -3.33668858e-01 9.45608318e-01
9.82187510e-01 5.60071111e-01 -7.74949431e-01 9.79809821e-01
-6.27497509e-02 -2.10736409e-01 -4.09783721e-01 7.44756386e-02
-2.50435144e-01 2.45602444e-01 4.30846483e-01 1.31172597e+00
1.95150927e-01 -1.25934228e-01 6.46647930e-01 4.35171127e-01
3.19498152e-01 4.08136934e-01 -1.06935039e-01 6.83319211e-01
4.01939213e-01 1.29129565e+00 -5.52226663e-01 -2.61699021e-01
-5.55792391e-01 1.08671355e+00 4.06272709e-01 1.55968428e-01
-1.00283635e+00 -3.08481246e-01 5.30352116e-01 3.10019881e-01
3.97510141e-01 -1.27735093e-01 -2.00280696e-01 -1.08735001e+00
2.04104096e-01 -8.29686701e-01 5.24944961e-01 -4.64747846e-01
-1.12918627e+00 5.55221200e-01 1.71215519e-01 -1.36701655e+00
-1.08023934e-01 -5.08728087e-01 -3.13234389e-01 1.08843744e+00
-1.14447379e+00 -1.80921304e+00 -5.06148398e-01 9.71735418e-01
5.65251768e-01 7.55658373e-02 6.66189790e-01 4.36353952e-01
-6.21582389e-01 8.89838159e-01 -6.25915945e-01 3.71124238e-01
7.60090590e-01 -1.05641270e+00 2.59994209e-01 7.95495450e-01
2.75244284e-03 7.80516624e-01 6.16638422e-01 -9.26239133e-01
-1.23118782e+00 -9.92339373e-01 8.50594819e-01 -5.66537261e-01
-5.60422763e-02 -3.44550848e-01 -3.96995425e-01 6.67869449e-01
-2.06443995e-01 1.44430757e-01 5.86044610e-01 3.37102592e-01
-1.19013079e-01 -1.44696400e-01 -1.09221911e+00 4.87016886e-01
1.13940239e+00 -2.89290518e-01 -5.82585096e-01 6.66040182e-02
4.41333264e-01 -7.69853354e-01 -1.03398645e+00 5.50449610e-01
1.05110419e+00 -8.40654314e-01 1.20620549e+00 -4.46615934e-01
2.57031441e-01 -5.88735521e-01 -8.15696046e-02 -1.08909070e+00
-7.59682178e-01 -3.64558220e-01 -3.43928009e-01 1.21979213e+00
4.66476828e-02 -3.93725783e-01 6.07568800e-01 5.12526214e-01
6.37864470e-02 -8.81892443e-01 -9.26014304e-01 -3.16410512e-01
-2.55855262e-01 -3.20573986e-01 9.77403164e-01 5.42529464e-01
2.03300510e-02 2.58324534e-01 -7.25190282e-01 4.32239771e-01
4.56269175e-01 1.05275117e-01 1.09079719e+00 -1.38408673e+00
-5.26470661e-01 -1.85739666e-01 -6.75466955e-01 -1.29088378e+00
-3.60423058e-01 -4.38460946e-01 -6.25824183e-02 -1.45181239e+00
4.27568734e-01 -1.16712965e-01 -3.00870150e-01 3.42468739e-01
-5.67247450e-01 2.74872720e-01 3.21130902e-01 1.65840074e-01
-8.31815064e-01 4.34067577e-01 1.50964713e+00 5.82677387e-02
-2.53320426e-01 1.86523616e-01 -6.94881022e-01 8.17352116e-01
5.66088855e-01 -1.38548046e-01 -1.66053772e-01 -3.32628518e-01
1.38312176e-01 -6.32781088e-02 5.45392513e-01 -1.49057639e+00
2.99323648e-01 -5.69032356e-02 1.20267856e+00 -9.13803875e-01
5.94173193e-01 -8.92136216e-01 2.56973892e-01 4.81582046e-01
1.75393163e-03 1.89629704e-01 7.72466585e-02 5.26547968e-01
-1.50777802e-01 1.94871485e-01 5.24870336e-01 -3.90526682e-01
-8.90369713e-01 5.90311587e-01 7.38152862e-02 -1.62613347e-01
9.55894470e-01 -5.10204017e-01 5.80602773e-02 -1.06317490e-01
-9.42940950e-01 4.04506803e-01 2.11964086e-01 5.90305209e-01
4.85679656e-01 -1.59717917e+00 -5.41668534e-01 4.96389955e-01
1.82198495e-01 1.49671942e-01 6.95431590e-01 8.52306664e-01
-3.27673942e-01 2.70220488e-01 -3.63914788e-01 -6.85970485e-01
-1.40067756e+00 4.75504607e-01 4.99432445e-01 -3.66079003e-01
-5.45645177e-01 1.01823616e+00 1.57478258e-01 -4.20783252e-01
2.56408751e-01 -1.75149813e-01 -7.04374850e-01 3.23089302e-01
6.53449595e-01 4.10994768e-01 -1.69609144e-01 -1.27619278e+00
-5.96982419e-01 1.12920034e+00 -1.13983497e-01 -1.03301257e-02
1.15208554e+00 -3.76799464e-01 8.07444975e-02 2.69446135e-01
9.74432230e-01 8.35639089e-02 -1.34967363e+00 -2.13276833e-01
-3.09917331e-01 -5.27229309e-01 -3.12329978e-01 -7.00091720e-01
-8.66825283e-01 6.66067183e-01 8.45633686e-01 -2.34409139e-01
1.09405947e+00 1.66715737e-02 6.63868904e-01 1.53155411e-02
7.43513346e-01 -1.50591934e+00 -7.77461529e-02 3.93506497e-01
9.40941870e-01 -1.15874803e+00 2.00105041e-01 -5.54380298e-01
-6.02894843e-01 1.03925025e+00 1.06496310e+00 -1.11672789e-01
7.18904793e-01 1.42096102e-01 -9.62444544e-02 -3.23318481e-01
-2.11459324e-01 -4.49025214e-01 6.36922061e-01 6.93518758e-01
3.67247611e-01 2.56614059e-01 -3.75396580e-01 1.14535069e+00
-3.59408081e-01 -1.22341461e-01 -2.04794273e-01 8.36512506e-01
-5.67267299e-01 -9.78225946e-01 -7.41745353e-01 2.96208292e-01
-3.45497698e-01 2.17059493e-01 -2.44433194e-01 9.31854665e-01
9.19584453e-01 8.12903583e-01 -8.67168829e-02 -6.91590428e-01
6.11023903e-01 -9.52295400e-03 6.70778573e-01 -4.90763158e-01
-9.33774710e-01 3.21147829e-01 -2.38257404e-02 -9.41062450e-01
-5.09643316e-01 -8.79204929e-01 -9.90218878e-01 -6.67919740e-02
-4.02696639e-01 -1.17738545e-01 3.29014271e-01 1.11408269e+00
2.92955726e-01 8.57234001e-01 2.68650442e-01 -1.33870745e+00
-2.62882382e-01 -1.20665038e+00 -5.72893202e-01 5.44483781e-01
1.79725409e-01 -1.00530112e+00 1.59098536e-01 -2.44954020e-01] | [7.198568344116211, -0.7396367788314819] |
e79d5ad0-3ce3-4cbc-8020-37a5b094f521 | low-cost-and-high-performance-data | 2101.02353 | null | https://arxiv.org/abs/2101.02353v1 | https://arxiv.org/pdf/2101.02353v1.pdf | Low-cost and high-performance data augmentation for deep-learning-based skin lesion classification | Although deep convolutional neural networks (DCNNs) have achieved significant accuracy in skin lesion classification comparable or even superior to those of dermatologists, practical implementation of these models for skin cancer screening in low resource settings is hindered by their limitations in computational cost and training dataset. To overcome these limitations, we propose a low-cost and high-performance data augmentation strategy that includes two consecutive stages of augmentation search and network search. At the augmentation search stage, the augmentation strategy is optimized in the search space of Low-Cost-Augment (LCA) under the criteria of balanced accuracy (BACC) with 5-fold cross validation. At the network search stage, the DCNNs are fine-tuned with the full training set in order to select the model with the highest BACC. The efficiency of the proposed data augmentation strategy is verified on the HAM10000 dataset using EfficientNets as a baseline. With the proposed strategy, we are able to reduce the search space to 60 and achieve a high BACC of 0.853 by using a single DCNN model without external database, suitable to be implemented in mobile devices for DCNN-based skin lesion detection in low resource settings. | ['Ronald X. Xu', 'Peng Yao', 'Zhihong Zhang', 'Peng Liu', 'Chi Zhang', 'Liang Xu', 'Honghong Liu', 'Pengfei Shao', 'Fan Zhang', 'Mengjuan Xu', 'Shuwei Shen'] | 2021-01-07 | null | null | null | null | ['skin-lesion-classification'] | ['medical'] | [ 6.30604327e-01 6.71146438e-02 -3.66721898e-01 -1.20136119e-01
-7.03287303e-01 -3.00337821e-01 2.84461886e-01 5.31224906e-01
-9.32377636e-01 5.62358975e-01 -2.79309094e-01 -5.02151728e-01
-2.36390695e-01 -9.69844341e-01 -2.66840249e-01 -7.24487424e-01
2.79774696e-01 2.73071676e-01 3.28226119e-01 8.56784731e-02
-4.11560722e-02 7.21910417e-01 -1.54820895e+00 3.12548190e-01
1.01837468e+00 1.23841512e+00 2.04947621e-01 8.15440834e-01
-1.31370693e-01 2.78464288e-01 -3.32744867e-01 -6.40899062e-01
3.40684503e-01 -1.88418210e-01 -8.19721580e-01 -2.89844662e-01
4.81119573e-01 -2.03067124e-01 -7.20152482e-02 8.64678800e-01
9.71927643e-01 -1.65141255e-01 1.97531015e-01 -6.66197181e-01
5.41149378e-02 9.58901867e-02 -4.36757267e-01 2.61581033e-01
-1.48370266e-01 1.61784202e-01 5.81134856e-01 -5.92760861e-01
6.77737176e-01 5.83376467e-01 1.04545319e+00 9.63788033e-01
-1.04860556e+00 -3.35344315e-01 -1.79205790e-01 8.96594003e-02
-1.45312190e+00 -2.69778609e-01 3.63555878e-01 -6.21684827e-02
9.32671785e-01 6.42381370e-01 1.00681686e+00 9.20969546e-01
-1.72894835e-01 4.84411210e-01 1.01802909e+00 -7.53693879e-01
3.09039950e-01 3.47780228e-01 -6.74866959e-02 8.20071697e-01
4.66596007e-01 -1.36726156e-01 -2.57606894e-01 -1.77270561e-01
8.13412428e-01 -1.71039149e-01 1.56349853e-01 5.51231131e-02
-5.95921874e-01 8.58702958e-01 6.88579082e-01 4.53631788e-01
-7.00159550e-01 4.98253703e-02 5.31918287e-01 -1.45791575e-01
2.92958349e-01 7.02163041e-01 -2.31170967e-01 1.15944959e-01
-8.35305810e-01 -3.49440128e-02 5.04333735e-01 2.43474543e-01
1.70566231e-01 -1.83823839e-01 -4.04875875e-01 1.16366160e+00
-1.36956885e-01 6.53626025e-02 7.04132736e-01 -2.99481899e-01
2.69174129e-01 9.09943700e-01 -2.15587586e-01 -6.30329847e-01
-6.89950764e-01 -8.19228113e-01 -1.12471092e+00 1.03527546e-01
4.69449282e-01 -2.90979922e-01 -1.33423924e+00 1.38960481e+00
5.62149644e-01 4.18047085e-02 -3.21847126e-02 7.80620277e-01
9.31912541e-01 3.93094532e-02 4.58005250e-01 7.51687214e-02
1.58253288e+00 -8.68798375e-01 -4.29687679e-01 -1.83131024e-01
9.52505291e-01 -5.39041936e-01 9.24010992e-01 3.60330135e-01
-1.03459477e+00 -4.35141683e-01 -1.04517019e+00 1.60082921e-01
-5.79459906e-01 6.71393692e-01 8.60430717e-01 1.08746564e+00
-1.15323579e+00 2.74084955e-01 -9.23728526e-01 -8.74883235e-01
7.94488907e-01 6.60341501e-01 -4.33057368e-01 2.15680264e-02
-1.09848964e+00 9.20498252e-01 5.41510046e-01 3.43203962e-01
-5.83069801e-01 -7.42189109e-01 -5.51081419e-01 -2.32952550e-01
2.90026605e-01 -6.05470598e-01 9.11854982e-01 -9.09817874e-01
-1.26473022e+00 1.02705526e+00 -3.53104956e-02 -8.64701509e-01
6.17455959e-01 -3.03122979e-02 -4.50413495e-01 3.13721508e-01
-2.33861387e-01 7.59277225e-01 4.08890218e-01 -7.14473724e-01
-7.34578967e-01 -3.24698836e-01 1.35220677e-01 3.19740564e-01
-9.07586753e-01 -1.05672032e-01 -9.21182871e-01 -4.33684856e-01
-9.12207738e-02 -9.45985496e-01 -6.39152110e-01 2.36112744e-01
-4.15940404e-01 1.20391078e-01 5.24269879e-01 -8.18168163e-01
1.42052007e+00 -1.79858780e+00 -3.40479314e-01 6.00545526e-01
4.00737636e-02 1.08944869e+00 -3.69418859e-01 1.09452367e-01
-1.27399847e-01 2.28058025e-01 2.16802265e-02 -2.03878149e-01
-7.58777440e-01 -3.08762901e-02 5.86205244e-01 2.94683516e-01
5.84730566e-01 9.72995937e-01 -6.64512455e-01 -5.86770236e-01
4.33582127e-01 7.06150830e-01 -4.85113353e-01 -5.91690354e-02
-5.92429284e-03 1.16533056e-01 -4.29944158e-01 1.02010489e+00
5.79541564e-01 -4.22537684e-01 3.63418072e-01 -3.94613266e-01
7.58720785e-02 -2.61079315e-02 -1.18241072e+00 1.44492459e+00
-5.98792553e-01 5.09294868e-01 1.37299478e-01 -7.06376970e-01
8.07975888e-01 3.00219029e-01 3.63527179e-01 -7.66119659e-01
1.26209810e-01 2.49905542e-01 2.54305780e-01 -6.41063750e-01
2.12378249e-01 2.17159912e-01 4.19558287e-01 6.62860414e-03
-1.22920237e-02 3.79443735e-01 3.63866478e-01 -2.32727632e-01
1.21138752e+00 -2.59463668e-01 4.42523271e-01 -5.17268889e-02
7.64587104e-01 1.81725696e-01 1.75358266e-01 9.15341437e-01
-1.37570113e-01 4.53250319e-01 2.73609191e-01 -7.09219038e-01
-9.92605865e-01 -5.13039291e-01 -3.62290531e-01 8.91088188e-01
-2.09213212e-01 -1.03255853e-01 -9.89115179e-01 -8.66022289e-01
-2.65637368e-01 2.16236159e-01 -9.46164548e-01 6.85476512e-02
-5.17864287e-01 -1.38213396e+00 8.54917645e-01 5.70690751e-01
7.63146579e-01 -1.00692368e+00 -7.39600301e-01 1.29425719e-01
1.25543013e-01 -9.23702717e-01 3.68786566e-02 3.49187970e-01
-7.75937915e-01 -1.25174451e+00 -8.96113157e-01 -8.17871213e-01
1.09244502e+00 -2.17507794e-01 5.71196735e-01 4.97160017e-01
-1.09622252e+00 -1.14556186e-01 -1.75865278e-01 -3.37897778e-01
-3.90295744e-01 5.28033435e-01 -2.75669187e-01 8.74171555e-02
2.72238880e-01 -8.11026990e-02 -9.46823955e-01 2.47620624e-02
-8.94036710e-01 2.83510704e-02 1.06327915e+00 1.10871983e+00
7.64450371e-01 -2.82237791e-02 5.52591860e-01 -8.97984028e-01
6.61323786e-01 -1.97090417e-01 -5.23229837e-01 3.45831126e-01
-6.93022251e-01 -2.95259178e-01 5.42355478e-01 -5.87483346e-01
-9.39588785e-01 4.84483808e-01 -5.71856678e-01 -1.97570808e-02
-3.00307214e-01 6.36638463e-01 2.22308114e-01 -5.75508773e-01
1.00330937e+00 1.14974856e-01 2.04207122e-01 -5.10663509e-01
-1.35100067e-01 5.39796531e-01 4.11402375e-01 -2.56821420e-02
4.43562925e-01 2.70435095e-01 3.86736691e-01 -8.20283651e-01
-6.69147193e-01 -4.38740879e-01 -7.01775432e-01 -1.98669836e-01
8.10712874e-01 -7.50507295e-01 -5.76522112e-01 6.76968455e-01
-7.42854953e-01 -2.28298143e-01 -2.47754887e-01 3.33004028e-01
7.82734081e-02 1.43007621e-01 -4.31763858e-01 -9.24686611e-01
-9.43138480e-01 -9.43425119e-01 7.04380095e-01 7.12656140e-01
-1.55818775e-01 -1.06055856e+00 -2.09927484e-01 2.55001485e-01
6.69110060e-01 4.74965513e-01 9.30972815e-01 -1.01802170e+00
-7.14334175e-02 -9.31805968e-01 -3.32299769e-01 3.88480544e-01
2.70147115e-01 -7.79872239e-02 -9.78261113e-01 -2.83309013e-01
-6.33960307e-01 -2.91065037e-01 7.46942461e-01 5.55553138e-01
1.62113512e+00 -4.69646789e-02 -5.06267190e-01 6.69678628e-01
1.83603466e+00 2.55034685e-01 6.65379941e-01 4.58023518e-01
3.57636392e-01 3.83387476e-01 4.16340828e-01 1.99182369e-02
-3.15372944e-02 4.95237917e-01 5.76494873e-01 -7.81081021e-01
-4.72233653e-01 -1.34902224e-01 -5.45917869e-01 1.70750935e-02
-4.24669892e-01 -8.30282718e-02 -1.08634531e+00 8.42297792e-01
-1.38012922e+00 -5.57210505e-01 3.18399295e-02 2.41140604e+00
7.83422768e-01 3.20063263e-01 2.99407989e-01 2.40914106e-01
7.52233446e-01 -2.89266080e-01 -5.96680522e-01 -5.14371753e-01
-8.52739662e-02 5.84436536e-01 6.37850225e-01 2.23732859e-01
-1.28404844e+00 7.31715441e-01 6.28766918e+00 1.05267954e+00
-1.35275531e+00 6.03479445e-02 9.38389480e-01 -1.22732952e-01
3.82222265e-01 -5.11995375e-01 -9.67117369e-01 3.16125721e-01
8.97884846e-01 4.16332752e-01 2.95974851e-01 8.04304719e-01
-1.13460690e-01 -2.70590246e-01 -6.86849177e-01 6.55657291e-01
-1.13199010e-01 -1.75215518e+00 -1.33814272e-02 3.56610790e-02
6.37888014e-01 -3.41600776e-02 1.90608963e-01 -8.50073397e-02
-1.57509103e-01 -1.23372710e+00 -1.88437909e-01 3.44684482e-01
1.13013077e+00 -7.67589211e-01 1.09581137e+00 1.51503965e-01
-1.01434708e+00 -2.33031586e-01 -3.16949114e-02 3.83192986e-01
-2.42509469e-01 3.46267730e-01 -1.35380089e+00 4.24676538e-01
6.75757647e-01 -1.55087374e-02 -9.76173818e-01 1.35156918e+00
2.71091878e-01 5.73843598e-01 -6.16018951e-01 -5.89040995e-01
3.54566425e-01 2.01322466e-01 1.95957497e-01 1.31244135e+00
2.20397487e-01 -2.30952114e-01 -3.55353445e-01 4.31987673e-01
5.06462827e-02 3.53549570e-01 -1.78171739e-01 -1.48255140e-01
4.57392335e-01 1.56848228e+00 -8.55039299e-01 1.75375119e-02
-1.09401621e-01 7.35543907e-01 1.37948409e-01 -3.86461578e-02
-5.37219584e-01 -5.91820240e-01 2.90486485e-01 2.63746321e-01
1.83262646e-01 4.27336365e-01 -4.23104376e-01 -3.41048658e-01
5.77136725e-02 -7.67273486e-01 6.83034360e-01 -1.46930888e-01
-9.95117545e-01 9.32716429e-01 -3.37862015e-01 -9.82335031e-01
-4.39261906e-02 -7.45305538e-01 -6.99721575e-01 1.09844387e+00
-1.47302330e+00 -1.50902236e+00 -6.62234783e-01 4.95275944e-01
3.38187128e-01 -2.38618314e-01 1.16399181e+00 3.81144196e-01
-7.76370645e-01 1.15610361e+00 -1.34904891e-01 1.73788950e-01
2.56371409e-01 -1.09906197e+00 3.11944485e-01 6.58409476e-01
-3.95672083e-01 4.08813208e-01 1.53054088e-01 -6.38166547e-01
-1.04835117e+00 -1.20500326e+00 7.19583094e-01 -3.17025706e-02
2.94175118e-01 -2.58429915e-01 -6.80888474e-01 -9.63610560e-02
-2.66975552e-01 2.63634831e-01 9.77503240e-01 1.11357898e-01
5.28443307e-02 -2.89772183e-01 -1.62483561e+00 6.63758934e-01
6.26637697e-01 -2.47165889e-01 3.60439658e-01 5.09897649e-01
2.99805015e-01 -5.24858057e-01 -1.01126707e+00 7.10774541e-01
7.81163096e-01 -5.37090003e-01 1.02033770e+00 -6.64889455e-01
3.21783543e-01 1.04716107e-01 2.25794241e-01 -8.31368864e-01
3.17008048e-02 -2.89043576e-01 -3.54105085e-02 9.88488734e-01
6.48512185e-01 -6.23037696e-01 1.29032886e+00 4.64333326e-01
2.84199536e-01 -1.47040534e+00 -1.03856683e+00 -3.53986412e-01
-1.94084615e-01 -7.91558847e-02 4.63258356e-01 6.06489122e-01
-4.69472140e-01 -1.72601372e-01 -2.26134509e-01 1.19242489e-01
3.69871676e-01 -4.46064472e-01 3.90540481e-01 -1.15382445e+00
-1.33839771e-02 -4.36958611e-01 -4.38472927e-01 -2.82294065e-01
-4.80041146e-01 -7.57082880e-01 -1.94356844e-01 -1.58748424e+00
1.38337389e-01 -7.56720304e-01 -5.28042138e-01 6.86638117e-01
-1.42904222e-01 7.26754308e-01 -1.45381734e-01 -1.87490165e-01
-1.98378369e-01 -2.10076630e-01 1.09585094e+00 1.01969965e-01
-4.93806660e-01 1.61744282e-01 -6.90487504e-01 6.97039485e-01
8.92237544e-01 -7.71255940e-02 -1.51948318e-01 -1.08462952e-01
9.37717967e-03 -2.24117383e-01 4.24533755e-01 -1.17061543e+00
3.93985778e-01 -3.81737314e-02 9.25245464e-01 -6.09928727e-01
5.07513225e-01 -7.45270908e-01 1.66271806e-01 1.09709132e+00
-5.36413014e-01 -3.03922474e-01 5.90244114e-01 2.71152973e-01
7.78554529e-02 -5.86689353e-01 9.48980689e-01 -4.42078002e-02
-5.46570897e-01 3.39560688e-01 -2.36973435e-01 -6.48151338e-01
1.16915274e+00 -5.02067447e-01 -3.17117423e-01 1.57966584e-01
-9.17297065e-01 1.12899095e-02 1.55390576e-01 2.34311178e-01
3.98588389e-01 -1.17858708e+00 -5.15759468e-01 3.00934255e-01
1.61820486e-01 -4.73612025e-02 5.35075963e-01 8.54956746e-01
-8.38211358e-01 5.86268723e-01 -3.99732322e-01 -4.45634454e-01
-1.63101208e+00 1.18823528e-01 6.68041945e-01 -7.18706548e-01
-3.03360939e-01 1.02162158e+00 -5.41881979e-01 -4.18678582e-01
4.93159860e-01 7.99520016e-02 -4.59593385e-01 -5.73486276e-02
5.36522686e-01 4.34161484e-01 6.10762358e-01 -1.52863652e-01
-4.30751622e-01 3.59534293e-01 -3.20664674e-01 2.29949087e-01
1.27744937e+00 4.50726241e-01 1.68013256e-02 -4.14444923e-01
9.81717527e-01 -2.25932807e-01 -7.78817356e-01 -3.80822793e-02
-1.20571949e-01 -3.40071470e-01 3.38993788e-01 -1.41633344e+00
-1.03682756e+00 5.38000643e-01 1.39885736e+00 7.96936154e-02
1.58500874e+00 -2.95317888e-01 5.78058600e-01 4.76642698e-01
-1.01435915e-01 -1.26425159e+00 -9.56089795e-02 -5.88294640e-02
6.67138696e-01 -1.21293628e+00 1.64886191e-02 -4.60860878e-01
-5.61000407e-01 9.89083052e-01 1.00407732e+00 6.01715520e-02
4.42377537e-01 2.01879993e-01 5.41660450e-02 -2.20688134e-01
-5.69237888e-01 -4.45195466e-01 5.71844757e-01 7.65454829e-01
3.29103976e-01 -1.29706636e-01 -6.53786778e-01 3.36696208e-01
2.46813118e-01 8.82564709e-02 6.12439625e-02 8.18232536e-01
-2.87849028e-02 -1.17458415e+00 -7.38066807e-02 9.83861327e-01
-7.59115815e-01 -1.09216437e-01 -4.33787376e-01 9.69641209e-01
3.43408883e-01 6.54575467e-01 1.25430375e-01 -3.56176525e-01
3.70377839e-01 1.52790062e-02 3.61464053e-01 -4.30869192e-01
-1.04238570e+00 2.28477223e-03 4.10575092e-01 -4.62663174e-01
-2.81534612e-01 -4.13535416e-01 -7.45777309e-01 1.50126815e-01
-6.09800220e-01 -1.69831082e-01 1.15914321e+00 6.91363215e-01
3.44860762e-01 5.02723575e-01 3.53414536e-01 -1.98778048e-01
-3.60806644e-01 -9.87293303e-01 -1.72368765e-01 4.95428406e-02
1.11294985e-01 -2.19085693e-01 1.75275251e-01 -2.75123060e-01] | [15.660407066345215, -2.9455208778381348] |
183045e5-615c-4276-aefc-7a646b076e93 | a-global-context-mechanism-for-sequence | 2305.19928 | null | https://arxiv.org/abs/2305.19928v4 | https://arxiv.org/pdf/2305.19928v4.pdf | Supplementary Features of BiLSTM for Enhanced Sequence Labeling | Sequence labeling tasks require the computation of sentence representations for each word within a given sentence. A prevalent method incorporates a Bi-directional Long Short-Term Memory (BiLSTM) layer to enhance the sequence structure information. However, empirical evidence Li (2020) suggests that the capacity of BiLSTM to produce sentence representations for sequence labeling tasks is inherently limited. This limitation primarily results from the integration of fragments from past and future sentence representations to formulate a complete sentence representation. In this study, we observed that the entire sentence representation, found in both the first and last cells of BiLSTM, can supplement each the individual sentence representation of each cell. Accordingly, we devised a global context mechanism to integrate entire future and past sentence representations into each cell's sentence representation within the BiLSTM framework. By incorporating the BERT model within BiLSTM as a demonstration, and conducting exhaustive experiments on nine datasets for sequence labeling tasks, including named entity recognition (NER), part of speech (POS) tagging, and End-to-End Aspect-Based sentiment analysis (E2E-ABSA). We noted significant improvements in F1 scores and accuracy across all examined datasets. | ['Hongguang Sun', 'Kun Shen', 'Conglei Xu'] | 2023-05-31 | null | null | null | null | ['sentiment-analysis', 'part-of-speech-tagging', 'named-entity-recognition-ner', 'chinese-named-entity-recognition'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 5.45117199e-01 -7.11067691e-02 -7.01410025e-02 -5.83432198e-01
-7.76218474e-01 -7.80614376e-01 3.03111941e-01 2.67641425e-01
-6.92847669e-01 9.24416721e-01 4.74535286e-01 -4.17859584e-01
7.43438244e-01 -6.68423474e-01 -4.84713376e-01 -4.47786003e-01
3.13210458e-01 -2.43873730e-01 2.63375521e-01 -2.88262546e-01
4.56428200e-01 2.16893375e-01 -1.04997206e+00 8.14417481e-01
6.85229897e-01 7.31598079e-01 6.12157702e-01 8.46028447e-01
-7.93191791e-01 9.71383274e-01 -8.27480316e-01 -5.78820705e-01
-4.85783011e-01 -5.75602770e-01 -9.95509088e-01 -5.22063151e-02
6.22017160e-02 1.88804284e-01 1.82426646e-02 1.02485645e+00
5.01495838e-01 3.09711903e-01 4.56305683e-01 -6.78842425e-01
-7.42496192e-01 5.65307140e-01 -5.95435023e-01 5.90341389e-01
4.09466326e-01 1.24709018e-01 1.20095766e+00 -9.17217016e-01
5.81307054e-01 9.22272861e-01 7.90449798e-01 7.59934068e-01
-7.76369393e-01 -3.43275040e-01 3.84041131e-01 -7.01993704e-02
-9.95342493e-01 -2.92663693e-01 5.58141291e-01 -2.74589956e-01
1.85355115e+00 5.97335510e-02 5.28649747e-01 9.85751987e-01
8.51211965e-01 8.65441084e-01 7.76869178e-01 -5.97812831e-01
5.46287522e-02 -1.57969266e-01 5.71193576e-01 6.27704501e-01
-4.87524047e-02 -2.77096719e-01 -6.79741323e-01 1.52420681e-02
3.59485269e-01 -4.07814644e-02 2.97535181e-01 5.99898338e-01
-1.10159385e+00 6.37979686e-01 1.75071359e-01 8.18644881e-01
-5.95270634e-01 1.96503669e-01 1.01470792e+00 8.47261474e-02
5.99994838e-01 8.31172764e-02 -7.96807885e-01 -2.59487927e-01
-8.16364586e-01 -4.66574728e-01 5.90352178e-01 7.51061976e-01
5.50130486e-01 2.08485156e-01 -5.52662492e-01 7.67574608e-01
3.43650967e-01 1.44593403e-01 9.99919832e-01 -5.27065456e-01
5.43828428e-01 5.80283999e-01 -2.05230743e-01 -7.51116157e-01
-2.91862637e-01 -5.16505003e-01 -6.19679928e-01 -4.74435896e-01
6.33879900e-02 -4.49105442e-01 -9.20294166e-01 1.97939658e+00
1.58330411e-01 7.16185346e-02 3.61432433e-01 4.57900941e-01
8.29596758e-01 8.96486998e-01 7.83384323e-01 -3.05302620e-01
1.66534114e+00 -1.06214833e+00 -8.29048634e-01 -7.15995431e-01
1.18167889e+00 -8.09137762e-01 8.14668715e-01 -1.17202111e-01
-9.27904785e-01 -7.77911425e-01 -1.11566770e+00 -2.97898918e-01
-5.03951907e-01 1.52610511e-01 5.00711560e-01 6.79323256e-01
-1.15689838e+00 3.53631347e-01 -7.16637492e-01 -4.61781830e-01
1.18209168e-01 1.57998994e-01 -2.19593033e-01 2.16130197e-01
-1.53160548e+00 1.08241379e+00 5.28945267e-01 4.66754824e-01
-5.46576560e-01 -2.78856635e-01 -1.01599729e+00 5.89700835e-03
-2.27595910e-01 -7.41653740e-01 1.46766376e+00 -9.42417204e-01
-1.18660831e+00 9.56930637e-01 -7.21098065e-01 -5.45027316e-01
-4.07047987e-01 -6.52116984e-02 -4.27972645e-01 5.96796051e-02
1.82915360e-01 8.19525182e-01 4.75509763e-01 -8.01228642e-01
-6.39435410e-01 -4.09550220e-01 -7.83510655e-02 2.44094983e-01
-2.95530111e-01 5.09247303e-01 2.53060400e-01 -7.18184590e-01
-1.19350836e-01 -4.97063249e-01 -4.11566108e-01 -4.83917385e-01
-1.37313500e-01 -5.58288872e-01 4.60005730e-01 -9.98388410e-01
1.47761655e+00 -2.07443213e+00 -4.18798119e-01 -6.15560412e-01
-3.13543677e-01 4.09293920e-01 -4.19476271e-01 6.98119044e-01
-3.19976211e-01 2.90742546e-01 -3.04759264e-01 -5.09343207e-01
-2.07005844e-01 3.01721781e-01 -4.43462849e-01 1.02807745e-01
4.69096720e-01 1.14486921e+00 -1.00675893e+00 -6.72171295e-01
4.81011458e-02 3.67397219e-01 4.25862893e-02 -7.65739381e-02
-2.65442759e-01 3.00617039e-01 -3.63109529e-01 3.79776776e-01
3.62230569e-01 -1.38185427e-01 3.30874562e-01 5.60448207e-02
-1.75994843e-01 9.10506070e-01 -4.87889618e-01 1.96264207e+00
-6.12648785e-01 5.35770416e-01 -3.28676283e-01 -9.40427601e-01
1.05921865e+00 7.93776751e-01 1.73003525e-01 -7.92972863e-01
1.11905403e-01 2.59509236e-01 5.73888011e-02 -3.68980438e-01
6.58573508e-01 -6.84352577e-01 -3.92109632e-01 5.45616567e-01
2.60244012e-01 3.04151118e-01 4.02258426e-01 8.69410113e-02
1.02569199e+00 1.62364393e-01 4.92240220e-01 -7.56879374e-02
8.75494719e-01 -5.23381773e-03 6.30612195e-01 4.42489684e-01
-5.36372244e-01 5.27762592e-01 2.79974759e-01 -3.26965868e-01
-9.38092291e-01 -7.23201632e-01 2.58527100e-01 1.37884402e+00
-3.91612530e-01 -2.52535671e-01 -8.49703133e-01 -8.98311496e-01
-6.59355402e-01 1.16572380e+00 -3.50202054e-01 -2.26974130e-01
-7.19779730e-01 -8.22227061e-01 8.91120493e-01 8.47438812e-01
3.83960515e-01 -1.68101203e+00 -5.87537766e-01 6.54120505e-01
-5.13054013e-01 -9.90906060e-01 -9.15324926e-01 4.49082434e-01
-1.00363433e+00 -5.69572508e-01 -6.45863295e-01 -1.37394667e+00
5.23102701e-01 3.22041363e-02 1.00017536e+00 -1.38678387e-01
-2.45545991e-02 3.92014571e-02 -3.69837433e-01 -2.59116858e-01
-4.40696031e-01 2.18400493e-01 -2.43719697e-01 -2.45306447e-01
7.02521443e-01 -3.70245665e-01 -3.52332681e-01 -2.30502322e-01
-7.29123473e-01 2.09183320e-02 5.26143670e-01 7.89545953e-01
4.13695425e-01 -3.57340664e-01 1.02317142e+00 -5.70622087e-01
6.94458544e-01 -3.38996202e-01 6.49581105e-02 4.73177135e-01
-1.79048732e-01 9.40861702e-02 6.20281219e-01 -1.62298664e-01
-1.40881240e+00 4.10410576e-02 -8.62568080e-01 3.82373869e-01
-2.44621009e-01 7.49251664e-01 3.22231129e-02 3.91340137e-01
2.92405427e-01 8.79078507e-01 -6.64656535e-02 -2.45082036e-01
2.95377672e-01 7.93168664e-01 3.02043200e-01 -2.73822904e-01
-9.79507118e-02 5.28435446e-02 -2.99143881e-01 -8.69777620e-01
-1.42363727e+00 -4.44277853e-01 -8.29855740e-01 -5.61565682e-02
1.06796706e+00 -8.87897670e-01 -7.95033693e-01 4.77501601e-01
-1.78880191e+00 2.15516314e-02 -2.24898770e-01 2.79656380e-01
-3.09547782e-01 7.65901685e-01 -1.07022214e+00 -9.28103626e-01
-9.44245160e-01 -8.26997459e-01 9.64153230e-01 2.81287402e-01
-5.91949403e-01 -1.21217966e+00 2.63554245e-01 2.46703312e-01
1.38760746e-01 -6.98606446e-02 7.81012118e-01 -8.10696781e-01
2.49297127e-01 -2.52540469e-01 -4.16298620e-02 6.73643053e-01
1.81751028e-01 -1.57824084e-01 -9.78716254e-01 -1.65837910e-02
4.57263738e-01 -1.74353287e-01 1.03884077e+00 2.96780229e-01
6.64629042e-01 -2.32848912e-01 -1.42669350e-01 -1.17537968e-01
1.33059323e+00 5.05192339e-01 6.65473104e-01 1.97587624e-01
5.25866032e-01 7.47095346e-01 7.15409935e-01 1.21950932e-01
5.23638010e-01 -1.09973261e-02 2.58622393e-02 1.55262560e-01
-2.43165463e-01 -3.18235129e-01 9.49619710e-01 1.46177447e+00
5.43072402e-01 -4.83043104e-01 -7.41810024e-01 8.39261711e-01
-1.54674768e+00 -9.76856411e-01 -2.88754106e-01 1.71990025e+00
9.87882376e-01 2.45450959e-01 -2.51710922e-01 -6.42811134e-02
9.78314340e-01 4.30466384e-01 -4.12937105e-01 -1.06517482e+00
-1.72019899e-01 1.71496481e-01 2.65280247e-01 2.85397440e-01
-8.59434128e-01 1.08493125e+00 7.04596376e+00 1.03819990e+00
-9.37292159e-01 2.57348984e-01 8.63439977e-01 3.23889077e-01
-3.67752284e-01 -2.94530578e-02 -1.17327785e+00 5.77230275e-01
1.47490776e+00 -7.73672387e-02 -1.53611898e-01 5.60020983e-01
2.05532759e-01 -2.82202274e-01 -7.80778885e-01 5.50190628e-01
6.82381541e-02 -1.23987532e+00 1.93743825e-01 -2.99749404e-01
5.44759810e-01 7.20526427e-02 -2.03454792e-01 4.92388994e-01
1.14643008e-01 -8.27793479e-01 6.80013776e-01 4.24845964e-01
6.03377044e-01 -7.20854819e-01 9.66639042e-01 6.47497952e-01
-1.56391549e+00 9.54654068e-02 -3.22814524e-01 -3.85370195e-01
5.65788805e-01 5.91775239e-01 -8.03849161e-01 5.98076999e-01
1.37488350e-01 6.69035375e-01 -3.93840402e-01 5.75993359e-01
-4.65765774e-01 7.34311104e-01 5.40049858e-02 -4.12006557e-01
5.56914151e-01 -7.36764744e-02 2.39699364e-01 1.66081715e+00
2.71778584e-01 5.00792339e-02 -6.74706325e-02 5.00887871e-01
-9.15031731e-02 8.87180194e-02 -4.62767750e-01 -6.02153182e-01
3.39352161e-01 1.24973035e+00 -9.35232520e-01 -5.06982923e-01
-5.98628879e-01 1.12592769e+00 4.41917330e-01 1.39978439e-01
-8.25872719e-01 -4.66599405e-01 5.36030114e-01 -5.74284494e-01
4.59966987e-01 -3.26933146e-01 -8.25405538e-01 -9.38895345e-01
2.16714237e-02 -3.66058022e-01 5.64573884e-01 -7.43961573e-01
-1.32317901e+00 7.30605364e-01 -7.07442462e-01 -8.70297551e-01
-2.01162845e-01 -4.80263829e-01 -8.02293718e-01 1.22031093e+00
-1.45676589e+00 -1.10467291e+00 4.30021763e-01 -6.59086974e-03
1.00138175e+00 1.93312317e-01 9.86955285e-01 2.60640353e-01
-7.21747458e-01 3.93171638e-01 -7.97681063e-02 3.33000660e-01
3.34001720e-01 -1.00187159e+00 8.62524271e-01 1.08544958e+00
2.19730258e-01 8.70252013e-01 4.57313478e-01 -8.86054099e-01
-1.01118112e+00 -1.09153724e+00 1.80625534e+00 -3.39543819e-01
5.96048355e-01 -2.73104876e-01 -7.93271244e-01 8.26784015e-01
3.99098754e-01 -3.51701856e-01 1.02760959e+00 -1.76255301e-01
-3.63323428e-02 2.08554685e-01 -1.05371845e+00 4.68403965e-01
8.51766169e-01 -8.63026500e-01 -1.00004029e+00 4.88551296e-02
1.18622947e+00 -4.15816419e-02 -7.98611104e-01 2.78110534e-01
4.56856400e-01 -5.53127825e-01 6.90553010e-01 -6.19885385e-01
4.92355615e-01 -3.96139652e-01 -2.05850556e-01 -9.75918770e-01
-1.77370638e-01 -1.86308011e-01 -1.11375526e-01 1.46591675e+00
7.22171664e-01 -5.68111122e-01 5.43776393e-01 3.40148002e-01
-5.33337712e-01 -7.71422625e-01 -8.76120985e-01 -6.65266991e-01
1.51118666e-01 -5.91540694e-01 3.57939541e-01 5.87061822e-01
2.47970611e-01 9.05646980e-01 -1.42375723e-01 -1.33195609e-01
2.34570891e-01 -2.29727160e-02 -1.93659753e-01 -5.76892614e-01
5.56618124e-02 -2.26605996e-01 -1.14837969e-02 -1.34036875e+00
3.88164997e-01 -1.14854228e+00 4.42850471e-01 -1.98239005e+00
4.32127506e-01 1.42984569e-01 -5.48255146e-01 5.18278301e-01
-3.98245394e-01 1.61714047e-01 1.15149602e-01 -1.77209064e-01
-6.97499692e-01 6.14919424e-01 1.33650613e+00 1.09963857e-01
1.25275090e-01 -2.38395259e-01 -8.06741476e-01 6.64843917e-01
9.82186317e-01 -6.99061155e-01 -1.05942473e-01 -5.45320630e-01
3.56981426e-01 1.97719976e-01 -1.93441331e-01 -7.48652399e-01
3.21403354e-01 -5.35296313e-02 4.54370946e-01 -1.10358369e+00
2.11446553e-01 -2.96195567e-01 -2.65696079e-01 7.62120306e-01
-5.23887694e-01 1.35318235e-01 3.41006756e-01 4.27488834e-01
-3.60981256e-01 -6.93494737e-01 5.00558615e-01 -4.85873491e-01
-1.03118372e+00 -1.74587928e-02 -9.97519314e-01 -4.73891646e-02
9.95141506e-01 -3.83883923e-01 -1.83892310e-01 9.84648466e-02
-9.01341438e-01 2.54471868e-01 2.50345189e-02 3.50871801e-01
6.67474747e-01 -1.04786837e+00 -4.75163519e-01 -8.52070898e-02
-1.04626276e-01 -3.06764394e-01 5.00710845e-01 6.98271513e-01
-3.56797338e-01 8.20644259e-01 -6.60062507e-02 -2.38251522e-01
-1.16177094e+00 6.05919182e-01 1.73982933e-01 -6.60025060e-01
-1.60905451e-01 1.12438250e+00 2.65456617e-01 -6.57643974e-01
-6.16038889e-02 -3.73253822e-01 -5.86371899e-01 2.35196590e-01
5.42925239e-01 1.58273280e-01 -6.55265301e-02 -7.93421507e-01
-6.67324483e-01 3.27866673e-01 -1.36994943e-01 -1.97091475e-01
9.88519490e-01 -5.52922964e-01 -4.04451638e-01 8.96765172e-01
1.22612035e+00 -3.88646901e-01 -8.77038002e-01 -1.56515390e-02
5.37949026e-01 3.69228721e-01 -3.44945043e-01 -7.40143061e-01
-5.60942054e-01 1.19832945e+00 4.57060337e-02 -6.44722953e-02
9.63010252e-01 -2.37928659e-01 1.45942104e+00 1.76960588e-01
3.26551050e-01 -1.00448585e+00 1.02279022e-01 1.14965844e+00
4.91618216e-01 -9.07085717e-01 -5.53864121e-01 -2.51416445e-01
-9.33010340e-01 1.22068548e+00 5.95557451e-01 -6.53715432e-02
3.93140823e-01 2.34500811e-01 -6.50688112e-02 -5.28968237e-02
-1.01297057e+00 -8.83338600e-02 1.64536051e-02 1.95586860e-01
1.05014968e+00 -4.27643321e-02 -7.77059078e-01 8.30459774e-01
-2.03671400e-02 -1.05371308e-02 3.65999520e-01 1.24547732e+00
-7.49034286e-01 -1.14035034e+00 5.62379025e-02 4.40337598e-01
-7.87856102e-01 -5.30852437e-01 -2.45726585e-01 2.84688976e-02
1.11672454e-01 1.18588102e+00 1.40989050e-01 -2.48120829e-01
3.18764150e-02 6.17474973e-01 4.51657265e-01 -1.13551509e+00
-1.03140163e+00 -1.16546705e-01 3.39294314e-01 -1.23293705e-01
-4.69697148e-01 -7.04114258e-01 -1.72525465e+00 7.57842734e-02
-2.20158339e-01 1.41823038e-01 6.15468323e-01 1.30675590e+00
3.26662481e-01 8.37313831e-01 4.56021667e-01 -3.45547289e-01
-5.57878494e-01 -1.30196142e+00 -4.25395846e-01 -5.45104183e-02
2.37883165e-01 2.58882367e-03 -6.47586361e-02 2.34553650e-01] | [10.035993576049805, 9.517949104309082] |
7b1979df-0153-46d8-b1e6-f10a90968265 | mordecai-3-a-neural-geoparser-and-event | 2303.13675 | null | https://arxiv.org/abs/2303.13675v1 | https://arxiv.org/pdf/2303.13675v1.pdf | Mordecai 3: A Neural Geoparser and Event Geocoder | Mordecai3 is a new end-to-end text geoparser and event geolocation system. The system performs toponym resolution using a new neural ranking model to resolve a place name extracted from a document to its entry in the Geonames gazetteer. It also performs event geocoding, the process of linking events reported in text with the place names where they are reported to occur, using an off-the-shelf question-answering model. The toponym resolution model is trained on a diverse set of existing training data, along with several thousand newly annotated examples. The paper describes the model, its training process, and performance comparisons with existing geoparsers. The system is available as an open source Python library, Mordecai 3, and replaces an earlier geoparser, Mordecai v2, one of the most widely used text geoparsers (Halterman 2017). | ['Andrew Halterman'] | 2023-03-23 | null | null | null | null | ['toponym-resolution'] | ['natural-language-processing'] | [-4.15117949e-01 2.00186819e-01 -5.29920869e-02 -2.66712397e-01
-1.28650701e+00 -7.14530945e-01 1.05960381e+00 9.41613793e-01
-8.33958149e-01 8.12910438e-01 9.83724952e-01 9.52236354e-02
-2.64828473e-01 -1.17221987e+00 -7.16788173e-01 -1.61429211e-01
-6.22603185e-02 9.69230592e-01 3.35626453e-01 -2.05229968e-01
3.20493340e-01 2.86972076e-01 -1.66745293e+00 3.93800974e-01
5.63576102e-01 8.95710826e-01 -8.02437589e-02 6.55039012e-01
-3.21996123e-01 9.48236287e-01 -7.36326456e-01 -2.87504196e-01
-3.14269125e-01 6.88311681e-02 -8.27643692e-01 -1.22585714e+00
6.42178655e-01 1.63269490e-01 -7.46704996e-01 5.99046350e-01
6.36462569e-01 2.56209493e-01 6.83220088e-01 -9.98180687e-01
-9.74740326e-01 1.29204488e+00 -2.40364075e-01 7.51238763e-01
7.26584315e-01 -5.66041529e-01 1.24007821e+00 -1.21942687e+00
9.87074971e-01 9.42788422e-01 9.67596233e-01 -1.47279933e-01
-6.00933909e-01 -6.35246634e-01 -3.79958868e-01 3.02596658e-01
-1.95457196e+00 -2.10510030e-01 2.45390348e-02 -4.64386195e-01
1.67479718e+00 2.01435894e-01 6.59952760e-02 9.81555402e-01
1.40002951e-01 3.73903841e-01 4.15410429e-01 -3.49481165e-01
4.18216288e-01 -4.49179322e-01 2.51301169e-01 3.77982587e-01
2.78800219e-01 -2.15814576e-01 -9.99806583e-01 -5.45610011e-01
2.39225477e-01 -6.75629685e-03 -2.04006061e-02 1.65593222e-01
-1.35059428e+00 4.65505451e-01 6.25342131e-01 4.16773677e-01
-7.91649222e-01 2.90894568e-01 3.49497586e-01 -2.79629767e-01
6.59141958e-01 5.71640372e-01 -5.34358144e-01 -3.83405268e-01
-1.12694895e+00 7.18548357e-01 9.10213590e-01 9.40955877e-01
7.08717108e-01 -3.66043299e-01 -4.25031483e-01 7.95452237e-01
4.06884104e-01 3.84518355e-01 7.35036790e-01 -5.45411289e-01
9.18426991e-01 8.02695632e-01 2.78189719e-01 -9.92686272e-01
-7.16707468e-01 -3.60540032e-01 -1.13943145e-01 -4.79016364e-01
3.45925033e-01 -1.71385780e-01 -6.91373050e-01 1.42834353e+00
3.59637022e-01 5.23712516e-01 1.51137546e-01 5.40597796e-01
1.34427595e+00 9.27924871e-01 6.98862851e-01 3.88002783e-01
1.78434086e+00 -3.60041559e-01 -6.84014380e-01 -3.09122592e-01
8.01889718e-01 -5.63066661e-01 6.58814788e-01 -1.17002897e-01
-7.73229122e-01 -7.90850595e-02 -9.16582644e-01 -3.14353913e-01
-1.35400844e+00 -2.85388473e-02 5.98796248e-01 2.26727188e-01
-9.93668258e-01 5.97056627e-01 -7.21823514e-01 -9.47823167e-01
3.65378037e-02 -5.77742606e-02 -5.41913092e-01 1.17981613e-01
-1.86950600e+00 9.93108273e-01 1.08852506e+00 -3.92716110e-01
-3.50624740e-01 -1.06579638e+00 -1.15664017e+00 2.06812322e-01
1.96313515e-01 -4.72898185e-01 1.41009188e+00 3.43654275e-01
-8.38492751e-01 1.06381917e+00 -1.21354572e-02 -6.61682725e-01
-4.00746204e-02 -3.32009882e-01 -1.07405722e+00 9.98274311e-02
8.23887408e-01 4.25518662e-01 1.35657072e-01 -2.48171940e-01
-1.16060221e+00 -3.46323311e-01 -2.94598460e-01 2.86078930e-01
2.35561728e-02 4.44123864e-01 -5.30232906e-01 -7.57823765e-01
-1.91230103e-02 -4.31451380e-01 2.34258384e-01 -6.28259778e-01
-3.03954840e-01 -6.97681487e-01 2.57963210e-01 -9.12110388e-01
1.63859606e+00 -2.11520314e+00 -4.59476680e-01 1.05408497e-01
3.40402015e-02 -2.17957899e-01 3.00067633e-01 1.18325889e+00
-3.44550133e-01 -8.40897951e-03 -1.01440400e-01 -8.30203667e-02
2.59016454e-01 -1.04940116e-01 -7.27153182e-01 4.24679905e-01
5.86751401e-02 8.99242580e-01 -1.32189119e+00 -4.99482870e-01
8.83342922e-02 3.88208449e-01 -3.01106900e-01 -4.08148095e-02
-8.82673562e-02 -2.60374159e-01 -3.15817088e-01 6.46439254e-01
3.38840306e-01 -2.36943260e-01 -1.09203607e-01 -6.48240745e-02
-4.43523288e-01 9.13597167e-01 -1.22325134e+00 1.73875344e+00
-3.44914109e-01 5.88775158e-01 -7.19243944e-01 -3.60915452e-01
7.61301458e-01 5.14522970e-01 4.93801743e-01 -6.92147374e-01
-9.73455384e-02 3.40713978e-01 -9.76977766e-01 -4.48735476e-01
1.39407921e+00 4.89615649e-01 -8.62171888e-01 3.58954966e-01
3.55671883e-01 3.05963486e-01 6.54104650e-01 4.64018941e-01
1.40446424e+00 1.37360752e-01 6.95314705e-01 -1.52765363e-01
2.78317213e-01 4.63159472e-01 3.15018654e-01 9.23907876e-01
1.10122085e-01 7.83703148e-01 2.08965480e-01 -4.23405647e-01
-9.94073033e-01 -1.20559394e+00 -3.95366639e-01 1.43574965e+00
-2.19915047e-01 -9.42796588e-01 -6.34981692e-01 -5.37420154e-01
2.78175443e-01 1.25559092e+00 -8.59486639e-01 1.16685294e-01
-6.13473594e-01 -6.91056311e-01 1.20280647e+00 6.55566692e-01
3.71774077e-01 -1.25496733e+00 -4.64253098e-01 4.40348506e-01
-4.47007954e-01 -7.68552899e-01 -4.14005309e-01 1.81819543e-01
-6.75148219e-02 -1.10727489e+00 -5.63939810e-01 -5.31852961e-01
8.88365582e-02 -1.78802699e-01 1.45346057e+00 -5.43631911e-01
-2.19758064e-01 2.43731216e-01 -5.44418693e-01 -4.46513295e-01
3.57634313e-02 6.31031096e-01 2.22681817e-02 -2.80962437e-01
9.96386111e-01 -4.56543803e-01 -5.69639981e-01 3.48428427e-03
-9.04185474e-01 -2.85336822e-01 -2.11245939e-02 2.86760509e-01
6.48277164e-01 -2.26050794e-01 7.50677943e-01 -9.38288271e-01
4.51836079e-01 -1.32104218e+00 -6.03353739e-01 2.21582264e-01
-4.25624162e-01 1.40464142e-01 3.26893330e-01 1.96522206e-01
-1.08703470e+00 3.36097218e-02 -3.33773345e-01 1.74811348e-01
-2.70191938e-01 1.35520554e+00 1.25346735e-01 6.56461835e-01
1.31533742e+00 -6.80741444e-02 -9.34498847e-01 -7.32648075e-01
7.48632431e-01 9.63038802e-01 1.29249525e+00 -2.97790706e-01
5.22531211e-01 3.65559489e-01 -6.26592875e-01 -4.38838989e-01
-1.05579388e+00 -8.55192304e-01 -4.73784506e-01 2.13507693e-02
9.05066073e-01 -1.25237739e+00 -4.73789752e-01 3.51121545e-01
-1.02834618e+00 1.53713331e-01 -3.20876867e-01 5.30502081e-01
-2.14286759e-01 -2.17107341e-01 -3.54686409e-01 -2.52633423e-01
-4.22736377e-01 -6.70132600e-03 1.27617157e+00 5.78577161e-01
-5.98399878e-01 -8.43216956e-01 5.92251897e-01 -3.78428288e-02
3.67832959e-01 4.89611566e-01 5.63234389e-01 -1.34781802e+00
-6.55412674e-02 -6.16160750e-01 -2.14827418e-01 -8.06008637e-01
-4.16141786e-02 -9.58235636e-02 -8.80763829e-01 2.41819710e-01
-7.16558456e-01 1.09377317e-01 9.02941287e-01 2.28436619e-01
8.51609349e-01 -3.43611330e-01 -5.74842930e-01 5.42299330e-01
1.40375018e+00 7.99283087e-02 6.10702038e-01 9.62332606e-01
5.93128979e-01 2.56959498e-01 4.88723308e-01 5.25990188e-01
1.00716448e+00 5.91742635e-01 3.79241616e-01 4.48534638e-01
1.17133804e-01 -7.79577434e-01 6.43638754e-03 4.46269631e-01
4.30332094e-01 -5.59899747e-01 -1.48979700e+00 9.45637882e-01
-2.03843808e+00 -1.19868696e+00 -1.38447061e-01 2.07112026e+00
9.14636910e-01 -2.17674375e-01 -1.32880628e-01 -1.93303958e-01
1.00483632e+00 4.23368394e-01 -2.50754356e-01 -6.80560768e-02
-3.08333606e-01 2.12399557e-01 9.52098787e-01 2.39074662e-01
-1.44816720e+00 1.23259163e+00 6.69494438e+00 6.23184562e-01
-7.75553048e-01 2.00995952e-01 5.22119505e-03 -1.18081793e-01
6.73206244e-03 1.03780115e-02 -1.28622866e+00 6.02298260e-01
1.57080853e+00 -6.11078680e-01 3.23919326e-01 9.26789343e-01
-1.50717586e-01 -3.08243513e-01 -1.02966332e+00 1.06641257e+00
2.89302677e-01 -1.81170475e+00 -1.82982296e-01 -3.18389475e-01
3.10005933e-01 6.99827135e-01 -4.16694999e-01 5.57341576e-01
8.69763136e-01 -8.48537982e-01 1.04839921e+00 7.63634443e-01
8.21359932e-01 -7.18083024e-01 6.30129635e-01 1.23995021e-02
-1.33555079e+00 1.10251522e-02 -1.67176977e-01 1.30407184e-01
4.16485935e-01 5.87367952e-01 -1.06669176e+00 6.68746889e-01
1.12137139e+00 9.29451048e-01 -9.48210180e-01 1.29094648e+00
-3.86584789e-01 6.44547403e-01 -5.90313733e-01 2.90856622e-02
1.33368507e-01 5.53772569e-01 6.37512863e-01 1.36145139e+00
6.49602234e-01 2.28401157e-03 -2.48033315e-01 5.06148994e-01
-4.27067041e-01 9.51749831e-02 -6.23380005e-01 -1.86596662e-01
1.30397558e+00 1.22960460e+00 -4.97057110e-01 -5.16164362e-01
-1.67795688e-01 6.68553352e-01 5.45360029e-01 1.74771994e-01
-8.25896502e-01 -1.20740044e+00 3.47515643e-01 7.10681081e-02
2.10444957e-01 1.91117264e-03 -1.49845019e-01 -1.10002506e+00
-2.83207566e-01 -2.16299146e-01 1.33313894e+00 -1.42879593e+00
-1.30609131e+00 5.29346347e-01 2.97445923e-01 -1.04534924e+00
-8.16845119e-01 -3.20646703e-01 -4.43497688e-01 9.49887812e-01
-1.18464100e+00 -1.06253958e+00 -2.88230598e-01 4.93362069e-01
2.66915746e-03 -1.44411042e-01 9.94478464e-01 8.31397533e-01
-4.46743131e-01 4.89468068e-01 3.73164237e-01 4.62326229e-01
1.08054507e+00 -1.49764550e+00 1.01208329e+00 7.37452209e-01
2.35103518e-01 7.91046798e-01 7.45831072e-01 -1.02641606e+00
-9.74731445e-01 -1.51754749e+00 1.86036694e+00 -9.16898191e-01
1.16377866e+00 -4.21473354e-01 -1.06319666e+00 1.07830739e+00
2.68728346e-01 -1.29637137e-01 6.29700363e-01 1.91860646e-01
-5.29891729e-01 2.49013156e-02 -9.48126256e-01 4.69421744e-01
9.96379554e-01 -8.23148012e-01 -1.26486826e+00 4.88491923e-01
7.03724861e-01 -8.52415681e-01 -9.67103839e-01 -8.61115605e-02
2.16548964e-01 -1.50325909e-01 1.02826273e+00 -4.14372444e-01
2.47043446e-01 -6.18338406e-01 -2.31351346e-01 -1.29880488e+00
-3.30528080e-01 -4.38217580e-01 -2.90056139e-01 1.59087598e+00
7.01444983e-01 -8.45311165e-01 3.19879293e-01 4.02303755e-01
-3.33468050e-01 -1.99122587e-03 -1.35566914e+00 -5.34910202e-01
-3.03355213e-02 -4.39581335e-01 1.03659070e+00 1.14795470e+00
3.56667072e-01 2.11314946e-01 -9.63029563e-02 5.59827685e-01
2.39054322e-01 -2.80433804e-01 3.04804116e-01 -1.50240827e+00
1.80016264e-01 -1.87351719e-01 -3.09968054e-01 -2.74257630e-01
-9.50520858e-02 -1.26807296e+00 1.72634453e-01 -1.91225278e+00
-3.22192252e-01 -2.38921732e-01 -5.20621419e-01 9.69674408e-01
-1.79364786e-01 3.27003986e-01 -3.81739885e-01 3.97694319e-01
-8.45608234e-01 3.22503328e-01 -2.39016131e-01 2.61135101e-02
-2.03934327e-01 -5.40758431e-01 -8.47152233e-01 5.34559131e-01
5.75882852e-01 -1.00648427e+00 1.68982401e-01 -5.03403306e-01
1.12248123e+00 -7.63838589e-02 2.22137898e-01 -1.20720875e+00
6.86451912e-01 1.37578368e-01 3.70920002e-01 -1.25156927e+00
-1.38026997e-02 -4.14185673e-01 3.55852216e-01 -8.62967372e-02
-4.68190432e-01 6.36585414e-01 2.21336201e-01 4.77918357e-01
-3.05625349e-01 -1.81574047e-01 1.96773276e-01 -4.87129241e-02
-1.18665743e+00 1.25868261e-01 -7.77512431e-01 5.50057173e-01
8.01544845e-01 1.10905707e-01 -8.89196396e-01 -3.48289385e-02
-6.08100832e-01 3.51000249e-01 1.36431023e-01 9.21933889e-01
5.18726744e-02 -1.60407937e+00 -7.55135000e-01 -2.97642708e-01
7.59240448e-01 -2.04332903e-01 1.84064329e-01 4.12813753e-01
-8.98983359e-01 6.80401862e-01 7.30508193e-02 -1.51904121e-01
-5.84350824e-01 4.21401650e-01 2.96868384e-01 -3.38280141e-01
-8.07394683e-01 6.17790401e-01 -3.32104772e-01 -6.33841872e-01
1.80848077e-01 -2.67187804e-01 -7.30783761e-01 4.59633052e-01
1.07967114e+00 4.13969874e-01 5.30020535e-01 -6.86098635e-01
-8.26056182e-01 6.96628615e-02 -1.98567435e-02 -4.48708117e-01
1.51697040e+00 -3.74779664e-02 -1.01359643e-01 5.62836111e-01
9.09874439e-01 1.37281641e-01 -5.30242980e-01 -5.83493471e-01
6.76685154e-01 -8.74550864e-02 -9.67915729e-03 -1.06186950e+00
-3.44487101e-01 2.81612784e-01 5.22583604e-01 2.00236365e-01
6.77393019e-01 3.54181200e-01 6.80853069e-01 5.96036911e-01
2.16297150e-01 -1.35989475e+00 -5.52228153e-01 1.24534523e+00
8.49980533e-01 -6.99566007e-01 -1.28676042e-01 4.76949587e-02
-4.08469468e-01 6.11462176e-01 2.99585521e-01 -1.70375854e-01
7.62324035e-01 4.70027961e-02 -6.31532967e-02 -4.66532677e-01
-5.97376406e-01 -3.06718528e-01 2.78927058e-01 4.61534232e-01
4.87880260e-01 -1.81860402e-01 -3.69122386e-01 8.25407445e-01
-6.58127964e-01 -6.05228804e-02 3.55254501e-01 9.89745080e-01
-3.23846817e-01 -6.20047390e-01 -3.47022057e-01 4.59662944e-01
-8.01225126e-01 -6.23185575e-01 -2.72667974e-01 7.29270220e-01
3.59505229e-02 7.03351736e-01 5.53932786e-01 -1.66723922e-01
4.82755512e-01 4.30231124e-01 -4.27691132e-01 -7.54814744e-01
-1.00120676e+00 -6.08845294e-01 5.11857986e-01 -5.47198474e-01
1.47250950e-01 -9.23059940e-01 -1.66485989e+00 -1.81699738e-01
1.19335316e-01 3.94067377e-01 8.66354167e-01 8.61088157e-01
9.57692325e-01 4.71284598e-01 8.63232166e-02 -6.10059917e-01
2.41566941e-01 -1.08565116e+00 -5.08450031e-01 2.48437941e-01
-2.00409461e-02 -5.70834279e-01 -4.88617197e-02 -1.71018720e-01] | [9.3195161819458, 9.070025444030762] |
15a5f77d-100c-4aca-891c-f74edb27564a | tool-flank-wear-prediction-using-high | 2212.13905 | null | https://arxiv.org/abs/2212.13905v1 | https://arxiv.org/pdf/2212.13905v1.pdf | Tool flank wear prediction using high-frequency machine data from industrial edge device | Tool flank wear monitoring can minimize machining downtime costs while increasing productivity and product quality. In some industrial applications, only a limited level of tool wear is allowed to attain necessary tolerances. It may become challenging to monitor a limited level of tool wear in the data collected from the machine due to the other components, such as the flexible vibrations of the machine, dominating the measurement signals. In this study, a tool wear monitoring technique to predict limited levels of tool wear from the spindle motor current and dynamometer measurements is presented. High-frequency spindle motor current data is collected with an industrial edge device while the cutting forces and torque are measured with a rotary dynamometer in drilling tests for a selected number of holes. Feature engineering is conducted to identify the statistical features of the measurement signals that are most sensitive to small changes in tool wear. A neural network based on the long short-term memory (LSTM) architecture is developed to predict tool flank wear from the measured spindle motor current and dynamometer signals. It is demonstrated that the proposed technique predicts tool flank wear with good accuracy and high computational efficiency. The proposed technique can easily be implemented in an industrial edge device as a real-time predictive maintenance application to minimize the costs due to manufacturing downtime and tool underuse or overuse. | ['I. Lazoglu', 'E. Emekli', 'U. Uresin', 'T. Pehlivan', 'G. Burun', 'M. R. Chehrehzad', 'C. Besirova', 'G. Kecibas', 'D. Bilgili'] | 2022-12-12 | null | null | null | null | ['feature-engineering'] | ['methodology'] | [ 2.04693705e-01 -5.78649700e-01 -1.61108553e-01 -7.05744550e-02
1.21775351e-01 1.45878553e-01 -2.17902631e-01 -1.21555917e-01
4.77896743e-02 3.45236808e-01 -7.97105014e-01 1.00609593e-01
-6.11571372e-01 -3.85672301e-01 -3.31803679e-01 -5.09001255e-01
7.56896064e-02 2.21267879e-01 9.13541019e-02 -1.67967767e-01
5.05675435e-01 8.32931101e-01 -2.01845050e+00 1.27868205e-01
5.97665489e-01 1.34103799e+00 9.59199548e-01 5.84422052e-01
3.79126906e-01 8.68257880e-01 -8.62973571e-01 8.48063767e-01
-2.97070947e-02 -2.93821245e-01 -5.18091023e-01 5.87872922e-01
-4.24208164e-01 -3.02515894e-01 -4.98589836e-02 8.41128349e-01
1.51938066e-01 2.89928764e-01 4.26491499e-01 -1.05007160e+00
-4.97141749e-01 -5.73194306e-03 -5.52048564e-01 2.70317733e-01
1.21573634e-01 -1.71557873e-01 1.66469708e-01 -1.08979189e+00
5.25524020e-01 6.93565786e-01 7.06737101e-01 -4.74910326e-02
-9.13657427e-01 -4.43512380e-01 -6.07646167e-01 4.84442145e-01
-9.68212366e-01 -2.04485193e-01 1.15294433e+00 -6.18518353e-01
1.03812695e+00 2.32737809e-01 6.22163951e-01 7.05903530e-01
1.63117993e+00 3.27320158e-01 9.19917464e-01 -4.49768841e-01
2.74922997e-01 5.75105399e-02 3.20183933e-01 5.46769738e-01
-8.60334858e-02 4.59634870e-01 -4.23362553e-01 1.72018260e-01
1.03293073e+00 4.64094818e-01 -2.68679172e-01 -2.44488135e-01
-7.28589058e-01 7.95412838e-01 3.07708859e-01 5.88066578e-01
-5.57092011e-01 -8.62052664e-02 8.25817108e-01 9.23792839e-01
3.87198031e-01 8.14259231e-01 -7.83898175e-01 -5.33576667e-01
-9.75936294e-01 -2.07199931e-01 8.09901118e-01 6.86994195e-01
2.16250449e-01 5.46937823e-01 4.00046587e-01 1.01976097e+00
-2.57013813e-02 3.09776694e-01 8.76388192e-01 -8.28369319e-01
-1.14431992e-01 3.32725555e-01 2.68090487e-01 -1.10172319e+00
-8.29392314e-01 -4.12139833e-01 -6.99567914e-01 5.49566805e-01
-2.31087089e-01 -1.15060516e-01 -6.21199191e-01 9.17780578e-01
-1.04993224e-01 -4.84869838e-01 -4.25852388e-01 9.05810952e-01
2.53303349e-01 6.28243208e-01 -4.91778523e-01 -6.79288507e-01
1.15289354e+00 -8.02398860e-01 -1.59195471e+00 -4.62593257e-01
5.89938581e-01 -9.21400845e-01 1.04626250e+00 7.34328508e-01
-8.43576849e-01 -1.03093624e+00 -1.75039613e+00 1.82615981e-01
-2.85908014e-01 8.65058959e-01 3.33772331e-01 -5.22265919e-02
-3.44449818e-01 1.38292086e+00 -1.27040529e+00 -3.60248871e-02
-2.18885973e-01 3.10968965e-01 -2.33127117e-01 3.03035975e-01
-1.19489324e+00 1.67687559e+00 1.01502098e-01 7.32127845e-01
-4.88612741e-01 -5.71409702e-01 -8.41584802e-01 -3.00176173e-01
2.75891930e-01 -8.96141157e-02 1.48760819e+00 -6.34668469e-01
-2.02447057e+00 1.73275724e-01 7.11056888e-02 -3.91970426e-01
2.07845613e-01 -8.25779915e-01 -7.08690345e-01 -1.28450617e-01
-3.83064225e-02 -2.83042490e-01 1.29931486e+00 -7.67873049e-01
-3.73324364e-01 -4.78899539e-01 -9.34519827e-01 -2.65743881e-01
-7.96871930e-02 6.86559305e-02 3.19388449e-01 -4.17256474e-01
1.59854829e-01 -1.09931314e+00 8.04175511e-02 -1.46078259e-01
-2.33285159e-01 4.29127216e-02 1.48853576e+00 -8.78736258e-01
1.15739810e+00 -2.25957274e+00 -6.86972290e-02 1.63855493e-01
-4.41418022e-01 -5.02866842e-02 3.50538105e-01 6.51992083e-01
4.18654131e-03 -8.24323356e-01 -3.41466554e-02 1.15422986e-01
-4.89032477e-01 2.82097280e-01 -2.52954941e-02 8.11993718e-01
2.36369193e-01 4.29254442e-01 -5.58422446e-01 2.65467137e-01
6.20035231e-01 1.71163410e-01 3.69863063e-01 2.65634388e-01
2.83971936e-01 1.07908562e-01 -1.06462501e-01 7.09623694e-01
1.25278920e-01 1.24455519e-01 -2.95462847e-01 -3.52507144e-01
-2.98943818e-01 -1.00934424e-01 -9.13237274e-01 1.60778809e+00
-1.28428769e+00 9.62226748e-01 4.97271568e-01 -7.84898281e-01
1.64151025e+00 3.96566600e-01 2.33950078e-01 -7.77359605e-01
5.64396262e-01 7.47250438e-01 1.98130265e-01 -8.28272760e-01
4.43475157e-01 -1.91908613e-01 -1.24777826e-02 1.23001061e-01
3.26728933e-02 -8.22208047e-01 -6.37578741e-02 -7.82641888e-01
8.77539814e-01 5.21850884e-02 1.78559184e-01 -4.48007554e-01
3.04635227e-01 -1.27927467e-01 5.08633614e-01 5.73813915e-02
1.35301188e-01 -3.68068297e-03 -9.84782502e-02 -4.55091685e-01
-1.02996063e+00 -5.84051490e-01 -3.02731872e-01 8.27172220e-01
9.48481336e-02 2.23167658e-01 -2.76161820e-01 2.22913146e-01
2.20671102e-01 6.61175251e-01 -5.90250790e-01 -9.25995588e-01
-5.46927989e-01 -2.77152002e-01 -1.46151423e-01 8.56515467e-01
5.81369437e-02 -1.21320236e+00 -1.52709889e+00 6.60901666e-01
5.72785139e-01 -6.95515215e-01 -4.74854320e-01 9.60503995e-01
-1.34344256e+00 -1.05175698e+00 -1.46345899e-01 -9.42555785e-01
4.66062516e-01 -2.15127096e-02 4.32231933e-01 -2.02777714e-01
-7.90294647e-01 -1.66883349e-01 -6.53058514e-02 -7.60454357e-01
-4.24944013e-01 -1.15720958e-01 4.95518535e-01 -3.83920580e-01
3.39023530e-01 -3.74784231e-01 -5.10977358e-02 5.25797307e-01
-3.32780361e-01 -1.89762205e-01 8.07701588e-01 1.44021487e+00
4.40148354e-01 8.88998151e-01 9.91565466e-01 -6.16560102e-01
1.03045070e+00 -3.88671160e-01 -5.35895109e-01 -3.89398336e-01
-9.14009869e-01 -3.40661615e-01 1.08216906e+00 -6.96331143e-01
-1.16043627e+00 -6.28832132e-02 2.92832911e-01 -8.32395196e-01
1.27259508e-01 9.56737041e-01 -2.43811589e-03 7.11227804e-02
3.68551791e-01 -3.74447331e-02 7.69777477e-01 -6.94123983e-01
-2.65822679e-01 1.05839670e+00 8.93471599e-01 -3.66663970e-02
3.14620614e-01 1.53757492e-02 1.78783357e-01 -1.34163094e+00
-4.48570609e-01 -3.41427267e-01 -6.44840240e-01 -4.98078108e-01
3.19875717e-01 -5.74285805e-01 -5.74402750e-01 7.75052786e-01
-9.17139232e-01 -2.40686119e-01 -3.29521209e-01 8.03104401e-01
-7.79267788e-01 7.82015696e-02 -1.35918033e+00 -8.67440283e-01
-7.25081325e-01 -1.13942325e+00 7.14320362e-01 1.22894183e-01
-8.06195617e-01 -6.08754277e-01 -4.06888187e-01 -3.86399180e-02
5.82644582e-01 4.73059356e-01 7.40436733e-01 -4.54523005e-02
3.38464737e-01 -8.62203300e-01 4.13484663e-01 8.10209036e-01
9.30299759e-01 1.07908189e-01 -6.15487635e-01 -4.33464646e-01
1.20942366e+00 -7.28912055e-02 2.25702330e-01 7.03857303e-01
8.74349654e-01 1.47867158e-01 -3.40740502e-01 1.57296851e-01
1.27882588e+00 1.01123023e+00 1.10703819e-01 4.16255683e-01
4.30081755e-01 6.67412341e-01 1.43997622e+00 4.17874247e-01
-7.93792009e-01 6.26262486e-01 1.81933254e-01 4.65463512e-02
3.10560703e-01 -9.82416514e-03 5.69981098e-01 1.40522730e+00
1.43640116e-01 4.51167822e-01 -4.60350901e-01 5.56414247e-01
-1.32867479e+00 -3.98360789e-01 -3.68421495e-01 2.00518489e+00
7.78687358e-01 4.23782289e-01 -4.12572861e-01 1.07303524e+00
7.58037686e-01 -2.31245264e-01 -7.59805620e-01 -1.23066866e+00
4.61123377e-01 1.91938326e-01 4.04675990e-01 2.94744879e-01
-8.67999136e-01 1.10406578e-01 5.35903740e+00 6.98753297e-01
-1.49152398e+00 -1.27821892e-01 -2.17588339e-02 -4.12196964e-01
6.62907898e-01 -4.62424606e-01 -2.99778849e-01 8.49291801e-01
1.42125535e+00 -2.82885712e-02 2.56025612e-01 1.23285306e+00
5.63682377e-01 -5.16123354e-01 -1.27936876e+00 8.50259304e-01
-1.22276813e-01 -9.64545250e-01 -8.60385299e-01 -1.99191093e-01
3.45558167e-01 -2.69555628e-01 3.05555109e-02 -1.98157933e-02
-6.56938910e-01 -8.00665379e-01 6.39439344e-01 7.51538515e-01
7.80610979e-01 -1.00972331e+00 1.16907656e+00 5.06948709e-01
-9.74118829e-01 -6.25658453e-01 -3.54086578e-01 -7.42603481e-01
4.55211818e-01 1.06361604e+00 -8.05775106e-01 4.52422708e-01
6.65273011e-01 7.72794902e-01 -1.04607798e-01 5.85925102e-01
8.14437568e-02 3.29242259e-01 -3.68914872e-01 -2.28793249e-01
-4.75945063e-02 -3.12378496e-01 3.77424926e-01 6.23710692e-01
6.68818057e-01 -5.35168648e-01 1.41148753e-02 7.47506976e-01
5.07997036e-01 -2.08745256e-01 -8.77768159e-01 -6.16872758e-02
6.86669827e-01 1.01293373e+00 -5.86507976e-01 3.36813480e-01
-1.86295539e-01 8.70340288e-01 -4.22952563e-01 -4.36032228e-02
-5.13893425e-01 -1.17490757e+00 3.45422775e-01 2.01012537e-01
-6.44293725e-02 -4.54232156e-01 -4.55398023e-01 -1.80351362e-02
3.45571578e-01 -4.74071056e-01 -2.97635198e-01 -9.23534393e-01
-1.15860653e+00 4.06070203e-01 -2.40244821e-01 -1.48455524e+00
-5.27152777e-01 -7.74958014e-01 -7.55302906e-01 9.66951787e-01
-6.79430008e-01 -5.19096375e-01 -1.58074424e-01 -3.50771099e-02
1.28343296e+00 -1.39032736e-01 7.36687839e-01 6.96852133e-02
-6.44954085e-01 -1.37781069e-01 3.32179576e-01 -5.40565252e-01
4.82103914e-01 -9.18578327e-01 5.44966385e-02 5.31318367e-01
-4.25935209e-01 5.14735758e-01 1.20531380e+00 -8.89649928e-01
-1.69163334e+00 -1.01074505e+00 6.54063523e-01 -3.44959535e-02
7.85999775e-01 -1.79553926e-01 -1.16714597e+00 3.24876934e-01
-4.77690771e-02 3.14871259e-02 2.09979996e-01 -4.03091371e-01
8.33297789e-01 -1.27290756e-01 -9.67397094e-01 5.35518350e-03
3.46457183e-01 -7.27547705e-01 -9.47164834e-01 1.04259014e-01
2.28901476e-01 -5.71469069e-01 -1.33494079e+00 6.08766437e-01
6.47866547e-01 -4.37038571e-01 3.44326049e-01 1.90661121e-02
3.99501175e-01 -2.06679389e-01 4.00148988e-01 -1.78883827e+00
-4.01216716e-01 -5.62651992e-01 -6.02862597e-01 9.11444247e-01
-4.82912250e-02 -5.61960399e-01 4.64001209e-01 1.80178061e-01
-5.17520308e-01 -1.34483659e+00 -1.03923953e+00 -1.15896916e+00
-4.21045989e-01 -2.03190088e-01 1.42948896e-01 5.03384352e-01
4.26146537e-01 3.86386603e-01 -4.78558928e-01 5.57513982e-02
2.06299834e-02 1.46283507e-01 3.31002861e-01 -1.68729317e+00
4.36533317e-02 8.62015933e-02 -3.70739996e-01 -4.11282092e-01
1.91504374e-01 -3.45062256e-01 6.47211850e-01 -1.19145548e+00
-4.28617269e-01 -4.00846712e-02 -2.41269022e-01 -1.31327748e-01
2.63351083e-01 -3.12340319e-01 -1.52359486e-01 9.94679481e-02
4.95717615e-01 7.24260688e-01 1.24957490e+00 -9.77321267e-02
-4.92831379e-01 5.35953939e-01 3.72439295e-01 7.69351602e-01
7.17249870e-01 -2.31688961e-01 -6.35252297e-01 -2.26766184e-01
-5.55365421e-02 5.98249435e-01 7.71576017e-02 -1.30323160e+00
2.98124433e-01 2.45051876e-01 6.86640799e-01 -8.37292671e-01
4.09284264e-01 -1.13868558e+00 6.18936956e-01 9.68042552e-01
4.05969843e-03 2.56201267e-01 3.91629100e-01 4.40636814e-01
-6.53678775e-01 -2.88367778e-01 8.34230781e-01 2.59242326e-01
-4.66124982e-01 -2.73229629e-01 -8.86848867e-01 -9.57106590e-01
9.34800565e-01 -6.16666317e-01 7.05394447e-02 -5.18108606e-02
-6.99737549e-01 7.05831125e-02 2.67007321e-01 9.35999095e-01
7.62542844e-01 -1.23660016e+00 -1.96412444e-01 6.44364893e-01
-2.15949029e-01 -2.31201440e-01 5.94534457e-01 1.04448771e+00
-4.20520574e-01 6.36001050e-01 -6.95597887e-01 -5.76004684e-01
-1.10923684e+00 6.41992450e-01 3.88552517e-01 1.04076639e-01
-8.49978507e-01 7.94170737e-01 -6.01707160e-01 1.19399495e-01
-1.24603920e-01 -7.88322508e-01 -1.31408647e-01 2.50775784e-01
2.57842779e-01 9.10121620e-01 8.38686824e-01 -4.12536979e-01
-2.03710929e-01 5.24919629e-01 3.98877449e-02 5.03557086e-01
1.35600126e+00 -1.30443633e-01 -3.23320143e-02 1.37649488e+00
1.20448267e+00 -4.70893770e-01 -1.31068099e+00 3.64974707e-01
3.20351005e-01 -3.03685457e-01 5.36174834e-01 -5.92927337e-01
-1.11966085e+00 6.15885556e-01 9.88565803e-01 1.98924020e-01
1.18740702e+00 -4.57514793e-01 1.06776571e+00 4.60827261e-01
8.68512511e-01 -1.81004798e+00 -1.30106211e-01 3.54250818e-01
1.37581384e+00 -7.21515954e-01 4.45874743e-02 -7.28203654e-02
-4.26580667e-01 1.54343832e+00 5.60930967e-01 -4.96045858e-01
8.35470140e-01 8.06026220e-01 1.00195006e-01 -3.33034605e-01
-8.22812021e-01 7.53442585e-01 6.15478903e-02 3.76811415e-01
3.14499319e-01 6.79054111e-02 -3.03513527e-01 8.41311634e-01
-1.74108043e-01 4.70225066e-01 7.43270218e-01 1.42587256e+00
-5.38294852e-01 -6.04303777e-01 -5.86733043e-01 1.00525916e+00
-6.74696565e-01 6.61496878e-01 8.14844444e-02 9.53847170e-01
1.73201710e-02 1.15374494e+00 3.24939221e-01 -6.05963409e-01
6.58349812e-01 1.26143754e-01 4.11689162e-01 -6.97602808e-01
-3.99840176e-01 5.11226729e-02 4.54634279e-02 -7.15371490e-01
1.00808583e-01 -5.99691570e-01 -1.32615387e+00 2.48104751e-01
-1.04570234e+00 1.33229256e-01 8.64010155e-01 9.13091958e-01
2.87450790e-01 1.16122818e+00 8.12541544e-01 -1.16988814e+00
-7.11517751e-01 -1.71114826e+00 -1.46518886e+00 -2.19544936e-02
4.95254457e-01 -1.45955896e+00 -6.06460452e-01 -3.62765454e-02] | [6.8292107582092285, 2.318666696548462] |
bdc8e31d-ec1e-4695-87da-72fb471b3223 | end-to-end-training-of-neural-retrievers-for | 2101.00408 | null | https://arxiv.org/abs/2101.00408v2 | https://arxiv.org/pdf/2101.00408v2.pdf | End-to-End Training of Neural Retrievers for Open-Domain Question Answering | Recent work on training neural retrievers for open-domain question answering (OpenQA) has employed both supervised and unsupervised approaches. However, it remains unclear how unsupervised and supervised methods can be used most effectively for neural retrievers. In this work, we systematically study retriever pre-training. We first propose an approach of unsupervised pre-training with the Inverse Cloze Task and masked salient spans, followed by supervised finetuning using question-context pairs. This approach leads to absolute gains of 2+ points over the previous best result in the top-20 retrieval accuracy on Natural Questions and TriviaQA datasets. We also explore two approaches for end-to-end supervised training of the reader and retriever components in OpenQA models. In the first approach, the reader considers each retrieved document separately while in the second approach, the reader considers all the retrieved documents together. Our experiments demonstrate the effectiveness of these approaches as we obtain new state-of-the-art results. On the Natural Questions dataset, we obtain a top-20 retrieval accuracy of 84, an improvement of 5 points over the recent DPR model. In addition, we achieve good results on answer extraction, outperforming recent models like REALM and RAG by 3+ points. We further scale up end-to-end training to large models and show consistent gains in performance over smaller models. | ['Bryan Catanzaro', 'William L Hamilton', 'Wei Ping', 'Neel Kant', 'Mohammad Shoeybi', 'Mostofa Patwary', 'Devendra Singh Sachan'] | 2021-01-02 | null | https://aclanthology.org/2021.acl-long.519 | https://aclanthology.org/2021.acl-long.519.pdf | acl-2021-5 | ['triviaqa'] | ['miscellaneous'] | [ 2.12146193e-01 2.34302863e-01 4.26130323e-03 -2.37388030e-01
-1.75832176e+00 -7.91374624e-01 5.73839605e-01 3.53799343e-01
-6.65934920e-01 5.62532067e-01 3.39503884e-01 -5.11036515e-01
-4.82908100e-01 -6.95165634e-01 -7.86662459e-01 -2.64240921e-01
1.49779767e-01 8.85630071e-01 6.15523636e-01 -7.01525092e-01
3.57410878e-01 -1.61676090e-02 -1.54168248e+00 6.99549377e-01
9.51375127e-01 1.10282993e+00 -4.00908776e-02 9.60943520e-01
-2.25492597e-01 1.20637774e+00 -6.63308978e-01 -6.24639332e-01
8.67423266e-02 -2.68475354e-01 -1.34175837e+00 -4.69964713e-01
9.47628617e-01 -5.78636587e-01 -3.45030814e-01 3.56989592e-01
7.57800102e-01 2.38221392e-01 7.04474032e-01 -4.22715485e-01
-1.07403231e+00 7.37791657e-01 -3.84198904e-01 4.19386894e-01
6.28206611e-01 4.47899103e-02 1.65363777e+00 -9.19470489e-01
6.22141957e-01 1.01588714e+00 3.66624475e-01 5.61480820e-01
-1.24347889e+00 -4.21106815e-01 -6.00499250e-02 2.26011902e-01
-1.12729025e+00 -6.03451014e-01 5.06145477e-01 1.05386013e-02
1.41052628e+00 2.95343369e-01 5.02747903e-03 6.80904508e-01
-3.45128775e-01 1.06053150e+00 9.78251040e-01 -8.38813305e-01
3.60805355e-02 1.68188885e-02 7.13514984e-01 4.88053501e-01
-2.47921236e-02 2.69115623e-02 -4.29111332e-01 -2.91627854e-01
1.53604329e-01 -3.17512631e-01 -2.69940168e-01 -1.42901778e-01
-1.00400972e+00 8.93249631e-01 7.36486375e-01 3.31238747e-01
-3.81705105e-01 1.47005081e-01 2.50942200e-01 7.41629004e-01
5.82016051e-01 1.03462851e+00 -6.38346016e-01 2.51105074e-02
-1.15912032e+00 7.07611978e-01 1.03313923e+00 7.23792315e-01
6.53913558e-01 -5.06142855e-01 -7.06907749e-01 1.23107815e+00
1.42671123e-01 5.79752862e-01 4.12533641e-01 -1.04547906e+00
6.36871994e-01 5.56702673e-01 1.45800531e-01 -8.00626755e-01
-2.37182006e-01 -5.36193728e-01 -2.12404132e-01 -4.64100808e-01
6.77363336e-01 -6.24929294e-02 -9.94707823e-01 1.55919504e+00
1.12991914e-01 -3.98413211e-01 2.63612390e-01 7.53099978e-01
1.15788937e+00 7.43692875e-01 1.31448179e-01 -2.25274991e-02
1.58351243e+00 -1.24034655e+00 -6.83642626e-01 -2.59547174e-01
8.88357878e-01 -7.11638391e-01 1.20946395e+00 3.88448089e-01
-1.35704100e+00 -3.64258379e-01 -9.19282615e-01 -6.26933992e-01
-5.39155662e-01 1.58387855e-01 4.00994778e-01 4.13701445e-01
-1.17777991e+00 4.10395205e-01 -2.40286767e-01 -2.61975676e-01
3.57593030e-01 3.93908203e-01 7.48596862e-02 -3.88553500e-01
-1.40885234e+00 9.76198435e-01 3.20859283e-01 -2.07729772e-01
-8.33498001e-01 -7.51543224e-01 -3.26530755e-01 2.66848356e-01
6.12685502e-01 -8.06472600e-01 1.68858814e+00 -5.71681678e-01
-1.21346796e+00 1.08829498e+00 -9.38419998e-02 -8.02283049e-01
2.94533446e-02 -6.43615901e-01 -1.83725983e-01 4.99437481e-01
-1.30374357e-02 9.46993291e-01 6.37676954e-01 -1.11982000e+00
-4.74614531e-01 -3.73341560e-01 4.86654609e-01 3.65493536e-01
-4.70094442e-01 2.47803718e-01 -4.53797787e-01 -4.85243589e-01
6.04003221e-02 -6.95196033e-01 6.27951920e-02 -2.85639614e-01
-1.61558613e-02 -8.08304608e-01 3.14217061e-01 -9.09018457e-01
1.36550939e+00 -1.86499429e+00 1.11839563e-01 -4.02444005e-02
3.67406338e-01 2.77662307e-01 -5.71862102e-01 4.20395046e-01
1.33405298e-01 2.25501489e-02 -1.80513099e-01 -2.14938909e-01
1.42672226e-01 -1.02083990e-02 -7.78631985e-01 -1.23226792e-01
3.30832899e-01 1.30458355e+00 -8.93423319e-01 -4.99634683e-01
-5.03654301e-01 1.15948536e-01 -6.42820835e-01 4.48336542e-01
-7.11998343e-01 -5.28999791e-02 -4.68837082e-01 5.93444347e-01
2.05034554e-01 -3.94496292e-01 -2.31558904e-01 1.89282700e-01
4.86297727e-01 9.59505975e-01 -6.13296926e-01 1.75550652e+00
-4.68207270e-01 5.50601065e-01 -3.92880067e-02 -5.72826147e-01
9.75053370e-01 3.79602790e-01 3.00061181e-02 -1.13702261e+00
4.55748243e-03 5.57695985e-01 -9.06974450e-02 -5.06457746e-01
9.02579308e-01 -4.32312451e-02 -3.25649865e-02 7.48437107e-01
3.85057211e-01 -1.95103794e-01 3.39886755e-01 6.42669261e-01
1.33207083e+00 5.44936769e-03 -1.68821931e-01 -2.05263689e-01
4.36832905e-01 1.38982728e-01 -3.74112815e-01 1.15528870e+00
8.97789001e-02 7.15167344e-01 4.38060254e-01 -1.06004626e-01
-7.53239214e-01 -9.63273823e-01 4.90341112e-02 1.80126500e+00
-1.75868154e-01 -3.42326134e-01 -8.15914690e-01 -8.40615869e-01
4.03353684e-02 8.64399374e-01 -5.53992748e-01 -2.25247294e-01
-8.75556350e-01 -4.38990802e-01 8.19945157e-01 5.43017149e-01
2.61736572e-01 -1.30344367e+00 -4.51175421e-01 8.99928883e-02
-5.10340452e-01 -1.01076794e+00 -1.88338786e-01 2.64186293e-01
-1.13864934e+00 -7.45550752e-01 -1.07033193e+00 -9.04257238e-01
2.59089082e-01 3.03109527e-01 1.75905824e+00 4.02025491e-01
7.12244138e-02 4.63024706e-01 -6.52129889e-01 -2.24187240e-01
-1.59651339e-01 8.13886940e-01 -4.66770083e-01 -4.33417290e-01
5.73829532e-01 -2.86754072e-01 -7.34729469e-01 2.05507889e-01
-1.00102222e+00 -3.78882229e-01 6.60827756e-01 8.10558617e-01
3.76366824e-01 -7.03783631e-01 1.06541443e+00 -9.37881768e-01
1.08537972e+00 -5.13262808e-01 -3.20194900e-01 4.86510456e-01
-6.93158627e-01 2.44161859e-01 3.59230340e-01 -4.09997910e-01
-9.23417985e-01 -4.40238029e-01 -3.28507155e-01 -1.79201901e-01
1.78305749e-02 6.36941493e-01 3.01390707e-01 1.64617240e-01
1.25170898e+00 5.82670234e-02 -1.85478717e-01 -6.08631611e-01
7.18618035e-01 8.99556577e-01 4.11630481e-01 -7.18099654e-01
7.04717577e-01 3.37364972e-02 -6.57090366e-01 -5.16450226e-01
-1.36104512e+00 -7.59098947e-01 -3.29742283e-01 9.88123268e-02
6.91659451e-01 -8.90717626e-01 -3.56348604e-01 1.12175994e-01
-1.13561094e+00 -3.06269944e-01 -4.76272643e-01 -1.48224356e-02
-4.71321851e-01 1.28672779e-01 -7.50864685e-01 -6.96526706e-01
-8.32968354e-01 -7.86425412e-01 1.25214720e+00 1.13708034e-01
-3.61144066e-01 -8.18781674e-01 3.08983535e-01 1.01777935e+00
7.24627972e-01 -3.65112066e-01 1.10246348e+00 -1.38076103e+00
-7.29557753e-01 -2.34691188e-01 -3.15764189e-01 2.86926538e-01
-3.39621812e-01 -6.72815979e-01 -1.12052929e+00 -2.37955809e-01
-1.36075974e-01 -9.76439059e-01 1.38345683e+00 -5.01921140e-02
8.52096081e-01 -2.51496643e-01 -1.16294019e-01 2.65953671e-02
1.23804653e+00 -2.05793619e-01 8.10997605e-01 4.16264266e-01
4.30661649e-01 9.25251663e-01 6.18767202e-01 -2.41289139e-01
4.88068521e-01 3.76118183e-01 2.43894979e-01 1.76221609e-01
-2.20283136e-01 -2.74673730e-01 1.90450281e-01 7.37802386e-01
1.45988524e-01 -2.95425653e-01 -9.34705555e-01 1.01312864e+00
-1.55200279e+00 -7.66448975e-01 4.21442389e-02 2.19425225e+00
1.17559838e+00 1.68800831e-01 1.34149417e-01 1.80197824e-02
2.09158957e-01 2.96206683e-01 -4.33396012e-01 -3.42527181e-01
-6.29788414e-02 7.93842554e-01 1.95470974e-01 4.46426511e-01
-8.66171598e-01 1.12856722e+00 6.54634905e+00 9.37109590e-01
-5.12093186e-01 8.64224136e-02 6.08265758e-01 -3.16993386e-01
-4.22186971e-01 1.30453959e-01 -9.46044266e-01 -1.15765557e-01
1.28114021e+00 2.39684001e-01 5.11977077e-01 5.67262173e-01
-5.48076749e-01 -1.78527102e-01 -1.25436604e+00 5.80480337e-01
4.32728976e-01 -1.11371362e+00 2.86801279e-01 -2.72748977e-01
7.55337417e-01 1.43795446e-01 -6.65253997e-02 9.27527905e-01
2.03934550e-01 -1.10081196e+00 4.39634383e-01 5.19113123e-01
4.65854466e-01 -6.36582136e-01 7.07375348e-01 4.93312508e-01
-5.60369968e-01 -1.98061347e-01 -4.04268295e-01 6.11601472e-02
1.30567618e-03 1.75990745e-01 -7.65611470e-01 4.55879182e-01
7.00314462e-01 2.78757960e-01 -1.01727247e+00 8.98169339e-01
-2.85899132e-01 8.31401885e-01 -3.31292480e-01 -5.02488136e-01
2.89126545e-01 2.93229282e-01 4.18923110e-01 1.04338300e+00
-2.52204657e-01 3.01152259e-01 -2.59161353e-01 6.93707228e-01
-5.19044995e-01 3.18323225e-01 -3.46217304e-01 -5.65673076e-02
3.26856762e-01 1.01304948e+00 -2.50128657e-01 -4.76260334e-01
-2.00729430e-01 6.60751164e-01 7.95870185e-01 4.27525669e-01
-5.30517578e-01 -7.54227042e-01 -1.64186329e-01 1.98280647e-01
6.35345340e-01 1.27783298e-01 -3.48304696e-02 -9.91422296e-01
3.10351402e-01 -1.40969574e+00 7.58260608e-01 -8.32039535e-01
-1.41054738e+00 5.84710240e-01 1.74597930e-02 -6.92671359e-01
-4.75026160e-01 -6.20261729e-01 -1.73291028e-01 7.94828057e-01
-1.78343999e+00 -1.09926856e+00 7.07006231e-02 3.93358022e-01
6.28832221e-01 -3.05612553e-02 9.14967358e-01 3.19959044e-01
5.03742099e-02 8.30091655e-01 2.62021631e-01 2.93340832e-01
9.62037921e-01 -1.31413138e+00 2.30492130e-01 4.55487877e-01
4.52829599e-01 9.45253134e-01 4.72414076e-01 -3.33428621e-01
-1.45148778e+00 -6.05824590e-01 1.18127263e+00 -1.02891314e+00
7.64964044e-01 -2.28022948e-01 -1.17311454e+00 5.99056602e-01
6.22356594e-01 -2.81307042e-01 6.23457015e-01 6.40439987e-01
-7.22222209e-01 -2.11584210e-01 -8.69350016e-01 4.69741762e-01
6.67390287e-01 -7.04976201e-01 -1.29709435e+00 4.72335130e-01
1.18059015e+00 -4.45613474e-01 -8.68128538e-01 5.53241670e-01
5.45167446e-01 -7.17023075e-01 1.05914593e+00 -8.23960960e-01
6.80259407e-01 3.71396281e-02 -3.00572038e-01 -1.02353024e+00
-4.36290503e-02 -3.83017868e-01 -5.36529899e-01 1.17617488e+00
8.13739002e-01 -4.68415886e-01 5.14001369e-01 4.19927657e-01
6.62501752e-02 -1.04969823e+00 -7.76603162e-01 -4.42466110e-01
7.09155619e-01 -2.20712215e-01 3.48881096e-01 5.60689270e-01
-3.48135643e-02 9.63062346e-01 8.70057121e-02 -2.28031933e-01
2.04316869e-01 2.53170520e-01 6.84457421e-01 -1.07848096e+00
-4.79345143e-01 -4.71279830e-01 2.29067221e-01 -1.60520089e+00
4.05744836e-02 -9.36019957e-01 1.01989329e-01 -1.78204083e+00
3.26364130e-01 -3.30550909e-01 -3.56337547e-01 3.16358149e-01
-4.38893139e-01 2.21284166e-01 1.26721084e-01 2.71422952e-01
-1.26730633e+00 4.28489953e-01 1.10630822e+00 -2.20319390e-01
-2.08643600e-01 -1.82979465e-01 -1.09056532e+00 2.67864764e-01
6.79607093e-01 -5.17330766e-01 -3.98034751e-01 -9.18789089e-01
5.69447339e-01 1.64517820e-01 3.48560065e-01 -9.13143754e-01
4.10138577e-01 5.05373359e-01 8.86586234e-02 -7.86776900e-01
3.38718086e-01 -5.24795711e-01 -8.19003284e-01 1.17561676e-01
-9.52984393e-01 -3.73080629e-03 3.03155988e-01 5.31291962e-01
-2.65194356e-01 -4.85796481e-01 4.48937595e-01 -1.74698412e-01
-2.57310987e-01 -1.61000371e-01 -1.15890980e-01 7.29339123e-01
3.23109120e-01 1.40687644e-01 -7.45543778e-01 -6.30823493e-01
-4.75779891e-01 5.89071393e-01 -1.45289190e-02 5.39613605e-01
6.01016879e-01 -9.23491180e-01 -1.04720807e+00 -2.89571315e-01
3.77220541e-01 2.13190224e-02 7.88021758e-02 6.76753998e-01
-4.12136048e-01 8.35035861e-01 4.08329993e-01 -4.98705745e-01
-1.11308026e+00 5.62359154e-01 2.86834002e-01 -9.44507420e-01
-7.86522627e-02 9.32148814e-01 -1.38775736e-01 -8.23109031e-01
4.57896084e-01 -1.89090863e-01 -5.29836833e-01 3.05768043e-01
6.40082300e-01 3.31842393e-01 4.46620017e-01 -3.48333389e-01
9.69438069e-03 4.98312533e-01 -6.72696173e-01 -3.61536145e-01
1.24108553e+00 6.81649372e-02 -2.69260615e-01 2.77432650e-01
1.38085365e+00 2.07743067e-02 -5.28640866e-01 -6.32196903e-01
3.18048149e-01 -8.83790180e-02 2.20931508e-02 -1.29046893e+00
-6.86541915e-01 8.44535053e-01 4.85165536e-01 2.63440192e-01
1.26038265e+00 3.75284046e-01 9.81799126e-01 1.04984856e+00
1.72209606e-01 -8.67397904e-01 2.92721689e-01 8.87943685e-01
9.93032575e-01 -1.18313813e+00 -4.85828891e-02 9.58424136e-02
-4.16788012e-01 7.28331149e-01 5.29638529e-01 -3.45974833e-01
2.90338635e-01 -4.20782179e-01 1.01771168e-01 -5.18079698e-01
-1.06905127e+00 -3.93277735e-01 6.66734219e-01 9.32666510e-02
5.69860816e-01 -3.25176686e-01 -2.89680451e-01 6.87134445e-01
-3.76438588e-01 -8.26885924e-02 3.89198177e-02 1.08227944e+00
-6.95794880e-01 -8.58613789e-01 -3.24932665e-01 6.78870440e-01
-7.89638162e-01 -5.75061083e-01 -6.31150544e-01 6.80126309e-01
-5.21477282e-01 1.25023830e+00 -5.42457588e-02 -1.11974917e-01
6.17506623e-01 5.26295543e-01 5.24304450e-01 -6.43775284e-01
-1.14425206e+00 -1.05835222e-01 4.30259526e-01 -3.35076720e-01
-2.35209316e-01 -3.19048822e-01 -1.04750776e+00 1.46422803e-01
-8.09898317e-01 4.75565434e-01 2.95328677e-01 9.51830506e-01
5.25839210e-01 4.01674151e-01 3.98757249e-01 -1.93012029e-01
-1.06191933e+00 -1.36791778e+00 -5.91947809e-02 2.99418390e-01
4.84973401e-01 -2.73799479e-01 -5.15390456e-01 -1.47001043e-01] | [11.331021308898926, 7.952814102172852] |
9fffca76-78a1-4464-94ec-193c1000d91d | on-the-effectiveness-of-image-manipulation | 2304.09414 | null | https://arxiv.org/abs/2304.09414v1 | https://arxiv.org/pdf/2304.09414v1.pdf | On the Effectiveness of Image Manipulation Detection in the Age of Social Media | Image manipulation detection algorithms designed to identify local anomalies often rely on the manipulated regions being ``sufficiently'' different from the rest of the non-tampered regions in the image. However, such anomalies might not be easily identifiable in high-quality manipulations, and their use is often based on the assumption that certain image phenomena are associated with the use of specific editing tools. This makes the task of manipulation detection hard in and of itself, with state-of-the-art detectors only being able to detect a limited number of manipulation types. More importantly, in cases where the anomaly assumption does not hold, the detection of false positives in otherwise non-manipulated images becomes a serious problem. To understand the current state of manipulation detection, we present an in-depth analysis of deep learning-based and learning-free methods, assessing their performance on different benchmark datasets containing tampered and non-tampered samples. We provide a comprehensive study of their suitability for detecting different manipulations as well as their robustness when presented with non-tampered data. Furthermore, we propose a novel deep learning-based pre-processing technique that accentuates the anomalies present in manipulated regions to make them more identifiable by a variety of manipulation detection methods. To this end, we introduce an anomaly enhancement loss that, when used with a residual architecture, improves the performance of different detection algorithms with a minimal introduction of false positives on the non-manipulated data. Lastly, we introduce an open-source manipulation detection toolkit comprising a number of standard detection algorithms. | ['Walter J. Scheirer', 'Jason Schlessman', 'Grant Jensen', 'Daniel Moreira', 'Priscila Saboia', 'Rosaura G. VidalMata'] | 2023-04-19 | null | null | null | null | ['image-manipulation-detection', 'image-manipulation'] | ['computer-vision', 'computer-vision'] | [ 6.93598807e-01 -4.51875597e-01 2.32997239e-01 3.52157503e-02
-4.31648254e-01 -5.63082457e-01 7.74851024e-01 6.96594715e-01
-1.76176742e-01 2.32906397e-02 -3.73225778e-01 -1.47212774e-01
4.98954952e-02 -7.77311504e-01 -8.86005700e-01 -7.94913471e-01
-3.33067119e-01 -9.44742234e-04 4.58684921e-01 -2.49419108e-01
5.02218723e-01 7.50840008e-01 -2.05736065e+00 6.48380995e-01
5.24896622e-01 1.17911589e+00 -1.36665821e-01 8.35841954e-01
6.65420815e-02 4.65408534e-01 -1.00554121e+00 -2.09865555e-01
4.64854300e-01 -2.74272323e-01 -3.25558096e-01 2.35109031e-01
9.14060593e-01 -4.84169722e-01 -2.52785474e-01 1.18080783e+00
2.23252892e-01 -1.04062399e-02 6.83412135e-01 -1.32484663e+00
-4.35418487e-01 2.68988997e-01 -6.98954821e-01 6.39658749e-01
3.67300570e-01 5.30618370e-01 5.89456558e-01 -8.21758270e-01
6.81474090e-01 1.08044815e+00 5.28319299e-01 9.18751881e-02
-1.25511515e+00 -5.30285776e-01 7.05117285e-02 2.64033496e-01
-1.14329207e+00 -4.78495598e-01 8.19464445e-01 -6.13353789e-01
8.99299800e-01 3.30471516e-01 2.65810072e-01 1.42307937e+00
3.57618272e-01 6.53971672e-01 8.54185522e-01 -6.08912945e-01
5.58477454e-02 -4.54778858e-02 1.15914166e-01 6.82153881e-01
4.93452728e-01 8.80537257e-02 -3.89713287e-01 -9.92036238e-02
3.99307013e-01 -1.02264069e-01 -3.69278163e-01 -5.76989353e-01
-1.38233387e+00 6.00539327e-01 3.17287952e-01 7.02095509e-01
-5.02260327e-01 -1.52904410e-02 8.35128486e-01 5.25397837e-01
3.01427752e-01 7.95174599e-01 -2.38076806e-01 -7.15368763e-02
-1.08479464e+00 7.99348131e-02 4.33714122e-01 7.23857462e-01
6.80836499e-01 1.37301341e-01 -3.44874233e-01 3.13547045e-01
-1.82293206e-01 2.18510672e-01 3.38271797e-01 -4.10010636e-01
2.63518631e-01 8.77169132e-01 2.47615762e-02 -1.49449563e+00
-3.16721141e-01 -3.23459774e-01 -9.08363342e-01 5.35874844e-01
7.93741643e-01 3.53412569e-01 -1.00854564e+00 1.26070559e+00
2.05354735e-01 7.97662064e-02 -2.54981130e-01 5.99181831e-01
2.47088283e-01 1.86495587e-01 -1.11895412e-01 -4.37328331e-02
1.28099394e+00 -6.59818172e-01 -7.81069636e-01 -1.20542966e-01
7.21777499e-01 -6.97627187e-01 1.19772494e+00 5.94253361e-01
-6.43192649e-01 -6.29593492e-01 -1.27807200e+00 2.14010060e-01
-1.06362128e+00 1.63686529e-01 2.71580458e-01 7.89902747e-01
-7.25879312e-01 7.92114556e-01 -8.15647304e-01 -4.40113425e-01
5.05479455e-01 2.30729476e-01 -5.75777233e-01 2.52545595e-01
-8.27326596e-01 9.91685569e-01 5.83894134e-01 1.59335852e-01
-1.06420457e+00 -5.87421060e-01 -8.30535173e-01 2.87330300e-02
6.67506695e-01 6.10662512e-02 7.51584232e-01 -1.27202582e+00
-8.56799901e-01 1.11841369e+00 -3.32606137e-02 -5.03667116e-01
8.85313153e-01 -2.85158366e-01 -6.58103883e-01 2.64649481e-01
-5.02049364e-02 1.33924112e-01 1.59797215e+00 -1.38726902e+00
-5.39246023e-01 -2.94554323e-01 3.75489108e-02 -4.31018114e-01
-4.76257384e-01 3.27335209e-01 -3.57846379e-01 -9.13573325e-01
-1.84489205e-01 -5.97318172e-01 1.44236043e-01 2.30351776e-01
-5.06701946e-01 1.99316144e-01 1.31267679e+00 -7.88950861e-01
1.36130929e+00 -2.44074440e+00 -2.13428721e-01 3.37891638e-01
2.91432410e-01 6.88164890e-01 -3.38405907e-01 3.55912924e-01
-3.87732953e-01 2.55949378e-01 -3.72092664e-01 -2.27887854e-01
-1.57120787e-02 -1.70616597e-01 -3.90476346e-01 8.82072031e-01
6.54564023e-01 6.57771528e-01 -9.07758236e-01 -1.24345355e-01
6.64127529e-01 1.81889340e-01 -1.35456711e-01 1.10243820e-01
-2.02792495e-01 3.96654040e-01 3.20545882e-02 1.02525437e+00
4.75345701e-01 1.34146363e-01 -4.26979363e-01 -2.80246407e-01
-6.59318566e-02 -1.05758302e-01 -1.18320727e+00 1.31165886e+00
-1.78076372e-01 8.45694900e-01 1.79079235e-01 -1.01008749e+00
6.81308270e-01 7.33742118e-03 3.43031794e-01 -6.07756853e-01
2.85988927e-01 3.06794822e-01 1.90683246e-01 -7.53304124e-01
6.68895483e-01 4.64477539e-01 2.15115651e-01 3.64676565e-01
-9.85332578e-02 3.24958086e-01 5.01092196e-01 1.23555763e-02
1.47081375e+00 -1.27096474e-01 3.38831514e-01 3.26517113e-02
6.39866173e-01 -1.68271095e-01 2.08892792e-01 1.25586927e+00
-4.65542287e-01 6.67140245e-01 6.34764910e-01 -3.72861594e-01
-1.13054633e+00 -9.24676120e-01 -2.16471970e-01 9.15365219e-01
5.92478439e-02 -3.85564864e-01 -6.79153800e-01 -1.07758141e+00
2.11938679e-01 5.54662108e-01 -8.78695250e-01 -4.70513433e-01
-5.45425832e-01 -7.49222159e-01 7.00009286e-01 2.70470858e-01
5.35376072e-01 -1.22772896e+00 -9.29122746e-01 3.44186723e-02
6.76963180e-02 -1.15523803e+00 -2.99858730e-02 1.87783763e-01
-4.17621583e-01 -1.36556280e+00 -3.58399153e-01 -5.13161719e-01
7.36417651e-01 2.66097993e-01 8.90383899e-01 4.32504743e-01
-6.66242421e-01 3.16369653e-01 -5.56295276e-01 -4.74377990e-01
-8.02354991e-01 -1.88357502e-01 1.50291622e-01 3.22524786e-01
4.24378455e-01 -1.70449495e-01 -3.00579131e-01 2.93041080e-01
-1.16340268e+00 -5.70375264e-01 6.39201403e-01 6.08910203e-01
3.15793246e-01 5.13669193e-01 3.13181192e-01 -5.47669232e-01
5.95816791e-01 -4.37854975e-01 -5.34212053e-01 2.10451052e-01
-2.58235484e-01 -1.16841299e-02 6.49595261e-01 -8.04007053e-01
-6.18184745e-01 5.82674444e-02 1.90562487e-01 -7.90824413e-01
-6.25683069e-01 2.23804876e-01 -2.59161860e-01 -2.87916034e-01
9.52138424e-01 2.79265940e-01 8.39818344e-02 -4.32617337e-01
1.40669227e-01 5.18310010e-01 6.28301799e-01 -6.03922792e-02
1.06529522e+00 5.48676848e-01 3.53665575e-02 -1.07434404e+00
-2.86168247e-01 -6.70232892e-01 -7.48208404e-01 -3.11692625e-01
6.52571380e-01 -3.90635431e-01 -2.89564222e-01 1.04089797e+00
-9.63791609e-01 -2.69546717e-01 -2.02619061e-01 -4.49189395e-02
-2.75357693e-01 8.92373919e-01 -2.38407388e-01 -8.75431836e-01
-9.46313217e-02 -1.40622878e+00 1.26938307e+00 -1.92107096e-01
-3.46687526e-01 -6.69395268e-01 -4.47460771e-01 1.97074914e-04
6.13329828e-01 7.81282127e-01 9.67837095e-01 -9.66677427e-01
-4.64635283e-01 -7.89165974e-01 -1.10211447e-01 5.52748919e-01
4.22678947e-01 4.94936943e-01 -1.14376044e+00 -4.21968400e-01
-8.70072618e-02 2.64673959e-02 9.10271168e-01 1.30224088e-02
1.32031095e+00 -1.44420250e-03 -3.38463485e-01 3.80338728e-01
1.24693000e+00 -2.88342871e-02 6.50948226e-01 6.50591373e-01
6.52804434e-01 5.13060749e-01 6.56925023e-01 2.94243366e-01
-4.26128000e-01 7.56801903e-01 8.93682599e-01 -7.82287791e-02
6.11689351e-02 2.19297811e-01 5.87562144e-01 1.66982412e-02
1.53373897e-01 -4.40694511e-01 -9.23023701e-01 6.14579737e-01
-1.56949234e+00 -1.00381315e+00 -4.01749313e-01 2.49478149e+00
3.29248637e-01 3.75470012e-01 2.07449242e-01 7.34612823e-01
9.70735371e-01 2.97717482e-01 -4.75972533e-01 -4.30796564e-01
-2.95307457e-01 1.77216649e-01 5.52136481e-01 -1.04710639e-01
-1.79942930e+00 6.92183435e-01 5.85617542e+00 7.36643851e-01
-1.21675432e+00 -2.31186911e-01 3.08702320e-01 8.70767236e-02
5.30186415e-01 -4.00239050e-01 -4.47137028e-01 7.67584324e-01
8.55993569e-01 3.94139856e-01 2.15983093e-01 8.08566034e-01
3.11292410e-01 -4.22189206e-01 -1.22375643e+00 9.38331008e-01
4.14954782e-01 -1.07757246e+00 1.19138218e-01 -5.48382886e-02
2.67942876e-01 -3.38807344e-01 1.11879788e-01 1.63406298e-01
-3.74343365e-01 -1.02339816e+00 7.88505852e-01 2.81633109e-01
5.64024866e-01 -5.45466065e-01 8.64363849e-01 2.06859410e-01
-9.69142795e-01 -3.33627939e-01 -3.45790684e-02 -1.71250012e-02
-7.78887719e-02 6.45874262e-01 -6.24378920e-01 4.95127022e-01
7.92833507e-01 6.12142265e-01 -1.02990198e+00 1.10402799e+00
-2.41155878e-01 3.99642557e-01 -3.78687918e-01 3.86174560e-01
1.41842112e-01 2.20398933e-01 8.17120016e-01 1.45563483e+00
1.63594350e-01 -6.57297909e-01 1.15596153e-01 8.91734302e-01
-4.11300063e-02 3.08715794e-02 -8.24148595e-01 -1.22573905e-01
3.31557661e-01 1.19200420e+00 -1.05107033e+00 -1.75023183e-01
-3.60200524e-01 1.06529474e+00 1.20026663e-01 1.41778708e-01
-8.21980298e-01 -5.80434322e-01 6.68823421e-01 1.62373289e-01
3.94241124e-01 -2.12819979e-01 -7.81470090e-02 -1.11727846e+00
5.45171142e-01 -1.27494776e+00 3.13743025e-01 -4.71179247e-01
-1.20759678e+00 2.13398740e-01 -9.64710563e-02 -1.33161080e+00
2.50824983e-03 -8.76272023e-01 -6.69079959e-01 5.34966767e-01
-1.28641891e+00 -1.06570017e+00 -6.34339213e-01 3.75735998e-01
4.77300793e-01 -8.94338861e-02 6.64078653e-01 3.53603303e-01
-8.07204425e-01 7.50630856e-01 -9.46053565e-02 5.21593869e-01
8.93396676e-01 -1.15294421e+00 5.80598056e-01 1.58491039e+00
4.66486923e-02 5.49116313e-01 8.15770626e-01 -6.19534016e-01
-1.27668285e+00 -1.21104705e+00 2.93069929e-01 -6.74287021e-01
7.99035847e-01 -6.31376266e-01 -1.29878986e+00 5.95393121e-01
-1.12019800e-01 2.12306723e-01 1.62633613e-01 -2.58641601e-01
-4.04957473e-01 2.39756033e-01 -1.17594755e+00 5.05402088e-01
7.71860898e-01 -6.24243498e-01 -4.34203297e-01 2.84049869e-01
1.28412366e-01 -3.44433933e-01 -5.46214581e-01 7.58920252e-01
3.67809176e-01 -1.05438876e+00 9.57409203e-01 -5.04810035e-01
4.14097667e-01 -5.24756610e-01 1.12856485e-01 -1.00858569e+00
-2.47642547e-01 -6.28183007e-01 -4.15294468e-01 1.21688354e+00
9.06082988e-02 -4.33422804e-01 4.99416739e-01 2.03744739e-01
-7.12480694e-02 -4.17759597e-01 -8.62691760e-01 -9.97213483e-01
-2.41731420e-01 -5.51937401e-01 3.30379337e-01 1.12020755e+00
-2.48642266e-01 -3.94043088e-01 -2.64383376e-01 4.64546859e-01
2.52003819e-01 -2.77047753e-01 9.89596128e-01 -9.31564927e-01
-2.76699662e-02 -6.06014073e-01 -1.21064687e+00 -3.81505281e-01
3.00365333e-02 -5.49746811e-01 1.63192049e-01 -8.39748442e-01
-1.57147035e-01 -1.79024026e-01 -2.86162972e-01 5.33439517e-01
-4.05057013e-01 3.93688500e-01 2.12642010e-02 1.67114928e-01
-3.45893115e-01 9.03273746e-02 6.29542828e-01 -3.23937923e-01
-7.87313432e-02 -2.14014381e-01 -1.21184848e-01 7.39178717e-01
8.48629773e-01 -3.88644874e-01 1.93194207e-02 -2.21578047e-01
-1.57888569e-02 -7.41683364e-01 6.42325044e-01 -1.53654242e+00
-2.24210009e-01 2.93356866e-01 4.12543863e-01 -5.31958103e-01
2.93755084e-02 -8.32125306e-01 -1.97606817e-01 4.74414974e-01
-3.05887878e-01 1.83773994e-01 3.82414728e-01 6.79331243e-01
-4.70964387e-02 -2.76995152e-01 1.00690639e+00 -3.79016511e-02
-9.65914786e-01 -7.72355646e-02 -7.70124078e-01 -3.53816599e-01
1.36432385e+00 -4.63216126e-01 -3.24284703e-01 -1.46598682e-01
-4.89418626e-01 -1.17381275e-01 9.10939455e-01 7.59770572e-01
4.55876201e-01 -8.04470241e-01 -4.82735693e-01 4.91544217e-01
4.61581379e-01 -1.93307027e-01 1.00195371e-01 9.98021781e-01
-6.49686158e-01 -1.84131581e-02 -2.99930334e-01 -7.82075047e-01
-1.49914968e+00 1.08808196e+00 3.67002219e-01 -2.76268601e-01
-7.52522647e-01 5.20757675e-01 -7.04792514e-02 -3.24740540e-03
4.42343563e-01 -5.51155269e-01 -2.07020104e-01 1.99863002e-01
8.09690773e-01 4.75246370e-01 4.49405670e-01 -7.38949895e-01
-3.98487806e-01 2.32744843e-01 -2.18973413e-01 5.04095376e-01
8.40124428e-01 2.07272828e-01 -7.00685084e-02 3.82978886e-01
1.10844028e+00 5.17546423e-02 -9.36987042e-01 -5.40974066e-02
2.74891794e-01 -6.28422260e-01 4.60813828e-02 -7.02912748e-01
-1.11009622e+00 8.90270352e-01 8.45907092e-01 4.55441713e-01
1.25286460e+00 -3.64721149e-01 4.58516210e-01 3.59060526e-01
1.21476874e-01 -9.49957311e-01 1.89230487e-01 3.64039242e-01
9.29723620e-01 -1.37371325e+00 7.21240267e-02 -4.22423959e-01
-1.84067026e-01 1.31543112e+00 5.81915915e-01 -1.85036629e-01
2.69891590e-01 2.80059308e-01 -1.87259130e-02 -2.74109066e-01
-1.93851471e-01 -1.97865769e-01 1.23902299e-01 5.96604705e-01
9.85213816e-02 -2.48305127e-01 4.73807342e-02 -1.58153363e-02
8.29462111e-02 -3.33848506e-01 6.18238926e-01 1.36987400e+00
-3.09593707e-01 -7.48261690e-01 -7.22183824e-01 8.31073284e-01
-6.18406713e-01 1.32999226e-01 -8.51934910e-01 9.70270038e-01
3.75528485e-01 9.75749493e-01 2.58885592e-01 -3.68632078e-01
3.88600260e-01 1.34088725e-01 2.52969980e-01 -3.73283744e-01
-8.69974256e-01 -2.63688982e-01 -6.43741190e-02 -8.85698020e-01
-2.90588081e-01 -7.85637319e-01 -7.77281821e-01 -1.66744798e-01
-5.82609951e-01 -4.92187709e-01 4.97166336e-01 9.03748691e-01
4.99539137e-01 6.60056651e-01 2.92784542e-01 -1.03516793e+00
-4.96774435e-01 -9.81238008e-01 -3.71105760e-01 9.89745915e-01
6.32947028e-01 -8.94907296e-01 -7.60119498e-01 8.66125599e-02] | [12.25423526763916, 1.0311890840530396] |
be5e9f32-ca2b-4d7c-b21a-e91b3b74cf6e | radar-based-respiratory-rate-monitoring-in | 2203.05075 | null | https://arxiv.org/abs/2203.05075v2 | https://arxiv.org/pdf/2203.05075v2.pdf | Radar-based Respiratory Rate Monitoring in Standing Position | Estimating human vital signs in a contactless non-invasive method using radar provides a convenient method in the medical field to conduct several health checkups easily and quickly. In addition to monitoring while sitting and sleeping, the standing position has aroused interest for both the industrial and medical fields. However, it is more challenging due to the micro motions induced by the body for balancing that may cause false respiratory rate estimation. In this work, we focus on the measurement of the respiratory rate of a standing person accurately with the capability of heavy breath detection and estimation. Multiple estimation approaches are presented and compared, including spectral estimation, deep-learning-based approaches, and adaptive peak selection with Kalman filtering. The latest technique is showing the best performance with an absolute error rate of 1.5 bpm, when compared to a Vernier Go Direct\textsuperscript{\textregistered} respiration belt. | ['Urs Schneider', 'Christoph Wasser', 'Dominik Alscher', 'Marco F. Huber', 'Omar Metwally', 'Tassneem Helal', 'Fady Aziz'] | 2022-03-09 | null | null | null | null | ['respiratory-rate-estimation'] | ['medical'] | [ 4.11535114e-01 -1.00777522e-01 1.17228344e-01 -3.01721022e-02
-4.20367837e-01 -1.40811494e-02 -5.99118359e-02 -3.09260469e-02
-5.56195378e-01 9.99069750e-01 -1.06360644e-01 -3.81706767e-02
-3.28126341e-01 -2.22918808e-01 2.49445975e-01 -9.07848001e-01
-2.15082243e-02 3.72103900e-01 2.79884756e-04 7.93463923e-03
-1.27108157e-01 3.94661248e-01 -1.20202351e+00 -3.55171978e-01
8.35763276e-01 1.00105941e+00 1.25906259e-01 7.64790952e-01
2.94649124e-01 5.04451573e-01 -9.36680973e-01 -1.23702563e-01
-1.31727103e-02 -8.36775601e-01 -2.34120376e-02 -3.72078210e-01
2.03442082e-01 -4.70877618e-01 -1.67174041e-01 6.65296495e-01
1.05043638e+00 5.28070033e-02 4.87935066e-01 -8.31102312e-01
3.89694422e-01 3.72918993e-01 -3.78170043e-01 4.85525131e-01
6.32045746e-01 7.95102194e-02 1.89346403e-01 -3.50906998e-01
7.70313814e-02 3.12393397e-01 1.10260105e+00 6.43808961e-01
-9.85248744e-01 -7.61524022e-01 -6.86464071e-01 3.17709506e-01
-1.36891222e+00 -7.22116888e-01 7.97376394e-01 -3.50089073e-01
6.82427526e-01 8.45945895e-01 9.42800522e-01 8.86497200e-01
6.19718492e-01 -9.68470201e-02 1.38432813e+00 -2.54686505e-01
1.74704075e-01 3.12965274e-01 -1.53502375e-02 5.25192678e-01
6.82352304e-01 -4.06316482e-02 -6.24690115e-01 -1.35543481e-01
6.60392880e-01 2.12050706e-01 -6.91847205e-01 -2.71354914e-02
-1.28646076e+00 4.19017464e-01 3.90150547e-02 7.20045865e-01
-6.95224047e-01 3.30665082e-01 9.17326510e-02 -6.57027811e-02
1.24835603e-01 4.87158716e-01 -3.28865647e-01 -6.82465553e-01
-1.41552413e+00 9.34959278e-02 1.10881412e+00 3.11189324e-01
-1.99244410e-01 1.95615202e-01 -3.38501781e-01 3.43387723e-01
4.48210210e-01 9.35042202e-01 4.26731080e-01 -7.90647686e-01
1.50830492e-01 2.94221900e-02 3.59479338e-01 -8.25966716e-01
-1.10282040e+00 -4.83577400e-01 -9.56893682e-01 -1.84310272e-01
3.92030209e-01 -3.33782583e-01 -3.12437773e-01 1.16478038e+00
6.57037973e-01 3.92717384e-02 -2.03628644e-01 1.04936194e+00
9.37903404e-01 3.91685963e-01 -1.20036984e-02 -8.24724674e-01
1.56606293e+00 -3.01422477e-01 -1.37238860e+00 -1.57128915e-01
-4.37284447e-03 -6.04904115e-01 4.46190953e-01 6.98766708e-01
-1.02718258e+00 -5.64987183e-01 -1.20072782e+00 3.69164079e-01
1.27397835e-01 1.80216044e-01 3.02852057e-02 1.21813607e+00
-6.83915317e-01 8.99539888e-01 -1.05274701e+00 -3.22773337e-01
-1.85383335e-01 2.45721593e-01 1.26873612e-01 4.18834418e-01
-1.23094583e+00 1.14045906e+00 -1.56166911e-01 4.43711936e-01
9.76514444e-02 -5.84492683e-01 -5.82473099e-01 -8.83750692e-02
1.44840345e-01 -7.61382878e-01 1.21102774e+00 2.45321002e-02
-2.01156163e+00 4.83313024e-01 -2.19324395e-01 -5.75944543e-01
9.52114701e-01 -5.06860912e-01 -6.82293594e-01 4.89081621e-01
-3.89062762e-01 -2.75787145e-01 1.00926387e+00 -3.95735323e-01
3.51605378e-02 -3.86041731e-01 -6.16419435e-01 7.21348673e-02
2.07950529e-02 -5.83837517e-02 3.34962271e-02 -2.91385531e-01
4.27343190e-01 -7.99743593e-01 6.09828494e-02 1.15422485e-02
-2.77524501e-01 1.55300125e-01 4.49857116e-01 -9.76313353e-01
1.52432919e+00 -1.85148036e+00 -1.22912213e-01 1.59478664e-01
3.54766577e-01 3.03480685e-01 9.49009120e-01 3.91617328e-01
1.84169084e-01 -1.99238807e-01 -3.52002174e-01 -1.46501794e-01
-2.55334407e-01 -2.33558670e-01 1.98727682e-01 1.02881718e+00
-4.77311581e-01 6.52567089e-01 -6.37268901e-01 -6.09719276e-01
5.84608376e-01 6.33598804e-01 1.56850651e-01 4.26873237e-01
5.62671065e-01 9.17263269e-01 -2.61423677e-01 6.44335330e-01
4.88403648e-01 -8.41138735e-02 2.41712049e-01 -5.64924300e-01
-2.90576279e-01 5.77829480e-01 -1.36398983e+00 1.35433948e+00
-4.80073214e-01 6.21320307e-01 2.23420173e-01 -8.06492567e-01
9.85927284e-01 7.26783097e-01 8.25206220e-01 -6.58956409e-01
2.05351934e-01 3.13255370e-01 1.61689699e-01 -1.18374860e+00
1.74015313e-01 -8.30468059e-01 1.39920160e-01 3.78981680e-01
-2.63577759e-01 -3.59938711e-01 -1.84341341e-01 -4.08011466e-01
1.02596962e+00 1.76642418e-01 8.25911045e-01 -3.53572547e-01
5.00943005e-01 -4.62501645e-01 2.94093758e-01 4.89173412e-01
-5.16453505e-01 5.02926171e-01 -1.52752295e-01 -2.15245172e-01
-3.52052063e-01 -9.05714989e-01 -5.37332416e-01 2.69477546e-01
2.21304893e-01 -2.02108249e-01 -5.36052227e-01 -4.99505401e-02
5.53808063e-02 8.11022341e-01 -2.42634863e-01 -2.39041969e-01
-6.50973678e-01 -7.32567668e-01 5.82128882e-01 3.90060693e-01
5.50819874e-01 -9.65412140e-01 -1.57766736e+00 3.22883636e-01
-4.89957660e-01 -1.02161705e+00 -1.89542890e-01 2.45169446e-01
-1.01709175e+00 -9.73692477e-01 -9.64951575e-01 2.07473248e-01
-1.37428865e-01 -1.63573116e-01 9.53363955e-01 -1.86763465e-01
-8.12053323e-01 5.48415780e-01 -1.86119769e-02 -5.92325389e-01
-2.70104229e-01 -6.45201728e-02 2.92849928e-01 -2.07261667e-01
2.17950493e-01 -6.16176307e-01 -1.01742482e+00 3.72808874e-01
-1.50033221e-01 -1.39934897e-01 2.90792793e-01 8.67747962e-02
1.24926627e-01 -4.07322943e-01 5.16177952e-01 -5.55454373e-01
4.62139368e-01 -3.09055358e-01 -4.28884834e-01 -2.15194687e-01
-6.99302197e-01 -2.99930781e-01 2.35097617e-01 -1.95990860e-01
-7.16035306e-01 -1.12540700e-01 -3.43382984e-01 -1.16357006e-01
-2.12313309e-01 -1.21072819e-02 1.54820174e-01 -1.91636905e-02
7.98993409e-01 2.02960804e-01 2.83569068e-01 -2.63513923e-01
-1.91094920e-01 5.36001563e-01 6.65891349e-01 1.50537983e-01
8.45223725e-01 3.26699346e-01 3.58829051e-01 -1.36650920e+00
-4.17851448e-01 -7.05427706e-01 -5.65444410e-01 -6.75563216e-01
1.29305434e+00 -6.05984092e-01 -1.02484763e+00 4.16771024e-01
-8.59248996e-01 -1.20494910e-01 -4.47748750e-01 1.06737518e+00
-5.57050765e-01 5.98645747e-01 -3.54576975e-01 -1.41121578e+00
-8.03404391e-01 -5.86392522e-01 7.81563163e-01 5.23153603e-01
-8.27053308e-01 -8.73152077e-01 1.66667163e-01 5.24541736e-01
9.32314932e-01 8.70951831e-01 8.14712867e-02 -2.93511152e-01
-5.64943664e-02 -5.49392700e-01 3.95207912e-01 4.87469099e-02
4.13792968e-01 -3.56473833e-01 -1.19829607e+00 -1.40290797e-01
6.69120073e-01 -1.21146869e-02 4.40485954e-01 5.31921923e-01
9.27682698e-01 -1.49523383e-02 -5.21750093e-01 5.90884387e-01
1.21086526e+00 3.70697409e-01 6.01401925e-01 -9.62575972e-02
3.27013522e-01 3.56005400e-01 4.32293981e-01 7.12145686e-01
-9.80884209e-02 5.51064372e-01 3.88945162e-01 -9.46497321e-02
2.17549860e-01 3.36982131e-01 1.86011389e-01 7.78041363e-01
-6.25143051e-01 -2.40382060e-01 -5.03602684e-01 -1.20258726e-01
-1.28888607e+00 -1.10510480e+00 -5.49052835e-01 2.76786470e+00
7.10015357e-01 4.02173512e-02 2.07007259e-01 6.99008465e-01
5.89163721e-01 1.85370371e-01 -4.64110404e-01 -2.39394695e-01
5.33979118e-01 4.88767743e-01 6.27277195e-01 2.27304623e-01
-1.00376570e+00 -2.48292461e-01 6.40057611e+00 1.02240965e-01
-1.22676826e+00 3.25973988e-01 1.10427225e-02 -1.74209252e-01
1.95835054e-01 -5.88947773e-01 -7.74648905e-01 7.21644402e-01
1.30626678e+00 -3.23976390e-02 1.16914406e-01 5.33785880e-01
4.11187202e-01 -6.23116612e-01 -8.62866044e-01 1.26883566e+00
2.51588404e-01 -6.59482300e-01 -9.17822063e-01 4.57178876e-02
7.08050234e-03 -2.78088778e-01 -2.97130704e-01 -1.01053836e-02
-1.03563535e+00 -8.53880286e-01 3.56684834e-01 1.04476273e+00
1.07866180e+00 -4.57580745e-01 7.61556864e-01 6.75585628e-01
-1.13293052e+00 2.52040029e-01 2.10321955e-02 -1.58070192e-01
5.09289384e-01 1.02361512e+00 -7.97606885e-01 5.68314791e-01
3.85090739e-01 1.73740551e-01 -7.18150362e-02 1.49115741e+00
-2.54485309e-01 8.06258798e-01 -7.05746114e-01 -3.01760912e-01
-5.15278399e-01 -3.37389231e-01 7.41780281e-01 1.01449895e+00
6.29276872e-01 4.30087894e-01 -1.98815286e-01 8.85226727e-01
3.65663767e-01 -8.99748579e-02 -4.87954348e-01 2.03377619e-01
1.18998215e-01 1.33149362e+00 -7.30895758e-01 -1.86524227e-01
-1.20646611e-01 9.24621046e-01 -8.67195368e-01 2.61540283e-02
-1.26550055e+00 -7.10103631e-01 2.03205913e-01 8.86096656e-01
-2.39716664e-01 -3.61767292e-01 -1.70467213e-01 -8.06716859e-01
3.70933488e-02 -5.85635900e-01 1.66914258e-02 -5.28095365e-01
-5.86870611e-01 2.63413966e-01 2.75349617e-01 -1.19545686e+00
-4.95253235e-01 -3.22250545e-01 -5.60444474e-01 1.04275405e+00
-1.19462621e+00 -4.33632582e-02 -9.21627522e-01 3.50634962e-01
3.23346168e-01 4.21131074e-01 9.38571930e-01 6.05549335e-01
-3.05427343e-01 3.47732991e-01 -1.31772459e-01 -4.98071581e-01
6.27393901e-01 -1.26820624e+00 4.08071764e-02 4.62833643e-01
-3.05239975e-01 5.48376679e-01 1.03047562e+00 -4.64569956e-01
-1.45327175e+00 -5.81693828e-01 9.82535779e-01 -2.66115606e-01
2.44628057e-01 -1.06666181e-02 -8.13526154e-01 -8.33770111e-02
3.18180434e-02 -3.77703011e-02 6.88543200e-01 -4.30121303e-01
5.71554601e-01 -4.66257274e-01 -1.37932229e+00 1.95131190e-02
6.33843005e-01 -3.48973304e-01 -6.79272294e-01 3.57698709e-01
1.35557517e-01 -8.32625568e-01 -9.56732035e-01 4.79693323e-01
1.00638890e+00 -1.08973241e+00 7.87371933e-01 3.33800375e-01
-3.08282703e-01 -2.75383592e-01 3.94097596e-01 -7.93577075e-01
-1.99469194e-01 -1.16332936e+00 -4.53082472e-01 8.57771039e-01
-1.03085395e-03 -1.01338243e+00 5.86446822e-01 4.96317774e-01
2.07337379e-01 -5.71474314e-01 -1.19706202e+00 -6.84023738e-01
-7.95987606e-01 -2.12600246e-01 -3.99211720e-02 5.17834246e-01
7.96040148e-02 2.23517120e-01 -5.70014775e-01 -3.08394451e-02
7.48396635e-01 -6.82140980e-03 4.19873714e-01 -1.45263660e+00
-4.34123039e-01 1.04923479e-01 -2.66318202e-01 -7.27729082e-01
-6.22663200e-01 -2.42863536e-01 3.42983186e-01 -1.69580901e+00
-2.11681411e-01 -2.70363837e-02 -3.99068356e-01 -7.99487159e-02
-2.83408642e-01 4.27861035e-01 -3.52799408e-02 -3.80326211e-02
-2.27973089e-01 1.36037767e-01 9.03881073e-01 2.45950669e-01
-4.55650032e-01 9.53319013e-01 -5.17300963e-02 7.09987640e-01
8.21383178e-01 -5.46039522e-01 -2.82746941e-01 4.00347143e-01
2.37178013e-01 5.12223125e-01 3.28449070e-01 -1.64873922e+00
1.19487189e-01 2.19463453e-01 6.48829162e-01 -7.42376387e-01
7.58977473e-01 -8.50927234e-01 6.92970991e-01 1.06213617e+00
1.98543310e-01 1.81119338e-01 1.14884563e-01 3.79852414e-01
1.46607220e-01 -3.55039358e-01 1.03891599e+00 -3.41364682e-01
8.79845768e-02 -8.22154507e-02 -6.78047121e-01 5.78213446e-02
7.30439663e-01 -5.49762070e-01 -9.22541022e-02 -6.01144314e-01
-8.35103989e-01 -8.12276080e-02 -5.43793924e-02 -1.43432423e-01
4.81993854e-01 -9.10158277e-01 -3.40192616e-01 9.98623595e-02
-2.72481024e-01 -3.90935242e-01 3.76806259e-01 1.65666533e+00
-5.93467474e-01 4.55838561e-01 2.91933026e-02 -6.52445853e-01
-1.37672019e+00 1.52430817e-01 7.96266794e-01 -1.56336755e-01
-8.23084116e-01 5.95608413e-01 -5.11907220e-01 3.91351193e-01
3.22893769e-01 -5.47853589e-01 -3.79500926e-01 2.93663025e-01
5.03650844e-01 9.36500788e-01 3.13530385e-01 -2.15375721e-01
-6.39022291e-01 7.85563707e-01 8.20972800e-01 -1.13443427e-01
7.30002463e-01 -2.22413808e-01 1.25311211e-01 9.71153975e-01
7.40777612e-01 3.77062410e-01 -6.16736591e-01 3.56618971e-01
9.72801063e-04 -1.57331452e-01 -1.96974635e-01 -8.35600436e-01
-6.55107200e-01 9.50749218e-01 1.04389024e+00 5.12272656e-01
1.12481129e+00 -4.84425962e-01 7.08547652e-01 3.62030715e-01
3.12628776e-01 -9.04841840e-01 -4.25022431e-02 -2.71884054e-01
6.72787309e-01 -7.32176721e-01 4.28138942e-01 -3.05252314e-01
-3.81629616e-01 1.22395587e+00 2.48752192e-01 1.69127673e-01
7.64259696e-01 3.09530675e-01 2.83872604e-01 -2.88334638e-02
-1.35352507e-01 -1.25116929e-01 3.33311290e-01 5.85441053e-01
9.06389654e-01 1.31937396e-02 -7.77152061e-01 1.94913581e-01
-1.60678342e-01 3.53435695e-01 4.15505350e-01 8.82079422e-01
-6.73926890e-01 -5.48964441e-01 -6.24684870e-01 5.95072806e-01
-9.36423957e-01 2.13517308e-01 -4.75647040e-02 9.13338065e-01
7.65429139e-02 1.10304832e+00 -9.93078351e-02 9.53340977e-02
5.82987368e-01 5.06504059e-01 7.87933588e-01 -3.78492892e-01
-8.56046379e-01 2.36814782e-01 2.11678386e-01 -7.21134663e-01
-7.76018202e-01 -6.88868284e-01 -1.32654250e+00 2.40969956e-01
-4.90104377e-01 2.52859801e-01 1.03621674e+00 8.19331646e-01
-5.26012331e-02 8.60617876e-01 3.59006912e-01 -6.98022246e-01
-7.33689487e-01 -1.17973447e+00 -8.38765562e-01 -1.01871535e-01
6.28294289e-01 -5.87134778e-01 -5.45743942e-01 -1.75417632e-01] | [13.915900230407715, 2.984098196029663] |
748df93e-79cb-4d05-963b-7d26ee2a19f7 | learning-robust-visual-representations-using | 1906.04547 | null | https://arxiv.org/abs/1906.04547v1 | https://arxiv.org/pdf/1906.04547v1.pdf | Learning robust visual representations using data augmentation invariance | Deep convolutional neural networks trained for image object categorization have shown remarkable similarities with representations found across the primate ventral visual stream. Yet, artificial and biological networks still exhibit important differences. Here we investigate one such property: increasing invariance to identity-preserving image transformations found along the ventral stream. Despite theoretical evidence that invariance should emerge naturally from the optimization process, we present empirical evidence that the activations of convolutional neural networks trained for object categorization are not robust to identity-preserving image transformations commonly used in data augmentation. As a solution, we propose data augmentation invariance, an unsupervised learning objective which improves the robustness of the learned representations by promoting the similarity between the activations of augmented image samples. Our results show that this approach is a simple, yet effective and efficient (10 % increase in training time) way of increasing the invariance of the models while obtaining similar categorization performance. | ['Peter König', 'Alex Hernández-García', 'Tim C. Kietzmann'] | 2019-06-11 | null | https://openreview.net/forum?id=B1elqkrKPH | https://openreview.net/pdf?id=B1elqkrKPH | null | ['object-categorization'] | ['computer-vision'] | [ 6.37650013e-01 1.48005694e-01 -1.32941594e-02 -6.22939587e-01
3.34867805e-01 -5.80045283e-01 8.91476691e-01 1.11919761e-01
-1.01649904e+00 2.45213985e-01 1.77841693e-01 -1.89315885e-01
-2.65320063e-01 -4.89633322e-01 -8.36539209e-01 -7.04756200e-01
-1.03096096e-02 -8.27936009e-02 2.82307006e-02 -1.57497838e-01
4.53683019e-01 9.70617115e-01 -1.77266634e+00 4.15396422e-01
5.21858513e-01 9.85103309e-01 -1.84409663e-01 4.84989017e-01
9.35585499e-02 6.51700258e-01 -2.46046335e-01 -8.90887901e-02
6.08955383e-01 -5.60305893e-01 -9.78209078e-01 1.47336170e-01
8.32662404e-01 -9.90409181e-02 -4.18587267e-01 1.17295313e+00
2.13062376e-01 4.64312285e-01 9.65076029e-01 -1.41062546e+00
-1.04985309e+00 3.35538089e-01 -1.67704687e-01 5.92462063e-01
-2.21604243e-01 1.70040444e-01 9.41324592e-01 -8.33787918e-01
5.60065925e-01 1.34135568e+00 5.38424671e-01 8.24451923e-01
-1.62411404e+00 -3.95976931e-01 1.87064052e-01 1.82816505e-01
-1.18392694e+00 -6.99638546e-01 6.30521178e-01 -4.88258898e-01
8.23665023e-01 1.82049096e-01 5.15018046e-01 9.26838875e-01
1.54945627e-01 3.68499517e-01 1.08037364e+00 -4.60024446e-01
2.72741824e-01 1.52915999e-01 2.43020222e-01 4.27409410e-01
4.82096493e-01 2.22709522e-01 -4.58071530e-01 8.15124363e-02
9.63873267e-01 -3.46616469e-02 -8.05945247e-02 -6.89261615e-01
-1.30091333e+00 6.97864056e-01 1.12134206e+00 5.30305505e-01
-4.65495974e-01 2.31669754e-01 5.03650665e-01 4.16554034e-01
2.12463155e-01 9.31871593e-01 -4.71476555e-01 3.80090505e-01
-4.04057682e-01 3.69380228e-02 2.06414416e-01 6.10125244e-01
6.94178760e-01 5.09056568e-01 -1.54447988e-01 9.37251389e-01
2.75917202e-01 1.17559470e-01 8.70354056e-01 -1.11690867e+00
-8.94471481e-02 7.26739705e-01 -4.68915075e-01 -1.02882731e+00
-5.44059455e-01 -5.14131010e-01 -9.54254866e-01 6.61442935e-01
6.26095414e-01 3.66456360e-01 -1.00330913e+00 2.11154461e+00
-1.94615778e-02 -2.38115653e-01 1.01873979e-01 7.39217222e-01
4.87427026e-01 1.77595899e-01 5.97978830e-01 1.47280563e-02
1.27517068e+00 -6.03973508e-01 -4.96914893e-01 -3.75596672e-01
4.49410111e-01 -4.51432347e-01 1.11594331e+00 9.32234228e-02
-9.84653413e-01 -9.43013787e-01 -1.14972937e+00 -2.41385534e-01
-7.37212121e-01 -1.76032588e-01 7.25380242e-01 5.93764663e-01
-9.58741844e-01 7.16214061e-01 -5.77475429e-01 -6.40266955e-01
9.02985871e-01 6.54162943e-01 -8.13588381e-01 1.34292364e-01
-6.56503260e-01 1.04291117e+00 6.06679678e-01 -3.69457044e-02
-7.88959980e-01 -8.71993363e-01 -8.27140093e-01 3.55871826e-01
-2.67990708e-01 -4.75922167e-01 9.83847737e-01 -1.66108978e+00
-1.15374613e+00 1.27129090e+00 -1.10686943e-01 -5.58863401e-01
3.27914119e-01 1.25839099e-01 4.64165844e-02 3.76796387e-02
-1.04170837e-01 1.37954867e+00 1.11187005e+00 -1.34537053e+00
-2.24538460e-01 -4.80229676e-01 -2.46533290e-01 -9.33134183e-02
-8.16876411e-01 5.04531190e-02 4.20230001e-01 -8.94993842e-01
5.08330643e-01 -7.89339483e-01 -2.66051412e-01 4.22013193e-01
9.44454446e-02 -2.53775507e-01 5.63640118e-01 -4.66282785e-01
5.19271970e-01 -2.41934204e+00 7.41128698e-02 3.19583416e-01
1.13479920e-01 2.88400531e-01 -6.42791033e-01 -4.55483794e-02
-6.78844392e-01 3.45044136e-01 -5.55466235e-01 -7.62724429e-02
-1.15910590e-01 4.98067796e-01 -4.10731941e-01 7.10884690e-01
6.10137641e-01 1.13950729e+00 -6.09927177e-01 -6.95789680e-02
7.22869262e-02 4.63805169e-01 -8.41525614e-01 1.29948650e-03
1.40226394e-01 5.97076356e-01 1.55206025e-01 1.47867918e-01
6.38794661e-01 7.01002330e-02 5.23355417e-02 -1.58190206e-01
-4.17737244e-03 2.83467114e-01 -7.75131047e-01 1.46069825e+00
-1.01995230e-01 1.04652131e+00 -3.32912952e-01 -1.51674271e+00
8.29476833e-01 1.16976433e-01 2.82607794e-01 -8.60644996e-01
5.60458779e-01 5.92721924e-02 8.47903192e-01 -2.68801302e-01
2.47550085e-01 -2.16613024e-01 2.87199616e-01 4.52150464e-01
6.10820651e-01 8.21115524e-02 9.14030969e-02 -5.53958118e-02
7.04188347e-01 -8.04115310e-02 2.74127692e-01 -6.93393111e-01
5.35967231e-01 -2.02914566e-01 2.47171149e-01 8.52718949e-01
-2.42483526e-01 4.37335372e-01 3.89897257e-01 -5.69410920e-01
-1.40461826e+00 -1.00659347e+00 -4.33184355e-01 1.31918883e+00
-1.75486863e-01 9.34007391e-02 -8.20018828e-01 -3.56206030e-01
-1.08400229e-02 5.07892072e-01 -1.20573318e+00 -7.54341424e-01
-5.92455208e-01 -7.53225505e-01 6.57083333e-01 8.20357025e-01
6.76808536e-01 -1.43109906e+00 -7.28152394e-01 9.19708237e-03
2.02796891e-01 -1.27194357e+00 -5.20669110e-02 4.61573869e-01
-1.14298701e+00 -9.22592878e-01 -5.37389874e-01 -9.15831089e-01
1.25248373e+00 2.20487148e-01 6.01814985e-01 3.02856863e-01
-5.40655434e-01 4.65072632e-01 -1.52400002e-01 -5.64819336e-01
-5.64167976e-01 7.15020448e-02 4.44572717e-01 2.40101784e-01
3.41464311e-01 -6.92609906e-01 -2.96205938e-01 2.76376218e-01
-1.32630789e+00 -3.63596469e-01 3.51712376e-01 8.45626175e-01
3.31500292e-01 -3.46391320e-01 6.24335408e-01 -4.20639843e-01
6.98311865e-01 -1.58117890e-01 -4.29212958e-01 -1.43207282e-01
-5.94117761e-01 4.01175499e-01 6.75392091e-01 -8.58490884e-01
-7.38324404e-01 1.57587662e-01 -1.55497804e-01 -2.63152212e-01
-6.15184247e-01 1.81480631e-01 1.47938788e-01 -6.72928452e-01
1.18062103e+00 2.24665120e-01 2.90624857e-01 -4.48810726e-01
3.27437103e-01 2.35272616e-01 6.91882432e-01 -3.73201966e-01
9.84738231e-01 6.97931886e-01 2.56871641e-01 -9.81162369e-01
-4.65305954e-01 -3.22146177e-01 -1.13563347e+00 -6.77021444e-02
8.92293692e-01 -4.08651769e-01 -8.10445428e-01 3.78438443e-01
-1.15613675e+00 -2.57828355e-01 -6.59930885e-01 4.33936179e-01
-7.15946913e-01 2.44300082e-01 -1.56576633e-01 -5.13148606e-01
-7.57427770e-05 -7.00641394e-01 4.07253206e-01 1.65106386e-01
-3.72183889e-01 -8.77427936e-01 -8.61870795e-02 -1.27966091e-01
6.33874774e-01 7.18525946e-02 1.16746974e+00 -9.32209134e-01
-2.56574124e-01 -1.83427438e-01 -4.62810785e-01 8.27458978e-01
3.08754593e-01 -5.41903116e-02 -1.33281362e+00 -2.23365337e-01
-6.73795342e-02 -2.62698650e-01 1.21921241e+00 3.69452953e-01
1.39961004e+00 -3.79428506e-01 1.34166688e-01 7.17815161e-01
1.14750183e+00 1.81129798e-01 8.02775145e-01 5.83838761e-01
3.49465549e-01 9.38063681e-01 -9.76528600e-02 -4.67008613e-02
-3.80884826e-01 3.31853211e-01 4.51993793e-01 -4.11812603e-01
-2.40691692e-01 4.78110425e-02 1.55791163e-01 3.49385798e-01
-3.94864619e-01 1.71328932e-01 -7.52760231e-01 6.61986291e-01
-1.58266902e+00 -1.03519261e+00 1.63962260e-01 2.27478027e+00
5.82620084e-01 1.18038870e-01 8.32240283e-02 5.48415482e-01
3.91141415e-01 -1.89446121e-01 -5.80853105e-01 -7.61842251e-01
-4.35143948e-01 4.29022133e-01 3.89863491e-01 2.56423801e-01
-1.10629785e+00 9.32402372e-01 7.30288935e+00 2.02073440e-01
-1.30390644e+00 -2.53114756e-02 4.29168940e-01 2.78301358e-01
1.26642495e-01 -3.08600783e-01 -1.52586520e-01 -8.26741382e-03
6.82091415e-01 -1.57565683e-01 5.43037295e-01 9.10933733e-01
-4.90800403e-02 2.21321911e-01 -1.47540534e+00 7.25332916e-01
1.89541578e-01 -1.36575103e+00 5.20898342e-01 2.86244422e-01
7.14433670e-01 -3.67702469e-02 4.25024509e-01 1.29482791e-01
-1.02253109e-01 -1.11732388e+00 6.89744592e-01 3.64809990e-01
4.76356953e-01 -5.25216401e-01 5.29915750e-01 2.09431797e-01
-5.70688963e-01 -3.58250797e-01 -6.96445227e-01 -3.43555659e-01
-3.92098635e-01 -2.30980232e-01 -6.04646862e-01 -2.09110394e-01
7.85009682e-01 5.64640522e-01 -1.04389620e+00 1.25357234e+00
-1.28141329e-01 3.44719887e-01 -1.17976129e-01 1.00272641e-01
2.83628672e-01 1.59574319e-02 3.96931231e-01 1.07861292e+00
-8.73738825e-02 -4.11549248e-02 -4.45267141e-01 9.92998302e-01
-3.77345324e-01 1.71008036e-01 -1.03471684e+00 -1.34352058e-01
-9.08842459e-02 1.06247497e+00 -9.67643619e-01 -3.56879771e-01
-2.00060681e-01 9.28204358e-01 2.99223274e-01 5.20341873e-01
-5.15764058e-01 -2.46952236e-01 9.23940122e-01 -3.72974575e-02
1.21688999e-01 -3.51578534e-01 -6.15488708e-01 -9.04187739e-01
-1.05410635e-01 -8.61033380e-01 1.68285280e-01 -7.15804577e-01
-1.31127179e+00 6.82318211e-01 -2.59770430e-03 -1.05809581e+00
4.70454991e-02 -1.30507958e+00 -3.84670198e-01 5.89445651e-01
-1.37315679e+00 -9.49713290e-01 -3.83589625e-01 6.67548478e-01
4.73767370e-01 -2.95889139e-01 9.22077239e-01 5.87153323e-02
-1.87610567e-01 7.64014304e-01 1.63166188e-02 3.61908495e-01
3.94626766e-01 -9.44830954e-01 4.18807983e-01 9.01347637e-01
4.82782155e-01 1.02111208e+00 5.67049444e-01 -1.82053782e-02
-8.95073056e-01 -8.46555293e-01 6.71998322e-01 -3.24698269e-01
4.79343563e-01 -5.02196610e-01 -1.30204964e+00 7.10950673e-01
3.49761307e-01 1.59364656e-01 6.49832785e-01 -1.71751417e-02
-8.95185232e-01 -4.99358624e-02 -1.14929974e+00 7.36167252e-01
1.19909048e+00 -6.26191199e-01 -9.01354015e-01 2.05661163e-01
4.40807492e-01 3.17229450e-01 -6.62904918e-01 5.81289768e-01
6.81365848e-01 -6.02828622e-01 1.18352318e+00 -1.44931960e+00
4.02601004e-01 -9.26427618e-02 -2.62993664e-01 -1.28536320e+00
-6.99362457e-01 -1.07038103e-01 3.84497255e-01 7.74041057e-01
3.49120229e-01 -1.02328992e+00 4.72938180e-01 5.68970799e-01
-5.10891862e-02 -1.44574031e-01 -1.18560588e+00 -9.98050451e-01
4.73643571e-01 -2.55728602e-01 2.07664490e-01 1.18763065e+00
-3.19850184e-02 2.80953020e-01 1.13691486e-01 -5.36959469e-02
5.63798845e-01 -3.88387084e-01 5.86599052e-01 -1.52668858e+00
2.61871099e-01 -9.05056417e-01 -8.57487381e-01 -6.45928502e-01
3.89919966e-01 -1.15318036e+00 1.76412351e-02 -1.12527812e+00
5.86355454e-04 -2.17973441e-01 -7.14670002e-01 8.64003599e-01
1.43833086e-01 5.99424720e-01 3.13108444e-01 1.88099742e-01
2.36489763e-03 4.50017869e-01 1.25764942e+00 -1.41605869e-01
3.50485817e-02 -2.13657200e-01 -7.76539147e-01 8.59554589e-01
9.59406078e-01 -5.93395710e-01 -1.64088860e-01 -5.58783531e-01
1.25372410e-01 -1.10367525e+00 7.07772732e-01 -1.09542418e+00
5.14680743e-02 6.22846372e-02 7.81481922e-01 2.10681155e-01
1.28906175e-01 -9.75305736e-01 -4.18480545e-01 7.55480766e-01
-9.02541459e-01 5.29238991e-02 6.91531479e-01 2.78046906e-01
-1.62974954e-01 -4.32432204e-01 1.35955238e+00 -6.38878196e-02
-8.93128812e-01 2.00470507e-01 -8.03184211e-01 -2.05570266e-01
7.18420684e-01 -4.05019879e-01 -3.23038757e-01 -1.01272732e-01
-7.43927896e-01 -4.14585263e-01 3.02095741e-01 7.21088290e-01
5.47358990e-01 -1.27654994e+00 -6.18021548e-01 6.21831119e-01
2.17941329e-01 -6.20208919e-01 -1.90677211e-01 6.92778409e-01
-3.43859583e-01 3.86531651e-01 -1.08759129e+00 -6.23128891e-01
-1.17893732e+00 8.72542679e-01 6.01586998e-01 4.42369699e-01
-2.94627696e-01 9.17080164e-01 5.54551005e-01 -5.50924897e-01
2.03586325e-01 -5.33117831e-01 -3.00879091e-01 4.26334813e-02
3.81914914e-01 7.88444430e-02 1.95815325e-01 -8.33772838e-01
-4.28151578e-01 6.06316507e-01 -3.31900753e-02 1.21445851e-02
1.35627460e+00 2.68838584e-01 -3.23756814e-01 2.38046616e-01
1.36758482e+00 -4.36758876e-01 -9.76336122e-01 -2.05426648e-01
1.32431582e-01 -5.01694500e-01 -1.63057774e-01 -4.60937083e-01
-9.23321903e-01 1.20025730e+00 8.70592177e-01 4.17745173e-01
1.21339691e+00 -9.91906300e-02 -1.02455162e-01 7.41209567e-01
-5.05871363e-02 -8.96380246e-01 2.50362098e-01 5.09049296e-01
1.28496647e+00 -1.27640986e+00 -7.36642480e-02 -2.01176003e-01
-4.26492900e-01 1.03768981e+00 8.94739330e-01 -5.25553107e-01
5.29506147e-01 -1.75267279e-01 5.59540428e-02 3.22839953e-02
-4.37385380e-01 -3.13306153e-01 5.03664553e-01 8.51047993e-01
3.57117623e-01 -4.05355692e-01 -2.72965163e-01 1.55819384e-02
-1.17474616e-01 -4.86337692e-01 2.86077350e-01 7.88497388e-01
-3.80768806e-01 -9.20115471e-01 -2.58719742e-01 2.65591651e-01
-3.36330593e-01 -7.30509534e-02 -5.57973504e-01 9.04639006e-01
2.57693380e-01 4.62570995e-01 4.45606768e-01 -9.52122286e-02
5.04318118e-01 3.72836620e-01 6.06088579e-01 -4.93873507e-01
-7.35208333e-01 -3.84159416e-01 -4.46667582e-01 -3.81083906e-01
-5.86455107e-01 -4.49602813e-01 -1.10388851e+00 6.79681078e-02
-1.36794746e-01 -2.21768841e-01 8.84987891e-01 8.53862286e-01
2.47396424e-01 6.82717085e-01 5.33671200e-01 -9.59922135e-01
-4.76073503e-01 -1.12421525e+00 -2.93387145e-01 8.98791313e-01
4.99000460e-01 -7.84609199e-01 -5.05206645e-01 4.82841551e-01] | [9.656601905822754, 2.385835886001587] |
8691f741-4b12-4769-a88b-9193a636caa3 | a-reranking-model-for-discourse-segmentation | null | null | https://aclanthology.org/W12-1623 | https://aclanthology.org/W12-1623.pdf | A Reranking Model for Discourse Segmentation using Subtree Features | null | ['Nguyen Le Minh', 'Ngo Xuan Bach', 'Akira Shimazu'] | 2012-07-01 | null | null | null | ws-2012-7 | ['discourse-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] | [-7.394907474517822, 3.655766248703003] |
559add4e-5df0-4ba8-b6f9-d5b05f91d843 | knowledge-aware-deep-framework-for | 2106.03455 | null | https://arxiv.org/abs/2106.03455v2 | https://arxiv.org/pdf/2106.03455v2.pdf | Knowledge-aware Deep Framework for Collaborative Skin Lesion Segmentation and Melanoma Recognition | Deep learning techniques have shown their superior performance in dermatologist clinical inspection. Nevertheless, melanoma diagnosis is still a challenging task due to the difficulty of incorporating the useful dermatologist clinical knowledge into the learning process. In this paper, we propose a novel knowledge-aware deep framework that incorporates some clinical knowledge into collaborative learning of two important melanoma diagnosis tasks, i.e., skin lesion segmentation and melanoma recognition. Specifically, to exploit the knowledge of morphological expressions of the lesion region and also the periphery region for melanoma identification, a lesion-based pooling and shape extraction (LPSE) scheme is designed, which transfers the structure information obtained from skin lesion segmentation into melanoma recognition. Meanwhile, to pass the skin lesion diagnosis knowledge from melanoma recognition to skin lesion segmentation, an effective diagnosis guided feature fusion (DGFF) strategy is designed. Moreover, we propose a recursive mutual learning mechanism that further promotes the inter-task cooperation, and thus iteratively improves the joint learning capability of the model for both skin lesion segmentation and melanoma recognition. Experimental results on two publicly available skin lesion datasets show the effectiveness of the proposed method for melanoma analysis. | ['Jun Liu', 'Yuqian Zhao', 'Henghui Ding', 'Xudong Jiang', 'XiaoHong Wang'] | 2021-06-07 | null | null | null | null | ['melanoma-diagnosis', 'skin-lesion-segmentation', 'clinical-knowledge'] | ['computer-vision', 'medical', 'miscellaneous'] | [ 4.76771861e-01 -3.13485175e-01 -1.98594689e-01 -3.13979566e-01
-9.73808348e-01 -4.00303423e-01 3.19073915e-01 1.61051065e-01
-4.90670383e-01 5.00764012e-01 -4.68501374e-02 -1.70733228e-01
-3.93687010e-01 -6.51365161e-01 2.54243854e-02 -1.36676824e+00
6.50296628e-01 -1.89379171e-01 1.19889088e-01 2.79907674e-01
1.24914810e-01 6.41942084e-01 -9.82859015e-01 3.44980806e-01
1.23948538e+00 1.06986272e+00 4.23826963e-01 6.91776037e-01
-1.50113791e-01 8.89596283e-01 -1.90159872e-01 -1.85702309e-01
-6.87946007e-02 -3.99381876e-01 -8.26270938e-01 7.88349092e-01
-1.01062968e-01 -2.85666466e-01 -9.38601717e-02 9.98438179e-01
6.63023055e-01 -2.05459088e-01 7.72517085e-01 -6.14779532e-01
-2.33202949e-01 1.30791560e-01 -9.84464526e-01 1.09793007e-01
-1.19938828e-01 3.42650771e-01 6.60082877e-01 -6.53814018e-01
4.19918686e-01 5.76926410e-01 5.92089713e-01 4.37827617e-01
-5.78753114e-01 -3.83745790e-01 -3.03379446e-02 2.53034025e-01
-1.43008459e+00 -1.91372707e-01 8.27290237e-01 -5.13005197e-01
2.87457585e-01 3.90755951e-01 6.37114286e-01 6.15432203e-01
1.93596020e-01 1.07883263e+00 1.13416159e+00 -4.13598359e-01
-3.35527472e-02 2.55743116e-01 1.58459708e-01 1.03256929e+00
-6.05121106e-02 -1.82370305e-01 -2.81513065e-01 2.79785898e-02
9.99862790e-01 3.01845074e-01 -2.52158344e-01 -5.99117354e-02
-9.24143076e-01 4.53308672e-01 5.77112615e-01 3.55016947e-01
-5.52225590e-01 -2.01794431e-01 4.13693607e-01 -5.82058989e-02
4.04041409e-01 -8.82207230e-02 -1.73747182e-01 3.45223874e-01
-7.73433626e-01 -4.16597366e-01 4.53380316e-01 3.78492684e-03
6.91165924e-01 -4.52607185e-01 -2.99228460e-01 8.74739587e-01
4.58305418e-01 2.91746825e-01 5.77754974e-01 -2.85423130e-01
1.01002574e-01 1.02576780e+00 -1.55617520e-01 -8.04916024e-01
-4.27689523e-01 -5.76032460e-01 -1.16986251e+00 -9.78178233e-02
3.20438176e-01 -5.29721677e-01 -1.01587689e+00 1.08264077e+00
7.48793006e-01 4.14296031e-01 3.05762053e-01 1.06693220e+00
8.87053668e-01 2.34616116e-01 4.38678592e-01 -1.56894207e-01
1.40890920e+00 -9.62643683e-01 -5.13206303e-01 1.02612749e-01
7.79520810e-01 -8.15214515e-01 3.61832678e-01 1.37346417e-01
-8.08069050e-01 -5.29973388e-01 -7.38312960e-01 -3.96959513e-04
-8.85492116e-02 8.29443276e-01 9.46146131e-01 3.80886495e-01
-6.63761556e-01 6.63890243e-02 -1.01293457e+00 -4.17651355e-01
8.15621138e-01 4.67535406e-01 -5.07209122e-01 -1.65297434e-01
-9.01320100e-01 7.26771295e-01 3.82373214e-01 6.44008994e-01
-7.31987357e-01 -5.39119303e-01 -6.45383835e-01 -4.48065668e-01
4.00518864e-01 -8.41919124e-01 8.83353412e-01 -1.16207755e+00
-1.49122834e+00 9.74894881e-01 -2.88624257e-01 -2.40625098e-01
5.06049037e-01 1.54606849e-01 -3.22207838e-01 3.97855252e-01
-2.73934871e-01 5.58447719e-01 9.29139376e-01 -1.04993594e+00
-8.58614683e-01 -5.21890759e-01 -2.30851453e-02 5.35412490e-01
-5.79891384e-01 -1.38581932e-01 -5.09992421e-01 -2.59018868e-01
-4.10699286e-02 -4.82299507e-01 -5.34370482e-01 2.84197748e-01
-5.66617966e-01 -2.81312168e-01 5.95830798e-01 -1.06173491e+00
1.20897734e+00 -2.18460417e+00 2.52803952e-01 5.72528064e-01
4.04622376e-01 6.27460659e-01 -4.43238467e-02 1.03181474e-01
1.44976586e-01 -6.09034300e-02 -2.70850211e-01 -2.39209309e-01
-3.92843843e-01 1.54236043e-02 3.86422694e-01 5.24553895e-01
5.86719215e-01 1.08587480e+00 -6.18685663e-01 -9.31512952e-01
4.16720450e-01 5.09289086e-01 -5.19214645e-02 2.25252628e-01
1.47381425e-01 5.63171327e-01 -7.87538052e-01 8.57179046e-01
8.34345698e-01 -4.94933814e-01 2.82265604e-01 -5.14157057e-01
1.95568465e-02 -5.07180274e-01 -9.08860385e-01 1.65341413e+00
-6.41404688e-01 2.22290397e-01 3.12912315e-01 -7.88033843e-01
7.58536398e-01 4.44985658e-01 5.06027162e-01 -4.10972893e-01
3.16711366e-01 2.28480726e-01 6.85482547e-02 -1.07253659e+00
-9.41842198e-02 -1.07107960e-01 4.18917298e-01 2.97209769e-01
-9.91636142e-02 5.93883693e-02 4.89792228e-02 -6.55158013e-02
8.38276505e-01 -5.68266027e-02 5.46918631e-01 1.35553420e-01
1.13098907e+00 9.32414085e-03 6.24397099e-01 4.35203798e-02
-3.41603398e-01 2.38934591e-01 2.98635960e-01 -2.70745486e-01
-4.42063749e-01 -1.05985010e+00 -1.34998441e-01 6.35529935e-01
2.92228371e-01 2.15054259e-01 -6.94320083e-01 -9.85892713e-01
-6.05908595e-02 -1.71511009e-01 -8.81440043e-01 -1.03791080e-01
-2.81140804e-01 -1.16393936e+00 4.45668399e-01 6.12309694e-01
8.75402570e-01 -7.82777905e-01 -3.64116490e-01 7.38201886e-02
6.26360178e-02 -7.43353546e-01 -4.18286711e-01 -1.17678575e-01
-6.25924885e-01 -1.47528565e+00 -1.03551102e+00 -1.13183475e+00
1.07793581e+00 2.73365438e-01 2.52968401e-01 2.81967700e-01
-9.95018959e-01 3.01248848e-01 -1.45459592e-01 -1.26445562e-01
-2.97794610e-01 -8.39197368e-04 -5.54413795e-01 6.94389105e-01
3.48416924e-01 -9.68360677e-02 -6.68177068e-01 1.12740651e-01
-9.72512066e-01 1.84315518e-01 1.33699000e+00 1.01396704e+00
6.39704585e-01 4.00040567e-01 5.96319735e-01 -8.72717679e-01
6.46027446e-01 -3.71280998e-01 -2.52966434e-01 7.25434303e-01
3.86264583e-04 -4.26124752e-01 2.97002554e-01 -1.39951527e-01
-1.64815712e+00 4.37302589e-01 -1.60058007e-01 -1.04174189e-01
-4.39649075e-01 8.26843798e-01 -1.89350784e-01 -3.85012895e-01
3.28114241e-01 5.00065565e-01 4.23358500e-01 -3.19937587e-01
7.44827986e-02 9.20838714e-01 5.47109663e-01 2.10551322e-02
4.68848586e-01 5.42966723e-01 1.53332353e-01 -8.11682403e-01
-8.55217814e-01 -8.39931369e-01 -7.14507759e-01 -1.69458315e-01
1.13294578e+00 -9.69724715e-01 -6.41369700e-01 1.10300756e+00
-9.08836126e-01 -5.07693477e-02 -4.36722115e-02 4.75858688e-01
-1.89702604e-02 7.60515392e-01 -7.86118150e-01 -8.40208590e-01
-5.44098020e-01 -1.16648400e+00 1.10948873e+00 7.83928633e-01
2.19431311e-01 -1.57977307e+00 -6.69281185e-02 6.71901703e-01
1.60266668e-01 3.23376328e-01 7.14828014e-01 -4.92103100e-01
-4.91507351e-01 -4.27296042e-01 -7.34575450e-01 6.31316543e-01
6.63547933e-01 1.87725753e-01 -8.73643875e-01 -8.64972621e-02
-1.89977095e-01 -4.16763693e-01 1.15508091e+00 3.66548568e-01
1.00024617e+00 3.18241030e-01 -4.67200190e-01 6.06033802e-01
1.49570119e+00 5.01419678e-02 3.13427120e-01 -2.37625659e-01
9.79870200e-01 7.99440503e-01 6.06962740e-01 4.51536894e-01
5.75809479e-01 1.36152059e-01 2.91034669e-01 -8.02380860e-01
-3.28261077e-01 1.32914558e-01 -5.50636239e-02 9.26740766e-01
-1.00039005e-01 -3.10823712e-02 -8.79241884e-01 6.27602398e-01
-1.71943712e+00 -4.59593356e-01 -6.07452057e-02 1.91263080e+00
9.62584317e-01 -3.21839631e-01 -2.09147960e-01 5.15843630e-02
7.79662788e-01 -1.92189395e-01 -6.56626403e-01 3.62339020e-02
-1.01635858e-01 1.38818949e-01 3.13678205e-01 3.15166086e-01
-1.39709115e+00 7.27276862e-01 5.07680750e+00 1.28784001e+00
-1.41961503e+00 -1.78129449e-02 6.71321452e-01 3.03734809e-01
2.04808190e-02 -3.93097132e-01 -6.04366839e-01 3.54067892e-01
-1.35398889e-02 4.01291661e-02 1.17006479e-02 2.23110661e-01
2.23409310e-01 -3.74546230e-01 -7.16124535e-01 8.83131981e-01
-4.62302603e-02 -1.24389863e+00 5.56618907e-02 8.35782196e-03
7.85651445e-01 -5.15885353e-01 2.82103568e-01 -2.56919205e-01
-7.65411034e-02 -8.84181857e-01 -3.87279332e-01 1.23122835e+00
7.47641683e-01 -8.98088574e-01 1.18678641e+00 2.90745229e-01
-1.14313853e+00 -7.58718848e-02 -9.18930024e-02 4.22029048e-01
-9.57050249e-02 7.25721180e-01 -1.17253089e+00 1.07333982e+00
-2.01927442e-02 9.98305917e-01 -8.48952830e-01 1.37634361e+00
-3.05804938e-01 4.75384653e-01 -8.23497847e-02 -1.12269949e-02
3.67489815e-01 1.66954026e-02 1.92982480e-01 1.08331537e+00
3.66025642e-02 -8.36773887e-02 3.67851913e-01 6.64403617e-01
2.58638024e-01 3.85621369e-01 9.80128348e-03 -1.65342152e-01
1.25213385e-01 1.79363954e+00 -7.07311451e-01 -2.10000798e-02
-2.11258963e-01 9.54899609e-01 1.13363206e-01 2.75212914e-01
-6.49360478e-01 -4.71285433e-01 2.77287781e-01 -4.43877012e-01
9.72731188e-02 2.00458497e-01 -2.48253196e-01 -1.01502395e+00
-1.14087008e-01 -4.40129042e-01 3.85808587e-01 -2.93317586e-01
-1.56266606e+00 3.32270414e-01 -5.67949533e-01 -1.14756167e+00
2.60720681e-02 -5.75084150e-01 -1.04073250e+00 1.08184934e+00
-1.95099199e+00 -1.82320642e+00 -5.63166320e-01 8.34844291e-01
1.97442055e-01 -6.88400939e-02 6.41788900e-01 1.51064977e-01
-1.05389082e+00 6.26454413e-01 -1.53412372e-01 3.65726113e-01
7.42145479e-01 -1.15457606e+00 -3.72543603e-01 6.99669302e-01
-3.74044806e-01 3.77643168e-01 -3.14376622e-01 -6.46863580e-01
-1.16742992e+00 -1.31260955e+00 5.87568820e-01 1.05469361e-01
6.63165390e-01 2.30540231e-01 -7.53679037e-01 2.13984713e-01
1.67954713e-01 9.34661105e-02 1.14513433e+00 -2.35116571e-01
8.51971284e-02 -1.90357909e-01 -1.14274633e+00 5.31480789e-01
4.76855665e-01 -6.62701845e-01 -2.00479805e-01 5.71083188e-01
2.46364176e-01 -3.65070611e-01 -9.45355117e-01 5.56081951e-01
5.85491598e-01 -5.92249095e-01 8.16402733e-01 -4.36306059e-01
3.18479329e-01 -3.24857712e-01 1.19754486e-01 -1.09847677e+00
-1.60372585e-01 -1.58620417e-01 2.96566367e-01 1.21119761e+00
3.18144053e-01 -5.61063588e-01 9.32790101e-01 3.09700280e-01
3.57414074e-02 -1.21013308e+00 -6.67404771e-01 -1.16383344e-01
-4.42704223e-02 -5.49616441e-02 1.75396562e-01 9.06776011e-01
-1.55743241e-01 -2.15199273e-02 -1.13854229e-01 3.39292854e-01
6.86053514e-01 1.87649846e-01 4.20526475e-01 -9.29116666e-01
-2.29863286e-01 -4.45404589e-01 -4.59550858e-01 -6.87109053e-01
-3.29938382e-02 -9.99155939e-01 -1.90696836e-01 -1.72247589e+00
4.99322504e-01 -3.50850910e-01 -7.28956759e-01 5.04430056e-01
-6.40185297e-01 2.77376264e-01 1.22188460e-02 -1.07195629e-02
-7.74814844e-01 1.37722671e-01 1.81296504e+00 -2.23069206e-01
-2.02519462e-01 2.00841621e-01 -6.91491783e-01 8.17979395e-01
7.02517807e-01 1.59904242e-01 -2.79425055e-01 -4.50980127e-01
-2.81653166e-01 1.77776292e-01 7.79805720e-01 -9.22909021e-01
7.90613353e-01 -9.89140347e-02 6.18656397e-01 -5.76952517e-01
1.65456042e-01 -6.08344316e-01 -2.87417442e-01 5.92256665e-01
-2.86072761e-01 -7.59396136e-01 8.03947076e-02 6.62604570e-01
-4.40117538e-01 -1.60247549e-01 7.20463812e-01 -1.71752438e-01
-9.40431118e-01 4.85053599e-01 -4.07096684e-01 -4.60187435e-01
1.41146648e+00 -1.96569622e-01 -1.97345406e-01 3.13350618e-01
-1.09243977e+00 4.94518906e-01 7.14964494e-02 7.64064351e-03
8.00036311e-01 -9.84998107e-01 -9.24833834e-01 1.96059883e-01
6.68308511e-02 1.57184154e-01 1.02961457e+00 1.48191404e+00
-4.79930490e-01 2.29380086e-01 -1.69859216e-01 -6.64896190e-01
-1.47953784e+00 1.02803059e-01 5.06385863e-01 -6.87831759e-01
-3.17231379e-03 1.22454166e+00 1.51357755e-01 -3.35247010e-01
6.33978173e-02 -3.19433846e-02 -4.55389470e-01 3.01166505e-01
5.27964592e-01 2.58545369e-01 8.22918043e-02 -5.01107335e-01
-3.96660626e-01 8.57640207e-01 -5.27216196e-01 4.38709229e-01
8.68852198e-01 -2.29130507e-01 -4.27071929e-01 -1.30612895e-01
1.12024248e+00 -7.22406879e-02 -1.17221546e+00 -5.00324249e-01
-2.38826528e-01 -2.04056263e-01 3.59507591e-01 -1.17615950e+00
-1.32450676e+00 1.07605350e+00 7.38761783e-01 -1.88845605e-01
1.48286867e+00 -3.45473021e-01 8.71223748e-01 2.20358923e-01
8.15139934e-02 -9.11456406e-01 -5.11253774e-02 -4.45493944e-02
4.27184582e-01 -1.37978458e+00 2.92725768e-02 -8.04539144e-01
-7.95546174e-01 1.19814432e+00 6.29119992e-01 -9.36596617e-02
7.24068344e-01 2.11235926e-01 3.42868328e-01 4.08003293e-02
-6.21761560e-01 -4.89730835e-01 4.70034778e-01 5.53977787e-01
1.45699605e-01 2.78343230e-01 -3.14905792e-01 7.55806863e-01
6.62333846e-01 2.06110075e-01 4.09472138e-02 8.89523745e-01
-4.21178550e-01 -1.19055235e+00 -8.61849487e-02 4.83243495e-01
-3.95634383e-01 -6.88530058e-02 -5.79694033e-01 5.99943459e-01
6.21446550e-01 8.12427223e-01 -1.53269216e-01 -3.82168770e-01
-2.02919275e-01 -1.30395591e-01 5.84664583e-01 -5.79911947e-01
-7.90483952e-01 3.19687694e-01 -2.49424741e-01 -2.19717935e-01
-6.91090941e-01 -5.77847779e-01 -1.25097954e+00 2.28689581e-01
-4.92102385e-01 -3.33285481e-02 6.61491573e-01 1.25995386e+00
1.49309188e-01 7.07852960e-01 9.60141480e-01 -2.52266496e-01
-4.16800439e-01 -6.54395342e-01 -7.74570286e-01 7.71396011e-02
3.05707127e-01 -3.07249069e-01 -8.98508132e-02 1.07922383e-01] | [15.64020824432373, -2.9317216873168945] |
dc8c8142-53d5-4caa-b9d4-b66f44851755 | watching-the-news-towards-videoqa-models-that | 2211.05588 | null | https://arxiv.org/abs/2211.05588v1 | https://arxiv.org/pdf/2211.05588v1.pdf | Watching the News: Towards VideoQA Models that can Read | Video Question Answering methods focus on commonsense reasoning and visual cognition of objects or persons and their interactions over time. Current VideoQA approaches ignore the textual information present in the video. Instead, we argue that textual information is complementary to the action and provides essential contextualisation cues to the reasoning process. To this end, we propose a novel VideoQA task that requires reading and understanding the text in the video. To explore this direction, we focus on news videos and require QA systems to comprehend and answer questions about the topics presented by combining visual and textual cues in the video. We introduce the ``NewsVideoQA'' dataset that comprises more than $8,600$ QA pairs on $3,000+$ news videos obtained from diverse news channels from around the world. We demonstrate the limitations of current Scene Text VQA and VideoQA methods and propose ways to incorporate scene text information into VideoQA methods. | ['C. V. Jawahar', 'Dimosthenis Karatzas', 'Minesh Mathew', 'Soumya Jahagirdar'] | 2022-11-10 | null | null | null | null | ['video-question-answering'] | ['computer-vision'] | [ 9.97341275e-02 -8.87744203e-02 6.63824379e-02 -7.51194060e-01
-8.14226985e-01 -7.35757530e-01 6.72528505e-01 4.61296178e-02
-4.67877984e-01 6.13193929e-01 6.98646426e-01 -2.40836680e-01
2.08472833e-01 -6.03823781e-01 -9.85343993e-01 -1.33914456e-01
2.51509905e-01 1.02310367e-01 3.23465228e-01 -6.09929204e-01
3.59068096e-01 -1.53439254e-01 -1.64162099e+00 1.25369251e+00
4.76147622e-01 1.19442606e+00 2.92557329e-01 1.05237639e+00
-2.63891697e-01 2.02879858e+00 -7.67698884e-01 -8.71923506e-01
1.29692793e-01 -8.81578088e-01 -1.11563659e+00 4.61187452e-01
1.18314338e+00 -9.84747827e-01 -9.67194438e-01 1.02351439e+00
7.40987509e-02 3.09598535e-01 3.89540046e-01 -1.49140358e+00
-8.41668308e-01 5.85208952e-01 -2.24204198e-01 8.97059262e-01
1.20271277e+00 3.81066084e-01 1.36171591e+00 -8.55414391e-01
1.03133690e+00 1.38575995e+00 2.10884050e-01 6.73590720e-01
-5.92657685e-01 -3.39690804e-01 5.22377312e-01 9.00721014e-01
-9.91420448e-01 -7.60487914e-01 9.63700056e-01 -5.06336570e-01
9.03986752e-01 3.42362881e-01 9.36313450e-01 1.44681346e+00
9.28743109e-02 1.34337914e+00 7.25632429e-01 -1.95417210e-01
3.08684886e-01 -1.84103101e-01 9.95664224e-02 1.04472435e+00
-2.35617101e-01 -2.53636509e-01 -1.16316128e+00 3.63380983e-02
6.33302212e-01 1.74053889e-02 -4.49100852e-01 -3.30954432e-01
-1.33503520e+00 9.12519157e-01 2.78303087e-01 8.55616629e-02
-4.72992778e-01 6.11448944e-01 5.92865050e-01 5.35872877e-01
2.20391065e-01 3.03934157e-01 -9.60524902e-02 -5.93562126e-01
-7.94553936e-01 3.89338613e-01 6.77105963e-01 1.06652880e+00
4.53601331e-01 8.16118866e-02 -3.40942293e-01 3.23287338e-01
4.33052182e-01 8.59125435e-01 -2.02760585e-02 -1.66250908e+00
8.94128084e-01 4.32658255e-01 1.85479715e-01 -1.46177495e+00
-3.86283062e-02 3.42102855e-01 -6.53872266e-02 -2.00916678e-01
6.01738691e-01 -3.43085229e-02 -7.56683826e-01 1.50892150e+00
2.24355087e-01 1.37659594e-01 8.40213075e-02 1.12223935e+00
1.43470454e+00 9.29283679e-01 2.74875194e-01 -2.31473312e-01
1.73531413e+00 -1.06010485e+00 -1.06311309e+00 -3.38811010e-01
3.58895212e-01 -6.02619946e-01 1.38776612e+00 5.53436041e-01
-1.40662909e+00 -4.58486944e-01 -9.04385149e-01 -6.12535298e-01
-1.85910419e-01 -1.46675587e-01 4.37971860e-01 4.72419709e-01
-1.01413691e+00 -3.42830010e-02 -4.06185925e-01 -3.14129919e-01
7.50480831e-01 -1.92940176e-01 -1.47328496e-01 -7.07496762e-01
-1.36621869e+00 5.24220884e-01 -1.25603974e-01 1.13885477e-01
-1.39544213e+00 -4.37103301e-01 -1.13888407e+00 -7.66395479e-02
8.15022171e-01 -8.86224389e-01 1.67296374e+00 -1.62923455e+00
-1.36124718e+00 8.43035698e-01 -2.99241751e-01 -5.07270515e-01
4.45930243e-01 -6.72721863e-01 -4.10457373e-01 1.30906713e+00
7.33372271e-02 7.56543458e-01 1.10857642e+00 -1.13377643e+00
-7.48565674e-01 -2.02176318e-01 1.10481358e+00 5.01726747e-01
-3.75320688e-02 1.59272790e-01 -5.51973641e-01 -7.65570045e-01
-2.77648002e-01 -5.53733349e-01 1.44020036e-01 2.70815998e-01
9.81588513e-02 2.05398388e-02 9.24015701e-01 -1.03205669e+00
9.31253612e-01 -2.13374329e+00 2.28097796e-01 -2.04634294e-01
4.59038883e-01 -1.89809710e-01 -2.04527691e-01 5.78228593e-01
3.01220924e-01 -2.14615047e-01 2.20392942e-01 -5.61417826e-02
7.00692683e-02 4.33450311e-01 -5.60384452e-01 4.78550315e-01
1.11296184e-01 1.08879530e+00 -1.21150017e+00 -7.84408808e-01
2.69357949e-01 4.45958167e-01 -8.11769366e-01 1.67066514e-01
-8.01454365e-01 1.97193623e-01 -5.96592069e-01 7.48443425e-01
2.29388058e-01 -3.30384463e-01 -1.23623088e-01 -3.56655300e-01
2.77026981e-01 2.66853392e-01 -6.99782014e-01 2.12661457e+00
5.95827866e-03 1.33978200e+00 3.50285694e-02 -9.15105462e-01
1.45033494e-01 4.46387053e-01 4.12413061e-01 -1.13804603e+00
2.03579158e-01 -3.77057970e-01 -3.43465805e-01 -1.14879823e+00
6.39042437e-01 -1.34574771e-01 -1.49731725e-01 4.50898200e-01
1.58288509e-01 -3.12958926e-01 4.77132231e-01 8.90205681e-01
1.11036825e+00 2.10680410e-01 1.05399594e-01 9.66512561e-02
3.91067505e-01 4.73140001e-01 1.11837789e-01 9.70691741e-01
-7.24621236e-01 5.01211584e-01 5.31273782e-01 -5.53374827e-01
-8.11794639e-01 -1.02615047e+00 5.56148231e-01 1.62245595e+00
5.03672063e-01 -6.22437537e-01 -7.31566727e-01 -6.89850807e-01
-4.67094541e-01 7.46084869e-01 -7.88603604e-01 1.05916530e-01
-6.33888423e-01 9.03410614e-02 4.26236719e-01 7.22627699e-01
7.14224815e-01 -1.02387512e+00 -9.76119518e-01 -1.56153604e-01
-1.11650848e+00 -1.51341736e+00 -5.27517855e-01 -5.84114790e-01
-5.54050088e-01 -1.29736745e+00 -5.11666238e-01 -5.67264318e-01
3.87199372e-01 6.26660585e-01 1.55017984e+00 1.70047715e-01
-2.90150911e-01 1.41809416e+00 -9.20352340e-01 -2.67275929e-01
-1.52155936e-01 -6.61791563e-01 -3.55003595e-01 -2.29584016e-02
4.06859487e-01 5.97245507e-02 -6.51816487e-01 2.07647666e-01
-8.38398218e-01 2.14989662e-01 1.50168806e-01 6.79844499e-01
1.82389751e-01 -2.10758653e-02 1.73633665e-01 -6.03216946e-01
3.95793527e-01 -5.01417875e-01 -1.05599396e-01 4.22804266e-01
2.92880028e-01 -2.18362510e-01 2.49954045e-01 -3.30701947e-01
-1.51605701e+00 -7.32730255e-02 3.81927080e-02 -5.04716396e-01
-1.71574295e-01 3.59155118e-01 -7.05737025e-02 2.17531696e-01
7.07080722e-01 2.10518003e-01 -2.48141661e-01 3.45437616e-01
6.87827587e-01 2.45464772e-01 6.64007425e-01 -7.39083946e-01
3.98688644e-01 1.06885266e+00 -4.23619449e-01 -1.06481111e+00
-1.08491158e+00 -7.16303468e-01 -3.01462054e-01 -8.95935893e-01
1.24119663e+00 -1.19071054e+00 -1.05068529e+00 2.66486239e-02
-1.37046421e+00 -3.28219652e-01 -4.58066553e-01 3.16331059e-01
-1.08345044e+00 6.32465959e-01 -9.03109491e-01 -7.13169336e-01
-3.20872478e-02 -1.05642402e+00 1.16037297e+00 -1.66737631e-01
-1.41398370e-01 -7.08009720e-01 -3.91467273e-01 1.09878612e+00
1.66181117e-01 -1.46038504e-02 5.17273784e-01 -2.67569870e-01
-8.30879807e-01 1.38299108e-01 -4.28433806e-01 8.83042719e-03
-2.25610241e-01 -1.71889812e-02 -9.12956774e-01 5.66956811e-02
2.99151123e-01 -7.31304288e-01 7.25771368e-01 4.32775140e-01
1.01696932e+00 -3.63487363e-01 1.76563010e-01 4.84752618e-02
1.22920477e+00 4.44238693e-01 9.10093904e-01 2.06957102e-01
5.90282977e-01 7.95816779e-01 9.12300229e-01 5.30584753e-01
7.83027589e-01 4.11649704e-01 6.78777099e-01 1.76681191e-01
-2.28674352e-01 -3.04495931e-01 7.99203336e-01 5.15809119e-01
-2.65688568e-01 -6.23701632e-01 -8.67713451e-01 5.27934253e-01
-1.90567994e+00 -1.68506551e+00 -1.22595981e-01 1.21366787e+00
4.09846634e-01 -6.91423342e-02 9.32510197e-02 3.74961160e-02
4.07478333e-01 4.50590640e-01 -2.81357557e-01 -2.24720761e-01
-1.82164595e-01 -3.29526424e-01 -1.25021428e-01 3.57547522e-01
-1.05685556e+00 8.30415368e-01 6.90991259e+00 6.02793634e-01
-6.50759041e-01 2.56927535e-02 5.29316366e-01 -3.94269794e-01
-4.40813065e-01 -1.90487355e-01 -2.49334186e-01 3.82172614e-02
7.63005257e-01 1.30931318e-01 4.53561246e-01 6.31010532e-01
2.11696148e-01 -5.04771352e-01 -1.31987131e+00 1.25775862e+00
8.45322251e-01 -1.56350923e+00 4.79292840e-01 -5.50077319e-01
4.61761892e-01 -3.38283330e-01 1.95916519e-01 3.25639993e-01
9.84150544e-02 -9.19830918e-01 1.18130910e+00 5.82179904e-01
4.52599198e-01 -4.54392463e-01 5.74782848e-01 9.41127166e-02
-1.19200361e+00 -3.34699959e-01 -8.12171027e-02 -3.49405438e-01
5.79210281e-01 1.26592452e-02 -4.97469097e-01 3.80134553e-01
1.14001536e+00 9.48646188e-01 -6.19455934e-01 4.72358167e-01
-1.85710847e-01 4.68079925e-01 9.51312333e-02 -7.95764923e-02
3.17141712e-01 -1.40188141e-02 4.47573602e-01 9.85399425e-01
5.17590381e-02 8.64724994e-01 -1.05325192e-01 5.15521169e-01
-1.44829974e-01 -1.37288058e-02 -5.96191585e-01 -3.64342153e-01
-1.28083855e-01 5.06703913e-01 -8.02710116e-01 -6.16913855e-01
-9.79376495e-01 1.13456094e+00 7.05127837e-03 6.42513275e-01
-1.13714564e+00 -8.45551938e-02 4.27143008e-01 5.09949848e-02
3.49355996e-01 -2.69062757e-01 3.24880809e-01 -1.54811907e+00
9.84644368e-02 -1.29337418e+00 8.76479685e-01 -1.73367143e+00
-1.22535563e+00 4.23669070e-01 1.09598339e-01 -1.00184023e+00
-2.23702297e-01 -8.74949515e-01 -1.50658280e-01 -2.71891177e-01
-1.41315091e+00 -1.04060543e+00 -6.27070665e-01 1.14597273e+00
1.33653319e+00 1.43603459e-01 2.67124653e-01 3.01828504e-01
1.85730338e-01 2.64662271e-03 -4.49706495e-01 3.30248266e-01
7.65464187e-01 -9.13510025e-01 5.73441684e-02 7.81688869e-01
5.66984832e-01 4.36448395e-01 1.05888283e+00 -5.05207062e-01
-1.82042372e+00 -4.39501286e-01 5.92570841e-01 -9.61039722e-01
8.02029490e-01 -2.72285819e-01 -5.86318493e-01 9.58564878e-01
7.61465728e-01 -1.78181589e-01 7.98788011e-01 -2.28301078e-01
-6.01993084e-01 1.24213412e-01 -9.61903572e-01 7.82072902e-01
1.29408681e+00 -1.03646326e+00 -1.20480871e+00 2.69833058e-01
9.88662362e-01 -4.84804809e-01 -4.42722708e-01 9.16982591e-02
6.51297927e-01 -9.64105904e-01 1.24068344e+00 -9.67756450e-01
9.15797710e-01 -1.28243014e-01 -7.27192819e-01 -8.52701247e-01
2.73051530e-01 -2.75672913e-01 -2.71771610e-01 8.51736546e-01
8.58121216e-02 1.58296347e-01 7.33851075e-01 7.14936435e-01
-5.09567782e-02 -8.38389471e-02 -8.03901315e-01 -1.06317878e-01
-3.09026062e-01 -9.34633911e-01 1.13295466e-01 1.01423061e+00
4.12410975e-01 3.25749278e-01 -6.25280321e-01 5.96598396e-03
2.54266441e-01 9.41075161e-02 7.66503572e-01 -8.12429547e-01
-2.68838227e-01 -6.13896959e-02 -5.79668880e-01 -1.57349265e+00
1.33755967e-01 -1.84919208e-01 2.43571639e-01 -1.59353197e+00
4.47867572e-01 5.09202957e-01 1.87332571e-01 5.06893881e-02
6.18222319e-02 2.91268796e-01 4.16614681e-01 -4.05091047e-02
-1.50299799e+00 5.29986918e-01 1.52902448e+00 -3.79782796e-01
3.10170799e-01 -6.73575580e-01 -5.12546122e-01 8.41437340e-01
3.60723138e-01 -5.12301885e-02 -6.84274554e-01 -9.39506412e-01
7.31451511e-01 6.46871924e-01 7.77183414e-01 -7.60978758e-01
3.85715008e-01 -4.46784973e-01 3.83692265e-01 -7.08218634e-01
7.98151314e-01 -9.62600768e-01 -3.19173127e-01 2.42291182e-01
-5.63003421e-01 3.27129602e-01 2.58379459e-01 1.06808054e+00
-5.20671785e-01 3.32032666e-02 1.41848832e-01 -3.69690388e-01
-1.29967177e+00 1.55033797e-01 -9.18715000e-01 4.59871829e-01
1.10529649e+00 -2.50030339e-01 -8.54683936e-01 -1.17697978e+00
-8.19261193e-01 4.86094981e-01 3.32461715e-01 4.54817414e-01
1.16064954e+00 -1.20752609e+00 -6.10356629e-01 -2.49074951e-01
4.56081092e-01 -3.67704451e-01 8.02332640e-01 6.06317937e-01
-8.23812783e-01 4.12965477e-01 -3.35683644e-01 -6.06500924e-01
-1.35359526e+00 8.62729967e-01 1.44565478e-01 5.60117066e-01
-6.68449283e-01 9.65732038e-01 4.91942585e-01 3.57294887e-01
3.54192376e-01 -5.14270663e-01 -3.99934679e-01 1.58100083e-01
8.76626015e-01 2.66516060e-01 -4.47870195e-01 -1.07074308e+00
-2.10656881e-01 4.20027524e-01 2.07139924e-01 -3.59828055e-01
9.23257053e-01 -7.84855187e-01 1.66071787e-01 6.09470904e-01
1.01259303e+00 -7.61596188e-02 -1.32468784e+00 -1.77989781e-01
-2.83944428e-01 -8.41778398e-01 -1.85326681e-01 -6.70649946e-01
-9.51799572e-01 1.09340358e+00 2.59439200e-01 1.54336408e-01
1.13315535e+00 4.65043515e-01 8.09612870e-01 7.50436902e-01
3.94595176e-01 -1.28354836e+00 1.01885986e+00 2.70068765e-01
1.00543189e+00 -1.52403104e+00 9.33948010e-02 -4.98012811e-01
-1.10114229e+00 1.15433037e+00 5.64284265e-01 4.76596206e-02
3.68693620e-01 -2.85180539e-01 3.03805321e-01 -7.94671476e-01
-1.02090025e+00 -2.66292930e-01 3.52397949e-01 5.52158356e-01
1.86385036e-01 -2.93633550e-01 4.22883146e-02 4.42484498e-01
-8.83877501e-02 6.62358254e-02 7.66311586e-01 1.11416125e+00
-5.41999280e-01 -2.48989969e-01 -3.90937686e-01 1.85364842e-01
-5.78791499e-01 -1.36041656e-01 -4.94458586e-01 7.16809511e-01
-1.60334781e-01 1.48422325e+00 1.94924921e-01 -5.29504009e-02
2.48321861e-01 -5.52761406e-02 6.94368958e-01 -3.31752181e-01
-2.80648053e-01 -2.26280272e-01 4.16598767e-01 -1.06494498e+00
-1.09189630e+00 -7.76285768e-01 -1.25734150e+00 -3.67390573e-01
9.77555569e-03 1.04706787e-01 3.03398907e-01 1.13495779e+00
-1.36874150e-02 5.55117548e-01 3.93457599e-02 -5.77063441e-01
1.16309643e-01 -3.41701359e-01 -2.74750710e-01 7.09154427e-01
6.08986378e-01 -6.25749588e-01 -4.79375243e-01 8.28725755e-01] | [10.456852912902832, 1.002577304840088] |
2084d044-fb1c-40f7-9d7e-d4050864995d | text2shape-deep-retrieval-model-generating | 2302.06341 | null | https://arxiv.org/abs/2302.06341v1 | https://arxiv.org/pdf/2302.06341v1.pdf | Text2shape Deep Retrieval Model: Generating Initial Cases for Mechanical Part Redesign under the Context of Case-Based Reasoning | Retrieving the similar solutions from the historical case base for new design requirements is the first step in mechanical part redesign under the context of case-based reasoning. However, the manual retrieving method has the problem of low efficiency when the case base is large. Additionally, it is difficult for simple reasoning algorithms (e.g., rule-based reasoning, decision tree) to cover all the features in complicated design solutions. In this regard, a text2shape deep retrieval model is established in order to support text description-based mechanical part shapes retrieval, where the texts are for describing the structural features of the target mechanical parts. More specifically, feature engineering is applied to identify the key structural features of the target mechanical parts. Based on the identified key structural features, a training set of 1000 samples was constructed, where each sample consisted of a paragraph of text description of a group of structural features and the corresponding 3D shape of the structural features. RNN and 3D CNN algorithms were customized to build the text2shape deep retrieval model. Orthogonal experiments were used for modeling turning. Eventually, the highest accuracy of the model was 0.98; therefore, the model can be effective for retrieving initial cases for mechanical part redesign. | ['Pingyu Jiang', 'Wentao Yong', 'Maolin Yang', 'Tianshuo Zang'] | 2023-02-13 | null | null | null | null | ['feature-engineering'] | ['methodology'] | [-3.52178693e-01 -1.79466575e-01 -2.00786129e-01 -1.88014984e-01
-7.10360706e-01 -4.35375124e-01 1.79508049e-02 -7.68719055e-03
4.31184262e-01 2.28631243e-01 1.88683018e-01 -9.00173262e-02
-7.39640236e-01 -1.01147318e+00 -5.15755415e-01 -4.35465395e-01
3.88033837e-01 6.99052870e-01 -1.45343676e-01 -5.75721204e-01
5.59423804e-01 8.31607461e-01 -2.02218986e+00 4.93594497e-01
5.10636210e-01 1.32004523e+00 5.33084512e-01 4.01320755e-02
-5.20479977e-01 3.86529654e-01 -7.00476587e-01 -1.65236160e-01
1.17954142e-01 3.86887160e-03 -6.69207454e-01 -1.16928682e-01
2.10330576e-01 -7.06528544e-01 -5.05346119e-01 4.71365184e-01
7.07538843e-01 2.13563889e-01 7.79326260e-01 -1.07395899e+00
-9.53676283e-01 6.67144060e-01 -1.46907315e-01 -5.11403859e-01
7.17227697e-01 -7.58794844e-02 1.03431261e+00 -1.30517352e+00
7.07732797e-01 1.20702350e+00 5.24987102e-01 3.98073703e-01
-3.97952497e-01 -5.21995723e-01 7.02066496e-02 3.98991346e-01
-1.50071239e+00 -5.84471691e-03 1.36286497e+00 -4.03498828e-01
1.18757510e+00 1.76592156e-01 1.16625834e+00 5.60109437e-01
5.14737666e-01 1.00710821e+00 7.90263992e-03 -2.19847322e-01
2.03287542e-01 -1.32371947e-01 2.84953807e-02 4.91584480e-01
1.19696945e-01 -1.31317213e-01 -2.64254093e-01 1.88282561e-02
8.37495387e-01 4.24353212e-01 7.24086352e-03 -3.24517667e-01
-7.87776172e-01 5.98723948e-01 7.21139908e-01 3.58510524e-01
-4.13386792e-01 1.84900045e-01 5.01281619e-01 2.80690432e-01
1.71914071e-01 7.49292016e-01 -6.56344533e-01 -9.89245772e-02
-7.85877645e-01 4.97648925e-01 5.88453948e-01 1.55317795e+00
6.20265663e-01 6.63631782e-02 -1.60586327e-01 9.86660361e-01
4.18883473e-01 6.48896277e-01 4.95531023e-01 -8.13639343e-01
5.33564508e-01 1.09737945e+00 1.28774373e-02 -1.24032545e+00
-3.94555420e-01 -3.40043992e-01 -8.15237045e-01 -2.55253702e-01
-4.49711651e-01 1.98856354e-01 -8.52523088e-01 1.13202226e+00
2.00305998e-01 -7.74469793e-01 -7.37965703e-02 8.48218441e-01
1.20188820e+00 7.44575799e-01 -2.65650243e-01 1.14203848e-01
1.09967613e+00 -5.60273707e-01 -7.99007714e-01 8.17093998e-02
6.61382794e-01 -9.35710490e-01 8.64046574e-01 2.51589715e-01
-8.33080888e-01 -8.14867318e-01 -1.20562136e+00 -1.59236282e-01
-6.08665884e-01 7.52377450e-01 7.76362598e-01 -1.56366229e-02
-5.62883556e-01 6.86160147e-01 -1.52338058e-01 -1.26823202e-01
2.13207528e-01 3.55356365e-01 4.58561517e-02 -2.76654184e-01
-1.29889882e+00 9.88922238e-01 6.18434370e-01 7.32732594e-01
-7.64477372e-01 -6.37488902e-01 -8.86681080e-01 1.64565757e-01
4.02405143e-01 -8.21802795e-01 1.05371308e+00 -3.13077420e-01
-1.50482249e+00 3.45971078e-01 3.60801011e-01 2.28473663e-01
2.03716993e-01 -4.56898153e-01 -5.33693016e-01 -2.75219344e-02
-7.32324049e-02 5.01127958e-01 8.95203888e-01 -1.26176643e+00
-3.07154566e-01 -1.60526022e-01 2.94176549e-01 2.12131709e-01
-1.53645933e-01 -3.28802466e-01 -5.47604799e-01 -7.00739622e-01
3.75490874e-01 -7.45686948e-01 3.72096114e-02 1.90124765e-01
-4.91978645e-01 -5.12104392e-01 7.40769565e-01 -8.32012057e-01
1.40578437e+00 -2.33847427e+00 2.01409355e-01 3.14719707e-01
1.77096665e-01 -7.86409155e-02 -1.42148733e-01 9.30426121e-01
-1.79817140e-01 2.92081796e-02 1.89341292e-01 3.63623887e-01
2.95975089e-01 -1.24129720e-01 -4.84755903e-01 -2.18638837e-01
2.28239849e-01 1.00283098e+00 -6.30379796e-01 -3.51641268e-01
4.48762774e-01 2.68800199e-01 -5.90641439e-01 1.08218275e-01
-4.46852058e-01 -2.72545695e-01 -1.03036714e+00 9.46454048e-01
4.50514555e-01 -1.84075739e-02 -1.64973252e-02 -8.04871798e-01
1.06642097e-01 -4.61907052e-02 -1.03996360e+00 1.70036793e+00
-7.45924532e-01 4.05169487e-01 -4.96603042e-01 -6.21351123e-01
1.40482199e+00 1.60189822e-01 7.37056315e-01 -6.02749467e-01
3.97997350e-01 2.67350405e-01 -1.00411266e-01 -9.03593004e-01
6.66311741e-01 9.73259285e-02 -3.16020936e-01 2.61531323e-01
-3.16552818e-01 -9.16551173e-01 4.36927117e-02 -9.17718336e-02
8.31495941e-01 4.79312241e-01 6.91263452e-02 9.73366499e-02
3.36909384e-01 2.43509710e-01 2.44318098e-01 2.56693125e-01
5.49385786e-01 7.21491098e-01 -3.53352688e-02 -8.33370090e-01
-1.08782327e+00 -7.28454232e-01 1.17424168e-01 5.62960684e-01
2.71134734e-01 -7.50730932e-01 -2.99612641e-01 -2.08878741e-01
1.84519246e-01 6.96714580e-01 -3.95197362e-01 -6.59089744e-01
-5.76856852e-01 -1.83830373e-02 3.53878923e-02 7.32868075e-01
6.71564877e-01 -1.17293441e+00 -5.24006903e-01 3.78721476e-01
4.78240401e-02 -4.73193765e-01 -1.70792043e-01 -1.51479453e-01
-9.19031143e-01 -1.27051842e+00 -6.31181180e-01 -1.06106985e+00
7.37891436e-01 2.08036780e-01 8.12790036e-01 5.26559353e-01
-4.63854909e-01 2.16380045e-01 -4.49272037e-01 -2.21292436e-01
-3.36892098e-01 2.40970522e-01 -1.39145032e-01 -6.78418338e-01
8.34057331e-02 -2.41757274e-01 -5.63685715e-01 3.63564551e-01
-9.51872051e-01 3.63802493e-01 8.85247469e-01 6.84783280e-01
6.23247206e-01 5.74997306e-01 4.10170853e-01 -9.73455459e-02
8.80698323e-01 -1.01668693e-01 -3.71423751e-01 6.57564163e-01
-3.67636323e-01 7.45027289e-02 7.62466252e-01 -6.70716703e-01
-8.19629550e-01 -1.78032108e-02 -1.12973601e-01 -6.46673262e-01
1.59368604e-01 1.10495746e+00 -3.39782476e-01 3.83873969e-01
2.41317153e-01 2.24440441e-01 1.02007538e-02 -6.56720042e-01
4.44533706e-01 8.02713037e-01 -9.87823755e-02 -8.85721743e-01
7.84747720e-01 -2.06237972e-01 -1.02657406e-02 -5.62239110e-01
-5.38699627e-01 1.38880894e-01 -6.16692066e-01 -5.74838698e-01
3.90179396e-01 -6.06385887e-01 -5.39496541e-01 5.29070854e-01
-1.31984580e+00 2.12608367e-01 -3.41675669e-01 3.35518241e-01
-5.78628182e-01 1.74753740e-01 -6.89833522e-01 -5.67813039e-01
-7.52710879e-01 -1.27058792e+00 1.35382652e+00 2.50587948e-02
-3.98412496e-01 -3.29377264e-01 -5.32267511e-01 2.52572834e-01
5.35131395e-01 -3.93208629e-03 1.63257515e+00 -3.36000472e-01
-5.11256635e-01 -8.85152578e-01 2.20492221e-02 2.84806669e-01
4.17550325e-01 5.29404700e-01 -4.70042408e-01 -2.77738851e-02
-4.37216423e-02 -1.50262535e-01 2.15077564e-01 3.38272840e-01
1.44825089e+00 -1.27942622e-01 -4.95745212e-01 2.61304304e-02
1.21279407e+00 7.76100218e-01 6.90656364e-01 1.87497154e-01
6.84969604e-01 6.20859146e-01 1.11162031e+00 6.79636002e-01
5.82406335e-02 5.07083356e-01 1.96654558e-01 2.60712594e-01
-6.44897744e-02 -4.47786808e-01 -2.45264918e-02 1.11413097e+00
1.26679644e-01 -6.70447275e-02 -9.21003401e-01 2.49072239e-01
-1.67838812e+00 -6.67701006e-01 1.81493491e-01 1.74580705e+00
6.33786082e-01 2.65781254e-01 -3.26823026e-01 2.85868526e-01
7.35487819e-01 -2.84932405e-01 -8.51282895e-01 -5.60450740e-02
2.55726367e-01 -1.58384368e-01 -1.48234457e-01 -1.73256889e-01
-4.98648226e-01 7.33148634e-01 5.98917294e+00 1.10313165e+00
-8.09424043e-01 -7.38727510e-01 2.59842426e-01 -2.35030532e-01
-5.55837333e-01 -4.53462005e-02 -7.17998266e-01 2.32232690e-01
1.33705258e-01 -1.99149534e-01 4.73155707e-01 1.13172126e+00
-7.15211704e-02 1.90274283e-01 -1.26689231e+00 1.14919150e+00
1.59553245e-01 -1.58128822e+00 6.77838981e-01 -2.03393415e-01
4.25337642e-01 -6.08759940e-01 -1.11136921e-02 4.73354012e-01
-5.32714315e-02 -7.29120135e-01 8.07106972e-01 8.71371448e-01
8.99822056e-01 -8.09543014e-01 8.64642620e-01 1.12770580e-01
-1.39737260e+00 -3.07715327e-01 -4.09531921e-01 2.27486059e-01
-2.40259096e-02 5.84774077e-01 -6.76015735e-01 9.64812398e-01
6.00768447e-01 9.33771908e-01 -3.78992349e-01 9.03341532e-01
5.81756309e-02 -1.97610468e-01 -5.38383543e-01 -4.98484820e-01
-6.11924492e-02 1.15411565e-01 1.02916032e-01 2.42057830e-01
8.71980429e-01 -4.90899235e-02 -9.40862522e-02 1.13578689e+00
3.21647190e-02 2.01327384e-01 -7.29191005e-01 -2.66808599e-01
6.89500809e-01 9.83808219e-01 -5.41157186e-01 -2.13007659e-01
5.44434972e-02 3.75060678e-01 -4.28388752e-02 3.95328194e-01
-7.74682879e-01 -6.38211727e-01 1.79655254e-01 1.38248861e-01
5.22173762e-01 -1.31276071e-01 -1.32357571e-02 -6.35055542e-01
3.98602545e-01 -7.92794347e-01 1.06898509e-01 -1.47970200e+00
-1.42398274e+00 3.23329180e-01 4.26971793e-01 -1.75801909e+00
-2.23543659e-01 -6.49300277e-01 -3.70941848e-01 6.08732820e-01
-7.93056071e-01 -1.27053559e+00 -3.30948055e-01 9.51488093e-02
9.00841892e-01 -3.37119251e-01 6.51118100e-01 2.77439028e-01
-5.71721792e-01 2.10515752e-01 1.38020054e-01 1.09700166e-01
2.30469748e-01 -4.59395498e-01 1.45631403e-01 3.14535536e-02
-3.67355764e-01 1.01150656e+00 3.94877374e-01 -9.70704019e-01
-2.05561614e+00 -7.40506411e-01 4.51070517e-01 -1.38119325e-01
5.39696634e-01 -1.51358441e-01 -6.03704870e-01 2.37040073e-01
-3.87625962e-01 -4.95179087e-01 5.42367578e-01 -1.07053645e-01
6.02958798e-02 -1.87057242e-01 -9.81692016e-01 7.77509987e-01
1.03727436e+00 -5.51793754e-01 -7.56235898e-01 3.87792498e-01
9.18107092e-01 -6.43158436e-01 -1.30492067e+00 6.59580290e-01
8.51836801e-01 -8.39409530e-02 9.45483685e-01 -5.16170979e-01
8.71757329e-01 -3.94969374e-01 -3.57193738e-01 -1.08762312e+00
-4.88408506e-01 -1.21407390e-01 -2.29209200e-01 1.25294530e+00
2.87632883e-01 -2.74263263e-01 4.39365417e-01 7.16591179e-01
-4.53548700e-01 -1.20119226e+00 -7.76673019e-01 -5.09787142e-01
-1.46892086e-01 -4.77364153e-01 1.30693161e+00 4.62922961e-01
-2.01138809e-01 3.04652959e-01 4.22529988e-02 -1.19612731e-01
-2.35922877e-02 8.07991147e-01 6.49330020e-01 -1.24331903e+00
8.26891661e-02 -6.88435614e-01 -5.11344075e-01 -1.27891278e+00
9.05341953e-02 -8.76356781e-01 6.53931722e-02 -1.93818748e+00
-5.46456687e-02 -4.47223932e-01 -2.06883758e-01 5.64413548e-01
3.50145608e-01 -4.77459937e-01 6.96584880e-02 3.47406358e-01
-2.00790659e-01 9.60433722e-01 1.66679919e+00 -8.81573796e-01
-1.72189832e-01 8.44007730e-02 -5.67115128e-01 3.29783618e-01
7.70680666e-01 -1.48925439e-01 -5.31795442e-01 -4.90959287e-01
8.31204414e-01 2.52775908e-01 7.95159414e-02 -9.18566227e-01
2.80552506e-01 -1.02478012e-01 6.29870057e-01 -1.42084074e+00
3.15466434e-01 -1.29552102e+00 6.29797995e-01 5.46966434e-01
-3.83176744e-01 1.78301781e-01 3.70109320e-01 2.17569023e-01
-1.38283834e-01 -4.95690197e-01 1.60793997e-02 6.21536863e-04
-8.95376980e-01 3.19142967e-01 -3.05403054e-01 -4.28495079e-01
6.93507254e-01 -3.69937122e-01 -2.06448436e-01 -1.90944225e-01
-5.38027883e-01 2.89320379e-01 2.73615748e-01 9.12643552e-01
1.39663696e+00 -1.76355946e+00 -3.49399984e-01 3.31491828e-01
3.94954711e-01 3.39475781e-01 5.96934736e-01 1.68523237e-01
-8.22881103e-01 3.49666327e-01 -1.09855339e-01 -2.99746126e-01
-9.32738781e-01 7.73695171e-01 4.69216257e-01 2.14784577e-01
-8.38780820e-01 3.22966069e-01 -2.01190218e-01 -2.94916362e-01
2.19998628e-01 -7.33342230e-01 -3.23545307e-01 9.20106396e-02
1.38019159e-01 2.11380750e-01 1.54411733e-01 -2.43653044e-01
-4.07314837e-01 1.09783196e+00 -1.92085251e-01 3.49600196e-01
1.52390850e+00 2.16869503e-01 -2.37638593e-01 3.61292988e-01
1.22549808e+00 -5.00140429e-01 -7.59857893e-01 -5.45201935e-02
-2.35828623e-01 -1.65326416e-01 6.37568384e-02 -8.09916735e-01
-1.36926293e+00 6.06096566e-01 4.34622049e-01 -5.45587502e-02
1.02645040e+00 1.37165457e-01 9.83346164e-01 9.71612096e-01
6.44445181e-01 -1.33475125e+00 6.57241419e-02 5.07865012e-01
1.57101464e+00 -7.50486910e-01 2.95374930e-01 -4.27574605e-01
-1.96908042e-01 1.58286786e+00 6.44020259e-01 3.81887406e-02
8.08610141e-01 5.71332164e-02 -3.72282743e-01 -8.70504856e-01
-5.13485134e-01 2.05415621e-01 6.48651600e-01 3.19275379e-01
2.55388528e-01 -5.16478084e-02 1.00646533e-01 8.01633596e-01
-4.94185299e-01 7.09754452e-02 -2.31558651e-01 1.21569264e+00
-3.01610500e-01 -7.26006508e-01 -3.20593491e-02 7.59589255e-01
3.16346318e-01 1.44475838e-03 -5.79064786e-01 1.00813758e+00
-2.95287799e-02 7.40837395e-01 -2.21507564e-01 -8.30091000e-01
7.46288002e-01 -1.23585857e-01 5.10030329e-01 -6.14835083e-01
-5.22204280e-01 -1.43658713e-01 -6.31079897e-02 -3.68979543e-01
-3.16372588e-02 -2.24976957e-01 -1.50317061e+00 -2.28868589e-01
-7.12725878e-01 -5.20536955e-03 7.65091300e-01 1.02161872e+00
4.25436437e-01 7.90997982e-01 7.27070928e-01 -7.68175721e-01
-5.38644552e-01 -9.26079392e-01 -4.45412189e-01 3.32416058e-01
-1.65990815e-01 -8.75221610e-01 -1.15935549e-01 -3.51927519e-01] | [5.927919387817383, 3.1253557205200195] |
5a25dbfa-a5a4-4ee5-936f-68c5f42644f4 | metricprompt-prompting-model-as-a-relevance | 2306.08892 | null | https://arxiv.org/abs/2306.08892v1 | https://arxiv.org/pdf/2306.08892v1.pdf | MetricPrompt: Prompting Model as a Relevance Metric for Few-shot Text Classification | Prompting methods have shown impressive performance in a variety of text mining tasks and applications, especially few-shot ones. Despite the promising prospects, the performance of prompting model largely depends on the design of prompt template and verbalizer. In this work, we propose MetricPrompt, which eases verbalizer design difficulty by reformulating few-shot text classification task into text pair relevance estimation task. MetricPrompt adopts prompting model as the relevance metric, further bridging the gap between Pre-trained Language Model's (PLM) pre-training objective and text classification task, making possible PLM's smooth adaption. Taking a training sample and a query one simultaneously, MetricPrompt captures cross-sample relevance information for accurate relevance estimation. We conduct experiments on three widely used text classification datasets across four few-shot settings. Results show that MetricPrompt outperforms manual verbalizer and other automatic verbalizer design methods across all few-shot settings, achieving new state-of-the-art (SOTA) performance. | ['Wanxiang Che', 'Weinan Zhang', 'Hongyuan Dong'] | 2023-06-15 | null | null | null | null | ['classification-1', 'text-classification', 'few-shot-text-classification'] | ['methodology', 'natural-language-processing', 'natural-language-processing'] | [ 3.94386441e-01 -1.68474931e-02 -6.35680318e-01 -4.87457275e-01
-1.09132659e+00 -1.61770746e-01 9.40387666e-01 6.27649248e-01
-8.04591238e-01 4.93779391e-01 5.63407362e-01 -2.33123437e-01
-1.41196534e-01 -3.73499662e-01 1.33206144e-01 -2.84992903e-01
6.41134381e-01 6.26289368e-01 4.24678594e-01 -4.29449230e-01
6.86359048e-01 -3.20784092e-01 -1.51603580e+00 2.64328212e-01
9.78113472e-01 7.77903318e-01 4.69100267e-01 9.58928704e-01
-6.14300489e-01 7.19734073e-01 -8.62349749e-01 -3.56090486e-01
-1.01667531e-01 -4.53682959e-01 -9.81824994e-01 -2.80744702e-01
2.85224527e-01 -2.09071815e-01 -1.46995351e-01 8.04569542e-01
9.06858385e-01 6.36503100e-01 7.54777491e-01 -1.15219414e+00
-8.47030520e-01 1.13209164e+00 -6.50318265e-01 5.63210547e-01
6.65286720e-01 1.32303149e-01 1.41425920e+00 -1.25468850e+00
3.89779747e-01 1.19847238e+00 5.27347803e-01 5.47015071e-01
-9.79368210e-01 -5.64290404e-01 2.91601896e-01 5.28629363e-01
-1.23193824e+00 -5.14018476e-01 5.60946584e-01 -3.40359539e-01
1.21153641e+00 4.29427087e-01 3.19636166e-01 1.15468991e+00
1.39282137e-01 9.89232302e-01 7.37084031e-01 -8.73868644e-01
3.99824142e-01 1.74444571e-01 9.34783518e-01 3.24016780e-01
2.93562077e-02 -2.45899424e-01 -8.75030220e-01 -3.96538794e-01
5.80668598e-02 6.32571951e-02 -3.00313979e-01 3.12217504e-01
-9.72920775e-01 8.63144636e-01 -1.93003520e-01 3.55493426e-01
-2.63493329e-01 -2.07146272e-01 5.10201454e-01 6.17604911e-01
5.68943381e-01 7.08632708e-01 -4.15339917e-01 -3.78016531e-01
-1.14785850e+00 1.66280925e-01 6.92186058e-01 1.20148599e+00
4.32684541e-01 -2.02404127e-01 -1.23110390e+00 1.10183787e+00
2.25429296e-01 3.20531696e-01 1.15498352e+00 -2.24123657e-01
6.72413170e-01 7.65485466e-01 1.12304285e-01 -7.06356287e-01
-4.50745851e-01 -1.80689424e-01 -5.93047023e-01 -4.22312468e-01
5.82803749e-02 -5.33001497e-02 -6.10551655e-01 1.34433544e+00
3.62135470e-01 -1.73057467e-01 -7.88656771e-02 9.03075874e-01
1.14088917e+00 7.64031410e-01 4.56787109e-01 -4.91726190e-01
1.69726944e+00 -1.13797688e+00 -9.63629663e-01 -4.56052482e-01
9.79075193e-01 -9.63548958e-01 1.95325971e+00 2.70108104e-01
-8.23375404e-01 -5.08821785e-01 -1.12586546e+00 -3.92013669e-01
-1.77634224e-01 1.44127861e-01 4.79623765e-01 4.78752106e-01
-6.49135292e-01 4.55641299e-01 -3.44836980e-01 -6.35916173e-01
1.93288982e-01 3.31898890e-02 -7.59842282e-04 6.22548908e-02
-1.51023281e+00 7.61940122e-01 2.72477001e-01 -6.85228467e-01
-4.77578282e-01 -8.32776666e-01 -6.26738846e-01 4.13575172e-01
6.86959863e-01 -6.64697528e-01 1.85935998e+00 -5.03073812e-01
-1.55251479e+00 7.34874964e-01 -1.84861511e-01 -2.59004891e-01
3.54668260e-01 -5.20582616e-01 -4.56985444e-01 -3.39052416e-02
1.94924355e-01 3.57426345e-01 1.05662465e+00 -5.49803317e-01
-7.75932431e-01 -8.47996995e-02 -3.24466556e-01 4.67825592e-01
-9.39477026e-01 4.59244728e-01 -3.78238171e-01 -7.68011332e-01
-3.97930652e-01 -1.80744320e-01 -1.68625027e-01 -6.98058456e-02
-3.91490042e-01 -8.68792832e-01 7.84257531e-01 -2.52381831e-01
2.01248550e+00 -1.93356681e+00 -1.69699833e-01 -2.69286275e-01
2.89536148e-01 5.51483572e-01 -4.66919601e-01 7.13322997e-01
1.06313117e-02 -3.99878360e-02 9.98453051e-02 -6.17507815e-01
3.19985747e-01 -1.41885251e-01 -4.74674523e-01 1.92533731e-01
-7.24964514e-02 1.14204562e+00 -1.04573107e+00 -9.15385962e-01
2.14522734e-01 -2.13914309e-02 -3.30606431e-01 5.84729910e-01
-3.68070811e-01 -3.29501659e-01 -5.54388344e-01 4.09473240e-01
2.33713001e-01 -4.72684681e-01 -1.71786070e-01 3.07556838e-01
1.22460179e-01 5.25018036e-01 -1.02908170e+00 1.63195753e+00
-4.22403455e-01 5.37438393e-01 -3.16775620e-01 -6.88262522e-01
1.09547353e+00 3.69330406e-01 2.49317899e-01 -8.38499308e-01
4.05179769e-01 -2.81043380e-01 -3.00779402e-01 -7.09495544e-01
1.12217057e+00 -8.10569152e-02 -1.61377266e-01 1.07163513e+00
1.50064871e-01 -5.81455454e-02 2.81857789e-01 5.99841952e-01
1.19512546e+00 -2.79387772e-01 1.06379783e+00 -4.12142962e-01
4.21629161e-01 -1.05322033e-01 1.90245852e-01 1.03558922e+00
-4.78020221e-01 5.56875527e-01 3.60021651e-01 -1.11733556e-01
-6.20428205e-01 -5.10271966e-01 3.86700630e-02 2.04114318e+00
2.70155847e-01 -8.74553025e-01 -6.46036386e-01 -8.57612312e-01
-1.39769778e-01 1.12289512e+00 -7.53359735e-01 -4.21767086e-01
-2.87314028e-01 -6.44931853e-01 4.00723904e-01 4.64246333e-01
4.64405641e-02 -1.03089845e+00 -7.35506296e-01 3.45288068e-01
-2.71254808e-01 -8.31291795e-01 -1.16154242e+00 3.61364663e-01
-5.42143404e-01 -9.71235394e-01 -6.30678594e-01 -5.21263778e-01
2.75918305e-01 8.83293569e-01 1.24585795e+00 1.03091799e-01
-2.56051034e-01 4.10092205e-01 -9.59774435e-01 -4.94942665e-01
-1.86632931e-01 4.54795063e-01 -3.30912024e-02 -3.14852029e-01
8.78133476e-01 -2.76757121e-01 -4.37349528e-01 4.42360908e-01
-8.56107414e-01 6.66119903e-02 2.97736228e-01 1.02821326e+00
1.18337996e-01 -4.61607188e-01 7.64025033e-01 -9.26143110e-01
1.38255322e+00 -6.27766848e-01 -2.79510468e-01 5.43292165e-01
-1.19604409e+00 1.76388115e-01 8.06848824e-01 -7.44695604e-01
-1.11786807e+00 -6.03878021e-01 -4.19017002e-02 -5.10726348e-02
6.21918179e-02 5.27603030e-01 7.72788227e-02 5.98784208e-01
1.27127814e+00 5.48940264e-02 -3.72782350e-01 -4.65211511e-01
3.61509740e-01 1.05778944e+00 3.06817383e-01 -5.35748601e-01
5.01138389e-01 -9.26176161e-02 -5.60859740e-01 -8.33187461e-01
-1.26587272e+00 -1.02074444e+00 -3.58976036e-01 -1.54376477e-01
3.72195125e-01 -6.17808461e-01 -3.63580257e-01 -1.28939413e-02
-1.24099410e+00 -4.17731732e-01 -5.02675653e-01 3.00530434e-01
-1.02008328e-01 4.21744376e-01 -3.90830159e-01 -7.91079104e-01
-1.06804454e+00 -7.69477129e-01 1.19267154e+00 3.06542724e-01
-8.80724669e-01 -1.07663715e+00 3.11814517e-01 1.43819556e-01
4.69484270e-01 -5.45132637e-01 7.34681666e-01 -1.15664399e+00
1.67812318e-01 -3.44025940e-01 -2.76584655e-01 -1.74544483e-01
1.50350314e-02 -2.42345914e-01 -1.05864286e+00 -1.52217075e-01
8.21712017e-02 -5.08408606e-01 8.47223580e-01 2.44101658e-01
9.57073092e-01 -4.89434779e-01 -4.08990860e-01 3.78707290e-01
9.86971021e-01 -4.04336788e-02 3.48930299e-01 2.82081425e-01
5.22912323e-01 6.19070172e-01 1.21345651e+00 9.15172994e-01
2.69736618e-01 8.89594436e-01 -4.07257341e-02 1.89718261e-01
-1.41703635e-01 -3.67126882e-01 3.66247118e-01 9.09398437e-01
4.16322231e-01 -5.24346530e-01 -7.75991738e-01 2.43361726e-01
-2.16626954e+00 -1.04167891e+00 -4.01849560e-02 2.00942397e+00
1.23593962e+00 3.82348970e-02 1.77987307e-01 2.10488230e-01
7.52362490e-01 2.80320823e-01 -4.24359173e-01 -3.33781898e-01
9.02156830e-02 1.64179504e-01 -5.55975065e-02 7.70420194e-01
-9.48598146e-01 1.08728039e+00 6.45192194e+00 1.24301851e+00
-9.27040994e-01 3.26616287e-01 2.54187882e-01 -2.60327220e-01
-3.95817280e-01 -1.34604841e-01 -1.16240418e+00 4.84441906e-01
8.39905977e-01 -8.10144842e-01 1.34764254e-01 9.97986078e-01
2.32040569e-01 -1.34455100e-01 -1.24204063e+00 1.12457573e+00
3.96564662e-01 -9.77583647e-01 5.54287173e-02 -4.56937283e-01
5.43345153e-01 -3.21265966e-01 -4.19445969e-02 8.20979357e-01
3.99723351e-01 -9.33506072e-01 6.78768754e-01 3.72088134e-01
9.43971336e-01 -4.80092347e-01 6.86576307e-01 6.20543122e-01
-1.13584781e+00 -3.93501557e-02 -4.15887207e-01 -2.91183591e-01
1.65820703e-01 7.22132266e-01 -1.03666174e+00 2.04357803e-01
3.92094553e-01 6.51829779e-01 -7.32841730e-01 6.93469346e-01
-3.94666225e-01 7.03330159e-01 -5.37191378e-03 -5.96884727e-01
5.58548719e-02 2.51344472e-01 5.87875724e-01 1.59601581e+00
3.46062928e-02 3.51324260e-01 3.02215040e-01 6.32407129e-01
-8.79313499e-02 5.13743460e-01 -2.19974428e-01 -5.99220805e-02
9.48385954e-01 1.47629273e+00 -5.81027865e-01 -5.72536051e-01
-2.65808046e-01 7.85361826e-01 4.83787328e-01 1.81903630e-01
-6.93263710e-01 -7.71635056e-01 3.12960327e-01 -4.39487882e-02
-8.09004828e-02 2.02874690e-01 -5.21501482e-01 -1.15213299e+00
1.16613507e-02 -8.59554946e-01 6.58300996e-01 -5.72149694e-01
-1.36325979e+00 4.50877666e-01 1.82951421e-01 -1.16575408e+00
-3.82640958e-01 -3.69666845e-01 -9.98733819e-01 7.76860118e-01
-1.40141189e+00 -8.90106618e-01 -4.57667798e-01 3.25691372e-01
1.21291757e+00 -2.46284455e-01 7.26792574e-01 2.98310234e-03
-6.37611389e-01 1.13902533e+00 -2.17562735e-01 -3.32653701e-01
1.35720062e+00 -1.16087306e+00 6.57537162e-01 8.21487725e-01
3.09834555e-02 6.70679212e-01 9.23607290e-01 -5.80950797e-01
-1.28028989e+00 -1.02291083e+00 1.30488503e+00 -6.14513993e-01
6.32129788e-01 -3.00889939e-01 -1.14256847e+00 2.80903548e-01
2.61288404e-01 -1.91031054e-01 8.97876620e-01 2.85382390e-01
-4.01390672e-01 1.73812211e-01 -8.56099069e-01 8.05990160e-01
1.10718381e+00 -4.51662272e-01 -9.91582990e-01 3.34032685e-01
1.04175305e+00 -2.04774380e-01 -4.63210523e-01 -1.19624570e-01
3.49697322e-01 -6.02162898e-01 6.56114817e-01 -5.98793745e-01
5.09383321e-01 7.92428330e-02 -3.07710972e-02 -1.37609267e+00
-5.28278649e-01 -1.04236960e+00 -6.18402898e-01 1.42866433e+00
2.37431139e-01 -4.95332539e-01 4.52844113e-01 6.87914908e-01
-2.02021897e-01 -8.30213666e-01 -6.49712980e-01 -7.16749072e-01
-1.44698650e-01 -4.98622090e-01 6.46147668e-01 1.05048704e+00
8.56611669e-01 1.18256140e+00 -3.34183633e-01 -6.77663445e-01
3.08335304e-01 -1.83095202e-01 8.72615933e-01 -1.30467522e+00
-4.96014655e-01 -7.20044672e-01 1.72627002e-01 -1.31935763e+00
1.28951609e-01 -9.30547595e-01 2.16737911e-01 -1.42599225e+00
5.27116835e-01 -1.67692184e-01 -1.61375359e-01 7.83144712e-01
-1.04524875e+00 -3.26934278e-01 5.48389517e-02 2.36626804e-01
-1.05914187e+00 9.21375692e-01 1.10273290e+00 -2.51293808e-01
-4.95170385e-01 -1.33147076e-01 -1.05076444e+00 4.31087196e-01
7.47312784e-01 -6.19943500e-01 -6.78160131e-01 -3.07713002e-01
1.53428406e-01 -9.83986771e-04 -1.16519287e-01 -5.53673148e-01
6.42056406e-01 -4.85022694e-01 2.29342114e-02 -5.22749960e-01
-4.63914536e-02 -3.58086795e-01 -6.61919236e-01 1.07790031e-01
-9.85445619e-01 2.16146499e-01 -1.59596335e-02 4.58735168e-01
1.11670405e-01 -6.07186198e-01 8.10831606e-01 2.62265980e-01
-7.35086739e-01 3.79638731e-01 -3.23057860e-01 6.27269447e-01
7.90515780e-01 -2.08366737e-01 -5.96184850e-01 -3.12568009e-01
4.34206091e-02 4.66171116e-01 2.86796600e-01 6.64258540e-01
6.31714582e-01 -1.16683960e+00 -7.16903687e-01 1.66097015e-01
6.67795300e-01 -2.41653770e-01 1.19323581e-01 7.62405753e-01
2.34464556e-01 4.50753003e-01 2.18860656e-01 -4.21863496e-01
-1.40206420e+00 6.79847598e-01 -1.65518105e-01 -5.83260238e-01
-8.32395256e-01 8.62272680e-01 1.93054806e-02 -1.44237682e-01
5.41384161e-01 -1.78916827e-01 -5.08586586e-01 2.28064865e-01
1.18442690e+00 5.13882697e-01 1.36065930e-01 -8.50045010e-02
-8.34770575e-02 2.17173100e-01 -6.04302764e-01 -2.86439300e-01
8.53075624e-01 -2.88617194e-01 2.96152562e-01 5.43669760e-01
9.89271283e-01 -2.92695105e-01 -1.02738571e+00 -6.76386774e-01
3.32086623e-01 -4.58250076e-01 2.20186964e-01 -6.78674400e-01
-2.39326805e-01 8.23938072e-01 1.10293351e-01 1.42131835e-01
8.29051375e-01 -1.28160000e-01 8.95756721e-01 6.07800722e-01
1.87588319e-01 -1.51592243e+00 4.95464057e-01 9.42972720e-01
7.33799100e-01 -1.38294160e+00 1.60875872e-01 -9.85419378e-02
-8.59645784e-01 9.10362959e-01 8.33616197e-01 2.58452952e-01
4.25207347e-01 2.36803025e-01 1.37936667e-01 -1.66618377e-01
-1.34149575e+00 -9.82655287e-02 5.49015343e-01 5.73237479e-01
7.04189003e-01 1.24349827e-02 -5.78015447e-01 1.21606839e+00
-3.69655758e-01 -1.32704005e-01 2.16951922e-01 9.95554984e-01
-9.16969299e-01 -9.60157990e-01 -3.10124546e-01 6.30094230e-01
-9.99800861e-02 -4.12514657e-01 -5.33679247e-01 5.28024077e-01
-4.83806163e-01 1.19641995e+00 -1.47302061e-01 -4.41433460e-01
3.36017817e-01 2.25656837e-01 1.77233204e-01 -1.04420984e+00
-8.44606340e-01 -2.09503308e-01 1.23793352e-02 -4.02350098e-01
1.65776938e-01 -4.92682248e-01 -1.17272735e+00 -3.20003569e-01
-6.95408881e-01 4.39745903e-01 2.14845553e-01 1.20616686e+00
3.38727266e-01 5.68296671e-01 5.93300939e-01 -7.12852061e-01
-1.21999860e+00 -1.55052614e+00 -4.38618362e-01 4.27902281e-01
1.30082071e-01 -7.12594748e-01 -3.21418405e-01 -8.86202306e-02] | [10.730785369873047, 7.755161285400391] |
2c7d1b89-40c5-48cf-86be-733351ef1823 | esresnet-environmental-sound-classification | 2004.07301 | null | https://arxiv.org/abs/2004.07301v1 | https://arxiv.org/pdf/2004.07301v1.pdf | ESResNet: Environmental Sound Classification Based on Visual Domain Models | Environmental Sound Classification (ESC) is an active research area in the audio domain and has seen a lot of progress in the past years. However, many of the existing approaches achieve high accuracy by relying on domain-specific features and architectures, making it harder to benefit from advances in other fields (e.g., the image domain). Additionally, some of the past successes have been attributed to a discrepancy of how results are evaluated (i.e., on unofficial splits of the UrbanSound8K (US8K) dataset), distorting the overall progression of the field. The contribution of this paper is twofold. First, we present a model that is inherently compatible with mono and stereo sound inputs. Our model is based on simple log-power Short-Time Fourier Transform (STFT) spectrograms and combines them with several well-known approaches from the image domain (i.e., ResNet, Siamese-like networks and attention). We investigate the influence of cross-domain pre-training, architectural changes, and evaluate our model on standard datasets. We find that our model out-performs all previously known approaches in a fair comparison by achieving accuracies of 97.0 % (ESC-10), 91.5 % (ESC-50) and 84.2 % / 85.4 % (US8K mono / stereo). Second, we provide a comprehensive overview of the actual state of the field, by differentiating several previously reported results on the US8K dataset between official or unofficial splits. For better reproducibility, our code (including any re-implementations) is made available. | ['Jörn Hees', 'Andrey Guzhov', 'Federico Raue', 'Andreas Dengel'] | 2020-04-15 | null | null | null | null | ['environmental-sound-classification', 'sound-classification'] | ['audio', 'audio'] | [ 1.71533182e-01 -3.52359474e-01 3.73979092e-01 -2.61640221e-01
-8.56338084e-01 -5.89915693e-01 5.28024256e-01 -1.74133956e-01
-6.54071152e-01 5.99533379e-01 3.24515730e-01 -1.11883253e-01
-2.19657093e-01 -5.90250731e-01 -5.26432693e-01 -6.61146998e-01
-1.78407773e-01 7.39031062e-02 4.85359907e-01 -4.29774612e-01
2.20324978e-01 3.27421039e-01 -1.87587929e+00 4.66860533e-01
5.64581871e-01 1.07366526e+00 1.34863973e-01 9.06496882e-01
9.69546586e-02 4.50448811e-01 -7.58271694e-01 -3.03660393e-01
2.54818648e-01 -4.59165603e-01 -1.06257713e+00 -4.96478945e-01
7.77715445e-01 -1.13119408e-01 -3.87562305e-01 9.80331182e-01
8.89156461e-01 1.43967986e-01 4.76549298e-01 -1.16853714e+00
-3.60298216e-01 5.64661682e-01 -2.76489705e-01 3.81428480e-01
1.91891342e-01 1.20855346e-01 1.13811409e+00 -7.59963155e-01
4.59176660e-01 1.00736725e+00 9.93122995e-01 2.16153577e-01
-1.26300502e+00 -1.01952374e+00 -7.76270479e-02 6.00402832e-01
-1.46955907e+00 -7.14396000e-01 6.69385552e-01 -5.02800405e-01
1.02183139e+00 3.44203472e-01 5.43590128e-01 1.19805908e+00
-1.14982076e-01 5.20196438e-01 1.29792798e+00 -5.55165529e-01
3.55646670e-01 -3.60484608e-02 2.13603228e-02 9.30650458e-02
-4.55767885e-02 3.24607730e-01 -7.60849953e-01 -7.75023503e-03
5.03132582e-01 -6.14448905e-01 -5.32623112e-01 -1.89506367e-01
-1.04456043e+00 6.77760303e-01 4.22297537e-01 6.79435313e-01
-4.02819132e-03 1.52355924e-01 6.01235211e-01 5.49558103e-01
5.52096963e-01 4.84197021e-01 -5.15772700e-01 -5.25985718e-01
-1.29970145e+00 4.07453030e-01 7.48631179e-01 5.60704112e-01
6.62063479e-01 2.79005617e-01 3.28896880e-01 1.26469588e+00
6.14401549e-02 5.12607872e-01 9.11358416e-01 -1.09812057e+00
5.41769385e-01 -1.52543128e-01 -8.74031410e-02 -1.16006613e+00
-4.56499398e-01 -5.92943668e-01 -6.92713082e-01 3.09288055e-01
4.14132059e-01 -4.98342328e-02 -7.59507537e-01 1.65425766e+00
-5.71684772e-03 4.23274964e-01 -2.56040762e-03 8.81760418e-01
9.38403428e-01 5.92797816e-01 -1.87781960e-01 1.97678342e-01
1.08721542e+00 -8.91797841e-01 -4.25673395e-01 -8.68686512e-02
2.98477679e-01 -8.96945179e-01 1.10114217e+00 5.79028249e-01
-7.81792164e-01 -8.25912714e-01 -1.08931696e+00 2.39233121e-01
-6.55645311e-01 4.67114411e-02 1.47772983e-01 7.72953093e-01
-1.36274445e+00 9.07698512e-01 -7.78569400e-01 -6.27240360e-01
1.57456279e-01 3.29286456e-01 -3.98280144e-01 1.72615901e-01
-1.37891340e+00 6.73321366e-01 2.30953962e-01 -1.60480186e-01
-6.76443160e-01 -8.60710263e-01 -5.69985449e-01 -1.59092396e-01
2.27636248e-01 -2.40395799e-01 1.54364634e+00 -9.50909793e-01
-1.76747286e+00 6.58627570e-01 2.05322400e-01 -6.66972280e-01
6.11913860e-01 -5.26542485e-01 -6.32523477e-01 3.08391899e-01
2.59520728e-02 6.48661435e-01 8.31542790e-01 -1.07263637e+00
-8.80663276e-01 -2.06603974e-01 7.86326975e-02 -2.63191685e-02
-4.21660542e-01 1.12751320e-01 -8.02096277e-02 -8.08789253e-01
-1.54107600e-01 -1.09909332e+00 4.38087061e-02 -3.27539653e-01
-1.44444898e-01 8.72101542e-03 6.38905764e-01 -6.59452140e-01
1.32672679e+00 -2.46857285e+00 -5.16236685e-02 -1.08658858e-01
-5.04731089e-02 4.59598899e-01 -2.27642477e-01 6.04737222e-01
-4.37198490e-01 7.61491954e-02 -4.34621036e-01 -3.79804671e-01
3.58938612e-02 4.05225679e-02 -4.34542179e-01 4.88058567e-01
-4.08649519e-02 4.07511353e-01 -9.07756627e-01 -1.31446794e-01
3.33235830e-01 4.80053961e-01 -6.20155692e-01 -1.87934086e-01
2.66261667e-01 2.80565381e-01 4.22054082e-02 2.44759232e-01
6.45052969e-01 2.33507350e-01 -1.85191333e-01 -2.62818515e-01
-5.21522701e-01 3.50890636e-01 -1.46347904e+00 1.74356735e+00
-7.85799325e-01 9.61690247e-01 1.92176625e-01 -1.02401638e+00
5.71720362e-01 5.05079627e-01 4.68713284e-01 -6.95339739e-01
-8.55089128e-02 6.07331097e-01 1.07871816e-01 -4.31562006e-01
6.80561066e-01 -7.42472336e-02 1.28283918e-01 2.45965078e-01
2.97129601e-01 -3.23386729e-01 1.67503908e-01 -7.35301003e-02
9.32111204e-01 1.07715381e-02 2.45890215e-01 -3.76084864e-01
4.57926065e-01 -1.49780199e-01 1.76550612e-01 7.08530664e-01
-2.92809486e-01 1.21587038e+00 1.81716964e-01 -2.38960475e-01
-9.87428427e-01 -8.96936893e-01 -4.38535750e-01 1.11720181e+00
-1.48919538e-01 -5.96288443e-01 -7.97502398e-01 -4.51452583e-01
-3.47437002e-02 6.68562412e-01 -6.23756647e-01 2.43236888e-02
-6.20436847e-01 -7.10789859e-01 9.73272383e-01 5.80784440e-01
7.18496323e-01 -1.09519839e+00 -7.96406686e-01 3.49710584e-01
-2.10603088e-01 -1.09692049e+00 -6.81555830e-03 3.28153670e-01
-5.74071586e-01 -8.59815598e-01 -9.47115123e-01 -3.84627879e-01
-2.33756378e-01 3.06221217e-01 1.15200543e+00 -3.30230296e-01
-1.14112154e-01 5.52848518e-01 -6.47679687e-01 -3.59302610e-01
-4.59473938e-01 3.45603973e-01 1.08147420e-01 1.05564654e-01
1.12353049e-01 -9.49897170e-01 -5.66028774e-01 3.93228799e-01
-9.01223004e-01 -1.27675220e-01 3.45064819e-01 7.72006035e-01
3.51400793e-01 1.26099914e-01 6.75517380e-01 -6.02037072e-01
4.33279514e-01 -3.52119535e-01 -3.95120710e-01 -2.61704236e-01
-5.17030895e-01 -2.39140436e-01 7.83637822e-01 -2.04928517e-01
-8.80544126e-01 -1.83886573e-01 -5.35448849e-01 -3.30779940e-01
-6.88939393e-01 3.81330371e-01 1.84950173e-01 -1.37016578e-02
8.65756631e-01 2.08170936e-01 -2.54872829e-01 -8.26785922e-01
1.16952866e-01 1.04561508e+00 5.95993757e-01 -4.60527986e-01
6.78057790e-01 5.53211629e-01 -3.64925086e-01 -1.13937294e+00
-7.26414382e-01 -5.07238090e-01 -5.27586341e-01 -2.27318108e-01
8.27922463e-01 -8.21051896e-01 -2.12027565e-01 7.80635953e-01
-8.25249135e-01 -4.28704798e-01 -4.27837104e-01 7.07519174e-01
-4.41390157e-01 2.32201889e-01 -4.11986470e-01 -8.12438846e-01
-1.53476540e-02 -1.07104743e+00 1.10735059e+00 1.36650177e-02
-4.68842924e-01 -7.83087492e-01 3.44311744e-01 2.59364784e-01
7.56286502e-01 -3.99976457e-03 5.32452166e-01 -6.40663564e-01
-3.43286023e-02 7.84884319e-02 -1.79357931e-01 7.03382909e-01
-3.43064032e-02 5.02908491e-02 -1.60205054e+00 -1.24050997e-01
-5.90597391e-02 -3.20041567e-01 1.16821611e+00 3.10226262e-01
9.97039914e-01 7.21394494e-02 1.42213836e-01 5.20701289e-01
1.29111743e+00 2.52425015e-01 6.27538860e-01 5.81633151e-01
3.76533359e-01 7.20458031e-01 2.81098992e-01 2.58706182e-01
3.02099049e-01 8.82710755e-01 4.58991885e-01 6.46848651e-03
-5.02219141e-01 -8.53313431e-02 4.02166516e-01 1.00199747e+00
-3.37234050e-01 -7.55885690e-02 -1.00016499e+00 7.26170182e-01
-1.50310349e+00 -9.68329430e-01 -1.58396453e-01 2.05801868e+00
6.84371293e-01 1.53089210e-03 2.26273790e-01 8.25434148e-01
5.69681823e-01 4.24003899e-01 -1.74007997e-01 -2.74120182e-01
-1.09211758e-01 6.02048397e-01 3.11383307e-01 3.65842849e-01
-1.36826432e+00 7.54386127e-01 6.38613224e+00 1.19004965e+00
-1.45586336e+00 1.75582647e-01 4.85200614e-01 -1.42147347e-01
1.84404686e-01 -2.97343105e-01 -5.18020451e-01 4.74541366e-01
1.18463266e+00 1.49356306e-01 5.71297050e-01 6.32764399e-01
1.11904040e-01 -1.53363705e-01 -9.56264555e-01 1.14971244e+00
7.85444006e-02 -1.02176332e+00 -4.27148491e-01 4.06482071e-03
6.52894795e-01 4.76731449e-01 2.33521193e-01 3.56904417e-01
1.36263043e-01 -9.58472073e-01 1.07825768e+00 1.76599264e-01
1.00486982e+00 -6.04891062e-01 8.55045199e-01 2.61455718e-02
-1.31617880e+00 -1.28010407e-01 -2.60898232e-01 -1.49166033e-01
-2.06420049e-02 5.72465122e-01 -4.02649850e-01 9.25831497e-01
1.39163566e+00 7.54313290e-01 -5.23035705e-01 1.07010460e+00
1.26692941e-02 1.02604127e+00 -5.62003016e-01 2.11385325e-01
3.64638060e-01 2.39367876e-02 6.92618549e-01 1.55093229e+00
4.53944027e-01 -3.32932442e-01 -1.25006706e-01 4.62378412e-01
1.45441145e-01 3.92415583e-01 -4.68538493e-01 2.02282995e-01
2.07420319e-01 1.10381484e+00 -5.85176945e-01 -1.46378160e-01
-5.84100485e-01 6.83045506e-01 -1.38471453e-02 3.26267183e-01
-8.64416063e-01 -6.32857323e-01 8.37614775e-01 6.15211800e-02
5.31445384e-01 -7.03645796e-02 -2.70877868e-01 -9.34869468e-01
-4.68219034e-02 -9.49598968e-01 3.83600444e-01 -8.66467893e-01
-1.19593024e+00 8.21593344e-01 2.11312965e-01 -1.39921606e+00
-1.07212946e-01 -4.75415021e-01 -3.85009706e-01 6.74175024e-01
-1.81384408e+00 -6.81153834e-01 -3.20861489e-01 2.66045064e-01
5.66332459e-01 -2.46257842e-01 8.88261378e-01 7.11140931e-01
-2.31548727e-01 6.17862165e-01 3.37510109e-01 1.23457611e-01
9.21346486e-01 -1.20275772e+00 4.16874796e-01 5.80632150e-01
5.15943468e-01 3.30239296e-01 7.78958142e-01 6.12213183e-03
-6.45833075e-01 -1.00076258e+00 7.81723619e-01 -1.79806039e-01
8.79011571e-01 -3.03282261e-01 -8.43075693e-01 2.28401437e-01
3.70706528e-01 -3.01564913e-02 5.98278582e-01 8.31495374e-02
-5.10407209e-01 -4.55178976e-01 -8.73530507e-01 5.14180958e-01
1.13171482e+00 -5.13533831e-01 -4.84480292e-01 -1.19005814e-01
5.15285611e-01 -3.96671146e-01 -7.56443024e-01 3.55265319e-01
7.81455815e-01 -1.59140134e+00 9.79403496e-01 -2.03398243e-01
5.77698231e-01 -2.92028606e-01 -5.51775157e-01 -1.55818653e+00
-3.54025334e-01 -4.36891854e-01 3.05097133e-01 1.25191081e+00
3.21504682e-01 -8.78934205e-01 4.65127170e-01 -2.14579701e-02
-4.21390980e-01 -5.88721991e-01 -1.28129852e+00 -9.62725937e-01
2.84814060e-01 -9.56156611e-01 6.46734357e-01 9.35442030e-01
-2.49083742e-01 2.64229804e-01 -2.56258726e-01 5.43503873e-02
1.72096863e-01 1.04297973e-01 7.80125082e-01 -1.26883590e+00
-3.42088938e-01 -6.17626905e-01 -6.99117005e-01 -8.78554344e-01
4.71406095e-02 -7.72931099e-01 -4.08581346e-02 -1.25747859e+00
-1.47750139e-01 -3.70935112e-01 -4.30527061e-01 4.36175644e-01
2.56054193e-01 6.79406643e-01 2.79105484e-01 1.69321612e-01
-3.22185546e-01 4.64182436e-01 8.93494725e-01 5.17767342e-03
-1.55734926e-01 -8.94241109e-02 -5.75843036e-01 8.60310137e-01
1.02146995e+00 -4.51278031e-01 -2.01458991e-01 -5.05789459e-01
8.85420218e-02 -3.04129869e-01 6.32545650e-01 -1.66802120e+00
1.17423162e-01 1.66652203e-01 1.30365819e-01 -3.56683403e-01
3.88509244e-01 -8.10357749e-01 2.31778696e-01 2.45985284e-01
-3.53596509e-01 -1.27794266e-01 2.30395898e-01 4.16787922e-01
-7.05786765e-01 -3.41207027e-01 1.04685342e+00 -6.72355145e-02
-7.16911316e-01 -1.99009463e-01 -3.87131274e-01 3.05276871e-01
5.92346907e-01 -3.16262007e-01 -1.87426612e-01 -6.13594592e-01
-6.42034054e-01 -3.16730469e-01 2.71624714e-01 4.45477098e-01
1.03097059e-01 -1.08709776e+00 -8.37147415e-01 1.90019473e-01
1.13497414e-02 -2.37459913e-01 5.34319103e-01 9.59249496e-01
-5.26096761e-01 4.83948350e-01 -1.36632621e-01 -6.89495921e-01
-1.28841567e+00 -6.40623793e-02 4.93047804e-01 -1.63753375e-01
-6.70817673e-01 7.97915697e-01 2.12916404e-01 -4.80191261e-01
2.00123921e-01 -4.25091356e-01 -3.84340316e-01 2.99626529e-01
4.33455169e-01 6.17174149e-01 4.34445947e-01 -8.03541780e-01
-4.26382154e-01 7.29893029e-01 2.20525250e-01 -3.82766426e-01
1.69374287e+00 1.24909125e-01 1.87170282e-01 7.06379950e-01
1.41116869e+00 2.49522671e-01 -9.74010587e-01 -1.29822940e-01
-2.54362971e-01 -3.91092986e-01 9.98403430e-02 -7.88823962e-01
-1.28923738e+00 1.16150045e+00 8.34856391e-01 5.70226610e-01
1.29055643e+00 -1.20550871e-01 6.59356356e-01 1.56766549e-01
3.11159253e-01 -1.22531426e+00 -1.05224714e-01 7.79606640e-01
1.06470847e+00 -9.84572768e-01 -1.62852973e-01 -2.19109565e-01
-4.47042018e-01 1.13376451e+00 3.31292540e-01 -1.23901457e-01
7.99388647e-01 3.39991391e-01 3.38461429e-01 1.02165259e-01
-4.92557615e-01 -1.91606879e-01 3.23092081e-02 7.13593602e-01
5.44737816e-01 -1.00033090e-01 -3.12470105e-02 7.00437784e-01
-8.12140942e-01 -1.15542233e-01 4.50036645e-01 8.12497437e-01
-3.13708752e-01 -9.42562878e-01 -5.30585110e-01 3.16462755e-01
-7.90947258e-01 -1.99974284e-01 -3.19754511e-01 8.96190464e-01
3.18138987e-01 1.05335319e+00 -1.25848919e-01 -4.84914422e-01
5.72262049e-01 2.33032137e-01 2.68672675e-01 -4.34878618e-01
-7.96768129e-01 1.23910993e-01 1.46419108e-01 -5.21325290e-01
-5.53682327e-01 -7.69889057e-01 -8.12408924e-01 -1.38692558e-01
-2.66294539e-01 2.86579996e-01 6.62201822e-01 5.70232511e-01
3.33830863e-01 7.63333738e-01 3.68651837e-01 -1.14192712e+00
-5.37321448e-01 -1.02696741e+00 -7.86530912e-01 3.27382505e-01
4.75449473e-01 -6.86063766e-01 -7.42637038e-01 1.99848443e-01] | [15.215742111206055, 5.12770938873291] |
4769030f-d4a8-4443-86eb-54b677f4f233 | rl4real-reinforcement-learning-for-register | 2204.02013 | null | https://arxiv.org/abs/2204.02013v3 | https://arxiv.org/pdf/2204.02013v3.pdf | RL4ReAl: Reinforcement Learning for Register Allocation | We aim to automate decades of research and experience in register allocation, leveraging machine learning. We tackle this problem by embedding a multi-agent reinforcement learning algorithm within LLVM, training it with the state of the art techniques. We formalize the constraints that precisely define the problem for a given instruction-set architecture, while ensuring that the generated code preserves semantic correctness. We also develop a gRPC based framework providing a modular and efficient compiler interface for training and inference. Our approach is architecture independent: we show experimental results targeting Intel x86 and ARM AArch64. Our results match or out-perform the heavily tuned, production-grade register allocators of LLVM. | ['Rohit Aggarwal', 'Anilava Kundu', 'Ramakrishna Upadrasta', 'Albert Cohen', 'Siddharth Jain', 'S. VenkataKeerthy'] | 2022-04-05 | null | null | null | null | ['hierarchical-reinforcement-learning'] | ['methodology'] | [-1.04260638e-01 1.40311822e-01 -1.30005693e+00 -2.14124456e-01
-6.87904298e-01 -6.07265413e-01 4.62665766e-01 -1.27969058e-02
-3.10806036e-01 8.97878230e-01 4.12078239e-02 -1.39217663e+00
3.60070825e-01 -9.03836012e-01 -9.43932593e-01 -2.82892525e-01
-3.12590271e-01 4.27107573e-01 8.85436758e-02 -6.11923337e-01
4.83593971e-01 3.42664540e-01 -1.60108185e+00 3.92338008e-01
4.30968612e-01 6.04038060e-01 -2.65535921e-01 1.26280487e+00
1.60820372e-02 1.40029407e+00 -6.54660344e-01 -1.44357011e-01
1.59112453e-01 -2.10216418e-01 -1.12719703e+00 -2.47384161e-01
2.01015010e-01 -3.19993824e-01 -2.60365307e-01 8.45209777e-01
2.24799186e-01 -2.59964108e-01 3.79221141e-01 -1.38009572e+00
-2.44274586e-01 1.10817003e+00 -7.55096555e-01 1.83550462e-01
5.30866832e-02 4.34489638e-01 1.17431355e+00 8.91721994e-02
3.97739202e-01 1.13120937e+00 4.33362931e-01 7.82409012e-01
-1.05462098e+00 -3.82166147e-01 -1.46404402e-02 -9.53245610e-02
-1.17092669e+00 -6.06780231e-01 5.02917826e-01 -2.01328665e-01
1.91681528e+00 2.29540169e-01 3.00164014e-01 9.27979827e-01
1.06623268e+00 8.67431641e-01 1.17016649e+00 -8.41376305e-01
5.60298383e-01 -2.68306702e-01 3.44390988e-01 1.22904193e+00
2.40580991e-01 8.24489772e-01 -4.72515188e-02 -5.94317436e-01
5.83544135e-01 -3.60439360e-01 3.46801370e-01 -5.69164097e-01
-1.20910299e+00 1.14349318e+00 2.22222045e-01 3.83722991e-01
1.51249483e-01 1.28026950e+00 9.23660517e-01 6.04360521e-01
-2.24801749e-01 5.53209722e-01 -8.28846693e-01 -3.61744434e-01
-8.96341980e-01 2.49550149e-01 1.25758433e+00 8.61597598e-01
1.02269089e+00 5.93612790e-01 1.82520017e-01 9.10360068e-02
7.08322525e-01 4.14373368e-01 7.11583018e-01 -9.55836654e-01
1.59427509e-01 3.71480525e-01 -3.45289856e-01 -3.49781215e-01
-5.25893271e-01 -4.22035486e-01 -3.19383055e-01 7.27876425e-01
-7.36231580e-02 -3.26626420e-01 -4.31408256e-01 1.60106266e+00
1.38118923e-01 3.26979369e-01 5.08679390e-01 5.42434454e-01
2.10965618e-01 5.51419020e-01 1.59249395e-01 -1.08827442e-01
1.36792302e+00 -1.24623144e+00 -4.01104867e-01 -5.63184083e-01
1.17432368e+00 -4.52941000e-01 8.19851995e-01 2.75379688e-01
-9.24781084e-01 -6.59562290e-01 -1.86885977e+00 3.15360874e-01
3.46270092e-02 1.47094473e-01 1.49174213e+00 1.20660913e+00
-1.44860792e+00 7.09923625e-01 -1.16119993e+00 5.15859246e-01
6.43590465e-02 1.13855731e+00 -3.73015255e-02 5.07142365e-01
-8.28177392e-01 7.50447929e-01 7.55095661e-01 -4.24981326e-01
-1.20836830e+00 -8.25543821e-01 -1.20622897e+00 -2.62172490e-01
3.37552279e-01 -8.10490131e-01 1.76147068e+00 -1.43509614e+00
-2.06922174e+00 9.97622967e-01 1.84515521e-01 -8.05302203e-01
-1.98128030e-01 -1.11770090e-02 -5.90313733e-01 -3.54280740e-01
-2.06441283e-01 4.52892244e-01 9.46976602e-01 -1.17443407e+00
-7.53603935e-01 -1.26708850e-01 2.76162893e-01 -3.24262440e-01
1.08801708e-01 1.52076101e-02 1.02984875e-01 -3.18875968e-01
-9.59348381e-01 -1.15515208e+00 -4.59843338e-01 -8.30053508e-01
-1.54496759e-01 -4.61497828e-02 7.59866714e-01 -2.01189041e-01
1.22304749e+00 -1.97837937e+00 2.78522938e-01 2.49546155e-01
3.67180139e-01 9.34622902e-03 -1.84253991e-01 1.13049723e-01
3.67631093e-02 -7.33283460e-02 -3.98174953e-03 -2.63335044e-03
5.24834335e-01 5.80234826e-01 -5.76510251e-01 6.85068905e-01
1.21146716e-01 1.14551318e+00 -1.18495119e+00 -5.84630013e-01
5.83006889e-02 -1.09753720e-01 -1.06016612e+00 3.44253689e-01
-8.05320382e-01 2.40023416e-02 -6.06401026e-01 8.09803486e-01
1.14412524e-01 -1.11144066e-01 7.79388905e-01 9.05097574e-02
-1.30731300e-01 4.83361036e-02 -7.78393745e-01 1.97378671e+00
-1.19628024e+00 4.72048104e-01 -1.06023081e-01 -7.80401289e-01
7.63921022e-01 -1.27931684e-01 8.99139643e-02 -8.07358205e-01
2.89996415e-01 4.59340483e-01 3.34470391e-01 -1.98863223e-02
8.39010239e-01 1.13345422e-01 -8.40549290e-01 9.96810257e-01
1.82969831e-02 -2.70738184e-01 -5.81303351e-02 -2.75250554e-01
1.47158480e+00 6.91052616e-01 6.94124758e-01 -6.39564395e-01
8.04462254e-01 1.73249707e-01 3.40379596e-01 7.80092001e-01
-2.63798475e-01 -4.83617544e-01 7.21263528e-01 -6.43301487e-01
-1.06409550e+00 -8.01801264e-01 1.57662764e-01 1.69283521e+00
-3.08404624e-01 -3.92360747e-01 -9.98496890e-01 -1.09991658e+00
-5.79540581e-02 1.03783000e+00 -6.02012515e-01 -4.68123823e-01
-1.27656281e+00 -8.63482058e-01 9.92816210e-01 6.57823622e-01
2.72853702e-01 -1.03764939e+00 -1.25506175e+00 2.84356624e-01
6.55555725e-01 -7.81167388e-01 -4.31474537e-01 7.20324218e-01
-8.17341328e-01 -9.09375787e-01 2.38926545e-01 -9.71915424e-01
2.92057872e-01 -4.31765437e-01 2.03478718e+00 5.14845073e-01
-5.16955316e-01 2.94468045e-01 -8.66508037e-02 1.52725410e-02
-1.40428650e+00 5.39865911e-01 -1.86411455e-01 -6.91023886e-01
3.31783235e-01 -4.36383456e-01 -2.66791523e-01 8.52724016e-02
-7.34823465e-01 -2.39623055e-01 8.36732864e-01 1.25697255e+00
5.55133462e-01 -1.31437540e-01 3.00958097e-01 -1.03605354e+00
3.46513093e-01 -3.81657094e-01 -1.08916688e+00 -3.26874256e-02
-6.64419949e-01 8.81325603e-01 8.97029579e-01 3.44746336e-02
-6.72727287e-01 1.87807485e-01 -4.35449451e-01 -1.62329838e-01
-9.97889340e-02 6.89787045e-02 -1.92748085e-01 -5.45795262e-01
8.95640612e-01 -5.77628836e-02 2.17856124e-01 3.40965211e-01
5.57984471e-01 5.54786801e-01 4.76449728e-01 -1.45857882e+00
5.70545137e-01 2.44624674e-01 1.88957438e-01 -2.00173661e-01
-4.81762201e-01 2.68312931e-01 -2.89426982e-01 2.08401784e-01
5.03707588e-01 -6.74907327e-01 -1.30597544e+00 -7.04263849e-03
-7.29658723e-01 -1.19239855e+00 -3.37881595e-01 -1.10031269e-01
-1.17726088e+00 1.38203442e-01 -7.19629228e-01 -3.54627639e-01
-5.95146537e-01 -1.90406561e+00 1.44026411e+00 8.72120112e-02
-3.69042337e-01 -1.21477556e+00 7.40043521e-01 -1.10491119e-01
6.15889728e-01 6.12358339e-02 1.40016866e+00 -6.13328934e-01
-6.20233536e-01 4.52925742e-01 2.11349815e-01 -1.00978157e-02
-8.09089467e-02 8.36796314e-02 -8.89675915e-01 -5.21254897e-01
-3.15334022e-01 -5.78438580e-01 5.29140353e-01 1.46370783e-01
1.18915021e+00 -2.59880275e-01 -4.82211113e-01 7.46849060e-01
1.75410795e+00 1.70467962e-02 5.12644351e-01 8.90156269e-01
5.53586960e-01 3.70893092e-03 6.48540139e-01 3.90580207e-01
3.15306008e-01 7.28554308e-01 8.36037397e-01 1.76689535e-01
-1.07658602e-01 4.71383408e-02 7.37441540e-01 5.46259582e-01
2.13442951e-01 2.44675264e-01 -1.12342942e+00 2.81368464e-01
-1.84075403e+00 -4.10217017e-01 3.37593138e-01 1.87314260e+00
1.36684680e+00 2.85284609e-01 9.41654295e-02 2.55941540e-01
-1.47874197e-02 1.79361343e-01 -3.36331755e-01 -1.31054986e+00
4.26822186e-01 1.06830370e+00 1.03865838e+00 8.23496699e-01
-1.07904649e+00 1.47348392e+00 7.21140766e+00 9.81662154e-01
-1.09624970e+00 -3.15881446e-02 5.94085276e-01 3.04492176e-01
-3.96072984e-01 2.15850547e-01 -1.10780466e+00 -6.93916008e-02
1.90368164e+00 -2.31377259e-01 7.35742450e-01 1.21500790e+00
-3.29109520e-01 -3.69230434e-02 -1.59538734e+00 4.17358458e-01
-7.63370544e-02 -1.50632596e+00 -3.30676466e-01 2.65092170e-03
6.75964296e-01 5.74490316e-02 2.35190734e-01 9.74071860e-01
9.13636267e-01 -1.37235272e+00 7.71606565e-01 5.48924655e-02
6.83627307e-01 -1.45103121e+00 7.12513804e-01 -1.42267393e-02
-9.21020150e-01 -1.51944056e-01 -2.06212878e-01 -1.65426582e-01
-5.78499198e-01 1.75243821e-02 -9.07216370e-01 3.44648778e-01
2.11413756e-01 4.46034878e-01 -7.40528941e-01 3.35345685e-01
-3.10203761e-01 4.81792271e-01 2.89998144e-01 -3.63617361e-01
3.71957332e-01 3.55574399e-01 -8.32584500e-02 1.41043925e+00
-1.22058064e-01 -4.64454323e-01 4.34400380e-01 6.21121883e-01
-1.12512022e-01 -1.42551988e-01 -4.66972291e-01 5.10359630e-02
2.47887656e-01 1.26549745e+00 -6.60290718e-01 -3.44756544e-01
-4.62390333e-01 7.10913420e-01 4.97303188e-01 -1.43461511e-01
-1.16017210e+00 -2.86024570e-01 9.99338806e-01 -5.20500600e-01
4.28030282e-01 -4.54234242e-01 -4.17187542e-01 -9.77729261e-01
-5.37475109e-01 -1.64403415e+00 2.18976662e-01 -3.63622531e-02
-6.30931377e-01 6.47215426e-01 -8.07905644e-02 -6.05564475e-01
-9.50097561e-01 -9.24051225e-01 -4.29640323e-01 3.51423323e-01
-1.84941494e+00 -1.08907819e+00 3.50934058e-01 3.36124629e-01
2.89407164e-01 -8.50853205e-01 1.16327834e+00 6.02109469e-02
-5.80711365e-01 1.12223887e+00 -1.37026101e-01 -1.90025032e-01
2.97016501e-01 -1.56489980e+00 7.69529343e-01 5.16570330e-01
-8.94574597e-02 7.09944248e-01 7.55148053e-01 -3.43565822e-01
-2.59445024e+00 -1.12978768e+00 1.18089065e-01 -2.53122032e-01
1.31693196e+00 -1.31172061e-01 -2.98036575e-01 1.17015100e+00
6.00544572e-01 3.24643523e-01 8.01139235e-01 1.09508298e-01
-5.03084958e-01 8.80668759e-02 -7.70670235e-01 7.02767730e-01
6.87205791e-01 -5.10283887e-01 -5.19923031e-01 3.49357843e-01
7.09874570e-01 -9.41306412e-01 -1.09619677e+00 1.46440566e-01
2.64307886e-01 -8.17744732e-01 1.00288749e+00 -6.98943198e-01
5.99474907e-01 -3.54720861e-01 -5.30195832e-01 -1.35486686e+00
-1.60511076e-01 -8.41954648e-01 -8.67327332e-01 6.15806222e-01
4.10015941e-01 -5.17684400e-01 1.15701532e+00 -1.88356459e-01
-4.41335350e-01 -6.24046504e-01 -6.76974297e-01 -6.89199567e-01
5.51842272e-01 -3.51144552e-01 9.69252229e-01 5.99955797e-01
2.01522544e-01 3.93862247e-01 -1.97503328e-01 1.29683793e-01
6.23373151e-01 5.72219670e-01 7.08658993e-01 -3.84117365e-01
-1.35477173e+00 -6.71346486e-01 -3.26764107e-01 -7.16844261e-01
1.31380451e+00 -1.16791499e+00 2.75596082e-02 -3.08585346e-01
9.34173465e-02 -5.76906264e-01 -2.63687849e-01 9.00024056e-01
-3.19977626e-02 8.50391239e-02 -2.46354759e-01 -1.40997007e-01
-9.01104808e-01 2.48829857e-01 7.55265117e-01 -3.19222480e-01
2.95317098e-02 -2.14139879e-01 -7.06386507e-01 5.25098443e-01
9.22305584e-01 -2.37649471e-01 -2.24291518e-01 -3.71300787e-01
5.09015203e-01 2.84735769e-01 1.68030456e-01 -9.86482799e-01
-7.28325620e-02 -3.55936885e-01 2.58053653e-02 -2.50793435e-02
-3.42787862e-01 -6.85500324e-01 -1.17865495e-01 1.10698652e+00
-2.75387824e-01 6.00020051e-01 7.88988769e-01 1.57490686e-01
-6.84004948e-02 -4.14109915e-01 7.97630489e-01 -1.82862468e-02
-1.26409841e+00 -1.84254441e-02 -6.27876639e-01 9.66474488e-02
1.38867688e+00 3.62461597e-01 -4.21640962e-01 3.52731586e-01
-3.25747162e-01 1.06787860e-01 8.06961834e-01 3.67897093e-01
3.42945725e-01 -1.16706145e+00 -6.39394641e-01 5.29081941e-01
2.07241803e-01 -7.97097802e-01 -2.87082553e-01 9.28449035e-02
-9.74490404e-01 4.74546343e-01 -6.39438331e-01 -3.93033177e-01
-1.11889684e+00 8.47024143e-01 6.03070796e-01 -9.04848516e-01
-2.69784570e-01 3.59165847e-01 -1.54719174e-01 -6.09467626e-01
-3.47303540e-01 -4.03985023e-01 8.71412754e-02 -6.70021057e-01
5.62793851e-01 -3.92115824e-02 2.87411004e-01 -1.58921316e-01
-5.34558713e-01 2.78449357e-01 -2.62731910e-01 -2.11624373e-02
1.08555663e+00 5.52869260e-01 -4.04927403e-01 5.63120097e-02
1.25370872e+00 1.01744890e-01 -8.20177376e-01 6.00113645e-02
2.84090340e-01 1.08651988e-01 6.34818673e-01 -4.74979192e-01
-1.02464557e+00 4.79265302e-01 5.45038164e-01 -2.29110926e-01
8.43562007e-01 -2.87027180e-01 7.13379264e-01 6.58898950e-01
9.60400343e-01 -8.43953669e-01 1.04504183e-01 7.88586915e-01
1.69732526e-01 -1.05416238e+00 2.72737235e-01 2.28881668e-02
-1.31443903e-01 1.43593395e+00 7.12734818e-01 -5.97152114e-01
2.45544374e-01 1.20910811e+00 -1.88121319e-01 7.85926506e-02
-1.03327739e+00 1.09600797e-02 -8.25992003e-02 6.90245390e-01
5.96082270e-01 2.83323467e-01 -1.68741420e-01 3.84478480e-01
-3.57313663e-01 8.84454250e-02 7.24756241e-01 1.60414588e+00
-4.27068800e-01 -1.92058182e+00 -4.31057453e-01 1.92526817e-01
-6.34115338e-01 -1.42255396e-01 1.44392289e-02 1.18745756e+00
-3.57560426e-01 5.01531124e-01 -1.41446784e-01 -4.56892312e-01
-1.82212945e-02 -1.45333767e-01 1.15945053e+00 -6.58195257e-01
-1.18333137e+00 -3.27040851e-01 4.14079368e-01 -8.03232193e-01
-2.96254028e-02 -3.49505484e-01 -1.49200487e+00 -5.91537833e-01
1.51809439e-01 2.68968403e-01 5.95202148e-01 8.00378799e-01
3.73777092e-01 1.09021533e+00 7.84833550e-01 -8.87267530e-01
-8.96873355e-01 7.51311518e-03 -3.71533871e-01 -4.73057300e-01
4.71816808e-01 -3.44746470e-01 8.04016367e-02 -1.71407372e-01] | [7.8390703201293945, 7.519133567810059] |
a71fc672-0618-4aaa-98d8-2fb98f88bf1f | improving-word-translation-via-two-stage | null | null | https://openreview.net/forum?id=ycgOlOnbbMq | https://openreview.net/pdf?id=ycgOlOnbbMq | Improving Word Translation via Two-Stage Contrastive Learning | Word translation or bilingual lexicon induction (BLI) is a key cross-lingual task, aiming to bridge the lexical gap between different languages. In this work, we propose a robust and effective two-stage contrastive learning framework for the BLI task. As Stage C1, we propose to refine standard cross-lingual linear maps between static word embeddings (WEs) via a contrastive learning objective; we also show how to integrate it into the self-learning procedure for even more refined cross-lingual maps. In Stage C2, we conduct BLI-oriented contrastive fine-tuning of mBERT, unlocking its word translation capability. We also show that static WEs induced from the 'C2-tuned' mBERT complement static WEs from Stage C1. Comprehensive experiments on standard BLI datasets for diverse languages and different experimental setups demonstrate substantial gains achieved by our framework. While the BLI method from Stage C1 already yields substantial gains over all state-of-the-art BLI methods in our comparison, even stronger improvements are met with the full two-stage framework: e.g., we report gains for 112/112 BLI setups, spanning 28 language pairs.
| ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['multilingual-word-embeddings', 'pretrained-multilingual-language-models', 'multilingual-nlp'] | ['methodology', 'natural-language-processing', 'natural-language-processing'] | [ 1.27893448e-01 -2.05431551e-01 -7.33232915e-01 -4.07080770e-01
-1.39631152e+00 -8.07936549e-01 8.31294358e-01 6.64878413e-02
-6.05415523e-01 7.40368545e-01 4.48929131e-01 -6.50847495e-01
2.00774893e-01 -3.34108829e-01 -9.04218197e-01 -3.44724953e-01
1.23195678e-01 6.13163531e-01 -9.44784209e-02 -5.75300753e-01
-1.64410084e-01 1.52879968e-01 -8.42805922e-01 2.78769046e-01
9.86522198e-01 5.33973753e-01 1.65057660e-03 2.31161803e-01
-3.56848538e-01 1.44926518e-01 -1.75268844e-01 -6.38170421e-01
3.90737474e-01 -4.82680529e-01 -7.61285365e-01 -2.96198845e-01
6.09245420e-01 1.98564723e-01 3.00804526e-03 1.11010754e+00
4.84897375e-01 -2.86972910e-01 5.43475688e-01 -1.09939373e+00
-9.40104485e-01 1.05207896e+00 -7.22200036e-01 2.18792826e-01
2.28911519e-01 2.50786319e-02 1.47681808e+00 -1.46242976e+00
6.59663320e-01 1.37933457e+00 8.71990979e-01 4.40322936e-01
-1.56893194e+00 -1.02767563e+00 4.62348223e-01 2.67227471e-01
-1.51860774e+00 -5.85460067e-01 7.19048858e-01 -3.26932847e-01
1.25551713e+00 9.99363065e-02 2.95745671e-01 1.10470629e+00
1.96403354e-01 1.02949679e+00 1.44251430e+00 -8.72727513e-01
-2.95758486e-01 4.35345232e-01 2.82824367e-01 5.64800084e-01
3.77489597e-01 2.20069066e-01 -8.65557134e-01 1.48003086e-01
3.27680469e-01 -4.01745498e-01 -4.77188490e-02 -3.03983122e-01
-1.64286578e+00 8.74977529e-01 2.55752563e-01 6.17311358e-01
1.32733643e-01 -1.61401421e-01 5.10405898e-01 6.41241252e-01
8.72125447e-01 4.64309812e-01 -7.91909277e-01 6.02074973e-02
-8.69462132e-01 -1.11748260e-02 5.39934516e-01 1.13141143e+00
1.18585086e+00 -1.96514398e-01 -2.65362263e-01 1.03186941e+00
2.55847812e-01 5.71780860e-01 6.13857567e-01 -1.37954846e-01
7.81092584e-01 5.28100252e-01 -2.81035572e-01 -4.44176525e-01
-4.07358050e-01 -5.88591933e-01 -6.97780430e-01 -1.75738499e-01
2.59747535e-01 -7.30356574e-02 -6.75535977e-01 2.10552025e+00
5.59939705e-02 7.49559850e-02 1.22614674e-01 4.87572283e-01
5.14542401e-01 6.35889590e-01 -4.89190817e-02 -8.16640630e-02
1.38362324e+00 -1.34978914e+00 -6.23684645e-01 -5.88540137e-01
9.22468543e-01 -9.72607315e-01 1.53797054e+00 8.63611922e-02
-9.39869106e-01 -7.46311247e-01 -1.19622982e+00 -2.58609831e-01
-5.90023458e-01 4.29888278e-01 6.84962571e-01 5.96928000e-01
-1.26770020e+00 1.47130296e-01 -6.79548979e-01 -4.16263103e-01
6.93738461e-02 3.27734351e-01 -4.42200810e-01 -2.75814354e-01
-1.49874985e+00 1.20782757e+00 3.73354465e-01 -4.39866409e-02
-5.22615850e-01 -1.05949545e+00 -1.09112489e+00 -2.65775055e-01
1.95544481e-01 -3.78929853e-01 8.81804287e-01 -8.70183051e-01
-1.56149173e+00 1.31172454e+00 -3.80342990e-01 -3.00573617e-01
3.70783389e-01 -3.35913181e-01 -4.59262997e-01 -5.05986333e-01
4.23090637e-01 8.89749348e-01 4.60678160e-01 -1.19474530e+00
-6.17773592e-01 -2.23074049e-01 -1.08126119e-01 3.10569167e-01
-4.21509176e-01 2.78340876e-01 -6.55154109e-01 -7.69927382e-01
-1.89169049e-01 -1.16957784e+00 1.77974477e-02 -3.31301600e-01
-2.89932132e-01 -3.88734698e-01 3.04806352e-01 -7.94643044e-01
1.36429739e+00 -1.97131300e+00 3.23348284e-01 3.94136598e-03
-8.88034478e-02 4.14119035e-01 -5.96282065e-01 3.66053879e-01
-2.18172416e-01 5.44358306e-02 -3.65236670e-01 -7.58555770e-01
2.29384303e-01 2.77599245e-01 -1.98439598e-01 4.34571356e-01
4.69995350e-01 1.35190368e+00 -8.60239029e-01 -2.96107441e-01
4.81219962e-02 3.90291065e-01 -6.82074428e-01 8.86285454e-02
7.75113925e-02 5.71767628e-01 1.90206796e-01 5.21711946e-01
6.23213708e-01 1.01904556e-01 4.80771989e-01 -3.95721942e-01
-2.97664702e-01 7.24893928e-01 -7.84578383e-01 2.21761274e+00
-9.78533030e-01 5.13711035e-01 -9.08503979e-02 -1.16330969e+00
8.87144148e-01 6.86492547e-02 1.99134380e-01 -1.00299239e+00
-1.69886008e-01 6.24999285e-01 1.58588216e-01 1.02263264e-01
4.09755558e-01 -2.59453565e-01 -5.13466835e-01 4.54944760e-01
5.09475231e-01 -1.51715772e-02 2.46993676e-01 -6.18946925e-02
6.27943754e-01 5.04150510e-01 4.33777869e-01 -8.74651194e-01
7.57664979e-01 -1.20032899e-01 6.37151003e-01 4.79243159e-01
-1.42844409e-01 2.36489996e-01 2.18802825e-01 -3.33886981e-01
-9.57720578e-01 -1.20337605e+00 -2.81965405e-01 1.44813406e+00
1.70301795e-01 -4.21732396e-01 -5.57535529e-01 -7.98905194e-01
-9.21727046e-02 6.81246281e-01 -5.95957339e-01 -1.67533740e-01
-9.31387424e-01 -1.21075559e+00 5.84006846e-01 5.16374946e-01
4.02376503e-01 -7.34757125e-01 2.00765997e-01 1.88238785e-01
-3.80785286e-01 -1.21798956e+00 -9.26804066e-01 3.50745261e-01
-4.19580311e-01 -5.24310708e-01 -5.94989955e-01 -1.27792025e+00
3.88065964e-01 2.92389154e-01 1.43475449e+00 -3.07228416e-01
-1.16308630e-02 -4.09762282e-03 -3.14085752e-01 -1.46188438e-01
-5.12053907e-01 6.66430712e-01 3.29932779e-01 -9.75511670e-02
5.95300972e-01 -4.39300209e-01 -2.87851244e-01 1.88960150e-01
-5.21912098e-01 1.66421458e-01 5.16361773e-01 1.06242847e+00
6.88584566e-01 -4.26305383e-01 5.94057620e-01 -9.23931062e-01
5.96912146e-01 -3.46920729e-01 -5.90355515e-01 5.67743480e-01
-8.01346421e-01 4.06059980e-01 5.49204409e-01 -4.58994925e-01
-8.70800078e-01 -3.76552612e-01 -3.16812456e-01 1.55873284e-01
2.07183972e-01 4.95636165e-01 -4.55740452e-01 -1.10063486e-01
4.57912564e-01 9.62812454e-02 -2.88747758e-01 -6.22105896e-01
8.62566113e-01 7.11432517e-01 5.35134315e-01 -9.23437595e-01
1.02840137e+00 5.10592908e-02 -5.54588377e-01 -3.56768757e-01
-9.51345444e-01 -4.88407612e-01 -1.01403999e+00 3.24365586e-01
6.67441070e-01 -1.32404208e+00 -2.04676166e-01 3.27007085e-01
-1.15024447e+00 -6.00536525e-01 -8.81418511e-02 5.35188615e-01
-2.81927556e-01 9.83290598e-02 -6.56610668e-01 -2.03952923e-01
-4.74097252e-01 -1.40042675e+00 1.24008417e+00 -2.69126594e-01
-2.85220832e-01 -1.45076466e+00 5.65625668e-01 2.55612195e-01
4.81369585e-01 -2.52421945e-01 1.24098361e+00 -7.73689449e-01
-1.66603118e-01 2.51426131e-01 -3.85078162e-01 6.06094897e-01
5.13336837e-01 -5.11619389e-01 -8.39336395e-01 -6.92380726e-01
-3.09830636e-01 -4.11886901e-01 8.66118670e-01 -8.64133518e-03
5.06737173e-01 -2.85394222e-01 -1.83170408e-01 9.60858345e-01
1.65489709e+00 -8.88302475e-02 3.21108043e-01 4.22241539e-01
9.14496720e-01 4.90865350e-01 4.30576444e-01 -3.51371527e-01
7.70195603e-01 9.95715678e-01 -2.93558866e-01 -5.94210982e-01
-5.35128415e-01 -4.56395090e-01 8.38447452e-01 1.79999101e+00
2.38653049e-01 1.92461565e-01 -1.06309354e+00 8.80905211e-01
-1.58211696e+00 -3.45436126e-01 1.92265496e-01 2.28018713e+00
1.48212945e+00 2.16887131e-01 -7.11045638e-02 -1.62343293e-01
3.97403181e-01 1.23875298e-01 -4.27989900e-01 -6.10294700e-01
-3.68097007e-01 6.98840439e-01 6.84720933e-01 1.10596621e+00
-1.06983399e+00 1.67147100e+00 6.21736908e+00 1.03107631e+00
-1.24500799e+00 5.96015334e-01 4.34627801e-01 -1.42377838e-02
-7.19929814e-01 5.80366366e-02 -1.25249600e+00 2.04072654e-01
8.90496731e-01 -3.01987648e-01 6.85976326e-01 4.94599491e-01
-1.63106382e-01 3.15945596e-01 -1.34068406e+00 8.60574186e-01
3.63351792e-01 -1.14767718e+00 1.08622134e-01 -3.48808207e-02
1.10330939e+00 4.88813549e-01 1.70256332e-01 7.82883108e-01
4.05305654e-01 -8.43797028e-01 6.55224085e-01 -1.78158164e-01
1.44296801e+00 -9.04389679e-01 6.03956103e-01 1.75575111e-02
-1.40955627e+00 3.49298626e-01 -2.98907518e-01 -4.05877717e-02
2.34705620e-02 4.87546176e-01 -5.76126039e-01 7.86046565e-01
4.10277635e-01 8.90570462e-01 -7.15807438e-01 2.52916455e-01
-4.48486090e-01 8.16111445e-01 -9.43430662e-02 3.07536662e-01
4.34092402e-01 -5.02917767e-01 3.34328890e-01 1.70050514e+00
1.29296437e-01 -6.76135361e-01 2.21491009e-01 6.67760551e-01
-2.97074616e-01 6.19493365e-01 -6.55073524e-01 5.55207916e-02
4.23040479e-01 1.15213490e+00 -3.43525380e-01 -3.63054276e-01
-8.33645940e-01 1.28444731e+00 8.23151112e-01 3.20151478e-01
-8.92059326e-01 -3.06477368e-01 1.17398679e+00 -1.89569533e-01
2.09063858e-01 -3.44420612e-01 -3.80093664e-01 -1.62199831e+00
-3.64543274e-02 -1.18478978e+00 1.62341878e-01 -1.37718886e-01
-1.46747589e+00 7.51360536e-01 9.29314084e-03 -1.01717174e+00
-2.64636487e-01 -9.36436236e-01 -3.83511484e-01 1.15602672e+00
-2.05786872e+00 -1.76306689e+00 2.97696888e-01 5.63779712e-01
6.93124175e-01 -3.94641966e-01 1.02168429e+00 5.44786692e-01
-6.67030990e-01 1.30675030e+00 2.60586470e-01 1.89767659e-01
1.21962249e+00 -1.18322408e+00 9.06594694e-01 1.12550628e+00
6.77724004e-01 8.58087182e-01 2.23075375e-01 -4.67040181e-01
-1.35332453e+00 -1.26377511e+00 1.48444641e+00 -4.30210263e-01
1.16657245e+00 -9.98182654e-01 -8.15659881e-01 1.03391564e+00
5.68369567e-01 -1.19916447e-01 7.73586333e-01 5.78386784e-01
-7.46153653e-01 -2.64437288e-01 -5.97246528e-01 8.45162690e-01
1.29285884e+00 -8.94518256e-01 -7.01623559e-01 3.19312066e-01
1.05011678e+00 -1.50229052e-01 -7.73331344e-01 6.27232790e-01
4.69432890e-01 -4.84403819e-01 1.10923362e+00 -7.02947319e-01
2.66761154e-01 -1.27566740e-01 -3.57525587e-01 -1.70892954e+00
-4.18889374e-01 -7.76700854e-01 3.99655193e-01 1.45103693e+00
8.86595964e-01 -8.22969615e-01 1.56144291e-01 -1.72871202e-01
-1.57288641e-01 -6.92294538e-01 -1.03747892e+00 -1.08065796e+00
9.38334823e-01 -5.38572073e-01 5.70789278e-01 1.21694422e+00
1.06768645e-01 7.69100845e-01 -4.70637143e-01 -5.55316992e-02
6.17324650e-01 1.77114815e-01 7.01977968e-01 -6.88611507e-01
-3.57873559e-01 -6.70637369e-01 -6.31849989e-02 -1.20542526e+00
6.06097698e-01 -1.63023913e+00 1.50104687e-01 -1.25455832e+00
4.69959468e-01 -5.96851468e-01 -5.68763494e-01 5.72536469e-01
-7.35529244e-01 6.88714921e-01 1.57092556e-01 1.11636356e-01
-3.84404033e-01 4.76969063e-01 9.15013850e-01 -3.33240360e-01
-2.30504245e-01 -4.50032890e-01 -8.92796814e-01 4.98098284e-01
7.57062912e-01 -3.35012525e-01 -4.05663073e-01 -8.87044847e-01
2.75550485e-01 -6.79566205e-01 -2.08903641e-01 -5.66984892e-01
-5.42544154e-03 5.49062304e-02 -7.93958604e-02 -3.89506996e-01
-9.67483371e-02 -3.48993778e-01 -2.69785970e-01 2.27690175e-01
-3.51705521e-01 2.82727093e-01 5.46428919e-01 2.20015142e-02
-3.89745206e-01 1.97898448e-01 7.29785144e-01 1.60184637e-01
-6.50211215e-01 2.09117904e-01 -4.87575680e-02 5.83034039e-01
4.18941468e-01 2.57014245e-01 -3.03929057e-02 1.03904262e-01
-2.71595091e-01 1.65199146e-01 3.58171344e-01 7.93657660e-01
5.93227930e-02 -1.76863694e+00 -1.08651209e+00 5.62813163e-01
6.66438401e-01 -3.60014319e-01 -3.34941596e-01 1.07795215e+00
-1.52875856e-01 6.18441939e-01 -5.33149503e-02 -4.80473816e-01
-9.96192336e-01 5.01652598e-01 5.44788204e-02 -8.84372234e-01
-4.24831271e-01 1.00466216e+00 6.54966533e-01 -1.06042826e+00
2.16212720e-02 -4.92873490e-01 2.08772883e-01 1.23039678e-01
3.51757526e-01 -3.38787995e-02 3.04155141e-01 -8.61146152e-01
-6.74248397e-01 8.52354825e-01 -3.76709670e-01 -3.75179648e-01
1.22273433e+00 -3.82199794e-01 -1.48519009e-01 6.91104949e-01
1.58365452e+00 4.66554195e-01 -8.87184203e-01 -6.98973715e-01
2.68162042e-01 -1.35351690e-02 -1.08242534e-01 -1.00558186e+00
-8.19747746e-01 1.00299370e+00 4.13380623e-01 -5.67624688e-01
1.05726397e+00 4.07024100e-02 8.79632235e-01 2.10945923e-02
5.50845385e-01 -1.04265368e+00 -2.24276245e-01 7.03862727e-01
7.91773915e-01 -1.45103049e+00 -2.51329124e-01 -2.31523022e-01
-6.07189298e-01 7.50040770e-01 3.17891777e-01 -1.69614609e-02
5.56532025e-01 4.73123342e-01 4.23679382e-01 2.44101360e-01
-5.82144678e-01 -3.41158241e-01 5.15556395e-01 3.17069501e-01
7.34774411e-01 2.80815005e-01 -7.68286943e-01 6.96389139e-01
-3.44978362e-01 -3.71969908e-01 -1.40642077e-01 4.93266732e-01
9.12869573e-02 -1.73727727e+00 -5.78577369e-02 -1.55511275e-01
-2.46037290e-01 -7.60136366e-01 -3.20334822e-01 9.70355093e-01
2.65401512e-01 7.25011528e-01 -1.66449416e-02 -4.74523932e-01
2.16988906e-01 2.68318206e-01 6.83385789e-01 -6.25170708e-01
-5.81367970e-01 4.86514568e-02 9.37847793e-02 -6.16067111e-01
-3.34542245e-01 -5.92121899e-01 -8.33866298e-01 -9.36341211e-02
-1.62742987e-01 -8.08071420e-02 6.98013961e-01 1.08350289e+00
1.91978380e-01 2.81208426e-01 6.53524935e-01 -6.67928159e-01
-4.15544152e-01 -1.06769145e+00 -7.81134292e-02 4.78937417e-01
2.22383201e-01 -5.66374958e-01 -2.90167630e-01 -1.85200311e-02] | [11.025527000427246, 10.025758743286133] |
d2d63281-62b0-4cc2-a949-61ad7e654c7c | ambiguity-aware-multi-object-pose | 2211.00960 | null | https://arxiv.org/abs/2211.00960v1 | https://arxiv.org/pdf/2211.00960v1.pdf | Ambiguity-Aware Multi-Object Pose Optimization for Visually-Assisted Robot Manipulation | 6D object pose estimation aims to infer the relative pose between the object and the camera using a single image or multiple images. Most works have focused on predicting the object pose without associated uncertainty under occlusion and structural ambiguity (symmetricity). However, these works demand prior information about shape attributes, and this condition is hardly satisfied in reality; even asymmetric objects may be symmetric under the viewpoint change. In addition, acquiring and fusing diverse sensor data is challenging when extending them to robotics applications. Tackling these limitations, we present an ambiguity-aware 6D object pose estimation network, PrimA6D++, as a generic uncertainty prediction method. The major challenges in pose estimation, such as occlusion and symmetry, can be handled in a generic manner based on the measured ambiguity of the prediction. Specifically, we devise a network to reconstruct the three rotation axis primitive images of a target object and predict the underlying uncertainty along each primitive axis. Leveraging the estimated uncertainty, we then optimize multi-object poses using visual measurements and camera poses by treating it as an object SLAM problem. The proposed method shows a significant performance improvement in T-LESS and YCB-Video datasets. We further demonstrate real-time scene recognition capability for visually-assisted robot manipulation. Our code and supplementary materials are available at https://github.com/rpmsnu/PrimA6D. | ['Ayoung Kim', 'Jee-Hwan Ryu', 'Jeongyun Kim', 'Myung-Hwan Jeon'] | 2022-11-02 | null | null | null | null | ['object-slam', 'scene-recognition', '6d-pose-estimation', 'robot-manipulation'] | ['computer-vision', 'computer-vision', 'computer-vision', 'robots'] | [ 1.46227047e-01 -4.52141687e-02 -2.06648245e-01 -3.78284425e-01
-5.14509916e-01 -5.50514638e-01 3.28347683e-01 -2.11716115e-01
-1.37596279e-01 3.98543239e-01 -2.31602728e-01 1.82190359e-01
-3.77908975e-01 -2.97480434e-01 -9.83918130e-01 -6.57308877e-01
3.40669274e-01 9.45183039e-01 2.70286743e-02 1.85590774e-01
3.76644641e-01 7.84432888e-01 -1.48469782e+00 -2.93027669e-01
7.29993641e-01 1.30097175e+00 6.98808968e-01 4.00302947e-01
3.08583587e-01 2.33061433e-01 -4.07834560e-01 -5.23213595e-02
5.82919538e-01 2.71202624e-01 -3.27092648e-01 5.32139242e-01
4.52991992e-01 -6.00556314e-01 -5.05105495e-01 1.17491722e+00
3.58813077e-01 -1.42505509e-03 5.85044682e-01 -1.52463341e+00
-2.69035935e-01 1.59494534e-01 -5.98781586e-01 -3.31389308e-01
5.16550004e-01 1.00018032e-01 7.34268486e-01 -1.17427969e+00
6.73122823e-01 1.18867517e+00 3.54002982e-01 2.76389092e-01
-1.03246605e+00 -5.53871810e-01 2.19391108e-01 3.86199802e-01
-1.56371462e+00 -3.84930909e-01 9.77868974e-01 -4.94711787e-01
5.10766983e-01 9.43500921e-02 5.65090597e-01 1.09289110e+00
1.09217502e-01 7.82862067e-01 7.76101112e-01 -7.43828863e-02
5.21940663e-02 -9.42556411e-02 -2.23585874e-01 4.50086236e-01
7.19597101e-01 -1.50238812e-01 -6.09703720e-01 1.40001714e-01
8.87930512e-01 2.35299036e-01 -4.49380010e-01 -1.15704513e+00
-1.60011280e+00 3.63889307e-01 3.82072002e-01 -4.29041386e-01
-3.08065355e-01 9.59352124e-03 -1.99470110e-02 -3.02413534e-02
5.93205094e-02 3.89394879e-01 -4.97083187e-01 -1.60160363e-01
-1.30576357e-01 2.22313449e-01 6.57648385e-01 1.74751365e+00
7.84736514e-01 4.08437885e-02 2.60891289e-01 5.07358849e-01
6.42981410e-01 9.88941312e-01 9.88052487e-02 -1.24775028e+00
6.72942996e-01 4.86981422e-01 4.84102130e-01 -1.30231512e+00
-4.59013432e-01 -3.30480516e-01 -6.61377728e-01 1.89371467e-01
4.48560268e-01 7.58810565e-02 -7.17763186e-01 1.50213015e+00
6.79460883e-01 1.11496568e-01 -4.10852693e-02 1.19145179e+00
5.98476529e-01 3.83403778e-01 -6.86600566e-01 -2.36103609e-01
1.24588454e+00 -8.04329157e-01 -7.27266133e-01 -4.82340574e-01
1.66234806e-01 -9.19804156e-01 5.61483502e-01 4.96270686e-01
-7.76948571e-01 -3.35282683e-01 -1.29389560e+00 -1.06682166e-01
1.38979442e-02 4.36791450e-01 3.99636239e-01 1.57704577e-01
-4.60625887e-01 3.23989272e-01 -1.02120149e+00 -2.37333164e-01
1.30006179e-01 5.12093246e-01 -6.18065119e-01 -3.17176223e-01
-5.54312587e-01 1.14195430e+00 7.09986746e-01 5.02778053e-01
-8.91420543e-01 -4.34144795e-01 -1.09228849e+00 -3.40609640e-01
1.10380828e+00 -6.59179270e-01 1.22431910e+00 -1.50277555e-01
-1.62112975e+00 5.89024782e-01 -2.31295340e-02 -4.11977507e-02
6.01736963e-01 -5.37892997e-01 1.03589810e-01 7.05481768e-02
-2.27005146e-02 4.91724283e-01 1.00305641e+00 -1.59417260e+00
-4.59244370e-01 -8.60213637e-01 6.19416647e-02 6.85853779e-01
1.98442787e-01 -5.44977903e-01 -6.97330296e-01 -3.46918106e-01
9.84512985e-01 -1.15027928e+00 -7.13116378e-02 4.83995169e-01
-4.19737101e-01 6.95002091e-04 1.15557730e+00 -4.49370086e-01
3.76822084e-01 -2.06706691e+00 5.17875254e-01 6.14459664e-02
1.29427789e-02 -2.67816875e-02 -6.24561310e-02 6.95762113e-02
2.47069195e-01 -3.92558813e-01 3.35263051e-02 -5.55301428e-01
1.51734486e-01 4.79766458e-01 -3.21801722e-01 9.10564840e-01
1.61501333e-01 7.54470766e-01 -7.68584907e-01 -3.56029540e-01
5.07397950e-01 4.76439297e-01 -3.36079568e-01 3.14438969e-01
-3.07206631e-01 6.76383257e-01 -6.11110747e-01 1.02817643e+00
1.00234544e+00 -7.16686174e-02 1.37923926e-01 -6.74980700e-01
-6.62073120e-02 -1.86344180e-02 -1.71451879e+00 2.06222630e+00
-2.70766079e-01 3.34369183e-01 3.29792798e-01 -8.79277468e-01
1.04288745e+00 7.52464160e-02 6.64326787e-01 -1.68663666e-01
3.71879309e-01 3.30538362e-01 -1.74516678e-01 -4.53388065e-01
5.20993173e-01 3.39261234e-01 -7.45815039e-02 1.69080719e-02
6.65421411e-02 -8.43794346e-01 -6.21170998e-02 -1.45824820e-01
5.69657624e-01 5.19082427e-01 4.49171364e-01 1.12507358e-01
3.28174442e-01 -8.95838067e-02 8.34124327e-01 5.63584328e-01
-1.66319773e-01 8.19971859e-01 1.19152516e-01 -1.71199605e-01
-1.14554787e+00 -1.08812928e+00 -2.71091729e-01 1.64275303e-01
8.10148716e-01 -2.18201429e-01 -1.45810023e-01 -2.77401447e-01
3.98095459e-01 3.83545071e-01 -1.71586633e-01 -7.06178099e-02
-5.72016716e-01 -4.74865168e-01 -1.76906556e-01 3.09608310e-01
3.65278155e-01 -5.19165933e-01 -7.93150783e-01 -1.41966417e-02
-3.62901211e-01 -1.56056488e+00 -3.33956301e-01 8.01876187e-02
-7.78378487e-01 -1.16573632e+00 -3.76241833e-01 -5.31960368e-01
7.54043162e-01 6.39346302e-01 5.27180374e-01 -3.00669611e-01
-3.01438123e-01 6.32915616e-01 -2.65020221e-01 -5.21954000e-01
-9.07272194e-03 -2.86802024e-01 6.30592287e-01 2.75291502e-02
8.16596821e-02 -6.31302536e-01 -5.09980679e-01 6.89855754e-01
-5.64547420e-01 1.14898138e-01 6.71721935e-01 7.02442706e-01
7.89132893e-01 -1.11745326e-02 -1.08065465e-02 -3.18726338e-02
-1.53928831e-01 -2.79030770e-01 -1.01546264e+00 3.63992304e-02
-3.51745754e-01 1.31057957e-02 1.11595407e-01 -5.84827125e-01
-1.04858398e+00 6.26309752e-01 3.28393042e-01 -9.58808482e-01
-1.22469164e-01 4.50776011e-01 -5.77748239e-01 -2.52570331e-01
2.35127315e-01 2.93272901e-02 2.67875791e-01 -5.03780901e-01
1.28207251e-01 6.76454067e-01 6.97119057e-01 -7.76257575e-01
1.04581988e+00 6.48468792e-01 4.28058326e-01 -7.30970860e-01
-8.15746903e-01 -5.54073989e-01 -9.05864298e-01 -4.35518742e-01
6.48342431e-01 -1.05003572e+00 -1.02474153e+00 5.46969950e-01
-1.31422675e+00 2.00799689e-01 -1.20035321e-01 9.28036392e-01
-7.98710167e-01 6.28507018e-01 -1.00412942e-01 -8.34987938e-01
5.85629195e-02 -1.49714875e+00 1.41282868e+00 9.61375013e-02
2.05525205e-01 -2.21181870e-01 -3.92847240e-01 5.29423594e-01
-8.20753202e-02 3.81850719e-01 3.97521853e-01 -2.47597694e-01
-1.22407258e+00 -2.16551021e-01 -1.59810185e-01 3.24141048e-02
1.89481482e-01 -4.71608266e-02 -5.59322178e-01 -3.56845468e-01
3.59520406e-01 -3.09727371e-01 3.82956058e-01 2.82313168e-01
9.62707102e-01 -7.92337433e-02 -3.04067761e-01 7.54633129e-01
1.34146702e+00 1.65832281e-01 1.90816000e-01 4.21829015e-01
9.66446161e-01 6.18698418e-01 1.19730616e+00 7.67841935e-01
3.91349465e-01 1.12149179e+00 1.16424239e+00 7.85732508e-01
1.64404169e-01 -2.03842402e-01 2.53116488e-01 8.54829729e-01
-4.90030423e-02 -2.10499838e-01 -8.74156654e-01 2.80565739e-01
-2.03155804e+00 -3.19048434e-01 -9.53004211e-02 2.29255819e+00
4.47653145e-01 5.92944473e-02 -4.22076702e-01 -1.04865141e-01
7.63390541e-01 4.78609093e-02 -1.09617281e+00 4.00839299e-01
-1.46069929e-01 -6.08128667e-01 5.66632211e-01 5.11283457e-01
-8.78939748e-01 7.19019175e-01 4.98102188e+00 5.21302879e-01
-1.09272552e+00 -2.62058824e-01 -1.92042664e-01 -9.32874829e-02
2.60474719e-02 1.95148796e-01 -1.00550556e+00 1.19591907e-01
2.05101028e-01 -7.05729201e-02 3.53301287e-01 1.03068197e+00
-1.62128747e-01 -2.63875514e-01 -1.28703761e+00 1.33892584e+00
3.77689719e-01 -8.80856514e-01 2.49291789e-02 6.75554648e-02
4.69503134e-01 3.92928235e-02 9.03661996e-02 -1.11106567e-01
-1.91966727e-01 -5.28221726e-01 1.03584111e+00 5.31113207e-01
5.51487207e-01 -4.14519995e-01 5.67130148e-01 6.75628006e-01
-1.09028506e+00 -1.15606427e-01 -5.42704940e-01 -1.12169594e-01
4.11589473e-01 5.26064217e-01 -1.04565287e+00 9.06012893e-01
6.89012349e-01 8.89268041e-01 -3.43557835e-01 1.12276483e+00
-3.61240119e-01 -2.19718963e-01 -6.91604316e-01 1.44709215e-01
-3.30606431e-01 -3.72541755e-01 1.09833372e+00 2.90600151e-01
6.51419163e-01 1.77582353e-01 3.49987656e-01 7.58198619e-01
1.80786058e-01 -3.15734059e-01 -5.35997689e-01 1.52552336e-01
6.34273231e-01 1.15198600e+00 -6.01383507e-01 -7.46141141e-03
-3.08791012e-01 8.79008174e-01 1.27811700e-01 1.73731565e-01
-8.03301513e-01 1.75037384e-02 7.78426468e-01 -2.85332263e-01
4.16950285e-01 -6.88085914e-01 -2.04418734e-01 -1.60246170e+00
5.81015587e-01 -7.78166890e-01 -4.90473583e-02 -1.18809497e+00
-9.95282173e-01 2.12342948e-01 3.02186161e-01 -1.72149682e+00
-1.46616861e-01 -1.04679298e+00 1.35375127e-01 5.55532277e-01
-1.28173518e+00 -1.11952531e+00 -5.75864613e-01 3.57409149e-01
6.60950124e-01 9.53105688e-02 5.21040320e-01 -3.33279818e-02
-3.87497604e-01 1.24788381e-01 3.22145186e-02 -1.90118000e-01
7.65838861e-01 -9.72746074e-01 -1.11172162e-01 7.54613578e-01
4.31158319e-02 4.36358184e-01 8.56300771e-01 -6.97131038e-01
-2.14816928e+00 -8.80531788e-01 3.89641881e-01 -6.38253689e-01
5.77344418e-01 -5.57390690e-01 -6.03006780e-01 8.69361639e-01
-4.23941791e-01 2.32548147e-01 -1.28766552e-01 -2.57099181e-01
-2.45327383e-01 -1.48757279e-01 -1.00893688e+00 5.22067368e-01
1.27702820e+00 -3.31624180e-01 -6.01599157e-01 4.21100140e-01
8.92087340e-01 -1.20664084e+00 -1.07227457e+00 9.43191767e-01
7.12080300e-01 -5.74036002e-01 1.14447188e+00 -8.46777558e-02
5.06978929e-02 -7.46478736e-01 -5.35789788e-01 -1.06137621e+00
7.57105183e-03 -4.09665018e-01 -3.77236813e-01 9.55306768e-01
-7.98471570e-02 -6.28723443e-01 8.21583331e-01 5.96983314e-01
-2.45313779e-01 -5.61985254e-01 -1.08463871e+00 -9.29481506e-01
-5.78790843e-01 -4.86544043e-01 6.10447645e-01 6.43822908e-01
-3.39933693e-01 1.38447836e-01 -6.44857645e-01 9.75908220e-01
8.71880114e-01 4.93800402e-01 1.18257797e+00 -1.26744986e+00
-1.28202304e-01 -4.70392779e-02 -8.28033984e-01 -1.56859112e+00
1.89764947e-01 -4.89549577e-01 4.59741861e-01 -1.13675141e+00
7.52825364e-02 -3.49640667e-01 1.87549517e-01 1.47920087e-01
1.19529918e-01 1.01441912e-01 4.39268261e-01 3.93910766e-01
-4.92193550e-01 8.55307460e-01 1.42151260e+00 -2.36399949e-01
6.37869313e-02 1.35839298e-01 -2.11472392e-01 8.27436745e-01
6.62862480e-01 -3.63249928e-01 -3.39492768e-01 -7.49254525e-01
2.24017739e-01 3.75828177e-01 4.90048975e-01 -1.04443800e+00
4.32827115e-01 -3.27615291e-01 3.37437153e-01 -1.07139373e+00
9.29408789e-01 -1.39348543e+00 4.94469404e-01 2.74396777e-01
1.25061795e-01 2.76813773e-03 9.28049441e-03 7.76220918e-01
-1.31540075e-01 -3.93298268e-01 3.81426781e-01 -5.44473976e-02
-8.54713500e-01 6.67781591e-01 1.15598448e-01 -3.99838448e-01
1.14849591e+00 -3.48347604e-01 -2.80370772e-01 -3.22210312e-01
-6.86342597e-01 4.26479071e-01 7.42308974e-01 6.53256714e-01
8.30523670e-01 -1.33990932e+00 -4.86343235e-01 2.78213382e-01
4.49596256e-01 8.92331779e-01 2.44073629e-01 9.41781163e-01
-4.29434240e-01 4.68487561e-01 -2.11819395e-01 -1.28004479e+00
-1.22889006e+00 5.82855403e-01 1.55759186e-01 3.54545712e-01
-4.03150618e-01 5.58652043e-01 1.45318061e-01 -6.93265736e-01
3.76472265e-01 -4.94733214e-01 6.41456246e-02 -1.89671248e-01
1.36834741e-01 3.64280969e-01 -7.01895133e-02 -9.11776125e-01
-4.05770987e-01 9.62029696e-01 1.09906808e-01 7.81439096e-02
1.30825698e+00 -5.70456803e-01 -8.75482038e-02 5.22905052e-01
1.07614660e+00 -1.89923272e-01 -1.67161047e+00 -4.25246269e-01
-1.51597857e-01 -8.42554569e-01 -2.17467070e-01 -3.14191014e-01
-8.54245007e-01 6.94954991e-01 4.17644680e-01 -3.63377750e-01
9.01449919e-01 7.49457106e-02 3.13633919e-01 8.49464178e-01
8.42939079e-01 -9.10699964e-01 2.03744292e-01 6.46568060e-01
1.30307317e+00 -1.56696141e+00 3.62955540e-01 -8.12860727e-01
-3.87992442e-01 1.21365047e+00 8.66239846e-01 5.21714846e-03
5.21897793e-01 1.25066355e-01 -7.50289112e-02 -9.07725946e-04
-3.24782908e-01 5.80234043e-02 3.80751878e-01 5.82008660e-01
-4.17542040e-01 2.20224559e-02 1.50560945e-01 1.83314264e-01
-1.32245600e-01 -3.04186732e-01 5.29241025e-01 1.11014533e+00
-3.75680447e-01 -8.75408769e-01 -6.74451530e-01 7.91246071e-02
4.68053110e-02 4.57022697e-01 -8.73103812e-02 8.16268802e-01
8.24288577e-02 7.34820426e-01 -5.67256995e-02 -3.25881690e-01
4.64654922e-01 -3.08376640e-01 8.29026341e-01 -6.18248701e-01
4.45381314e-01 1.12344563e-01 -5.41283600e-02 -8.12935591e-01
-5.92181265e-01 -9.00116265e-01 -1.05162764e+00 1.02402672e-01
-6.46124363e-01 -2.20968679e-01 1.04816055e+00 8.70584130e-01
2.99651772e-01 9.28574279e-02 4.41230774e-01 -1.41049540e+00
-8.86658013e-01 -8.72642040e-01 -6.17664576e-01 2.28128567e-01
5.28102458e-01 -1.20243311e+00 -4.88017440e-01 -1.94701031e-01] | [7.386322021484375, -2.517674446105957] |
a9229bb4-6a0c-45c8-90d4-d1b34f054af3 | layoutgpt-compositional-visual-planning-and | 2305.15393 | null | https://arxiv.org/abs/2305.15393v1 | https://arxiv.org/pdf/2305.15393v1.pdf | LayoutGPT: Compositional Visual Planning and Generation with Large Language Models | Attaining a high degree of user controllability in visual generation often requires intricate, fine-grained inputs like layouts. However, such inputs impose a substantial burden on users when compared to simple text inputs. To address the issue, we study how Large Language Models (LLMs) can serve as visual planners by generating layouts from text conditions, and thus collaborate with visual generative models. We propose LayoutGPT, a method to compose in-context visual demonstrations in style sheet language to enhance the visual planning skills of LLMs. LayoutGPT can generate plausible layouts in multiple domains, ranging from 2D images to 3D indoor scenes. LayoutGPT also shows superior performance in converting challenging language concepts like numerical and spatial relations to layout arrangements for faithful text-to-image generation. When combined with a downstream image generation model, LayoutGPT outperforms text-to-image models/systems by 20-40% and achieves comparable performance as human users in designing visual layouts for numerical and spatial correctness. Lastly, LayoutGPT achieves comparable performance to supervised methods in 3D indoor scene synthesis, demonstrating its effectiveness and potential in multiple visual domains. | ['William Yang Wang', 'Xin Eric Wang', 'Sugato Basu', 'Xuehai He', 'Arjun Akula', 'Varun Jampani', 'Tsu-Jui Fu', 'Wanrong Zhu', 'Weixi Feng'] | 2023-05-24 | null | null | null | null | ['indoor-scene-synthesis'] | ['computer-vision'] | [ 1.70064092e-01 4.15424973e-01 2.80596972e-01 -3.12203079e-01
-5.92875600e-01 -9.54831243e-01 9.08855021e-01 7.13880137e-02
2.02205345e-01 5.99773943e-01 3.63650143e-01 -6.82138681e-01
2.66285717e-01 -9.33804095e-01 -8.77179861e-01 5.61856776e-02
2.95828581e-01 4.96654540e-01 -1.10889599e-01 -2.55940139e-01
2.88704991e-01 6.75483823e-01 -1.53341389e+00 6.22859418e-01
1.11133552e+00 3.64989132e-01 6.59214973e-01 9.94749308e-01
-4.81875271e-01 7.78094113e-01 -8.51124108e-01 -1.25687104e-02
2.12374419e-01 -5.07166684e-01 -5.91405809e-01 2.52811372e-01
6.04632854e-01 -3.62204760e-01 -2.07151547e-02 5.37325978e-01
4.55450624e-01 1.77861556e-01 9.37188506e-01 -1.49209964e+00
-1.18305624e+00 4.16848034e-01 -3.13663960e-01 -4.38574553e-01
9.89538193e-01 7.32163191e-01 8.05862486e-01 -9.11191404e-01
8.34518492e-01 1.67675161e+00 3.08525056e-01 5.06155193e-01
-1.64965355e+00 -3.66373181e-01 3.79427612e-01 -4.21402037e-01
-1.34958541e+00 -3.78939092e-01 5.07364988e-01 -7.47257411e-01
1.20299137e+00 3.39740634e-01 8.40908110e-01 1.22523463e+00
-2.14499459e-02 7.40501702e-01 1.20081794e+00 -4.78587091e-01
2.56763339e-01 2.69195378e-01 -5.33847511e-01 8.50568175e-01
7.36707449e-02 6.67807683e-02 -4.57109481e-01 1.77395537e-01
1.51366997e+00 -3.15596581e-01 -1.23888545e-01 -4.18890357e-01
-1.35963333e+00 6.18793488e-01 5.80609798e-01 -3.10563799e-02
-9.79310200e-02 3.95711243e-01 -1.02992386e-01 1.89340010e-01
1.18728399e-01 1.18677044e+00 -9.55047160e-02 -2.24862248e-02
-7.31932461e-01 6.43243253e-01 6.60771191e-01 1.81547606e+00
5.43541014e-01 1.90708578e-01 -6.14784062e-01 5.84855497e-01
3.69799733e-01 8.69938016e-01 -4.47165668e-02 -1.13574386e+00
8.04550052e-01 6.76297426e-01 3.84584576e-01 -9.98099923e-01
-3.41335952e-01 -6.00078255e-02 -6.52880967e-01 7.09295928e-01
3.42533082e-01 -1.62680581e-01 -1.02097726e+00 1.54251873e+00
7.36537278e-02 -6.33870840e-01 2.06191409e-02 8.40278566e-01
9.17186201e-01 1.12728918e+00 3.58766913e-01 2.49405071e-01
1.15015757e+00 -1.03228462e+00 -4.85721946e-01 -4.28665340e-01
5.07951617e-01 -7.97873020e-01 1.79626727e+00 1.21541455e-01
-1.42994952e+00 -8.25811088e-01 -7.53608286e-01 -4.26928192e-01
-4.47934926e-01 3.43752772e-01 5.17843902e-01 3.12137336e-01
-1.37100518e+00 3.09835255e-01 -5.23322582e-01 -4.77074206e-01
3.10812801e-01 -1.61989573e-02 -1.12161696e-01 -2.13663001e-02
-6.67429626e-01 8.55417609e-01 3.18926901e-01 -3.19057196e-01
-1.01810193e+00 -9.04155374e-01 -1.28558087e+00 4.93844822e-02
1.33979961e-01 -1.11978662e+00 1.49117899e+00 -5.59579730e-01
-1.49519837e+00 5.19753814e-01 -1.97784454e-01 4.64098714e-02
7.58827865e-01 4.94958088e-02 1.56704083e-01 -4.19478677e-02
3.66577208e-01 1.40970540e+00 7.23508418e-01 -1.62278962e+00
-3.02479804e-01 2.61152893e-01 4.16836113e-01 5.95553994e-01
1.82238504e-01 -4.35459852e-01 -3.63515586e-01 -7.15002894e-01
-8.79863948e-02 -7.93878198e-01 -5.59287906e-01 4.11565691e-01
-7.71655262e-01 8.75343662e-03 8.12975883e-01 -3.93009067e-01
8.90991271e-01 -1.85018313e+00 3.44176024e-01 2.48959914e-01
3.41538377e-02 5.79305962e-02 -2.96871006e-01 8.71542096e-01
6.92554042e-02 4.13817316e-01 1.12131633e-01 -6.07113361e-01
4.33989644e-01 1.25122935e-01 -4.20226783e-01 -3.88084441e-01
3.19116443e-01 1.41736042e+00 -1.01508749e+00 -5.40492892e-01
7.24349320e-01 3.13806266e-01 -7.60857821e-01 5.76817632e-01
-7.65129805e-01 6.82670116e-01 -3.45228136e-01 4.37738895e-01
2.76328683e-01 -5.56260824e-01 3.42952430e-01 -8.73696357e-02
-1.25355437e-01 2.52646506e-01 -8.98383975e-01 1.88305593e+00
-9.14408624e-01 7.93618381e-01 -2.98031718e-01 -2.71843940e-01
9.46114600e-01 8.41718614e-02 -2.85970181e-01 -9.46647525e-01
-2.53619075e-01 -8.50397274e-02 -4.30375725e-01 -5.29140353e-01
7.53670335e-01 6.84022307e-02 -4.21348572e-01 6.37406409e-01
-2.62827188e-01 -1.17452133e+00 3.61411422e-01 5.02636850e-01
6.29254162e-01 6.35190725e-01 3.18557322e-01 -2.57556856e-01
-9.93154496e-02 1.31817326e-01 -2.86792397e-01 9.72036242e-01
6.04247987e-01 7.49180257e-01 6.65789783e-01 -2.42744654e-01
-1.51905775e+00 -1.33604884e+00 4.23032880e-01 9.00376201e-01
6.46905601e-02 -6.50903940e-01 -9.12784219e-01 -3.37744564e-01
-9.48416144e-02 1.16742849e+00 -4.33807909e-01 2.31673867e-01
-5.09437203e-01 -3.93056357e-03 3.86641115e-01 7.74358869e-01
4.92795080e-01 -1.31765842e+00 -9.03981447e-01 -3.47888395e-02
-8.26673657e-02 -1.19790959e+00 -6.21398151e-01 -1.69640303e-01
-7.18223870e-01 -7.34173596e-01 -9.12949324e-01 -8.92390668e-01
1.30033016e+00 3.25292140e-01 1.34973431e+00 9.14465077e-03
-2.66057611e-01 5.27681172e-01 -2.00727046e-01 -3.26218396e-01
-8.73863757e-01 -7.37463264e-03 -3.25620621e-01 -8.35841000e-01
-6.10032916e-01 -6.07161641e-01 -4.82691944e-01 2.98512578e-01
-9.00396705e-01 1.26323533e+00 5.46487927e-01 5.01406848e-01
3.55837405e-01 -2.68559068e-01 1.66339189e-01 -6.97351217e-01
1.07124591e+00 -8.27835314e-03 -6.85036659e-01 3.38444054e-01
-3.91873956e-01 3.59759241e-01 8.64959359e-01 -4.16113466e-01
-1.09775794e+00 2.25177668e-02 1.43330991e-01 -2.33067557e-01
-5.21510005e-01 2.36839041e-01 -2.23035455e-01 2.18856335e-01
8.36889684e-01 1.20707192e-01 -3.21456909e-01 -2.17398122e-01
1.06497419e+00 1.44978613e-01 3.94529462e-01 -9.90138352e-01
1.09175372e+00 -3.03155258e-02 -3.94193605e-02 -7.97588646e-01
-3.63147020e-01 3.00375253e-01 -5.29440284e-01 -1.07178427e-01
9.25990224e-01 -7.12212563e-01 -6.24352694e-01 2.14123011e-01
-1.41228378e+00 -9.87879872e-01 -4.59202498e-01 -6.55979142e-02
-8.29058528e-01 -4.92776297e-02 -5.30851901e-01 -6.73973322e-01
7.59042799e-02 -1.36620426e+00 1.48226964e+00 3.03664774e-01
-7.65222788e-01 -1.05226421e+00 -3.47441643e-01 -1.56180831e-02
2.69659489e-01 5.62253773e-01 1.29615366e+00 2.84402609e-01
-1.07834685e+00 2.86399037e-01 -4.31180388e-01 -1.88099459e-01
2.91884273e-01 8.06355327e-02 -7.52066612e-01 -4.43325937e-02
-9.00668144e-01 -3.54452133e-01 1.57649562e-01 2.87971258e-01
1.06090438e+00 -6.04997873e-01 -3.68432045e-01 6.32219136e-01
1.24483871e+00 4.93574172e-01 6.82125330e-01 6.51284009e-02
1.01157749e+00 5.91324985e-01 5.03878057e-01 3.74380291e-01
5.07671356e-01 7.00592697e-01 4.49409187e-02 -5.54258049e-01
-4.27982181e-01 -1.13179195e+00 7.33206943e-02 3.89404863e-01
1.37927786e-01 -5.51078975e-01 -8.83683980e-01 3.44456196e-01
-1.75932300e+00 -7.79163361e-01 -8.65200907e-02 1.79685748e+00
8.15315008e-01 1.45433247e-01 4.21125814e-02 -2.12103412e-01
3.11809868e-01 1.04723372e-01 -3.42476070e-01 -7.48837471e-01
1.60893667e-02 8.35466161e-02 2.04011977e-01 6.47088528e-01
-5.69461882e-01 1.22428954e+00 7.03912067e+00 6.27319336e-01
-8.79933417e-01 -4.17120427e-01 7.59599328e-01 -2.44135689e-03
-9.66943145e-01 4.19681109e-02 -5.19958317e-01 1.98881060e-01
4.05940801e-01 4.06271107e-02 6.89999521e-01 7.71150231e-01
7.54610360e-01 -1.36111498e-01 -1.38352919e+00 1.07974398e+00
-1.59741074e-01 -1.73144019e+00 5.72436094e-01 4.05878685e-02
9.89613891e-01 -7.51884997e-01 2.70958871e-01 2.58297235e-01
7.40810513e-01 -1.61361659e+00 1.22776890e+00 5.06220222e-01
1.32885027e+00 -6.12333417e-01 -1.19879305e-01 4.56403524e-01
-1.29192650e+00 2.69350857e-01 -2.09796727e-02 -2.82905281e-01
5.30473471e-01 4.93720993e-02 -1.44775689e+00 2.73519129e-01
4.11705315e-01 4.04777855e-01 -7.20137835e-01 7.33447134e-01
-6.46309197e-01 4.67555709e-02 -5.38909845e-02 -3.89353484e-01
2.83018351e-01 -5.73450793e-03 2.48056546e-01 1.43278730e+00
4.76303220e-01 1.26703754e-01 3.13422978e-01 1.60645795e+00
3.44558954e-01 -1.06810331e-01 -1.11136115e+00 -3.66422474e-01
5.19229352e-01 9.50352848e-01 -8.38857889e-01 -5.68049967e-01
1.46851391e-01 1.14328015e+00 2.07684830e-01 7.90735483e-01
-9.35653567e-01 -4.93738413e-01 5.06390989e-01 4.25850391e-01
8.05010274e-02 -8.23332071e-01 -6.18556499e-01 -8.29177916e-01
-1.81134120e-01 -1.01892459e+00 -3.81328076e-01 -1.72878456e+00
-8.33328426e-01 6.22825325e-01 4.59682822e-01 -1.16063452e+00
-4.93408144e-01 -7.53641367e-01 -5.62041938e-01 1.08021975e+00
-9.14847434e-01 -1.46305418e+00 -6.68571055e-01 3.97255361e-01
8.48256290e-01 1.17835596e-01 8.38614047e-01 -2.51709133e-01
-5.62500283e-02 3.24939013e-01 -5.95889986e-01 7.06163119e-04
4.04488951e-01 -1.58859158e+00 1.29960823e+00 7.24968612e-01
1.47158444e-01 8.13805163e-01 7.17694879e-01 -7.83136725e-01
-1.36518085e+00 -1.14414394e+00 6.51243448e-01 -6.98201120e-01
9.71559137e-02 -7.24629462e-01 -3.91632378e-01 6.13768160e-01
4.51875120e-01 -5.15682578e-01 1.80970684e-01 -2.23285720e-01
-3.69163781e-01 4.44032788e-01 -1.02701521e+00 1.32388234e+00
1.51244462e+00 -6.43783867e-01 -4.68037784e-01 3.79547775e-01
1.10059810e+00 -8.85752141e-01 -5.34795344e-01 -3.37561369e-02
4.17437911e-01 -8.29468906e-01 1.18747020e+00 -5.31936228e-01
7.51064658e-01 -7.28424311e-01 -2.48052389e-03 -1.58413017e+00
-4.50769901e-01 -1.01318240e+00 2.83085316e-01 1.00315750e+00
6.18388832e-01 -3.23223084e-01 3.84276927e-01 6.71520472e-01
-2.98410445e-01 -2.96325237e-01 -4.10672463e-02 -6.23245120e-01
-4.39088941e-02 -5.35380006e-01 8.11566055e-01 6.38811469e-01
1.44736052e-01 3.95435512e-01 -1.75491691e-01 2.15089321e-03
3.60970289e-01 2.29244888e-01 1.18104851e+00 -7.07555592e-01
-3.57523918e-01 -5.78746676e-01 7.36251920e-02 -1.50078881e+00
-3.14130485e-02 -8.54897916e-01 1.42107844e-01 -2.23895502e+00
-1.96179494e-01 -6.67962551e-01 7.61490166e-01 5.29852808e-01
7.32870176e-02 1.06584914e-01 5.80789924e-01 -6.55269325e-02
-4.71842498e-01 3.74939144e-01 1.95420468e+00 -1.87078744e-01
-4.66167927e-01 -3.46199572e-01 -9.33200359e-01 5.56907177e-01
5.47675669e-01 1.49898767e-01 -9.33004916e-01 -6.95274770e-01
2.97795564e-01 3.33795667e-01 6.84532881e-01 -9.51260567e-01
4.21457440e-02 -6.71723247e-01 8.10975373e-01 -4.78656441e-01
3.39355350e-01 -3.90260518e-01 3.99930477e-01 1.44574761e-01
-6.87018156e-01 4.12803113e-01 4.71189708e-01 1.80915311e-01
2.68713236e-01 1.61464229e-01 4.39792991e-01 -5.27386904e-01
-6.60580873e-01 -4.81985137e-02 -6.41587198e-01 1.02272712e-01
8.08136344e-01 -4.41624254e-01 -5.52261531e-01 -6.82784855e-01
-6.88151240e-01 1.86563551e-01 8.54035854e-01 6.33405209e-01
9.49328899e-01 -1.50486398e+00 -3.34450126e-01 3.21614861e-01
1.32368162e-01 3.45364451e-01 1.39856756e-01 1.53976038e-01
-9.27951157e-01 6.48472667e-01 -2.39173040e-01 -6.90886915e-01
-1.03954172e+00 6.11436784e-01 1.60035208e-01 5.63321039e-02
-7.64706254e-01 6.77002728e-01 8.60781968e-01 -4.91418898e-01
7.06119835e-02 -9.88748729e-01 2.68646181e-01 -4.42892939e-01
4.96525645e-01 -1.90340111e-03 -4.79404122e-01 -1.42048851e-01
7.25683868e-02 6.61717653e-01 4.34146106e-01 -6.06093049e-01
8.52487564e-01 -6.23131841e-02 3.51972729e-01 2.85669953e-01
8.16291273e-01 5.21432497e-02 -1.70219219e+00 3.03572178e-01
-3.72773170e-01 -4.83005106e-01 -5.65890610e-01 -1.05213439e+00
-3.72406363e-01 9.43143964e-01 2.95929294e-02 1.63293928e-01
7.56394506e-01 2.99322400e-02 3.60798419e-01 5.38290977e-01
6.19489789e-01 -7.84527123e-01 6.54362798e-01 3.94357085e-01
1.41235673e+00 -9.34740722e-01 -2.30179384e-01 -4.87296820e-01
-9.49712574e-01 1.08298337e+00 9.18040097e-01 2.36247443e-02
-4.20515276e-02 5.28556883e-01 1.26330899e-02 -1.44472599e-01
-6.69160068e-01 -7.70443007e-02 6.37974083e-01 1.13986897e+00
4.44858730e-01 2.18965277e-01 4.21220154e-01 2.04821886e-03
-6.97025716e-01 -1.55763119e-01 4.65796262e-01 9.10334408e-01
-2.72771478e-01 -1.09968376e+00 -3.81112456e-01 7.52277533e-03
5.09750664e-01 -3.02298009e-01 -4.11687523e-01 9.52878177e-01
6.60813525e-02 8.92818689e-01 9.38824490e-02 -1.90644383e-01
4.76583749e-01 -2.81274170e-01 7.05152154e-01 -9.69119489e-01
-4.21538740e-01 2.26421468e-02 3.15655679e-01 -5.87914348e-01
1.00518558e-02 -4.51641858e-01 -1.29921412e+00 -2.47329250e-01
4.64863062e-01 -5.93317896e-02 5.07487059e-01 7.44180441e-01
4.61346030e-01 8.00226629e-01 1.73050404e-01 -1.24072707e+00
-7.13782758e-02 -7.21379161e-01 -1.04486704e-01 3.48066300e-01
1.69676065e-01 -4.56528634e-01 1.48151055e-01 3.68807614e-01] | [11.22646713256836, -0.21706973016262054] |
8a59cc5e-9f76-4619-91d7-f8e262f261d6 | a-multi-head-convolutional-neural-network-1 | 2205.15994 | null | https://arxiv.org/abs/2205.15994v1 | https://arxiv.org/pdf/2205.15994v1.pdf | A Multi-Head Convolutional Neural Network Based Non-Intrusive Load Monitoring Algorithm Under Dynamic Grid Voltage Conditions | In recent times, non-intrusive load monitoring (NILM) has emerged as an important tool for distribution-level energy management systems owing to its potential for energy conservation and management. However, load monitoring in smart building environments is challenging due to high variability of real-time load and varied load composition. Furthermore, as the volume and dimensionality of smart meters data increases, accuracy and computational time are key concerning factors. In view of these challenges, this paper proposes an improved NILM technique using multi-head (Mh-Net) convolutional neural network (CNN) under dynamic grid voltage conditions. An attention layer is introduced into the proposed CNN model, which helps in improving estimation accuracy of appliance power consumption. The performance of the developed model has been verified on an experimental laboratory setup for multiple appliance sets with varied power consumption levels, under dynamic grid voltages. Moreover, the effectiveness of the proposed model has been verified on widely used UK-DALE data, and its performance has been compared with existing NILM techniques. Results depict that the proposed model accurately identifies appliances, power consumptions and their time-of-use even during practical dynamic grid voltage conditions. | ['T. S. Bhatti', 'B. K. Panigrahi', 'Ashu Verma', 'Lokesh Panwar', 'Himanshu Grover'] | 2022-05-31 | null | null | null | null | ['non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'time-series'] | [-5.03004611e-01 -5.12978375e-01 -3.82751487e-02 -1.98192775e-01
-2.32976183e-01 -3.08573484e-01 4.36626643e-01 1.12869740e-01
-5.75146861e-02 8.50957394e-01 -4.65781875e-02 1.86937526e-02
-2.61496782e-01 -1.04848266e+00 -4.58285585e-02 -9.86617446e-01
-3.05134326e-01 3.67558897e-01 -5.47439337e-01 -5.31755835e-02
-1.36465102e-01 6.60710812e-01 -1.43657506e+00 -2.71098614e-01
7.23755538e-01 1.27464795e+00 2.40814403e-01 3.72925907e-01
3.31113309e-01 8.70271385e-01 -1.15829873e+00 2.95739353e-01
1.78210363e-01 -5.33594340e-02 -5.14780939e-01 -6.84676841e-02
-1.63092583e-01 -6.88380837e-01 -1.22043125e-01 1.08028078e+00
8.91473353e-01 2.27857172e-01 4.04578418e-01 -1.62443316e+00
-4.48030829e-01 7.36353517e-01 -3.78944278e-01 5.75614154e-01
6.45620674e-02 2.39592358e-01 7.09412158e-01 -9.38344076e-02
-2.22760409e-01 5.20071745e-01 6.92169905e-01 2.44249403e-02
-1.21641946e+00 -8.90603602e-01 -2.03671142e-01 1.02295947e+00
-1.41740608e+00 3.81330057e-04 1.08829665e+00 -2.30433464e-01
1.75576556e+00 2.77607739e-01 9.46757317e-01 8.00953209e-01
3.78325164e-01 6.76976919e-01 1.06055188e+00 -2.04647735e-01
6.14934921e-01 1.16827503e-01 7.44940192e-02 -1.43271804e-01
3.92618299e-01 -3.87727879e-02 -2.80157086e-02 7.76750147e-02
1.52042672e-01 -4.79238071e-02 -4.63522583e-01 1.11508548e-01
-5.01548827e-01 8.74539435e-01 5.63960791e-01 1.07337737e+00
-6.44474983e-01 5.41221872e-02 7.00219691e-01 1.55202784e-02
4.17027563e-01 2.68652290e-01 -5.95402539e-01 -3.36009562e-01
-1.14730549e+00 -1.03496969e-01 8.27921808e-01 7.83971310e-01
3.91445309e-01 8.56243789e-01 2.31278822e-01 5.28609633e-01
1.15742534e-01 3.50588024e-01 1.03342128e+00 -3.45861167e-01
1.12792246e-01 8.57389390e-01 2.52986960e-02 -8.87735248e-01
-1.12911558e+00 -6.39722168e-01 -1.64330137e+00 4.00107712e-01
-2.40459979e-01 -1.72423959e-01 -3.36453676e-01 1.52809298e+00
1.26081988e-01 1.41596317e-01 -1.00811191e-01 5.79126954e-01
7.53350675e-01 7.83606946e-01 2.22098276e-01 -4.67782855e-01
1.41065574e+00 -4.86542851e-01 -1.28318775e+00 2.94527322e-01
3.85172695e-01 -2.31242120e-01 6.10982716e-01 4.28452492e-01
-7.31627882e-01 -5.91365874e-01 -1.27520049e+00 1.59529597e-01
-8.90603781e-01 2.77604461e-01 2.16594994e-01 6.70271933e-01
-9.35932398e-01 6.00886524e-01 -7.70074010e-01 -5.19865155e-01
6.08408988e-01 5.60411930e-01 -3.80178243e-02 6.06981933e-01
-1.46069300e+00 1.29413366e+00 7.52737403e-01 4.70880061e-01
-6.61958933e-01 -7.30417490e-01 -7.59722888e-01 4.09811169e-01
-1.59354612e-01 -2.06716120e-01 1.28985643e+00 -4.66552854e-01
-1.57452250e+00 -8.79493207e-02 4.30413425e-01 -9.26046610e-01
5.20113230e-01 -2.04858929e-01 -9.18487549e-01 -1.27729595e-01
-1.81958020e-01 -3.50507684e-02 4.28953290e-01 -8.98203433e-01
-7.88193643e-01 -3.30353230e-01 -3.21521223e-01 -3.77621464e-02
-4.06526893e-01 -3.10572475e-01 6.08748555e-01 -2.24700466e-01
-6.31293178e-01 -3.98555964e-01 1.04542293e-01 -7.18638062e-01
-5.05075634e-01 -6.63065434e-01 1.56292760e+00 -9.20649469e-01
1.11112070e+00 -1.69981909e+00 -4.41715360e-01 2.77527183e-01
1.43457633e-02 6.87427223e-01 5.33663213e-01 5.37241280e-01
-2.97561198e-01 -1.78449243e-01 -1.43683434e-01 -2.58083552e-01
4.54021066e-01 2.83986211e-01 1.98041826e-01 6.71953738e-01
-8.18025917e-02 1.11660910e+00 -5.09022713e-01 -1.44944945e-02
1.41827309e+00 8.68657649e-01 3.32680732e-01 2.63417542e-01
-1.64130688e-01 4.64358032e-01 -2.02526882e-01 3.28226179e-01
7.58792579e-01 -2.32212692e-01 2.82399237e-01 -8.26750159e-01
-2.28820086e-01 1.67535201e-01 -1.38361263e+00 1.15706027e+00
-9.21770215e-01 8.63207400e-01 -1.01056490e-02 -1.41941476e+00
7.35091925e-01 7.17977762e-01 8.09788644e-01 -1.22018886e+00
7.35373318e-01 -1.00952432e-01 -1.40775397e-01 -4.82722163e-01
2.56640226e-01 2.97207296e-01 2.54405767e-01 4.04825300e-01
1.62071094e-01 -3.53401108e-03 2.07384586e-01 -3.29595447e-01
8.54813993e-01 -2.03890473e-01 8.60237360e-01 -6.05524957e-01
6.92259848e-01 -3.91605526e-01 4.27185059e-01 9.63230282e-02
-2.89301723e-01 -3.25383127e-01 -9.43904296e-02 -7.82099485e-01
-9.58342433e-01 -5.67142904e-01 -4.85938877e-01 4.05552328e-01
-2.27234319e-01 -7.36964047e-02 -7.40644991e-01 -2.02646703e-01
-6.52474314e-02 1.38810217e+00 -5.43418765e-01 4.37658839e-02
-8.01483572e-01 -1.37409544e+00 2.32630193e-01 7.26369977e-01
1.08844507e+00 -1.28274453e+00 -1.28771746e+00 7.15179324e-01
-1.15405858e-01 -1.11388206e+00 1.61922157e-01 6.39025271e-01
-3.65453780e-01 -1.36103618e+00 -1.66473553e-01 -5.81126928e-01
2.00930089e-01 -2.43755460e-01 1.57019389e+00 -1.70501918e-02
-6.43197358e-01 6.44525141e-02 -1.76747039e-01 -6.29510701e-01
-2.34383881e-01 3.62296492e-01 1.38262436e-01 -2.92790264e-01
8.52302790e-01 -1.01018596e+00 -8.48731279e-01 1.78057820e-01
-7.40716159e-01 -2.02589929e-01 1.73414826e-01 5.26019156e-01
1.59944355e-01 1.09324872e+00 1.24082530e+00 -3.89330804e-01
6.57504618e-01 -6.88040137e-01 -1.31593406e+00 -1.24716451e-02
-1.16688693e+00 -6.03456616e-01 1.04246175e+00 -2.22301826e-01
-7.77979434e-01 -1.83046982e-01 -3.27660292e-01 -4.18925658e-02
-6.02877676e-01 -3.09051480e-02 -5.99621952e-01 1.93317980e-01
-1.39140770e-01 2.73003399e-01 -4.80597883e-01 -6.10146046e-01
1.78573787e-01 8.03908825e-01 6.72010183e-01 -8.01855512e-03
8.04805636e-01 2.64459759e-01 1.19738601e-01 -8.37904036e-01
-2.29847983e-01 -2.70162284e-01 -7.17964947e-01 -1.75851882e-01
1.02419937e+00 -9.82801855e-01 -1.55983675e+00 1.02248538e+00
-8.42509508e-01 -3.49831372e-01 -4.64559257e-01 2.61566311e-01
-2.31240258e-01 1.62967473e-01 -7.13193059e-01 -9.75429237e-01
-1.10692751e+00 -1.33303249e+00 6.10943854e-01 5.49491584e-01
-3.88112694e-01 -1.35951030e+00 -1.84667304e-01 1.02224365e-01
1.07965040e+00 5.29169917e-01 1.11972880e+00 -7.94699728e-01
-3.71356040e-01 -4.65709567e-01 2.78733973e-03 6.57578111e-01
6.19281054e-01 -3.91487867e-01 -1.16607141e+00 -6.30922258e-01
4.14609849e-01 -2.98690256e-02 -1.30985066e-01 4.85355765e-01
1.19878602e+00 -4.79502827e-01 -8.06045439e-03 4.94764805e-01
2.03790355e+00 5.23663044e-01 4.97989208e-01 5.91736913e-01
5.85976183e-01 -9.44809020e-02 -1.94039360e-01 7.40985632e-01
5.11403024e-01 6.07897580e-01 8.26300681e-01 -2.72994518e-01
1.21479005e-01 3.06537539e-01 -3.00704632e-02 1.04195094e+00
3.34684670e-01 -4.84546185e-01 -2.81525970e-01 7.28576541e-01
-1.59473836e+00 -9.98056054e-01 -1.18028507e-01 1.64433873e+00
4.59096581e-01 -7.54837543e-02 2.87533879e-01 9.14345026e-01
4.97792870e-01 1.87519059e-01 -9.11732435e-01 -4.67159212e-01
-1.10929713e-01 3.58575284e-01 5.32288313e-01 1.05112463e-01
-1.01452756e+00 -8.08627307e-02 5.56181335e+00 7.48729646e-01
-1.14962125e+00 2.85928041e-01 4.64978844e-01 -2.95259170e-02
4.59501415e-01 -8.62670004e-01 -5.65792620e-01 8.98167014e-01
1.19202340e+00 -4.45975631e-01 5.50032020e-01 1.10346091e+00
5.05847275e-01 -3.34565669e-01 -1.15839911e+00 1.10208452e+00
-1.78120621e-02 -1.00485754e+00 -5.06818295e-01 9.08891410e-02
7.92452991e-01 3.57878268e-01 -3.06282669e-01 1.38401568e-01
5.29197812e-01 -9.89656091e-01 2.02382654e-01 1.23427898e-01
3.16295892e-01 -1.15137196e+00 1.42886662e+00 3.54481369e-01
-1.68375003e+00 -4.01013225e-01 -7.35502467e-02 -1.07159607e-01
3.90016228e-01 8.58489752e-01 -8.20695460e-01 9.18259144e-01
1.05414712e+00 4.39647764e-01 -3.84495556e-01 7.65960038e-01
-1.81533530e-01 8.22942019e-01 -6.83925390e-01 2.77295597e-02
2.78105475e-02 -2.42034420e-01 -9.60640311e-02 1.05320120e+00
2.57622659e-01 -1.38757408e-01 5.76272532e-02 8.96282673e-01
-1.21780261e-01 -5.22940420e-02 -4.29500401e-01 4.84383106e-01
4.56493169e-01 1.79826128e+00 -6.01350367e-01 -4.09585178e-01
-3.97053838e-01 6.80616975e-01 -1.73182085e-01 1.82756484e-01
-7.82584131e-01 -1.51194662e-01 7.47666478e-01 -1.82592377e-01
2.69077331e-01 1.23916358e-01 -1.30784810e-01 -7.57287145e-01
-9.76825282e-02 -5.27524292e-01 3.63602072e-01 -7.12186158e-01
-1.52694643e+00 4.26921993e-01 1.38678208e-01 -1.06956255e+00
-5.36249578e-01 -3.26115608e-01 -1.11345780e+00 8.88067603e-01
-1.84000039e+00 -1.11405635e+00 -6.11262262e-01 6.07814252e-01
8.72184217e-01 -1.75285578e-01 1.12023878e+00 7.38643944e-01
-9.57798421e-01 4.91782874e-01 5.70502937e-01 2.42914557e-01
-3.24880391e-01 -1.51688111e+00 3.94742072e-01 7.59226739e-01
-2.20360592e-01 -2.25318149e-01 6.38988853e-01 -2.92633772e-01
-1.29805768e+00 -1.48132646e+00 3.56974304e-01 1.06045283e-01
5.63246906e-01 -4.06836390e-01 -9.63383496e-01 7.49688268e-01
9.49256599e-01 -1.89805612e-01 7.65811503e-01 -4.31847095e-01
1.69233739e-01 -3.77766043e-01 -1.74174345e+00 -1.24263301e-01
4.24720496e-02 -3.43729109e-01 -4.61368322e-01 4.32485938e-01
1.10641971e-01 -7.58784041e-02 -1.36261868e+00 4.04503167e-01
2.46495098e-01 -8.71154606e-01 5.90489507e-01 1.65916219e-01
-4.23985988e-01 -5.81477225e-01 -2.38091871e-01 -1.78184366e+00
-4.46923763e-01 -4.05965477e-01 -6.64753973e-01 1.58302855e+00
-3.19269627e-01 -9.23354805e-01 4.42546010e-01 4.34212416e-01
5.63235059e-02 -6.54324532e-01 -1.35435236e+00 -6.53001428e-01
-1.37825459e-01 -4.91197586e-01 1.45392108e+00 1.02517462e+00
4.63672541e-03 1.88331902e-01 -2.76135117e-01 4.82003301e-01
7.04873919e-01 -1.53628048e-02 5.08829355e-01 -1.26768780e+00
1.20191522e-01 -4.66953188e-01 -6.91237986e-01 -1.35475755e-01
5.61138056e-02 -5.07619917e-01 -2.48189449e-01 -1.71995616e+00
-1.27913579e-01 1.11163393e-01 -7.30091989e-01 4.80464518e-01
3.11410069e-01 4.25699502e-01 1.99778646e-01 -8.55430812e-02
-1.49500012e-01 6.34267926e-01 4.09544110e-01 -3.77162009e-01
1.45705208e-01 -5.71337454e-02 5.87878264e-02 7.69947350e-01
1.47664475e+00 3.79211195e-02 -4.86870378e-01 -1.68149561e-01
-1.08477943e-01 -3.26694071e-01 2.74040401e-01 -1.46510434e+00
6.08878657e-02 4.04683262e-01 9.33781624e-01 -1.22186542e+00
1.84291884e-01 -1.51831567e+00 7.09327579e-01 7.02717662e-01
2.23489776e-01 6.08588636e-01 3.43630224e-01 2.49135960e-02
5.70547804e-02 6.04357421e-02 9.38720465e-01 -2.65685748e-02
-7.82148898e-01 7.64303431e-02 -4.03737485e-01 -5.79486370e-01
1.35118270e+00 -9.61783379e-02 -5.20053148e-01 -2.38057941e-01
-3.66187096e-01 5.38560450e-01 1.55476987e-01 5.79333544e-01
-2.21908633e-02 -1.53826010e+00 -4.61678565e-01 4.07135874e-01
-1.97441801e-01 8.64039585e-02 3.90267074e-01 5.85518658e-01
-3.37150872e-01 5.58468223e-01 -1.27281889e-01 -5.80314040e-01
-9.84558702e-01 5.83198130e-01 7.43922472e-01 -3.67106825e-01
-8.81904721e-01 -1.74252130e-02 -2.85986632e-01 -8.18219930e-02
3.06275398e-01 -6.89652205e-01 -5.35546541e-01 3.48320127e-01
6.75843358e-01 9.61969078e-01 7.55331159e-01 -7.56800592e-01
-2.19516024e-01 5.59276402e-01 2.47132987e-01 8.01872730e-01
1.52817464e+00 -3.07821572e-01 -1.10049918e-01 5.74647963e-01
1.28908265e+00 -7.32533097e-01 -6.90703690e-01 1.66966811e-01
-4.32443731e-02 1.46274269e-02 3.68153721e-01 -1.12983644e+00
-1.57656431e+00 6.25102818e-01 1.31396198e+00 8.09699774e-01
1.57363236e+00 -4.58452165e-01 1.07907307e+00 1.01188734e-01
5.14995694e-01 -1.34005523e+00 -6.32779956e-01 -6.66441247e-02
4.44917321e-01 -1.11913371e+00 -9.75777023e-03 3.42352152e-01
1.51638091e-01 1.02397490e+00 4.82612252e-01 6.43931702e-02
9.58549678e-01 6.37867808e-01 1.25979185e-01 -2.36450344e-01
-4.61879700e-01 1.82258308e-01 -2.40720183e-01 9.42533433e-01
1.48584932e-01 4.76901263e-01 -2.65753791e-02 2.81564027e-01
-5.61984956e-01 3.06174606e-01 4.03816879e-01 8.18146825e-01
-5.08987978e-02 -5.57118118e-01 -4.17263448e-01 6.79347634e-01
-7.10480154e-01 2.38684654e-01 5.92396200e-01 9.31407094e-01
4.67994452e-01 1.35213876e+00 2.96422839e-01 -1.76424757e-01
4.12499368e-01 1.33574799e-01 -3.34738158e-02 1.75524071e-01
-8.55497837e-01 -2.63336241e-01 -3.19812596e-01 -3.53693366e-01
-3.85507613e-01 -4.11098868e-01 -1.14710712e+00 -5.58882296e-01
-6.37558579e-01 2.37599194e-01 1.11310792e+00 1.09678352e+00
9.86864343e-02 1.10866737e+00 1.04089200e+00 -1.00010836e+00
-5.21358073e-01 -1.48909593e+00 -9.35129702e-01 5.18028438e-01
3.64367187e-01 -5.85195541e-01 -4.81705487e-01 -2.27020800e-01] | [6.0285210609436035, 2.6076648235321045] |
f7d9405a-67e0-4082-8f2e-a2520553d440 | unbiased-multi-modality-guidance-for-image | 2208.11844 | null | https://arxiv.org/abs/2208.11844v1 | https://arxiv.org/pdf/2208.11844v1.pdf | Unbiased Multi-Modality Guidance for Image Inpainting | Image inpainting is an ill-posed problem to recover missing or damaged image content based on incomplete images with masks. Previous works usually predict the auxiliary structures (e.g., edges, segmentation and contours) to help fill visually realistic patches in a multi-stage fashion. However, imprecise auxiliary priors may yield biased inpainted results. Besides, it is time-consuming for some methods to be implemented by multiple stages of complex neural networks. To solve this issue, we develop an end-to-end multi-modality guided transformer network, including one inpainting branch and two auxiliary branches for semantic segmentation and edge textures. Within each transformer block, the proposed multi-scale spatial-aware attention module can learn the multi-modal structural features efficiently via auxiliary denormalization. Different from previous methods relying on direct guidance from biased priors, our method enriches semantically consistent context in an image based on discriminative interplay information from multiple modalities. Comprehensive experiments on several challenging image inpainting datasets show that our method achieves state-of-the-art performance to deal with various regular/irregular masks efficiently. | ['Tiejian Luo', 'Libo Zhang', 'Dawei Du', 'Yongsheng Yu'] | 2022-08-25 | null | null | null | null | ['image-inpainting'] | ['computer-vision'] | [ 6.66305661e-01 2.12281451e-01 -9.75068510e-02 -4.30546701e-01
-1.28516006e+00 -3.14843923e-01 1.32855847e-01 -2.36952543e-01
-1.55595362e-01 7.70164549e-01 1.88041463e-01 7.93087631e-02
1.54307172e-01 -7.28702128e-01 -1.11018944e+00 -7.00088918e-01
6.03178978e-01 4.17186379e-01 3.25869501e-01 -1.82186380e-01
2.60979652e-01 3.50056827e-01 -1.32623100e+00 6.69749260e-01
1.10066569e+00 1.14165854e+00 5.59697807e-01 3.75293821e-01
-2.18118325e-01 8.00831914e-01 -1.77468568e-01 -3.26146781e-01
2.96956867e-01 -5.08131444e-01 -7.10718274e-01 5.38110971e-01
5.55606425e-01 -6.26073897e-01 -4.24953282e-01 1.17227018e+00
3.00332606e-01 8.22455287e-02 6.17192328e-01 -8.69033694e-01
-7.99132347e-01 4.34832573e-01 -1.09432781e+00 -3.45906094e-02
1.50655642e-01 2.56854534e-01 6.81833923e-01 -1.13092566e+00
5.73372602e-01 1.26012456e+00 6.50243938e-01 2.91505128e-01
-1.39151287e+00 -4.72935826e-01 3.25334400e-01 1.66353285e-01
-1.13680851e+00 -4.99312341e-01 1.39777339e+00 -1.21960327e-01
2.70279735e-01 2.59827465e-01 4.34018731e-01 1.01109600e+00
1.03985682e-01 1.10600150e+00 1.25560462e+00 -2.06205457e-01
-2.91156061e-02 -1.12978294e-01 -2.97446251e-01 6.97178662e-01
2.47686803e-02 3.31627764e-02 -3.25502753e-01 7.38830343e-02
1.10977733e+00 2.61877060e-01 -4.75979507e-01 -2.56174415e-01
-1.03185356e+00 6.31315112e-01 5.06939769e-01 8.31267834e-02
-6.66434884e-01 7.38536566e-02 7.04011768e-02 -2.60192454e-02
7.40555704e-01 3.99650559e-02 -3.89103085e-01 4.06836540e-01
-1.30807734e+00 1.43527418e-01 1.83806598e-01 7.88680553e-01
1.01565635e+00 1.19503498e-01 -3.60990942e-01 1.05204761e+00
2.42663398e-01 2.92123526e-01 1.69948041e-01 -1.18552113e+00
5.35591662e-01 4.93950337e-01 4.03607368e-01 -9.59912121e-01
-3.73885743e-02 -5.84039509e-01 -1.13331544e+00 1.33191824e-01
5.00261486e-01 1.53188512e-01 -1.26999485e+00 1.51678252e+00
5.04248261e-01 3.31921577e-01 -3.50094318e-01 1.22598374e+00
6.70422494e-01 7.59274244e-01 4.50336486e-02 -2.25492865e-01
1.28336513e+00 -1.46343076e+00 -7.84053385e-01 -6.90342486e-01
-7.11165816e-02 -9.34019804e-01 1.06392145e+00 4.05610025e-01
-1.39739776e+00 -5.30027807e-01 -6.31029308e-01 -5.88504076e-01
1.64106473e-01 1.69246346e-01 5.48148692e-01 5.87330461e-02
-8.06122720e-01 5.31170368e-01 -8.17058146e-01 9.31884944e-02
8.78729403e-01 -4.03354391e-02 -4.43867326e-01 -6.24684572e-01
-8.40590119e-01 6.11510098e-01 3.85883927e-01 4.89264488e-01
-1.09366405e+00 -6.72737241e-01 -9.90116537e-01 -2.51058768e-02
4.55415785e-01 -7.42463648e-01 9.05184507e-01 -1.27244937e+00
-1.28244305e+00 8.77478004e-01 -2.82199711e-01 -1.08885296e-01
7.49522567e-01 -2.26238459e-01 3.36112157e-02 5.22311270e-01
3.84215564e-01 7.70267785e-01 1.20130968e+00 -1.74191284e+00
-5.09151042e-01 -3.86154175e-01 -1.73207805e-01 2.39777386e-01
4.04440574e-02 -1.66663721e-01 -8.43647122e-01 -1.26897073e+00
3.57350469e-01 -3.82407963e-01 -3.28866363e-01 4.38651860e-01
-5.30492961e-01 2.68411964e-01 8.96421254e-01 -1.44549561e+00
8.37684572e-01 -1.93187094e+00 4.56559628e-01 -1.54511973e-01
1.26974821e-01 4.08811383e-02 -2.46832564e-01 -6.95618941e-03
4.00936007e-02 -1.84266940e-01 -8.25636506e-01 -6.43009841e-01
-2.84090638e-01 3.74078482e-01 -3.64186972e-01 5.11497319e-01
4.00549680e-01 1.02001834e+00 -8.04869771e-01 -6.47436678e-01
4.46847767e-01 7.73490965e-01 -5.37727714e-01 5.05943418e-01
-3.82589608e-01 9.26494658e-01 -4.79600787e-01 1.11283672e+00
1.06111240e+00 -2.92987049e-01 -2.68555254e-01 -5.72409153e-01
1.85478315e-01 -2.30437145e-01 -1.03397453e+00 2.27830458e+00
-5.98542333e-01 3.35432976e-01 6.61063373e-01 -1.42702270e+00
6.02505445e-01 1.40200675e-01 4.37665284e-01 -6.12467170e-01
2.43998915e-01 3.03183764e-01 -5.70645332e-01 -6.23626351e-01
3.04645807e-01 -3.05306375e-01 1.04479149e-01 2.42701650e-01
1.83430333e-02 -1.37300193e-01 -1.42385185e-01 -3.93701941e-02
6.74663901e-01 4.86856222e-01 -1.86094090e-01 9.06022638e-02
4.87079173e-01 -2.21975788e-01 8.22327197e-01 5.54703772e-01
-3.27986442e-02 1.27822459e+00 2.23994479e-01 -4.66358900e-01
-1.19910765e+00 -9.67123151e-01 -1.32372808e-02 8.47427070e-01
4.05046195e-01 7.01367259e-02 -7.89009273e-01 -5.80069602e-01
-2.68520683e-01 6.86559558e-01 -6.22219920e-01 -1.76440671e-01
-6.15209699e-01 -6.92049980e-01 2.26983577e-01 4.48899329e-01
8.59683573e-01 -1.20363879e+00 -3.57267588e-01 4.36109245e-01
-7.63066530e-01 -1.26727903e+00 -8.83578479e-01 4.22493927e-02
-9.89788532e-01 -1.08345354e+00 -1.07669663e+00 -1.04862785e+00
1.00186312e+00 4.33494806e-01 1.11786008e+00 1.68002144e-01
-2.95657694e-01 5.55806682e-02 -2.74806827e-01 2.30819553e-01
-2.61216491e-01 -2.58885115e-01 -6.96357608e-01 5.10508835e-01
-2.81027317e-01 -8.83279026e-01 -1.02344310e+00 2.76169658e-01
-1.41246116e+00 5.27359307e-01 9.92388904e-01 1.20717323e+00
9.73581016e-01 1.04347065e-01 3.87074709e-01 -9.47082520e-01
1.59859851e-01 -4.72264409e-01 -3.55361342e-01 4.59805340e-01
-2.09848389e-01 1.28801301e-01 6.25556052e-01 -4.61924493e-01
-1.39836740e+00 3.30955416e-01 -2.98838228e-01 -8.72783422e-01
-3.02910566e-01 3.04123193e-01 -4.88694012e-01 -1.05440594e-01
2.53491253e-01 7.70299137e-01 -6.38598204e-02 -7.27818847e-01
5.31538248e-01 4.91214633e-01 7.21884489e-01 -8.21125984e-01
6.77235842e-01 7.39546955e-01 -1.96299925e-01 -4.83238459e-01
-1.09799266e+00 -5.14875874e-02 -4.62882549e-01 -1.39414608e-01
7.43092120e-01 -1.10650826e+00 -1.12217255e-01 7.92306185e-01
-1.23410594e+00 -4.24558043e-01 -1.36999726e-01 -1.29822735e-02
-5.40531754e-01 7.12535381e-01 -8.62088203e-01 -5.69721162e-01
-4.10670847e-01 -1.25740421e+00 1.63377202e+00 1.46061540e-01
3.12098056e-01 -6.94251597e-01 -4.59233522e-01 9.65313315e-01
3.57654899e-01 5.23287714e-01 7.36150324e-01 1.14287175e-01
-9.06302989e-01 3.50153865e-03 -4.79453713e-01 4.83232617e-01
1.18187301e-01 -3.88939559e-01 -9.85099435e-01 -2.69570470e-01
1.87161520e-01 -4.97757286e-01 1.28455269e+00 4.69611079e-01
1.51936376e+00 -4.73215461e-01 -2.41444513e-01 9.00259554e-01
1.52148747e+00 -5.66599369e-02 6.98929310e-01 -4.40558642e-02
1.04601371e+00 6.67960167e-01 7.17998326e-01 2.45402813e-01
3.55062664e-01 4.02651697e-01 5.74423015e-01 -4.39680099e-01
-3.44144046e-01 -5.24554968e-01 1.33935198e-01 2.87557602e-01
1.28985152e-01 -2.16908470e-01 -4.96864974e-01 7.32492268e-01
-1.78945351e+00 -7.09923387e-01 3.05400454e-02 1.90202761e+00
1.15018594e+00 1.54362731e-02 -1.98859930e-01 -9.07485709e-02
9.11242604e-01 3.21453571e-01 -8.19280922e-01 3.18986177e-01
-3.27244401e-01 1.80911317e-01 3.68294775e-01 7.46497810e-01
-9.76390719e-01 1.10683942e+00 5.03836393e+00 1.35227001e+00
-1.06904542e+00 3.78566056e-01 1.31718457e+00 1.00812770e-01
-6.15960300e-01 1.81850761e-01 -3.84392172e-01 6.66202128e-01
1.49095669e-01 5.45275629e-01 5.56778133e-01 5.95276952e-01
1.70915335e-01 -3.92925948e-01 -7.43605971e-01 1.10307765e+00
1.63574845e-01 -1.40450549e+00 1.98887452e-01 -2.54617065e-01
9.61820126e-01 -1.70865878e-01 9.23441257e-03 -1.80747323e-02
3.51498947e-02 -9.97087240e-01 1.02994919e+00 6.08246446e-01
1.01476669e+00 -7.35872447e-01 4.23320502e-01 3.82408679e-01
-1.04237676e+00 -9.26944241e-02 -3.91836405e-01 1.91219836e-01
4.56690043e-01 8.99085402e-01 -1.02919176e-01 6.51863635e-01
6.79404974e-01 8.36030483e-01 -4.28037852e-01 9.31069493e-01
-3.91531259e-01 4.13731962e-01 -3.38320047e-01 7.58664191e-01
1.10080175e-01 -2.91025400e-01 4.61683899e-01 7.83857822e-01
2.64580220e-01 2.85988629e-01 1.52089059e-01 1.16735029e+00
-8.94504488e-02 -2.18698069e-01 -3.64189446e-01 1.25081003e-01
2.08501115e-01 1.39107502e+00 -9.47150111e-01 -3.41493249e-01
-4.22456384e-01 1.44526172e+00 4.01849478e-01 5.84418356e-01
-8.17281067e-01 -2.63617598e-02 1.79498106e-01 3.20510834e-01
4.73519892e-01 1.56350322e-02 -5.17667234e-01 -1.47509968e+00
2.39676550e-01 -8.17298293e-01 2.94988036e-01 -1.11741972e+00
-1.34260094e+00 6.13074660e-01 -2.75494546e-01 -1.17818630e+00
2.10016266e-01 -1.89730495e-01 -6.34776473e-01 9.55989361e-01
-1.77609921e+00 -1.67124987e+00 -4.13505316e-01 6.83268785e-01
8.43624532e-01 2.63172120e-01 4.24182832e-01 4.88401145e-01
-5.98214805e-01 3.66790622e-01 -1.72049515e-02 3.74341235e-02
7.13521183e-01 -9.03267205e-01 -3.98582220e-02 1.07522440e+00
-9.37034115e-02 1.86227530e-01 7.27357447e-01 -7.70442188e-01
-1.26126397e+00 -1.37059534e+00 3.10177743e-01 -3.60582322e-02
1.46511033e-01 -1.61680996e-01 -1.06984377e+00 6.51626885e-01
3.23958576e-01 4.27363157e-01 -4.63140942e-02 -5.16110003e-01
-2.10823953e-01 -1.36970535e-01 -1.26918674e+00 4.86444026e-01
1.04339921e+00 -3.78433764e-01 -3.83131564e-01 3.98675591e-01
6.81101620e-01 -6.27522230e-01 -5.87332487e-01 4.97892588e-01
1.75586954e-01 -1.14777124e+00 1.20099580e+00 -2.79889494e-01
8.05831969e-01 -5.00949621e-01 -8.17271769e-02 -1.14812052e+00
-1.03742607e-01 -5.60266733e-01 -9.94274765e-02 1.22418153e+00
3.25187147e-02 -2.33236864e-01 7.94306934e-01 5.83004296e-01
-3.35940212e-01 -9.25661683e-01 -8.07360172e-01 -1.95991993e-01
-7.56789818e-02 -3.84190053e-01 2.84817100e-01 8.82584393e-01
-7.49608755e-01 2.61468798e-01 -8.06476116e-01 2.41214469e-01
1.01569271e+00 6.07178867e-01 4.72162127e-01 -8.24763000e-01
-3.88039529e-01 -2.63038695e-01 1.27713501e-01 -1.34066737e+00
2.15298310e-01 -6.90696895e-01 2.03791678e-01 -1.69833004e+00
2.48172596e-01 -3.76818895e-01 -1.57248437e-01 5.65425277e-01
-4.43411738e-01 6.51272118e-01 -1.77356601e-01 1.55119374e-01
-5.38934171e-01 9.72956538e-01 1.80321836e+00 -3.86517584e-01
6.55060485e-02 -2.04801366e-01 -9.14958060e-01 7.78989255e-01
5.58185577e-01 -5.17539203e-01 -4.57603365e-01 -6.12503111e-01
6.62647635e-02 5.88690162e-01 7.75944173e-01 -5.93210459e-01
1.28215075e-01 -3.23956221e-01 7.40612328e-01 -7.13301599e-01
5.04709184e-01 -9.01387274e-01 2.11837873e-01 1.63988814e-01
-1.22075126e-01 -4.42641795e-01 1.12496980e-01 8.27982187e-01
-5.58584273e-01 -3.23269628e-02 1.10774696e+00 -4.13668334e-01
-6.17997706e-01 6.05085194e-01 7.05882087e-02 2.05622524e-01
8.26375842e-01 -2.50478774e-01 8.83012637e-02 -3.43491614e-01
-7.99608171e-01 2.05343828e-01 6.37773871e-01 2.81530291e-01
8.65662813e-01 -1.21636987e+00 -7.38912225e-01 3.34485561e-01
-1.79303110e-01 7.98768997e-01 8.89299870e-01 8.90951753e-01
-6.57776415e-01 -1.10544585e-01 -4.10108745e-01 -6.57078028e-01
-8.66086125e-01 5.38992286e-01 4.25028086e-01 -2.71243870e-01
-7.83324063e-01 9.49224651e-01 5.79212129e-01 -1.98774174e-01
1.62463978e-01 -2.75354713e-01 1.56710848e-01 -1.06842659e-01
4.03345585e-01 7.12498557e-04 -1.12869851e-01 -6.56328559e-01
-1.75641179e-02 7.53945827e-01 -2.08183482e-01 -9.81499106e-02
1.57436335e+00 -3.67388159e-01 -3.27039868e-01 9.54691246e-02
1.03835094e+00 -1.04414515e-01 -1.96985567e+00 -3.96966279e-01
-4.56531227e-01 -7.16197312e-01 2.72629529e-01 -7.71089494e-01
-1.60408390e+00 9.31740999e-01 4.24588203e-01 -3.32777917e-01
1.53325224e+00 -1.24567993e-01 1.23678279e+00 -1.91568702e-01
3.40144426e-01 -1.03320074e+00 1.42011687e-01 -6.09569736e-02
1.28002357e+00 -1.51098239e+00 -7.84976128e-03 -6.63487971e-01
-6.26096547e-01 9.53079283e-01 7.12417066e-01 -3.24571371e-01
6.00605071e-01 2.93551534e-02 -4.04825527e-03 1.89099740e-03
-4.09214079e-01 1.43795997e-01 3.30382705e-01 4.31504548e-01
5.58263510e-02 -1.87731594e-01 -2.59384215e-01 7.08382607e-01
4.62269157e-01 -8.60692188e-02 3.59453827e-01 7.32677042e-01
-3.09904635e-01 -1.05928564e+00 -5.84688306e-01 3.43404263e-01
-5.12750268e-01 -2.15837404e-01 -4.54411991e-02 4.84325826e-01
2.35798106e-01 8.53005528e-01 -1.59764037e-01 1.10903308e-02
3.73547971e-02 -1.76631570e-01 5.74120522e-01 -4.99245822e-01
-2.78797805e-01 5.70090950e-01 -2.39063457e-01 -7.15638161e-01
-4.85878676e-01 -5.62677562e-01 -1.04595959e+00 2.92472932e-02
-1.06408648e-01 -2.45661408e-01 2.09308133e-01 9.44729209e-01
4.04163033e-01 6.30306005e-01 4.31380689e-01 -1.21891749e+00
-1.54277042e-01 -1.06215501e+00 -6.30519331e-01 6.07039630e-01
4.67652857e-01 -5.57383537e-01 -2.66977251e-01 2.68479347e-01] | [11.273797035217285, -1.2614222764968872] |
1b8afcad-31b8-45aa-820f-db45b61cba11 | mathematical-and-preclinical-investigation-of | 2004.11325 | null | https://arxiv.org/abs/2004.11325v1 | https://arxiv.org/pdf/2004.11325v1.pdf | Mathematical and Preclinical Investigation of Respiratory Sinus Arrhythmia Effects on Cardiac Output | Respiratory sinus arrhythmia (RSA) is heart rate variability in synchrony with respiration although its functional significance not clear. The loss of sinus arrhythmia may indicate underlying heart failure or disease; therefore, there would be a great advantage of knowing how it works and affects the cardio-respiratory system, especially by providing a mathematical model. To this end, Windkessel model and cardiovascular partial differential equations are used to obtain cardiac output based on the elasticity of left ventricle, which is related to RSA. By solving the corresponding equations, it would be possible to propose a new model to predict the RSA effects on cardiac output. | ['Sahar Rahbar'] | 2020-04-23 | null | null | null | null | ['heart-rate-variability'] | ['medical'] | [-1.86248764e-01 -5.76707013e-02 -1.58359692e-01 1.65260568e-01
7.73599565e-01 -5.25991321e-01 -1.73122823e-01 -2.69547045e-01
-3.80112120e-04 8.82739007e-01 -4.09561768e-02 -3.39100927e-01
-9.13705602e-02 -6.90200031e-01 1.18449032e-01 -7.02271640e-01
-7.95548409e-02 5.72122000e-02 -7.32150152e-02 -8.24599639e-02
2.89238870e-01 1.00654900e+00 -7.88082361e-01 -4.24450845e-01
5.60950279e-01 5.16695440e-01 -3.66138224e-03 7.53845870e-01
1.77978158e-01 7.84962296e-01 -7.35187590e-01 2.51606852e-01
-5.32030873e-02 -1.22473967e+00 -5.40838599e-01 -3.24372023e-01
-6.64079607e-01 -5.32632411e-01 -3.47858667e-02 2.08328560e-01
6.15173161e-01 -1.41267344e-01 6.85593367e-01 -1.01472282e+00
7.53955217e-03 4.64595020e-01 -1.55909464e-01 6.44908190e-01
1.04146287e-01 1.54922027e-02 3.89671206e-01 -7.39739835e-01
2.73325533e-01 7.02599943e-01 8.15435827e-01 6.83307290e-01
-1.17055583e+00 -5.00448465e-01 -5.54959714e-01 -3.41585040e-01
-1.37569344e+00 -6.22698776e-02 9.15512860e-01 -3.46182317e-01
4.03770328e-01 4.74051207e-01 1.28783274e+00 5.12489200e-01
7.08572567e-01 -8.29829127e-02 1.02268350e+00 -3.08726549e-01
-1.77293867e-01 3.12189668e-01 2.35591456e-01 4.18780357e-01
8.79275680e-01 2.38607183e-01 -3.14891160e-01 -2.52607852e-01
1.08201826e+00 -2.63905916e-02 -8.05910707e-01 -2.30122685e-01
-1.24266481e+00 4.27509755e-01 6.28782660e-02 7.79307842e-01
-3.57159108e-01 3.35081041e-01 4.00417119e-01 1.52428940e-01
-1.71032399e-01 5.52213609e-01 -5.20976543e-01 -2.53539205e-01
-8.31502020e-01 1.40093833e-01 8.06508243e-01 -2.00424320e-03
2.85184115e-01 4.20001835e-01 1.62616149e-01 5.50457239e-01
6.02511585e-01 6.86355352e-01 5.68494678e-01 -1.11557639e+00
-8.72045830e-02 5.37589908e-01 -3.17967236e-02 -1.00260949e+00
-7.10325062e-01 -6.14099920e-01 -1.10977674e+00 1.53236315e-01
4.18705016e-01 -1.90549001e-01 -6.60307333e-03 1.46105874e+00
3.20495903e-01 2.81644583e-01 2.44346112e-01 8.95261884e-01
6.82389319e-01 3.69893551e-01 1.12633631e-01 -9.00723696e-01
1.11503518e+00 -1.13012359e-01 -1.06824398e+00 4.34750587e-01
7.06953526e-01 -6.82049274e-01 6.09719515e-01 1.57868609e-01
-8.71274710e-01 -7.01520443e-01 -1.01477385e+00 3.32522482e-01
4.04055603e-02 -4.56507541e-02 2.08946675e-01 9.27674353e-01
-7.06641078e-01 8.53768587e-01 -9.82865155e-01 -2.33831361e-01
-2.20076784e-01 -5.08174412e-02 7.28179067e-02 7.98667669e-01
-1.31497598e+00 1.23688126e+00 -2.74693910e-02 5.35067081e-01
-2.75820464e-01 -8.38953137e-01 -5.33515513e-01 4.42837290e-02
-2.48673096e-01 -1.21872687e+00 5.23231864e-01 -3.55894327e-01
-1.36356318e+00 5.86115599e-01 -2.01316923e-01 -1.98022813e-01
4.31779027e-01 -1.85431521e-02 -3.30111474e-01 3.05134147e-01
-3.32365751e-01 -2.02946782e-01 4.31188405e-01 -1.15011466e+00
3.22765470e-01 -3.97483706e-01 -5.83968461e-01 1.25995249e-01
1.30990759e-01 -1.53656444e-02 5.98166287e-01 -6.42020226e-01
5.01515687e-01 -1.06361878e+00 -1.09670289e-01 -1.22748524e-01
-1.45655394e-01 2.72112697e-01 8.12914371e-01 -7.64549077e-01
1.56186163e+00 -2.13882327e+00 -5.27011901e-02 3.51498216e-01
3.29329550e-01 4.69244033e-01 6.03685379e-01 7.37708330e-01
-2.00714111e-01 5.52947223e-01 -3.73590916e-01 4.86497372e-01
-6.82554126e-01 2.48322010e-01 -3.52027714e-01 5.86631060e-01
9.55625698e-02 7.14982331e-01 -5.11896908e-01 -4.77694422e-01
2.62060672e-01 7.38012135e-01 -1.47032384e-02 1.00437254e-01
6.30503476e-01 9.12196398e-01 -6.18449569e-01 3.55481178e-01
6.63836658e-01 -2.68146861e-02 3.58438402e-01 -6.79877102e-02
-3.22202414e-01 1.57814100e-01 -9.58025217e-01 8.23844612e-01
-2.56488062e-02 4.73584265e-01 -1.76731661e-01 -9.61607873e-01
1.34562743e+00 9.51905191e-01 8.26979756e-01 -2.69317538e-01
1.31466180e-01 2.77481467e-01 4.57366049e-01 -6.64747000e-01
-1.03602082e-01 -8.01568151e-01 3.79105836e-01 6.18693471e-01
-5.51572680e-01 -3.76679540e-01 -2.24571735e-01 -2.70969961e-02
7.57775486e-01 2.78401375e-01 6.13017201e-01 -6.01697683e-01
9.68704879e-01 -3.22220594e-01 7.71057010e-01 1.75037965e-01
-3.87413651e-01 3.81256044e-01 7.39420235e-01 -6.79579794e-01
-6.61003530e-01 -9.53836739e-01 -5.25045455e-01 -4.13203627e-01
9.85403955e-02 -1.83877833e-02 -4.81713474e-01 4.76182736e-02
5.11422716e-02 6.49150968e-01 -2.37797216e-01 -3.15488994e-01
-8.87508571e-01 -9.20190036e-01 5.76206803e-01 2.48899952e-01
3.26331139e-01 -1.32140231e+00 -1.16343033e+00 4.56170648e-01
-1.55965701e-01 -7.51044571e-01 6.39045164e-02 -1.93280905e-01
-1.62007964e+00 -1.26971281e+00 -7.09521234e-01 -1.57460630e-01
3.73587817e-01 1.63278341e-01 8.79655600e-01 7.49170423e-01
-7.48127937e-01 1.25609905e-01 7.48517960e-02 -5.15400112e-01
-8.02340269e-01 -1.30701989e-01 2.76835501e-01 -2.82652736e-01
-3.64618152e-01 -9.61642921e-01 -8.66209805e-01 4.03111339e-01
-5.83789647e-01 -3.35228980e-01 1.89416394e-01 1.58135086e-01
4.31031406e-01 -1.90652102e-01 1.18552721e+00 -6.26264274e-01
5.45009375e-01 -3.28673095e-01 -3.46301287e-01 -8.59342590e-02
-1.07442856e+00 -2.50687242e-01 5.59185207e-01 -1.15970578e-02
-7.77850389e-01 -3.35282296e-01 1.03800580e-01 -1.27754971e-01
5.03769219e-02 1.84232160e-01 1.42553389e-01 9.90270972e-02
5.93179345e-01 3.94834220e-01 3.64850074e-01 -2.85737067e-01
-2.83219576e-01 3.21621090e-01 2.69724101e-01 -1.76981360e-01
8.13287675e-01 1.66107893e-01 7.54920006e-01 -1.06674504e+00
-4.00220692e-01 -3.30334693e-01 -6.17642045e-01 -3.62019092e-01
9.15913105e-01 -5.06719112e-01 -9.46395695e-01 3.34488988e-01
-9.48840380e-01 -2.77443320e-01 -4.34840292e-01 1.05208242e+00
-6.13109708e-01 7.20372498e-01 -3.71866614e-01 -1.11356378e+00
-4.97024089e-01 -5.25629759e-01 -8.81920457e-02 1.92766294e-01
-5.13634443e-01 -1.16336489e+00 5.33912539e-01 1.41057909e-01
5.62110603e-01 6.97686195e-01 9.50650811e-01 -1.68419212e-01
-3.20182174e-01 1.00642510e-01 2.27798015e-01 5.41908979e-01
3.72515291e-01 6.07854486e-01 -5.92442989e-01 -5.36682270e-02
7.12353647e-01 4.18409944e-01 2.77167976e-01 6.96038842e-01
9.22065735e-01 5.76887615e-02 -3.25408250e-01 5.55972159e-01
1.53621864e+00 5.76130629e-01 8.78082275e-01 -1.00776359e-01
4.09095019e-01 5.43642640e-01 5.44261694e-01 4.98666734e-01
5.23431832e-03 4.16469157e-01 6.85711727e-02 -2.20959589e-01
-1.98047101e-01 1.32140607e-01 1.47871196e-01 9.45646703e-01
-6.78018868e-01 1.64152056e-01 -7.95811832e-01 1.89904675e-01
-1.29231679e+00 -1.09498310e+00 -9.56300080e-01 2.41578197e+00
6.62915885e-01 -7.88976550e-02 2.56934494e-01 5.96265078e-01
5.08827448e-01 -1.57105535e-01 -2.14441463e-01 -7.27524221e-01
3.34052183e-02 2.10281879e-01 2.26425588e-01 5.16890883e-01
-1.91980749e-01 9.95625481e-02 7.75473452e+00 -5.92518687e-01
-1.39555848e+00 -3.83978039e-01 4.87227142e-01 4.78403687e-01
-2.69530058e-01 2.18782127e-01 -5.97408354e-01 5.97202659e-01
1.08447862e+00 -4.96164292e-01 1.45717070e-01 2.44951949e-01
8.14204812e-01 -2.43496597e-01 -6.36098683e-01 7.54267037e-01
-2.47745544e-01 -8.63504231e-01 -3.61655623e-01 1.67461082e-01
-1.40671255e-02 -5.31961441e-01 -3.95276070e-01 -1.67563036e-01
-8.67257953e-01 -7.39766359e-01 -7.43188113e-02 1.03785050e+00
6.65256798e-01 -5.24336696e-01 7.58684695e-01 5.12298346e-01
-1.04777050e+00 2.42356956e-01 -1.61405489e-01 -1.26090705e-01
2.03836292e-01 9.01923537e-01 -1.08609211e+00 3.16769511e-01
1.85534894e-01 3.71845216e-01 -2.76342988e-01 1.12123930e+00
-2.06154108e-01 8.75167489e-01 -2.94230521e-01 2.23452926e-01
-7.34815955e-01 -4.33552146e-01 7.65751898e-01 5.00773072e-01
5.58025181e-01 4.31723654e-01 -5.47681630e-01 1.33968508e+00
4.03726637e-01 3.99384141e-01 -9.10687625e-01 -1.29143119e-01
2.97604024e-01 1.14056039e+00 -8.28944206e-01 -1.09345824e-01
-2.58473933e-01 4.24537510e-01 -6.28710687e-01 2.14582741e-01
-7.80010879e-01 -1.42789215e-01 5.03811955e-01 5.68116188e-01
-3.52752298e-01 -1.64748490e-01 -7.37145185e-01 -1.03492045e+00
-2.21858561e-01 -4.17435110e-01 4.06325310e-02 -6.29807174e-01
-3.90326023e-01 3.47598314e-01 -1.82988588e-02 -1.17737496e+00
-2.66793877e-01 -2.67598890e-02 -6.96446419e-01 1.48005509e+00
-1.36010492e+00 -5.12279153e-01 -3.20058048e-01 2.79167503e-01
-4.35045362e-02 1.61895379e-01 1.19415939e+00 7.60651156e-02
-5.08363307e-01 1.18004054e-01 -6.56769514e-01 -1.69999167e-01
6.20097995e-01 -1.16799152e+00 -1.40079305e-01 5.38831651e-01
-5.50322175e-01 8.03409398e-01 7.98609316e-01 -7.08969772e-01
-8.06454897e-01 -4.91721928e-01 1.32261884e+00 -3.66906673e-01
2.49992028e-01 3.59508485e-01 -1.05480576e+00 1.60844833e-01
-1.22491091e-01 2.87120908e-01 9.27916288e-01 -4.91589785e-01
4.14409220e-01 -4.47786599e-01 -1.00094855e+00 3.32119524e-01
5.36190569e-01 -2.58865446e-01 -7.08804667e-01 -3.08788046e-02
3.67000133e-01 4.32295687e-02 -1.36215413e+00 6.87794566e-01
7.29699850e-01 -9.46219563e-01 9.65694249e-01 -1.20585881e-01
4.97044213e-02 -4.96444255e-01 6.29765928e-01 -9.63458955e-01
-9.29587185e-02 -9.08690512e-01 3.17064300e-02 1.08948541e+00
3.48209769e-01 -1.09096205e+00 2.65027076e-01 5.48311293e-01
-1.72881258e-03 -6.24706447e-01 -6.75150335e-01 -6.34932518e-01
-9.46315527e-02 3.04182291e-01 3.54070634e-01 1.12550998e+00
1.78787425e-01 2.28416547e-01 -3.33849072e-01 -7.84127936e-02
3.64739925e-01 3.27280276e-02 5.49412489e-01 -1.52482808e+00
-1.26477718e-01 -1.60645306e-01 3.39588188e-02 -8.51207078e-02
-3.05709928e-01 -5.60837507e-01 -5.00081599e-01 -1.53561091e+00
-2.24752471e-01 -8.54480684e-01 -6.34582281e-01 -6.97416216e-02
-3.07868719e-01 -3.26192901e-02 6.48829192e-02 4.25117105e-01
7.01816857e-01 1.96359113e-01 1.70432031e+00 8.99842143e-01
-8.78380895e-01 4.84419137e-01 -4.67590570e-01 4.62106049e-01
1.14999747e+00 -8.22357297e-01 -6.05181754e-01 6.18522942e-01
3.22953403e-01 9.20226097e-01 2.76781976e-01 -9.38912034e-01
-4.31687623e-01 -1.78484946e-01 4.28131223e-01 -5.32596171e-01
5.89175522e-02 -8.27050805e-01 8.97957146e-01 1.26867771e+00
-4.31171283e-02 4.13712800e-01 -1.57330409e-02 5.16429693e-02
-1.60902381e-01 -4.73479092e-01 8.98104846e-01 -1.46869346e-01
4.26514745e-01 2.14790311e-02 -8.57233822e-01 -3.60679403e-02
1.12229347e+00 -4.81661141e-01 -8.18888694e-02 -8.69026780e-02
-9.55200613e-01 -4.40428197e-01 3.62790555e-01 -7.24960938e-02
5.53290963e-01 -1.04954612e+00 -5.49074411e-01 1.76814497e-01
-3.22156191e-01 -3.11802208e-01 3.09715152e-01 1.38067532e+00
-1.27460372e+00 5.79254508e-01 -4.23301011e-01 -4.01140898e-01
-1.19566715e+00 5.04193306e-01 9.03902650e-01 1.99641623e-02
-6.98672950e-01 2.23794132e-01 -4.48509082e-02 1.23618267e-01
-3.35739255e-01 -3.70772660e-01 -6.39466584e-01 -1.27734721e-01
2.62439042e-01 6.42922997e-01 -5.18475294e-01 -2.61603594e-01
-5.51688194e-01 8.56417298e-01 8.95257771e-01 2.06721753e-01
7.88567662e-01 -3.33631098e-01 -5.67917466e-01 8.53684545e-01
6.25948429e-01 4.91251349e-01 -5.48329115e-01 3.00939679e-01
-3.30187649e-01 -2.42209852e-01 -8.37805942e-02 -6.88015819e-01
-9.79956448e-01 8.31774116e-01 5.76732516e-01 5.05301416e-01
1.15045202e+00 -4.30511028e-01 6.15838826e-01 2.08764225e-02
1.60108004e-02 -7.62833774e-01 2.08284017e-02 -3.50097865e-01
9.74699914e-01 -6.94433928e-01 2.13988408e-01 -7.47478962e-01
-5.58967412e-01 1.46319509e+00 2.62527376e-01 -2.48139873e-01
1.06318450e+00 3.02916855e-01 3.99784297e-01 1.00399777e-01
-5.76815665e-01 3.39636117e-01 -5.37976958e-02 5.01459777e-01
9.75274682e-01 6.88323677e-02 -1.42713392e+00 3.37531149e-01
-1.41427711e-01 4.26049203e-01 8.93116534e-01 5.19676268e-01
-4.82769877e-01 -1.25165653e+00 -3.10300171e-01 1.45396665e-01
-8.28649759e-01 2.72224098e-01 -2.03684390e-01 9.43903923e-01
8.68076757e-02 7.37932205e-01 -3.24761271e-01 2.76818484e-01
2.87795126e-01 5.05974591e-01 3.62352282e-01 -4.25596893e-01
-6.76837146e-01 2.27622777e-01 -1.12694286e-01 -3.26705068e-01
-5.96200168e-01 -7.72817552e-01 -1.67457116e+00 -7.15541691e-02
-4.21204567e-01 2.09516898e-01 7.01094806e-01 6.46284521e-01
7.80081153e-02 6.07139409e-01 8.42943251e-01 1.97104216e-01
-4.22651142e-01 -9.43419933e-01 -1.11223447e+00 -1.04221683e-02
3.48779857e-01 -4.63475585e-01 -8.36353123e-01 1.70942508e-02] | [14.06822395324707, 3.021376609802246] |
99bc8379-fac2-4d2b-9deb-92026e092fef | camera-fingerprint-a-new-perspective-for | 1610.07728 | null | http://arxiv.org/abs/1610.07728v1 | http://arxiv.org/pdf/1610.07728v1.pdf | Camera Fingerprint: A New Perspective for Identifying User's Identity | Identifying user's identity is a key problem in many data mining
applications, such as product recommendation, customized content delivery and
criminal identification. Given a set of accounts from the same or different
social network platforms, user identification attempts to identify all accounts
belonging to the same person. A commonly used solution is to build the
relationship among different accounts by exploring their collective patterns,
e.g., user profile, writing style, similar comments. However, this kind of
method doesn't work well in many practical scenarios, since the information
posted explicitly by users may be false due to various reasons. In this paper,
we re-inspect the user identification problem from a novel perspective, i.e.,
identifying user's identity by matching his/her cameras. The underlying
assumption is that multiple accounts belonging to the same person contain the
same or similar camera fingerprint information. The proposed framework, called
User Camera Identification (UCI), is based on camera fingerprints, which takes
fully into account the problems of multiple cameras and reposting behaviors. | ['Shikui Wei', 'Xiang Jiang', 'Xindong Wu', 'Yao Zhao', 'Ruizhen Zhao'] | 2016-10-25 | null | null | null | null | ['product-recommendation'] | ['miscellaneous'] | [ 2.78981447e-01 -4.67025459e-01 -2.89661258e-01 -3.30034822e-01
-1.42957583e-01 -8.31980884e-01 4.72995967e-01 3.44612926e-01
-2.05628276e-01 1.97946370e-01 5.89428656e-03 -1.71739161e-01
-1.40265182e-01 -7.78351486e-01 -2.74513215e-01 -5.36493242e-01
5.88811815e-01 2.22206935e-01 8.66382569e-02 2.64132291e-01
6.61354601e-01 4.54499781e-01 -1.27806246e+00 1.52214959e-01
5.86821914e-01 8.86529684e-01 1.60829335e-01 2.99950749e-01
-2.49569863e-01 4.84501123e-01 -3.04240733e-01 -8.00149143e-01
3.90562974e-02 -3.24950516e-01 -5.78534126e-01 2.83557326e-01
2.79077888e-01 -5.74570000e-01 -7.35992789e-02 1.46323538e+00
-6.64048865e-02 6.40095100e-02 3.85507017e-01 -1.31511986e+00
-7.24746823e-01 4.42291796e-01 -9.66229618e-01 -5.88606521e-02
6.23262823e-01 -4.94719267e-01 8.84584486e-01 -5.93467355e-01
2.23073646e-01 1.06221676e+00 7.98985839e-01 5.52886724e-01
-1.05724204e+00 -8.95365655e-01 -2.06681956e-02 1.49840757e-01
-1.59983599e+00 -3.87642890e-01 7.23689497e-01 -5.26967347e-01
-1.10258967e-01 3.57071102e-01 2.46111855e-01 8.89139950e-01
-3.63458157e-01 6.91166759e-01 9.39679861e-01 -4.12512571e-01
-3.21895778e-02 9.77851629e-01 6.49322927e-01 3.31705779e-01
6.00917399e-01 -4.33282912e-01 -2.28936225e-01 -5.58833241e-01
5.65914929e-01 6.86410546e-01 -9.98643972e-03 -2.76702434e-01
-9.46139514e-01 7.73472846e-01 -7.87246004e-02 5.39841115e-01
-8.52382407e-02 -4.76951778e-01 2.10602835e-01 -1.24867875e-02
6.44937232e-02 3.21296066e-01 -2.87600923e-02 -1.15582654e-02
-8.06224763e-01 6.63212314e-02 7.14131474e-01 9.66487765e-01
1.11559737e+00 -6.84418261e-01 3.99025112e-01 9.10056531e-01
5.85156500e-01 2.17671201e-01 6.86778009e-01 -6.34298503e-01
3.84494036e-01 9.60160613e-01 5.17138600e-01 -1.69792473e+00
1.62804872e-02 -8.72692615e-02 -8.35509479e-01 -3.45763654e-01
6.46882772e-01 -2.36413211e-01 -8.56340304e-02 1.61236107e+00
3.04797202e-01 4.48706388e-01 -7.68481418e-02 6.42227709e-01
5.36876738e-01 2.97565848e-01 -1.32788226e-01 -1.89416364e-01
1.47350168e+00 -4.43831235e-01 -7.10404813e-01 -1.70077369e-01
2.88068831e-01 -8.47071290e-01 5.30604303e-01 3.42736870e-01
-5.37400246e-01 -6.98595643e-01 -6.59005702e-01 4.45612937e-01
-3.85257035e-01 4.69767004e-01 5.65125406e-01 1.08039713e+00
-4.08948392e-01 4.78287041e-01 -2.91838706e-01 -9.83938873e-01
-6.44953502e-03 4.62483317e-01 -4.13557500e-01 -8.08831155e-02
-9.63979900e-01 3.54505390e-01 1.44553915e-01 9.58230719e-02
-7.90150687e-02 -1.65240824e-01 -4.60329473e-01 -1.01346821e-01
4.23283488e-01 -2.70129085e-01 1.07021487e+00 -1.29353416e+00
-1.27816463e+00 9.26461697e-01 -4.81234819e-01 -8.25520903e-02
5.22786200e-01 -1.38096556e-01 -8.19826543e-01 -2.60833561e-01
2.38839284e-01 -2.43378669e-01 7.83621073e-01 -1.21776867e+00
-7.30989456e-01 -8.38718832e-01 1.27344355e-01 -8.60911533e-02
-6.56676054e-01 3.91384840e-01 -7.31470406e-01 -3.89221996e-01
2.21029535e-01 -1.14850748e+00 2.04071864e-01 -5.35556197e-01
-7.45603144e-01 -1.87575385e-01 9.53327477e-01 -6.98054433e-01
1.52684724e+00 -2.23411942e+00 -2.37275228e-01 7.79028654e-01
1.82734475e-01 2.70916760e-01 5.02264678e-01 5.63249946e-01
2.66152788e-02 3.58327627e-01 9.08726156e-02 -4.19984579e-01
-2.44626641e-01 -4.89582345e-02 -8.04053470e-02 4.70871180e-01
-4.02050793e-01 2.46446565e-01 -7.93156981e-01 -5.45421600e-01
8.90126750e-02 7.73005709e-02 -1.95554599e-01 2.44244993e-01
4.12924796e-01 5.25655210e-01 -7.75210857e-01 5.09792805e-01
8.97941470e-01 -4.05117482e-01 6.38961196e-01 -1.94124237e-01
-2.37360209e-01 -2.84456015e-01 -1.62299168e+00 1.21105576e+00
-2.51240492e-01 5.31708837e-01 -2.95106880e-02 -6.74072504e-01
9.19301212e-01 3.71605754e-01 5.32302141e-01 -2.26078093e-01
7.60076642e-02 3.03700745e-01 -1.07184343e-01 -6.00953579e-01
6.61886096e-01 1.58182099e-01 -4.67233807e-02 8.94073069e-01
-4.20135528e-01 1.03904843e+00 7.37014785e-02 1.94446295e-02
7.32389748e-01 -3.39744836e-01 5.29425204e-01 4.12024036e-02
9.57468987e-01 -3.84773940e-01 4.37053621e-01 8.33107591e-01
-1.49405122e-01 5.65581322e-01 2.53971130e-01 -1.23303100e-01
-9.37023163e-01 -5.69203258e-01 -1.12346664e-01 7.83135056e-01
5.87008655e-01 -3.95722508e-01 -8.90545905e-01 -5.41051745e-01
8.54659900e-02 6.71392232e-02 -5.16461253e-01 9.20570418e-02
-2.52531797e-01 -6.43384576e-01 5.80727339e-01 1.47729918e-01
7.56368995e-01 -6.47647321e-01 -1.59855306e-01 2.72450775e-01
-3.63841712e-01 -1.13141859e+00 -6.08770251e-01 -7.97918320e-01
-5.26500881e-01 -1.25895619e+00 -8.40036571e-01 -5.53561568e-01
7.69322813e-01 9.03763533e-01 4.29717898e-01 4.87991840e-01
8.61310661e-02 3.13125879e-01 -3.49811047e-01 -7.68548027e-02
-3.89419019e-01 1.40340716e-01 4.64320064e-01 1.06136918e+00
7.61781216e-01 -2.95318872e-01 -5.92662930e-01 8.44617546e-01
-8.75270009e-01 -1.92716941e-01 3.65771413e-01 3.06548566e-01
1.24736195e-02 5.11800528e-01 3.43099177e-01 -1.39914906e+00
6.77923977e-01 -8.13806415e-01 -2.26587504e-01 6.02722824e-01
-5.42944908e-01 -4.03960466e-01 5.20711780e-01 -5.60905576e-01
-1.12890422e+00 1.33954793e-01 9.39667523e-02 -3.34566325e-01
-4.93874609e-01 3.00211430e-01 -4.56079721e-01 -1.30321324e-01
1.75906211e-01 1.30907163e-01 3.63211259e-02 -7.82052338e-01
-1.37487710e-01 1.18353808e+00 4.01684046e-01 -3.06729108e-01
6.91097796e-01 4.26279694e-01 -5.10099173e-01 -8.87003601e-01
-6.18909359e-01 -1.21583450e+00 -6.04783535e-01 -3.10422391e-01
8.41405690e-01 -6.95384026e-01 -1.17826891e+00 8.89294565e-01
-1.22662079e+00 5.37237883e-01 6.84485972e-01 5.42455554e-01
9.66824219e-02 9.72540617e-01 -3.11483979e-01 -1.01475275e+00
4.06394489e-02 -9.76052165e-01 6.39221609e-01 6.17863536e-01
-5.48425913e-01 -9.75781620e-01 1.02889083e-01 5.35453558e-01
2.06835300e-01 1.41052112e-01 6.58267498e-01 -9.01358008e-01
-3.96114379e-01 -7.07508266e-01 -4.83327925e-01 2.31878713e-01
5.99673450e-01 1.89443842e-01 -9.01938260e-01 -1.72048867e-01
-4.98567931e-02 3.25589925e-01 1.03036247e-01 -4.07053493e-02
1.08423376e+00 -4.68682796e-01 -5.03543019e-01 2.71628827e-01
1.53993511e+00 4.02279496e-01 6.17925763e-01 2.72279561e-01
9.83221948e-01 8.14960957e-01 2.30306253e-01 7.32661784e-01
2.14983240e-01 8.70351911e-01 1.82252988e-01 1.48435444e-01
7.52547264e-01 -3.96973848e-01 7.53829479e-02 2.78535694e-01
-1.67877689e-01 -3.67030740e-01 -7.54306257e-01 7.49994516e-02
-1.91373456e+00 -1.16938305e+00 -2.91680872e-01 2.52708244e+00
1.69170439e-01 -1.00709543e-01 4.69368309e-01 2.19148666e-01
1.58292079e+00 -3.37086439e-01 -6.06111646e-01 -1.98808938e-01
3.46537381e-01 -3.66886407e-01 7.25766420e-01 2.41863668e-01
-7.66935825e-01 3.72746378e-01 4.83678055e+00 3.91393423e-01
-7.74488211e-01 6.56909049e-02 6.39559686e-01 5.25583029e-01
-1.71600401e-01 3.14120293e-01 -1.15823591e+00 9.58356380e-01
5.65715313e-01 -7.08430186e-02 6.78051293e-01 6.75597131e-01
1.80708647e-01 -2.14980751e-01 -9.61666882e-01 1.24439001e+00
3.32603425e-01 -1.16363311e+00 -1.99624673e-01 5.60783863e-01
4.51092333e-01 -6.15956485e-01 -3.34949978e-02 -1.41391292e-01
4.75965403e-02 -4.59439218e-01 6.38063073e-01 5.11167169e-01
5.68150699e-01 -6.32097960e-01 6.18669569e-01 5.63176215e-01
-1.18707883e+00 -2.50373632e-01 -1.75602257e-01 1.51952714e-01
5.81235625e-03 2.75358289e-01 -6.10536933e-01 4.15968388e-01
5.74700594e-01 8.66041541e-01 -6.38385236e-01 1.09516263e+00
3.03438008e-01 4.08356905e-01 -2.48827592e-01 -7.36988895e-03
2.45379191e-02 -6.71734989e-01 8.85235369e-02 8.24065447e-01
6.09558582e-01 8.33545327e-02 -4.04991992e-02 7.72331715e-01
-8.65928233e-02 2.99006134e-01 -6.42119944e-01 -8.14952999e-02
5.41063070e-01 1.30808461e+00 -7.24757731e-01 -2.31422976e-01
-7.97246337e-01 1.09772205e+00 5.81902862e-02 2.51330763e-01
-7.80461371e-01 -3.01939130e-01 7.07754195e-01 3.33709508e-01
-1.66851461e-01 -2.26363111e-02 1.14626894e-02 -1.10147798e+00
1.04546644e-01 -7.31275916e-01 3.03861201e-01 -5.24860978e-01
-1.29386020e+00 6.38279095e-02 -2.27913126e-01 -1.52058589e+00
-4.32519540e-02 -3.79485726e-01 -9.37358677e-01 9.06430602e-01
-1.05361986e+00 -9.69774365e-01 -5.19210577e-01 8.16570640e-01
1.42249346e-01 -2.13262096e-01 6.51562929e-01 6.42500877e-01
-1.06300902e+00 8.26644123e-01 4.60675567e-01 4.76276577e-01
7.08649933e-01 -7.88223386e-01 2.49127954e-01 7.85084307e-01
5.40538132e-02 1.26399446e+00 4.59204704e-01 -5.89120090e-01
-1.41598916e+00 -6.92080915e-01 1.12990284e+00 -5.45241356e-01
6.61713183e-01 -2.10835829e-01 -8.11278522e-01 6.67734742e-01
-1.14512958e-01 -3.25786650e-01 1.10494399e+00 2.81598389e-01
-1.82435364e-01 -1.57735541e-01 -1.13603806e+00 3.43250364e-01
6.72950327e-01 -8.00166905e-01 4.71266136e-02 2.59680867e-01
-7.20347241e-02 1.96721442e-02 -8.84958148e-01 -2.42445245e-01
9.43895400e-01 -1.12100244e+00 8.65671813e-01 -4.67675507e-01
1.88612252e-01 -3.58785450e-01 -5.10613434e-02 -6.84410691e-01
-1.83126181e-01 -5.96313059e-01 2.23150119e-01 2.03373003e+00
1.02069579e-01 -7.19654858e-01 9.47303176e-01 1.22355998e+00
5.34695923e-01 -1.18138328e-01 -5.35771072e-01 -5.27643859e-01
-6.79082572e-01 -2.50970274e-01 1.06867242e+00 1.11833227e+00
1.48149028e-01 2.02098131e-01 -7.59139359e-01 5.32675505e-01
6.44299686e-01 1.66956544e-01 1.05829954e+00 -1.48115766e+00
-4.66194183e-01 -2.44196773e-01 -2.64580786e-01 -9.30957437e-01
5.89523874e-02 -3.88700008e-01 -4.09992456e-01 -9.81514096e-01
5.96431851e-01 -8.97370696e-01 -1.85306862e-01 -2.08629686e-02
4.35448885e-02 2.13973969e-01 1.35392338e-01 9.06957507e-01
-4.28432077e-01 -4.08627063e-01 6.92369819e-01 1.47234108e-02
-2.92890906e-01 7.96764314e-01 -9.66007352e-01 7.84711361e-01
7.44690359e-01 -3.08462590e-01 -3.27963442e-01 -3.60092491e-01
3.98797423e-01 2.00151384e-01 3.97489518e-01 -9.65679705e-01
4.11202103e-01 -2.17198476e-01 2.30005726e-01 -1.84658021e-01
9.08948779e-02 -1.16146171e+00 5.83082855e-01 9.74915400e-02
-2.77130574e-01 3.19713980e-01 -2.98812807e-01 8.56548429e-01
-2.23069847e-01 -7.09825933e-01 6.57058179e-01 -5.17666817e-01
-8.01692426e-01 3.45237345e-01 -3.88374120e-01 -3.78173292e-01
1.03453159e+00 -6.22003317e-01 -2.12775603e-01 -1.72130078e-01
-5.68967938e-01 -9.03974622e-02 8.59126866e-01 3.87394518e-01
3.57355267e-01 -1.31650174e+00 -2.36546531e-01 1.62788808e-01
4.63218898e-01 -6.32715821e-01 3.63154650e-01 4.44748074e-01
-1.67454153e-01 9.53181311e-02 -1.44712329e-01 -5.43980300e-01
-1.71740925e+00 4.68755782e-01 1.10238373e-01 5.81105007e-03
-1.04485825e-01 3.59261245e-01 9.51792598e-02 -1.84425250e-01
1.41937643e-01 2.22938508e-01 -7.18134820e-01 3.15723330e-01
6.99070275e-01 4.68973368e-01 -5.55546470e-02 -1.10930288e+00
-4.74546909e-01 8.78562212e-01 -2.80187100e-01 7.05613121e-02
6.93948865e-01 -5.31057835e-01 -1.13008246e-01 3.73837620e-01
1.44743359e+00 1.73254609e-01 -6.68628454e-01 -5.35050869e-01
8.52431059e-02 -9.49438572e-01 -4.08756524e-01 -6.24483451e-02
-1.00792480e+00 6.35541677e-01 5.96536756e-01 6.57703519e-01
6.74086452e-01 -2.72171795e-01 9.97591376e-01 2.33644068e-01
5.31433642e-01 -1.07502389e+00 -1.56353131e-01 7.01096728e-02
-3.28763835e-02 -1.51873863e+00 2.79932115e-02 -3.88213277e-01
-6.34027660e-01 1.18265176e+00 6.04370475e-01 2.32881978e-01
7.46323168e-01 -4.29223448e-01 -2.77690858e-01 8.70716870e-02
1.48997039e-01 5.72459772e-02 4.58077826e-02 2.74491161e-01
3.72160524e-01 1.34057641e-01 -1.54875994e-01 5.47446787e-01
1.65761590e-01 -8.84076133e-02 6.71643317e-01 7.23765016e-01
-2.56772339e-01 -1.34765911e+00 -5.62930763e-01 4.97696757e-01
-6.51675165e-01 1.88342631e-01 -4.38251257e-01 4.24620330e-01
2.92859346e-01 1.17247105e+00 9.34775844e-02 -6.44360244e-01
1.79721802e-01 -1.28756329e-01 1.30510390e-01 -6.13395452e-01
-5.09568512e-01 -2.99425364e-01 -1.99239478e-01 4.26352210e-03
-7.68823206e-01 -1.04503298e+00 -7.10837543e-01 -8.76998305e-01
-5.00485122e-01 1.31453708e-01 7.07581162e-01 1.00626433e+00
9.84727591e-02 -3.68113101e-01 1.12017310e+00 -3.99263889e-01
-2.00435400e-01 -5.02079427e-01 -8.69265974e-01 8.56936514e-01
1.20110072e-01 -5.90609193e-01 -2.49200881e-01 1.60120547e-01] | [14.742260932922363, 1.025586485862732] |
4c3a0606-815d-4b28-9a67-3c406c316dd8 | an-evolutionary-forest-for-regression | null | null | https://ieeexplore.ieee.org/document/9656554 | https://ieeexplore.ieee.org/document/9656554 | An Evolutionary Forest for Regression | Random forest (RF) is a type of ensemble-based machine learning method that has been applied to a variety of machine learning tasks in recent years. This article proposes an evolutionary approach to generate an oblique RF for regression problems. More specifically, our method induces an oblique RF by transforming the original feature space to a new feature space through the evolutionary feature construction method. To speed up the searching process, the proposed method evaluates each set of features based on a decision tree (DT) rather than an RF. In order to obtain an RF, we archive top-performing features and corresponding trees during the search. In this way, both the features and the forest can be constructed simultaneously in a single run. The proposed evolutionary forest is applied to 117 benchmark problems with different characteristics and compared with some state-of-the-art regression methods, including several variants of the RF and gradient boosted DTs (GBDTs). The experimental results suggest that the proposed method outperforms the existing RF and GBDT methods. | ['Hu Zhang', 'Aimin Zhou', 'Hengzhe Zhang'] | 2021-11-20 | null | null | null | ieee-transactions-on-evolutionary-computation-2 | ['penn-machine-learning-benchmark'] | ['miscellaneous'] | [ 0.6468028 -0.44984007 -0.06936543 -0.4526189 -0.34097192 -0.0927317
0.5507391 -0.2144817 -0.2745524 1.1015406 -0.22103955 -0.23604092
-0.396027 -1.0484512 -0.19615489 -1.1953784 0.15915368 0.41521594
0.1832017 -0.18789954 0.5320397 0.5067131 -1.9274981 0.01553491
1.4429646 1.0499848 0.19394045 0.29191306 -0.1250771 0.27331913
-0.6057105 -0.18609765 0.25859708 -0.46644908 -0.4874203 -0.17624398
-0.11192816 0.21311131 0.11033199 0.5907475 0.5443082 0.34946227
0.8991305 -1.2820064 -0.35681012 0.44306365 -0.97334343 0.11106803
0.2409036 -0.21237527 0.87548566 -1.0250274 0.55913645 1.0484215
0.59531283 0.34694868 -1.1910087 -1.0221041 0.35189262 0.52177685
-1.3796369 -0.14412242 0.8573495 -0.3897537 0.6508236 0.42883128
0.9087774 0.8285775 0.50606173 0.6271471 1.5674863 -0.73812634
0.32168826 0.14578138 0.2280807 0.7814062 0.38254863 0.3960259
-0.63007605 -0.43885598 0.27014092 0.10827513 -0.33633634 -0.44276813
-1.0025638 0.98940414 0.49801013 0.19529825 -0.57797354 -0.32675964
0.20169802 0.18416883 0.68567145 0.2625369 -0.5443487 0.30888563
-0.809456 0.36638668 0.67377573 0.46230507 0.8661556 -0.03717596
-0.32529202 0.9476434 0.30239937 0.3622342 0.6037319 -0.13039681
0.49067658 0.7769102 -0.02829327 -0.9656251 -0.35451525 -0.85053575
-0.9767411 0.41044122 -0.12737143 -0.27019873 -0.94552267 1.3409802
0.77212226 0.20809402 0.15490349 0.5951645 0.7252997 0.7005865
-0.05852693 -0.64532036 1.023446 -1.1086956 -0.5393613 -0.04554996
0.34128425 -0.67914975 0.6831567 0.6213565 -0.34628853 -0.7135077
-1.2879956 0.6216394 -0.21226752 0.31961045 0.774047 0.7464927
-0.67170167 0.7335841 -0.5300422 -0.19709066 0.23047687 0.46767193
-0.2810222 -0.15223576 -1.0137699 0.90292996 0.3804186 0.5012948
-0.5066834 -0.19064926 -0.500423 -0.18989319 0.40111095 -1.0005906
0.90634984 -1.0249709 -1.5868013 0.3178177 -0.6262411 0.04652558
0.41651654 -0.37181312 -0.5199453 -0.46697637 -0.01461841 0.09341081
1.0995616 -1.0939521 -1.0492498 -0.7220225 -0.5041351 0.45521504
-0.2642169 -0.01379527 0.1748664 -0.6130144 0.29791492 -1.1390352
-0.4051731 -0.6846557 -0.4728516 -0.46266896 0.92437005 -0.5536965
1.4838862 -1.6581019 0.51495224 0.58684504 -0.07422571 0.14132458
0.06347235 0.28481406 -0.03674946 0.08363099 -0.514006 0.19858405
-0.5592837 0.0443617 -0.05067297 0.27960548 0.13462295 0.37753257
-0.7489789 -0.69884264 -0.02598713 0.25283307 -0.25031608 0.18515763
0.04017372 0.791051 -0.844984 0.7157817 0.78785646 0.37881687
0.22191153 0.11775334 -0.34968004 0.05820123 -1.1943492 1.1964455
-0.69836855 0.24579485 -0.5446482 -1.0960628 1.6499103 0.14139982
0.45067218 -0.31511617 -0.05834242 0.38455027 0.20484881 -0.41040283
0.43677503 -0.02819756 0.08758383 0.56786597 -0.10453881 0.05604691
0.19403426 -0.49232554 0.96383226 0.6362416 0.70300037 -0.03228381
0.9272106 0.06716289 0.9260561 0.47353172 0.4626805 0.08408967
-0.04542482 -0.66783136 -0.7100349 -0.70988643 -0.31673867 1.1603092
-0.01662336 -0.3148096 -0.38896707 -0.97947973 0.07302976 0.8730045
-0.64693373 -0.21970518 -0.600011 -1.2503287 0.00999938 0.2304889
0.8395807 -1.1958836 -0.66547275 0.4424549 -0.23721904 -0.3407331
-0.0954401 0.34475228 -1.1814003 -0.9370552 -0.4602662 -0.68945336
0.66892684 0.3221706 0.9385025 0.07418095 -0.27987203 -0.46493533
-0.84088904 -0.5381743 -0.4640632 0.44696638 -0.08503631 0.23336415
0.28429803 -0.7567599 -0.49655828 0.5973065 -0.44595593 0.16217522
1.0089122 1.2703307 0.7455055 0.45555288 0.71154183 -1.0282393
0.655614 -0.5777138 -0.67008734 0.43682778 -1.1540405 0.4898079
0.7450993 -0.45585647 -1.3666346 0.32965726 0.07037138 0.03950579
0.10534469 0.5475265 -0.08334819 -0.19740787 0.59571713 0.40430576
-0.3103173 -0.6125164 0.10287812 1.0147104 0.10028245 -0.5901097
0.8842975 -0.04044128 0.36589295 -0.25559115 -0.54035985 -0.20182586
-0.7768594 -0.31139097 0.42104658 -0.4116002 -0.3240637 0.48013446
-0.89372504 0.34598747 0.07138579 0.5412456 -0.37322786 -0.10654331
0.1978688 -1.0764971 -0.7846587 -1.0375779 0.80278236 0.33832696
-0.14077432 -0.3296132 0.41704097 0.09608014 0.19302332 0.58782303
1.0761921 -0.6129589 -0.13124564 -0.19084434 0.30258194 0.04391243
0.30209652 0.40363312 -0.72352743 -0.22808036 -0.0786524 0.05554796
1.0278009 0.18260597 1.0712491 -0.2349827 -0.7966433 0.6802211
1.3776486 0.8376565 0.3918084 0.8418224 0.23123801 0.5317429
1.1799884 0.5828863 0.03292757 0.80388457 0.18484296 -0.04041033
0.3241643 -0.2802444 0.24160616 0.75348324 -0.72234607 -0.03497841
-0.6208231 0.10348079 -2.0011477 -1.0010824 -0.09478811 2.438937
0.7415711 -0.12279183 0.06874261 0.5941136 1.1005749 0.02801261
-0.80973005 -0.7344441 0.07557446 0.68781435 0.14890395 -0.05341537
-1.1610368 0.63102126 5.625163 0.7140682 -1.2064749 -0.14440762
0.5448165 0.07697196 -0.17497678 0.13099496 -0.89897007 0.35031748
0.57912487 -0.43745407 0.5079951 0.99118817 -0.02402627 -0.16679469
-0.8079612 0.6672795 -0.2195936 -0.8287354 -0.05108365 -0.03024112
0.90318036 -0.2899072 -0.04219015 0.40354124 0.3669133 -1.0957061
0.4754738 0.52963203 0.79031074 -1.0283833 0.96051294 0.42690355
-1.4539806 -0.403867 -0.3435239 0.22487704 -0.18001069 0.7216074
-0.915328 1.1655031 0.8598463 0.5474675 -0.68922615 1.1131097
-0.3459406 0.5613527 -0.26977074 -0.48445922 -0.10777436 -0.45180193
0.59819835 0.8750442 0.720622 -0.1533178 0.05504452 0.51141965
0.2561787 0.59165627 -0.5776361 0.41750175 0.73472995 1.3986591
-0.38788864 0.04293555 -0.11016691 0.71029663 0.26663494 0.08589824
-0.7775163 -0.59723586 0.07594138 -0.20408976 0.46955308 0.25836018
-0.23662996 -0.9717939 0.14666669 -0.95004505 0.48633614 -0.57358474
-1.080507 1.1695119 0.10757418 -1.5553992 -0.5588124 -0.07406285
-0.5957946 1.0125282 -1.3817668 -0.84821385 -0.6762211 0.61202407
0.544591 -0.46628618 1.0291041 -0.14154817 -0.6200699 0.42431724
0.44727796 -0.49264684 0.6809125 -0.95256275 0.19461188 0.61205614
0.08387335 0.513081 0.5947179 -0.6245219 -1.1046026 -1.092959
0.6506469 0.3291027 0.14007019 0.07950458 -0.5952216 0.36765745
-0.01279541 -0.12531793 0.69762325 0.3440106 -0.08751344 -0.58426774
-1.3789418 0.41821545 1.1810067 0.37681335 -0.44793016 -0.04187065
0.2620835 -0.1702265 -0.7121328 0.8612882 1.0666991 -1.0420507
0.49904004 -0.38449922 0.3646187 -0.58967376 0.00664856 -1.8275958
-0.55733764 -0.44025397 -0.00961477 1.2908539 0.52471614 -1.0330621
0.5804811 0.00867527 0.09995913 -1.3069625 -0.99982387 -0.683897
-0.25020772 0.12597482 1.0529226 0.5428979 -0.417263 0.48943198
-0.2588017 -0.2456445 0.44515303 0.6721693 0.90906054 -1.819275
-0.39206535 -0.21954294 -0.42009476 -0.38496143 0.06362952 -0.96914
-0.04106148 -1.3154827 0.36851588 -1.0882071 -0.58645546 0.42474616
-0.4634954 -0.10913707 0.11812899 0.3438632 0.02268003 0.7011065
1.168899 -0.01215518 -0.48764428 0.6376026 -0.7404664 0.51403415
1.0140036 -0.80798054 -0.29330212 0.18358958 -0.17639968 0.10146347
-0.14728819 -0.9325864 -0.11368545 -0.33318344 0.7158742 -0.7777157
0.0500653 -0.8138431 0.62719893 0.6713358 -0.11884023 0.32876375
-0.18315504 0.6559017 -0.22819272 -0.37793216 0.610819 0.23595902
-0.5564688 0.13913 -0.11863294 -0.5730354 1.3002809 -0.39751512
-0.2265576 0.01953353 -0.5366416 0.01692218 0.1311872 0.3311663
0.7523497 -1.2630763 -1.0194767 0.2807487 0.1305765 0.03241615
-0.06416394 0.81155545 -0.21902841 0.22861883 -0.40879935 -0.45184898
-1.6752961 0.3548444 0.0762607 -0.90905213 -0.54017633 0.57946754
-0.1353122 -0.5501985 -0.1133122 0.27908206 -0.6397247 -0.10522177
0.16295278 0.60629225 0.13674557 -0.41373217 -0.57189137 0.9253035
-0.11324544 -0.10427921 1.5890754 0.32748225 -0.2538214 0.35229093
0.8586014 0.1299171 -0.7391052 -0.20597593 0.13951242 -0.58430254
0.00609965 -0.98736334 -1.2166629 0.28242692 0.6952235 -0.13258116
1.6472325 -0.53585446 0.5345464 0.44009626 0.7018271 -0.90159464
-0.44289476 0.02494917 1.0509576 -0.9367961 0.50603074 -0.45552063
-0.71999925 1.4797629 0.6724172 -0.07301338 0.49474013 0.14490366
-0.35157987 0.3824307 -0.9690702 -0.12402861 0.24801384 0.5612131
0.4573631 0.1332635 -0.99481034 0.5836241 -0.4638503 0.3365448
0.00583894 0.97809625 -0.3819296 -1.4785452 -0.56613016 0.7164696
-0.1487494 -0.05370494 -0.42352083 0.7884877 0.39601952 1.1938891
-0.1563037 -1.0402473 0.30689755 0.16196682 0.5862279 -0.33752456
-0.81274444 -0.20536187 0.2958908 -0.13248758 -0.5105288 -0.7829478
-0.9143824 -0.15974087 -0.9981756 0.48914957 0.84514135 0.7157026
0.26188082 0.58852804 1.4335971 -0.7669254 -0.47727042 -1.14781
-0.4415542 -0.09426711 -0.1399416 -1.2509116 -0.3021587 -0.2836923 ] | [8.293801307678223, 4.21299934387207] |
6730c7e7-47ae-40ff-8d6e-f23b4db7b5e1 | analysis-of-resource-efficient-predictive | null | null | https://aclanthology.org/2020.sustainlp-1.18 | https://aclanthology.org/2020.sustainlp-1.18.pdf | Analysis of Resource-efficient Predictive Models for Natural Language Processing | In this paper, we presented an analyses of the resource efficient predictive models, namely Bonsai, Binary Neighbor Compression(BNC), ProtoNN, Random Forest, Naive Bayes and Support vector machine(SVM), in the machine learning field for resource constraint devices. These models try to minimize resource requirements like RAM and storage without hurting the accuracy much. We utilized these models on multiple benchmark natural language processing tasks, which were sentimental analysis, spam message detection, emotion analysis and fake news classification. The experiment results shows that the tree-based algorithm, Bonsai, surpassed the rest of the machine learning algorithms by achieve higher accuracy scores while having significantly lower memory usage. | ['Ambesh Shekhar', 'Raj Pranesh'] | null | null | https://openreview.net/forum?id=4Dowguaqed | https://openreview.net/pdf?id=4Dowguaqed | emnlp-sustainlp-2020-11 | ['news-classification'] | ['natural-language-processing'] | [ 1.63942218e-01 -1.48169160e-01 -4.89933103e-01 -5.10318220e-01
2.45018005e-01 -1.05486840e-01 7.27017939e-01 4.16933328e-01
-5.65354586e-01 1.11942589e+00 8.83277729e-02 -8.09626043e-01
-3.28950584e-01 -7.25283980e-01 -4.60283384e-02 -5.52778959e-01
1.49168000e-01 5.54870367e-01 3.11797321e-01 -1.71812907e-01
9.67108548e-01 4.77850735e-01 -1.62242055e+00 5.80471635e-01
7.76468396e-01 1.28984880e+00 -3.01387012e-01 4.75484967e-01
-1.89190149e-01 1.20799065e+00 -4.65108782e-01 -6.51658714e-01
3.20586041e-02 -1.05220005e-02 -9.53912079e-01 -6.53318644e-01
-2.75728852e-01 -1.63758919e-01 -9.85263884e-02 1.01030517e+00
1.96678445e-01 -9.80872214e-02 9.03231382e-01 -1.42995811e+00
-2.25817457e-01 7.67727315e-01 -7.25332558e-01 3.66517425e-01
3.39618921e-01 -4.91632730e-01 5.34129679e-01 -5.43001771e-01
5.41050792e-01 1.37385643e+00 7.82124639e-01 5.86325943e-01
-7.11435616e-01 -9.11847949e-01 -4.41412926e-01 5.58695197e-01
-1.21854496e+00 -1.34794876e-01 3.79453957e-01 -1.95377856e-01
1.49638546e+00 7.47674704e-01 5.40751040e-01 7.96204507e-01
7.88245201e-01 6.26838207e-01 1.48872674e+00 -7.77883947e-01
4.55582798e-01 8.28335404e-01 8.08250606e-01 5.68080902e-01
6.33421421e-01 1.71338052e-01 -7.23039448e-01 -8.05458903e-01
-2.27427930e-01 3.52968842e-01 1.64701238e-01 2.65351802e-01
-5.47556043e-01 1.11263931e+00 1.63548172e-01 5.19395769e-01
-2.15097278e-01 -3.96629907e-02 7.39514649e-01 3.41126174e-01
5.61366498e-01 2.40702257e-01 -8.84513617e-01 -2.90293634e-01
-9.55359280e-01 1.19731076e-01 1.06331885e+00 8.46565545e-01
2.91160550e-02 4.34583575e-02 1.25075862e-01 5.66797912e-01
4.75157559e-01 4.37068969e-01 1.18939531e+00 -3.74563783e-01
2.17246622e-01 7.29645073e-01 7.82194585e-02 -1.46907127e+00
-4.43824708e-01 -2.16940880e-01 -1.12873948e+00 -1.60055548e-01
-1.75392851e-01 1.40554816e-01 -1.04185629e+00 8.08707237e-01
1.95857719e-01 2.34926082e-02 1.33940622e-01 4.88648653e-01
7.14050829e-01 7.85541058e-01 3.83690476e-01 -4.98920202e-01
1.30546582e+00 -1.13911557e+00 -8.19477081e-01 1.69688079e-03
9.22573209e-01 -1.13288522e+00 7.40503609e-01 8.91831875e-01
-4.18028682e-01 -2.61325210e-01 -1.19582963e+00 4.68219459e-01
-6.80998981e-01 -8.30641687e-02 1.41835570e+00 1.06686509e+00
-7.07710862e-01 9.53037322e-01 -6.84814215e-01 -3.84746522e-01
5.24702728e-01 4.98490036e-01 -1.72875747e-01 -6.54220805e-02
-1.04260421e+00 1.20252693e+00 7.15904534e-01 -5.25347829e-01
-1.86017439e-01 5.03822863e-02 -3.46485585e-01 1.11486398e-01
-1.09320901e-01 -1.94724634e-01 9.88252878e-01 -1.01873541e+00
-1.18463862e+00 7.84625113e-01 -8.43152255e-02 -1.07992876e+00
8.31946507e-02 -1.49352103e-01 -7.03967035e-01 -1.47076234e-01
-5.23007751e-01 3.73802125e-01 6.61217630e-01 -7.52587438e-01
-7.55306840e-01 -7.73565173e-01 -5.85469663e-01 -9.27678570e-02
-4.31764007e-01 2.50163406e-01 4.64236170e-01 -2.56053567e-01
4.06142980e-01 -9.87052619e-01 -3.84782791e-01 -6.12252772e-01
-4.06551987e-01 -2.21954480e-01 1.37187088e+00 -6.65321827e-01
1.51260591e+00 -1.93998384e+00 -3.24680001e-01 4.51512814e-01
-9.78966877e-02 4.45440680e-01 6.30904675e-01 7.62439072e-02
1.54528618e-01 3.57360959e-01 2.19758078e-01 2.99479425e-01
-5.01083672e-01 5.63861728e-01 -3.68892103e-01 4.12023038e-01
-5.45854151e-01 5.25265336e-01 -4.03980881e-01 -8.54673862e-01
3.13468397e-01 8.62346664e-02 -3.46003205e-01 -3.29168528e-01
6.17091544e-03 -3.96974713e-01 -2.73144096e-01 1.05494237e+00
7.45502889e-01 -2.06198156e-01 2.57955551e-01 -8.94877017e-02
1.89237535e-01 1.67034283e-01 -8.13284934e-01 8.27473998e-01
-3.02286237e-01 3.61422837e-01 -3.08541536e-01 -1.19282019e+00
1.11410332e+00 2.61683390e-02 5.86387143e-02 -6.81035578e-01
6.28695548e-01 3.92663121e-01 3.28008570e-02 -5.13189197e-01
5.18893719e-01 -1.61101475e-01 -7.14475885e-02 1.69949368e-01
-2.02128395e-01 1.78322062e-01 -3.45313400e-01 2.23342180e-01
9.97415185e-01 -4.40274507e-01 8.96839499e-01 -4.21019912e-01
6.30925179e-01 4.80138689e-01 3.46871287e-01 5.85602105e-01
-4.49124575e-01 -2.66182095e-01 2.03541085e-01 -7.15977132e-01
-8.55197966e-01 -4.25178975e-01 -2.91017741e-01 1.07426572e+00
3.62603635e-01 -5.36626935e-01 -6.70097828e-01 -7.22357690e-01
1.72647193e-01 1.12533557e+00 -6.25616759e-02 -2.71865845e-01
-2.07897112e-01 -1.25297368e+00 3.13564062e-01 2.37514630e-01
6.88111246e-01 -7.58701622e-01 -8.53483498e-01 -8.46449509e-02
9.22014415e-02 -7.33652651e-01 4.36958760e-01 6.42095447e-01
-1.37343657e+00 -6.09431922e-01 2.96351492e-01 -4.16224688e-01
4.29866582e-01 1.00221366e-01 9.76830661e-01 3.09994787e-01
-4.04831380e-01 -2.83351213e-01 -8.23063910e-01 -7.53257632e-01
-3.22377235e-01 -6.17867671e-02 3.83043051e-01 -4.59662050e-01
9.30354893e-01 -5.08423805e-01 -3.86773467e-01 1.78861901e-01
-5.77026606e-01 -3.15715699e-03 7.52539217e-01 1.26586318e+00
5.40592730e-01 6.81286275e-01 1.74770355e-01 -1.38358653e+00
5.27755857e-01 -6.05363786e-01 -3.12255144e-01 -9.80296265e-03
-1.51147056e+00 4.02162448e-02 7.23801434e-01 -4.26842898e-01
-9.19846892e-01 1.77210882e-01 -4.54376414e-02 7.63775855e-02
-9.65847075e-03 3.61116588e-01 4.15678211e-02 -2.53066570e-01
8.54865670e-01 2.46358350e-01 -7.47812539e-02 -4.85358596e-01
-4.56664115e-02 1.50419927e+00 1.16517395e-01 8.37329999e-02
3.14776272e-01 4.17421103e-01 8.21655914e-02 -6.21214807e-01
-7.50248015e-01 -5.58174968e-01 -3.76883447e-01 -3.41599770e-02
5.77535629e-01 -5.05199075e-01 -6.59107625e-01 4.67176408e-01
-1.12225187e+00 7.11680114e-01 2.46432170e-01 5.42383432e-01
-9.99391004e-02 2.94277370e-01 -8.03215802e-01 -1.37959194e+00
-1.12888825e+00 -6.57386422e-01 4.37854409e-01 3.84347618e-01
-6.29677832e-01 -6.32181764e-01 -5.75093567e-01 8.09955060e-01
5.68579614e-01 8.24394524e-02 1.28995335e+00 -1.41956210e+00
-4.38545011e-02 -6.26609743e-01 -2.40352124e-01 3.16524595e-01
-4.16541934e-01 7.28552118e-02 -6.75809860e-01 -7.87340701e-02
3.51035863e-01 -2.43317023e-01 7.80660152e-01 1.59874335e-01
1.44710302e+00 -8.68510365e-01 -7.94730544e-01 4.52486634e-01
1.64796758e+00 8.47598791e-01 8.15469444e-01 6.26223087e-01
8.55896622e-02 3.10237586e-01 1.14401424e+00 6.15270495e-01
-2.00576216e-01 3.34845513e-01 5.42957902e-01 4.10106212e-01
5.73707283e-01 -3.28187704e-01 6.85197785e-02 1.10068607e+00
1.29340235e-02 -3.16378325e-01 -8.83997798e-01 3.58925201e-02
-1.87137079e+00 -8.38793516e-01 -3.65942508e-01 1.84422874e+00
6.57111645e-01 6.19741976e-01 -1.30172297e-01 9.96887207e-01
4.68256801e-01 -2.12586045e-01 -3.10790598e-01 -1.15229738e+00
4.25295644e-02 3.01556796e-01 1.10091662e+00 1.90826640e-01
-1.12331474e+00 1.01116538e+00 6.39492798e+00 1.41130519e+00
-1.05455530e+00 3.03797543e-01 1.21772456e+00 -8.28570500e-03
-5.72404861e-02 1.72756374e-01 -1.10802937e+00 7.42163301e-01
1.50599015e+00 -5.10788225e-02 3.21750998e-01 1.58159173e+00
-1.68151721e-01 -5.45851767e-01 -5.22724748e-01 1.05901361e+00
7.57417306e-02 -1.32111561e+00 8.33689496e-02 -2.36573294e-01
2.23109767e-01 -9.10786092e-02 -2.65906960e-01 3.54388416e-01
2.28921816e-01 -1.34782231e+00 2.07540452e-01 3.97147954e-01
3.75144213e-01 -9.78764653e-01 1.38021147e+00 6.91179395e-01
-2.85231546e-02 -3.99423361e-01 -6.58707559e-01 -6.35368586e-01
-2.18052551e-01 9.46783841e-01 -8.29574645e-01 1.52450487e-01
1.01885831e+00 3.43580633e-01 -4.60854173e-01 5.83457053e-01
4.65759277e-01 1.09131384e+00 -5.95219851e-01 -9.48713779e-01
1.60389990e-01 2.29064445e-03 2.38844737e-01 1.07897937e+00
2.01869875e-01 4.81083333e-01 -1.22523934e-01 1.48448616e-01
5.71862102e-01 4.68327254e-01 -5.47927856e-01 1.16134711e-01
6.04324102e-01 1.08101654e+00 -1.04507744e+00 -3.70630860e-01
-1.55262068e-01 7.58712351e-01 -2.81182915e-01 -4.64819938e-01
-9.01570082e-01 -5.04926682e-01 -5.38381822e-02 3.50013196e-01
-8.55952129e-02 1.90227643e-01 -1.04982173e+00 -7.90548444e-01
-3.40996802e-01 -8.02337945e-01 6.01018488e-01 -7.66085386e-01
-1.08928537e+00 7.37153292e-01 -4.00964953e-02 -9.74211752e-01
-1.65580094e-01 -9.12207305e-01 -2.52792090e-01 4.16751623e-01
-8.96122277e-01 -7.69063532e-01 -7.56366691e-03 5.47284067e-01
3.08428615e-01 -8.05103719e-01 9.09360111e-01 2.69866794e-01
-4.12954181e-01 5.33935845e-01 4.09739673e-01 -4.38484251e-01
1.40207946e-01 -7.19074249e-01 -3.11276913e-01 1.65685147e-01
4.98199314e-02 3.87854934e-01 1.00267017e+00 -7.77597010e-01
-1.16298091e+00 -6.42770290e-01 1.11438000e+00 -6.23696186e-02
5.01043439e-01 -5.00585958e-02 -4.41368967e-01 1.69929266e-01
-6.43611550e-02 -2.07476571e-01 8.50770116e-01 -1.78431887e-02
-1.09687306e-01 -2.52505988e-01 -1.83375192e+00 1.82636768e-01
4.41210091e-01 2.27700382e-01 -3.48240644e-01 6.55148029e-01
6.76524103e-01 -1.26950443e-02 -7.05920875e-01 3.53786081e-01
5.19029438e-01 -1.09162378e+00 9.51470971e-01 -1.02815640e+00
7.71106064e-01 2.16787815e-01 -5.29030740e-01 -6.29540741e-01
-1.88125148e-01 -3.34978439e-02 -4.40926909e-01 1.17324066e+00
6.94177210e-01 -7.43390441e-01 1.12569976e+00 8.33412290e-01
7.42032647e-01 -1.23434508e+00 -1.15780580e+00 -5.98986387e-01
-1.19188040e-01 -3.37670386e-01 4.61833447e-01 1.00511849e+00
4.06548798e-01 6.77488923e-01 -7.60740519e-01 -5.07862270e-01
3.23674858e-01 9.29011106e-02 2.94165701e-01 -1.25527775e+00
-3.40083927e-01 -9.68574658e-02 -9.69746470e-01 -3.75247359e-01
-5.95793352e-02 -8.03914368e-01 -6.24546051e-01 -9.68580186e-01
4.78734016e-01 -5.47110260e-01 -2.30052382e-01 5.41545093e-01
1.88742876e-01 9.96929500e-03 2.05377322e-02 5.34879744e-01
-4.85296965e-01 -1.37471883e-02 6.69773400e-01 -1.91377848e-01
-2.70854239e-03 3.94586384e-01 -7.29460955e-01 1.21513617e+00
1.02045107e+00 -8.62619102e-01 -4.02646124e-01 1.75740883e-01
1.13371693e-01 3.31840754e-01 -1.19289219e-01 -9.57653940e-01
2.07992509e-01 -3.96247923e-01 5.89050055e-01 -7.33463883e-01
1.41758397e-01 -1.22297502e+00 1.56845525e-01 1.23763299e+00
-2.68726021e-01 1.89297497e-01 7.35338181e-02 7.01570451e-01
-1.83441952e-01 -8.68605852e-01 1.06910825e+00 -7.11031631e-02
-6.56222820e-01 -2.66080439e-01 -4.63110745e-01 -6.50781214e-01
1.44659686e+00 -2.88416803e-01 -4.73236620e-01 -2.71206856e-01
-5.09581804e-01 -2.48061642e-01 3.26242819e-02 1.55193731e-01
6.97175801e-01 -1.00070322e+00 -3.33871305e-01 4.49525900e-02
-1.85050458e-01 -5.98000824e-01 -1.25835985e-01 7.14806855e-01
-1.06834900e+00 7.80624986e-01 -3.91802579e-01 -3.16626936e-01
-1.90513408e+00 7.04390466e-01 -1.37054711e-03 -7.63908088e-01
-2.01216996e-01 9.72324669e-01 -9.15890932e-01 -3.96322370e-01
2.85503298e-01 1.22755580e-01 -5.52659273e-01 -4.38086450e-01
2.62485504e-01 8.03149819e-01 1.59843385e-01 -6.44904673e-01
-7.27215409e-01 -5.82396658e-03 -5.43998003e-01 2.10159376e-01
1.14534473e+00 1.68669030e-01 -5.55514693e-01 9.30975452e-02
9.44517076e-01 -2.50949055e-01 1.95217967e-01 3.28350812e-02
4.35008764e-01 -6.34880900e-01 4.16200250e-01 -1.38743651e+00
-7.52952754e-01 5.43096960e-01 9.27523017e-01 3.01424652e-01
1.21156609e+00 -4.34290290e-01 1.03633130e+00 4.84849930e-01
7.58586466e-01 -1.44840407e+00 -5.18472135e-01 2.93583989e-01
2.00717062e-01 -1.27600729e+00 6.00515485e-01 -5.95636427e-01
-8.21623027e-01 9.72479105e-01 7.36319661e-01 -1.02380566e-01
1.34555614e+00 7.75777400e-01 -4.57754463e-01 9.95229334e-02
-1.22490966e+00 5.40800691e-01 -8.58453810e-02 5.42171180e-01
3.71129364e-01 3.96542370e-01 -1.20969081e+00 1.18407273e+00
-5.78199446e-01 2.26022869e-01 3.24018180e-01 9.03343260e-01
-7.51265585e-01 -7.80862272e-01 -4.60398585e-01 1.36426413e+00
-9.33211744e-01 -2.61186898e-01 -6.67502463e-01 6.51227236e-01
2.54961312e-01 1.08966827e+00 -1.86235040e-01 -1.20486796e+00
-1.49593592e-01 8.04267302e-02 1.06183305e-01 6.30870834e-02
-6.09473050e-01 -4.27438855e-01 3.19837719e-01 -3.10139149e-01
-2.91844219e-01 -4.12826926e-01 -1.20014536e+00 -9.19544220e-01
-7.34593511e-01 4.35732156e-01 1.10397351e+00 1.03335440e+00
3.40013593e-01 6.05802890e-03 3.26127887e-01 -2.53586292e-01
-9.03079748e-01 -1.08351028e+00 -6.08383119e-01 3.40395510e-01
-3.87517333e-01 -6.10051274e-01 -6.90012455e-01 -2.08068654e-01] | [8.372835159301758, 4.86386251449585] |
2cbeadda-44f1-4c6e-be7b-30143aff4774 | visual-lidar-odometry-and-mapping-with | 2304.08978 | null | https://arxiv.org/abs/2304.08978v2 | https://arxiv.org/pdf/2304.08978v2.pdf | Visual-LiDAR Odometry and Mapping with Monocular Scale Correction and Visual Bootstrapping | This paper presents a novel visual-LiDAR odometry and mapping method with low-drift characteristics. The proposed method is based on two popular approaches, ORB-SLAM and A-LOAM, with monocular scale correction and visual-bootstrapped LiDAR poses initialization modifications. The scale corrector calculates the proportion between the depth of image keypoints recovered by triangulation and that provided by LiDAR, using an outlier rejection process for accuracy improvement. Concerning LiDAR poses initialization, the visual odometry approach gives the initial guesses of LiDAR motions for better performance. This methodology is not only applicable to high-resolution LiDAR but can also adapt to low-resolution LiDAR. To evaluate the proposed SLAM system's robustness and accuracy, we conducted experiments on the KITTI Odometry and S3E datasets. Experimental results illustrate that our method significantly outperforms standalone ORB-SLAM2 and A-LOAM. Furthermore, regarding the accuracy of visual odometry with scale correction, our method performs similarly to the stereo-mode ORB-SLAM2. | ['Junzheng Wang', 'Ni Ou', 'Hanyu Cai'] | 2023-04-18 | null | null | null | null | ['motion-compensation', 'visual-odometry'] | ['computer-vision', 'robots'] | [-1.72807127e-01 -3.18643123e-01 -2.89950252e-01 -4.83280301e-01
-5.93917906e-01 -5.37521422e-01 6.87276959e-01 1.64881602e-01
-6.15854383e-01 1.04857779e+00 -3.49932313e-01 -1.18375339e-01
-1.85549229e-01 -7.66244769e-01 -6.31257296e-01 -3.67802799e-01
8.93882960e-02 1.16394353e+00 5.42619407e-01 -2.22261727e-01
5.62283456e-01 1.10422158e+00 -1.69962633e+00 -6.54319048e-01
1.03077364e+00 6.73424363e-01 2.61118293e-01 5.34670174e-01
-1.65453747e-01 3.20463851e-02 -1.92173630e-01 -6.69626892e-02
5.76722503e-01 1.08913288e-01 -2.02616587e-01 -3.53650488e-02
1.04017878e+00 -4.36294615e-01 4.73034382e-02 1.05830824e+00
5.14435709e-01 -1.22428216e-01 5.75760067e-01 -1.50157309e+00
-1.79510117e-01 -2.24000677e-01 -6.96819425e-01 -2.91995883e-01
5.37280858e-01 1.95526257e-01 4.17290717e-01 -1.17332530e+00
8.87714744e-01 1.37033510e+00 1.05113268e+00 -1.90730497e-01
-1.61717212e+00 -8.00834894e-01 -4.86632347e-01 1.27781555e-01
-1.95485544e+00 -2.92201042e-01 3.57834965e-01 -6.90815330e-01
8.71067286e-01 -1.90966591e-01 5.30361593e-01 4.67425644e-01
4.86590236e-01 -2.74848640e-01 1.17712057e+00 -3.59603882e-01
2.28866130e-01 3.02572370e-01 -9.25928876e-02 7.12426722e-01
1.16956127e+00 4.23486024e-01 -7.22602129e-01 -2.13711917e-01
6.29609644e-01 5.54267690e-02 -1.50898427e-01 -1.23932481e+00
-1.15056884e+00 7.86756635e-01 6.33503377e-01 -2.29873553e-01
-2.62911975e-01 1.54047579e-01 1.01429738e-01 2.31100798e-01
1.12728663e-01 1.22909434e-01 -6.63802773e-02 -5.96077070e-02
-1.18530858e+00 2.31547087e-01 4.32590425e-01 1.33687806e+00
1.51667762e+00 2.47198403e-01 6.69090986e-01 3.44924510e-01
6.27218127e-01 1.18607533e+00 4.16337758e-01 -8.71794105e-01
3.85062337e-01 7.23602831e-01 3.89167428e-01 -8.26489806e-01
-4.38737512e-01 -2.56351065e-02 -5.24893045e-01 6.52996063e-01
5.13459183e-02 2.83690244e-01 -1.00889647e+00 9.41975594e-01
4.90724385e-01 1.34710073e-02 1.52585417e-01 7.92202353e-01
4.55937445e-01 1.29424199e-01 -4.16973472e-01 -1.32827982e-01
8.27947021e-01 -2.79411644e-01 -6.84301972e-01 -4.25292790e-01
4.91876096e-01 -1.06193066e+00 6.76001847e-01 2.95754552e-01
-5.78562021e-01 -8.27668846e-01 -1.40495801e+00 -1.19059727e-01
-2.59476006e-01 9.18538496e-02 1.82268679e-01 5.74065506e-01
-1.16116214e+00 6.23768151e-01 -7.10276067e-01 -9.15133417e-01
-2.75963992e-01 4.85904247e-01 -4.89122659e-01 2.59594969e-03
-7.88247406e-01 1.42229664e+00 5.85360885e-01 -7.82300532e-02
-5.92547357e-01 -3.57493430e-01 -9.98018384e-01 -5.60215056e-01
-3.62114124e-02 -5.22324443e-01 8.96715581e-01 -2.34011501e-01
-1.38455760e+00 9.01565313e-01 -5.71205556e-01 -6.26571119e-01
7.41257012e-01 -2.96946526e-01 -1.89167321e-01 -8.45324993e-02
5.73570132e-01 7.89539337e-01 6.05475128e-01 -1.57180047e+00
-6.64099157e-01 -6.34415090e-01 -7.32201219e-01 1.74620330e-01
5.88111460e-01 -7.38882959e-01 -1.12238176e-01 1.54469550e-01
1.07066810e+00 -1.13006330e+00 -5.41810058e-02 1.99857652e-01
-9.48432535e-02 4.53046322e-01 8.24465036e-01 -1.74174756e-01
5.89956403e-01 -2.09282064e+00 -3.67346704e-01 5.01796842e-01
-4.08376157e-01 -3.95871773e-02 5.76689206e-02 6.74350619e-01
2.03789860e-01 -1.90971851e-01 -6.82123378e-02 -3.48564953e-01
-1.82919905e-01 6.07304454e-01 -4.79375184e-01 9.46062446e-01
-6.54403418e-02 5.74895978e-01 -7.84332573e-01 -5.59670866e-01
7.12778211e-01 4.69926775e-01 -1.82734519e-01 -1.63480252e-01
2.21412629e-01 5.49926221e-01 8.67251083e-02 8.84218574e-01
1.27823877e+00 5.14039993e-01 1.49759665e-01 -1.00004345e-01
-7.10357666e-01 3.04695278e-01 -1.76652730e+00 1.71220839e+00
-4.73864496e-01 5.61131120e-01 9.48969424e-02 -1.31498098e-01
1.66475797e+00 -1.15006581e-01 2.71223813e-01 -6.64896309e-01
-1.28924832e-01 9.14581180e-01 -2.63220757e-01 -1.66188017e-01
1.12433827e+00 -2.50241131e-01 2.08385557e-01 -1.53490946e-01
-7.74083808e-02 -9.58111525e-01 6.96734041e-02 -1.87302555e-03
4.16007102e-01 6.74332678e-01 8.61306131e-01 -3.48923743e-01
7.39530981e-01 3.97502005e-01 5.73370099e-01 4.54957753e-01
-1.65360540e-01 4.43455160e-01 -1.32163689e-01 -3.36925656e-01
-1.18526185e+00 -1.52260125e+00 -6.33316934e-01 9.97472256e-02
7.14097261e-01 -1.09722480e-01 4.63965982e-02 -1.12995915e-01
8.44766498e-01 5.34351885e-01 -1.84843987e-01 1.30400434e-01
-4.32910442e-01 -2.41795048e-01 3.21938515e-01 2.19942182e-01
4.97838527e-01 -4.25735921e-01 -6.44859612e-01 2.37397417e-01
7.39773437e-02 -9.62146699e-01 1.43774956e-01 1.50213182e-01
-1.39599168e+00 -1.02626097e+00 -1.29205540e-01 -5.68743646e-01
4.29130197e-01 5.73051453e-01 7.00638533e-01 -1.72551736e-01
-1.35182530e-01 9.42628235e-02 -1.68433666e-01 -3.95409137e-01
-4.07591373e-01 -4.30371836e-02 5.57370484e-01 -2.70854294e-01
5.25702357e-01 -8.20049882e-01 -1.99795648e-01 6.65622175e-01
-2.83404291e-01 -4.34689224e-01 5.36970794e-01 4.45355684e-01
8.90843451e-01 -6.37129664e-01 9.47484151e-02 -3.53777379e-01
3.55265327e-02 -9.88865793e-02 -1.30737138e+00 -3.63615513e-01
-1.19888139e+00 2.38450274e-01 -2.98343468e-02 -1.08300649e-01
-6.24845862e-01 5.91156363e-01 1.36653215e-01 -4.10972297e-01
-9.35574025e-02 1.13794655e-01 8.26399215e-03 -6.44237816e-01
9.84748483e-01 1.57437339e-01 2.81250685e-01 -5.23694873e-01
3.18694562e-01 8.94746661e-01 8.38500679e-01 -3.43183875e-01
1.42779541e+00 1.01575005e+00 6.85436130e-01 -1.13565981e+00
-1.45034522e-01 -9.06377494e-01 -1.30236423e+00 -1.07106738e-01
5.52717745e-01 -1.26186836e+00 -5.24049044e-01 1.30415842e-01
-1.15566528e+00 3.02493542e-01 -3.40498179e-01 8.81902277e-01
-6.36739492e-01 6.78399205e-01 3.57284546e-02 -9.95741606e-01
-2.69776992e-02 -1.22573626e+00 1.26551104e+00 5.77130690e-02
-2.20375791e-01 -5.94011843e-01 6.24967873e-01 -7.27983043e-02
-2.23550666e-02 4.96818900e-01 2.00240105e-01 -1.33814067e-01
-1.07624876e+00 -3.08887839e-01 -2.42292300e-01 -7.78436139e-02
3.39763500e-02 1.23614207e-01 -8.73256087e-01 -6.09025240e-01
-2.63528675e-01 -3.93133163e-02 5.81878901e-01 6.84646517e-02
-2.31621087e-01 3.72801900e-01 -4.68126357e-01 8.95450950e-01
2.02716613e+00 2.29566559e-01 6.75024331e-01 1.05874300e+00
5.23633659e-01 2.90771306e-01 1.33375907e+00 3.27648371e-01
4.73848820e-01 1.01049948e+00 9.60097849e-01 1.75480187e-01
1.69272527e-01 -5.78192711e-01 2.82171041e-01 5.36719084e-01
-4.19514738e-02 4.90208805e-01 -1.10582221e+00 6.52675629e-01
-1.85792565e+00 -6.53120100e-01 -8.56636047e-01 2.82044506e+00
1.80717289e-01 2.13825151e-01 -1.69111788e-01 1.37329787e-01
6.40876114e-01 1.24417149e-01 -3.63996953e-01 -7.12745607e-01
1.24364011e-01 1.49793640e-01 1.25660193e+00 1.01157653e+00
-7.06405282e-01 9.78665352e-01 5.47414875e+00 -2.42469124e-02
-1.12311196e+00 2.53346980e-01 -9.70527351e-01 2.24099860e-01
-2.27835458e-02 5.17868280e-01 -1.29875362e+00 2.44193718e-01
9.64501381e-01 -1.40074894e-01 1.87855750e-01 1.10747290e+00
3.09078068e-01 -5.97014070e-01 -7.77031004e-01 1.06334627e+00
6.31470755e-02 -1.04650450e+00 3.63081880e-02 4.44331974e-01
6.71062887e-01 4.56835091e-01 -4.07202125e-01 1.81172162e-01
2.30443984e-01 -5.03640711e-01 8.47029030e-01 3.22957516e-01
9.99015391e-01 -7.23365188e-01 8.82283747e-01 4.90886211e-01
-1.60234571e+00 2.18808681e-01 -7.39975631e-01 -3.37077022e-01
3.34924221e-01 5.00795543e-01 -1.48409843e+00 1.02420127e+00
5.33767700e-01 6.67390227e-01 -7.16829717e-01 1.23906577e+00
-4.91710156e-01 -2.33592331e-01 -5.21024644e-01 2.88479716e-01
7.45679289e-02 -5.26854455e-01 6.77432418e-01 9.33013022e-01
7.57470250e-01 -6.66611016e-01 2.82655805e-01 5.31751752e-01
3.81192952e-01 1.53217211e-01 -1.19170249e+00 6.15594745e-01
8.28539729e-01 8.98248553e-01 -2.44174898e-01 -2.62378216e-01
-3.68493907e-02 7.08614647e-01 -1.45917103e-01 7.28773624e-02
-6.27591789e-01 -3.33971620e-01 9.47396815e-01 6.18660867e-01
2.18625322e-01 -7.56024003e-01 -4.83535737e-01 -8.86382937e-01
-4.15666737e-02 -2.92540699e-01 -7.30533749e-02 -1.14769280e+00
-5.75118124e-01 3.16261619e-01 5.88298589e-02 -1.94221568e+00
-5.64619005e-01 -5.99685431e-01 -2.23522723e-01 1.10708499e+00
-1.70284450e+00 -1.15889442e+00 -7.61587262e-01 1.95260197e-01
2.09257856e-01 1.01629585e-01 5.13038516e-01 1.67175516e-01
3.38451177e-01 -7.80124217e-02 3.80976468e-01 -3.88643235e-01
1.22772956e+00 -1.10056472e+00 5.50085306e-01 9.80980277e-01
7.96277151e-02 5.76003373e-01 9.22249794e-01 -1.19399178e+00
-1.23468423e+00 -1.10524225e+00 1.10466635e+00 -5.24783790e-01
6.22022092e-01 -2.37024605e-01 -8.40866446e-01 9.79466498e-01
-2.96284974e-01 -5.19633926e-02 -2.06883609e-01 -1.06649548e-01
-1.53036401e-01 -5.03235281e-01 -1.40568626e+00 2.32539654e-01
8.39449227e-01 -6.20732248e-01 -8.08205009e-01 9.32847336e-02
3.35187554e-01 -7.31324852e-01 -7.55141914e-01 6.09351337e-01
6.56826437e-01 -1.34061921e+00 9.62044477e-01 3.16463768e-01
-4.85305578e-01 -9.72676218e-01 -3.78759384e-01 -1.07420206e+00
-1.95761956e-02 -3.90963435e-01 4.29893285e-01 1.19932353e+00
-1.14923760e-01 -1.07541800e+00 7.62923002e-01 -2.86491781e-01
5.99327832e-02 1.00755863e-01 -1.39253616e+00 -1.25363743e+00
-3.93786728e-01 -2.74314731e-01 5.06678164e-01 7.28415132e-01
-4.92908925e-01 1.85848713e-01 -2.18237668e-01 7.40667701e-01
1.05531859e+00 1.66029826e-01 1.60494232e+00 -1.80017173e+00
3.03712815e-01 2.56206065e-01 -1.12847579e+00 -7.63412058e-01
-1.27187278e-02 -8.75291705e-01 1.58466607e-01 -1.44518542e+00
-3.25056821e-01 -1.17181405e-01 3.85818273e-01 1.28373224e-02
4.46172088e-01 4.81988907e-01 1.57304361e-01 6.15382433e-01
1.73858747e-01 3.71654481e-01 7.00862050e-01 2.19107404e-01
-4.35923994e-01 -7.10565373e-02 3.97994339e-01 8.77753496e-01
4.18924391e-01 -8.81548345e-01 -1.78710863e-01 -1.82116359e-01
1.30811289e-01 -9.04577598e-03 5.58846176e-01 -1.41288257e+00
1.72974423e-01 -1.92948923e-01 3.01253617e-01 -1.30327749e+00
5.85554719e-01 -9.95871246e-01 5.75863242e-01 9.41022217e-01
6.09409153e-01 3.89002740e-01 1.15159623e-01 4.95077759e-01
-2.29680941e-01 -3.57566088e-01 1.10874152e+00 5.87671362e-02
-9.99780536e-01 4.49987240e-02 -1.04147546e-01 -4.09023643e-01
9.56516862e-01 -8.23129356e-01 -3.78934205e-01 -3.33152741e-01
-3.79978031e-01 1.42067656e-01 1.34484017e+00 2.39724398e-01
6.74683154e-01 -1.45136642e+00 -3.95137906e-01 5.44145525e-01
5.35861313e-01 2.07558960e-01 -3.19281131e-01 1.08700895e+00
-9.94739115e-01 5.66524565e-01 -4.19897288e-01 -1.31384289e+00
-1.21617448e+00 4.25600499e-01 3.34400028e-01 2.09207937e-01
-3.47129822e-01 1.27426431e-01 -4.26947743e-01 -8.52141917e-01
-2.11607944e-02 -5.43513834e-01 2.47409090e-01 1.15835983e-02
1.16933994e-02 6.71034873e-01 1.74204454e-01 -1.09029293e+00
-8.24914038e-01 1.33882225e+00 5.10843277e-01 -4.88160789e-01
7.34321594e-01 -4.83757824e-01 -4.63528372e-02 8.24405670e-01
8.93103063e-01 4.08721119e-01 -8.97325933e-01 4.87537608e-02
4.22694206e-01 -8.20754290e-01 -1.08478472e-01 -2.02160925e-01
-1.67417228e-01 9.48562980e-01 1.04830563e+00 -3.29574406e-01
4.86885488e-01 -4.67867851e-01 1.77566558e-01 5.24171948e-01
9.66628730e-01 -8.13118637e-01 -3.51157933e-01 7.10839808e-01
7.94490397e-01 -1.32479930e+00 6.01303220e-01 -2.64086306e-01
-3.91330868e-01 1.20621121e+00 4.94443536e-01 -2.71701247e-01
1.99253380e-01 1.82351455e-01 3.87813032e-01 1.24699786e-01
-7.96478614e-02 -5.55632949e-01 -1.02172963e-01 7.89090037e-01
3.40306014e-02 -1.70186102e-01 -4.24962103e-01 -7.16895998e-01
-2.15365306e-01 2.10063055e-01 7.04378545e-01 9.99160111e-01
-1.03522980e+00 -1.12235594e+00 -9.92131114e-01 -1.34335175e-01
5.44647217e-01 1.60980687e-01 -3.97251487e-01 1.50605536e+00
2.75214255e-01 6.12447262e-01 2.74825603e-01 -3.85892719e-01
6.52132750e-01 2.88645923e-01 3.88801157e-01 -5.08782148e-01
-1.94963217e-01 -5.36579080e-02 5.54384328e-02 -7.77964532e-01
-2.98547238e-01 -7.64610350e-01 -1.50560021e+00 -3.81123990e-01
-3.75873208e-01 1.20065525e-01 1.40245903e+00 4.92121816e-01
3.26140404e-01 -2.54070073e-01 5.42003572e-01 -1.16046095e+00
-7.61106193e-01 -8.70729685e-01 -7.30329096e-01 8.63154083e-02
5.52423596e-01 -1.01184297e+00 -5.46891093e-01 -2.98226178e-01] | [7.353085041046143, -2.1595497131347656] |
03be7ba4-98df-4055-8324-c5c40b5d17d7 | development-of-a-realistic-crowd-simulation | 2304.13403 | null | https://arxiv.org/abs/2304.13403v1 | https://arxiv.org/pdf/2304.13403v1.pdf | Development of a Realistic Crowd Simulation Environment for Fine-grained Validation of People Tracking Methods | Generally, crowd datasets can be collected or generated from real or synthetic sources. Real data is generated by using infrastructure-based sensors (such as static cameras or other sensors). The use of simulation tools can significantly reduce the time required to generate scenario-specific crowd datasets, facilitate data-driven research, and next build functional machine learning models. The main goal of this work was to develop an extension of crowd simulation (named CrowdSim2) and prove its usability in the application of people-tracking algorithms. The simulator is developed using the very popular Unity 3D engine with particular emphasis on the aspects of realism in the environment, weather conditions, traffic, and the movement and models of individual agents. Finally, three methods of tracking were used to validate generated dataset: IOU-Tracker, Deep-Sort, and Deep-TAMA. | ['Michał Staniszewski', 'Elżbieta Macioszek', 'Dominik Golba', 'Michał Cogiel', 'Bartosz Bizoń', 'Adam Cygan', 'Nicola Messina', 'Luca Ciampi', 'Agnieszka Szczęsna', 'Paweł Foszner'] | 2023-04-26 | null | null | null | null | ['multiple-people-tracking', 'unity'] | ['computer-vision', 'computer-vision'] | [-5.35407484e-01 -3.45997125e-01 5.25728047e-01 9.19713452e-02
8.15126970e-02 -5.28646350e-01 9.82817173e-01 6.47321194e-02
-7.56147146e-01 1.13473856e+00 2.16835588e-01 -1.11169733e-01
3.24466765e-01 -9.23422635e-01 -5.38354576e-01 -4.69279915e-01
-1.25989765e-01 8.36036921e-01 7.55764663e-01 -4.82086271e-01
-1.25728935e-01 8.13689828e-01 -2.04236436e+00 -7.24644884e-02
4.86811668e-01 4.32706296e-01 1.81988582e-01 1.06656909e+00
-1.57590851e-01 9.09531891e-01 -1.24015903e+00 -1.18165009e-01
4.10961092e-01 -1.43655434e-01 -6.25624284e-02 7.34238178e-02
2.60942370e-01 -2.07290396e-01 -2.74165362e-01 5.14226079e-01
9.43590879e-01 2.47181505e-01 2.30619267e-01 -1.75235438e+00
-9.77728814e-02 1.19569607e-01 -1.19536392e-01 4.58079696e-01
9.76916611e-01 9.60878074e-01 -3.24750870e-01 -4.33220237e-01
7.44080603e-01 1.47404778e+00 8.65670204e-01 6.85997486e-01
-8.08177888e-01 -6.90000832e-01 -7.60044158e-02 -4.67221513e-02
-1.29747331e+00 -2.63852119e-01 4.39262241e-01 -9.63089287e-01
7.73498714e-01 1.39450684e-01 1.20683348e+00 1.54785049e+00
5.92583045e-02 6.22178316e-01 1.32879806e+00 -4.81754810e-01
7.82303512e-01 2.56783515e-01 -6.56184778e-02 5.81576943e-01
5.82832634e-01 4.61790413e-01 -5.56688905e-01 -2.98811704e-01
6.38822913e-01 -2.57120669e-01 1.31954381e-03 -2.53416002e-01
-1.24641800e+00 6.76035225e-01 2.91616440e-01 2.38913566e-01
-4.44775969e-01 1.80592030e-01 5.37615359e-01 -5.15559502e-02
2.68551946e-01 6.71965182e-02 -1.37826517e-01 -4.03202802e-01
-8.01755488e-01 9.79538441e-01 7.61493921e-01 1.06792903e+00
6.29872203e-01 3.70974123e-01 -3.99301171e-01 -1.11536197e-01
4.96436805e-01 9.81935501e-01 5.49369931e-01 -1.11072171e+00
3.14707279e-01 7.55599082e-01 7.13722169e-01 -1.02621412e+00
-7.98576951e-01 3.05705130e-01 -3.55038583e-01 5.09344399e-01
6.09494269e-01 -8.62852216e-01 -7.03209937e-01 1.29310894e+00
7.36353934e-01 5.42622149e-01 -1.53704554e-01 9.11205053e-01
1.20471382e+00 4.23478156e-01 3.39321494e-01 9.68825892e-02
1.01451898e+00 -5.88531554e-01 -6.37630999e-01 7.20553324e-02
7.09682405e-01 -2.71948129e-01 8.03660989e-01 -1.84069678e-01
-1.02776778e+00 -8.56614172e-01 -5.09108424e-01 5.98628640e-01
-8.96845877e-01 -2.12093145e-01 5.47010422e-01 1.13086665e+00
-1.16067338e+00 -6.45482019e-02 -7.80468822e-01 -5.56530118e-01
3.31345677e-01 3.84686708e-01 -1.78862035e-01 2.01045364e-01
-1.24113166e+00 1.02270710e+00 1.55737996e-01 -3.75292867e-01
-1.18563819e+00 -8.44170809e-01 -7.85640001e-01 -5.62683403e-01
-8.28672573e-02 -8.54939640e-01 1.16391313e+00 -6.61586225e-01
-1.51468897e+00 9.74734604e-01 -3.46840285e-02 -6.11888647e-01
1.06678498e+00 -7.02457875e-03 -5.19116163e-01 -1.15193464e-01
4.16800410e-01 5.90505183e-01 2.00616285e-01 -1.40522563e+00
-7.67838359e-01 -3.15167159e-01 6.02940097e-02 -1.08728781e-01
3.62418979e-01 4.46052134e-01 -1.52355909e-01 -2.58646578e-01
-1.04315186e+00 -9.64437723e-01 -5.04719377e-01 -6.28906041e-02
-2.37322059e-02 3.82211283e-02 1.03648007e+00 -4.10433352e-01
6.98631585e-01 -1.66085291e+00 -5.68318009e-01 1.35909900e-01
1.73027068e-01 9.29221511e-01 1.28331989e-01 6.18320465e-01
4.89012867e-01 -2.39298716e-01 1.71030238e-01 -3.27639073e-01
4.38904501e-02 1.66624829e-01 1.32132888e-01 4.53871638e-01
-2.98832536e-01 8.72687161e-01 -1.08339989e+00 -6.26611292e-01
8.53553355e-01 5.98841727e-01 -1.62255183e-01 1.39321938e-01
-3.06812316e-01 8.99545550e-01 -3.26069206e-01 4.22482103e-01
6.69987142e-01 3.18773538e-01 -2.61842817e-01 3.93122613e-01
-6.62763774e-01 -2.81218350e-01 -1.47860658e+00 1.16441119e+00
-4.13205743e-01 9.40054178e-01 8.11822042e-02 -2.54615694e-01
7.84967005e-01 3.50839198e-01 4.99370992e-01 -8.15407097e-01
4.39202040e-01 -1.73511803e-01 -3.72700781e-01 -8.95085633e-01
5.17652750e-01 3.31183642e-01 -4.07724306e-02 2.74347842e-01
-2.50548363e-01 1.31352246e-01 5.20563900e-01 1.02776205e-02
9.47318733e-01 1.75972611e-01 1.04598895e-01 -2.22736657e-01
5.64402282e-01 6.55030131e-01 2.92871296e-01 8.73580933e-01
-6.21199727e-01 1.84409305e-01 -5.03457300e-02 -8.43003869e-01
-1.08303368e+00 -8.66146088e-01 3.36245984e-01 1.00711060e+00
2.30776086e-01 1.07063666e-01 -1.08306611e+00 -4.39880013e-01
2.24962249e-01 6.97431266e-01 -6.28638387e-01 3.30265194e-01
-7.91100502e-01 -6.04058146e-01 9.32328463e-01 4.62302029e-01
6.02383316e-01 -1.19553173e+00 -1.29751527e+00 1.65245831e-01
6.07439652e-02 -1.25562060e+00 7.81621784e-02 -6.91779435e-01
-3.11687917e-01 -1.21052456e+00 -6.54980004e-01 -2.95935184e-01
3.90560180e-01 3.99175614e-01 1.19554937e+00 1.76674709e-01
-2.43801996e-01 8.20362628e-01 -3.14493984e-01 -9.99481559e-01
-4.24574733e-01 -1.91686779e-01 3.00226361e-01 -1.38250366e-01
7.21710861e-01 -2.38762438e-01 -3.70447576e-01 3.84029806e-01
-6.65471435e-01 -1.25703081e-01 -2.75149047e-01 -7.70394057e-02
-4.45585418e-03 -1.82675228e-01 3.67934138e-01 -2.31921673e-01
9.38249052e-01 -5.10222375e-01 -9.62950230e-01 -3.39333564e-02
2.68834203e-01 -4.99687970e-01 4.58230019e-01 -5.64771235e-01
-1.03586054e+00 1.43445894e-01 2.27446675e-01 -3.01735729e-01
-7.13850558e-01 -2.08341196e-01 -3.03071797e-01 -2.01365456e-01
1.03584409e+00 4.32794727e-02 9.53150094e-02 -1.13568753e-02
1.24530725e-01 6.45708084e-01 3.23815078e-01 -4.25151616e-01
7.07010508e-01 9.21664834e-01 -7.08551481e-02 -1.18338501e+00
-2.71197557e-01 -2.35361680e-01 -7.19203055e-01 -9.92164195e-01
8.07182908e-01 -1.13826895e+00 -1.13063252e+00 7.90429831e-01
-1.12320924e+00 -8.13296914e-01 -4.13089782e-01 4.96395707e-01
-2.39659324e-01 8.98054764e-02 -1.91794530e-01 -1.07936299e+00
-1.04352776e-02 -1.05079603e+00 1.06095159e+00 7.12266624e-01
-2.32391104e-01 -1.33022702e+00 7.20908344e-01 1.82217121e-01
5.95727205e-01 7.91635931e-01 -2.05078468e-01 -5.34037590e-01
-5.40437162e-01 -2.83543944e-01 5.15290722e-02 -3.09454799e-01
-2.48847187e-01 3.04749042e-01 -9.45419610e-01 -7.22174644e-02
-5.32876015e-01 3.30154598e-02 2.56640971e-01 5.13158202e-01
3.66084516e-01 9.08122137e-02 -6.53121591e-01 1.32309258e-01
1.19312024e+00 2.07083300e-01 4.57870901e-01 6.27715290e-01
7.50687838e-01 4.97818410e-01 4.91095155e-01 7.22579658e-01
8.39546502e-01 8.81167829e-01 1.87552959e-01 3.32109188e-03
-2.80283004e-01 -3.30219060e-01 3.87113303e-01 5.92892617e-02
-5.65652728e-01 -4.10590529e-01 -1.19981062e+00 4.90601927e-01
-1.83876944e+00 -1.38839972e+00 -9.37225521e-01 2.00675297e+00
1.87560111e-01 -2.12718427e-01 8.81980598e-01 -7.59133417e-03
8.77592206e-01 -1.66758239e-01 -5.30807339e-02 -1.44925117e-01
-1.27455831e-01 -8.68497640e-02 5.59768975e-01 7.00013936e-01
-8.85300696e-01 1.01446235e+00 7.33871555e+00 3.99410635e-01
-9.97884452e-01 2.26277635e-01 -1.58018678e-01 -2.63207525e-01
1.77690625e-01 -2.81363070e-01 -1.07527423e+00 1.02324271e+00
1.22188306e+00 -3.04071188e-01 1.84333608e-01 8.63417387e-01
8.72233629e-01 -5.08927047e-01 -5.87465346e-01 6.10056877e-01
4.77246009e-03 -1.52535319e+00 -2.29874939e-01 1.12493299e-01
8.27100575e-01 2.91358083e-01 -4.45835590e-01 5.11483252e-01
9.00390625e-01 -8.00754905e-01 8.92108202e-01 8.87553930e-01
2.80042171e-01 -5.08768380e-01 9.04932678e-01 7.35324562e-01
-1.07193530e+00 5.99061288e-02 -2.98188746e-01 -5.76182425e-01
4.79438871e-01 3.76307309e-01 -9.58693624e-01 1.25063047e-01
7.31535912e-01 -8.81637447e-03 -8.34212780e-01 1.46980047e+00
1.74701601e-01 4.11730349e-01 -6.14080191e-01 -5.02936721e-01
1.93826795e-01 1.82891577e-01 7.24016905e-01 1.65799332e+00
1.19848892e-01 -2.26965457e-01 4.72534001e-01 5.25069892e-01
3.84930253e-01 -4.40626502e-01 -1.00100911e+00 6.96158469e-01
6.09572113e-01 9.35368299e-01 -5.29816151e-01 -3.26797545e-01
-2.04289362e-01 3.31102848e-01 -1.07792713e-01 3.63198876e-01
-1.19167936e+00 2.06718534e-01 6.93219304e-01 6.79154098e-01
-3.53114232e-02 -4.41033036e-01 3.72545607e-02 -6.26905859e-01
-1.76946402e-01 -7.66914129e-01 4.23581004e-02 -9.71050084e-01
-8.81467223e-01 3.88610065e-01 5.31359494e-01 -1.07595503e+00
-4.11201119e-01 -5.55114448e-01 -8.95113945e-01 7.45468736e-01
-1.08120489e+00 -1.24794757e+00 -1.09772098e+00 8.57910514e-01
2.38391906e-01 -6.65417433e-01 5.50019681e-01 3.56152833e-01
-5.19011199e-01 9.42909420e-02 -4.70176339e-02 2.24722490e-01
8.00172240e-02 -9.06857014e-01 5.83752811e-01 6.24869108e-01
-3.74517411e-01 3.27714421e-02 8.95705521e-01 -9.23257530e-01
-1.04521787e+00 -1.20534706e+00 4.95655298e-01 -1.00111485e+00
4.67255354e-01 -4.46197867e-01 -3.54554802e-01 6.86255932e-01
2.14577034e-01 3.42852831e-01 5.21705866e-01 -6.83561623e-01
3.37114185e-01 9.85836983e-02 -1.48723698e+00 6.03825927e-01
8.95131111e-01 9.39194579e-03 -2.18800783e-01 2.93666065e-01
4.51524764e-01 -4.83254910e-01 -4.37960327e-01 4.38203104e-02
3.66181433e-01 -1.41626489e+00 9.16777492e-01 -5.33186495e-01
-3.07214528e-01 -6.10380709e-01 2.58276518e-03 -1.34773576e+00
-6.83610365e-02 -7.97706842e-01 -3.22942972e-01 1.26521695e+00
5.38020395e-02 -5.21996140e-01 9.06748593e-01 8.00101459e-01
2.31678620e-01 5.70551201e-04 -8.83660495e-01 -8.71931911e-01
-1.78719237e-01 -3.68053317e-01 1.06989992e+00 6.11129820e-01
-3.10542375e-01 -9.72774401e-02 -1.68809891e-01 2.13661820e-01
6.92269146e-01 -5.97173333e-01 1.75548518e+00 -1.31092310e+00
3.62565219e-01 -2.82682210e-01 -8.83122623e-01 -4.02019441e-01
1.38374314e-01 -1.43384948e-01 -1.47111177e-01 -1.56325567e+00
-2.84914553e-01 -5.13343930e-01 6.88044012e-01 -2.11754009e-01
-2.71329675e-02 -2.24595536e-02 3.58880818e-01 -5.84811147e-04
-7.74407446e-01 4.78667974e-01 1.05498147e+00 3.09693396e-01
-2.63826966e-01 3.55472565e-01 4.90442244e-03 9.28762496e-01
9.83384073e-01 -4.48020399e-01 -1.07720889e-01 -3.33383232e-01
-3.37386318e-02 -6.89261034e-02 8.80175233e-01 -1.68849206e+00
4.89303678e-01 -2.18344465e-01 4.56619889e-01 -6.68559134e-01
3.01959544e-01 -9.58908558e-01 3.86657625e-01 6.54848099e-01
5.49552776e-02 4.45166409e-01 4.69542056e-01 3.20642203e-01
2.89624751e-01 -1.80709347e-01 5.58592439e-01 -5.62736094e-01
-7.39856482e-01 6.25753179e-02 -8.62423599e-01 3.36002022e-01
1.81399751e+00 -5.80561996e-01 -5.03528059e-01 -3.72242779e-01
-4.26293492e-01 3.49719942e-01 7.69184113e-01 3.49568069e-01
2.29169965e-01 -1.24822199e+00 -8.40744674e-01 1.02677904e-01
-7.10795596e-02 -1.67843521e-01 4.60039638e-02 4.26142722e-01
-1.02078485e+00 3.48520547e-01 -5.11931837e-01 -6.44478321e-01
-1.29984176e+00 5.92012942e-01 5.88286519e-01 -1.10816032e-01
-5.18235326e-01 4.41625118e-01 -3.67864609e-01 -8.35136354e-01
3.96059692e-01 -1.27920061e-01 -3.17829937e-01 -1.92668095e-01
9.57392514e-01 1.07506073e+00 -2.43068069e-01 -1.11648667e+00
-5.33116519e-01 2.93987095e-01 8.27114284e-01 -2.61473924e-01
9.14777577e-01 -3.75019796e-02 3.93742025e-01 3.32416594e-01
3.86537462e-01 1.48881659e-01 -1.52735782e+00 1.14184283e-01
-5.91754243e-02 -4.86909270e-01 -2.31379107e-01 -4.98116761e-01
-7.89129496e-01 5.64465344e-01 9.03851688e-01 4.19145942e-01
3.97126406e-01 -1.64750353e-01 5.53825498e-01 2.07350314e-01
8.09964418e-01 -1.12712014e+00 -1.32970706e-01 4.01110172e-01
5.16756833e-01 -1.26942444e+00 -1.51152313e-01 -3.00154071e-02
-7.89197683e-01 5.78193247e-01 9.17745531e-01 -1.14157528e-01
5.54597974e-01 7.24804580e-01 3.48305076e-01 -3.75360429e-01
-4.44241762e-01 -7.21199453e-01 -3.18449169e-01 1.73609340e+00
8.49522185e-03 2.62991518e-01 2.20326871e-01 2.09045142e-01
-4.10943210e-01 4.47046667e-01 6.70227945e-01 9.61499393e-01
-5.48159182e-01 -5.66153824e-01 -1.08353186e+00 -9.62839052e-02
-1.60393178e-01 6.60834789e-01 -5.72763503e-01 1.21675336e+00
6.63349152e-01 1.24566126e+00 2.52932698e-01 -1.34875864e-01
6.22126937e-01 -3.51066530e-01 4.15923446e-01 -1.84486508e-01
-1.03239083e+00 -7.77206063e-01 1.68933004e-01 -3.52830917e-01
-8.27564299e-01 -8.60524178e-01 -1.19780672e+00 -9.24292028e-01
-5.28654791e-02 4.31311892e-05 7.99104929e-01 9.62958455e-01
3.98889571e-01 4.29481566e-01 2.57574320e-01 -1.54224694e+00
3.69847417e-01 -9.01526630e-01 -9.15645957e-02 3.51629257e-01
6.36521801e-02 -8.97687197e-01 -1.29698932e-01 7.01238140e-02] | [8.247453689575195, -1.1653778553009033] |
e1c51b12-40ed-4b44-90aa-24b2835f0c4f | before-and-after-default-information-and | 2208.07163 | null | https://arxiv.org/abs/2208.07163v2 | https://arxiv.org/pdf/2208.07163v2.pdf | Before and after default: information and optimal portfolio via anticipating calculus | Default risk calculus plays a crucial role in portfolio optimization when the risky asset is under threat of bankruptcy. However, traditional stochastic control techniques are not applicable in this scenario, and additional assumptions are required to obtain the optimal solution in a before-and-after default context. We propose an alternative approach using forward integration, which allows to avoid one of the restrictive assumptions, the Jacod density hypothesis. We demonstrate that, in the case of logarithmic utility, the weaker intensity hypothesis is the appropriate condition for optimality. Furthermore, we establish the semimartingale decomposition of the risky asset in the filtration that is progressively enlarged to accommodate the default process, under the assumption of the existence of the optimal portfolio. This work aims to provide valueable insights for developing effective risk management strategies when facing default risk. | ["Bernardo D'Auria", 'Giulia Di Nunno', 'José A. Salmerón'] | 2022-07-05 | null | null | null | null | ['portfolio-optimization'] | ['time-series'] | [-3.02750133e-02 -7.92810023e-02 3.06707378e-02 4.20760922e-02
-1.55643120e-01 -6.32141590e-01 1.37908652e-01 -1.20432645e-01
-4.20919389e-01 9.59832072e-01 -3.67268324e-02 -7.33712137e-01
-4.75713789e-01 -1.13095582e+00 -1.43869475e-01 -1.05395114e+00
2.54660130e-01 1.37496114e-01 -7.37160025e-03 -6.68213665e-02
3.73365462e-01 7.87762821e-01 -1.25662112e+00 -3.78765076e-01
1.00600338e+00 1.17153800e+00 4.25648652e-02 4.29224938e-01
-1.51036486e-01 6.87731326e-01 -3.49164724e-01 -7.01794922e-01
5.40528476e-01 -4.62035179e-01 -4.86910433e-01 1.37888327e-01
-4.11515743e-01 -5.92497349e-01 2.48338878e-01 1.17725599e+00
3.06212395e-01 4.43185270e-01 8.26481104e-01 -8.26785743e-01
-2.80439049e-01 3.63598853e-01 -4.01199520e-01 4.26030785e-01
-7.01856688e-02 -5.02145439e-02 1.14449680e+00 -6.35503948e-01
1.54517367e-01 7.83098757e-01 3.74495298e-01 3.97687346e-01
-1.11604333e+00 -1.03096269e-01 1.65692657e-01 -1.63119137e-01
-1.00988650e+00 -1.68073565e-01 6.13445997e-01 -4.76394236e-01
6.00300550e-01 1.77285314e-01 4.39993620e-01 8.33430529e-01
5.32440901e-01 2.74215788e-01 9.79715466e-01 -4.77672517e-01
6.04096711e-01 7.13324407e-03 2.06413627e-01 -3.72240879e-02
8.45579088e-01 7.60962218e-02 -2.79040243e-02 -3.96300256e-01
9.68770742e-01 1.40322492e-01 -4.63418603e-01 -3.25580090e-01
-5.51985681e-01 1.09461772e+00 -2.61052817e-01 4.33404624e-01
-7.77054965e-01 -2.41758958e-01 8.40463191e-02 4.63176906e-01
4.11573052e-01 2.60912120e-01 -1.78498045e-01 -1.69986784e-01
-8.85247648e-01 4.37978715e-01 1.15392947e+00 4.20490623e-01
9.24807414e-02 2.92628825e-01 -1.30610347e-01 1.97897539e-01
3.81477922e-01 4.42032814e-01 6.48403913e-02 -1.33865845e+00
7.64696419e-01 1.99370265e-01 7.79665828e-01 -5.68326414e-01
-2.17138026e-02 -6.97870851e-01 -7.06149340e-01 8.09416771e-01
9.38662589e-01 -4.07716513e-01 -1.41613558e-01 1.63916397e+00
1.56863958e-01 -1.15215197e-01 2.47116238e-01 5.05317211e-01
-5.98644376e-01 3.32419932e-01 -1.86449483e-01 -9.05117989e-01
1.02063239e+00 -2.52991825e-01 -8.89646709e-01 2.52190262e-01
3.06201428e-01 -3.43266815e-01 1.06984413e+00 6.46516502e-01
-1.29765189e+00 2.20806882e-01 -8.63044381e-01 6.29994392e-01
1.55377999e-01 -3.11741322e-01 2.63404995e-01 1.15652764e+00
-7.49394476e-01 9.08594191e-01 -7.69120514e-01 -1.16223589e-01
4.58444543e-02 -5.76948673e-02 1.82984963e-01 3.68283868e-01
-1.07960069e+00 1.03918064e+00 -1.96601711e-02 2.45043218e-01
-5.79581618e-01 -5.72001636e-01 -3.24859381e-01 4.85456526e-01
4.29164469e-01 -5.89247286e-01 1.21347070e+00 -8.01656485e-01
-1.56551087e+00 3.73379648e-01 3.56854707e-01 -7.14120328e-01
1.13661766e+00 -2.98983961e-01 1.71913952e-01 3.75064909e-01
-7.21501783e-02 -8.16415370e-01 7.12784886e-01 -6.72902524e-01
-4.94891524e-01 -3.97971720e-01 4.15006757e-01 2.12862834e-01
-3.74331862e-01 2.86586255e-01 5.21039486e-01 -9.27834392e-01
8.76616221e-04 -5.53483844e-01 -3.37569058e-01 -7.95408264e-02
-3.11078201e-03 2.68907044e-02 3.89120579e-02 -5.89000106e-01
1.21108270e+00 -1.90203953e+00 -1.38742179e-01 3.28208864e-01
-2.35771224e-01 -1.28753230e-01 7.34836340e-01 6.23880446e-01
-1.33741856e-01 1.85787648e-01 -5.08453846e-01 -2.46251762e-01
1.97261736e-01 -9.51919779e-02 -7.74843812e-01 5.57761610e-01
4.58892845e-02 2.24807829e-01 -4.57237959e-01 -1.52778506e-01
-9.97910202e-02 2.09366009e-01 -4.05741423e-01 1.53351784e-01
-4.42874357e-02 4.58919466e-01 -6.92722201e-01 3.87939125e-01
6.49124146e-01 -4.22876235e-03 2.17476457e-01 3.57911587e-01
-5.23292005e-01 5.41845746e-02 -1.58281434e+00 7.52342820e-01
-5.33125699e-01 5.52916666e-03 3.70873541e-01 -1.13726604e+00
7.46097624e-01 5.40097237e-01 4.23850954e-01 -3.44543636e-01
1.59873083e-01 3.44745845e-01 -1.44196957e-01 -3.93791437e-01
2.21252203e-01 -1.02954674e+00 1.10853322e-01 7.61571765e-01
-4.99803334e-01 1.99851707e-01 1.06992073e-01 -2.51078963e-01
8.08697224e-01 8.48170519e-02 2.67590731e-01 -6.59329176e-01
4.41310585e-01 -5.37716091e-01 7.93774784e-01 5.94323814e-01
-1.45927444e-01 4.73458439e-01 1.09085727e+00 -2.81910803e-02
-7.96689212e-01 -1.22548187e+00 -3.90933126e-01 5.31757057e-01
-2.44521692e-01 4.38799948e-01 -7.97379792e-01 -4.18254167e-01
2.04080373e-01 1.16277528e+00 -5.57840228e-01 4.62781303e-02
-5.14144599e-01 -1.23148727e+00 1.92342475e-01 5.32928944e-01
4.56232667e-01 -8.57143998e-01 -1.06984770e+00 4.39615935e-01
4.69814688e-02 -5.10974526e-01 -2.51078904e-01 2.62666553e-01
-1.10929751e+00 -1.11370909e+00 -1.22247148e+00 9.70756635e-03
4.77951527e-01 -1.30945042e-01 7.88847506e-01 -1.61756035e-02
1.81733534e-01 5.04887283e-01 -1.54128462e-01 -1.63324505e-01
-2.67521113e-01 -3.57496500e-01 7.20530283e-03 5.44411778e-01
-1.99656039e-02 -6.42282605e-01 -6.99144483e-01 2.11367965e-01
-9.48775947e-01 -6.33538783e-01 3.41641679e-02 6.10333204e-01
2.25332603e-01 5.06116509e-01 9.14192557e-01 -4.22009289e-01
7.32247770e-01 -5.34563899e-01 -9.53445911e-01 3.95674199e-01
-8.94147217e-01 1.20086871e-01 6.27102315e-01 -1.71754658e-01
-1.74918056e+00 -3.89637917e-01 3.87501493e-02 -1.15095578e-01
5.16275726e-02 5.09115636e-01 -6.77970648e-01 6.89897910e-02
9.33729932e-02 -1.40186734e-02 -1.06127843e-01 -9.47491467e-01
-4.14599329e-02 4.24329340e-01 2.48145893e-01 -7.86715746e-01
5.86989343e-01 6.54693723e-01 3.81584823e-01 -6.39281094e-01
-6.97664261e-01 -4.29195799e-02 -6.87401652e-01 -3.27834189e-02
8.30224752e-01 -3.33254039e-01 -7.44165421e-01 5.44787109e-01
-9.61683571e-01 -1.02412961e-01 -5.37156820e-01 6.56816363e-01
-9.22398567e-01 6.51348770e-01 -7.40254641e-01 -1.54755771e+00
-1.41940430e-01 -8.49242210e-01 1.27908513e-02 3.40062737e-01
6.23270348e-02 -1.28322184e+00 9.55714732e-02 1.28175884e-01
4.19317454e-01 5.29908717e-01 9.16088820e-01 -6.83339119e-01
-5.38028598e-01 -3.05127293e-01 1.94821209e-01 5.92429698e-01
1.88323081e-01 -2.69456170e-02 -5.13736308e-01 -9.51641649e-02
9.68693256e-01 2.64441222e-01 6.95447624e-01 5.40979147e-01
4.47653741e-01 -3.33652377e-01 2.14286253e-01 4.32700127e-01
1.60583425e+00 4.07852918e-01 5.83074570e-01 6.67095423e-01
-4.91381250e-02 1.09858930e+00 7.77211249e-01 9.44313109e-01
6.32281005e-02 4.29709584e-01 4.25855160e-01 5.28905749e-01
7.76506424e-01 1.66927665e-01 3.44783098e-01 8.95127580e-02
-4.60616916e-01 -2.06475183e-01 -7.94555545e-01 4.65688020e-01
-1.68151629e+00 -1.14927745e+00 -1.21588133e-01 2.77260447e+00
7.15189397e-01 3.69783521e-01 2.60750145e-01 2.42228732e-01
7.48880684e-01 -1.18920714e-01 -3.56279790e-01 -4.65617865e-01
-3.07049125e-01 2.22722992e-01 6.16409659e-01 6.42062068e-01
-8.57833385e-01 1.53473660e-01 6.31213570e+00 3.88303518e-01
-7.01224804e-01 -7.37239495e-02 5.65797865e-01 -7.67968222e-02
-4.73163784e-01 3.16811532e-01 -8.57512832e-01 6.67321503e-01
8.40542614e-01 -6.45713389e-01 -4.02573422e-02 6.69432521e-01
6.97392166e-01 -3.25946271e-01 -6.61410987e-01 2.48962998e-01
-3.89430255e-01 -6.54150188e-01 -2.38086298e-01 4.24342841e-01
2.55342126e-01 -9.76710737e-01 1.23676300e-01 -2.53689915e-01
2.27632988e-02 -6.69133067e-01 7.81243503e-01 1.01012301e+00
1.57793343e-01 -1.09726453e+00 7.49706209e-01 4.57521230e-01
-9.21267092e-01 -3.03983688e-01 -2.85847008e-01 -3.56374472e-01
7.92702973e-01 7.54099309e-01 2.22383812e-02 5.92340887e-01
3.61785710e-01 1.91671312e-01 -2.03391891e-02 1.16834307e+00
-1.72281966e-01 6.01078272e-01 -2.83087134e-01 3.52583617e-01
3.59174944e-02 -7.85908341e-01 6.55273795e-01 6.99812591e-01
6.31583214e-01 1.41061157e-01 -3.23029369e-01 1.08506763e+00
4.53148037e-01 2.10538238e-01 -4.88452315e-01 1.13877572e-01
1.19675890e-01 7.49327838e-01 -1.05865180e+00 -2.58660223e-02
-6.50320888e-01 8.55206311e-01 -2.41311476e-01 4.64908183e-01
-5.59376717e-01 -4.65420038e-01 5.73362529e-01 2.67096013e-01
5.08046329e-01 -2.72495389e-01 -7.13366389e-01 -1.26460803e+00
4.84914660e-01 -5.56762815e-01 6.14240468e-01 -8.37673154e-03
-1.25654769e+00 1.12893716e-01 1.80687442e-01 -1.05656040e+00
-4.27812457e-01 -5.72268307e-01 -1.27062035e+00 1.27300858e+00
-1.55897248e+00 -5.37604094e-01 3.82478982e-01 5.18844903e-01
1.91286638e-01 -8.84901509e-02 4.45009977e-01 7.02157095e-02
-7.53543735e-01 2.09224463e-01 6.34158194e-01 -3.08767796e-01
2.52545089e-01 -1.41633487e+00 -2.84687784e-02 1.08033729e+00
-4.76893455e-01 6.92784965e-01 9.68861461e-01 -9.29764092e-01
-9.59238470e-01 -5.49467921e-01 9.19132888e-01 -1.10533766e-01
9.93439078e-01 5.69932908e-02 -1.16431034e+00 4.27981794e-01
-7.00747455e-03 -5.96314073e-01 7.37733841e-01 -4.06543612e-01
2.16368586e-01 -4.90361974e-02 -1.15608561e+00 4.49999988e-01
4.49823081e-01 -2.82842636e-01 -7.72542953e-01 8.11454933e-03
3.78403693e-01 3.37540507e-01 -1.06228757e+00 1.79973289e-01
5.46205580e-01 -1.10955453e+00 8.96172941e-01 -3.59642774e-01
-3.29339094e-02 8.62423107e-02 -2.92718202e-01 -6.39930964e-01
-8.00292566e-02 -1.00616288e+00 -2.92615388e-02 1.61257994e+00
2.35795394e-01 -9.97195244e-01 7.57518470e-01 1.01658106e+00
3.96482855e-01 -7.74981320e-01 -1.35243917e+00 -1.12343121e+00
6.70304179e-01 -1.94023997e-01 5.12437701e-01 5.63793421e-01
-3.76271196e-02 -3.47675622e-01 -3.06584895e-01 -1.91063099e-02
9.58797097e-01 2.17838898e-01 -2.53565777e-02 -1.34639275e+00
-5.95652699e-01 -6.06205702e-01 2.18993023e-01 -5.97049236e-01
2.65136749e-01 -2.69096732e-01 -2.53705412e-01 -1.11364293e+00
-6.77215531e-02 -3.05705994e-01 -5.13105631e-01 -2.67896116e-01
-1.55007899e-01 -4.43890125e-01 3.62736493e-01 3.24070275e-01
1.79657459e-01 6.96006894e-01 8.62937868e-01 4.13857490e-01
-2.80242383e-01 9.47038174e-01 -7.58391738e-01 8.09053600e-01
9.38689232e-01 -2.57922620e-01 -4.80313927e-01 9.82896760e-02
3.96495461e-01 4.39146191e-01 3.88892710e-01 -4.38883990e-01
-2.91709062e-02 -7.26954818e-01 -1.80738747e-01 -4.99281943e-01
8.67908448e-02 -6.87347770e-01 2.09009260e-01 5.34698129e-01
-1.50368750e-01 1.36894539e-01 -3.16177070e-01 5.94707668e-01
1.79590315e-01 -1.12803984e+00 9.69228685e-01 -2.26361722e-01
1.42456651e-01 6.62994236e-02 -5.67022383e-01 1.35396972e-01
1.08260202e+00 -2.52462387e-01 1.40534580e-01 -4.17847723e-01
-9.21667635e-01 3.73222232e-02 6.70258820e-01 -2.04993591e-01
3.60191077e-01 -9.71891046e-01 -5.42002141e-01 -9.15164798e-02
-6.44109190e-01 -3.88939083e-01 1.62562817e-01 9.74194825e-01
-5.45233607e-01 4.04015213e-01 -3.37242149e-02 2.30602831e-01
-7.32450247e-01 6.18762255e-01 7.22221494e-01 -4.45212215e-01
-6.95688963e-01 2.59419411e-01 2.97680557e-01 6.03783071e-01
2.12765291e-01 -2.24210843e-01 -3.87755483e-01 4.06348705e-01
6.24598384e-01 8.14089596e-01 -2.93274224e-02 -3.43133509e-01
-7.14617595e-02 4.70572680e-01 2.52402276e-01 -6.12807274e-01
1.30505145e+00 -5.96863806e-01 -8.96217600e-02 5.66939175e-01
5.90963304e-01 7.90175796e-02 -1.52019596e+00 2.09924519e-01
5.19501328e-01 -4.33871388e-01 -5.43392859e-02 -8.08736831e-02
-8.97249222e-01 1.07906711e+00 1.81693077e-01 5.24663866e-01
1.07254946e+00 -4.20585573e-01 5.72713792e-01 1.68021873e-01
3.04466456e-01 -1.13094354e+00 -2.27748439e-01 1.62234679e-01
9.52213287e-01 -5.24129629e-01 -1.73811749e-01 -2.64018238e-01
-4.61972296e-01 1.31904197e+00 2.41589304e-02 -2.74579883e-01
1.02172768e+00 3.48957688e-01 -2.43629783e-01 2.52047151e-01
-4.29936200e-01 -1.81615770e-01 -2.56506622e-01 3.72304350e-01
3.03964972e-01 7.97034130e-02 -8.18071067e-01 1.01770413e+00
9.59551781e-02 -2.12342627e-02 8.08965743e-01 1.22608805e+00
-5.15059471e-01 -1.22846496e+00 -6.44231439e-01 1.99772894e-01
-1.19689155e+00 -1.21342570e-01 1.07337169e-01 5.65221965e-01
-3.11349750e-01 8.09945703e-01 -7.18027055e-02 5.62912822e-01
4.92855281e-01 4.16911721e-01 3.10839504e-01 -4.17444915e-01
-3.94558847e-01 2.93699384e-01 -1.87143296e-01 -4.65961732e-02
-3.09186935e-01 -1.06674433e+00 -8.79805505e-01 -2.14649200e-01
-5.67448676e-01 2.45769009e-01 1.70485497e-01 1.04327774e+00
-2.43055373e-01 1.27435595e-01 8.98825109e-01 -4.57500964e-01
-1.45967472e+00 -6.15302324e-01 -1.35742593e+00 4.51268628e-02
2.57365048e-01 -7.56571114e-01 -9.76781130e-01 -8.93066358e-03] | [4.94113302230835, 3.9389216899871826] |
796cf4c2-5740-406a-88dc-639993a560e2 | the-cloud-of-knowing-non-factive-al-ta-aknowa | null | null | https://aclanthology.org/Y16-3026 | https://aclanthology.org/Y16-3026.pdf | The Cloud of Knowing: Non-factive al-ta `know' (as a Neg-raiser) in Korean | null | ['Chungmin Lee', 'Seungjin Hong'] | 2016-10-01 | the-cloud-of-knowing-non-factive-al-ta-know | https://aclanthology.org/Y16-3026 | https://aclanthology.org/Y16-3026.pdf | paclic-2016-10 | ['rumour-detection'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.245980739593506, 3.6938414573669434] |
a7854d2a-fc3e-48e8-aa0c-5f71ec360401 | decanus-to-legatus-synthetic-training-for-2d | 2210.02231 | null | https://arxiv.org/abs/2210.02231v1 | https://arxiv.org/pdf/2210.02231v1.pdf | Decanus to Legatus: Synthetic training for 2D-3D human pose lifting | 3D human pose estimation is a challenging task because of the difficulty to acquire ground-truth data outside of controlled environments. A number of further issues have been hindering progress in building a universal and robust model for this task, including domain gaps between different datasets, unseen actions between train and test datasets, various hardware settings and high cost of annotation, etc. In this paper, we propose an algorithm to generate infinite 3D synthetic human poses (Legatus) from a 3D pose distribution based on 10 initial handcrafted 3D poses (Decanus) during the training of a 2D to 3D human pose lifter neural network. Our results show that we can achieve 3D pose estimation performance comparable to methods using real data from specialized datasets but in a zero-shot setup, showing the generalization potential of our framework. | ['David Picard', 'Yue Zhu'] | 2022-10-05 | null | null | null | null | ['3d-pose-estimation', '3d-human-pose-estimation'] | ['computer-vision', 'computer-vision'] | [ 2.19215319e-01 2.44487539e-01 3.46683860e-01 -3.45271379e-01
-8.37764740e-01 -5.96616864e-01 4.78148401e-01 -4.70557272e-01
-6.27590477e-01 9.47862089e-01 2.43967742e-01 1.91920012e-01
1.26504585e-01 -2.73172319e-01 -9.26743567e-01 -1.78415790e-01
-1.38061404e-01 1.20126665e+00 3.05547655e-01 -4.71836269e-01
-5.29856458e-02 5.89366794e-01 -1.55243301e+00 -1.12250239e-01
2.91823626e-01 8.00547659e-01 -5.99628985e-02 7.76699424e-01
5.36541462e-01 9.09710005e-02 -9.78716552e-01 -3.47450823e-01
7.75407493e-01 -5.16273737e-01 -7.54669845e-01 3.53515595e-01
6.74725831e-01 -5.50118506e-01 -1.67445138e-01 6.48455858e-01
1.03872859e+00 3.35654110e-01 6.18409634e-01 -1.38045669e+00
6.57780990e-02 -1.62451044e-02 -2.19560221e-01 -2.22918421e-01
8.48316014e-01 3.61828744e-01 4.88128841e-01 -7.32698679e-01
1.05453086e+00 1.17702162e+00 9.98731077e-01 8.47398758e-01
-1.21964860e+00 -3.73650968e-01 -2.57344097e-01 -1.63818628e-01
-1.55357647e+00 -2.78240919e-01 5.03770828e-01 -6.60833836e-01
9.74997401e-01 8.90468210e-02 1.06911147e+00 1.89070058e+00
6.74666613e-02 6.56559348e-01 9.20748234e-01 -5.12995958e-01
2.70253778e-01 -7.52256289e-02 -4.50138628e-01 5.55378199e-01
4.58439201e-01 1.12909928e-01 -4.86033022e-01 -6.40634671e-02
1.05603671e+00 -2.33955830e-01 -2.06871480e-01 -1.00474226e+00
-1.36678338e+00 5.63221693e-01 3.30230862e-01 -1.70975894e-01
-2.99511671e-01 1.93528250e-01 5.51515460e-01 -1.13207817e-01
1.20601267e-01 6.42643631e-01 -7.38572180e-01 -2.20308810e-01
-7.41414666e-01 9.45425868e-01 8.36846411e-01 1.28904247e+00
3.93073231e-01 -1.08719625e-01 -3.91771384e-02 3.29780400e-01
-1.50483459e-01 3.46212745e-01 4.37166840e-01 -1.09672606e+00
5.80808222e-01 4.28260803e-01 6.56678915e-01 -7.97617078e-01
-6.60395324e-01 -3.74320447e-01 -3.98667186e-01 3.07544440e-01
8.28306854e-01 -3.89568597e-01 -9.21075940e-01 1.74739552e+00
5.72145998e-01 -2.79233992e-01 -7.87431747e-02 1.28040564e+00
4.47406590e-01 1.68858081e-01 -4.62291241e-02 3.26960832e-01
1.05655754e+00 -6.66127920e-01 -2.93270677e-01 -5.38840413e-01
3.97641689e-01 -5.20691276e-01 1.10098171e+00 5.35520554e-01
-9.59017992e-01 -8.14013064e-01 -1.00443137e+00 -1.09678507e-01
-2.86334217e-01 1.90011159e-01 6.51789725e-01 7.32428014e-01
-5.85450411e-01 7.33583808e-01 -8.68915915e-01 -7.47253954e-01
2.12216496e-01 4.36923027e-01 -8.06311488e-01 2.17851251e-01
-1.28405833e+00 1.26263750e+00 7.40921736e-01 3.04405779e-01
-9.65176284e-01 -2.40631700e-01 -9.58096623e-01 -5.59013903e-01
7.26111293e-01 -8.49973977e-01 1.23982680e+00 -5.69204986e-01
-1.41605270e+00 1.25845277e+00 3.69837910e-01 -4.73732263e-01
1.12008464e+00 -8.81035924e-01 1.64675578e-01 -1.64767668e-01
7.69325867e-02 9.66457903e-01 6.03062868e-01 -1.15619636e+00
-2.78995156e-01 -5.17608583e-01 -6.39039576e-02 3.89845163e-01
1.68165207e-01 -3.33109230e-01 -6.12662435e-01 -5.33077896e-01
5.28725088e-02 -1.31251431e+00 -4.35412079e-01 5.13221696e-02
-3.74687463e-01 1.46582350e-01 3.78946036e-01 -8.18818033e-01
5.55361211e-01 -1.79098237e+00 5.64917088e-01 -8.39920938e-02
-5.27301282e-02 3.40378284e-01 1.31016672e-01 4.95852530e-01
2.82980632e-02 -1.73043489e-01 -4.99799363e-02 -2.82370448e-01
1.66862741e-01 2.94530928e-01 -1.19030870e-01 5.26508629e-01
2.56279618e-01 9.36016738e-01 -8.99337769e-01 -4.73976761e-01
3.65369469e-01 5.24429142e-01 -6.57613873e-01 4.31741446e-01
-2.18589753e-01 7.63119102e-01 -2.34124497e-01 5.62568188e-01
3.40332270e-01 -5.09097055e-03 2.19847679e-01 -2.27039024e-01
2.84590870e-01 1.40997112e-01 -1.45955682e+00 2.18796110e+00
-1.75231159e-01 3.57463688e-01 -2.70551234e-01 -6.69298351e-01
8.52420628e-01 2.73383558e-01 4.01646346e-01 -1.93453878e-01
5.82919776e-01 2.65222400e-01 -1.28788829e-01 -4.90348279e-01
5.42405605e-01 -2.88479537e-01 -5.21063149e-01 2.31505170e-01
4.58581865e-01 -4.74608958e-01 1.09130733e-01 -2.61758000e-01
8.89682412e-01 1.03791702e+00 3.92563045e-01 -1.06199063e-01
2.37601809e-02 2.60791332e-01 5.31106651e-01 5.56603134e-01
-5.71402907e-01 1.15234983e+00 2.59869576e-01 -8.15645337e-01
-1.44111657e+00 -1.23714662e+00 1.01735048e-01 7.32404530e-01
-2.36504115e-02 -3.98218840e-01 -9.46767151e-01 -6.29276931e-01
-2.15782924e-03 5.20418584e-01 -7.05469787e-01 -1.70274094e-01
-8.58153045e-01 -5.68449914e-01 7.69381225e-01 7.43508816e-01
4.83935118e-01 -8.67482841e-01 -1.31516850e+00 1.67151585e-01
-3.71793956e-01 -1.51887035e+00 -2.71939546e-01 2.82857746e-01
-7.18090355e-01 -1.03099251e+00 -8.60821307e-01 -6.36422634e-01
5.74029446e-01 -3.07702899e-01 1.30207360e+00 -4.07454580e-01
-6.23662949e-01 2.48261005e-01 -3.44328016e-01 -6.15065157e-01
-1.53991342e-01 9.48056132e-02 4.70531285e-01 -5.81242442e-01
2.63730496e-01 -2.81592339e-01 -4.90812391e-01 5.28598070e-01
-4.34353322e-01 7.99581110e-02 5.12023747e-01 8.71142983e-01
5.79309225e-01 -1.83911681e-01 1.92091167e-01 -5.75587749e-01
4.12147701e-01 6.76029474e-02 -4.92827088e-01 -8.62802379e-03
5.43926880e-02 2.42197178e-02 2.51043051e-01 -5.58866203e-01
-7.70093024e-01 5.98926425e-01 -2.14992315e-01 -3.51232797e-01
-5.41422904e-01 -1.85759842e-01 -2.39649907e-01 8.23687762e-02
1.21358645e+00 -9.43460464e-02 2.02349782e-01 -3.93635243e-01
2.39382342e-01 3.73077750e-01 8.11151385e-01 -7.79687345e-01
8.66178632e-01 3.31179708e-01 9.89698023e-02 -5.55478394e-01
-1.07238841e+00 -1.26541018e-01 -1.30592191e+00 -3.09399366e-01
9.77678776e-01 -1.12182069e+00 -5.66808105e-01 4.01578188e-01
-1.05806530e+00 -4.12911177e-01 -3.56770426e-01 6.03242218e-01
-1.18806672e+00 2.30213255e-01 -3.64734948e-01 -7.76987195e-01
-2.28344560e-01 -1.09147024e+00 1.49487340e+00 -1.54968584e-02
-9.94614303e-01 -5.54316759e-01 6.77633435e-02 4.35259223e-01
5.19608613e-03 1.14729548e+00 2.66761541e-01 -4.46568042e-01
-2.47765407e-01 -8.06818843e-01 3.31219822e-01 3.44181627e-01
-1.16747515e-02 -3.72264624e-01 -7.64120758e-01 -4.21919316e-01
-2.40055427e-01 -1.01163268e+00 1.23016834e-01 1.35895982e-01
7.27721870e-01 8.75618830e-02 -3.27706397e-01 4.60671246e-01
1.07824171e+00 -3.10473979e-01 5.00065565e-01 3.91717821e-01
5.80241680e-01 7.01905191e-01 7.80093074e-01 4.69839275e-01
5.94046116e-02 9.11712825e-01 3.18597108e-01 6.05787598e-02
-1.92995053e-02 -6.13971829e-01 8.67080241e-02 1.76059343e-02
-4.58765209e-01 -1.92567557e-01 -9.81504261e-01 3.92435342e-01
-1.74068809e+00 -7.75080085e-01 1.76306948e-01 2.44232011e+00
6.90692067e-01 5.53412914e-01 6.10617161e-01 2.73632854e-01
6.07073665e-01 -2.30879709e-01 -5.51089942e-01 4.05806005e-02
1.64674059e-01 2.07957223e-01 4.82515097e-01 1.80001646e-01
-1.09777272e+00 9.96401727e-01 6.90585184e+00 3.12939733e-01
-8.10549319e-01 -2.40162879e-01 1.64856970e-01 -2.84286827e-01
4.70238000e-01 -1.97300777e-01 -8.97609890e-01 1.24438606e-01
5.84336162e-01 2.08570227e-01 1.73100874e-01 1.03588581e+00
-9.35844257e-02 -3.19329314e-02 -1.36186290e+00 1.08192337e+00
2.21740633e-01 -7.12444305e-01 -1.79411441e-01 -2.33333074e-02
6.79276705e-01 -1.99519247e-01 -2.01357156e-01 4.05395091e-01
3.72018009e-01 -9.76394832e-01 9.67101514e-01 2.86823928e-01
8.90379608e-01 -6.57624662e-01 7.57058024e-01 8.00226331e-01
-7.33853757e-01 3.00366551e-01 -2.46844247e-01 -2.86628246e-01
3.14297795e-01 7.94470236e-02 -1.33285511e+00 5.44429660e-01
7.22195566e-01 1.75634086e-01 -6.03936672e-01 9.85794961e-01
-3.31338853e-01 3.60042751e-02 -4.79828477e-01 -1.66083500e-01
3.22584920e-02 2.27358907e-01 5.31579435e-01 9.43009317e-01
2.70342588e-01 -3.12171429e-01 3.29303026e-01 5.99815130e-01
2.20265478e-01 -3.07840973e-01 -7.83809841e-01 1.23844624e-01
1.06394179e-01 9.92073596e-01 -7.39605248e-01 -6.64368272e-02
2.44627088e-01 1.30182850e+00 2.60606915e-01 4.63004522e-02
-9.90579128e-01 -3.18471402e-01 4.14269686e-01 3.22535694e-01
1.11449540e-01 -6.65174305e-01 -1.17960125e-01 -1.15483403e+00
2.31404990e-01 -1.02975214e+00 3.63349557e-01 -8.26089740e-01
-1.08897936e+00 6.98887944e-01 5.33252895e-01 -1.41042042e+00
-8.54513645e-01 -8.41885507e-01 5.01810247e-03 7.69211888e-01
-5.53319991e-01 -9.88303423e-01 -5.50610840e-01 3.34934592e-01
4.82553840e-01 2.29316950e-03 1.00366139e+00 1.20181628e-01
-1.31984070e-01 5.00687897e-01 -6.18392944e-01 2.73743182e-01
8.70729625e-01 -1.13155067e+00 8.85910749e-01 5.35839200e-01
8.69657397e-02 5.15166223e-01 1.16023564e+00 -6.97424114e-01
-1.27135718e+00 -7.71384835e-01 7.62836456e-01 -1.22788596e+00
1.66515276e-01 -9.11044896e-01 -4.68481123e-01 9.21487093e-01
-2.15492025e-01 8.61948263e-03 4.97153133e-01 1.28315151e-01
-6.75308183e-02 2.28762150e-01 -1.26891172e+00 7.47971892e-01
1.60404897e+00 -1.44362599e-01 -9.22958732e-01 3.61418396e-01
5.11236012e-01 -9.65774000e-01 -7.08596408e-01 6.09240055e-01
9.01208043e-01 -8.30487669e-01 1.07956719e+00 -8.31735611e-01
2.29711786e-01 -3.54635358e-01 -3.06683034e-01 -1.20641458e+00
-7.58070126e-02 -6.94060206e-01 9.78953689e-02 5.86495161e-01
1.09876700e-01 -1.91461012e-01 1.17038476e+00 5.85740507e-01
1.67163923e-01 -5.85975587e-01 -8.21812034e-01 -9.97006834e-01
9.88315418e-03 -3.96712899e-01 4.31253880e-01 4.89852220e-01
-3.00484359e-01 5.16292095e-01 -7.28810608e-01 1.64004955e-02
7.94033051e-01 -1.72325313e-01 1.56660938e+00 -1.25121248e+00
-4.40232128e-01 7.85255581e-02 -7.73640871e-01 -9.86004353e-01
4.89685200e-02 -4.30063069e-01 3.42697501e-01 -1.32762277e+00
1.47599382e-02 -3.67429331e-02 3.58619303e-01 2.86974490e-01
-4.48369458e-02 3.95365030e-01 1.60742790e-01 -5.86032346e-02
-5.88207960e-01 5.36345243e-01 1.24356711e+00 3.20884287e-01
-9.76967439e-02 -6.01328723e-02 -2.70156920e-01 9.21104074e-01
6.09393835e-01 -3.69961530e-01 -3.79653722e-01 -3.43959063e-01
5.50420061e-02 7.45429695e-02 6.94580436e-01 -1.52948976e+00
-1.30394533e-01 -1.75220091e-02 9.98303294e-01 -7.25295544e-01
7.15029657e-01 -7.86792934e-01 4.58911121e-01 7.09676862e-01
-1.73367068e-01 8.29448029e-02 3.17741483e-01 4.29957449e-01
1.59542263e-01 -8.20980445e-02 6.15594387e-01 -6.63353980e-01
-8.22000384e-01 9.85526219e-02 4.39998433e-02 4.18960840e-01
1.09670281e+00 -5.39760470e-01 3.27200532e-01 -2.19256446e-01
-9.31124389e-01 -1.88792750e-04 6.41781688e-01 5.59695721e-01
3.85554165e-01 -1.37474608e+00 -6.46182418e-01 2.94720829e-01
1.93151206e-01 3.15802157e-01 -3.58316377e-02 2.66663909e-01
-8.36247921e-01 3.33233565e-01 -7.90557325e-01 -8.02648246e-01
-1.03316820e+00 3.49667460e-01 4.37941462e-01 -1.78100467e-01
-7.61996329e-01 8.35711718e-01 -2.15624258e-01 -9.21055913e-01
3.77908885e-01 -4.21533268e-03 4.19014156e-01 -2.93990135e-01
2.64364153e-01 3.21994931e-01 1.08337767e-01 -7.39083529e-01
-3.98521423e-01 5.93882978e-01 3.14309657e-01 -3.75457764e-01
1.23455870e+00 1.70778781e-01 6.45393491e-01 4.98806000e-01
1.09525013e+00 -2.84925878e-01 -1.58884025e+00 1.86528519e-01
-1.65491417e-01 -6.30153716e-01 -6.77028775e-01 -7.89378047e-01
-4.08477694e-01 6.81859314e-01 6.10888898e-01 -3.71455640e-01
6.39997661e-01 2.63151955e-02 8.10084581e-01 7.04714894e-01
9.13473308e-01 -1.45274985e+00 3.37558001e-01 4.62800443e-01
1.17037678e+00 -1.25502563e+00 1.19758405e-01 -3.29697877e-01
-7.94974625e-01 8.94509614e-01 8.49131167e-01 -3.42406958e-01
1.34415880e-01 1.33428127e-01 1.68276951e-01 -8.38969946e-02
-4.13984120e-01 -2.47096270e-01 3.47508639e-01 9.31297004e-01
3.33456904e-01 -3.38569134e-02 -1.53965220e-01 3.64163935e-01
-6.74892843e-01 1.66561157e-01 3.05370390e-01 1.21832752e+00
-2.43776575e-01 -1.01511204e+00 -5.44123590e-01 -7.42447898e-02
-2.92090237e-01 5.80492377e-01 -6.32536829e-01 1.32092917e+00
3.97172749e-01 4.56649512e-01 -2.03768373e-01 -6.25670552e-01
9.10561442e-01 2.94932008e-01 1.05683589e+00 -7.68006980e-01
-4.31413114e-01 -1.90940294e-02 2.71602362e-01 -5.09146929e-01
-1.46680444e-01 -7.60063350e-01 -1.07303905e+00 5.13763316e-02
-2.29605317e-01 -2.55485833e-01 6.26692712e-01 8.55689108e-01
2.42770284e-01 3.14931780e-01 1.06068626e-02 -1.60590255e+00
-8.76518786e-01 -1.01958466e+00 -3.87130558e-01 8.81389737e-01
-4.16690260e-02 -1.13447845e+00 -1.90501183e-01 1.44802228e-01] | [6.9679131507873535, -1.0077049732208252] |
33035a88-f265-4599-9554-56e2dcc3c056 | semi-supervised-visual-tracking-of-marine | 2302.07344 | null | https://arxiv.org/abs/2302.07344v1 | https://arxiv.org/pdf/2302.07344v1.pdf | Semi-Supervised Visual Tracking of Marine Animals using Autonomous Underwater Vehicles | In-situ visual observations of marine organisms is crucial to developing behavioural understandings and their relations to their surrounding ecosystem. Typically, these observations are collected via divers, tags, and remotely-operated or human-piloted vehicles. Recently, however, autonomous underwater vehicles equipped with cameras and embedded computers with GPU capabilities are being developed for a variety of applications, and in particular, can be used to supplement these existing data collection mechanisms where human operation or tags are more difficult. Existing approaches have focused on using fully-supervised tracking methods, but labelled data for many underwater species are severely lacking. Semi-supervised trackers may offer alternative tracking solutions because they require less data than fully-supervised counterparts. However, because there are not existing realistic underwater tracking datasets, the performance of semi-supervised tracking algorithms in the marine domain is not well understood. To better evaluate their performance and utility, in this paper we provide (1) a novel dataset specific to marine animals located at http://warp.whoi.edu/vmat/, (2) an evaluation of state-of-the-art semi-supervised algorithms in the context of underwater animal tracking, and (3) an evaluation of real-world performance through demonstrations using a semi-supervised algorithm on-board an autonomous underwater vehicle to track marine animals in the wild. | ['Yogesh Girdhar', 'T. Aran Mooney', 'Roger Hanlon', 'Nathan E. McGuire', 'Levi Cai'] | 2023-02-14 | null | null | null | null | ['visual-tracking'] | ['computer-vision'] | [ 2.07554456e-03 -3.57493460e-01 4.83383417e-01 -3.52752566e-01
-2.30695605e-01 -8.47980797e-01 3.91623914e-01 2.65295655e-01
-1.25682116e+00 7.51367629e-01 -2.46215984e-01 1.35359466e-01
-1.87234916e-02 -6.03289843e-01 -7.88805425e-01 -9.20090437e-01
-7.84582198e-01 5.44360995e-01 9.66481626e-01 -3.37334514e-01
1.02796987e-01 3.27180326e-01 -1.90747142e+00 -3.87361795e-01
4.03272152e-01 5.97507298e-01 6.01213217e-01 9.29801941e-01
6.42265528e-02 1.48163155e-01 -2.98716873e-01 -6.23415262e-02
3.69484603e-01 -1.90896675e-01 -9.95017812e-02 -1.61679611e-01
7.09996104e-01 -5.79686046e-01 3.64045203e-02 1.11354363e+00
7.31751084e-01 1.25753775e-01 2.90054113e-01 -8.24590981e-01
1.53733432e-01 5.71965635e-01 -1.84267774e-01 4.27170545e-01
1.18417526e-02 2.32178390e-01 7.65797794e-01 -5.80233276e-01
4.83707935e-01 9.09418046e-01 1.09681153e+00 8.34723234e-01
-1.03563952e+00 -8.31927538e-01 -1.13776438e-01 -1.43566638e-01
-9.73413944e-01 -6.11311436e-01 3.30064505e-01 -5.66006958e-01
6.59182072e-01 -7.15962127e-02 1.20882845e+00 8.61833572e-01
9.90355313e-02 4.93534476e-01 1.21156824e+00 -6.68561235e-02
4.42743510e-01 -5.16525581e-02 2.96611972e-02 6.15024090e-01
6.71822846e-01 5.46170712e-01 -8.70969057e-01 -4.18473572e-01
4.93557572e-01 2.51572251e-01 -5.11434615e-01 -5.84174693e-01
-1.02090669e+00 7.92215645e-01 3.22049618e-01 -2.47806087e-01
-8.24332982e-02 3.25286746e-01 5.18364549e-01 2.47996554e-01
5.52531183e-01 2.42508784e-01 -5.77573538e-01 -3.69652331e-01
-1.07623422e+00 1.13597900e-01 1.20208383e+00 1.00816715e+00
8.34639668e-01 3.26756001e-01 7.91583955e-01 5.35028636e-01
9.51336801e-01 1.09090436e+00 4.25633013e-01 -8.38491440e-01
-1.01917945e-01 3.55105907e-01 2.68787205e-01 -3.38331372e-01
-7.56300807e-01 -1.05465770e-01 -1.54344112e-01 7.08953559e-01
4.60832864e-01 -4.98728812e-01 -6.99217319e-01 1.27970934e+00
3.09538245e-01 1.46026075e-01 4.18328792e-01 1.02589726e+00
1.10653031e+00 3.85803908e-01 -5.87214082e-02 -1.26393229e-01
1.29417717e+00 -7.17449546e-01 -4.19853806e-01 -4.89501387e-01
6.47868395e-01 -4.33281243e-01 1.75099328e-01 4.65265624e-02
-6.50442839e-01 -1.88923135e-01 -1.02705753e+00 4.43285793e-01
-4.88686770e-01 -1.45103186e-01 5.51292717e-01 6.98690891e-01
-1.30857098e+00 5.53990602e-01 -1.65550268e+00 -8.52767885e-01
3.37855160e-01 3.71096253e-01 -5.31340122e-01 1.64836332e-01
-8.24103713e-01 8.10283184e-01 -1.24498188e-01 3.46079648e-01
-1.42538440e+00 -5.26488304e-01 -1.19722366e+00 -3.81389111e-01
-6.91719055e-02 -2.88085025e-02 1.31172669e+00 -6.94743872e-01
-1.29689252e+00 9.91677344e-01 2.55073696e-01 -6.31007671e-01
5.24475932e-01 -4.26913768e-01 -1.06697142e-01 1.98334098e-01
1.24562003e-01 7.86276340e-01 5.59322476e-01 -1.32962394e+00
-1.16529214e+00 -4.52634901e-01 -7.48597039e-03 2.66371429e-01
-3.13121408e-01 1.57898694e-01 -2.07596824e-01 -4.08503979e-01
8.95494670e-02 -1.16210282e+00 -3.46995860e-01 9.57726598e-01
2.64228463e-01 3.26275289e-01 1.15728605e+00 -3.46197575e-01
4.11785930e-01 -2.02413940e+00 -8.67432356e-02 -2.63677508e-01
-1.97600007e-01 6.33964002e-01 1.35634720e-01 7.35460579e-01
6.13582790e-01 -3.36449236e-01 -4.68281746e-01 -7.19989657e-01
-2.59725153e-01 7.84587860e-01 -4.32420596e-02 1.09774852e+00
-4.14924800e-01 2.01787382e-01 -1.31754136e+00 -5.73105395e-01
2.19476521e-01 3.32519233e-01 -5.49146414e-01 3.91357362e-01
1.72832817e-01 5.71090162e-01 -1.18204758e-01 1.11453962e+00
6.06388211e-01 5.36202133e-01 2.12473586e-01 2.64708459e-01
-1.09592843e+00 -1.59050360e-01 -1.07084429e+00 1.45172179e+00
-2.31994405e-01 1.12357044e+00 9.11917031e-01 -6.73067391e-01
8.88210595e-01 3.43328938e-02 3.82031590e-01 -6.19367920e-02
8.65148529e-02 3.28360170e-01 -2.29330771e-02 -7.18611717e-01
8.05845320e-01 -4.19281125e-01 1.67300373e-01 1.92009900e-02
1.23987131e-01 8.92110914e-02 2.88599014e-01 2.80928109e-02
1.01436794e+00 6.62037909e-01 1.53290272e-01 -8.17298889e-01
-9.62547585e-02 5.35692155e-01 6.65491998e-01 7.58403242e-01
-5.18464327e-01 4.96134251e-01 -3.71991366e-01 -5.95318377e-01
-7.48910129e-01 -8.75854254e-01 -7.13366449e-01 1.30614960e+00
7.39261925e-01 -9.97939780e-02 -3.77194911e-01 -1.91790387e-01
3.22713196e-01 -2.28209525e-01 -7.06839085e-01 3.14894795e-01
-4.65321183e-01 -6.25982225e-01 9.42929864e-01 8.02934051e-01
2.36460835e-01 -1.09426296e+00 -1.49990284e+00 4.53753114e-01
4.08264518e-01 -1.08291948e+00 4.73468229e-02 7.21567810e-01
-1.03656995e+00 -1.07914352e+00 -7.28821874e-01 -9.24012363e-01
8.07639480e-01 6.59001589e-01 6.63999438e-01 4.50873733e-01
-1.69391513e-01 4.44670171e-01 -8.49073768e-01 -5.71526051e-01
-2.33481944e-01 -4.00182784e-01 6.53941691e-01 -3.29717994e-01
6.94422722e-02 -4.65713292e-01 -6.63020670e-01 6.53262317e-01
-6.95503652e-01 -5.22656858e-01 3.69056106e-01 8.35511386e-01
2.66991019e-01 -6.31657839e-01 9.82030705e-02 -5.40096521e-01
-4.35329437e-01 -4.31547105e-01 -1.19328880e+00 -3.09315175e-01
7.51702860e-02 -2.88537741e-01 4.80312526e-01 -3.75195891e-01
-7.43525565e-01 4.52643454e-01 -2.08561376e-01 -1.09914273e-01
-1.79075167e-01 4.77850020e-01 3.09094936e-01 -7.09373236e-01
5.91059685e-01 4.34811562e-01 3.83640379e-01 -6.35352314e-01
-3.13108474e-01 8.56145740e-01 4.65569735e-01 -2.67883241e-01
9.80935037e-01 1.12916398e+00 -5.73671535e-02 -1.43023682e+00
-6.83399975e-01 -9.74329054e-01 -7.02393830e-01 -5.12916446e-01
8.41862679e-01 -1.34367239e+00 -5.41921973e-01 7.49922514e-01
-5.98024726e-01 -8.47898841e-01 -7.48171434e-02 9.17127788e-01
-1.90186366e-01 6.60216272e-01 -4.65160817e-01 -1.00164831e+00
-3.96871001e-01 -1.20512247e+00 1.23280168e+00 6.43227816e-01
3.50298017e-01 -1.15299666e+00 5.95877171e-01 6.68456033e-02
5.63992977e-01 1.01328291e-01 -6.20205045e-01 -7.63052166e-01
-2.76511729e-01 -2.21121773e-01 9.89693552e-02 6.76703602e-02
-8.68667569e-03 3.20027292e-01 -8.22971106e-01 -9.69101250e-01
-5.41110992e-01 -4.42470491e-01 8.68803024e-01 7.05387667e-02
4.24558809e-03 1.25994056e-01 -6.18234873e-01 8.34215641e-01
1.24670541e+00 -7.76730478e-02 1.28009439e-01 6.98855877e-01
4.62895840e-01 7.11807668e-01 9.35020804e-01 8.04720521e-01
5.73400080e-01 7.26084411e-01 9.57839131e-01 1.35606721e-01
9.30254832e-02 7.00943358e-03 6.77168489e-01 6.80857301e-01
-5.39015830e-01 -8.93916711e-02 -8.41374516e-01 8.57620060e-01
-1.60089469e+00 -9.39422369e-01 -6.47949696e-01 2.18536162e+00
4.46076691e-01 -1.11280996e-02 6.67449273e-03 -3.51099849e-01
5.72816312e-01 -1.33074284e-01 -3.56525153e-01 1.00210860e-01
-1.12736061e-01 -1.45426258e-01 1.24465740e+00 2.03495502e-01
-1.37112546e+00 8.91997576e-01 6.04811001e+00 -9.52628255e-02
-9.13300157e-01 -3.24645415e-02 -8.22566628e-01 2.26749390e-01
3.01375777e-01 4.25125390e-01 -1.46776521e+00 3.54507685e-01
7.56669283e-01 3.58725041e-01 -4.58962917e-02 1.11269915e+00
1.85158700e-01 -3.28606635e-01 -7.55468667e-01 4.52778131e-01
8.36592466e-02 -9.36291575e-01 -3.94920737e-01 4.48270887e-01
5.73607743e-01 7.05888808e-01 -7.47877479e-01 1.60579696e-01
4.54499185e-01 -2.18943819e-01 1.05218709e+00 4.36536461e-01
5.08241415e-01 -1.72526836e-01 1.20318544e+00 4.80841488e-01
-1.58013439e+00 2.33790856e-02 -8.36195946e-01 -6.26176119e-01
4.29744571e-01 -1.63384795e-01 -4.10613447e-01 1.32201746e-01
1.49692774e+00 8.16269100e-01 -4.76581663e-01 1.81400168e+00
-1.08127564e-01 7.27336824e-01 -9.02378798e-01 -5.00596404e-01
4.96028185e-01 -2.17944682e-01 8.89567196e-01 1.36611116e+00
4.53229934e-01 1.74977899e-01 4.19093817e-01 -1.74589474e-02
1.59530550e-01 -1.19138643e-01 -6.06808662e-01 2.65763134e-01
5.32712221e-01 1.42058074e+00 -7.98488617e-01 -9.58543718e-02
-6.04695141e-01 3.69199604e-01 1.36795327e-01 -2.88377792e-01
-5.50996840e-01 -2.82967299e-01 1.06086838e+00 6.96957335e-02
4.18308139e-01 -6.80396914e-01 3.36863428e-01 -9.70979929e-01
-5.83727896e-01 -2.01706156e-01 3.22267681e-01 -6.00512564e-01
-1.09783125e+00 4.55063432e-01 5.32297418e-03 -2.16313314e+00
2.40570977e-01 -7.34892428e-01 -6.39580846e-01 3.22499603e-01
-1.87613070e+00 -1.07540512e+00 -8.47615421e-01 5.93411736e-02
5.58238685e-01 -6.04770407e-02 7.56381691e-01 2.40878314e-01
-7.89313018e-02 2.82697678e-01 6.09979928e-01 2.79467106e-01
7.72963047e-01 -1.36855221e+00 1.01084620e-01 1.05224788e+00
2.47486886e-02 3.78501862e-01 1.16259491e+00 -9.07690763e-01
-1.88407362e+00 -1.04178345e+00 -7.79173449e-02 -2.85216987e-01
1.11748195e+00 -2.90922016e-01 -9.56816852e-01 7.98507452e-01
1.11349352e-01 4.68993098e-01 9.27984178e-01 -6.26729965e-01
4.57391500e-01 -6.43351525e-02 -9.67939973e-01 3.04725856e-01
9.52264130e-01 1.80419385e-01 -5.54984152e-01 2.80283809e-01
1.74408436e-01 -7.02707946e-01 -9.37208295e-01 4.19027179e-01
7.62106895e-01 -7.31019735e-01 6.86156750e-01 -2.31838390e-01
-7.31320828e-02 -9.45528507e-01 -1.37389049e-01 -1.24555731e+00
1.21777616e-01 -4.09210652e-01 4.79734510e-01 1.01084900e+00
1.97672039e-01 -5.67841470e-01 8.62302959e-01 -2.10481752e-02
-7.39624679e-01 7.68436864e-02 -1.15929151e+00 -1.07997668e+00
-1.81101680e-01 -3.77414562e-02 -2.64740825e-01 6.45826340e-01
-2.02999674e-02 -3.01545352e-01 -8.36967051e-01 9.90068734e-01
1.32626009e+00 2.33935356e-01 1.12163627e+00 -1.59939289e+00
-8.17924142e-02 1.14880100e-01 -1.01568484e+00 -9.86566842e-01
-1.16072660e-02 -4.00031537e-01 9.90403891e-01 -1.53028631e+00
-1.42994344e-01 -3.02428067e-01 4.90303129e-01 7.94119775e-01
1.92920476e-01 8.17361832e-01 -2.37903073e-01 4.67306942e-01
-6.67224109e-01 6.88616693e-01 7.83610284e-01 1.99267298e-01
5.78504987e-02 1.17261365e-01 7.83041269e-02 9.69653487e-01
4.41601336e-01 -8.16584468e-01 3.71321470e-01 -5.46360075e-01
-1.17044151e-01 -2.10274067e-02 4.09661442e-01 -1.43958247e+00
9.94449675e-01 1.42062470e-01 -9.01109427e-02 -3.15893680e-01
5.79867303e-01 -9.80303347e-01 4.01360765e-02 9.31009412e-01
2.82613605e-01 -1.48351774e-01 2.76578039e-01 8.24623883e-01
-3.12091857e-01 -8.55917335e-01 1.11082125e+00 -2.46109635e-01
-1.39521503e+00 7.63117224e-02 -9.10080373e-01 2.53882091e-02
1.06703663e+00 -4.74983245e-01 -6.10572100e-01 1.91778038e-02
-5.17070055e-01 5.58714092e-01 8.96988332e-01 1.49374604e-01
7.05091834e-01 -4.12407011e-01 -7.67300069e-01 9.08348411e-02
5.06183028e-01 1.04679041e-01 1.35363296e-01 7.39600360e-01
-1.48874474e+00 -4.04102325e-01 -5.25372386e-01 -8.87180984e-01
-1.58756101e+00 -6.45028278e-02 2.83888012e-01 5.90399206e-01
-8.33792567e-01 9.34911370e-01 -8.52797329e-02 -5.61420500e-01
1.91940531e-01 -5.21422699e-02 -3.48041683e-01 2.48738393e-01
6.34071827e-01 3.01551908e-01 -2.39193484e-01 -9.53511834e-01
-6.07183039e-01 8.93486500e-01 3.59748840e-01 1.46336049e-01
1.75241446e+00 -2.66740948e-01 3.52873057e-02 3.34987342e-01
6.29531443e-01 1.68140251e-02 -1.97238040e+00 -2.13418275e-01
-1.03169598e-01 -2.90168285e-01 1.35266498e-01 1.03796786e-02
-1.06708217e+00 7.87072659e-01 9.47828114e-01 3.11646640e-01
4.93421137e-01 1.46673352e-01 4.90430593e-01 6.26502573e-01
9.24300015e-01 -8.26056242e-01 -3.61273378e-01 6.32862151e-01
3.84351909e-01 -1.45010626e+00 3.32833976e-01 8.15764815e-02
-5.99035561e-01 1.24177551e+00 6.33485436e-01 -1.65579721e-01
5.92543960e-01 8.70851099e-01 7.36857593e-01 -2.10208058e-01
-4.66469765e-01 -8.17813814e-01 -4.43451822e-01 7.31219530e-01
7.77183846e-02 -1.87481269e-01 -2.38469467e-01 1.35140970e-01
-2.01552808e-01 -4.63420033e-01 9.60820556e-01 1.76396275e+00
-1.01968694e+00 -7.15900064e-01 -5.50392687e-01 1.47974208e-01
-4.33451146e-01 -2.89894734e-02 7.23020881e-02 8.00071180e-01
-4.53209020e-02 7.89257407e-01 3.09066977e-02 -9.09995064e-02
4.53202754e-01 -4.92837459e-01 2.87897442e-03 -7.18821287e-01
-5.75025678e-01 -8.92600268e-02 2.85251886e-01 -7.45551810e-02
-1.12786412e+00 -1.17158401e+00 -1.27716148e+00 1.01587519e-01
-6.05136514e-01 6.34535909e-01 1.07199621e+00 8.23323250e-01
-2.58542836e-01 8.39552805e-02 3.56965631e-01 -1.82080126e+00
-3.49936217e-01 -1.16993141e+00 -8.78857017e-01 -8.60687196e-02
1.62890881e-01 -1.04426110e+00 -7.77975976e-01 2.42983907e-01] | [8.083379745483398, -1.4233990907669067] |
569528e9-7ab0-403b-a0df-b98cf64484ac | self-supervised-pre-training-for-transformer | 2111.12084 | null | https://arxiv.org/abs/2111.12084v1 | https://arxiv.org/pdf/2111.12084v1.pdf | Self-Supervised Pre-Training for Transformer-Based Person Re-Identification | Transformer-based supervised pre-training achieves great performance in person re-identification (ReID). However, due to the domain gap between ImageNet and ReID datasets, it usually needs a larger pre-training dataset (e.g. ImageNet-21K) to boost the performance because of the strong data fitting ability of the transformer. To address this challenge, this work targets to mitigate the gap between the pre-training and ReID datasets from the perspective of data and model structure, respectively. We first investigate self-supervised learning (SSL) methods with Vision Transformer (ViT) pretrained on unlabelled person images (the LUPerson dataset), and empirically find it significantly surpasses ImageNet supervised pre-training models on ReID tasks. To further reduce the domain gap and accelerate the pre-training, the Catastrophic Forgetting Score (CFS) is proposed to evaluate the gap between pre-training and fine-tuning data. Based on CFS, a subset is selected via sampling relevant data close to the down-stream ReID data and filtering irrelevant data from the pre-training dataset. For the model structure, a ReID-specific module named IBN-based convolution stem (ICS) is proposed to bridge the domain gap by learning more invariant features. Extensive experiments have been conducted to fine-tune the pre-training models under supervised learning, unsupervised domain adaptation (UDA), and unsupervised learning (USL) settings. We successfully downscale the LUPerson dataset to 50% with no performance degradation. Finally, we achieve state-of-the-art performance on Market-1501 and MSMT17. For example, our ViT-S/16 achieves 91.3%/89.9%/89.6% mAP accuracy on Market1501 for supervised/UDA/USL ReID. Codes and models will be released to https://github.com/michuanhaohao/TransReID-SSL. | ['Rong Jin', 'Hao Li', 'Fan Wang', 'Yanxin Zhou', 'Feng Ding', 'Yi Xu', 'Pichao Wang', 'Hao Luo'] | 2021-11-23 | null | null | null | null | ['unsupervised-person-re-identification'] | ['computer-vision'] | [ 8.74700025e-02 -2.41345048e-01 1.96166020e-02 -6.96904898e-01
-4.77770656e-01 -2.94196129e-01 6.08724713e-01 -1.39002010e-01
-7.49747515e-01 6.97973847e-01 1.78815544e-01 -3.62808146e-02
-1.49054542e-01 -7.73867190e-01 -7.68090069e-01 -4.66948986e-01
3.78839433e-01 5.72761774e-01 1.20248996e-01 -2.03170702e-01
-1.02262244e-01 1.72693700e-01 -1.66805863e+00 1.56871706e-01
1.19857621e+00 6.20292127e-01 3.58681738e-01 3.28876376e-01
-2.20951568e-02 4.82125968e-01 -5.58714569e-01 -7.99727261e-01
4.34142023e-01 -2.59128571e-01 -7.94593155e-01 -8.37473795e-02
8.08195293e-01 -3.69115889e-01 -6.21125638e-01 1.12346268e+00
8.30236614e-01 2.44547687e-02 4.89161551e-01 -1.25809228e+00
-8.58974874e-01 6.22925282e-01 -6.40822768e-01 2.10889786e-01
2.95960605e-02 3.19770187e-01 1.55774772e-01 -1.11121166e+00
4.62568462e-01 1.28778911e+00 1.05221140e+00 8.13779891e-01
-1.30871856e+00 -1.20231485e+00 6.70226067e-02 4.21394199e-01
-1.63243723e+00 -6.29700840e-01 5.34707904e-01 -3.65643859e-01
6.68688655e-01 9.92731005e-03 4.85993773e-01 1.38464975e+00
-1.44189432e-01 7.65572131e-01 1.24984145e+00 -3.64983827e-01
2.82983147e-02 3.81698966e-01 4.84594464e-01 4.59257096e-01
2.81468630e-01 3.82733822e-01 -7.62216389e-01 1.43773213e-01
7.94475138e-01 7.62038082e-02 -9.30053443e-02 -1.35050118e-01
-1.04224122e+00 4.33521241e-01 5.92896104e-01 -4.92771119e-02
-1.50893882e-01 -3.63421530e-01 4.66870397e-01 4.59434807e-01
2.61064917e-01 2.64812440e-01 -5.48381925e-01 2.97772195e-02
-8.91131639e-01 7.87921846e-02 3.63296151e-01 1.19515908e+00
9.16292906e-01 -8.97406340e-02 -3.61286938e-01 1.22786522e+00
-1.57298967e-01 6.69953227e-01 5.80057085e-01 -4.64624882e-01
5.03742516e-01 6.93097174e-01 -1.38008252e-01 -4.37388927e-01
-2.98144877e-01 -8.31470490e-01 -1.25951314e+00 -1.64174199e-01
4.83970791e-01 3.66551126e-03 -1.33096445e+00 1.81204784e+00
2.36230910e-01 2.74805725e-01 9.26289558e-02 8.88071656e-01
9.54037607e-01 4.03525561e-01 3.54307383e-01 -2.44901457e-04
1.31068075e+00 -9.48345721e-01 -3.79407227e-01 -2.65242994e-01
4.95669484e-01 -6.02336407e-01 1.25715303e+00 1.28272817e-01
-7.60754168e-01 -1.24436200e+00 -9.23492193e-01 -1.07268251e-01
-5.19875348e-01 5.25172651e-01 2.10804194e-01 6.35284603e-01
-1.00257337e+00 2.94751644e-01 -5.40516973e-01 -7.41536915e-01
4.87772912e-01 5.07689953e-01 -4.36049759e-01 -4.09890622e-01
-1.34000504e+00 6.85135365e-01 4.24942046e-01 -6.77650347e-02
-8.39799225e-01 -1.09960175e+00 -6.74880862e-01 -1.34937033e-01
1.36062473e-01 -8.62476289e-01 9.92765307e-01 -9.32443917e-01
-1.19178092e+00 1.07016420e+00 -1.95425332e-01 -7.54057705e-01
7.63880610e-01 -2.34743550e-01 -6.52334630e-01 -1.79609105e-01
3.51188838e-01 1.08522534e+00 9.66091633e-01 -1.08938086e+00
-8.01527917e-01 -5.07468998e-01 -1.41665131e-01 1.20950334e-01
-7.30600715e-01 -7.74161071e-02 -7.71575689e-01 -8.00185502e-01
-2.82091498e-01 -9.66161966e-01 1.24351412e-01 -4.01351184e-01
-4.39839900e-01 -2.31317416e-01 7.10385799e-01 -7.42848933e-01
1.00902951e+00 -2.21405149e+00 -3.24891657e-01 1.30958945e-01
6.60101175e-02 7.02220619e-01 -4.04976308e-01 1.49112856e-02
-2.82023877e-01 -2.43559867e-01 -1.23717934e-01 -6.52492344e-01
-2.73574591e-01 1.11578405e-01 -1.60113722e-01 3.13016474e-01
4.43472527e-02 1.05127537e+00 -7.27654934e-01 -3.79645467e-01
2.98619479e-01 4.67644632e-01 -2.97738194e-01 1.98139206e-01
2.89453834e-01 6.82032824e-01 -1.00544140e-01 6.43803835e-01
1.02127278e+00 -4.99462672e-02 -1.60839438e-01 -4.04986262e-01
-1.87762961e-01 2.98191570e-02 -1.11580777e+00 1.70966029e+00
-2.46416509e-01 3.48743379e-01 -2.45081261e-01 -9.21178579e-01
1.03805280e+00 -1.69582456e-01 2.99754262e-01 -1.20828176e+00
-7.17880279e-02 -1.54870033e-01 -2.02185363e-01 -2.53230065e-01
5.44413149e-01 2.22086146e-01 -1.54175177e-01 1.45418197e-01
3.29504937e-01 6.89167082e-01 2.52447486e-01 1.80549905e-01
8.12034190e-01 3.00827879e-03 -1.92874402e-01 -4.30525810e-01
6.81640089e-01 1.32764310e-01 7.29888678e-01 1.05086255e+00
-3.06466699e-01 7.39381135e-01 -7.35690519e-02 -4.65775102e-01
-1.15230870e+00 -1.21068025e+00 -3.68031055e-01 1.11880016e+00
3.83462936e-01 -4.06612277e-01 -9.87823248e-01 -6.95945203e-01
2.05440864e-01 6.23117268e-01 -6.12697721e-01 -5.35137415e-01
-5.79549193e-01 -8.85337830e-01 8.08155239e-01 6.14207327e-01
1.25768328e+00 -7.18977213e-01 9.76992622e-02 2.55451519e-02
-1.79038197e-01 -1.07683504e+00 -6.91265345e-01 -2.20217407e-01
-6.25117779e-01 -1.03924334e+00 -8.98891449e-01 -9.12106812e-01
9.64312673e-01 5.07036328e-01 8.64434659e-01 -1.10381059e-01
-2.13100851e-01 4.18046266e-01 -2.05897361e-01 -3.48047644e-01
-2.14855507e-01 3.80975962e-01 5.87059915e-01 1.55511960e-01
6.98409975e-01 -3.63451809e-01 -7.21914530e-01 7.92317271e-01
-6.19099319e-01 3.44590753e-01 6.90194666e-01 1.12922657e+00
6.33302867e-01 2.08633855e-01 7.44778037e-01 -6.91846311e-01
4.28912133e-01 -2.41652831e-01 -5.86771905e-01 5.01740932e-01
-8.76992047e-01 -4.22347486e-02 5.94023705e-01 -7.62221873e-01
-1.34897554e+00 1.02865927e-01 3.82455252e-02 -4.98241693e-01
-2.25428060e-01 1.68521762e-01 -2.47209087e-01 -5.94016239e-02
8.48899186e-01 4.84453231e-01 -4.65523303e-02 -6.94303274e-01
1.33691043e-01 8.71277094e-01 1.10121906e+00 -5.53066850e-01
1.07347524e+00 2.62446672e-01 -5.82468867e-01 -6.14897668e-01
-7.70033300e-01 -4.87058848e-01 -7.19662189e-01 -1.06789738e-01
5.87038517e-01 -1.39309669e+00 -5.52349031e-01 9.35503662e-01
-7.27452695e-01 -3.54623020e-01 -2.43168771e-01 5.47689557e-01
-9.75645016e-05 1.96907505e-01 -6.07525468e-01 -5.81025541e-01
-6.58478498e-01 -9.05117035e-01 6.52647376e-01 6.11692250e-01
-3.19485702e-02 -5.01888871e-01 -2.67980456e-01 6.57637358e-01
5.41322291e-01 -4.05182511e-01 5.20546496e-01 -6.57713592e-01
-3.14460248e-01 -1.49500519e-01 -5.77243567e-01 5.53490877e-01
8.89533982e-02 -6.21283710e-01 -1.14577079e+00 -6.73683167e-01
-2.80103654e-01 -2.71542698e-01 1.11677372e+00 2.79390752e-01
1.13226700e+00 -8.92916918e-02 -3.59545022e-01 8.48866999e-01
1.18055117e+00 -2.27429103e-02 8.09362590e-01 5.25053203e-01
8.31797481e-01 4.92574960e-01 6.70315683e-01 2.81392962e-01
8.34174335e-01 5.82920671e-01 -1.91634268e-01 -1.93340212e-01
-5.98673403e-01 -7.19461858e-01 4.25729036e-01 4.87190843e-01
-1.12801932e-01 2.07710832e-01 -9.28440988e-01 5.88680863e-01
-1.74809790e+00 -8.18516850e-01 -4.85916734e-02 2.60929322e+00
9.26607490e-01 2.21485302e-01 3.26016575e-01 -9.28222463e-02
1.05120790e+00 -3.74955803e-01 -8.78352761e-01 3.06978971e-01
-3.10561150e-01 3.88254784e-02 9.32263434e-01 3.34378481e-01
-1.21166420e+00 1.15116513e+00 5.20496273e+00 1.03058922e+00
-1.09152210e+00 2.14073971e-01 6.55987680e-01 -2.23335922e-02
2.47463197e-01 -1.97389513e-01 -1.41993511e+00 7.74814725e-01
8.75172675e-01 -1.82352751e-01 5.36063373e-01 7.85178244e-01
1.12066157e-01 4.37212782e-03 -9.96643066e-01 1.34180927e+00
4.04840112e-02 -1.14733899e+00 1.52770817e-01 -1.02654167e-01
7.26799071e-01 1.11112840e-01 1.84602857e-01 7.15463698e-01
1.66968927e-01 -8.07699859e-01 6.11077487e-01 6.30623519e-01
1.16652751e+00 -7.78791428e-01 8.64001632e-01 3.44198644e-01
-1.13078701e+00 -2.35918134e-01 -5.85534036e-01 1.23753518e-01
-2.51952857e-01 5.37978113e-01 -8.15685332e-01 6.30357027e-01
1.18381631e+00 7.56165922e-01 -1.12146044e+00 1.11548316e+00
-9.03220847e-02 6.31800950e-01 -3.51281166e-01 4.42481130e-01
-2.96648830e-01 3.26234475e-02 3.63692045e-01 1.20307231e+00
1.35272548e-01 -7.56550208e-02 7.53532499e-02 8.10676336e-01
-1.68106079e-01 -3.12647521e-01 -2.94316947e-01 4.35006440e-01
7.05432951e-01 9.17498112e-01 -1.86175019e-01 -5.57300866e-01
-2.74026424e-01 1.08818042e+00 3.35945427e-01 5.56221128e-01
-9.14685011e-01 -2.98291683e-01 7.16893137e-01 3.04066390e-01
7.71886930e-02 1.11363211e-03 -3.19591701e-01 -1.30615425e+00
7.42067322e-02 -7.63061702e-01 5.40387392e-01 -5.27470171e-01
-1.65685403e+00 5.21361291e-01 1.60722598e-01 -1.22977984e+00
2.13667527e-02 -2.03732148e-01 -3.72368068e-01 1.04110014e+00
-1.58819222e+00 -1.46094465e+00 -6.06152177e-01 9.82160449e-01
5.42037845e-01 -5.29149830e-01 5.70308506e-01 6.42773151e-01
-8.93111289e-01 1.43629074e+00 1.92157254e-01 3.56229872e-01
1.22111595e+00 -9.66864109e-01 6.21812344e-01 1.06445205e+00
-3.46692115e-01 8.39865685e-01 3.64683002e-01 -8.60059619e-01
-1.38597441e+00 -1.52349246e+00 8.55056942e-01 -3.98533523e-01
1.99617490e-01 -5.48435152e-01 -9.82191026e-01 6.07693493e-01
-6.75429627e-02 -1.04626268e-01 4.53147411e-01 9.32235792e-02
-5.59978902e-01 -7.20749855e-01 -1.40920198e+00 4.49441910e-01
1.44922900e+00 -5.05293190e-01 -6.50988638e-01 6.11170903e-02
6.09931171e-01 -3.56645793e-01 -8.86052907e-01 5.23986220e-01
4.56242740e-01 -7.74720073e-01 1.32404709e+00 -2.60274380e-01
-2.34652795e-02 -4.56570536e-01 2.10393444e-01 -1.26685238e+00
-6.08534634e-01 -4.19253498e-01 1.90438837e-01 1.80076146e+00
1.17945626e-01 -8.15001369e-01 7.09222972e-01 7.94260979e-01
5.31544862e-03 -1.76992312e-01 -7.49347448e-01 -1.02013028e+00
-6.01111688e-02 -3.02753150e-01 8.43597174e-01 1.03892231e+00
-4.13906276e-01 4.34167743e-01 -5.07499278e-01 3.00406486e-01
1.09933794e+00 -1.75827146e-01 1.11209130e+00 -1.16213453e+00
9.27453041e-02 -1.67504877e-01 -1.94801971e-01 -1.02945614e+00
-2.96467897e-02 -9.78024244e-01 -2.86783338e-01 -1.18289030e+00
4.05088246e-01 -7.46663392e-01 -5.35546660e-01 7.48173356e-01
-2.56625086e-01 3.06302428e-01 2.38725990e-01 4.86198485e-01
-5.75843453e-01 5.75906932e-01 1.00446451e+00 -4.25612360e-01
-3.70384961e-01 -1.00782253e-01 -7.25266874e-01 3.58994037e-01
8.97367299e-01 -3.08486044e-01 -4.34261858e-01 -5.54782391e-01
-4.36143309e-01 -4.15503502e-01 6.57552004e-01 -1.35537231e+00
6.94708467e-01 2.02648461e-01 9.86111820e-01 -7.60250211e-01
1.59172431e-01 -4.76986706e-01 2.79085368e-01 3.82103860e-01
-3.16520065e-01 -8.45274702e-02 2.44410306e-01 3.72661799e-01
-1.54364817e-02 1.77634165e-01 9.08047974e-01 1.08601183e-01
-1.11879754e+00 5.30097008e-01 8.98370221e-02 -9.10253897e-02
6.66478515e-01 -3.27653021e-01 -6.26247346e-01 3.89925949e-02
-5.69782972e-01 6.22176468e-01 5.52672565e-01 6.30875111e-01
6.48194909e-01 -1.35040116e+00 -9.15894926e-01 6.99733019e-01
4.72424537e-01 -2.12068669e-03 6.52083397e-01 6.71907365e-01
5.16474387e-03 3.29610854e-01 -4.21711475e-01 -6.38055742e-01
-1.24029684e+00 6.03135407e-01 3.26947361e-01 -1.56147853e-01
-6.41344607e-01 7.44884491e-01 3.52237195e-01 -7.58221030e-01
4.68999177e-01 1.63562283e-01 -9.15099159e-02 -1.29991040e-01
8.49039495e-01 4.61031109e-01 -8.08658358e-03 -6.23947263e-01
-4.65407372e-01 4.07460123e-01 -5.42335272e-01 2.38133833e-01
1.15890050e+00 -2.41110206e-01 9.00965631e-02 -2.89175659e-01
1.09756613e+00 -4.35755104e-01 -1.35702693e+00 -6.02030635e-01
-1.14296295e-01 -4.18452233e-01 -1.61960840e-01 -1.02355731e+00
-8.92748237e-01 5.50947070e-01 1.13843679e+00 -3.52218896e-01
1.25829363e+00 -6.12463653e-02 9.10001397e-01 3.57738644e-01
5.03584564e-01 -1.30908215e+00 -9.55667049e-02 6.11183047e-01
7.29550600e-01 -1.30982530e+00 -1.59650832e-01 -2.21501142e-01
-5.78510344e-01 7.64022708e-01 1.02069783e+00 2.32403856e-02
4.37607586e-01 7.59032816e-02 -5.63006289e-02 2.08743632e-01
-3.03044081e-01 -3.58064175e-01 1.89668074e-01 9.33689833e-01
-9.81341377e-02 1.28039286e-01 -5.05275838e-02 8.57359886e-01
-4.70965117e-01 3.85665238e-01 1.28670022e-01 5.63982248e-01
-1.18793920e-02 -1.10015464e+00 -5.81906259e-01 4.95776296e-01
-5.19350953e-02 -1.73084497e-01 -2.13720620e-01 6.83672249e-01
4.38018322e-01 8.63248467e-01 7.52329174e-03 -8.36029828e-01
5.91496825e-01 -3.36192511e-02 4.21664625e-01 -2.82177776e-01
-5.96591294e-01 -1.89028963e-01 -9.20718908e-03 -2.64882535e-01
-2.92743236e-01 -5.50766349e-01 -9.64425325e-01 -6.77186608e-01
-1.09621629e-01 -2.88300887e-02 2.95546800e-01 7.69804597e-01
5.86518645e-01 2.36153498e-01 6.08035445e-01 -4.80563521e-01
-4.88252014e-01 -1.10616136e+00 -3.04765314e-01 5.62010229e-01
1.04723386e-01 -6.03563488e-01 -9.16652083e-02 1.63315073e-01] | [14.781160354614258, 1.0153462886810303] |
1a325582-36c8-43cc-8e89-5cd32fc35ffa | knowledge-transfer-for-melanoma-screening | 1703.07479 | null | http://arxiv.org/abs/1703.07479v1 | http://arxiv.org/pdf/1703.07479v1.pdf | Knowledge Transfer for Melanoma Screening with Deep Learning | Knowledge transfer impacts the performance of deep learning -- the state of
the art for image classification tasks, including automated melanoma screening.
Deep learning's greed for large amounts of training data poses a challenge for
medical tasks, which we can alleviate by recycling knowledge from models
trained on different tasks, in a scheme called transfer learning. Although much
of the best art on automated melanoma screening employs some form of transfer
learning, a systematic evaluation was missing. Here we investigate the presence
of transfer, from which task the transfer is sourced, and the application of
fine tuning (i.e., retraining of the deep learning model after transfer). We
also test the impact of picking deeper (and more expensive) models. Our results
favor deeper models, pre-trained over ImageNet, with fine-tuning, reaching an
AUC of 80.7% and 84.5% for the two skin-lesion datasets evaluated. | ['Flávia Vasques Bittencourt', 'Ramon Pires', 'Afonso Menegola', 'Sandra Avila', 'Eduardo Valle', 'Michel Fornaciali'] | 2017-03-22 | null | null | null | null | ['skin-cancer-classification'] | ['medical'] | [ 4.98285681e-01 3.26332808e-01 -3.00716162e-01 -2.06764594e-01
-9.53242362e-01 -4.55250710e-01 4.95015174e-01 4.11294959e-02
-8.38042080e-01 9.23491955e-01 2.98630446e-01 -5.66894472e-01
-1.27343029e-01 -6.69855356e-01 -9.34710622e-01 -6.81682467e-01
2.53826112e-01 3.44037563e-01 3.47638547e-01 -2.65717655e-01
6.20879903e-02 2.60381162e-01 -9.80611026e-01 8.41701627e-01
8.45555484e-01 7.26176500e-01 -5.08693093e-03 8.14688385e-01
1.05288781e-01 8.67855906e-01 -6.18723750e-01 -8.05852056e-01
1.33012878e-02 -3.24128568e-01 -1.22115719e+00 -6.06109202e-02
3.99673969e-01 -4.78967667e-01 5.55002280e-02 7.69739270e-01
6.76556230e-01 -4.82290328e-01 7.41363525e-01 -7.49656320e-01
-6.82401061e-01 2.97652155e-01 -5.71536779e-01 2.66067803e-01
-9.10821036e-02 4.90946352e-01 4.11345035e-01 -5.04232764e-01
8.21864843e-01 9.14697587e-01 1.42739820e+00 1.07076252e+00
-1.16798043e+00 -5.47727406e-01 -2.99695015e-01 1.45718604e-02
-9.49855685e-01 -3.92292529e-01 -9.66626853e-02 -7.24351287e-01
7.76981235e-01 2.59871811e-01 5.22776425e-01 1.43849003e+00
3.12860698e-01 4.87280846e-01 1.49182057e+00 -4.80674416e-01
-5.00169098e-02 5.73432148e-01 -8.49605538e-03 6.61466300e-01
3.11512232e-01 1.27100706e-01 -3.34833354e-01 -2.61563927e-01
6.48100317e-01 -1.75981179e-01 -2.62622207e-01 2.05791947e-02
-7.94288576e-01 8.44558060e-01 6.31864786e-01 3.24075997e-01
-4.16678190e-01 1.32456109e-01 5.95804572e-01 4.45873231e-01
8.28053057e-01 6.23979092e-01 -7.17707396e-01 1.71844810e-02
-8.30113232e-01 -2.37423465e-01 8.01083267e-01 3.12643379e-01
6.61404192e-01 -6.67958975e-01 -6.26606107e-01 8.60766947e-01
-5.86233810e-02 1.38576254e-02 6.76364660e-01 -3.76983821e-01
1.65499225e-01 6.88160598e-01 -1.97722211e-01 -1.74213469e-01
-5.58224916e-01 -5.64310908e-01 -5.66250920e-01 2.41426960e-01
7.44270205e-01 -7.36429930e-01 -1.35005271e+00 1.49279749e+00
2.17930913e-01 3.13556641e-01 1.16731420e-01 7.19140947e-01
7.74159849e-01 1.67645723e-01 5.90636551e-01 3.21447939e-01
1.25330710e+00 -1.08521461e+00 -2.04310566e-01 -3.03537786e-01
1.23222125e+00 -6.23908162e-01 9.01868105e-01 2.90395856e-01
-9.66649890e-01 -2.99364030e-01 -7.00053453e-01 -1.35496691e-01
-4.53545898e-01 2.14728862e-01 6.40257359e-01 7.76309133e-01
-1.34295905e+00 8.03315043e-01 -7.85522461e-01 -7.84927309e-01
8.20358694e-01 6.19396985e-01 -6.02278531e-01 -2.94442296e-01
-9.96485710e-01 1.22406507e+00 2.25908235e-01 -8.81796852e-02
-1.10271311e+00 -1.19581711e+00 -4.09674495e-01 -1.87927842e-01
9.03271511e-02 -1.09504020e+00 1.26250100e+00 -1.64637089e+00
-1.30874765e+00 1.34615958e+00 1.14449456e-01 -5.39280653e-01
9.02387798e-01 -2.49737322e-01 -5.83718419e-02 2.43379120e-02
7.31557831e-02 9.98633564e-01 8.74943435e-01 -1.01459289e+00
-7.37452388e-01 -3.56684536e-01 6.35049343e-02 -3.08887530e-02
-6.11093283e-01 -3.19129936e-02 -2.79934853e-01 -2.16657326e-01
-8.30293059e-01 -1.11745787e+00 -3.54600966e-01 1.61352947e-01
-3.88597369e-01 -1.52062654e-01 3.83169681e-01 -1.08439338e+00
7.59324551e-01 -2.17220402e+00 -1.73487235e-02 3.98879759e-02
2.53684908e-01 7.04431474e-01 -3.91315907e-01 3.42295885e-01
-1.77762210e-01 4.43079442e-01 -1.39999911e-01 -1.49984553e-01
-4.11882967e-01 -2.80716904e-02 2.14765519e-01 3.49059165e-01
7.26504743e-01 1.18118501e+00 -9.22264814e-01 -3.02171171e-01
-1.16705239e-01 7.07907498e-01 -5.56165278e-01 -2.72498447e-02
-6.19395375e-02 5.83361387e-01 -3.92075747e-01 3.84295970e-01
5.26203573e-01 -7.00018167e-01 -3.18949111e-02 -1.38922185e-01
1.68153141e-02 1.49476439e-01 -2.77288973e-01 1.73814487e+00
-4.34660494e-01 5.96346617e-01 -5.92413209e-02 -7.17129111e-01
4.71646577e-01 2.72788614e-01 3.71634245e-01 -6.24033034e-01
1.39103346e-02 3.07364970e-01 5.40045321e-01 -7.97871053e-01
5.18061146e-02 -2.82391101e-01 3.82869005e-01 2.46457696e-01
2.90396482e-01 3.49163085e-01 -2.42160901e-01 3.52924131e-02
1.72261965e+00 5.26079275e-02 1.11464508e-01 -3.83180529e-01
3.45088065e-01 4.09433931e-01 1.83878139e-01 7.73506463e-01
-1.18038654e-01 4.32527483e-01 5.78133583e-01 -3.68568540e-01
-8.21564376e-01 -6.78566456e-01 -3.90187740e-01 1.31850004e+00
-4.04166609e-01 -5.57691120e-02 -9.93306816e-01 -1.17929161e+00
2.46740967e-01 3.07831585e-01 -1.31917095e+00 -4.20499027e-01
-2.62107670e-01 -1.18763530e+00 8.64712298e-01 4.60201710e-01
3.29750836e-01 -9.85335231e-01 -4.35863763e-01 1.50160007e-02
1.71876643e-02 -7.38960207e-01 -1.86294109e-01 2.33929977e-01
-9.43878651e-01 -1.28767848e+00 -1.27921736e+00 -6.93606794e-01
8.30567896e-01 -9.02833864e-02 1.27579951e+00 3.82240593e-01
-6.19300961e-01 3.62096876e-01 -3.76374543e-01 -8.15429926e-01
-7.31097102e-01 5.23154855e-01 -6.45634770e-01 8.94005820e-02
4.47817743e-01 1.86434053e-02 -8.80959094e-01 1.73532948e-01
-8.16792071e-01 1.45030692e-01 1.28344226e+00 1.11076021e+00
3.50735635e-01 -4.28197652e-01 7.15790153e-01 -1.43845451e+00
7.31100202e-01 -4.94296938e-01 -8.89768153e-02 3.93031806e-01
-4.00480598e-01 -5.61702736e-02 1.09408088e-01 -4.20192122e-01
-1.05016696e+00 -1.26382083e-01 -2.19141796e-01 -1.61024883e-01
-1.40291408e-01 4.98473465e-01 4.46251690e-01 -4.85723495e-01
1.08861554e+00 -2.68229455e-01 2.42305756e-01 -3.60148758e-01
7.63645247e-02 5.43695867e-01 1.17967471e-01 -2.15561286e-01
4.78919774e-01 6.19186699e-01 3.02025639e-02 -5.54893136e-01
-1.06580997e+00 -4.15322334e-01 -5.83930314e-01 2.21093930e-02
1.15886641e+00 -7.87766218e-01 -2.56990910e-01 5.59832990e-01
-1.04517734e+00 -9.17525113e-01 -4.19855118e-01 3.28996837e-01
-1.01838291e-01 -4.20454629e-02 -7.97468364e-01 -3.38258207e-01
-4.59672511e-01 -1.10157466e+00 1.02989960e+00 1.56664640e-01
-2.84407526e-01 -1.39751995e+00 2.93777615e-01 2.76376873e-01
6.16746247e-01 3.20823312e-01 9.43539619e-01 -6.93150938e-01
-1.44580856e-01 -1.16521433e-01 -6.47355795e-01 4.11090702e-01
3.25055152e-01 -3.20678800e-01 -1.46746135e+00 -4.53158975e-01
-4.72493321e-01 -6.55405462e-01 1.41306913e+00 4.69493896e-01
1.28167653e+00 -6.75164908e-02 -8.06228399e-01 5.79253495e-01
1.43023908e+00 -2.51760334e-01 9.06045258e-01 3.93926233e-01
4.10712004e-01 8.70057464e-01 2.63737023e-01 -1.33052394e-01
1.49183095e-01 3.06673706e-01 3.54164898e-01 -7.65660226e-01
-6.03135645e-01 -1.16109900e-01 1.24262214e-01 2.18482330e-01
-3.15548837e-01 3.05604786e-02 -1.09501660e+00 6.15578055e-01
-1.61698198e+00 -5.26757479e-01 4.70190011e-02 2.28214741e+00
1.23403275e+00 1.10848956e-01 1.05084613e-01 -2.28916004e-01
5.86402357e-01 -4.99710172e-01 -6.28938735e-01 -4.14043367e-01
2.00094268e-01 6.59464300e-01 7.21983075e-01 4.08193648e-01
-1.02832818e+00 8.05062115e-01 6.38887358e+00 8.16776156e-01
-1.40596259e+00 5.46006382e-01 9.30314183e-01 1.18781112e-01
9.90828685e-03 -1.22978993e-01 -7.57736921e-01 3.95839661e-01
1.20047796e+00 3.24983984e-01 5.58409318e-02 2.85856992e-01
-9.93438587e-02 -1.05682865e-01 -1.30762506e+00 5.37388742e-01
3.38957980e-02 -1.31523442e+00 1.54925901e-02 3.16112518e-01
9.02568996e-01 4.03508127e-01 2.35686362e-01 4.72947747e-01
3.62662017e-01 -1.28489530e+00 -8.71920288e-02 7.41241813e-01
1.04260516e+00 -2.88934380e-01 1.05714166e+00 8.08663070e-02
-3.00304890e-01 2.86646411e-02 -1.67259380e-01 1.87682346e-01
-3.88564974e-01 6.43716693e-01 -1.48343468e+00 3.15162897e-01
7.60717690e-01 6.24475002e-01 -1.10289097e+00 1.27454138e+00
-8.34583640e-02 9.63218629e-01 1.63081750e-01 2.02075448e-02
4.47067469e-01 3.59904975e-01 -9.08379722e-03 1.56509614e+00
1.69878349e-01 -4.04491246e-01 -1.89165637e-01 6.12596929e-01
-2.60416269e-01 -1.71869504e-03 -5.43754339e-01 8.64101499e-02
-1.63266640e-02 1.38275671e+00 -3.45958322e-01 -1.38039500e-01
-4.61564839e-01 9.83081222e-01 4.36451346e-01 2.10042492e-01
-6.43113852e-01 -1.34993270e-01 2.51617610e-01 4.11713600e-01
1.11042917e-01 5.98890960e-01 -2.10714683e-01 -6.37183249e-01
-2.91381866e-01 -8.02817285e-01 6.05688989e-01 -6.56594694e-01
-1.65561175e+00 3.23294550e-01 -4.14554000e-01 -8.08419526e-01
-8.56388584e-02 -9.29639399e-01 -6.96099937e-01 1.01829755e+00
-1.96422148e+00 -1.38581312e+00 -4.66321588e-01 5.96144676e-01
1.06829412e-01 5.17122680e-03 9.66447175e-01 2.83572912e-01
-3.51057261e-01 9.10800934e-01 4.33082990e-02 2.96363175e-01
1.25756407e+00 -1.30816948e+00 9.59125012e-02 1.68974042e-01
-4.71707076e-01 2.05490530e-01 1.46214634e-01 -5.68121374e-01
-1.12219286e+00 -1.42057347e+00 7.05388069e-01 -7.14153469e-01
7.39230871e-01 1.57899838e-02 -1.03981924e+00 6.24273419e-01
2.78331339e-01 7.37049803e-02 1.00498807e+00 4.97046143e-01
-3.19482923e-01 -5.02914526e-02 -1.23713970e+00 3.89308214e-01
9.40925539e-01 -4.43946868e-01 -2.10136160e-01 4.68462467e-01
5.56364596e-01 -3.29980642e-01 -1.01626599e+00 5.51195145e-01
6.26885831e-01 -8.67427945e-01 8.01983893e-01 -1.07294035e+00
8.25707853e-01 2.96812177e-01 3.39706123e-01 -1.55412757e+00
-4.06454146e-01 -2.13212192e-01 4.86565828e-01 8.69081974e-01
9.02738333e-01 -7.84387052e-01 1.00183463e+00 5.25172114e-01
-2.10920990e-01 -9.21034694e-01 -7.96083868e-01 -4.73276824e-01
5.02963006e-01 2.16271475e-01 7.60509819e-02 1.13413620e+00
-2.07392856e-01 3.06186944e-01 -8.88458267e-02 -1.76872555e-02
3.06596041e-01 -5.52805960e-01 7.59830773e-01 -1.23199379e+00
-4.67683703e-01 -4.18122649e-01 -3.66287351e-01 -2.02612817e-01
-3.74734914e-03 -1.14595771e+00 -1.87094569e-01 -1.57045352e+00
6.00878835e-01 -4.36609507e-01 -5.33324182e-01 8.62210631e-01
-4.35228378e-01 4.48740065e-01 -1.37658585e-02 2.95007795e-01
-5.50289333e-01 -2.04553649e-01 1.48660719e+00 -3.36353213e-01
-1.53808877e-01 -1.43490195e-01 -9.58119392e-01 7.22835720e-01
9.35272396e-01 -5.87415397e-01 -6.71376288e-02 -7.38597333e-01
1.11898951e-01 -1.77219570e-01 6.83921278e-01 -9.74763632e-01
1.08086251e-01 1.46036237e-01 7.12385416e-01 1.34872526e-01
2.18135729e-01 -5.31563044e-01 1.66780986e-02 1.05386841e+00
-4.28864837e-01 -3.40225548e-01 6.73783600e-01 3.43989879e-01
6.17225952e-02 -4.44830984e-01 8.68651032e-01 -2.65272826e-01
-4.80824411e-01 1.32735655e-01 -1.34585917e-01 1.16515765e-02
9.81709003e-01 -5.63864261e-02 -7.19405532e-01 6.83745444e-02
-9.27363217e-01 5.28978407e-02 3.32399428e-01 2.46415138e-01
2.17711911e-01 -9.25551653e-01 -1.11610401e+00 -8.34902283e-03
3.37694436e-01 -1.32669017e-01 2.57645398e-01 1.14949143e+00
-3.71758133e-01 2.86336422e-01 -3.53130817e-01 -6.51656806e-01
-1.04533744e+00 2.04142347e-01 8.23541343e-01 -7.73696184e-01
-3.22467983e-01 1.04549360e+00 3.90647173e-01 -4.51276720e-01
9.20947492e-02 -1.53129250e-01 -8.81380066e-02 -2.64699776e-02
4.23603356e-01 3.27245831e-01 4.63416934e-01 1.02140784e-01
-2.44713724e-01 3.88448924e-01 -6.76131427e-01 1.40129477e-01
1.14316833e+00 5.29567719e-01 -1.90154746e-01 8.30400139e-02
1.10542667e+00 -2.82583386e-01 -1.32281399e+00 -5.48334420e-02
1.82698108e-02 -1.09386876e-01 9.46812630e-02 -1.56630063e+00
-8.35092306e-01 9.67713058e-01 9.97981131e-01 -1.38290063e-01
9.60684121e-01 -4.21501957e-02 4.57366318e-01 3.19667429e-01
1.04365662e-01 -8.78758967e-01 2.20549673e-01 2.39752054e-01
7.88696647e-01 -1.46736991e+00 -8.95159021e-02 -2.57336825e-01
-4.22236830e-01 8.03555489e-01 7.25176573e-01 -1.30556285e-01
6.21836722e-01 2.96763331e-01 1.97173014e-01 -2.19083801e-01
-7.13592052e-01 -2.84049362e-01 3.44799072e-01 7.28231132e-01
6.13890886e-01 7.30069429e-02 -2.80119240e-01 5.17638922e-01
2.39470527e-01 5.59165359e-01 3.24616015e-01 8.95101845e-01
-3.35922211e-01 -1.16715884e+00 -2.00554281e-01 9.80837107e-01
-8.04298759e-01 -2.50347286e-01 -8.31084073e-01 8.92117798e-01
4.02393579e-01 7.02381551e-01 -5.45920506e-02 -2.38174796e-01
2.13964432e-01 8.01504217e-03 5.69149256e-01 -9.05028820e-01
-1.10152209e+00 -2.69444734e-01 2.17535526e-01 -3.93561423e-01
-4.96831656e-01 -3.77059191e-01 -6.36850178e-01 -4.17341217e-02
-4.77234542e-01 1.16854325e-01 6.39013767e-01 8.00196826e-01
4.41355020e-01 7.25044966e-01 1.67400211e-01 -6.01815462e-01
-5.85186362e-01 -1.17057896e+00 -2.62272865e-01 4.79273587e-01
4.12193924e-01 -3.49839538e-01 -1.68412492e-01 8.10569152e-02] | [15.339889526367188, -2.72599458694458] |
2bed9f3d-82e1-447e-bf4d-4dc40094c4a3 | self-attentive-model-for-headline-generation | 1901.07786 | null | http://arxiv.org/abs/1901.07786v1 | http://arxiv.org/pdf/1901.07786v1.pdf | Self-Attentive Model for Headline Generation | Headline generation is a special type of text summarization task. While the
amount of available training data for this task is almost unlimited, it still
remains challenging, as learning to generate headlines for news articles
implies that the model has strong reasoning about natural language. To overcome
this issue, we applied recent Universal Transformer architecture paired with
byte-pair encoding technique and achieved new state-of-the-art results on the
New York Times Annotated corpus with ROUGE-L F1-score 24.84 and ROUGE-2
F1-score 13.48. We also present the new RIA corpus and reach ROUGE-L F1-score
36.81 and ROUGE-2 F1-score 22.15 on it. | ['Pavel Kalaidin', 'Valentin Malykh', 'Daniil Gavrilov'] | 2019-01-23 | null | null | null | null | ['headline-generation'] | ['natural-language-processing'] | [ 2.00179979e-01 7.64513433e-01 -3.32612813e-01 -2.84741312e-01
-1.45618677e+00 -6.95646584e-01 9.19128716e-01 3.54199767e-01
-3.25312138e-01 1.44970834e+00 1.09329724e+00 -3.15628260e-01
2.13831991e-01 -6.69349372e-01 -8.17792892e-01 -1.54574364e-01
-1.09731205e-01 7.10360050e-01 2.85114139e-01 -8.64940464e-01
5.98476350e-01 -1.24377400e-01 -1.14965427e+00 7.72741497e-01
9.40814495e-01 5.97827852e-01 1.31601736e-01 1.34550905e+00
3.04181632e-02 1.10985959e+00 -9.24370944e-01 -6.68432415e-01
-1.90865040e-01 -6.87639236e-01 -1.28403580e+00 -3.77544612e-01
6.83901489e-01 -2.11960852e-01 -4.35004920e-01 9.73399460e-01
6.73839509e-01 1.76474810e-01 8.83826196e-01 -8.41243446e-01
-6.30101919e-01 1.13449442e+00 -3.98544580e-01 6.38939917e-01
8.12562227e-01 -2.46260375e-01 1.57302344e+00 -4.67741102e-01
1.04050219e+00 1.05820060e+00 5.01537681e-01 6.76207304e-01
-7.80544639e-01 -3.17006886e-01 -2.00788856e-01 9.59330276e-02
-4.88673091e-01 -3.57324213e-01 4.49134529e-01 -9.34641343e-04
1.61543500e+00 3.65437508e-01 3.98306072e-01 1.15364766e+00
5.53649426e-01 1.22872722e+00 7.45836973e-01 -4.56252694e-01
-2.18813583e-01 -9.69185531e-02 3.66149813e-01 5.71438134e-01
1.62306316e-02 -3.30472469e-01 -5.31727314e-01 1.44624859e-01
2.82663226e-01 -6.34131193e-01 -3.23222518e-01 7.45023429e-01
-1.42200410e+00 1.06103063e+00 2.56723791e-01 4.15508240e-01
-4.19945031e-01 1.46389594e-02 9.92258012e-01 5.99512339e-01
5.59409559e-01 7.77682185e-01 -5.14594913e-01 -7.04204798e-01
-1.16366637e+00 7.14524388e-01 9.92647707e-01 1.16026962e+00
2.46911749e-01 2.34665528e-01 -7.95226276e-01 9.42694724e-01
-3.53543282e-01 5.89064837e-01 9.02284622e-01 -7.63185620e-01
9.11106169e-01 3.92127633e-01 1.59008667e-01 -6.93542659e-01
-5.10034263e-01 -7.09825754e-01 -1.02343178e+00 -6.00873888e-01
1.09367654e-01 -3.48690361e-01 -6.83716834e-01 1.62512410e+00
-2.45182604e-01 -4.14600551e-01 5.31422675e-01 4.37757581e-01
1.23666763e+00 1.25415075e+00 -4.18390512e-01 -5.20387292e-01
1.41916966e+00 -1.49874556e+00 -7.95860887e-01 -3.16824228e-01
7.78425932e-01 -1.03706884e+00 1.20374024e+00 3.38318169e-01
-1.41389430e+00 -4.77476031e-01 -1.01833391e+00 -4.27119493e-01
-5.25468774e-02 2.10062787e-01 4.00820047e-01 4.10245001e-01
-9.84407365e-01 5.56211233e-01 -1.96583971e-01 -4.33171242e-01
2.21499681e-01 -9.05409977e-02 -3.63104552e-01 1.55038357e-01
-1.47302628e+00 1.00706804e+00 6.27787471e-01 -5.49936175e-01
-7.56447792e-01 -5.78648865e-01 -6.21045113e-01 1.55403942e-01
-3.14437039e-02 -6.21660590e-01 1.90537715e+00 -4.98373002e-01
-1.61252558e+00 6.78364992e-01 -4.51664738e-02 -1.16596198e+00
6.47964478e-01 -6.48049653e-01 -6.16412222e-01 2.64732540e-01
3.22334200e-01 6.06488585e-01 4.53900456e-01 -7.14706242e-01
-1.00893092e+00 6.89314082e-02 1.08329795e-01 3.72476697e-01
-1.98555216e-01 6.56447411e-02 4.26444918e-01 -7.84417689e-01
-5.64769804e-01 -6.50780141e-01 5.45926616e-02 -1.18108320e+00
-6.27373576e-01 -4.78562623e-01 6.59381926e-01 -1.00628173e+00
1.56715643e+00 -1.42918777e+00 -1.40352368e-01 -4.70586807e-01
-9.53501463e-02 4.35611308e-01 -2.33067200e-01 7.92973161e-01
7.03638047e-02 -1.14153093e-02 -4.69896831e-02 2.20293496e-02
1.22166239e-01 -1.49199620e-01 -9.15808737e-01 9.03782099e-02
2.50884235e-01 1.06797683e+00 -1.14760351e+00 -4.12012070e-01
-2.36979142e-01 -1.49905430e-02 -7.32616246e-01 3.07932228e-01
-5.58700681e-01 -3.40193906e-03 -3.31237257e-01 1.49767816e-01
2.36745879e-01 -1.39325401e-02 -2.93803900e-01 -1.79561297e-03
-1.39251664e-01 8.30742955e-01 -5.18377364e-01 1.69700742e+00
-6.09005034e-01 9.38874841e-01 -6.35582089e-01 -9.77009296e-01
1.10003769e+00 4.83963937e-01 1.46579221e-01 -8.41054082e-01
1.53746128e-01 2.23384246e-01 -9.61991847e-02 -5.31493247e-01
1.47360909e+00 -2.92987317e-01 -5.17504871e-01 7.61314929e-01
3.79030854e-01 -3.25382352e-01 7.54618943e-01 5.88159204e-01
1.26731420e+00 9.83439106e-03 4.39279169e-01 -3.41164052e-01
6.02948546e-01 2.63134271e-01 1.76750228e-01 8.29542398e-01
2.10235164e-01 7.00520456e-01 6.02288485e-01 -3.17406833e-01
-1.56626785e+00 -5.91885328e-01 -1.63284484e-02 1.16867518e+00
-4.48594451e-01 -6.65388465e-01 -9.00374472e-01 -7.88925350e-01
-5.29402256e-01 1.54426217e+00 -4.95379388e-01 -1.96056038e-01
-9.88004565e-01 -6.11199617e-01 9.69520569e-01 3.62519264e-01
5.65326631e-01 -1.27625978e+00 -6.43778443e-01 6.02252364e-01
-8.91639829e-01 -1.14181221e+00 -6.97874606e-01 -1.13958322e-01
-9.25030708e-01 -7.47189939e-01 -1.06827366e+00 -7.95467436e-01
1.84054643e-01 -6.33073524e-02 1.55883729e+00 -3.93648773e-01
1.36712566e-01 -1.38665289e-01 -7.42018521e-01 -4.47946459e-01
-8.31210434e-01 7.14271426e-01 -3.43595952e-01 -5.75712800e-01
1.78013504e-01 -1.18790768e-01 -3.03091586e-01 -2.65845567e-01
-8.90649796e-01 3.06506991e-01 5.12665391e-01 1.13351202e+00
1.42076269e-01 -2.02546775e-01 1.19710886e+00 -1.11070192e+00
1.20364809e+00 -2.76985854e-01 -1.56011553e-02 2.48113155e-01
-3.93546075e-01 2.26132631e-01 1.12916803e+00 -2.51018226e-01
-1.15596092e+00 -6.56845987e-01 -4.42052394e-01 4.88803208e-01
7.56984875e-02 6.74145401e-01 3.53294641e-01 1.02292275e+00
1.01352620e+00 4.57622617e-01 -4.08583611e-01 -2.55439550e-01
5.93787611e-01 1.01539648e+00 9.13673937e-01 -2.44543865e-01
2.71801680e-01 -1.03544563e-01 -3.64787191e-01 -9.96017873e-01
-1.34056604e+00 -3.77391219e-01 -1.02563471e-01 -9.48056951e-02
5.66455483e-01 -8.39479387e-01 -3.07252198e-01 9.47467014e-02
-1.37030351e+00 -2.25803420e-01 -4.65762228e-01 2.68454492e-01
-8.35558772e-01 4.82457370e-01 -6.84658229e-01 -4.38537002e-01
-1.26784337e+00 -6.91581666e-01 9.42792892e-01 1.39204606e-01
-7.81895220e-01 -8.44891131e-01 3.58592480e-01 4.55120027e-01
4.16579276e-01 2.51096845e-01 1.00427282e+00 -9.88926589e-01
1.32223085e-01 -5.62227964e-01 -1.80100486e-01 4.04565692e-01
-1.65908262e-01 -1.93482265e-01 -7.02481568e-01 -2.86738753e-01
-9.83650461e-02 -6.26998782e-01 8.45641613e-01 3.95491093e-01
5.21241426e-01 -9.04259562e-01 5.51980510e-02 -6.16481341e-02
1.01325941e+00 -5.99903613e-02 8.80729139e-01 5.08577764e-01
2.10362390e-01 4.23888028e-01 6.05528891e-01 5.74396372e-01
5.26498377e-01 4.60508823e-01 -1.76680684e-02 4.24227148e-01
-4.75819916e-01 -7.00735271e-01 6.46844566e-01 1.27312195e+00
-2.60949265e-02 -7.21050441e-01 -5.71342349e-01 6.65967584e-01
-1.77677333e+00 -1.44922936e+00 -3.44865352e-01 1.82742071e+00
1.21337950e+00 5.01506984e-01 2.95394629e-01 1.65778667e-01
6.50433779e-01 4.57776576e-01 -1.20530367e-01 -8.48477960e-01
-3.88325959e-01 1.21040933e-01 3.00947249e-01 6.72141433e-01
-8.27418029e-01 1.16660380e+00 6.40055561e+00 9.62094545e-01
-9.76091027e-01 1.68607738e-02 5.09467125e-01 6.62650615e-02
-4.71828312e-01 -5.80962822e-02 -9.41017032e-01 5.01292109e-01
1.47918975e+00 -9.24816608e-01 6.24064654e-02 6.38340592e-01
-3.17593552e-02 -5.01454622e-02 -1.05603552e+00 7.06946909e-01
2.85612524e-01 -1.56190944e+00 3.96591783e-01 -4.41084504e-01
1.06149447e+00 1.98872373e-01 -3.22350353e-01 7.05040336e-01
4.99911904e-01 -9.94960308e-01 7.31545091e-01 4.01134759e-01
9.59166884e-01 -8.05741727e-01 9.42630231e-01 6.09755874e-01
-5.66059709e-01 1.98771954e-01 -5.22897959e-01 -1.88772693e-01
5.48001766e-01 7.40902603e-01 -1.23763227e+00 6.78808570e-01
3.13354403e-01 8.36139917e-01 -3.50049615e-01 7.86068261e-01
-3.60679209e-01 7.18001783e-01 -2.02065483e-01 -6.24863803e-01
5.34073174e-01 2.74261177e-01 8.23796988e-01 1.72103131e+00
5.17462790e-01 -4.72309999e-02 -1.02575071e-01 1.53339311e-01
-4.70732689e-01 2.66350418e-01 -5.41408181e-01 -1.60248086e-01
1.02402337e-01 9.99841928e-01 -3.36672574e-01 -6.57362342e-01
3.57280001e-02 9.89784241e-01 2.62933999e-01 -1.08214552e-02
-7.65308440e-01 -8.03955853e-01 -1.10601224e-01 3.51194516e-02
1.75933659e-01 5.58790825e-02 8.48544165e-02 -1.41440868e+00
-1.03748821e-01 -1.07806313e+00 4.57575023e-01 -7.99971104e-01
-9.38347578e-01 1.00954640e+00 4.53230739e-02 -1.19536400e+00
-8.58943045e-01 -1.75563127e-01 -6.76413774e-01 3.84362668e-01
-1.52592850e+00 -1.09045708e+00 7.11223334e-02 1.09231062e-01
1.28120387e+00 -7.05716431e-01 9.81708765e-01 1.50732771e-02
-6.97137341e-02 7.28495002e-01 3.36223871e-01 -2.01639123e-02
7.28246331e-01 -1.31813645e+00 6.13240480e-01 7.86945522e-01
6.47946149e-02 2.16994941e-01 1.25037980e+00 -6.07120991e-01
-1.14237356e+00 -9.94018555e-01 1.48446524e+00 -1.93754151e-01
7.40718186e-01 -6.31936407e-03 -6.15751803e-01 7.56465673e-01
8.51558506e-01 -7.61012375e-01 4.68968004e-01 -4.04592529e-02
-3.05619150e-01 3.28364298e-02 -1.18851388e+00 6.70435488e-01
7.37459838e-01 -2.60291398e-01 -1.21697652e+00 7.32303798e-01
9.57919300e-01 -5.34045696e-01 -9.45766807e-01 1.27257660e-01
3.93119603e-01 -8.54925513e-01 5.56204319e-01 -8.67706895e-01
1.15838432e+00 8.69802758e-02 -1.95260599e-01 -1.65950060e+00
-3.67736399e-01 -1.03261530e+00 -1.59451753e-01 1.32649791e+00
8.22892427e-01 -4.69235420e-01 5.42536795e-01 -6.14649914e-02
-5.81015944e-01 -4.60306287e-01 -8.58330607e-01 -6.21376157e-01
5.14843047e-01 -6.37533218e-02 3.32329243e-01 7.15165377e-01
5.06781936e-01 1.35175347e+00 -5.72553575e-01 -5.22544265e-01
3.29100490e-01 1.83419511e-01 8.53759587e-01 -8.42333734e-01
-1.68132320e-01 -5.36177695e-01 1.05269231e-01 -1.26917422e+00
2.65738845e-01 -1.19953346e+00 3.22041623e-02 -1.99519610e+00
1.56218499e-01 1.22677334e-01 2.69260913e-01 1.84628680e-01
-1.97933149e-02 5.27885603e-03 8.39069337e-02 -1.17941694e-02
-6.79014027e-01 7.11395681e-01 1.36730814e+00 -3.42739135e-01
5.51096611e-02 -9.07329470e-02 -8.20712507e-01 3.54075968e-01
1.14128375e+00 -3.52995396e-01 -4.39193368e-01 -4.78000194e-01
4.55355436e-01 4.51775014e-01 -3.03863645e-01 -9.36201274e-01
1.72427878e-01 9.96592566e-02 2.76784226e-02 -1.00290298e+00
5.02090566e-02 -1.44840404e-01 -2.77293712e-01 5.20268917e-01
-9.16388154e-01 4.45942730e-01 5.39342053e-02 3.04270536e-01
-5.15923977e-01 -6.16066158e-01 5.65928340e-01 -3.29730988e-01
-4.01874125e-01 -2.33914897e-01 -4.70522404e-01 8.75035524e-01
7.78809726e-01 -2.03323644e-02 -9.30657148e-01 -8.42769861e-01
-1.27236724e-01 2.10298523e-01 8.97376612e-02 5.96667051e-01
5.31377733e-01 -1.12141466e+00 -1.53036582e+00 -8.45542625e-02
5.69758303e-02 -7.57762939e-02 2.00125098e-01 4.26289827e-01
-7.56422043e-01 8.09140861e-01 -3.62204194e-01 -1.11809410e-01
-1.05130839e+00 2.01516688e-01 -9.51746851e-02 -8.77367973e-01
-8.28155935e-01 6.96688354e-01 -3.46785426e-01 -2.93910950e-01
-1.25846907e-01 -2.15490490e-01 -5.48224509e-01 3.18867862e-01
9.06044900e-01 7.19678879e-01 2.80819952e-01 -6.59933567e-01
1.18018515e-01 1.55556113e-01 -6.26314163e-01 -3.07997853e-01
1.45688546e+00 -8.49959999e-03 -1.72266364e-01 3.74905050e-01
1.27287412e+00 -9.13724378e-02 -5.05722761e-01 -9.30530503e-02
1.28800169e-01 -6.52841851e-02 -1.57893404e-01 -8.69937658e-01
-4.39425766e-01 7.87415743e-01 -5.17404862e-02 4.52345788e-01
7.46545613e-01 -1.72072828e-01 1.38187480e+00 6.71640098e-01
6.52225316e-02 -1.46645367e+00 2.81131595e-01 1.32468081e+00
1.23682106e+00 -9.38234985e-01 1.82030126e-01 9.93411914e-02
-9.86169040e-01 1.21720588e+00 2.45449632e-01 -4.43090975e-01
8.73256698e-02 1.56429887e-03 -1.37698874e-01 -3.90650481e-02
-1.13823736e+00 3.87206137e-01 1.50167748e-01 3.46083552e-01
1.03015006e+00 1.28693283e-01 -6.98882639e-01 5.43411374e-01
-1.23509765e+00 5.32901436e-02 1.11259580e+00 5.09842515e-01
-7.47216582e-01 -7.18565464e-01 3.36537510e-02 6.96056724e-01
-8.75314355e-01 -2.19120130e-01 -4.97745946e-02 4.27268565e-01
-7.77662456e-01 9.78891194e-01 -8.27071220e-02 -2.00992361e-01
3.78948659e-01 6.52817413e-02 4.73248810e-01 -5.61446488e-01
-7.78558373e-01 -8.64285380e-02 9.04721081e-01 7.49763474e-02
-3.00506085e-01 -5.17807126e-01 -1.24021673e+00 -4.60081995e-01
-2.57663369e-01 6.07814670e-01 3.90781015e-01 7.20168650e-01
1.03739522e-01 7.09300160e-01 5.64248443e-01 -5.49625278e-01
-9.25162256e-01 -1.62982333e+00 -2.91920394e-01 4.93306220e-01
5.16437352e-01 9.06709135e-02 1.14501333e-02 2.68435299e-01] | [12.402246475219727, 9.443326950073242] |
e1662d53-738a-4023-b37e-e02bce173907 | tensorformer-normalized-matrix-attention | 2306.15989 | null | https://arxiv.org/abs/2306.15989v1 | https://arxiv.org/pdf/2306.15989v1.pdf | Tensorformer: Normalized Matrix Attention Transformer for High-quality Point Cloud Reconstruction | Surface reconstruction from raw point clouds has been studied for decades in the computer graphics community, which is highly demanded by modeling and rendering applications nowadays. Classic solutions, such as Poisson surface reconstruction, require point normals as extra input to perform reasonable results. Modern transformer-based methods can work without normals, while the results are less fine-grained due to limited encoding performance in local fusion from discrete points. We introduce a novel normalized matrix attention transformer (Tensorformer) to perform high-quality reconstruction. The proposed matrix attention allows for simultaneous point-wise and channel-wise message passing, while the previous vector attention loses neighbor point information across different channels. It brings more degree of freedom in feature learning and thus facilitates better modeling of local geometries. Our method achieves state-of-the-art on two commonly used datasets, ShapeNetCore and ABC, and attains 4% improvements on IOU on ShapeNet. Our implementation will be released upon acceptance. | ['Kai Xu', 'Chenyang Zhu', 'Renjiao Yi', 'Zheng Qin', 'Hui Tian'] | 2023-06-28 | null | null | null | null | ['point-cloud-reconstruction'] | ['computer-vision'] | [ 9.19769928e-02 -3.00215751e-01 9.68531594e-02 -3.18338752e-01
-8.95523548e-01 -2.19896361e-01 5.23100138e-01 3.71148109e-01
-3.00226837e-01 3.76025021e-01 1.41500784e-02 -2.94210821e-01
3.14772762e-02 -1.25689542e+00 -1.09245539e+00 -5.95396578e-01
-2.92342268e-02 5.99003077e-01 2.38098100e-01 -1.38805196e-01
3.62096131e-01 9.50171590e-01 -1.68372059e+00 4.90787715e-01
7.56565094e-01 1.33948147e+00 2.55603492e-01 4.81324255e-01
-6.16310835e-01 2.93408304e-01 -9.46157053e-02 -2.54518211e-01
2.19808310e-01 4.07801569e-01 -6.20259404e-01 -1.60215154e-01
8.65563571e-01 -3.02453011e-01 -3.41347486e-01 9.20249641e-01
4.64694411e-01 4.32529636e-02 5.12202382e-01 -8.63878965e-01
-4.74526525e-01 1.45482048e-01 -7.83065736e-01 -1.89965039e-01
1.21143676e-01 7.34055042e-02 1.07494438e+00 -1.39443231e+00
4.34172630e-01 1.49842536e+00 7.01085925e-01 7.53660128e-02
-1.35925388e+00 -9.48195517e-01 1.41423732e-01 2.49680072e-01
-1.73787093e+00 -3.88272613e-01 9.11844969e-01 -2.55015761e-01
1.21269965e+00 6.28293276e-01 8.20853770e-01 6.26839221e-01
6.07025772e-02 6.51543081e-01 6.68487132e-01 3.13421004e-02
2.53250331e-01 -3.04298729e-01 -2.54820436e-01 4.44275498e-01
1.70940340e-01 -6.13083728e-02 -6.72127783e-01 -4.24382567e-01
1.52971244e+00 3.10257584e-01 -3.60481530e-01 -4.50814039e-01
-1.45272529e+00 9.29946899e-01 8.99965584e-01 1.10516772e-02
-5.64641774e-01 6.29620194e-01 2.10450754e-01 2.26252601e-01
6.97032213e-01 2.21160889e-01 -4.90803570e-01 -1.89739272e-01
-8.29897225e-01 4.05549198e-01 5.34885406e-01 1.09564006e+00
9.82958138e-01 7.25786528e-03 -2.51033064e-02 7.97058165e-01
6.06579900e-01 9.54957485e-01 -1.15497842e-01 -9.65640068e-01
5.01819789e-01 6.96761072e-01 -1.84897780e-02 -1.39653051e+00
-3.74928057e-01 -4.62579042e-01 -1.24023592e+00 3.81943315e-01
-1.02727719e-01 4.19065714e-01 -8.40166032e-01 1.07997751e+00
5.39487064e-01 6.42854691e-01 -5.31585872e-01 9.57345307e-01
1.04631150e+00 9.78067994e-01 -2.19544113e-01 1.59626052e-01
1.10827553e+00 -7.07238734e-01 -4.38006938e-01 -3.85268740e-02
4.65132922e-01 -9.50469017e-01 9.12937164e-01 5.20635605e-01
-1.12375128e+00 -4.72578466e-01 -9.06574965e-01 -4.90535468e-01
1.23215474e-01 -2.38612786e-01 1.02143991e+00 3.37414920e-01
-1.07286572e+00 7.05864489e-01 -1.24740696e+00 1.46732643e-01
7.00170577e-01 6.07687891e-01 -3.84141266e-01 -3.75216007e-01
-4.20287758e-01 2.40119815e-01 -5.78995705e-01 8.82062241e-02
-3.00023347e-01 -1.11928487e+00 -8.29792857e-01 1.57743558e-01
9.87585112e-02 -8.69360268e-01 1.00725019e+00 -3.39777291e-01
-1.32709086e+00 5.80416203e-01 -4.64472234e-01 -3.17741297e-02
3.13834429e-01 -1.65865839e-01 8.31410438e-02 -1.72763094e-01
9.75424647e-02 7.11885035e-01 8.81661117e-01 -1.33766234e+00
-3.50798607e-01 -6.84091032e-01 -1.73022360e-01 2.26086855e-01
-1.46963984e-01 -2.95706242e-01 -9.01674628e-01 -5.90377569e-01
7.10908473e-01 -7.32479572e-01 -5.69876492e-01 7.47014165e-01
-2.68707871e-01 -2.08007023e-01 8.88451159e-01 -2.40536600e-01
7.97644556e-01 -2.30532789e+00 1.37699798e-01 4.39110190e-01
5.87911308e-01 -1.60861135e-01 -1.43550396e-01 2.53405958e-01
-4.44685221e-02 9.59083531e-03 -2.26784781e-01 -7.80412257e-01
-1.52591839e-01 2.89492607e-01 -3.85789782e-01 6.57889903e-01
1.93262920e-01 7.85457850e-01 -6.82191908e-01 -2.26508662e-01
6.02409005e-01 1.01499200e+00 -1.18176115e+00 8.56501758e-02
-2.59879112e-01 4.24077392e-01 -6.06141090e-01 7.54420280e-01
1.21316099e+00 -6.18943930e-01 -1.68723628e-01 -6.25580907e-01
-2.09238827e-01 2.22128198e-01 -1.28720534e+00 1.99966443e+00
-7.10054100e-01 3.42411816e-01 2.17301369e-01 -5.41477442e-01
8.83959770e-01 1.21184289e-02 7.58669496e-01 -6.06722116e-01
1.16177991e-01 1.39644876e-01 -2.10341498e-01 1.69448435e-01
6.52585804e-01 1.51587054e-01 1.57787725e-01 1.89546138e-01
-3.95716220e-01 -5.99893868e-01 -5.42678833e-01 1.39923871e-01
1.15754139e+00 7.70204067e-02 7.47599360e-03 -2.97382712e-01
2.06727639e-01 -1.91259548e-01 3.85892272e-01 4.36867833e-01
4.94437456e-01 8.65055144e-01 4.75714244e-02 -5.83441794e-01
-9.75895405e-01 -1.05990076e+00 -3.71383041e-01 7.84568846e-01
2.68951058e-01 -6.81107759e-01 -2.11332917e-01 -1.91669807e-01
5.08421004e-01 4.51794237e-01 -3.89474094e-01 2.35950068e-01
-6.21109188e-01 -4.90506977e-01 2.85621453e-02 6.20795786e-01
2.64819771e-01 -6.98816299e-01 -3.95465642e-01 3.11253220e-01
1.18475137e-02 -8.90079141e-01 -2.40508750e-01 -5.07142954e-02
-1.25124729e+00 -8.37443590e-01 -5.62469840e-01 -2.56663293e-01
6.85258627e-01 6.53806210e-01 1.18029213e+00 4.80593741e-01
-1.39265195e-01 1.81401744e-01 -1.84910625e-01 -2.69641131e-01
2.16957286e-01 3.70421149e-02 -1.72599077e-01 1.45541191e-01
2.38974988e-01 -8.66108418e-01 -6.31566167e-01 3.46197486e-01
-8.80912006e-01 2.46349841e-01 5.17984509e-01 8.28470409e-01
1.06093287e+00 -3.48397434e-01 -6.20178320e-02 -7.53781438e-01
1.96555629e-02 -4.41754341e-01 -6.76282406e-01 -3.33385527e-01
-1.83848113e-01 5.82255833e-02 4.10746992e-01 -1.12958476e-02
-6.28541946e-01 1.36101887e-01 -5.85573554e-01 -8.96285474e-01
6.91779405e-02 3.89312029e-01 -1.15544692e-01 -7.99663186e-01
4.23690557e-01 3.86463515e-02 -6.17631599e-02 -8.00168037e-01
2.02744126e-01 3.26729327e-01 2.57947415e-01 -8.18766832e-01
6.74044430e-01 9.09122944e-01 2.08949819e-01 -1.23875034e+00
-2.48303026e-01 -4.72206205e-01 -4.02747363e-01 -6.20391816e-02
3.74834001e-01 -9.27321374e-01 -1.05109978e+00 5.29110312e-01
-1.56888878e+00 -8.53135884e-02 -3.08451861e-01 2.19825611e-01
-4.16646689e-01 2.55718857e-01 -4.42094713e-01 -6.57481074e-01
-3.17887843e-01 -1.42120731e+00 1.62655199e+00 -2.38623738e-01
1.01977788e-01 -7.23396003e-01 -2.27040261e-01 4.90782149e-02
6.82381570e-01 -3.74652371e-02 8.06763589e-01 1.20827533e-01
-1.49263203e+00 -1.08269259e-01 -6.86960340e-01 -4.80334312e-02
1.93507131e-02 -1.05683170e-01 -1.03515613e+00 -3.41449648e-01
-4.95692715e-02 -2.46093562e-03 9.52566803e-01 4.66842562e-01
1.59009516e+00 -1.20763741e-01 -5.35597384e-01 1.19667697e+00
1.54585159e+00 -3.33190471e-01 8.73031080e-01 -4.00658846e-02
1.25839710e+00 1.67721644e-01 4.04107332e-01 7.16646791e-01
5.88369191e-01 9.17064905e-01 8.28903377e-01 -2.57295191e-01
-1.63492098e-01 2.67529930e-03 -2.39875391e-01 1.16652596e+00
-3.28984559e-01 -5.92309721e-02 -9.56736028e-01 3.41280133e-01
-1.80000389e+00 -4.32665855e-01 -4.17222172e-01 2.25906396e+00
5.61046958e-01 -1.25240952e-01 -4.70791429e-01 8.69344771e-02
2.22133875e-01 3.56028110e-01 -5.81993043e-01 -1.99946743e-02
-7.27981329e-02 4.78436768e-01 7.02755392e-01 6.13850594e-01
-9.28569555e-01 9.34961677e-01 6.02407885e+00 1.21574128e+00
-1.36592865e+00 2.64305145e-01 5.90792477e-01 -2.93325722e-01
-7.85718858e-01 -2.30633244e-01 -6.64072633e-01 2.16274858e-01
4.69850659e-01 1.67203829e-01 3.58053058e-01 9.26882744e-01
-2.58122422e-02 2.78674159e-02 -1.05628669e+00 1.52013838e+00
-2.52020196e-03 -1.80978012e+00 2.62691617e-01 3.13400477e-01
5.69804132e-01 5.88606536e-01 4.92294431e-02 -4.05596948e-04
1.30547538e-01 -1.12904704e+00 6.21970236e-01 5.27364612e-01
1.19234300e+00 -7.61585057e-01 5.70343554e-01 1.13610007e-01
-1.41517317e+00 3.67769867e-01 -6.72820270e-01 -2.21807495e-01
3.29322904e-01 9.84657586e-01 -4.04699206e-01 5.89609385e-01
7.86212504e-01 8.88278365e-01 -2.27558002e-01 1.14792013e+00
3.83263767e-01 3.99694681e-01 -1.02213597e+00 9.05841738e-02
6.65440410e-02 -2.36251876e-01 5.66796064e-01 7.54236877e-01
4.99464840e-01 7.85454735e-02 2.37572506e-01 7.80084968e-01
-7.93474391e-02 3.18920046e-01 -5.91712832e-01 5.14766634e-01
4.77421224e-01 1.07985663e+00 -7.01675832e-01 -9.35941935e-02
-6.08226180e-01 8.71410012e-01 4.32340950e-01 1.22055270e-01
-5.84563076e-01 3.21437083e-02 1.19351017e+00 5.74876308e-01
4.27159965e-01 -5.96478105e-01 -8.52515280e-01 -1.04909730e+00
2.27044169e-02 -5.97205579e-01 -1.77429035e-01 -5.92178464e-01
-1.24081743e+00 5.43935061e-01 -2.86425978e-01 -1.30094051e+00
3.26848447e-01 -5.73428988e-01 -3.00783515e-01 9.09628808e-01
-1.60117733e+00 -1.14167130e+00 -5.21402061e-01 6.13879859e-01
4.51735109e-01 1.58448368e-01 8.81641746e-01 8.21971357e-01
-6.82506561e-02 5.27436137e-01 1.58638023e-02 -2.04204783e-01
4.49815452e-01 -8.50537241e-01 9.34005201e-01 3.82607698e-01
1.92514375e-01 6.00761235e-01 3.92523736e-01 -7.84467638e-01
-2.06481767e+00 -1.18709421e+00 6.00922048e-01 -3.44098866e-01
2.36731932e-01 -3.12160194e-01 -1.00159061e+00 5.00150204e-01
-3.19557995e-01 3.75907212e-01 3.05848122e-01 2.02110052e-01
-5.26878476e-01 -1.34916857e-01 -1.01686442e+00 4.52104241e-01
1.18314505e+00 -3.88259381e-01 8.09063762e-02 4.97807324e-01
8.49331558e-01 -7.61516452e-01 -9.32942748e-01 5.96443892e-01
4.19271559e-01 -9.00243104e-01 1.31284475e+00 -1.11475952e-01
4.16465312e-01 -4.01116341e-01 -5.27211785e-01 -9.94530022e-01
-7.65209854e-01 -3.58502090e-01 -1.10958077e-01 7.04539657e-01
2.11578771e-01 -6.08170271e-01 1.07819176e+00 4.81524915e-01
-3.15096438e-01 -1.09300554e+00 -1.20637596e+00 -4.39378619e-01
-1.92061827e-01 -1.02089906e+00 9.42043483e-01 9.61648107e-01
-6.53705955e-01 1.82340428e-01 -2.63773620e-01 3.79787147e-01
8.47571969e-01 2.37606853e-01 1.01911867e+00 -1.42389011e+00
-2.71537770e-02 -6.14118338e-01 -7.95699656e-01 -1.75977612e+00
-2.34692708e-01 -9.27585602e-01 -1.47146881e-01 -1.51520348e+00
-1.48798391e-01 -1.08121121e+00 1.24975771e-01 4.81344819e-01
1.36433020e-01 6.98886514e-01 1.30866468e-01 1.87380165e-01
-3.70480776e-01 9.94818807e-01 1.35848987e+00 -2.16382116e-01
7.39239380e-02 -1.29432335e-01 -5.20782292e-01 6.52346253e-01
3.48162562e-01 -4.27725017e-01 -5.35459667e-02 -9.77350473e-01
5.15881658e-01 -2.66064461e-02 4.72328305e-01 -1.01386058e+00
4.19751972e-01 -1.25804499e-01 4.41073507e-01 -1.12084687e+00
1.00559211e+00 -1.11594748e+00 3.45469356e-01 1.53415781e-02
3.76506895e-01 1.95689470e-01 3.33837628e-01 6.08726680e-01
-1.37272730e-01 4.15028304e-01 7.09367573e-01 -2.34517897e-03
-4.18810397e-01 9.68447208e-01 2.10107148e-01 -4.01678860e-01
6.84448898e-01 -1.93284467e-01 -8.08176473e-02 -4.56038952e-01
-3.74812186e-01 1.11259855e-01 7.28187859e-01 3.66735995e-01
9.27181244e-01 -1.57636321e+00 -9.45879996e-01 7.46052861e-01
2.29438525e-02 7.38344014e-01 6.47598565e-01 7.46804595e-01
-9.22461867e-01 2.74515808e-01 -8.18284985e-04 -1.20236146e+00
-1.04928768e+00 2.15935707e-01 -8.68286043e-02 9.84353498e-02
-1.05578279e+00 1.21034825e+00 4.76552188e-01 -5.64709783e-01
4.55898419e-02 -6.69447482e-01 2.00633094e-01 -2.68792838e-01
5.44193625e-01 3.25323641e-01 5.12422383e-01 -7.26205707e-01
-3.56255323e-01 1.05691576e+00 4.32299003e-02 1.39656499e-01
1.43930542e+00 1.23667337e-01 -3.34246069e-01 3.28471839e-01
1.25752640e+00 1.27828136e-01 -1.10777390e+00 -4.56475616e-01
-5.31357288e-01 -1.08314049e+00 6.83359265e-01 -1.68394521e-01
-1.41656125e+00 1.01380372e+00 3.44068408e-01 -1.89545788e-02
7.03342021e-01 -2.05257349e-02 1.02446508e+00 3.07713479e-01
9.14397299e-01 -4.25214231e-01 -3.90407592e-01 6.95269644e-01
1.27469730e+00 -1.10981178e+00 2.51130760e-01 -9.93661046e-01
-5.51314466e-02 9.03425694e-01 4.35169548e-01 -4.04425591e-01
1.08492982e+00 5.93855202e-01 -2.12889060e-01 -4.03838128e-01
-6.93406165e-01 1.91036642e-01 4.09810662e-01 4.41861153e-01
4.05756891e-01 2.09208637e-01 2.07328737e-01 3.85239214e-01
-2.54212022e-01 -2.15821892e-01 1.45158423e-02 7.69751251e-01
-3.06000084e-01 -1.14868069e+00 -5.58715105e-01 8.47424984e-01
2.81445943e-02 -2.93276817e-01 1.19197749e-01 3.42338711e-01
-2.47712195e-01 3.76807719e-01 5.65788388e-01 -3.77359092e-01
4.56763983e-01 -4.56608504e-01 5.78815639e-01 -6.32702529e-01
-4.40892458e-01 2.68665999e-01 -1.77489400e-01 -1.18876493e+00
-1.79604143e-01 -6.82018757e-01 -1.15833390e+00 -7.63178110e-01
-2.90657014e-01 -1.21227510e-01 9.76486444e-01 5.30650020e-01
9.14051592e-01 6.08886182e-01 4.88262743e-01 -1.74299145e+00
-2.48148397e-01 -7.87156999e-01 -4.29388970e-01 2.21203834e-01
2.80695111e-01 -9.25398588e-01 -1.92064673e-01 -4.64841932e-01] | [8.012418746948242, -3.5474042892456055] |
aed6051a-71a1-49a4-8890-7da81ddd659c | a-conceptual-model-for-end-to-end-causal | 2305.16165 | null | https://arxiv.org/abs/2305.16165v1 | https://arxiv.org/pdf/2305.16165v1.pdf | A Conceptual Model for End-to-End Causal Discovery in Knowledge Tracing | In this paper, we take a preliminary step towards solving the problem of causal discovery in knowledge tracing, i.e., finding the underlying causal relationship among different skills from real-world student response data. This problem is important since it can potentially help us understand the causal relationship between different skills without extensive A/B testing, which can potentially help educators to design better curricula according to skill prerequisite information. Specifically, we propose a conceptual solution, a novel causal gated recurrent unit (GRU) module in a modified deep knowledge tracing model, which uses i) a learnable permutation matrix for causal ordering among skills and ii) an optionally learnable lower-triangular matrix for causal structure among skills. We also detail how to learn the model parameters in an end-to-end, differentiable way. Our solution placed among the top entries in Task 3 of the NeurIPS 2022 Challenge on Causal Insights for Learning Paths in Education. We detail preliminary experiments as evaluated on the challenge's public leaderboard since the ground truth causal structure has not been publicly released, making detailed local evaluation impossible. | ['Andrew Lan', 'Aritra Ghosh', 'Hunter McNichols', 'Jaewook Lee', 'Wanyong Feng', 'Nischal Ashok Kumar'] | 2023-05-11 | null | null | null | null | ['causal-discovery', 'knowledge-tracing'] | ['knowledge-base', 'miscellaneous'] | [ 4.83831495e-01 4.59771305e-01 -3.24510515e-01 -4.73044574e-01
-5.10821760e-01 -7.18438089e-01 5.12993753e-01 2.37688795e-01
-1.94033876e-01 8.33091319e-01 6.91684246e-01 -8.02382708e-01
-1.02523911e+00 -7.92873204e-01 -1.20746648e+00 -3.88087720e-01
-7.14640990e-02 4.97652441e-01 8.80776569e-02 -4.64020133e-01
4.19101268e-01 6.11336119e-02 -1.62597990e+00 4.89718497e-01
1.19534659e+00 1.54834092e-01 3.24650437e-01 7.08934605e-01
2.54416019e-01 1.33986604e+00 -3.55095476e-01 -3.15337360e-01
-1.98719442e-01 -7.83751607e-01 -1.46575987e+00 -5.89644015e-01
7.47470260e-01 -1.08814254e-01 -2.95714229e-01 9.21626031e-01
3.52918655e-01 3.82529438e-01 4.85352665e-01 -1.20267260e+00
-7.00543880e-01 1.48551166e+00 -4.32779789e-01 4.38647002e-01
5.06215513e-01 -1.65194318e-01 1.08221817e+00 -5.19050285e-02
5.33997118e-01 1.56557572e+00 2.90572435e-01 5.77806175e-01
-1.26841974e+00 -8.66756797e-01 7.11516738e-01 8.52819681e-01
-8.50743651e-01 1.53230771e-01 5.09677887e-01 -7.68943250e-01
3.72072607e-01 1.93569705e-01 8.71001065e-01 1.54436314e+00
1.63410842e-01 9.64089811e-01 1.26210642e+00 -3.78334671e-01
-1.32280141e-01 -2.11956471e-01 5.68913043e-01 1.02987838e+00
5.26863523e-02 3.52144748e-01 -1.09059227e+00 -3.79933720e-03
8.61480832e-01 -3.83727938e-01 -3.04673553e-01 -2.40126833e-01
-1.15268588e+00 7.57206857e-01 5.54212868e-01 7.65904784e-02
-1.27112418e-01 7.85199404e-01 -2.04621360e-01 6.48495436e-01
-1.19152613e-01 6.80283844e-01 -4.79132921e-01 -4.18156981e-01
-4.07834619e-01 4.72757518e-01 6.26741230e-01 6.87938869e-01
2.72003412e-01 -1.52203232e-01 -3.86720836e-01 3.92326981e-01
4.02010858e-01 1.66233823e-01 2.18564987e-01 -9.63304341e-01
2.71256626e-01 6.50910139e-01 5.73923625e-02 -8.17207694e-01
-5.49059093e-01 -5.92042387e-01 -3.58364254e-01 -3.35773826e-01
6.50847197e-01 -4.29865867e-01 -7.71261156e-01 2.02232337e+00
3.03522408e-01 1.10354960e+00 -1.97212353e-01 1.00949621e+00
1.03825605e+00 3.49854618e-01 1.17817946e-01 1.33877158e-01
1.42896366e+00 -4.96394634e-01 -6.94410503e-01 5.66532724e-02
5.68035305e-01 -6.59511149e-01 1.12273002e+00 6.54462636e-01
-9.49867368e-01 -4.89313275e-01 -7.50654101e-01 -1.56880230e-01
-1.28828391e-01 -5.65158427e-02 8.89840543e-01 7.30168998e-01
-9.60688233e-01 5.97503722e-01 -7.93417931e-01 -1.43998161e-01
1.10443152e-01 3.70410055e-01 8.95882584e-03 -1.09478116e-01
-1.68940938e+00 6.58002198e-01 1.61105797e-01 1.72508642e-01
-1.49990618e+00 -1.51232505e+00 -4.95033264e-01 2.57299900e-01
6.45699084e-01 -8.06610644e-01 1.30623102e+00 -2.85377055e-01
-1.59676600e+00 3.11905563e-01 1.14479050e-01 -3.07152152e-01
3.59014004e-01 -5.51510632e-01 -3.33511531e-02 -3.59099507e-01
-8.25873986e-02 4.57666367e-01 3.32254171e-01 -1.08281863e+00
-9.51154947e-01 -1.90878540e-01 1.58696815e-01 2.03027576e-01
-1.92789376e-01 -1.88357785e-01 -1.89538509e-01 -4.48479116e-01
-1.17837466e-01 -1.05238223e+00 -3.75367910e-01 -1.08686101e+00
-4.43093389e-01 -6.26166165e-01 2.31749132e-01 -5.88052571e-01
1.33395267e+00 -1.63634229e+00 4.89642203e-01 5.42856038e-01
2.38282397e-01 -3.16251308e-01 -2.89743483e-01 2.49835446e-01
-4.79080498e-01 4.24493067e-02 3.47757727e-01 4.31364834e-01
-2.75571588e-02 -4.36221920e-02 -6.22608662e-01 7.22731277e-02
7.35626817e-02 8.37279975e-01 -1.67033577e+00 -6.33815229e-02
-9.25366133e-02 3.31836343e-01 -8.73841166e-01 1.71975464e-01
-3.21457416e-01 7.31529593e-01 -6.91029191e-01 9.82245281e-02
-1.71912178e-01 -1.59900844e-01 4.83356088e-01 2.29585961e-01
-3.11598241e-01 9.21856344e-01 -1.38777220e+00 1.63670325e+00
-6.66427970e-01 6.94885671e-01 -3.30496818e-01 -9.98943329e-01
6.88565314e-01 5.27262330e-01 3.97355646e-01 -5.49635530e-01
-1.33463040e-01 -7.80667812e-02 4.30088401e-01 -5.16899168e-01
5.57922661e-01 1.89492300e-01 -8.19118619e-02 5.52260220e-01
2.71217693e-02 1.27298698e-01 3.06972355e-01 4.55773115e-01
1.51828003e+00 3.42177719e-01 -3.03402930e-01 -5.88920474e-01
1.81098700e-01 -2.98950020e-02 5.48391819e-01 1.02214551e+00
3.08622748e-01 2.23384708e-01 1.08313835e+00 -2.24214911e-01
-2.30010882e-01 -9.42328453e-01 6.29059523e-02 1.39904308e+00
-9.01535004e-02 -3.77253026e-01 -5.64063370e-01 -7.49524176e-01
-1.35181593e-02 9.54062879e-01 -9.86219764e-01 -3.23752046e-01
-7.27078795e-01 -6.59006178e-01 4.38546985e-01 4.32953089e-01
1.36248276e-01 -9.49216247e-01 -4.49189723e-01 1.86151192e-01
-5.46273589e-01 -6.16257131e-01 -2.97414958e-01 3.23037714e-01
-6.69983745e-01 -1.42112482e+00 -1.52318448e-01 -6.16160512e-01
6.21576369e-01 2.88715839e-01 1.09611177e+00 3.45354602e-02
-2.19539300e-01 6.55664742e-01 -1.25852734e-01 -5.88113666e-01
-1.85258463e-01 2.09990591e-02 1.55608162e-01 -3.33742410e-01
2.37092644e-01 -4.97407705e-01 -6.73771501e-01 3.69482309e-01
-3.76022130e-01 1.29187435e-01 5.01044214e-01 8.87452424e-01
3.92353505e-01 3.21936667e-01 6.71849668e-01 -1.27966869e+00
1.01113582e+00 -6.34064913e-01 -7.62300313e-01 3.51043910e-01
-6.81686044e-01 3.29309434e-01 2.00077236e-01 -3.43291819e-01
-1.28499067e+00 -1.05339050e-01 4.23004068e-02 -1.96850061e-01
-4.58786003e-02 9.23533380e-01 7.83032849e-02 2.07793057e-01
7.64752448e-01 -1.17091395e-01 -4.90268350e-01 -5.63400209e-01
6.59793496e-01 -4.41517048e-02 5.84700167e-01 -1.23238385e+00
5.92096806e-01 -9.87615734e-02 8.58304128e-02 -4.21673059e-01
-1.38870597e+00 -4.50344712e-01 -5.80426276e-01 -2.95097291e-01
8.42398226e-01 -1.14011478e+00 -1.53891146e+00 -1.12170570e-01
-1.03041959e+00 -8.55009854e-01 -9.15056393e-02 7.07658470e-01
-4.66810793e-01 -3.82321239e-01 -5.18586516e-01 -4.12668049e-01
3.53894532e-01 -1.15786386e+00 4.63339627e-01 4.61668253e-01
-4.79712546e-01 -1.09115207e+00 3.12132806e-01 8.89166236e-01
-3.26554850e-02 -9.14973542e-02 1.30338717e+00 -3.66825730e-01
-7.20219016e-01 3.62027049e-01 5.00672087e-02 -1.61769465e-01
-1.23007290e-01 1.52448103e-01 -6.17752790e-01 -6.89682811e-02
-6.07382178e-01 -4.59748775e-01 1.09143794e+00 5.14810920e-01
1.45397353e+00 -2.32827678e-01 -3.95126909e-01 1.99211329e-01
8.05472195e-01 2.29869392e-02 2.15965509e-01 1.64036661e-01
1.18443894e+00 1.03042567e+00 5.11940062e-01 1.04601488e-01
8.75092924e-01 4.51371640e-01 3.66090119e-01 1.60198569e-01
-2.98256487e-01 -7.78570354e-01 3.70947599e-01 8.90689909e-01
-3.46697450e-01 -1.18348800e-01 -1.19320607e+00 7.70955443e-01
-2.04316425e+00 -1.15307879e+00 -7.38518178e-01 2.03471231e+00
1.41740859e+00 -2.51028948e-02 -7.24262521e-02 -3.04135621e-01
2.63639212e-01 -4.50281918e-01 -3.25170964e-01 -3.66106689e-01
5.32744527e-01 4.74704474e-01 4.04350877e-01 8.21404874e-01
-7.00221956e-01 1.11377287e+00 6.05758810e+00 6.22480631e-01
-7.07225025e-01 2.19760812e-03 4.74540710e-01 -2.53450662e-01
-8.10433149e-01 1.37774646e-01 -1.15492129e+00 1.87367335e-01
1.16379738e+00 -3.08439702e-01 4.32493895e-01 2.57349253e-01
4.67179686e-01 -1.77853438e-03 -1.38744783e+00 3.72451603e-01
-5.20414829e-01 -1.30850971e+00 -3.46839018e-02 -2.40755081e-02
1.24539280e+00 -2.66222566e-01 2.90335149e-01 5.59902072e-01
1.63561571e+00 -1.38111734e+00 4.11351770e-01 7.40018547e-01
3.33045930e-01 -1.08786857e+00 2.90824443e-01 2.28820756e-01
-7.75646806e-01 -2.35076100e-01 -2.73794621e-01 -3.38853806e-01
-3.79650533e-01 6.21716082e-01 -1.01778913e+00 5.65514088e-01
6.63830221e-01 6.66599154e-01 -4.26844448e-01 8.87323856e-01
-1.12276483e+00 1.42239606e+00 9.90128741e-02 -1.65949211e-01
1.81625083e-01 9.82398465e-02 3.75213206e-01 8.45245361e-01
2.60135412e-01 5.27605057e-01 -2.62203650e-03 1.03416872e+00
-1.88201830e-01 -2.58329958e-01 -3.23839486e-01 -5.65722995e-02
5.46983182e-01 1.10700536e+00 -6.27479613e-01 1.11496717e-01
-2.93161515e-02 3.68243963e-01 4.07519400e-01 4.79549676e-01
-7.02624798e-01 6.45343065e-02 8.49449158e-01 1.32437274e-02
-8.80411863e-02 8.41883291e-03 -3.06527406e-01 -8.33340168e-01
-7.18668520e-01 -9.32435632e-01 5.25533378e-01 -6.28056943e-01
-1.08635068e+00 -1.42566636e-01 1.53362691e-01 -6.64409339e-01
-2.65031904e-01 -2.74860024e-01 -6.12937391e-01 1.04492402e+00
-1.25461841e+00 -6.11763299e-01 -1.53928995e-01 5.86336195e-01
9.73219424e-02 3.84249091e-02 5.74606657e-01 1.64361507e-01
-9.63867068e-01 6.82416439e-01 -1.77740753e-01 -2.72795502e-02
9.75368321e-01 -1.67986536e+00 2.62554973e-01 8.18797469e-01
3.26756239e-01 1.04547048e+00 1.05186176e+00 -8.77861857e-01
-1.42918158e+00 -9.30070221e-01 8.53004813e-01 -8.70125949e-01
1.17279375e+00 -2.23694295e-01 -8.01710367e-01 8.66164625e-01
3.31029892e-01 -6.24996662e-01 9.03269172e-01 1.18973339e+00
-1.84515417e-01 1.97432935e-01 -2.30465591e-01 7.39048362e-01
1.28349781e+00 -1.19908661e-01 -7.31775761e-01 3.15934062e-01
9.00736988e-01 -5.95847785e-01 -9.31436479e-01 2.75806189e-01
3.65351170e-01 -5.08200884e-01 9.30187762e-01 -1.10111773e+00
1.30166101e+00 -1.90726399e-01 4.51838821e-01 -1.83349168e+00
-4.85686421e-01 -7.50217795e-01 -3.44561823e-02 9.74376678e-01
5.10565817e-01 -6.79368079e-02 1.10986686e+00 4.03556883e-01
-1.80971012e-01 -7.42640197e-01 -5.61639845e-01 -3.42665374e-01
1.93826571e-01 -4.73547429e-01 3.78684610e-01 1.62033272e+00
3.01392656e-02 7.26475537e-01 -4.02313858e-01 4.15051162e-01
6.87391520e-01 4.28236872e-01 7.03744292e-01 -1.50222242e+00
-4.41174626e-01 -5.58407426e-01 1.56994864e-01 -1.21746182e+00
2.53493577e-01 -1.01926243e+00 -3.87079082e-02 -1.61811233e+00
3.54243249e-01 -6.83548570e-01 -5.57967782e-01 9.02010560e-01
-6.47098064e-01 -4.73507702e-01 -1.66169152e-01 -4.65882182e-01
-4.06253755e-01 4.25325632e-01 1.70487452e+00 7.24756941e-02
-2.82430708e-01 7.47520477e-02 -1.19403493e+00 6.12306356e-01
4.69118357e-01 -6.35177970e-01 -1.06525457e+00 -4.31552619e-01
9.08311963e-01 3.22285056e-01 6.66139662e-01 -4.26717788e-01
5.19352436e-01 -6.80125713e-01 1.96522430e-01 -4.51213539e-01
-3.49036455e-01 -4.40653563e-01 -1.01178326e-01 5.23111105e-01
-9.37639475e-01 -1.01323582e-01 2.82427043e-01 6.82328463e-01
-1.46257609e-01 -1.75375238e-01 1.27154753e-01 9.82706100e-02
-8.33289266e-01 -1.30959481e-01 -3.71541172e-01 1.59827664e-01
6.99059784e-01 3.18010926e-01 -7.29219079e-01 -3.74559760e-01
-7.70538747e-01 9.57990587e-01 -4.79337692e-01 8.99004996e-01
5.31852365e-01 -1.30077648e+00 -9.84835386e-01 -1.98916614e-01
-2.46469468e-01 -2.69136654e-04 6.43485010e-01 7.70617604e-01
-1.35035008e-01 6.53633475e-01 -2.62919217e-01 -3.54247630e-01
-1.35023689e+00 2.13764504e-01 2.07672298e-01 -5.77119648e-01
-2.55446404e-01 1.61889458e+00 3.42503101e-01 -6.36526763e-01
3.23011696e-01 -3.01297009e-01 -6.17460310e-01 3.01524878e-01
5.24630487e-01 4.99197185e-01 -3.70834507e-02 2.22998366e-01
-7.40274265e-02 -1.04813417e-02 -2.46562973e-01 -1.45851254e-01
1.48988354e+00 2.08275661e-01 6.20776508e-03 5.14045000e-01
5.91079295e-01 1.17809176e-01 -1.18650174e+00 -2.05164656e-01
1.70712397e-01 -2.20790520e-01 3.02255183e-01 -1.29940760e+00
-8.76880586e-01 7.74470270e-01 4.00883138e-01 -3.44899669e-02
8.05911422e-01 -1.01582430e-01 1.85683504e-01 3.71787727e-01
2.76967406e-01 -8.20004463e-01 2.53195941e-01 7.62409627e-01
7.86955357e-01 -9.98315275e-01 7.09214509e-02 -5.03301382e-01
-2.58288354e-01 9.31339085e-01 8.80397737e-01 2.75967062e-01
5.53508759e-01 9.99961048e-02 -1.59424633e-01 -4.73011613e-01
-1.48274601e+00 -1.14603646e-01 6.56705797e-01 1.99434623e-01
1.01077712e+00 4.19294417e-01 -3.34328830e-01 7.66349494e-01
-7.44622648e-01 -6.78651258e-02 8.66447151e-01 5.31798005e-01
-2.70177633e-01 -1.27644014e+00 -3.36064190e-01 4.36486065e-01
-2.24392429e-01 -2.81244218e-01 -5.13328731e-01 5.73004961e-01
5.37007339e-02 9.83661830e-01 -2.55851112e-02 -5.86709917e-01
6.30357862e-01 -4.65112515e-02 8.87068868e-01 -8.00977528e-01
-7.96520829e-01 -3.37781042e-01 8.20645243e-02 -7.76794732e-01
-1.49089247e-01 -8.76858294e-01 -1.27647555e+00 -5.50604343e-01
1.49633586e-01 2.62157023e-01 3.98026407e-01 9.07402337e-01
1.56096533e-01 1.48171043e+00 2.33862996e-01 -3.63974236e-02
-3.25333267e-01 -7.90581226e-01 -1.62048250e-01 6.83152750e-02
3.55165377e-02 -8.90691280e-01 1.74243301e-01 1.60385862e-01] | [10.05357551574707, 7.201754570007324] |
1b97d25d-e099-48b8-a7bb-f3c959527ceb | a-spatiotemporal-multi-channel-learning | null | null | https://ieeexplore.ieee.org/abstract/document/9106397/ | https://ieeexplore.ieee.org/abstract/document/9106397/ | A Spatiotemporal Multi-Channel Learning Framework for Automatic Modulation Recognition | Automatic modulation recognition (AMR) plays a vital role in modern communication systems. This letter proposes a novel three-stream deep learning framework to extract the features from individual and combined in-phase/quadrature (I/Q) symbols of the modulated data. The proposed framework integrates one-dimensional (1D) convolutional, two-dimensional (2D) convolutional and long short-term memory (LSTM) layers to extract features more effectively from a time and space perspective. Experiments on the benchmark dataset show the proposed framework has efficient convergence speed and achieves improved recognition accuracy, especially for the signals modulated by higher dimensional schemes such as 16 quadrature amplitude modulation (16-QAM) and 64-QAM. | ['Yang Luo', 'Gerard Parr', 'Chunbo Luo', 'Jialang Xu'] | 2020-06-02 | null | null | null | ieee-wireless-communications-letters-2020-6 | ['automatic-modulation-recognition'] | ['time-series'] | [ 5.62775910e-01 -5.81033230e-01 -4.12970424e-01 -2.44971558e-01
-1.03366840e+00 1.81877002e-01 6.56816840e-01 -2.60794640e-01
-5.22236168e-01 8.88025582e-01 4.00023125e-02 -7.92676806e-01
-5.46116292e-01 -6.74401999e-01 -1.29832730e-01 -9.84456837e-01
-7.06278741e-01 -3.83607984e-01 -2.86872000e-01 -2.37769499e-01
4.39584047e-01 8.29963326e-01 -1.11992109e+00 5.18380284e-01
3.44407141e-01 1.42888236e+00 -2.43735686e-01 1.24547744e+00
1.01566710e-01 9.54830170e-01 -1.00052643e+00 -1.48799330e-01
1.21119916e-01 -3.15786451e-01 -5.41955352e-01 -2.06215501e-01
9.75364074e-02 -1.65134862e-01 -1.36222279e+00 8.80886793e-01
8.25955510e-01 -2.11217210e-01 6.61627233e-01 -9.44575191e-01
-5.29465973e-01 2.58466184e-01 -6.67683721e-01 6.80665255e-01
2.68404018e-02 2.14056119e-01 4.22079921e-01 -8.31609368e-01
-6.26543686e-02 1.19269204e+00 6.18405759e-01 1.12386502e-01
-5.88736653e-01 -8.36014390e-01 -5.26296020e-01 7.60722041e-01
-1.62299776e+00 -7.06514835e-01 5.33359349e-01 -7.60738701e-02
1.24905860e+00 2.33606324e-01 1.43734500e-01 7.88025916e-01
9.01945949e-01 5.28489769e-01 1.09631026e+00 -5.04526973e-01
-8.80060643e-02 -5.82865179e-01 8.63646641e-02 4.72326100e-01
-4.89141792e-02 3.85898769e-01 -4.65321422e-01 -1.47465572e-01
5.78039765e-01 -3.25208604e-01 -1.89557403e-01 3.90140891e-01
-1.16041708e+00 7.82631278e-01 4.42562222e-01 6.97925150e-01
-5.10613799e-01 5.93849778e-01 1.92509905e-01 8.34255397e-01
2.97020338e-02 2.17431068e-01 -2.13597447e-01 -5.33171296e-01
-1.06175327e+00 -2.59709537e-01 2.56904721e-01 7.85356164e-01
4.82549876e-01 6.37945533e-01 -4.81082678e-01 7.17694283e-01
5.46249807e-01 7.12885916e-01 7.65828192e-01 -5.35452485e-01
6.45186305e-01 -1.48552895e-01 -8.42375904e-02 -9.99325514e-01
-7.12578297e-01 -9.64952886e-01 -1.41540062e+00 -2.27204002e-02
1.01450104e-02 -1.51569098e-01 -1.28324282e+00 1.24325716e+00
-2.63356090e-01 3.40614587e-01 4.44550157e-01 7.78671622e-01
8.81132960e-01 1.03399706e+00 -1.65505260e-01 -3.58495981e-01
1.30051363e+00 -7.93094635e-01 -1.11248398e+00 -1.13470264e-01
5.65450311e-01 -1.05602241e+00 -1.18558899e-01 4.20111656e-01
-8.80192161e-01 -8.62184405e-01 -1.71099293e+00 2.59427652e-02
-4.23504859e-01 2.63607442e-01 4.66020852e-01 9.37415898e-01
-7.18683720e-01 6.56253517e-01 -2.63978601e-01 4.68905300e-01
8.23776901e-01 7.97220528e-01 -1.68763518e-01 -5.29363938e-02
-1.77879882e+00 8.59759450e-01 6.81604445e-02 4.13078249e-01
-5.99552810e-01 -2.59515464e-01 -7.39905357e-01 -3.89876217e-02
-3.93584460e-01 -3.48571420e-01 1.21126461e+00 -7.98865676e-01
-1.60270333e+00 3.38646919e-01 -2.14546427e-01 -7.47065067e-01
-1.48480847e-01 2.63832539e-01 -1.31841743e+00 3.36012989e-01
-4.68249440e-01 5.27920872e-02 1.47267592e+00 -5.03026366e-01
-7.98897326e-01 -4.50333804e-01 -3.01978171e-01 -8.18408057e-02
1.74655672e-02 1.15689367e-01 2.08838835e-01 -7.26772308e-01
2.42997706e-01 -4.51472789e-01 -1.43204108e-01 -4.62537229e-01
-4.13647182e-02 7.11987168e-02 1.12209833e+00 -7.81968355e-01
1.48778307e+00 -2.22870135e+00 -1.80717722e-01 1.65293321e-01
7.89153501e-02 9.99837697e-01 -2.25663826e-01 4.56240892e-01
-2.41117150e-01 -3.36257219e-01 -4.16286848e-02 -1.88525349e-01
-6.41064644e-02 -8.08461607e-02 -3.57819796e-01 8.28471601e-01
4.73682225e-01 1.04837286e+00 -5.67138791e-01 -1.02050900e-02
1.78436249e-01 5.81743181e-01 2.40364879e-01 -2.74339914e-01
2.33666554e-01 6.73465967e-01 -2.73678720e-01 5.67436457e-01
1.08023691e+00 -1.77384809e-01 2.99038254e-02 -3.71621251e-01
-1.39310762e-01 5.67039609e-01 -8.63416851e-01 1.36225975e+00
-7.74818778e-01 1.38853037e+00 -1.51123703e-01 -1.15323877e+00
8.98485065e-01 5.63277125e-01 3.15435022e-01 -1.54266107e+00
4.98149961e-01 5.11412084e-01 5.65193772e-01 -4.88616318e-01
6.11636579e-01 -1.47109792e-01 -1.40459120e-01 2.96895683e-01
2.79480666e-01 4.40771371e-01 -2.12136909e-01 -3.49501580e-01
8.72628927e-01 -6.58732057e-01 4.30757135e-01 4.75034527e-02
9.95857716e-01 -8.30956757e-01 2.87427187e-01 7.16480970e-01
-3.75580043e-01 2.72065759e-01 1.41086847e-01 -4.13330525e-01
-1.01254582e+00 -6.59713507e-01 -4.85872269e-01 3.87606651e-01
1.15848437e-01 -9.06613767e-02 -1.81968790e-02 -2.60719746e-01
-5.18503338e-02 5.51327467e-01 4.34987135e-02 -4.87798572e-01
-8.66581082e-01 -1.05998683e+00 1.04924297e+00 2.42387727e-01
8.28314424e-01 -6.55866504e-01 -3.93359244e-01 4.93238330e-01
1.40037194e-01 -1.16867197e+00 -1.51955694e-01 3.84610742e-01
-5.90385675e-01 -6.45839334e-01 -8.25322270e-01 -8.93319607e-01
1.17123917e-01 6.59976453e-02 6.05776310e-01 -4.11736339e-01
-3.53942424e-01 -1.59284323e-01 -2.45282367e-01 -4.09970991e-02
-3.30824584e-01 -2.00302228e-01 2.70244386e-02 5.05689919e-01
4.99974936e-01 -6.46525085e-01 -6.59434557e-01 7.74376541e-02
-5.57628691e-01 -2.57199526e-01 1.10339952e+00 7.39703357e-01
2.06185207e-01 2.82189637e-01 1.03789723e+00 -2.36234710e-01
6.99115098e-01 -4.90259349e-01 -4.55575258e-01 -8.04685652e-02
-3.30561072e-01 1.07312530e-01 7.19947040e-01 -2.71541834e-01
-5.41385353e-01 -4.88574862e-01 -6.35455787e-01 -9.13017243e-02
4.88841794e-02 4.64712501e-01 -7.59618804e-02 -6.26856327e-01
4.55757469e-01 6.91724122e-01 -3.22927743e-01 -3.80225450e-01
2.84813076e-01 1.76290023e+00 5.66376984e-01 -9.13826525e-02
8.13941658e-01 3.74704748e-01 3.85928541e-01 -1.20474684e+00
-2.55040467e-01 -2.23511249e-01 -3.69470090e-01 3.81502742e-03
4.03116494e-01 -1.16227865e+00 -6.60079002e-01 8.96454930e-01
-1.13601005e+00 2.24265531e-02 4.01390009e-02 8.17629755e-01
-4.10410523e-01 5.46189189e-01 -6.46463752e-01 -1.04228568e+00
-5.81318021e-01 -9.34632182e-01 1.13105178e+00 3.99314612e-01
2.17894554e-01 -6.96943104e-01 -3.46996456e-01 2.86182594e-02
9.49926496e-01 -9.49743688e-02 1.10794687e+00 -3.72245610e-01
-7.01462448e-01 -4.75871742e-01 -4.06299233e-01 3.44607443e-01
2.84502029e-01 -4.93845522e-01 -9.85479414e-01 -4.99503016e-01
2.20648125e-01 -4.35675047e-02 9.06088471e-01 4.39889163e-01
1.15450883e+00 -3.12153995e-01 -1.43335894e-01 8.36832106e-01
1.24186552e+00 8.45825791e-01 1.25455332e+00 -1.77526139e-02
2.23128229e-01 -3.01608235e-01 5.75623810e-01 5.68469465e-01
1.14904337e-01 7.65833735e-01 9.39951912e-02 5.48233353e-02
-3.42072278e-01 2.39458904e-01 2.37231329e-01 9.59274411e-01
4.07771409e-01 -2.62715220e-01 -5.80591679e-01 2.38092914e-01
-1.41878664e+00 -9.41525459e-01 -3.48751336e-01 2.18426919e+00
6.81028903e-01 2.01502606e-01 -3.62861067e-01 8.35134268e-01
5.74635565e-01 4.73522812e-01 -3.77748936e-01 -7.99236894e-01
-3.49551708e-01 9.63191628e-01 6.41542971e-01 6.36048138e-01
-1.37347758e+00 3.28998297e-01 6.32256222e+00 1.54276407e+00
-1.62673688e+00 8.93130228e-02 7.25253224e-01 2.34337062e-01
1.49316907e-01 -3.08957696e-01 -6.33144975e-01 6.87748313e-01
1.57776213e+00 6.05422333e-02 1.25442967e-01 3.78065109e-02
1.04941092e-01 1.90916599e-03 -7.65141308e-01 1.52005732e+00
-1.84152752e-01 -1.40925014e+00 3.91761474e-02 3.04731965e-01
4.78516638e-01 -9.67003033e-02 6.00422621e-01 5.68605125e-01
-6.47138178e-01 -1.48601317e+00 1.84647560e-01 8.41616571e-01
1.37724400e+00 -1.07163942e+00 8.82899821e-01 2.82057077e-01
-1.22595155e+00 -4.68596786e-01 -1.56841949e-01 -2.61529446e-01
1.08577022e-02 6.09514773e-01 -5.16776741e-01 9.02495563e-01
-9.29984450e-02 4.93978530e-01 -3.56436938e-01 1.01575494e+00
2.78380126e-01 4.76744175e-01 -9.19886082e-02 -1.17156610e-01
5.33094943e-01 1.10589445e-01 4.76110548e-01 1.23088026e+00
5.72094798e-01 5.75517640e-02 -4.66689944e-01 1.10719517e-01
-4.49065864e-02 -5.26936889e-01 -3.07799071e-01 -3.23374510e-01
4.55121338e-01 1.04546487e+00 -1.11164704e-01 -1.87264279e-01
-4.83807653e-01 7.66887426e-01 -3.99675548e-01 4.12906885e-01
-8.37169170e-01 -1.20182085e+00 7.34249234e-01 -2.90529162e-01
6.18931890e-01 -6.42537475e-01 -2.70149440e-01 -6.06475711e-01
-1.99214742e-01 -1.01339614e+00 3.43104273e-01 -3.85169566e-01
-1.07515764e+00 4.70316440e-01 -6.72323585e-01 -1.49746275e+00
-2.71330506e-01 -7.16362774e-01 -1.76427424e-01 1.26025307e+00
-2.07274699e+00 -8.52703512e-01 1.92175303e-02 5.63412786e-01
3.20268482e-01 -6.31081522e-01 9.13914502e-01 9.18943048e-01
-2.63574213e-01 9.84175861e-01 5.11502802e-01 2.47367486e-01
2.45773613e-01 -8.26149285e-01 5.84096730e-01 6.10906124e-01
1.60910606e-01 4.05369908e-01 2.61054099e-01 1.50282064e-03
-1.67681777e+00 -1.25236058e+00 9.77729201e-01 3.63157481e-01
3.33067775e-01 -1.05604428e-04 -3.80084962e-01 1.41272604e-01
3.54949325e-01 4.56038266e-02 9.27084982e-01 -5.04705846e-01
-1.99013963e-01 -3.01268637e-01 -1.13949931e+00 3.59944105e-01
6.31274164e-01 -7.85058916e-01 -1.55650526e-01 2.29231402e-01
3.03839743e-01 -5.31295419e-01 -1.10286725e+00 8.03605139e-01
5.08015215e-01 -7.62772024e-01 9.20457959e-01 -4.62346226e-01
1.31032150e-02 -4.71769631e-01 -6.04736090e-01 -1.02479708e+00
-2.66456008e-01 -1.07511175e+00 -8.76570046e-01 3.19288760e-01
3.07639331e-01 -7.68717706e-01 4.10572082e-01 -5.33312619e-01
-2.27408618e-01 -1.17492640e+00 -1.53355455e+00 -6.60000324e-01
6.12240583e-02 -5.08658051e-01 6.75579011e-01 5.45629919e-01
-2.87867397e-01 5.31848729e-01 -7.43991196e-01 6.00237489e-01
4.53480542e-01 1.78639680e-01 2.86189258e-01 -7.59229779e-01
-3.81733596e-01 -5.19877136e-01 -1.24890316e+00 -1.54904902e+00
-2.06394091e-01 -8.24103594e-01 -4.74226296e-01 -1.32689130e+00
-3.53659660e-01 -3.80591452e-01 -6.82065308e-01 -2.42395494e-02
3.45522314e-01 4.66271043e-01 -3.12344909e-01 -1.19821839e-01
-5.21149397e-01 8.03526044e-01 1.01791143e+00 -3.62168133e-01
3.05989504e-01 3.05447251e-01 -4.25488532e-01 2.36640230e-01
8.55455935e-01 -1.18715540e-01 1.14879072e-01 -3.58657986e-01
-5.79532534e-02 3.86461437e-01 1.17245704e-01 -1.56243765e+00
2.42160186e-01 3.59227270e-01 7.97857523e-01 -7.71278620e-01
7.80489743e-01 -3.64300162e-01 -1.28365040e-01 7.84383178e-01
-1.52480394e-01 -3.66934448e-01 1.87270507e-01 5.04184842e-01
-5.10137618e-01 -4.81135696e-02 1.14062440e+00 4.73549902e-01
-8.21481586e-01 4.62139040e-01 -8.22864890e-01 -3.86618257e-01
7.71848440e-01 -2.06899226e-01 -4.65656251e-01 -6.21825218e-01
-5.67715943e-01 -2.61040498e-02 -5.67469776e-01 5.54692268e-01
9.97891963e-01 -1.86217070e+00 -7.25483775e-01 6.61610186e-01
-2.13255286e-02 -6.17793858e-01 4.20839965e-01 1.02165365e+00
-4.83281314e-01 1.14365673e+00 -1.96401745e-01 -6.63530469e-01
-1.07168329e+00 5.17852940e-02 7.30872631e-01 -2.76077062e-01
-3.56139272e-01 8.47639024e-01 -7.39406168e-01 1.47984430e-01
3.06658983e-01 -3.22078764e-02 -3.47198129e-01 -1.25947326e-01
7.64851153e-01 3.24881643e-01 3.42647940e-01 -1.05739951e+00
-3.93390149e-01 6.26888812e-01 -7.09546879e-02 -1.53232127e-01
9.25532818e-01 4.88599464e-02 7.60163041e-03 3.18614125e-01
1.90490365e+00 -2.26318434e-01 -6.66367650e-01 -5.31989276e-01
2.36153215e-01 -5.85066378e-01 2.93417752e-01 -6.02403760e-01
-8.29938650e-01 1.30070281e+00 1.21039677e+00 1.83499396e-01
1.27528584e+00 -6.43836260e-01 1.35631537e+00 4.06693727e-01
5.48905611e-01 -1.01264703e+00 4.64776568e-02 7.82114267e-01
4.79042947e-01 -6.65556133e-01 -4.69745463e-03 3.21204245e-01
1.06531695e-01 1.58753717e+00 -8.31092522e-02 1.20584689e-01
8.56181622e-01 2.86955446e-01 8.93203914e-02 -2.01555230e-02
-6.45204246e-01 -2.85341144e-01 3.49637747e-01 5.24371207e-01
4.31096315e-01 1.04221746e-01 -3.78425539e-01 4.97272104e-01
-1.07734904e-01 -2.18173847e-01 4.87521142e-01 9.77655172e-01
-5.56885064e-01 -1.16596699e+00 -2.27579027e-01 7.39869535e-01
-5.75780749e-01 1.82582121e-02 8.54289085e-02 2.93438315e-01
1.83530927e-01 1.13572550e+00 1.17527634e-01 -7.96075046e-01
-2.83926781e-02 -3.49242287e-03 7.70261705e-01 -1.23976968e-01
-7.39244521e-02 -9.39036980e-02 2.75168896e-01 -3.62768531e-01
-1.95629761e-01 -1.77395776e-01 -1.23899758e+00 -1.71879008e-01
-2.14685380e-01 -7.31371939e-02 8.97166193e-01 1.08299971e+00
7.31107414e-01 9.53661501e-01 1.36089897e+00 -5.75008810e-01
-8.25587451e-01 -1.15655708e+00 -6.63017690e-01 -2.38383040e-01
1.10523427e+00 -2.89045066e-01 -4.60029254e-03 -4.29081589e-01] | [6.479672908782959, 1.4907599687576294] |
e42f61c1-2ac0-4c49-b156-dfff597e075c | neural-topic-modeling-with-deep-mutual | 2203.06298 | null | https://arxiv.org/abs/2203.06298v1 | https://arxiv.org/pdf/2203.06298v1.pdf | Neural Topic Modeling with Deep Mutual Information Estimation | The emerging neural topic models make topic modeling more easily adaptable and extendable in unsupervised text mining. However, the existing neural topic models is difficult to retain representative information of the documents within the learnt topic representation. In this paper, we propose a neural topic model which incorporates deep mutual information estimation, i.e., Neural Topic Modeling with Deep Mutual Information Estimation(NTM-DMIE). NTM-DMIE is a neural network method for topic learning which maximizes the mutual information between the input documents and their latent topic representation. To learn robust topic representation, we incorporate the discriminator to discriminate negative examples and positive examples via adversarial learning. Moreover, we use both global and local mutual information to preserve the rich information of the input documents in the topic representation. We evaluate NTM-DMIE on several metrics, including accuracy of text clustering, with topic representation, topic uniqueness and topic coherence. Compared to the existing methods, the experimental results show that NTM-DMIE can outperform in all the metrics on the four datasets. | ['Zheng Zhou', 'Dong Wang', 'Ning Ye', 'Guilin Qi', 'Tongtong Wu', 'Yuan-Fang Li', 'Xiaoqiu Lu', 'Kang Xu'] | 2022-03-12 | null | null | null | null | ['mutual-information-estimation', 'text-clustering', 'topic-models'] | ['methodology', 'natural-language-processing', 'natural-language-processing'] | [-2.11521581e-01 2.28021711e-01 -2.54031420e-01 -5.22884190e-01
-7.33921111e-01 -1.87038869e-01 7.49183714e-01 -3.86635661e-02
6.98539168e-02 5.50309539e-01 3.60524446e-01 1.93666905e-01
-4.24472243e-01 -1.01892948e+00 -4.29846793e-01 -8.53833735e-01
-2.93063164e-01 5.62364459e-01 -5.91605417e-02 2.10778907e-01
1.73501045e-01 -3.98288935e-01 -1.43862820e+00 2.80205727e-01
1.13344991e+00 9.72022593e-01 3.24644983e-01 1.62958831e-01
-4.50031281e-01 3.88465375e-01 -8.96388531e-01 -3.62657011e-02
-3.08367033e-02 -3.61000121e-01 -6.67393982e-01 6.88095316e-02
1.97277233e-01 -2.63382405e-01 -4.34553683e-01 1.02484155e+00
4.18105364e-01 5.34644186e-01 1.13202810e+00 -1.69520175e+00
-6.95595503e-01 9.54684973e-01 -6.63986087e-01 -8.79199877e-02
-1.39133170e-01 -3.30993891e-01 1.01538801e+00 -8.48311305e-01
6.75581574e-01 1.51017332e+00 5.70391119e-01 3.52337599e-01
-1.07839108e+00 -1.22301531e+00 5.37065089e-01 8.41044635e-02
-1.49099898e+00 1.07899621e-01 1.01992607e+00 -5.58584213e-01
3.48463237e-01 -3.24698277e-02 3.35986167e-01 1.19656146e+00
4.63686734e-01 1.12995028e+00 9.64259326e-01 6.00413382e-02
3.54333490e-01 5.26422620e-01 5.61271012e-01 1.59533426e-01
1.52528360e-01 -1.90476447e-01 -6.71153545e-01 -3.80482525e-01
5.50085962e-01 3.50710750e-01 -2.57904917e-01 -4.28744853e-01
-1.13297319e+00 1.23975241e+00 4.64983672e-01 3.62892777e-01
-2.82875001e-01 -3.23824100e-02 4.88281250e-01 2.28338853e-01
1.15170145e+00 2.00170442e-01 -3.66285682e-01 2.52139747e-01
-1.29383016e+00 2.78683692e-01 7.50277996e-01 9.96325195e-01
9.37280357e-01 8.86663795e-02 -2.38160074e-01 7.92330265e-01
5.93246162e-01 4.46667254e-01 1.21227527e+00 -6.50066137e-01
4.20966357e-01 7.30750620e-01 -4.23097640e-01 -1.43044126e+00
-1.72279283e-01 -6.01075292e-01 -1.22330761e+00 -1.28201291e-01
-3.41203541e-01 -4.06297386e-01 -8.88395727e-01 1.92713010e+00
3.53842109e-01 2.88346231e-01 4.44158226e-01 6.21196747e-01
1.10247183e+00 1.21995330e+00 1.29768878e-01 -1.92357481e-01
1.03335726e+00 -7.24242568e-01 -9.72325325e-01 -1.40298188e-01
3.79970342e-01 -3.95925850e-01 8.89702320e-01 3.51831496e-01
-5.77312887e-01 -3.82214934e-01 -9.64663744e-01 1.53065771e-01
-5.91311157e-01 8.93367156e-02 6.97732329e-01 5.36281109e-01
-8.92015934e-01 3.65420371e-01 -9.17032361e-01 -1.89613223e-01
5.76027989e-01 5.14403403e-01 -3.11662048e-01 2.04604715e-02
-1.53229737e+00 1.45371825e-01 9.91830885e-01 -5.25261700e-01
-1.00190544e+00 -8.25002789e-01 -7.90311992e-01 4.13037449e-01
1.48977861e-01 -6.08505666e-01 8.88807476e-01 -6.88393295e-01
-1.33111250e+00 5.13349593e-01 -1.97101057e-01 -6.14458501e-01
2.18712568e-01 -2.50685811e-01 -2.16598228e-01 5.90707059e-04
4.81825769e-01 1.06091511e+00 7.61060059e-01 -1.38526845e+00
-5.36856055e-01 -3.88932467e-01 -5.69929957e-01 2.60600895e-01
-8.54287624e-01 -2.55021334e-01 -3.62938315e-01 -1.03993940e+00
5.71198165e-01 -6.19480252e-01 -5.89799173e-02 -2.39628643e-01
-7.12461710e-01 -7.15953052e-01 1.65490580e+00 -6.45206690e-01
1.07882464e+00 -2.20409417e+00 -1.56850833e-02 2.58908540e-01
3.38587910e-01 -1.86557963e-01 2.54363090e-01 1.82559654e-01
4.47111614e-02 2.51080602e-01 -3.52999210e-01 -4.93139952e-01
2.39349484e-01 1.83877628e-02 -8.01585078e-01 2.94675618e-01
-5.07603064e-02 6.37671709e-01 -5.22923291e-01 -5.98995805e-01
3.60189304e-02 6.38031840e-01 -4.98054981e-01 2.46440813e-01
-4.82762367e-01 1.79043099e-01 -5.95440567e-01 1.22779891e-01
7.88459301e-01 -4.25907999e-01 -1.67663079e-02 1.58964217e-01
3.36026400e-01 4.45707083e-01 -1.20292711e+00 1.78272510e+00
-2.53943294e-01 1.01974928e+00 -2.63121605e-01 -8.89187694e-01
1.34375691e+00 5.11250973e-01 6.60458624e-01 -1.68085456e-01
1.72715709e-01 -1.85788259e-01 -2.79694766e-01 4.63480279e-02
7.58850574e-01 -1.41841426e-01 -1.49278015e-01 1.09020674e+00
4.05427963e-01 1.34777531e-01 -2.54605025e-01 5.41271329e-01
3.43683481e-01 -3.49516153e-01 6.31089061e-02 -6.19048297e-01
-1.69129968e-02 -2.96904951e-01 6.13799632e-01 6.47980154e-01
7.23328069e-02 6.47502601e-01 4.26523894e-01 -8.48429576e-02
-8.47223401e-01 -1.08361816e+00 -3.65124732e-01 1.21263063e+00
1.93110347e-01 -4.32070851e-01 -7.44918942e-01 -6.99778438e-01
-1.34548455e-01 7.65122294e-01 -7.98406363e-01 -5.58037579e-01
-4.75878939e-02 -7.40023613e-01 4.03574467e-01 3.72409612e-01
8.02520633e-01 -8.39810371e-01 -1.79556310e-01 1.57164365e-01
-3.03988993e-01 -4.29982752e-01 -4.31488335e-01 1.35384575e-01
-1.05336273e+00 -4.92457122e-01 -9.31256711e-01 -7.21391261e-01
3.86688739e-01 4.12221789e-01 7.75608122e-01 -5.85312665e-01
1.10668913e-02 1.70346677e-01 -2.61399418e-01 -5.29264510e-01
-3.67094986e-02 5.18535852e-01 8.70835111e-02 -8.77790302e-02
6.59945726e-01 -7.23654568e-01 -5.70351005e-01 3.19449455e-01
-1.04951406e+00 2.29747090e-02 3.52765411e-01 8.07107389e-01
3.85755926e-01 6.77025259e-01 8.72003078e-01 -9.81338501e-01
9.07728255e-01 -1.04433596e+00 -5.43398447e-02 -3.07228025e-02
-9.12285328e-01 -5.37127294e-02 8.73817801e-02 -6.92741632e-01
-1.25638974e+00 -6.07174754e-01 2.85662442e-01 -4.74038929e-01
-2.03454033e-01 7.01460183e-01 -5.05502284e-01 8.42064023e-01
3.67505670e-01 5.48430443e-01 -1.91256166e-01 -4.41921711e-01
4.97740120e-01 9.14140761e-01 5.30352235e-01 -4.89202529e-01
3.82243305e-01 6.31328106e-01 -6.74585998e-01 -6.75584733e-01
-8.18979919e-01 -6.89502954e-01 -3.77186984e-01 1.02520511e-01
7.75908709e-01 -1.12157106e+00 -2.14047879e-01 4.28120226e-01
-1.26584351e+00 -4.01900634e-02 -2.51193136e-01 7.01890767e-01
-5.27414262e-01 2.11355701e-01 -4.14287210e-01 -7.65782654e-01
-7.73380220e-01 -9.37005162e-01 1.14205885e+00 2.12840602e-01
-2.71657884e-01 -1.31593192e+00 1.56726599e-01 -2.07624674e-01
3.12182724e-01 1.11853495e-01 9.71754253e-01 -1.30852175e+00
-3.78967285e-01 -1.13923684e-01 -2.89932728e-01 2.79077917e-01
3.12069952e-01 -2.83011526e-01 -1.09063065e+00 -3.98889005e-01
4.13927943e-01 -1.13356717e-01 1.24244297e+00 6.16310477e-01
1.15410495e+00 -7.08641350e-01 -7.06730127e-01 6.39319658e-01
1.04782081e+00 1.79191694e-01 5.65269589e-01 5.58291912e-01
5.23206949e-01 7.26393223e-01 6.02708876e-01 3.60786378e-01
3.96967947e-01 2.01913714e-01 1.27265766e-01 9.96789038e-02
3.13865364e-01 -3.01987112e-01 2.93530464e-01 1.04501915e+00
6.04045093e-01 -5.53387344e-01 -9.35847998e-01 6.10238850e-01
-1.87884092e+00 -8.53412271e-01 2.01757669e-01 1.75685251e+00
9.14377689e-01 1.65512592e-01 -1.00158781e-01 2.18836218e-03
1.19454813e+00 1.83245853e-01 -7.31454670e-01 1.99031144e-01
-1.18761368e-01 -2.79792517e-01 -1.13888755e-01 5.32114469e-02
-1.50452590e+00 1.04814339e+00 6.03094769e+00 1.07610691e+00
-1.09752405e+00 3.95224541e-01 8.46928596e-01 -9.36809778e-02
-3.92784834e-01 -3.33411098e-01 -8.43356013e-01 7.70853460e-01
8.69051397e-01 -6.65454507e-01 -4.93773550e-01 1.41804254e+00
-6.69862628e-02 3.00810486e-01 -9.25614417e-01 6.87670171e-01
1.93656236e-01 -1.17387092e+00 4.10822153e-01 3.23180407e-01
1.26773584e+00 -2.12469324e-02 5.85260153e-01 5.70490122e-01
6.23169065e-01 -9.11159039e-01 2.26207405e-01 5.75599372e-01
3.93022001e-01 -1.07139885e+00 8.46439302e-01 4.92093861e-01
-8.46598625e-01 1.78436816e-01 -7.25872695e-01 3.88730973e-01
-1.12863421e-01 8.17875981e-01 -8.51584136e-01 4.49462026e-01
8.56738687e-01 9.11044836e-01 -2.05391601e-01 9.58720505e-01
1.27850249e-01 8.95847559e-01 -4.76426691e-01 7.36980513e-02
3.50836545e-01 -2.17058942e-01 7.80273736e-01 1.12301362e+00
4.27643865e-01 -3.53895068e-01 3.80079478e-01 1.21949983e+00
-2.86747873e-01 2.59246022e-01 -6.24359906e-01 1.36172116e-01
5.39938211e-01 9.83348072e-01 -8.06927741e-01 -5.09684265e-01
1.83419466e-01 7.73919880e-01 2.94714700e-02 4.02663171e-01
-3.80374670e-01 -5.27874291e-01 7.02299178e-01 -2.60837495e-01
2.44322002e-01 -1.45082414e-01 -3.38246405e-01 -1.09785676e+00
-1.41525641e-01 -6.36279821e-01 5.36160171e-01 -6.60198331e-01
-1.44561708e+00 6.49433374e-01 3.07820290e-01 -1.13662517e+00
-4.85555261e-01 -1.84653118e-01 -1.14326477e+00 8.06450307e-01
-1.05874121e+00 -1.23956692e+00 -1.65898874e-01 5.04846632e-01
7.29131222e-01 -6.70284867e-01 6.21091008e-01 -9.89972576e-02
-4.73346144e-01 6.91209912e-01 7.66770124e-01 1.68596860e-02
8.24184060e-01 -1.34656525e+00 4.06134397e-01 4.22558904e-01
2.27747187e-02 1.05337036e+00 5.42919993e-01 -8.48060310e-01
-5.25183141e-01 -1.54046237e+00 3.95897269e-01 -1.54606611e-01
4.85445440e-01 -6.40089691e-01 -1.42709613e+00 9.03472841e-01
4.78876144e-01 -8.14249277e-01 1.12516987e+00 3.34107786e-01
-4.59547073e-01 1.26713097e-01 -8.88650119e-01 3.34806561e-01
2.25785047e-01 -4.74281818e-01 -8.60273778e-01 4.97222483e-01
1.45679247e+00 1.34300768e-01 -9.58462536e-01 2.64367163e-01
3.53336304e-01 -5.07695615e-01 7.75070429e-01 -6.72789514e-01
5.61101317e-01 -2.54621599e-02 -2.66933501e-01 -1.42533302e+00
-9.94240046e-02 -3.83597672e-01 -2.30782077e-01 1.80620623e+00
3.29506546e-01 -5.80169082e-01 1.09604025e+00 4.11501616e-01
8.66856873e-02 -6.79921865e-01 -9.82980967e-01 -6.00369632e-01
6.53296411e-01 -4.16000307e-01 8.86310697e-01 1.47103536e+00
9.98734534e-02 4.54366922e-01 -4.30399835e-01 9.92913097e-02
5.43849289e-01 3.40681255e-01 6.73342526e-01 -1.61532223e+00
-1.79208424e-02 -4.85400349e-01 -3.87922317e-01 -1.07671034e+00
5.25203586e-01 -1.01271939e+00 2.72663403e-02 -1.60410929e+00
5.65046549e-01 -3.52886260e-01 -4.17055666e-01 2.05030471e-01
-3.24144334e-01 -2.54642367e-01 -5.99270687e-02 7.15793610e-01
-8.66234243e-01 1.22547710e+00 8.37260783e-01 -4.36012447e-01
-3.31749797e-01 1.79645255e-01 -8.99810731e-01 8.34360063e-01
1.01444054e+00 -7.48125315e-01 -5.19190252e-01 -1.25687242e-01
-1.37176931e-01 -3.16820651e-01 1.00707196e-01 -1.08027422e+00
5.65059185e-01 8.43879655e-02 5.51092744e-01 -1.32304502e+00
3.51509959e-01 -6.02284908e-01 -2.55947828e-01 2.22530916e-01
-8.48699450e-01 -4.34079617e-01 8.84504393e-02 9.57974315e-01
-3.96026194e-01 -1.67648599e-01 4.51430529e-01 -6.37942776e-02
-4.58558559e-01 5.94389617e-01 -5.32392979e-01 -4.59564812e-02
8.75268102e-01 1.19499452e-02 -3.99054945e-01 -7.12318480e-01
-6.60312355e-01 4.73537266e-01 1.31969422e-01 5.10165334e-01
6.33957207e-01 -1.51473451e+00 -7.23066628e-01 6.19516820e-02
3.91797200e-02 3.60255271e-01 4.87963647e-01 1.53821364e-01
1.98863581e-01 6.92977786e-01 -2.06386726e-02 -8.15127671e-01
-8.24033082e-01 5.81346691e-01 6.64529353e-02 -3.48715872e-01
-5.99915922e-01 6.40844166e-01 1.00571251e+00 -9.14959311e-01
3.90307277e-01 2.21265405e-02 -3.56935829e-01 3.35089453e-02
4.09040511e-01 2.47711763e-01 -2.67511308e-01 -3.29390466e-01
4.64873053e-02 1.67853892e-01 -6.87174976e-01 -2.18159407e-01
1.22099626e+00 -1.96284056e-01 -1.89942956e-01 7.18138039e-01
1.60259318e+00 -5.91056049e-01 -1.16537797e+00 -5.38942456e-01
1.24008646e-02 -3.67575558e-03 3.96369636e-01 -4.22943622e-01
-9.56086695e-01 1.19362092e+00 7.76953518e-01 2.99971282e-01
8.42104912e-01 2.30233207e-01 7.99720764e-01 6.07578516e-01
-2.14715406e-01 -9.88424599e-01 3.48040104e-01 4.57288861e-01
7.24902630e-01 -1.28494930e+00 7.72160962e-02 -3.33929211e-01
-6.12530112e-01 7.49533594e-01 8.34135532e-01 -1.96318388e-01
1.07608426e+00 -2.39842162e-01 -3.08839846e-02 -4.43322957e-01
-9.21027541e-01 2.11606979e-01 4.92630303e-01 6.79091811e-01
4.30422455e-01 5.19802608e-02 -3.24173234e-02 1.10767901e+00
-6.28107488e-01 -5.68831444e-01 7.53971562e-02 5.56029975e-01
-7.05465496e-01 -6.77766979e-01 -3.41559947e-01 7.59682119e-01
-4.93537486e-01 -1.50984958e-01 -4.95090246e-01 8.69399607e-01
-1.48899838e-01 7.63549447e-01 4.46107149e-01 -5.25133729e-01
-3.84756416e-01 1.94755003e-01 -6.09967053e-01 -8.26978564e-01
-1.64328694e-01 2.49768347e-01 -7.07193851e-01 9.46695954e-02
-2.19158679e-01 -6.87487483e-01 -1.06305349e+00 5.61712906e-02
-7.92521477e-01 6.68469191e-01 8.96302521e-01 9.53899443e-01
5.70032537e-01 6.62864089e-01 6.88451946e-01 -5.94340026e-01
-3.39446127e-01 -1.48063707e+00 -7.99090743e-01 1.39494523e-01
-9.08539444e-03 -8.18920195e-01 -3.23517561e-01 1.02722049e-01] | [10.394994735717773, 6.921274662017822] |
752ca3a8-40f7-4c2f-8367-f0e782b067b4 | a-preference-aware-meta-optimization | 2306.14421 | null | https://arxiv.org/abs/2306.14421v1 | https://arxiv.org/pdf/2306.14421v1.pdf | A Preference-aware Meta-optimization Framework for Personalized Vehicle Energy Consumption Estimation | Vehicle Energy Consumption (VEC) estimation aims to predict the total energy required for a given trip before it starts, which is of great importance to trip planning and transportation sustainability. Existing approaches mainly focus on extracting statistically significant factors from typical trips to improve the VEC estimation. However, the energy consumption of each vehicle may diverge widely due to the personalized driving behavior under varying travel contexts. To this end, this paper proposes a preference-aware meta-optimization framework Meta-Pec for personalized vehicle energy consumption estimation. Specifically, we first propose a spatiotemporal behavior learning module to capture the latent driver preference hidden in historical trips. Moreover, based on the memorization of driver preference, we devise a selection-based driving behavior prediction module to infer driver-specific driving patterns on a given route, which provides additional basis and supervision signals for VEC estimation. Besides, a driver-specific meta-optimization scheme is proposed to enable fast model adaption by learning and sharing transferable knowledge globally. Extensive experiments on two real-world datasets show the superiority of our proposed framework against ten numerical and data-driven machine learning baselines. The source code is available at https://github.com/usail-hkust/Meta-Pec. | ['Hao liu', 'Weijia Zhang', 'Siqi Lai'] | 2023-06-26 | null | null | null | null | ['total-energy', 'memorization'] | ['miscellaneous', 'natural-language-processing'] | [-2.20566586e-01 -3.03145468e-01 -8.50401163e-01 -7.21073747e-01
-8.69108975e-01 -2.20504582e-01 2.92188406e-01 2.37397602e-04
-2.84576744e-01 5.68644345e-01 3.43620211e-01 -2.40179852e-01
-3.70643616e-01 -9.03985739e-01 -8.49112630e-01 -8.88630688e-01
2.22799182e-01 4.63612229e-02 -1.81636363e-01 -2.35835239e-01
2.12241873e-01 1.68339565e-01 -1.57698810e+00 -1.49620950e-01
1.26594687e+00 8.38383198e-01 5.82817972e-01 1.36439532e-01
1.39341220e-01 4.04611081e-01 2.15320066e-01 -3.54861140e-01
2.68672816e-02 -1.84901208e-01 -2.98538774e-01 1.31295575e-02
-2.43309334e-01 -1.93175197e-01 -7.87997782e-01 7.55659163e-01
3.52808714e-01 6.55276239e-01 5.72025657e-01 -1.67570007e+00
-2.99508244e-01 6.84118092e-01 -1.90324187e-01 9.47462544e-02
-1.53528363e-01 5.29631376e-01 7.93958843e-01 -5.52001536e-01
1.94824003e-02 8.24930072e-01 3.62441957e-01 4.65090036e-01
-8.76267910e-01 -7.95612395e-01 5.39439082e-01 1.03808272e+00
-1.60813689e+00 -4.87351745e-01 1.25238121e+00 -8.51499811e-02
7.35973120e-01 4.18316931e-01 7.99031854e-01 8.69366884e-01
2.63788193e-01 1.15244853e+00 7.04304218e-01 1.19426534e-01
3.08266222e-01 5.35216033e-01 1.64081782e-01 5.29161394e-01
4.47297730e-02 1.08268164e-01 -4.68111455e-01 2.33641684e-01
-8.17473829e-02 3.01249355e-01 -4.70432341e-02 -1.51807815e-01
-1.01692939e+00 5.17038405e-01 5.23302853e-01 -7.06112012e-02
-6.20200694e-01 3.00478339e-01 3.48247975e-01 -1.15485013e-01
4.38887596e-01 -1.29207581e-01 -5.87979376e-01 -3.48339498e-01
-7.78436005e-01 2.46895269e-01 4.01663989e-01 1.05454111e+00
1.37515092e+00 9.48955864e-03 -3.59360695e-01 7.84257352e-01
3.22537631e-01 6.12492621e-01 5.09888172e-01 -9.85294938e-01
6.12127662e-01 7.45360255e-01 2.03883573e-01 -1.14068532e+00
-3.40877026e-01 -4.76689190e-01 -7.63375998e-01 -4.66185123e-01
-2.19370112e-01 -2.95874923e-01 -7.22971261e-01 1.80353117e+00
3.97480160e-01 6.73834920e-01 -1.21108547e-01 9.14745688e-01
4.52623129e-01 9.65524733e-01 3.41969162e-01 -1.86774656e-01
1.15250027e+00 -1.08064163e+00 -6.75755680e-01 -3.39726597e-01
8.33444059e-01 -7.58955553e-02 9.49013829e-01 8.20546150e-02
-8.79263163e-01 -5.53168535e-01 -8.87788594e-01 1.15961224e-01
-5.82129955e-01 2.98244298e-01 6.20004356e-01 7.82594085e-01
-5.42065024e-01 2.12914094e-01 -9.04302418e-01 -1.66856237e-02
4.66500610e-01 3.60862851e-01 1.60802037e-01 -1.79782107e-01
-1.29090023e+00 8.13037872e-01 6.32652521e-01 3.03118795e-01
-9.09138560e-01 -1.06339586e+00 -8.22777212e-01 1.01092495e-01
5.02875209e-01 -5.88679552e-01 1.05868876e+00 -5.81887126e-01
-1.53443241e+00 1.11189596e-01 -6.42189324e-01 -3.31091374e-01
3.92982602e-01 1.36373967e-01 -8.88659179e-01 -4.44607824e-01
-1.10927917e-01 4.91235316e-01 6.22750580e-01 -1.13253903e+00
-1.09040475e+00 -1.80943578e-01 -6.08412474e-02 3.44941407e-01
-6.21963739e-01 -6.30460620e-01 -8.56635332e-01 -2.81314224e-01
-4.43624914e-01 -1.11466229e+00 -3.08159441e-01 -6.36804283e-01
-3.91164511e-01 -5.45982242e-01 7.20221043e-01 -5.29619336e-01
1.67939961e+00 -2.11307025e+00 1.26399711e-01 4.15888458e-01
-1.13908075e-01 -2.30765939e-02 -1.70620680e-02 4.33734059e-01
3.70000303e-01 -4.59324867e-02 -2.00007647e-01 -4.45850432e-01
3.12228799e-01 3.02895516e-01 4.11948003e-03 4.24425453e-01
-3.43228392e-02 1.03908658e+00 -9.06208873e-01 -5.74119806e-01
6.62930906e-01 5.31210124e-01 -5.16478717e-01 1.86299115e-01
-1.05906598e-01 3.56318682e-01 -8.12607706e-01 4.92896855e-01
8.40791583e-01 7.82390404e-03 -1.25552472e-02 -4.67050731e-01
-3.51283312e-01 2.26253539e-01 -1.12284410e+00 1.76271129e+00
-1.06434608e+00 4.63962346e-01 -2.89665401e-01 -1.13723481e+00
6.10471189e-01 -1.12333205e-02 5.62395692e-01 -1.06436491e+00
1.47347867e-01 9.42405909e-02 -4.12138045e-01 -6.04017496e-01
6.80763781e-01 3.80309463e-01 -3.92701864e-01 2.74649531e-01
-2.92832583e-01 3.93566817e-01 2.36880779e-01 -3.52575779e-02
5.94329655e-01 6.11115396e-02 -7.47547522e-02 -2.45645374e-01
8.31392467e-01 2.58643236e-02 9.44990456e-01 5.88840395e-02
-2.85395741e-01 -1.05205461e-01 -4.02155705e-02 -3.39759231e-01
-6.20788693e-01 -7.97832429e-01 2.10008975e-02 1.08513272e+00
7.21702695e-01 -3.43228430e-01 -7.41357386e-01 -3.50393116e-01
4.83605191e-02 1.36291146e+00 -6.10362351e-01 -6.66589558e-01
-6.71115875e-01 -9.74650025e-01 2.91268397e-02 4.64122027e-01
6.96115315e-01 -6.70598447e-01 -5.09949386e-01 1.31317288e-01
-6.52296484e-01 -8.78408432e-01 -9.15487051e-01 -1.54630974e-01
-5.06537855e-01 -6.43395305e-01 -2.55402386e-01 -5.72305262e-01
6.32390082e-01 5.62646806e-01 7.10724711e-01 8.12468305e-02
9.52388272e-02 3.75175655e-01 -5.35015129e-02 -3.97910178e-01
-4.69030403e-02 6.17846549e-01 -7.22940685e-03 4.81905431e-01
8.43401492e-01 -4.45598453e-01 -1.21122205e+00 6.27602100e-01
-3.91854316e-01 3.46408904e-01 6.32982790e-01 4.19887394e-01
8.24167073e-01 4.78976160e-01 5.91441631e-01 -3.67769361e-01
3.98385704e-01 -1.21448016e+00 -4.39139277e-01 2.54431188e-01
-9.98939395e-01 8.23038593e-02 6.87119365e-01 -2.53755957e-01
-1.38740885e+00 8.95324200e-02 3.11905555e-02 -1.66532889e-01
-1.76010817e-01 4.44155723e-01 -7.13813961e-01 2.89178550e-01
-2.72935778e-02 6.88795626e-01 -3.20434898e-01 -3.24689716e-01
4.85856891e-01 7.64642358e-01 4.00986373e-01 -3.82562518e-01
7.79202163e-01 3.11994493e-01 -9.45771411e-02 -5.12403727e-01
-4.28222030e-01 -5.18113554e-01 -2.46879116e-01 -4.44890738e-01
5.91124415e-01 -1.18975210e+00 -9.20562625e-01 3.53726655e-01
-6.21431768e-01 -5.93400598e-01 1.25995159e-01 6.18736923e-01
-6.02171004e-01 -8.86358246e-02 1.09005675e-01 -8.94683003e-01
-1.28576681e-01 -1.21338904e+00 6.56606257e-01 5.41557193e-01
2.67416000e-01 -1.16892564e+00 9.73627716e-02 4.67695236e-01
3.57964844e-01 3.41810435e-02 7.16824293e-01 -1.01178847e-01
-8.55433047e-01 -9.81908068e-02 -6.19852990e-02 1.40680045e-01
3.76397580e-01 -1.82326168e-01 -7.71144986e-01 -1.97921544e-01
-4.87038553e-01 1.46209598e-01 8.63779366e-01 4.90599513e-01
1.60758829e+00 -4.59372967e-01 -8.60417724e-01 7.50221193e-01
1.54387438e+00 2.51010478e-01 4.95288551e-01 2.26164088e-01
8.29578996e-01 7.72976995e-01 8.71789634e-01 4.90301371e-01
1.39450538e+00 8.48626196e-01 4.34920073e-01 2.05758125e-01
6.93142321e-03 -5.62175810e-01 4.33829874e-01 9.16164517e-01
3.15937325e-02 -4.73256826e-01 -7.27256298e-01 1.03517497e+00
-2.12703395e+00 -9.47732031e-01 -7.01134875e-02 2.12864971e+00
4.92304265e-01 -4.89744842e-02 2.61183649e-01 -1.27649456e-01
5.55345118e-01 9.09996629e-02 -1.07352698e+00 -3.21207047e-01
2.47272670e-01 -5.57629168e-01 9.45392609e-01 3.83169353e-01
-7.07223594e-01 7.26496279e-01 4.64723539e+00 1.23681915e+00
-1.04914200e+00 4.39677238e-01 6.46446943e-01 -3.22921753e-01
-8.05252314e-01 -2.69444376e-01 -9.08812344e-01 1.00663126e+00
1.37223613e+00 -7.70617545e-01 7.83594191e-01 6.82732224e-01
8.36899281e-01 7.09154084e-02 -7.95519173e-01 1.00974572e+00
-2.65602410e-01 -1.24440479e+00 -1.11010693e-01 6.01310544e-02
7.34781086e-01 2.12555081e-01 1.74288273e-01 4.56957191e-01
1.27334520e-01 -7.85469890e-01 5.26733160e-01 8.58329654e-01
4.42054242e-01 -1.36620605e+00 4.44233775e-01 5.74425101e-01
-1.62916124e+00 -3.72075140e-01 -3.44120413e-01 2.81422734e-01
3.90187532e-01 4.97678697e-01 -5.67748427e-01 8.63379776e-01
6.63447618e-01 1.02265382e+00 -4.24429059e-01 8.44051242e-01
2.76589040e-02 8.50908875e-01 -2.50769734e-01 -1.52633384e-01
2.38229960e-01 -2.58044362e-01 3.59414488e-01 1.27977586e+00
4.53354299e-01 1.90688238e-01 -6.68457896e-02 7.16674984e-01
-1.48600891e-01 2.01075241e-01 -2.45733202e-01 2.54891902e-01
7.72297502e-01 1.48403108e+00 -2.36922354e-01 -7.28160664e-02
-4.52719897e-01 7.47729659e-01 1.14986844e-01 5.02483368e-01
-1.25349176e+00 -4.03447300e-01 1.08712280e+00 -1.03621617e-01
3.02873045e-01 -1.39083892e-01 -2.95047581e-01 -1.01545215e+00
5.64206354e-02 -1.52003840e-01 1.92779452e-01 -4.23405677e-01
-8.41614664e-01 1.78135589e-01 2.78292477e-01 -1.35837352e+00
-1.16476543e-01 -2.23193660e-01 -9.65307951e-01 7.43259788e-01
-2.07737136e+00 -1.20785820e+00 -7.08367586e-01 6.99115753e-01
7.59765744e-01 -1.28534555e-01 2.64770508e-01 6.01308644e-01
-1.16228259e+00 9.37349200e-01 2.49795362e-01 -3.50253195e-01
5.10725677e-01 -8.83074880e-01 4.19377059e-01 7.25338876e-01
-3.32583487e-01 2.81520665e-01 6.43216014e-01 -3.99771154e-01
-2.00770831e+00 -1.68980861e+00 6.90982461e-01 -4.17510808e-01
3.17732513e-01 -1.32054135e-01 -8.26584518e-01 5.11803269e-01
1.32067725e-01 -3.34763825e-01 7.37650096e-01 -1.79349288e-01
1.57599166e-01 -7.61247039e-01 -1.00084591e+00 6.74587011e-01
1.09696817e+00 -3.46156418e-01 -4.58346568e-02 1.89786792e-01
6.62814856e-01 -1.11380026e-01 -8.13549519e-01 1.54340774e-01
6.69433117e-01 -5.54938912e-01 8.72654736e-01 -1.99119046e-01
1.36430943e-02 -3.48488212e-01 -9.18158591e-02 -1.62450469e+00
-5.41359067e-01 -4.54748154e-01 -3.60150397e-01 1.24023902e+00
3.89697731e-01 -7.63846874e-01 7.42397189e-01 9.76127207e-01
-3.68127406e-01 -7.86967397e-01 -8.56321871e-01 -6.02154374e-01
-1.00579336e-01 -8.61559629e-01 1.31969559e+00 5.24113774e-01
-7.26950243e-02 -1.16013542e-01 -5.81000865e-01 4.23048526e-01
8.29218626e-01 1.92733601e-01 8.63499820e-01 -6.87274218e-01
5.57194836e-02 -5.45575917e-01 -1.36565417e-01 -1.06106186e+00
5.21850705e-01 -1.02751231e+00 1.53730109e-01 -1.53738105e+00
2.18603238e-01 -5.18156946e-01 -9.33091938e-01 3.33543539e-01
-2.59012520e-01 -7.34162182e-02 -1.57424808e-01 -1.15496386e-02
-6.00070953e-01 9.19897556e-01 8.79881442e-01 -3.94046724e-01
-4.28713411e-01 2.18316421e-01 -8.56021225e-01 2.48667166e-01
1.30522120e+00 -4.76185203e-01 -8.87006402e-01 -4.35281068e-01
1.61909595e-01 -1.40051544e-01 3.50912094e-01 -7.99106121e-01
5.44142306e-01 -7.66463697e-01 -1.04900278e-01 -8.02570462e-01
1.91754371e-01 -9.70132291e-01 4.58581507e-01 4.67648327e-01
-1.92476422e-01 -9.26483274e-02 2.32683882e-01 9.33843374e-01
4.08242121e-02 1.95505977e-01 3.20717484e-01 2.69704908e-01
-1.32136810e+00 8.09825778e-01 -4.32367265e-01 -2.31268510e-01
1.23853743e+00 -1.77685201e-01 -6.89920112e-02 -2.44431317e-01
-1.28175870e-01 1.06613016e+00 3.51675123e-01 8.16497087e-01
5.15340626e-01 -1.63480103e+00 -7.95752466e-01 7.50683472e-02
4.42462683e-01 -3.45897466e-01 9.49629843e-01 9.69268680e-01
1.55301973e-01 6.38874888e-01 -1.48847669e-01 -3.67083639e-01
-8.27876627e-01 6.06243908e-01 4.03793335e-01 5.48176691e-02
-3.17055911e-01 3.61980498e-01 7.33216247e-03 -4.05889332e-01
-8.04750249e-02 -2.34330878e-01 -1.38471588e-01 -6.65572137e-02
4.55224127e-01 9.90686178e-01 6.74486905e-02 -7.49967754e-01
-6.58228576e-01 4.77642804e-01 2.31163859e-01 2.34242007e-01
1.13980746e+00 -8.33944798e-01 3.69194835e-01 4.44840580e-01
1.38250232e+00 -2.46958002e-01 -1.54163539e+00 -2.97273755e-01
-2.81315953e-01 -4.97988671e-01 6.24338925e-01 -6.93887770e-01
-1.32099652e+00 6.99005067e-01 7.13930428e-01 -2.34459534e-01
1.43551219e+00 -2.15108931e-01 1.29455984e+00 3.24942559e-01
4.41183895e-01 -1.43586063e+00 -5.16164184e-01 -1.78464517e-01
4.58828896e-01 -1.48554003e+00 -2.68558830e-01 -9.04121101e-02
-7.49609888e-01 7.17100322e-01 6.28771424e-01 1.66935280e-01
8.79255950e-01 -2.57523656e-01 -2.52089500e-01 1.20695591e-01
-9.56373453e-01 -1.80151194e-01 5.60016274e-01 3.11527252e-01
-8.21126848e-02 5.33952773e-01 -2.48184457e-01 7.36649692e-01
-5.59061915e-02 1.97027639e-01 2.63839930e-01 6.24732971e-01
-3.24741423e-01 -1.01540792e+00 2.70876110e-01 4.88106072e-01
2.53012031e-01 3.07926178e-01 3.95458817e-01 3.83435190e-01
3.18829417e-01 1.17061234e+00 1.32196490e-02 -8.20566356e-01
4.18047190e-01 7.65202269e-02 8.49095359e-02 -1.23669565e-01
-2.38659918e-01 -3.79027396e-01 2.46370607e-03 -8.09562802e-01
-4.93169338e-01 -8.72251630e-01 -1.29954910e+00 -6.99866235e-01
-5.02575003e-02 2.82059938e-01 8.43665361e-01 9.66780901e-01
7.57343352e-01 6.48120165e-01 1.19496024e+00 -8.05511117e-01
-5.75023554e-02 -5.49905896e-01 -3.33658516e-01 1.57020852e-01
5.10074139e-01 -7.37462103e-01 -2.48467445e-01 2.18447652e-02] | [6.226416110992432, 1.8604621887207031] |
3f40446c-7090-4e8d-aa32-9bca1c05f509 | identification-explanation-and-clinical | 2301.08019 | null | https://arxiv.org/abs/2301.08019v1 | https://arxiv.org/pdf/2301.08019v1.pdf | Identification, explanation and clinical evaluation of hospital patient subtypes | We present a pipeline in which unsupervised machine learning techniques are used to automatically identify subtypes of hospital patients admitted between 2017 and 2021 in a large UK teaching hospital. With the use of state-of-the-art explainability techniques, the identified subtypes are interpreted and assigned clinical meaning. In parallel, clinicians assessed intra-cluster similarities and inter-cluster differences of the identified patient subtypes within the context of their clinical knowledge. By confronting the outputs of both automatic and clinician-based explanations, we aim to highlight the mutual benefit of combining machine learning techniques with clinical expertise. | ['Raul Santos-Rodriguez', 'Christopher J. McWilliams', 'Alexander Hepburn', 'Christopher P. Bourdeaux', 'Michael Ambler', 'Ranjeet S. Bhamber', 'Jeffrey N. Clark', 'Enrico Werner'] | 2023-01-19 | null | null | null | null | ['clinical-knowledge'] | ['miscellaneous'] | [ 9.12139267e-02 9.66951013e-01 -1.73846275e-01 -5.38889468e-01
-9.47830200e-01 -4.61179167e-01 4.22205240e-01 1.10263884e+00
-2.86634535e-01 4.57197398e-01 8.81655276e-01 -6.36379659e-01
-9.67788935e-01 -2.23882586e-01 4.65064831e-02 -5.77377319e-01
-8.74271542e-02 1.22747517e+00 -4.73800063e-01 2.35240206e-01
3.19397837e-01 4.73418564e-01 -1.31046331e+00 1.23567367e+00
7.52910614e-01 5.33400595e-01 -2.04331592e-01 8.16720128e-01
-7.64352083e-03 1.29595041e+00 -3.10470790e-01 -3.42990935e-01
1.57457665e-02 -6.40920818e-01 -1.24620509e+00 6.84854016e-02
-1.78618152e-02 4.46311273e-02 -1.62898332e-01 1.77793726e-01
5.39880633e-01 -5.85572869e-02 9.59285498e-01 -8.41905653e-01
-1.97554797e-01 7.30126023e-01 2.87355959e-01 5.13224363e-01
5.41007280e-01 3.28048259e-01 8.62519503e-01 -6.87848449e-01
8.19595873e-01 7.76763916e-01 7.74021029e-01 6.25770509e-01
-1.20020568e+00 -4.83327091e-01 -1.64375246e-01 3.21479976e-01
-1.06321347e+00 -1.22867800e-01 -2.83760168e-02 -9.66253996e-01
1.52350390e+00 3.57174635e-01 8.62306237e-01 5.31526804e-01
9.84533578e-02 2.39567950e-01 8.42070162e-01 -3.14493746e-01
2.32799858e-01 2.27918983e-01 9.36204009e-03 6.83029592e-01
1.71817131e-02 -8.20469558e-02 -3.32190871e-01 -6.51719451e-01
3.24536085e-01 3.34702402e-01 -1.89723536e-01 2.95668561e-02
-1.52825880e+00 7.73902118e-01 3.71220469e-01 2.26117522e-01
-4.51403439e-01 -3.88500094e-01 5.76707780e-01 -1.26955003e-01
3.76923651e-01 9.19550776e-01 -7.16334939e-01 -3.12908381e-01
-1.06939471e+00 -2.09997699e-01 7.64810920e-01 7.15687454e-01
2.48224035e-01 -5.09998620e-01 -1.98138431e-01 6.15113616e-01
3.72519940e-01 -1.15635553e-02 6.72158360e-01 -8.57761383e-01
1.18916370e-01 1.21845639e+00 -2.23659575e-01 -6.40919089e-01
-7.72626996e-01 -2.41003960e-01 -5.62512517e-01 -1.41185462e-01
8.09750520e-03 -1.15121350e-01 -7.47181952e-01 9.29441810e-01
2.50861883e-01 2.93139040e-01 4.34568465e-01 6.01510584e-01
1.07073295e+00 -2.00839803e-01 6.15702987e-01 -2.88878888e-01
1.28279030e+00 -7.94992089e-01 -6.28981233e-01 1.78355575e-01
1.51621342e+00 -5.88604450e-01 4.16619122e-01 2.29861706e-01
-9.98362124e-01 -1.34384364e-01 -3.12242150e-01 2.98678309e-01
-3.00676227e-01 -1.36575654e-01 3.64251524e-01 1.78744599e-01
-1.16365314e+00 6.19268596e-01 -9.96179461e-01 -5.80927789e-01
6.61137938e-01 6.74535871e-01 -6.03450656e-01 2.80199379e-01
-7.75875688e-01 1.06767511e+00 5.38295627e-01 -3.06648351e-02
-6.44731522e-01 -1.35953522e+00 -7.02950418e-01 8.38092864e-02
3.98247242e-02 -1.13339710e+00 1.31233895e+00 -6.66222513e-01
-7.28448987e-01 1.38611591e+00 -2.39644393e-01 -3.43360782e-01
4.19773757e-01 -1.96240172e-02 -4.20093536e-01 4.82524604e-01
-3.80390361e-02 3.66769075e-01 -1.56459197e-01 -9.61343706e-01
-1.15957975e+00 -2.23836094e-01 -3.52327079e-01 2.87340075e-01
3.36858407e-02 9.30277556e-02 2.17596471e-01 -3.98830354e-01
-2.88353767e-03 -8.87879014e-01 -5.14983356e-01 -3.74255210e-01
-2.20173866e-01 -3.69192123e-01 6.26776576e-01 -6.32033288e-01
1.40317035e+00 -2.12228227e+00 -5.05601335e-03 4.26672906e-01
8.52722168e-01 8.08116570e-02 5.86663067e-01 6.94058001e-01
-5.60057819e-01 2.95570016e-01 -1.85001150e-01 1.12026436e-02
-3.05656463e-01 3.06076288e-01 -5.67119475e-03 3.75786424e-01
4.76768762e-01 8.73281598e-01 -1.40420556e+00 -6.68407977e-01
3.87378573e-01 4.14765328e-01 -8.09492767e-01 7.12044001e-01
1.98209107e-01 8.59671295e-01 -2.81298816e-01 2.29721859e-01
-1.86341181e-01 -4.18263376e-01 5.68030655e-01 1.93909019e-01
1.50174767e-01 6.40939891e-01 -5.12104452e-01 1.17767310e+00
-3.09051245e-01 4.66441661e-01 -3.02159339e-01 -8.96289051e-01
6.46365047e-01 7.84262180e-01 7.73818433e-01 1.45050630e-01
-1.61735024e-02 3.57267827e-01 5.28990805e-01 -1.05266750e+00
-2.18024477e-01 -3.32677871e-01 3.04704279e-01 7.76949167e-01
-1.88732684e-01 -1.94887131e-01 -1.58772752e-01 5.32565594e-01
1.37822998e+00 -3.94361496e-01 7.13741362e-01 -5.99254131e-01
4.60395217e-01 4.42144066e-01 3.83537084e-01 6.67186916e-01
-1.26599595e-01 7.89475024e-01 4.28618670e-01 -8.41519833e-01
-1.09594870e+00 -6.79441392e-01 -3.50094080e-01 7.17845500e-01
-6.52809143e-01 -7.11917043e-01 -4.74881262e-01 -8.27694535e-01
7.69275948e-02 8.44867170e-01 -1.16005945e+00 -3.83206278e-01
-3.75822961e-01 -4.83114451e-01 4.88119453e-01 8.19743693e-01
-3.11968803e-01 -1.23820221e+00 -1.09143305e+00 3.64335805e-01
4.19942029e-02 -8.75319004e-01 -1.01577446e-01 2.85035878e-01
-9.44337666e-01 -1.56223845e+00 -3.04933280e-01 -8.14192355e-01
9.72111642e-01 -2.82885194e-01 1.33611560e+00 7.94377267e-01
-8.23503315e-01 7.20643461e-01 -3.76852244e-01 -4.59430575e-01
-8.81683528e-01 2.01353401e-01 -1.36890942e-02 -3.40558708e-01
6.80606306e-01 -1.95027038e-01 -9.13116097e-01 7.52801374e-02
-8.41762602e-01 2.44416013e-01 6.36803448e-01 8.89083207e-01
5.32569528e-01 -3.70950073e-01 1.67508975e-01 -1.38241410e+00
5.84722579e-01 -9.39330339e-01 3.46467078e-01 3.34372342e-01
-1.05409431e+00 1.22705720e-01 3.08265746e-01 -1.41669899e-01
-7.60474503e-01 5.50423376e-02 1.94397923e-02 -2.07513630e-01
-6.77422345e-01 5.56478560e-01 4.00318354e-01 4.44679558e-01
5.34880996e-01 1.41334664e-02 -1.89798456e-02 -4.48403135e-02
2.25449651e-01 9.42540944e-01 4.26548660e-01 -1.56241193e-01
3.49509448e-01 3.51825327e-01 5.85559085e-02 -3.74270529e-02
-9.84818578e-01 -8.75562489e-01 -1.06947052e+00 -4.17082570e-02
1.17636478e+00 -7.51762390e-01 -8.10785115e-01 -2.55156517e-01
-8.41143966e-01 -4.97474968e-01 -4.88648593e-01 6.96569800e-01
-6.58686578e-01 4.69058044e-02 -3.79796594e-01 -5.23732662e-01
-4.18685526e-01 -1.21581376e+00 1.08196533e+00 -2.18332447e-02
-1.14855659e+00 -1.41204214e+00 2.96083122e-01 4.41555023e-01
5.61849661e-02 1.94727331e-01 1.44354463e+00 -1.60946238e+00
1.76381901e-01 -1.07887454e-01 -1.44183353e-01 -2.92840093e-01
5.74872375e-01 3.24102789e-01 -7.04924226e-01 4.67874222e-02
-1.82077631e-01 1.42843440e-01 3.13142031e-01 2.00785324e-01
1.23081565e+00 -5.33462703e-01 -5.07643163e-01 6.28028452e-01
1.03551710e+00 2.30287388e-01 1.17805406e-01 5.23757100e-01
7.66842782e-01 1.02342439e+00 3.27256829e-01 5.09087622e-01
5.96396983e-01 2.39610940e-01 1.35482043e-01 -3.67915779e-01
1.98875755e-01 1.32268548e-01 -4.55391794e-01 7.33278394e-01
-3.02655637e-01 4.67718810e-01 -1.94173539e+00 6.65322363e-01
-1.98807395e+00 -8.28703105e-01 -4.88389581e-01 1.80116880e+00
8.41101170e-01 9.31130126e-02 4.62323911e-02 7.60018751e-02
4.36627746e-01 -7.80214489e-01 -9.02825147e-02 -9.28799450e-01
2.10473388e-01 8.12991261e-02 2.64606863e-01 2.24111944e-01
-7.16626346e-01 5.09829342e-01 7.71489239e+00 1.96911991e-01
-6.33613229e-01 -1.92392498e-01 1.09586036e+00 -2.32821241e-01
-1.90873995e-01 -1.44816726e-01 -7.68180192e-02 2.95364201e-01
1.32201016e+00 -1.15316920e-01 1.38398811e-01 7.82090604e-01
5.25555730e-01 -1.57837644e-02 -1.56574583e+00 7.75391281e-01
-6.61850721e-02 -1.54893970e+00 5.11222985e-03 1.44811459e-02
8.93601239e-01 -2.52112389e-01 -4.40625921e-02 -2.13796310e-02
4.23316091e-01 -1.38919652e+00 2.09527478e-01 6.87248230e-01
8.43297482e-01 -5.27917087e-01 1.15226471e+00 -6.95496425e-02
-8.19436967e-01 -3.36636126e-01 1.32918969e-01 -1.81318149e-02
-1.74856976e-01 2.44107515e-01 -1.77735877e+00 6.22953892e-01
7.68799484e-01 7.44977415e-01 -4.98852491e-01 1.03813767e+00
-4.81568158e-01 8.34214568e-01 1.77236751e-01 5.68638980e-01
4.04837906e-01 1.44793451e-01 2.17108503e-01 1.70705652e+00
6.57102764e-02 8.43971431e-01 7.05328910e-03 4.70093369e-01
4.08925563e-01 1.78118512e-01 -5.56514084e-01 -5.52921034e-02
3.48596066e-01 1.37248647e+00 -8.34579945e-01 -1.03992105e+00
-1.54355630e-01 4.49486494e-01 2.78727591e-01 3.80504616e-02
-3.54277283e-01 1.49063259e-01 6.02655888e-01 4.09902334e-01
-3.25459316e-02 3.52466136e-01 -6.51231289e-01 -6.62819564e-01
-7.37251282e-01 -1.09648144e+00 1.14370692e+00 -7.26124585e-01
-1.24710512e+00 7.37724483e-01 -9.78504568e-02 -1.09513438e+00
-6.30203366e-01 -3.71423960e-01 -6.82750583e-01 7.12124884e-01
-1.05053973e+00 -7.80680835e-01 -4.42568034e-01 3.85425359e-01
5.13636649e-01 -2.21422583e-01 1.28805947e+00 -1.10945836e-01
-5.08669615e-01 3.28121901e-01 1.21998869e-01 1.63778201e-01
7.90571034e-01 -1.54311466e+00 7.84189999e-02 -1.84454113e-01
-2.97921717e-01 7.86283016e-01 5.88602245e-01 -7.80473888e-01
-6.93122089e-01 -1.10468066e+00 1.32657123e+00 -7.50542104e-01
7.58069992e-01 6.53989837e-02 -1.15818894e+00 7.63304114e-01
1.87320128e-01 -2.10433498e-01 1.69710684e+00 -1.65285834e-03
-1.08617783e-01 6.37051523e-01 -1.13294852e+00 3.41173261e-01
1.06155396e+00 -5.61519086e-01 -1.05184519e+00 6.35337949e-01
1.55623406e-01 -2.84078509e-01 -1.36235332e+00 5.48079550e-01
5.32660246e-01 -7.82339692e-01 6.22259080e-01 -1.58791232e+00
7.51202524e-01 4.39819470e-02 4.26469922e-01 -1.26929247e+00
-3.71031523e-01 -5.90627670e-01 4.14230406e-01 7.11232781e-01
7.57990062e-01 -5.95210910e-01 5.09353399e-01 1.14368284e+00
-2.49341607e-01 -1.19397533e+00 -6.73531592e-01 8.00727829e-02
5.17367944e-02 8.95903930e-02 5.40446043e-01 1.31727052e+00
9.04259801e-01 -1.13894872e-01 5.23692667e-01 2.78579414e-01
1.14753477e-01 -1.75189767e-02 2.25171372e-01 -1.26489711e+00
-2.48748049e-01 -7.42380083e-01 -6.51779771e-01 2.36379772e-01
1.62624523e-01 -1.22716117e+00 -1.59415733e-02 -2.10239458e+00
6.92458451e-01 -4.64644313e-01 -6.82654142e-01 6.79002941e-01
-7.42377162e-01 1.57871135e-02 -1.58272713e-01 7.01298237e-01
-6.01831079e-01 -1.65710494e-01 5.45998991e-01 1.72863647e-01
-1.67411953e-01 -2.34970659e-01 -7.48299718e-01 6.51428998e-01
8.20716619e-01 -9.56718802e-01 -1.75807938e-01 -3.99869800e-01
1.03011765e-01 -1.47081196e-01 2.74725378e-01 -6.06988192e-01
2.02608839e-01 -2.35801637e-01 3.56629550e-01 -3.48544538e-01
-3.36442202e-01 -8.12569976e-01 4.56695288e-01 9.07216430e-01
-8.86282563e-01 5.14815509e-01 3.57211560e-01 1.34810209e-01
-1.71152607e-01 -7.25569623e-03 4.29054916e-01 -3.95984709e-01
-2.27313384e-01 -9.89409313e-02 -9.59602892e-01 9.70378146e-02
1.26788640e+00 -2.50351250e-01 -2.36779407e-01 -3.90000731e-01
-1.12693810e+00 2.14217082e-01 3.27016175e-01 1.42905921e-01
3.95529836e-01 -8.30968142e-01 -1.03898120e+00 -9.89697874e-02
3.23971570e-01 5.85725419e-02 3.68934870e-01 1.36134052e+00
-7.68743992e-01 7.54071712e-01 -5.22471666e-02 -6.30567789e-01
-1.38660443e+00 7.17000365e-01 4.14018065e-01 -4.01324540e-01
-8.48509550e-01 6.75274670e-01 4.54151869e-01 -4.52217847e-01
9.97773334e-02 -6.97727799e-02 -2.61689186e-01 -2.89434623e-02
4.81061399e-01 2.06956044e-01 1.76383346e-01 -6.47213221e-01
-6.73171937e-01 2.76095301e-01 -1.10501274e-01 9.25978050e-02
1.55193257e+00 9.19321030e-02 -1.59752518e-01 4.20558482e-01
9.01418507e-01 -2.56130308e-01 -5.84845960e-01 1.03185683e-01
4.72078264e-01 1.07350247e-02 -2.99969763e-01 -1.02341425e+00
-7.62867093e-01 8.19541514e-01 4.05445755e-01 8.60775784e-02
9.76769388e-01 3.09399366e-01 1.28036797e-01 6.57542795e-02
-4.86806870e-01 -7.97669649e-01 -2.25909725e-01 3.65810804e-02
4.18267816e-01 -1.29880893e+00 9.02767479e-02 -6.14098370e-01
-9.77390647e-01 1.37110865e+00 3.12365144e-01 1.32529333e-01
3.54038268e-01 2.04385072e-01 5.76599538e-01 -5.88565052e-01
-1.31009376e+00 -3.72841209e-02 5.04835248e-01 6.12860203e-01
7.62664080e-01 2.79602885e-01 -3.89036685e-01 7.19284654e-01
-3.68206233e-01 -1.72647119e-01 5.05895078e-01 8.56684268e-01
-7.74951428e-02 -9.48007405e-01 -7.73376003e-02 8.97986710e-01
-6.12102926e-01 -6.26494229e-01 -6.94875896e-01 7.74603367e-01
3.37995797e-01 9.32533383e-01 1.57967031e-01 -3.06809008e-01
3.63646418e-01 2.87902832e-01 -7.87575841e-02 -1.19297290e+00
-1.35580969e+00 4.25239950e-02 1.66286826e-01 -3.61656457e-01
-4.86875474e-01 -7.42741644e-01 -1.66230500e+00 3.05968553e-01
3.93443555e-02 5.38682997e-01 3.00714433e-01 1.11327064e+00
7.27864861e-01 8.62352192e-01 2.58292019e-01 -8.73405933e-02
-3.25159401e-01 -8.67665231e-01 -4.19526584e-02 8.20765257e-01
1.32969052e-01 -4.37877268e-01 -5.07656991e-01 4.13427800e-01] | [8.090123176574707, 6.521927833557129] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.