id stringlengths 6 16 ⌀ | text stringlengths 33 3.74k | category int64 0 19 | input_ids sequence | attention_mask sequence |
|---|---|---|---|---|
2009.0642 | completely self supervised crowd counting via distribution matching dense crowd counting challenging task demand million head annotation training model though existing self supervised approach could learn good representation require labeled data map feature end task density estimation mitigate issue proposed paradigm complete self supervision need even single labeled image input required train apart large set unlabeled crowd image approximate upper limit crowd count given dataset method dwells idea natural crowd follow power law distribution could leveraged yield error signal backpropagation density regressor first pretrained self supervision distribution prediction matched prior optimizing sinkhorn distance two experiment show result effective learning crowd feature delivers significant counting performance furthermore establish superiority method le data setting well code model approach available | 0 | [
101,
3294,
2969,
13588,
4306,
10320,
3081,
4353,
9844,
9742,
4306,
10320,
10368,
4708,
5157,
2454,
2132,
5754,
17287,
3508,
2731,
2944,
2295,
4493,
2969,
13588,
3921,
2071,
4553,
2204,
6630,
5478,
12599,
2951,
4949,
3444,
2203,
4708,
4304,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2010.13821 | wavelet flow fast training high resolution normalizing flow normalizing flow class probabilistic generative model allow fast density computation efficient sampling effective modelling complex distribution like image drawback among current method significant training cost sometimes requiring month gpu training time achieve state art result paper introduces wavelet flow multi scale normalizing flow architecture based wavelet wavelet flow explicit representation signal scale inherently includes model lower resolution signal conditional generation higher resolution signal ie super resolution major advantage wavelet flow ability construct generative model high resolution data eg image impractical previous model furthermore wavelet flow competitive previous normalizing flow term bit per dimension standard low resolution benchmark faster train | 0 | [
101,
4400,
7485,
4834,
3435,
2731,
2152,
5813,
3671,
6026,
4834,
3671,
6026,
4834,
2465,
4013,
3676,
27965,
4588,
11416,
6024,
2944,
3499,
3435,
4304,
22334,
8114,
16227,
4621,
19518,
3375,
4353,
2066,
3746,
4009,
5963,
2426,
2783,
4118,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1904.12782 | transversally elliptic complex cohomological field theory work continuation previous paper arxiv constructed supersymmetric yang mill theory manifold killing vector field isolated fixed point work expand mathematical aspect theory particular focus nature cohomological field theory well known donaldson witten theory twisted version sym also constructed using atiyah jeffrey construction theory concerned modulus space anti self dual gauge connection deformation theory controlled elliptic complex generally supersymmetry requires considering configuration look like either instantons anti instantons around fixed point call flipping instantons flipping instantons theory derived contact instantons novelty deformation theory controlled transversally elliptic complex demonstrate repeat atiyah jeffrey construction equivariant setting arrive lagrangian equivariant euler class relevant field space also obtained previous work arxiv show transversal ellipticity deformation complex crucial non degeneracy lagrangian calculability theory construction valid large class quasi toric manifold | 1 | [
101,
9099,
14028,
3973,
29413,
3375,
2522,
23393,
10091,
2492,
3399,
2147,
13633,
3025,
3259,
12098,
9048,
2615,
3833,
3565,
6508,
20058,
12412,
8675,
4971,
3399,
19726,
4288,
9207,
2492,
7275,
4964,
2391,
2147,
7818,
8045,
7814,
3399,
3327... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2105.00878 | malliavin rubel theorem small entire function exponential type given zero early p malliavin rubel gave complete description pair distribution positive point entire function exponential type vanishes entire function exponential type vanishes satisfies inequality everywhere imaginary axis extend result much general distribution complex point lying outside pair vertical angle containing imaginary axis point approach | 1 | [
101,
6670,
2401,
6371,
14548,
2884,
9872,
2235,
2972,
3853,
27258,
2828,
2445,
5717,
2220,
1052,
6670,
2401,
6371,
14548,
2884,
2435,
3143,
6412,
3940,
4353,
3893,
2391,
2972,
3853,
27258,
2828,
25887,
2229,
2972,
3853,
27258,
2828,
25887,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1906.04024 | odd cycle game connected rule study positional game two player maker breaker alternately select respectively previously unclaimed edge maker win succeeds claiming edge odd cycle breaker win otherwise improving result bednarska pikhurko show maker win odd cycle game furthermore introduce connected rule study odd cycle game maker breaker well client waiter variant | 1 | [
101,
5976,
5402,
2208,
4198,
3627,
2817,
2597,
2389,
2208,
2048,
2447,
9338,
24733,
23554,
7276,
4414,
3130,
4895,
25154,
2098,
3341,
9338,
2663,
21645,
6815,
3341,
5976,
5402,
24733,
2663,
4728,
9229,
2765,
2793,
11802,
8337,
14255,
10023,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
2102.10644 | surface tension destabiliser vortical interface study dynamic initially flat interface two immiscible fluid vortex situated show surface tension cause vorticity generation general curved interface creates velocity jump across interface increase quadratically time cause kelvin helmholtz instability surface tension thus act destabiliser vorticity creation winning tendency stabilize smoothing interfacial perturbation reduce surface energy show instability manifested within vortex core time larger weber number perturbation wavenumber destroying flow structure vorticity peel small scale structure away interface using energy balance provide growth time total interface length density difference fluid produce additional instability outside vortex core due centrifugal effect demonstrate importance mechanism two dimensional turbulence simulation prescribed initial interface show increase kinetic energy large wave number due creation small scale vortex action surface tension | 2 | [
101,
3302,
6980,
4078,
2696,
27965,
2121,
29536,
28228,
9289,
8278,
2817,
8790,
3322,
4257,
8278,
2048,
10047,
15630,
6895,
3468,
8331,
23463,
4350,
2265,
3302,
6980,
3426,
29536,
28228,
12972,
4245,
2236,
9203,
8278,
9005,
10146,
5376,
240... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1909.13561 | imagine leveraging emergent affordances tool synthesis paper explore richness information captured latent space vision based generative model model combine unsupervised generative learning task based performance predictor learn exploit task relevant object affordances given visual observation reaching task involving scenario stick like tool learned embedding generative model capture factor variation tool geometry eg length width shape performance predictor identifies sub manifold embedding correlate task success within variety scenario demonstrate traversing latent space via backpropagation performance predictor allows u imagine tool appropriate task hand result indicate affordances like utility reaching encoded along smooth trajectory latent space accessing emergent affordances considering high level performance criterion task success enables agent manipulate tool geometry targeted deliberate way | 0 | [
101,
5674,
15929,
16594,
12636,
3372,
8984,
26755,
6994,
10752,
3259,
8849,
4138,
2791,
2592,
4110,
2397,
3372,
2686,
4432,
2241,
11416,
6024,
2944,
2944,
11506,
4895,
6342,
4842,
11365,
2098,
11416,
6024,
4083,
4708,
2241,
2836,
16014,
295... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1902.08524 | acoustically modulated optical emission hexagonal boron nitride layer investigate effect surface acoustic wave atomic like optical emission defect center hexagonal boron nitride layer deposited surface linbo substrate dynamic strain field surface acoustic wave modulates emission line resulting intensity variation large oscillation emission energy amplitude almost mev systematic study dependence modulation acoustic wave power determine hydrostatic deformation potential defect center two dimensional material mev furthermore show dynamic piezoelectric field acoustic wave could contribute stabilization optical property center result show surface acoustic wave powerful tool modulate control electronic state two dimensional material | 2 | [
101,
6490,
3973,
16913,
8898,
9380,
15760,
2002,
18684,
20028,
8945,
4948,
9152,
18886,
3207,
6741,
8556,
3466,
3302,
6490,
4400,
9593,
2066,
9380,
15760,
21262,
2415,
2002,
18684,
20028,
8945,
4948,
9152,
18886,
3207,
6741,
14140,
3302,
11... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1904.05951 | prevalence persistent tangle article address persistent tangle tangle whose presence knot diagram force diagram knotted provide new method constructing persistent tangle technique rely mainly existence non trivial coloring tangle question main result article knot admitting non trivial coloring give rise persistent tangle furthermore discus persistent tangle non trivial | 1 | [
101,
20272,
14516,
24453,
3720,
4769,
14516,
24453,
24453,
3005,
3739,
12226,
16403,
2486,
16403,
27039,
3073,
2047,
4118,
15696,
14516,
24453,
6028,
11160,
3701,
4598,
2512,
20610,
22276,
24453,
3160,
2364,
2765,
3720,
12226,
17927,
2512,
20... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
2007.01974 | one sided topological conjugacy topological markov shift gauge action cuntz krieger algebra characterize topological conjugacy class one sided topological markov shift term associated cuntz krieger algebra gauge action potential | 1 | [
101,
2028,
11536,
24309,
9530,
9103,
3654,
5666,
24309,
28003,
2615,
5670,
7633,
2895,
12731,
3372,
2480,
1047,
7373,
4590,
11208,
2839,
4697,
24309,
9530,
9103,
3654,
5666,
2465,
2028,
11536,
24309,
28003,
2615,
5670,
2744,
3378,
12731,
33... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
2103.05146 | ore condition hamiltonicity tough graph let tough graph vertex shown bauer et al minimum degree greater hamiltonian term ore condition direction problem studied paper show degree sum two nonadjacent vertex greater hamiltonian | 1 | [
101,
10848,
4650,
5226,
28775,
3723,
7823,
10629,
2292,
7823,
10629,
19449,
3491,
17838,
3802,
2632,
6263,
3014,
3618,
5226,
2937,
2744,
10848,
4650,
3257,
3291,
3273,
3259,
2265,
3014,
7680,
2048,
2512,
4215,
3900,
13013,
19449,
3618,
5226... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
1906.09354 | boosting rule accuracy deep disease detection using class weight modifier many screening application primary goal radiologist assisting artificial intelligence rule certain finding classifier built application often trained large datasets derive label clinical note written patient quality positive finding described note often reliable lack mention finding always rule presence happens radiologist comment patient context exam example focusing trauma opposed chronic disease emergency room however disease finding ambiguity affect performance algorithm hence critical model ambiguity training propose scheme apply reasonable class weight modifier loss function mention case training experiment two different deep neural network architecture show proposed method result large improvement performance classifier specially negated finding baseline performance custom made dilated block network proposed paper show improvement comparison baseline densenet architecture benefit new proposed loss function weighting scheme chest ray image three highly common disease along negated counterpart included study | 0 | [
101,
12992,
2075,
3627,
10640,
2784,
4295,
10788,
2478,
2465,
3635,
16913,
18095,
2116,
11326,
4646,
3078,
3125,
2557,
10727,
13951,
7976,
4454,
3627,
3056,
4531,
2465,
18095,
2328,
4646,
2411,
4738,
2312,
2951,
13462,
2015,
18547,
3830,
66... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1910.0944 | speed convergence chernoff approximation solution evolution equation communication devoted establishing first step study speed error decrease dealing based chernoff theorem approximation one parameter semigroups provide solution evolution equation | 0 | [
101,
3177,
19143,
24188,
3630,
4246,
20167,
5576,
6622,
8522,
4807,
7422,
7411,
2034,
3357,
2817,
3177,
7561,
9885,
7149,
2241,
24188,
3630,
4246,
9872,
20167,
2028,
16381,
4100,
17058,
2015,
3073,
5576,
6622,
8522,
102
] | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
1912.02502 | monte carlo study spin chirality decoupling three dimensional heisenberg spin glass extensive equilibrium monte carlo simulation performed isotropic heisenberg sg model random nearest neighbor gaussian coupling particular interest chiral glass cg spin glass sg ordering model possibility spin chirality decoupling ie cg order setting higher temperature sg order suggested earlier still remains controversial simulate model maximum size linear dimension periodic open boundary condition bc locating cg sg transition temperature extrapolation variety independent physical quantity bc computed utilized get larger number degree freedom ndf thanks large ndf ndf succeed obtaining stable accurate estimate cg sg transition temperature sign size crossover observed larger cg correlation length progressively outgrows sg correlation length low temperature result provide strong numerical support spin chirality decoupling critical exponent associated cg sg transition evaluated use finite size scaling scaling correction cg transition get cg exponent consistently corresponding experimental exponent canonical sg implication chirality scenario experimental sg ordering discussed | 2 | [
101,
10125,
9758,
2817,
6714,
9610,
7941,
3012,
21933,
6279,
2989,
2093,
8789,
2002,
28992,
4059,
6714,
3221,
4866,
14442,
10125,
9758,
12504,
2864,
11163,
13181,
24330,
2002,
28992,
4059,
22214,
2944,
6721,
7205,
11429,
11721,
17854,
2937,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1911.08413 | igatelink gateway library linking iot edge fog cloud computing environment recent year internet thing iot growing popularity along increasingly important role played iot gateway mediating interaction among plethora heterogeneous iot device cloud service paper present igatelink open source android library easing development android application acting gateway iot device edgefogcloud computing environment thanks pluggable design module providing connectivity number device acting data source fogcloud framework easily reused different application using igatelink two case study replicating previous work healthcare image processing domain library proved effective adapting different scenario speeding development gateway application compared use conventional method | 3 | [
101,
1045,
5867,
13767,
11909,
3075,
11383,
22834,
2102,
3341,
9666,
6112,
9798,
4044,
3522,
2095,
4274,
2518,
22834,
2102,
3652,
6217,
2247,
6233,
2590,
2535,
2209,
22834,
2102,
11909,
2865,
3436,
8290,
2426,
20228,
11031,
6525,
21770,
106... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2011.06398 | spherical covering raying convex body constant width main goal work give construction certain spherical covering small dimension bezdek gy kiss showed existence origin symmetric covering unit sphere congruent spherical cap radius exceeding implies ray conjecture illumination conjecture convex body constant width constructed covering give construction fewer cap combination known result higher dimension due schramm completely settle mentioned conjecture convex body constant width dimension also show calculate covering radius given discrete point set sphere efficiently computer | 1 | [
101,
18970,
5266,
4097,
2075,
18309,
2303,
5377,
9381,
2364,
3125,
2147,
2507,
2810,
3056,
18970,
5266,
2235,
9812,
2022,
26494,
5937,
1043,
2100,
3610,
3662,
4598,
4761,
19490,
5266,
3131,
10336,
26478,
6820,
4765,
18970,
6178,
12177,
1700... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1809.02883 | noise noise ratio correlation length calculation near criticality finite random system possible define two type variance noise demonstrated ratio useful calculating correlation length infinite rather general random system function temperature numerical method obtaining variable relevant real space numerical renormalization simulation method matter correlation length obtained novel technique may used obtain directly critical correlation exponent rather indirectly using scaling relation often done method demonstrated applying random field ising model | 4 | [
101,
5005,
5005,
6463,
16902,
3091,
17208,
2379,
4187,
3012,
10713,
6721,
2291,
2825,
9375,
2048,
2828,
23284,
5005,
7645,
6463,
6179,
20177,
16902,
3091,
10709,
2738,
2236,
6721,
2291,
3853,
4860,
15973,
4118,
11381,
8023,
7882,
2613,
2686... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2010.04754 | explicit time mimetic discretiztions wave equation paper part program combine staggered time staggered spatial discretization continuum wave equation important property continuum proved using vector calculus proven analogous way discretized system spatial discretizations second order accurate mimetic time discretizations second order accurate explicit leapfrog scheme discretizations conserved quantity guarantee stability reasonable constraint time step conserved quantity closely related energy continuum equation | 0 | [
101,
13216,
2051,
2771,
11368,
2594,
5860,
13465,
10993,
9285,
4400,
8522,
3259,
2112,
2565,
11506,
14648,
2051,
14648,
13589,
5860,
13465,
3989,
22961,
4400,
8522,
2590,
3200,
22961,
4928,
2478,
9207,
19276,
10003,
19639,
2126,
5860,
13465,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2005.00921 | epidemic model sipherd application prediction spread covid infection india usa propose epidemic model sipherd three category infection carrier symptomatic purely asymptomatic exposed considered different rate transmission infection taken dependent lockdown social distancing rate detection infected carrier taken dependent test done per day model applied covid outbreak germany south korea validate predictive capability applied india united state prediction spread different lockdown situation testing coming month | 5 | [
101,
16311,
2944,
10668,
5886,
2094,
4646,
17547,
3659,
2522,
17258,
8985,
2634,
3915,
16599,
16311,
2944,
10668,
5886,
2094,
2093,
4696,
8985,
6839,
25353,
27718,
9626,
4588,
11850,
2004,
24335,
13876,
9626,
4588,
6086,
2641,
2367,
3446,
6... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2105.12605 | convergence lagrangian discretization barotropic fluid porous medium flow expressed lagrangian variable equation motion compressible barotropic fluid structure classical hamiltonian system potential energy given internal energy fluid dissipative counterpart system coincides porous medium equation cast form gradient flow internal energy motivated related variational structure propose particle method problem internal energy replaced moreau yosida regularization sense efficiently computed semi discrete optimal transport problem using modulated energy argument exploit convexity problem eulerian variable prove quantitative convergence estimate towards smooth solution verify estimate mean several numerical test | 1 | [
101,
19143,
2474,
17643,
3070,
2937,
5860,
13465,
3989,
3347,
4140,
18981,
2594,
8331,
18499,
3560,
5396,
4834,
5228,
2474,
17643,
3070,
2937,
8023,
8522,
4367,
4012,
20110,
7028,
3347,
4140,
18981,
2594,
8331,
3252,
4556,
5226,
2937,
2291,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1904.10322 | neural influence diffusion model social recommendation precise user item embedding learning key building successful recommender system traditionally collaborative filteringcf provides way learn user item embeddings user item interaction history however performance limited due sparseness user behavior data emergence online social network social recommender system proposed utilize user local neighbor preference alleviate data sparsity better user embedding modeling argue user social platform potential embedding influenced trusted user social influence recursively propagates diffuses social network user interest change recursive process nevertheless current social recommendation model simply developed static model leveraging local neighbor user without simulating recursive diffusion global social network leading suboptimal recommendation performance paper propose deep influence propagation model stimulate user influenced recursive social diffusion process social recommendation user diffusion process start initial embedding fuse related feature free user latent vector capture latent behavior preference key idea proposed model design layer wise influence propagation structure model user latent embeddings evolve social diffusion process continues show proposed model general could applied useritem attribute social network structure available finally extensive experimental result two real world datasets clearly show effectiveness proposed model performance improvement best baseline | 0 | [
101,
15756,
3747,
19241,
2944,
2591,
12832,
10480,
5310,
8875,
7861,
8270,
4667,
4083,
3145,
2311,
3144,
16755,
2121,
2291,
6964,
12317,
22910,
2278,
2546,
3640,
2126,
4553,
5310,
8875,
7861,
8270,
4667,
2015,
5310,
8875,
8290,
2381,
2174,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1905.01607 | performance evaluation ndn data plane using statistical model checking named data networking ndn emerging technology future internet architecture address weakness internet protocol ip since internet user application demonstrated ever increasing need high speed packet forwarding research group investigated different design implementation fast ndn data plane forwarders claimed capable achieving high throughput rate however correctness statement supported verification technique formal proof paper propose using formal model based approach overcome issue consider ndn dpdk prototype implementation forwarder realized nist leverage concurrency enhance overall quality service use approach improve design formally show achieve high throughput rate | 0 | [
101,
2836,
9312,
1050,
2094,
2078,
2951,
4946,
2478,
7778,
2944,
9361,
2315,
2951,
14048,
1050,
2094,
2078,
8361,
2974,
2925,
4274,
4294,
4769,
11251,
4274,
8778,
12997,
2144,
4274,
5310,
4646,
7645,
2412,
4852,
2342,
2152,
3177,
14771,
2... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2012.03303 | tridomain model potassium clearance optic nerve necturus accumulation potassium narrow space outside nerve cell classical subject biophysics received much attention recently may involved potassium accumulation textcolorblackincluding spreading depression perhaps migraine kind epilepsy even speculatively learning quantitative analysis likely help evaluate role potassium clearance extracellular space train action potential clearance involves three structure extend length nerve glia extracellular space axon need described system distributed space tradition used electrical potential cable equation nerve since work hodgkin three compartment model proposed optic nerve used study accumulation potassium clearance model allows convection diffusion electrical migration water ion depend data orkand et al ensure relevance model align parameter anatomy property membrane channel transporter model fit experimental data quite well aligned model show glia important role buffering potassium expected model show potassium cleared mostly convective flow syncytium glia driven osmotic pressure difference simplified model might possible must involve flow length optic nerve easy compartment model neglect flow model used structure quite different optic nerve might different distribution channel transporter three compartment generalized include fourth distributed compartment representing blood vessel deal glymphatic flow circulatory system | 5 | [
101,
13012,
9527,
8113,
2944,
18044,
14860,
22816,
9113,
26785,
20689,
2271,
20299,
18044,
4867,
2686,
2648,
9113,
3526,
4556,
3395,
16012,
15638,
2363,
2172,
3086,
3728,
2089,
2920,
18044,
20299,
3793,
18717,
28522,
18009,
12273,
7630,
4667,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2011.12104 | recurrent multi view alignment network unsupervised surface registration learning non rigid registration end end manner challenging due inherent high degree freedom lack labeled training data paper resolve two challenge simultaneously first propose represent non rigid transformation point wise combination several rigid transformation representation make solution space well constrained also enables method solved iteratively recurrent framework greatly reduces difficulty learning second introduce differentiable loss function measure shape similarity projected multi view depth image full framework trained end end without ground truth supervision extensive experiment several different datasets demonstrate proposed method outperforms previous state art large margin source code available | 0 | [
101,
28667,
29264,
4800,
3193,
12139,
2897,
4895,
6342,
4842,
11365,
2098,
3302,
8819,
4083,
2512,
11841,
8819,
2203,
2203,
5450,
10368,
2349,
16112,
2152,
3014,
4071,
3768,
12599,
2731,
2951,
3259,
10663,
2048,
4119,
7453,
2034,
16599,
505... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1909.11232 | sign language recognition analysis using multimodal data voice controlled personal home assistant amazon echo apple siri becoming increasingly popular variety application however benefit technology readily accessible deaf hard ofhearing dhh user objective study develop evaluate sign recognition system using multiple modality used dhh signer interact voice controlled device advancement depth sensor skeletal data used application like video analysis activity recognition despite similarity well studied human activity recognition use skeleton data sign language recognition rare unlike activity recognition sign language mostly dependent hand shape pattern work investigate feasibility using skeletal rgb video data sign language recognition using combination different deep learning architecture validate result large scale american sign language asl dataset user sample across sign named gmuasl collected dataset month publicly released hope spurring machine learning research towards providing improved accessibility digital assistant | 6 | [
101,
3696,
2653,
5038,
4106,
2478,
4800,
5302,
9305,
2951,
2376,
4758,
3167,
2188,
3353,
9733,
9052,
6207,
2909,
2072,
3352,
6233,
2759,
3528,
4646,
2174,
5770,
2974,
12192,
7801,
12419,
2524,
1997,
26560,
2075,
28144,
2232,
5310,
7863,
2... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1812.09299 | higgs coupling without higgs measurement higgs coupling constitute important part present standard model precision test collider article show modification higgs coupling induce energy growing effect specific amplitude involving longitudinally polarized vector boson initiate novel program study modification higgs coupling shell high energy rather higgs resonance analysis suggests channel complementary time competitive familiar shell measurement moreover high energy probe offer endless opportunity refinement improvement | 7 | [
101,
7632,
21314,
19780,
2302,
7632,
21314,
10903,
7632,
21314,
19780,
12346,
2590,
2112,
2556,
3115,
2944,
11718,
3231,
8902,
24198,
2099,
3720,
2265,
14080,
7632,
21314,
19780,
19653,
2943,
3652,
3466,
3563,
22261,
5994,
20134,
2135,
11508,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2107.13438 | dark matter search result pandax commissioning run report first dark matter search result using commissioning data pandax using time projection chamber tonne liquid xenon target exposure tonneyear candidate event identified within approximate electron equivalent energy window kev significant excess background observed data set stringent limit dark matter nucleon spin independent interaction lowest excluded cross section cl cm dark matter mass gev | 2 | [
101,
2601,
3043,
3945,
2765,
25462,
2595,
21612,
2448,
3189,
2034,
2601,
3043,
3945,
2765,
2478,
21612,
2951,
25462,
2595,
2478,
2051,
13996,
4574,
10228,
2638,
6381,
1060,
16515,
2078,
4539,
7524,
10228,
5420,
14644,
4018,
2724,
4453,
2306... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1912.00635 | hook variable cut join operator function young diagram parameterized help hook variable well known never studied big detail demonstrate adequate parameterization many physical application schur function conventional skew shifted satisfy kind determinant formula coordinate kptoda integrability related basis cut join operator actually expressed single hook diagram particular discus new type multi component kp function matisse function also demonstrate casimir operator responsible integrability single hook popular basis completed cycle distinguished especially simple coefficient corresponding expansion casimir operator also generate ruijsenaars hamiltonians however property broken naive macdonald deformation reason loss kptoda integrability related structure matrix model | 8 | [
101,
8103,
8023,
3013,
3693,
6872,
3853,
2402,
16403,
16381,
3550,
2393,
8103,
8023,
2092,
2124,
2196,
3273,
2502,
6987,
10580,
11706,
16381,
3989,
2116,
3558,
4646,
8040,
24572,
3853,
7511,
15315,
7974,
5429,
13225,
2785,
28283,
22311,
337... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2012.15792 | chiral correlators superconformal quiver consider family superconformal field theory four dimension defined orbifolds super yang mill theory compute chiralanti chiral correlation function perturbative level using matrix model approach arising supersymmetric localisation four sphere explicit field theory calculation flat space using superspace formalism implement highly efficient algorithm produce large number result finite value exploiting symmetry quiver reduce complexity mixing operator finally interplay field theory calculation allows isolate special observables deviate high order perturbation theory | 9 | [
101,
9610,
7941,
2522,
14343,
20051,
5668,
3565,
8663,
14192,
2389,
29049,
5136,
2155,
3565,
8663,
14192,
2389,
2492,
3399,
2176,
9812,
4225,
19607,
10128,
11614,
2015,
3565,
8675,
4971,
3399,
24134,
9610,
7941,
4630,
2072,
9610,
7941,
1690... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1904.05375 | scanner invariant representation diffusion mri harmonization purpose present work describe correction diffusion weighted mri site scanner bias using novel method based invariant representation theory method pooled imaging data multiple source subject variation source correcting bias become important imaging study increase size multi site case become common propose learning intermediate representation invariant siteprotocol variable technique adapted information theory based algorithmic fairness leveraging data processing inequality representation used create image reconstruction uninformative original source yet still faithful underlying structure implement use deep learning method based variational auto encoders vae construct scanner invariant encoding imaging data result evaluate method use training data miccai computational diffusion mri cdmri challenge harmonization dataset proposed method show improvement independent test data relative recently published baseline method subtask mapping data three different scanning context one separate target scanning context conclusion imaging study continue grow use pooled multi site imaging similarly increase invariant representation present strong candidate harmonization data | 3 | [
101,
26221,
23915,
6630,
19241,
27011,
25546,
3989,
3800,
2556,
2147,
6235,
18140,
19241,
18215,
27011,
2609,
26221,
13827,
2478,
3117,
4118,
2241,
23915,
6630,
3399,
4118,
19107,
2094,
12126,
2951,
3674,
3120,
3395,
8386,
3120,
6149,
2075,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2102.01495 | deep learning based joint precoder design antenna selection partially connected hybrid massive mimo system efficient resource allocation hybrid precoder design essential massive mimo system operating millimeter wave mmw domain owing higher energy efficiency lower complexity partially connected hybrid architecture letter propose joint deep convolutional neural network cnn based scheme precoder design antenna selection partially connected massive mimo hybrid system precoder design antenna selection formulated regression classification problem respectively cnn channel data fed first cnn network output subset selected antenna optimal spectral efficiency subset fed second cnn obtain block diagonal precoder partially connected architecture simulation result verifies superiority cnn based approach conventional iterative alternating minimization alt min algorithm moreover proposed scheme computationally efficient sensitive channel irregularity | 3 | [
101,
2784,
4083,
2241,
4101,
3653,
16044,
2099,
2640,
13438,
4989,
6822,
4198,
8893,
5294,
2771,
5302,
2291,
8114,
7692,
16169,
8893,
3653,
16044,
2099,
2640,
6827,
5294,
2771,
5302,
2291,
4082,
4971,
19198,
4400,
3461,
2860,
5884,
11427,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1506.00898 | extreme compressive sampling covariance estimation paper study problem estimating covariance collection vector using highly compressed measurement vector estimator based back projection compressive sample proposed analyzed distribution free analysis show observing single linear measurement vector one consistently estimate covariance matrix infinity spectral norm analysis lead precise rate convergence norm via information theoretic technique lower bound showing estimator minimax optimal infinity spectral norm estimation problem established result also specialized give matching upper lower bound estimating population covariance collection gaussian vector compressive measurement model analysis conducted paper show effective sample complexity problem scaled factor compression dimension ambient dimension application subspace learning principal component analysis learning distributed sensor network also discussed | 0 | [
101,
6034,
4012,
27484,
16227,
2522,
10755,
28335,
24155,
3259,
2817,
3291,
9765,
22835,
2522,
10755,
28335,
3074,
9207,
2478,
3811,
16620,
10903,
9207,
9765,
9581,
4263,
2241,
2067,
13996,
4012,
27484,
7099,
3818,
16578,
4353,
2489,
4106,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1908.09579 | analytic structure solution one dimensional burger equation modified dissipation use one dimensional burger equation illustrate effect replacing standard laplacian dissipation term general function laplacian hyperviscosity best known example equation hydrodynamics analyze asymptotic structure solution fourier space high wave number introducing approach applicable wide class hydrodynamical equation whose solution calculated limit vanishing reynolds number algebraic recursion relation involving iterated integration give detailed analysis analytic structure two different type dissipation hyperviscous exponentially growing dissipation term result obtained limit vanishing reynolds number validated high precision numerical simulation non zero reynolds number study bottleneck problem intermediate asymptotics phenomenon case burger equation arises one us dissipation term hyperviscosity growing faster high wave number standard laplacian dissipation term linearized solution well known boundary layer limit burger equation involving two numerically determined parameter give good description bottleneck region | 1 | [
101,
23521,
3252,
5576,
2028,
8789,
15890,
8522,
6310,
4487,
18719,
24952,
2239,
2224,
2028,
8789,
15890,
8522,
19141,
3466,
6419,
3115,
5001,
2721,
14483,
4487,
18719,
24952,
2239,
2744,
2236,
3853,
5001,
2721,
14483,
23760,
11365,
13186,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2005.07581 | convergence fractal periodic schrodinger equation consider fractal refinement carleson problem pointwise convergence solution periodic schrodinger equation initial datum sufficient solution corresponding every datum converge almost everywhere | 1 | [
101,
19143,
25312,
25572,
2140,
15861,
8040,
8093,
7716,
9912,
8522,
5136,
25312,
25572,
2140,
25416,
3170,
3672,
5529,
21421,
3291,
2391,
14244,
19143,
5576,
15861,
8040,
8093,
7716,
9912,
8522,
3988,
23755,
2819,
7182,
5576,
7978,
2296,
2... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
2104.04516 | whittaker vector algebra topological recursion identify whittaker vector module partition function higher airy structure implies gaiotto vector describing fundamental class equivariant cohomology suitable compactification modulus space bundle complex simple lie group computed non commutative version chekhov eynard orantin topological recursion formulate connection higher airy structure gaiotto vector type b explicitly construct topological recursion type arbitrary level type self dual level physic side mean nekrasov partition function pure four dimensional supersymmetric gauge theory accessed topological recursion method | 1 | [
101,
1059,
16584,
15166,
2099,
9207,
11208,
24309,
28667,
9236,
3258,
6709,
1059,
16584,
15166,
2099,
9207,
11336,
13571,
3853,
3020,
2250,
2100,
3252,
12748,
11721,
25185,
3406,
9207,
7851,
8050,
2465,
1041,
15549,
10755,
2937,
2102,
2522,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1705.01242 | non existence higgs field calabi yau manifold article study higgs vector bundle compact calabi yau manifold use yang mill higgs flow prove semistable higgs bundle vanishing chern class compact connected calabi yau manifold higgs field trivial particular vector bundle must semistable | 1 | [
101,
2512,
4598,
7632,
21314,
2492,
10250,
28518,
8038,
2226,
19726,
3720,
2817,
7632,
21314,
9207,
14012,
9233,
10250,
28518,
8038,
2226,
19726,
2224,
8675,
4971,
7632,
21314,
4834,
6011,
4100,
9153,
3468,
7632,
21314,
14012,
24866,
24188,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
2008.01245 | cautious active clustering consider problem classification point sampled unknown probability measure euclidean space study question querying class label small number judiciously chosen point able attach appropriate class label every point set approach consider unknown probability measure convex combination conditional probability class technique involves use highly localized kernel constructed hermite polynomial order create hierarchical estimate support constituent probability measure need make assumption nature probability measure know advance number class involved give theoretical guarantee measured score classification scheme example include classification hyper spectral image mnist classification | 0 | [
101,
17145,
3161,
9324,
2075,
5136,
3291,
5579,
2391,
18925,
4242,
9723,
5468,
25826,
2686,
2817,
3160,
23032,
2075,
2465,
3830,
2235,
2193,
18414,
14808,
19426,
4217,
2391,
2583,
22476,
6413,
2465,
3830,
2296,
2391,
2275,
3921,
5136,
4242,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1909.09796 | absolute algebraic geometry affine case develop algebraic geometry general segal gamma ring show new theory unifies two approach considered earlier geometry spec starting observation category obtained gluing together category commutative ring pointed commutative monoids used previous work define scheme naturally full subcategory category segal gamma ring paper develop affine case general algebraic geometry one distinctive feature spectrum speca gamma ring general grothendieck site rather point set endowed topology two striking feature new geometry natural domain cyclic homology homological algebra new operation make sense ordinary algebraic geometry available instance new context quotient ring multiplicative subgroup still gamma ring general theory applies thus adele class space give rise naturally gamma ring finally show theory special case toen vaqui general theory algebraic geometry spec | 1 | [
101,
7619,
17390,
10988,
21358,
23460,
2553,
4503,
17390,
10988,
2236,
16562,
2140,
13091,
3614,
2265,
2047,
3399,
4895,
14144,
2048,
3921,
2641,
3041,
10988,
28699,
3225,
8089,
4696,
4663,
1043,
7630,
2075,
2362,
4696,
4012,
28120,
8082,
3... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1905.06029 | non uniqueness parameter extracted resonant second order nonlinear optical spectroscopy experimental data second order nonlinear optical spectroscopy sfg dfg shg provide parameter relevant physical chemistry interface thin film show general equivalent set parameter fit experimental curve comprising resonant feature vibrational electronic origin example provide algorithm calculate set among appropriate selected main consequence deal existence ghost resonance need critical analysis fit result procedure search better set parameter coherent applied constraint | 2 | [
101,
2512,
4310,
2791,
16381,
15901,
24501,
7856,
3372,
2117,
2344,
27400,
9380,
25458,
6388,
2951,
2117,
2344,
27400,
9380,
25458,
16420,
2290,
1040,
2546,
2290,
14021,
2290,
3073,
16381,
7882,
3558,
6370,
8278,
4857,
2143,
2265,
2236,
566... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2101.01908 | factor modelling clustering high dimensional time series propose new unsupervised learning method clustering large number time series based latent factor structure cluster characterized cluster specific factor addition common factor impact time series concerned setting also offer flexibility time series may belong cluster consistency explicit convergence rate established estimation common factor cluster specific factor latent cluster numerical illustration simulated data well real data example also reported spin proposed new approach also advance significantly statistical inference factor model lam yao | 6 | [
101,
5387,
19518,
9324,
2075,
2152,
8789,
2051,
2186,
16599,
2047,
4895,
6342,
4842,
11365,
2098,
4083,
4118,
9324,
2075,
2312,
2193,
2051,
2186,
2241,
2397,
3372,
5387,
3252,
9324,
7356,
9324,
3563,
5387,
2804,
2691,
5387,
4254,
2051,
21... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1903.02072 | risk sensitive necessary sufficient optimality condition financial application fully coupled forward backward stochastic differential equation jump diffusion throughout paper focused aim problem optimal control risk sensitive performance functional system given fully coupled forward backward stochastic differential equation jump risk neutral control system used preliminary step admissible control convex optimal solution exists necessary well sufficient optimality condition risk sensitive performance proved end work illustrate main result giving example mean variance risk sensitive control problem applied cash flow market | 1 | [
101,
3891,
7591,
4072,
7182,
15502,
3012,
4650,
3361,
4646,
3929,
11211,
2830,
8848,
2358,
11663,
20875,
11658,
8522,
5376,
19241,
2802,
3259,
4208,
6614,
3291,
15502,
2491,
3891,
7591,
2836,
8360,
2291,
2445,
3929,
11211,
2830,
8848,
2358,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2005.09357 | color routing via cross polarized detuned plasmonic nanoantennas large area metasurfaces bidirectional nanoantennas key relevance advanced functionality implemented nanoscale particular color routing ultracompact flat optic configuration demonstrate novel approach avoiding complex collective geometry andor restrictive morphological parameter based cross polarized detuned plasmonic nanoantennas uniaxial quasi bimetallic configuration nanofabrication flat optic system controlled large area cm novel self organized technique exploiting ion induced nanoscale wrinkling instability glass template engineer tilted bimetallic nanostrip dimer nanoantennas feature broadband color routing superior light scattering directivity figure well described numerical simulation turn competitive response lithographic nanoantennas result demonstrate large area self organized metasurfaces implemented real world application flat optic color routing telecom photonics optical nanosensing | 2 | [
101,
3609,
16972,
3081,
2892,
11508,
3550,
20010,
9816,
2094,
20228,
3022,
8202,
2594,
28991,
12956,
9516,
2015,
2312,
2181,
18804,
26210,
12172,
2015,
7226,
7442,
7542,
2389,
28991,
12956,
9516,
2015,
3145,
21923,
3935,
15380,
7528,
28991,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2007.14755 | learning transferable push manipulation skill novel context paper concerned learning transferable forward model push manipulation applying novel context improve quality prediction critical information available propose learn parametric internal model push interaction similar human enables robot predict outcome physical interaction even novel context given desired push action human capable identify place finger new object produce predictable motion object achieve behaviour factorising learning two part first learn set local contact model represent geometrical relation robot pusher object environment learn set parametric local motion model predict contact change throughout push set contact motion model represent internal model adjusting shape distribution physical parameter modify internal model response uniform distribution yield coarse estimate information available novel context ie unbiased predictor accurate predictor learned specific environmentobject pair eg low frictionhigh mass ie biased predictor effectiveness approach shown simulated environment pioneer dx robot need predict push outcome novel object provide proof concept real robot train object cube cylinder total push various condition test object encompassing variety shape size physical parameter total predicted push outcome result show biased unbiased predictor reliably produce prediction line outcome carefully tuned physic simulator | 0 | [
101,
4083,
4651,
3085,
5245,
16924,
8066,
3117,
6123,
3259,
4986,
4083,
4651,
3085,
2830,
2944,
5245,
16924,
11243,
3117,
6123,
5335,
3737,
17547,
4187,
2592,
2800,
16599,
4553,
11498,
12589,
4722,
2944,
5245,
8290,
2714,
2529,
12939,
8957,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1909.07808 | chinese street view text large scale chinese text reading partially supervised learning existing text reading benchmark make difficult evaluate performance advanced deep learning model large vocabulary due limited amount training data address issue introduce new large scale text reading benchmark dataset named chinese street view text svt street view image least time large existing chinese text reading benchmark recognize chinese text wild keeping large scale datasets labeling cost effective propose annotate one part csvt dataset image location text label full annotation add image corresponding text interest region given weak annotation exploit rich information weakly annotated data design text reading network partially supervised learning framework enables localize recognize text learn fully weakly annotated data simultaneously localize best matched text proposal weakly labeled image propose online proposal matching module incorporated whole model spotting keyword region sharing parameter end end training compared fully supervised training algorithm model improve end end recognition performance remarkably score labeling cost proposed model also achieve state art result icdar rctw dataset demonstrates effectiveness proposed partially supervised learning framework | 0 | [
101,
2822,
2395,
3193,
3793,
2312,
4094,
2822,
3793,
3752,
6822,
13588,
4083,
4493,
3793,
3752,
6847,
10665,
2191,
3697,
16157,
2836,
3935,
2784,
4083,
2944,
2312,
16188,
2349,
3132,
3815,
2731,
2951,
4769,
3277,
8970,
2047,
2312,
4094,
3... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2005.04828 | pretraining federated text model next word prediction federated learning decentralized approach training model distributed device summarizing local change sending aggregate parameter local model cloud rather data research employ idea transfer learning federated training next word prediction nwp conduct number experiment demonstrating enhancement current baseline federated nwp model successful specifically compare federated training baseline randomly initialized model various combination pretraining approach including pretrained word embeddings whole model pretraining followed federated fine tuning nwp dataset stack overflow post realize lift performance using pretrained embeddings without exacerbating number required training round memory footprint also observe notable difference using centrally pretrained network especially depending datasets used research offer effective yet inexpensive improvement federated nwp pave way rigorous experimentation transfer learning technique federated learning | 0 | [
101,
3653,
23654,
2075,
7349,
16848,
3793,
2944,
2279,
2773,
17547,
7349,
16848,
4083,
11519,
7941,
3550,
3921,
2731,
2944,
5500,
5080,
7680,
7849,
6026,
2334,
2689,
6016,
9572,
16381,
2334,
2944,
6112,
2738,
2951,
2470,
12666,
2801,
4651,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2105.01427 | code channel paper collection result combinatorial property code channel z channel error fraction take input length binary codeword injects adversarial manner asymmetric error ie error zero bit flip known largest list decodable code channel error fraction exponential size le critical value call plotkin point constant size larger threshold list decoding plotkin point known equal unique decoding paper derive various result size largest code list decoding plotkin point particular show largest list decodable code plotkin point size also devise upper lower bound exponential size code list decoding plotkin point | 0 | [
101,
3642,
3149,
3259,
3074,
2765,
22863,
23207,
4818,
3200,
3642,
3149,
1062,
3149,
7561,
12884,
2202,
7953,
3091,
12441,
3642,
18351,
1999,
20614,
2015,
4748,
14028,
10980,
2140,
5450,
2004,
24335,
12589,
7561,
29464,
7561,
5717,
2978,
11... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2004.10038 | spectral gap diameter cayley graph obtain new bound connecting first non trivial eigenvalue laplace operator graph diameter graph effective graph small diameter graph number maximal path comparable expectation | 1 | [
101,
17435,
6578,
6705,
6187,
12844,
2100,
10629,
6855,
2047,
5391,
7176,
2034,
2512,
20610,
1041,
29206,
10175,
5657,
5001,
19217,
6872,
10629,
6705,
10629,
4621,
10629,
2235,
6705,
10629,
2193,
29160,
4130,
12435,
17626,
102
] | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
1912.05565 | exactly solvable model beyond cohomology symmetry protected topological phase construct exactly solvable commuting projector model dimensional symmetry protected topological phase spt outside cohomology classification spts model described decorated domain wall construction three fermion walker wang phase domain wall describe anomalous nature phase several way one interesting feature contrast cohomology phase effective symmetry dimensional boundary cannot described quantum circuit instead nontrivial quantum cellular automaton qca related property codimension two defect example termination domain wall trivial boundary carry nontrivial chiral central charge mod also construct gapped symmetric topologically ordered boundary state model constitutes anomalous symmetry enriched topological phase outside classification arxiv define corresponding anomaly indicator | 4 | [
101,
3599,
14017,
12423,
2944,
3458,
2522,
23393,
6779,
14991,
5123,
24309,
4403,
9570,
3599,
14017,
12423,
4012,
28120,
2075,
2622,
2953,
2944,
8789,
14991,
5123,
24309,
4403,
11867,
2102,
2648,
2522,
23393,
6779,
5579,
11867,
3215,
2944,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1901.08241 | location reference identification tweet emergency deep learning approach twitter recently used crisis communicate official provide rescue relief operation real time geographical location information event well user vitally important scenario identification geographic location one challenging task location information field user location place name tweet reliable extraction location information tweet text difficult contains lot non standard english grammatical error spelling mistake non standard abbreviation research aim extract location word used tweet using convolutional neural network cnn based model achieved exact matching score hamming loss score tweet related earthquake model able extract even three four word long location reference also evident exact matching score finding paper help early event localization emergency situation real time road traffic management localized advertisement various location based service | 0 | [
101,
3295,
4431,
8720,
1056,
28394,
2102,
5057,
2784,
4083,
3921,
10474,
3728,
2109,
5325,
10639,
2880,
3073,
5343,
4335,
3169,
2613,
2051,
10056,
3295,
2592,
2724,
2092,
5310,
8995,
2135,
2590,
11967,
8720,
9183,
3295,
2028,
10368,
4708,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2001.11555 | higher order estimate long time solution kahler ricci flow article study higher order regularity kahler ricci flow compact kahler manifold semi ample canonical line bundle proved using parabolic analogue hein tosattis work collapsing calabi yau metric generic fiber iitaka fibration biholomorphic flow converges locally smoothly away singular fiber negative kahler einstein metric base manifold particular proved ricci curvature flow uniformly bounded compact subset away singular fiber generic fiber biholomorphic | 1 | [
101,
3020,
2344,
10197,
2146,
2051,
5576,
10556,
13620,
26220,
6895,
4834,
3720,
2817,
3020,
2344,
3180,
3012,
10556,
13620,
26220,
6895,
4834,
9233,
10556,
13620,
19726,
4100,
20851,
18562,
2240,
14012,
4928,
2478,
11498,
18647,
21800,
2002,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2012.0017 | stochastic gravitational wave background quantum gravity among cosmological quantum gravity quantum gravity inspired scenario predict blue tilted primordial tensor spectrum explore five check whether generate stochastic gravitational wave background detectable present future interferometer non local quantum gravity string gas cosmology new ekpyrotic scenario brandenberger ho non commutative inflation multi fractional spacetimes show non local quantum gravity unobservable model reach strain sensitivity decigo ligo virgo kagra lisa einstein telescope quantum gravity model red tilted spectrum loop quantum cosmology exceptionally tiny quantum correction wheeler dewitt quantum cosmology found non detectable | 10 | [
101,
2358,
11663,
20875,
19790,
4400,
4281,
8559,
8992,
2426,
2522,
25855,
9966,
8559,
8992,
8559,
8992,
4427,
11967,
16014,
2630,
9939,
26927,
5302,
25070,
23435,
8674,
8849,
2274,
4638,
3251,
9699,
2358,
11663,
20875,
19790,
4400,
4281,
1... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2106.06976 | game gans game theoretical model generative adversarial network generative adversarial network promising research direction ai community recently attracts considerable attention due ability generating high quality realistic data gans competing game two neural network trained adversarial manner reach nash equilibrium despite improvement accomplished gans last year remain several issue solve way tackle issue make advance lead rising research interest paper review literature leverage game theory gans address game model relieve specific generative model challenge improve gans performance particular firstly review preliminary including basic gan model game theory background present taxonomy summarize state art solution three significant category modified game model modified architecture modified learning method classification based modification made basic model proposed approach game theoretic perspective classify category several subcategories following proposed taxonomy explore main objective class review recent work group finally discus remaining challenge field present potential future research topic | 0 | [
101,
2208,
25957,
2015,
2208,
9373,
2944,
11416,
6024,
4748,
14028,
10980,
2140,
2897,
11416,
6024,
4748,
14028,
10980,
2140,
2897,
10015,
2470,
3257,
9932,
2451,
3728,
17771,
6196,
3086,
2349,
3754,
11717,
2152,
3737,
12689,
2951,
25957,
2... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2011.04994 | aim challenge learned image signal processing pipeline paper review second aim learned isp challenge provides description proposed solution result participating team solving real world raw rgb mapping problem goal map original low quality raw image captured huawei device photo obtained canon dslr camera considered task embraced number complex computer vision subtasks image demosaicing denoising white balancing color contrast correction demoireing etc target metric used challenge combined fidelity score psnr ssim solution perceptual result measured user study proposed solution significantly improved baseline result defining state art practical image signal processing pipeline modeling | 0 | [
101,
6614,
4119,
4342,
3746,
4742,
6364,
13117,
3259,
3319,
2117,
6614,
4342,
2003,
2361,
4119,
3640,
6412,
3818,
5576,
2765,
8019,
2136,
13729,
2613,
2088,
6315,
1054,
18259,
12375,
3291,
3125,
4949,
2434,
2659,
3737,
6315,
3746,
4110,
2... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2106.10653 | practical assessment generalization performance robustness deep network via contrastive example training image data transformation suggested contrastive example complement testing set generalization performance evaluation deep neural network dnns work propose practical framework contre word contre mean versus french us contrastive example dnn generalization performance estimation specifically contre follows assumption contrastive learning robust dnn model good generalization performance capable extracting consistent set feature making consistent prediction image varying data transformation incorporating set randomized strategy well designed data transformation training set contre adopts classification error fisher ratio generated contrastive example ass analyze generalization performance deep model complement testing set show effectiveness efficiency contre extensive experiment done using various dnn model three open source benchmark datasets thorough ablation study applicability analysis experiment result confirm behavior deep model contrastive example strongly correlated testing set contre robust measure generalization performance complementing testing set various setting | 0 | [
101,
6742,
7667,
2236,
3989,
2836,
15873,
2791,
2784,
2897,
3081,
5688,
3512,
2742,
2731,
3746,
2951,
8651,
4081,
5688,
3512,
2742,
13711,
5604,
2275,
2236,
3989,
2836,
9312,
2784,
15756,
2897,
1040,
10695,
2015,
2147,
16599,
6742,
7705,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2001.04791 | iterative bayesian based localization mechanism industry vertical propose evaluate iterative localization mechanism employing bayesian inference estimate position target using received signal strength measurement probability density function target coordinate estimated bayesian network herein consider iterative procedure whereby predictor posterior distribution updated sequential order whenever new measurement made available performance mechanism assessed term respective root mean square error kernel density estimation target coordinate numerical result showed proposed iterative mechanism achieves increasingly better estimation target node position updating round bayesian network new input measurement | 6 | [
101,
2009,
25284,
3016,
25253,
2241,
2334,
3989,
7337,
3068,
7471,
16599,
16157,
2009,
25284,
2334,
3989,
7337,
15440,
3016,
25253,
28937,
10197,
2597,
4539,
2478,
2363,
4742,
3997,
10903,
9723,
4304,
3853,
4539,
13530,
4358,
3016,
25253,
2... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2009.08467 | waveform systematics gravitational wave inference tidal parameter equation state binary neutron star signal gravitational wave signal binary neutron star coalescence carry information star equation state tidal signature major issue inference tidal parameter directly equation state systematic error introduced waveform approximants use bottom approach based gauge invariant phase analysis fisher information matrix investigate waveform systematics help identifying bias parameter estimation mock analysis different binary indicates systematics current waveform model dominate statistical error signal noise ratio snr implies bias inference reduced tidal parameter larger statistical credible interval example neutron star radius could constrained level snr systematics level apply approach gw snr confirm significant systematic effect present using optimal frequency range analysis estimate neutron star radius km latter consistent electromagnetic informed prior recent nicer measurement exploring snr view third generation detector find current waveform model lead difference least sigma inference reduced tidal parameter value latter conclude current waveform model including numerical relativity insufficient infer equation state loudest potentially informative event observed advanced third generation detector | 11 | [
101,
4400,
14192,
11778,
2015,
19790,
4400,
28937,
15065,
16381,
8522,
2110,
12441,
20393,
2732,
4742,
19790,
4400,
4742,
12441,
20393,
2732,
5317,
28964,
4287,
2592,
2732,
8522,
2110,
15065,
8085,
2350,
3277,
28937,
15065,
16381,
3495,
8522,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2107.14624 | convergent star product cotangent bundle lie group connected real lie group consider canonical standard ordered star product arising canonical global symbol calculus based half commutator connection star product trivially converges polynomial function thanks homogeneity define nuclear frechet algebra certain analytic function standard ordered star product shown well defined continuous multiplication depending holomorphically deformation parameter nuclear frechet algebra realized completed projective tensor product nuclear frechet algebra entire function appropriate nuclear frechet algebra function passage weyl ordered star product ie gutt star product shown preserve function space yielding continuity gutt star product holomorphic dependence | 1 | [
101,
28314,
3372,
2732,
4031,
26046,
22043,
3372,
14012,
4682,
2177,
4198,
2613,
4682,
2177,
5136,
18562,
3115,
3641,
2732,
4031,
17707,
18562,
3795,
6454,
19276,
2241,
2431,
4012,
28120,
8844,
4434,
2732,
4031,
20610,
2135,
28314,
2015,
17... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1604.00719 | renormalization almost commuting pair paper give new prove hyperbolicity renormalization critical circle map using formalism almost commuting pair extend renormalization two dimensional dissipative map annulus small perturbation one dimensional critical circle map finally demontsrate two dimensional map lie stable set renormalization operator posse attractor topologically circle circle critical dynamic topologically smoothly conjugate rigid rotation | 1 | [
101,
17738,
17830,
22731,
2471,
4012,
28120,
2075,
3940,
3259,
2507,
2047,
6011,
23760,
18647,
3012,
17738,
17830,
22731,
4187,
4418,
4949,
2478,
5337,
2964,
2471,
4012,
28120,
2075,
3940,
7949,
17738,
17830,
22731,
2048,
8789,
4487,
18719,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1808.00031 | fast approximate clearance evaluation rover articulated suspension system present light weight body terrain clearance evaluation algorithm automated path planning nasa mar rover extraterrestrial path planning challenging due combination terrain roughness severe limitation computational resource path planning cluttered andor uneven terrain requires repeated safety check candidate path small interval predicting future rover state requires simulating vehicle settling terrain involves inverse kinematics problem iterative nonlinear optimization geometric constraint however expensive computation intractable slow spacecraft computer rad used curiosity mar rover upcoming mar rover propose approximate clearance evaluation ace algorithm obtains conservative bound vehicle clearance attitude suspension angle without iterative computation obtains bound estimating lowest highest height wheel may reach given underlying terrain calculating worst case vehicle configuration associated extreme wheel height bound guaranteed conservative hence ensuring vehicle safety autonomous navigation ace planned used part new onboard path planner mar rover paper describes algorithm detail validates claim conservatism fast computation experiment | 0 | [
101,
3435,
15796,
14860,
9312,
13631,
20742,
8636,
2291,
2556,
2422,
3635,
2303,
9291,
14860,
9312,
9896,
12978,
4130,
4041,
9274,
9388,
13631,
4469,
3334,
28533,
14482,
4130,
4041,
10368,
2349,
5257,
9291,
5931,
2791,
5729,
22718,
15078,
7... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1904.02077 | graph based nearest neighbor search promise failure recently graph based nearest neighbor search get popular large scale retrieval task attractiveness type approach lie superior performance known nearest neighbor search approach well genericness various metric paper role two strategy namely hierarchical structure graph diversification adopted key step graph based approach investigated find hierarchical structure could achieve much better logarithmic complexity scaling claimed original paper particularly high dimensional case moreover find similar high search speed efficiency one hierarchical structure could achieved support flat nn graph graph diversification finally point difficulty faced graph based search approach directly linked curse dimensionality | 0 | [
101,
10629,
2241,
7205,
11429,
3945,
4872,
4945,
3728,
10629,
2241,
7205,
11429,
3945,
2131,
2759,
2312,
4094,
26384,
4708,
8702,
2791,
2828,
3921,
4682,
6020,
2836,
2124,
7205,
11429,
3945,
3921,
2092,
12391,
2791,
2536,
12046,
3259,
2535,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1904.07733 | subjective assessment text complexity dataset german language paper present textcomplexityde dataset consisting sentence german language taken wikipedia article different article genre used developing text complexity predictor model automatic text simplification german language dataset includes subjective assessment different text complexity aspect provided german learner level addition contains manual simplification sentence provided native speaker subjective assessment simplified sentence participant target group subjective rating collected using laboratory study crowdsourcing approach | 0 | [
101,
20714,
7667,
3793,
11619,
2951,
13462,
2446,
2653,
3259,
2556,
3793,
9006,
19386,
3012,
3207,
2951,
13462,
5398,
6251,
2446,
2653,
2579,
16948,
3720,
2367,
3720,
6907,
2109,
4975,
3793,
11619,
16014,
2953,
2944,
6882,
3793,
21934,
2475... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2010.05127 | approximation algorithm stochastic minimum norm combinatorial optimization motivated need growing interest modeling uncertainty data introduce study em stochastic minimum norm optimization underlying combinatorial optimization problem cost involved em random variable given distribution feasible solution induces random multidimensional cost vector given certain objective function goal find solution depend realization cost minimizes expected objective value instance stochastic load balancing job random processing time need assigned machine induced cost vector machine load vector recently deterministic setting chakrabarty swamy citechakrabartysa considered fairly broad suite objective wherein seek minimize norm cost vector given em arbitrary monotone symmetric norm stochastic minimum norm optimization work broad class objective seek solution minimizes em expected norm induced cost vector give general framework devising algorithm stochastic minimum norm combinatorial optimization using obtain approximation algorithm stochastic minimum norm version load balancing spanning tree problem two key technical contribution work structural result independent interest connecting stochastic minimum norm optimization simultaneous optimization emphsmall collection expected norm showing tackle expected norm minimization leveraging technique used deal minimizing expected maximum circumventing difficulty posed non separable nature norm | 0 | [
101,
20167,
9896,
2358,
11663,
20875,
6263,
13373,
22863,
23207,
4818,
20600,
12774,
2342,
3652,
3037,
11643,
12503,
2951,
8970,
2817,
7861,
2358,
11663,
20875,
6263,
13373,
20600,
10318,
22863,
23207,
4818,
20600,
3291,
3465,
2920,
7861,
672... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1910.0465 | learning remember multi task teacher recent study catastrophic forgetting sequential learning typically focus fixing accuracy prediction previously learned task paper argue output neural network subject rapid change learning new data distribution network appear forget everything still contain useful representation towards previous task instead enforcing output accuracy stay propose reduce effect catastrophic forgetting representation level output layer quickly recovered later small number example towards goal propose experimental setup measure amount representational forgetting develop novel meta learning algorithm overcome issue proposed meta learner produce weight update sequential learning network mimicking multi task teacher network representation show meta learner improve learned representation new task maintaining good representation old task | 0 | [
101,
4083,
3342,
4800,
4708,
3836,
3522,
2817,
23546,
17693,
25582,
4083,
4050,
3579,
15887,
10640,
17547,
3130,
4342,
4708,
3259,
7475,
6434,
15756,
2897,
3395,
5915,
2689,
4083,
2047,
2951,
4353,
2897,
3711,
5293,
2673,
2145,
5383,
6179,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2001.11459 | almost sure convergence largest smallest eigenvalue high dimensional sample correlation matrix paper show largest smallest eigenvalue sample correlation matrix stemming independent observation dimensional time series iid component converge almost surely respectively truncated variance entry distribution almost slowly varying condition describe via moment property self normalized sum moreover empirical spectral distribution sample correlation matrix converge weakly probability marchenko pastur law extends result bai zhou compare behavior eigenvalue sample covariance sample correlation matrix argue latter seems robust particular case infinite fourth moment briefly address practical issue estimation extreme eigenvalue simulation study proof use method moment combined path shortening algorithm efficiently us structure sample correlation matrix calculate precise bound matrix norm believe new approach could use random matrix theory | 1 | [
101,
2471,
2469,
19143,
2922,
10479,
1041,
29206,
10175,
5657,
2152,
8789,
7099,
16902,
8185,
3259,
2265,
2922,
10479,
1041,
29206,
10175,
5657,
7099,
16902,
8185,
29217,
2981,
8089,
8789,
2051,
2186,
2462,
2094,
6922,
28314,
2471,
7543,
44... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1902.06318 | homotopy invariant via koszul duality show integral cohomology ring modulus space stable rational marked curve koszul answer open question manin using machinery koszul space developed berglund compute rational homotopy lie algebra space obtain estimate betti number free loop space case torsion coefficient also prove conjecture generalisation main result | 1 | [
101,
24004,
14399,
2100,
23915,
3081,
12849,
17112,
5313,
7037,
3012,
2265,
9897,
2522,
23393,
6779,
3614,
16913,
11627,
2686,
6540,
11581,
4417,
7774,
12849,
17112,
5313,
3437,
2330,
3160,
23624,
2078,
2478,
10394,
12849,
17112,
5313,
2686,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2101.0994 | supervised unsupervised approach controlling narrow lexical focus sequence sequence speech synthesis although sequence sequence s architecture become state art speech synthesis capable generating output approach perceptual quality natural sample limited lack flexibility come controlling output work present framework capable controlling prosodic output via set concise interpretable disentangled parameter apply framework realization emphatic lexical focus proposing variety architecture designed exploit different level supervision based availability labeled resource evaluate approach via listening test demonstrate able successfully realize controllable focus maintaining higher naturalness established baseline explore different approach compare synthesizing target voice without labeled data | 3 | [
101,
13588,
4895,
6342,
4842,
11365,
2098,
3921,
9756,
4867,
16105,
9289,
3579,
5537,
5537,
4613,
10752,
2348,
5537,
5537,
1055,
4294,
2468,
2110,
2396,
4613,
10752,
5214,
11717,
6434,
3921,
2566,
3401,
13876,
8787,
3737,
3019,
7099,
3132,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2002.08851 | adaptive finite time stability delayed system via aperiodically intermittent control quantized control brief set finite time stability fntsta theory dynamical system bounded time varying delay via aperiodically intermittent control aic quantized control qc general qc designed brief bound time varying delay required le infimum aic two phase method pm applied solve fntsta delayed system ie divide whole proof process two phase one phase process norm system error evolving initial value process norm system error evolving proving two phase use fnt realize whole fntsta system proved also design adaptive rule prove validity rigorously furthermore obtained theory used discus finite time synchronization fntsyn neural network application finally simulation given illustrate effectiveness theoretical result | 12 | [
101,
19293,
10713,
2051,
9211,
8394,
2291,
3081,
23957,
9488,
14808,
3973,
23852,
2491,
24110,
23355,
2491,
4766,
2275,
10713,
2051,
9211,
1042,
7666,
2696,
3399,
8790,
2389,
2291,
10351,
2051,
9671,
8536,
3081,
23957,
9488,
14808,
3973,
23... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2107.03121 | thermal conductivity prediction nanoparticle packed bed using modified johnson kendall robert model nanoparticle packed bed npbs demonstrated potential thermal insulation reducing thermal conductivity requires theoretical understanding thermal conduction till theoretical model framework effective medium approach ema widely developed thermal conductivity prediction npb model corresponding architecture parameter usually evaluated classical johnson kendall robert jkr model unfortunately size effect usually ignored jkr model resulting inferior ability accurately predict geometrical information npb work modified jkr model including size effect young modulus integrated ema model prediction experimental result int heat mass tran explained result developed model illustrates advantage prediction solid phase thermal conductivity k especially npb low porosity work provides modified jkr model improve accuracy existing ema model prediction npb | 2 | [
101,
9829,
6204,
7730,
17547,
28991,
19362,
4588,
2571,
8966,
2793,
2478,
6310,
3779,
14509,
2728,
2944,
28991,
19362,
4588,
2571,
8966,
2793,
27937,
5910,
7645,
4022,
9829,
25710,
8161,
9829,
6204,
7730,
5942,
9373,
4824,
9829,
6204,
3258,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2104.03169 | empowering prosumer community smart grid wireless communication federated edge learning exponential growth distributed energy resource enabling transformation traditional consumer smart grid prosumers transition present promising opportunity sustainable energy trading yet integration prosumers energy market imposes new consideration designing unified sustainable framework efficient use power communication infrastructure furthermore several issue need tackled adequately promote adoption decentralized renewable oriented system communication overhead data privacy scalability sustainability article present different aspect challenge addressed building efficient energy trading market relation communication smart decision making accordingly propose multi level pro decision framework prosumer community achieve collective goal since individual decision prosumers mainly driven individual self sufficiency goal framework prioritizes individual prosumers decision relies wireless network fast coordination among community member fact prosumer predicts energy production consumption make proactive trading decision response collective level request moreover collaboration community extended including collaborative training prediction model using federated learning assisted edge server prosumer home area equipment addition preserving prosumers privacy show evaluation training prediction model using federated learning yield high accuracy different energy resource reducing communication overhead | 3 | [
101,
7861,
23948,
4013,
23545,
2099,
2451,
6047,
8370,
9949,
4807,
7349,
16848,
3341,
4083,
27258,
3930,
5500,
2943,
7692,
12067,
8651,
3151,
7325,
6047,
8370,
4013,
23545,
2869,
6653,
2556,
10015,
4495,
9084,
2943,
6202,
2664,
8346,
4013,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2106.12799 | phase field model cell spreading dynamic suggest phase field model describe cell spreading flat substrate model simplified version minimal model developed model couple order parameter polarization orientation vector field actin network derive closed integro differential equation governing cell spreading dynamic flat substrate includes normal velocity membrane curvature volume relaxation rate function determined molecular effect subcell level adhesion effect equation easily solved numerically result agreement early fast phase observed experimentally also find agreement universal power law suggest cell adhesion contact area versus time behave early stage cell spreading dynamic slow next stage | 2 | [
101,
4403,
2492,
2944,
3526,
9359,
8790,
6592,
4403,
2492,
2944,
6235,
3526,
9359,
4257,
16305,
2944,
11038,
2544,
10124,
2944,
2764,
2944,
3232,
2344,
16381,
11508,
3989,
10296,
9207,
2492,
2552,
2378,
2897,
18547,
2701,
20014,
13910,
3217... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2107.09092 | learning sensor invariant embedding satellite data case study lake ice monitoring fusing satellite imagery acquired different sensor long standing challenge earth observation particularly across different modality optical synthetic aperture radar sar image explore joint analysis imagery different sensor light representation learning propose learn joint sensor invariant embedding feature representation within deep neural network application problem monitoring lake ice alpine lake reach temporal resolution requirement swiss global climate observing system gcos office combine three image source sentinel sar sar terra modis suomi npp viirs large gap optical sar domain sensor resolution make challenging instance sensor fusion problem approach classified feature level fusion learnt data driven manner proposed network architecture separate encoding branch image sensor feed single latent embedding ie common feature representation shared input subsequent processing step deliver comparable output irrespective sort input image used fusing satellite data map lake ice temporal resolution respectively miou score generalises well across different lake winter moreover set new state art determining important ice ice date target lake many case meeting gcos requirement | 0 | [
101,
4083,
13617,
23915,
7861,
8270,
4667,
5871,
2951,
2553,
2817,
2697,
3256,
8822,
11865,
7741,
5871,
13425,
3734,
2367,
13617,
2146,
3061,
4119,
3011,
8089,
3391,
2408,
2367,
16913,
23732,
9380,
12553,
18892,
7217,
18906,
3746,
8849,
410... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1911.05611 | uno uncertainty aware noisy multimodal fusion unanticipated input degradation fusion multiple sensor modality especially deep learning architecture active area study however explored aspect work whether method robust degradation across input modality especially must generalize degradation seen training work propose uncertainty aware fusion scheme effectively fuse input might suffer range known unknown degradation specifically analyze number uncertainty measure capture different aspect uncertainty propose novel way fuse degraded input scaling modality specific output softmax probability additionally propose novel data dependent spatial temperature scaling method complement existing uncertainty measure finally integrate uncertainty scaled output modality using probabilistic noisy fusion method photo realistic simulation environment airsim show method achieves significantly better result semantic segmentation task compared state art fusion architecture range degradation eg fog snow frost various type noise unknown training specifically improve upon state art mean iou various degradation abhinav valada rohit mohan wolfram burgard self supervised model adaptation multimodal semantic segmentation arxiv print arxiv aug arxiv arxiv cscv | 0 | [
101,
27776,
12503,
5204,
20810,
4800,
5302,
9305,
10077,
14477,
16778,
6895,
17585,
2094,
7953,
16627,
10077,
3674,
13617,
16913,
23732,
2926,
2784,
4083,
4294,
3161,
2181,
2817,
2174,
10641,
7814,
2147,
3251,
4118,
15873,
16627,
2408,
7953,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1912.04466 | doublade unknown vulnerability detection smart contract via abstract signature matching refined detection rule prosperity smart contract blockchain technology various security analyzer proposed academia industry address associated risk yet exist high quality benchmark smart contract vulnerability security research study propose approach towards building high quality vulnerability benchmark approach consists two part first improve recall propose search similar vulnerability automated way leveraging abstract vulnerability signature av second remove false positive fps due av based matching summarize detection rule existing tool apply refined rule considering various defense mechanism dm integrating av based code matching refined detection rule rdr approach achieves higher precision recall collected contract build benchmark consisting vulnerability covering five different vulnerability type identified together tool doublade three scanner additionally conduct comparison doublade others additional contract result show doublade yield better detection accuracy similar execution time | 0 | [
101,
2079,
12083,
27266,
2063,
4242,
18130,
10788,
6047,
3206,
3081,
10061,
8085,
9844,
15514,
10788,
3627,
14165,
6047,
3206,
3796,
24925,
2078,
2974,
2536,
3036,
17908,
2099,
3818,
16926,
3068,
4769,
3378,
3891,
2664,
4839,
2152,
3737,
68... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2006.01166 | role quantum coherence thermal machine oldest important application thermodynamics operation refrigeration well production useful energy part effort understand develop thermodynamics quantum regime focusing harnessing quantum effect operation review present recent development regarding role quantum coherence performance thermal machine device realising thermodynamic operation known intricate subject part largely model dependent review recent result allow u identify general tendency suggest future direction | 13 | [
101,
2535,
8559,
2522,
5886,
10127,
9829,
3698,
4587,
2590,
4646,
1996,
10867,
7716,
18279,
22924,
3169,
25416,
3089,
4590,
3370,
2092,
2537,
6179,
2943,
2112,
3947,
3305,
4503,
1996,
10867,
7716,
18279,
22924,
8559,
6939,
7995,
17445,
2075... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2004.01736 | data driven solution stochastic differential equation using maximum entropy basis function paper present data driven approach uncertainty propagation particular consider stochastic differential equation parametric uncertainty solution differential equation approximated using maximum entropy maxent basis function similar polynomial chaos expansion maxent basis function derived available data maximization information theoretic entropy therefore need specify basis function beforehand compare proposed maxent based approach existing method | 1 | [
101,
2951,
5533,
5576,
2358,
11663,
20875,
11658,
8522,
2478,
4555,
23077,
3978,
3853,
3259,
2556,
2951,
5533,
3921,
12503,
20594,
3327,
5136,
2358,
11663,
20875,
11658,
8522,
11498,
12589,
12503,
5576,
11658,
8522,
15796,
2094,
2478,
4555,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1910.00735 | temperature based hardware trojan ring oscillator based trngs true random number generator trngs essential component cryptographic design used generate private key encryption authentication used masking countermeasure work present mechanism design stealthy parametric hardware trojan ring oscillator based trng architecture proposed yang et al isscc trojan triggered malicious trng generates predictable non random output trojan require additional logic even single gate purely based subtle manipulation sub transistor level underlying concept disable entropy source high temperature trigger trojan ensuring trojan infected trng work correctly normal condition show attack performed trojan infected trng design attacker us stochastic markov chain model predict reduced entropy output | 0 | [
101,
4860,
2241,
8051,
23445,
3614,
9808,
6895,
4571,
4263,
2241,
19817,
3070,
2015,
2995,
6721,
2193,
13103,
19817,
3070,
2015,
6827,
6922,
19888,
13705,
2640,
2109,
9699,
2797,
3145,
21999,
27280,
2109,
7308,
2075,
4675,
4168,
3022,
5397,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1908.04737 | end end multi speaker speech recognition using speaker embeddings transfer learning paper present latest investigation end end automatic speech recognition asr overlapped speech propose train end end system conditioned speaker embeddings improved transfer learning clean speech proposed framework require parallel non overlapped speech material independent number speaker experimental result overlapped speech datasets show joint conditioning speaker embeddings transfer learning significantly improves asr performance | 0 | [
101,
2203,
2203,
4800,
5882,
4613,
5038,
2478,
5882,
7861,
8270,
4667,
2015,
4651,
4083,
3259,
2556,
6745,
4812,
2203,
2203,
6882,
4613,
5038,
2004,
2099,
17702,
5669,
4613,
16599,
3345,
2203,
2203,
2291,
22442,
5882,
7861,
8270,
4667,
20... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2001.02362 | n independent localized krylov bogoliubov de gennes method ultra fast numerical approach large scale inhomogeneous superconductors propose ultra fast numerical approach large scale inhomogeneous superconductors call localized krylov bogoliubov de gennes method lk bdg lk bdg method computational complexity local green function used calculate local density state mean field depend system size calculation cost self consistent calculation enables u open new avenue treating extremely large system million lattice site show power lk bdg method demonstrate self consistent calculation site penrose quasicrystal lattice vortex calculation site two dimensional nearest neighbor square lattice tight binding model many vortex also demonstrate take le second one cpu core calculate local density state whole energy range million site tight binding model | 4 | [
101,
1050,
2981,
22574,
1047,
23320,
4492,
22132,
10893,
12083,
4492,
2139,
8991,
5267,
4118,
11087,
3435,
15973,
3921,
2312,
4094,
1999,
23393,
23924,
14769,
3565,
8663,
8566,
24817,
16599,
11087,
3435,
15973,
3921,
2312,
4094,
1999,
23393,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2107.12990 | joint reconstruction dark energy modified growth evolution present first combined non parametric reconstruction three time dependent function capture departure standard cosmological model cdm expansion history gravitational effect matter light currently available combination background large scale structure data perform reconstruction without theory informed prior built general horndeski class scalar tensor theory correlate three function find combination data constrain combined eigenmodes three function respect prior allowing informative reconstruction cosmological model featuring non trivial time dependence interpret latter context well known tension datasets within cdm along discussing implication reconstruction modified gravity theory | 10 | [
101,
4101,
8735,
2601,
2943,
6310,
3930,
6622,
2556,
2034,
4117,
2512,
11498,
12589,
8735,
2093,
2051,
7790,
3853,
5425,
6712,
3115,
2522,
25855,
9966,
2944,
3729,
2213,
4935,
2381,
19790,
3466,
3043,
2422,
2747,
2800,
5257,
4281,
2312,
4... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2106.13367 | seanet towards knowledge graph based autonomic management software defined network automatic network management driven artificial intelligent technology heatedly discussed decade however current report mainly focus theoretic proposal architecture design work practical implementation real life network yet appear paper proposes effort toward implementation knowledge graph driven approach autonomic network management software defined network sdns termed seanet driven toco ontology seanet reprogrammed based mininet sdn emulator consists three core component knowledge graph generator sparql engine network management api knowledge graph generator represents knowledge telecommunication network management task formally represented ontology driven model expert experience network management rule formalized knowledge graph automatically inferenced sparql engine network management api able packet technology specific detail expose technology independent interface user experiment carried evaluate proposed work comparing commercial sdn controller ryu implemented language python evaluation result show seanet considerably faster circumstance ryu seanet code significantly compact benefit rdf reasoning seanet able achieve time complexity different scale knowledge graph traditional database achieve onlogn best developed network management api seanet enables researcher develop semantic intelligent application sdns | 0 | [
101,
5977,
3388,
2875,
3716,
10629,
2241,
8285,
3630,
7712,
2968,
4007,
4225,
2897,
6882,
2897,
2968,
5533,
7976,
9414,
2974,
9685,
2135,
6936,
5476,
2174,
2783,
3189,
3701,
3579,
14833,
13465,
2594,
6378,
4294,
2640,
2147,
6742,
7375,
26... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2104.09421 | generalization higher rank graph introduce call generalized higher rank graph class category equipped notion size extend higher rank graph also levi category introduced first author categorical setting graph group prove example generalized higher rank graph constructed using zappa szep product groupoids higher rank graph | 1 | [
101,
2236,
3989,
3020,
4635,
10629,
8970,
2655,
18960,
3020,
4635,
10629,
2465,
4696,
6055,
9366,
2946,
7949,
3020,
4635,
10629,
2036,
11902,
4696,
3107,
2034,
3166,
4937,
27203,
4292,
10629,
2177,
6011,
2742,
18960,
3020,
4635,
10629,
3833... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
1903.02657 | scattering mechanism modeling terahertz wireless communication paper provides analysis radio wave scattering frequency ranging microwave terahertz band eg ghz thz studying scattering power reradiated various type material different surface roughness first fundamental scattering reflection developed explained use wireless mobile radio effect scattering reflection coefficient rough surface investigated received power derived using two popular scattering model directive scattering d model radar cross section rcs model simulation wide range frequency material orientation two model measurement confirm accuracy d model ghz paper show scattering become prominent propagation mechanism frequency extend millimeter wave mmwave beyond time treated like simple reflection knowledge scattering effect critical appropriate realistic channel model support development massive multiple input multiple output mimo technique localization ray tracing tool design imaging future wireless system | 2 | [
101,
17501,
7337,
11643,
28774,
4430,
8743,
2480,
9949,
4807,
3259,
3640,
4106,
2557,
4400,
17501,
6075,
7478,
18302,
28774,
4430,
8743,
2480,
2316,
1041,
2290,
29066,
16215,
2480,
5702,
17501,
2373,
2128,
12173,
15070,
2536,
2828,
3430,
23... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2009.04447 | beyond observed connection link injection paper proposed textitlink injection novel method help differentiable graph machine learning model go beyond observed connection input data end end learning fashion find weak connection favor current task present input data via parametric link injection layer evaluate method node classification link prediction task using series state art graph convolution network result show link injection help variety model achieve better performance application empirical analysis show great potential method efficiently exploiting unseen connection injected link | 0 | [
101,
3458,
5159,
4434,
4957,
13341,
3259,
3818,
3793,
4183,
13767,
13341,
3117,
4118,
2393,
2367,
19210,
10629,
3698,
4083,
2944,
2175,
3458,
5159,
4434,
7953,
2951,
2203,
2203,
4083,
4827,
2424,
5410,
4434,
5684,
2783,
4708,
2556,
7953,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1906.01015 | silcc zoom co dark gas molecular cloud impact feedback magnetic field analyse co dark molecular gas content simulated molecular cloud silcc zoom project simulation reach resolution pc include co formation radiative stellar feedback magnetic field co dark gas found region local visual extinction number density cm gas temperature k co bright gas found number density cm temperature co dark gas fraction range scale inversely amount well shielded gas smaller magnetised molecular cloud show density chemical abundance along given line sight cannot properly determined projected quantity example pixel projected visual extinction co bright co dark attributed presence absence strong density enhancement along line sight producing synthetic co emission map simulation radmc show region intensity detection limit cloud factor around cm km spread factor implying similar uncertainty derived total mass even worse individual pixel based result suggest new approach determine mass relies availability co emission map reduces uncertainty cloud overall mass factor individual pixel ie sub pc scale factor | 10 | [
101,
9033,
15472,
2278,
24095,
2522,
2601,
3806,
8382,
6112,
4254,
12247,
8060,
2492,
20302,
23274,
2522,
2601,
8382,
3806,
4180,
23599,
8382,
6112,
9033,
15472,
2278,
24095,
2622,
12504,
3362,
5813,
7473,
2421,
2522,
4195,
10958,
9032,
602... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2106.08453 | gradient trained weight wide neural network align layerwise error scaled input correlation recent work examined deep neural network solve variety difficult problem incorporate statistic training data achieve success however existing result established limited setting work derive layerwise weight dynamic infinite width neural network nonlinear activation trained gradient descent show theoretically weight update aligned input correlation intermediate layer weighted error demonstrate empirically result also hold finite width wide network alignment result allows u formulate backpropagation free learning rule named align zero align ada theoretically achieve alignment backpropagation finally test learning rule benchmark problem feedforward recurrent neural network demonstrate wide network comparable performance backpropagation | 0 | [
101,
17978,
4738,
3635,
2898,
15756,
2897,
25705,
6741,
14244,
7561,
18953,
7953,
16902,
3522,
2147,
8920,
2784,
15756,
2897,
9611,
3528,
3697,
3291,
13265,
28093,
6553,
2731,
2951,
6162,
3112,
2174,
4493,
2765,
2511,
3132,
4292,
2147,
1854... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2006.10315 | similarity precursor solid state synthesis text mined scientific literature collecting analyzing vast amount information available solid state chemistry literature may accelerate understanding material synthesis however one major problem difficulty identifying material synthesis paragraph precursor target material study developed two step chemical named entity recognition cner model identify precursor target based information context around material entity using extracted data conducted meta analysis study similarity difference precursor context solid state synthesis quantify precursor similarity built substitution model calculate viability substituting one precursor another retaining target hierarchical clustering precursor demonstrate chemical similarity precursor extracted text data quantifying similarity precursor help provide foundation suggesting candidate reactant predictive synthesis model | 2 | [
101,
14402,
14988,
5024,
2110,
10752,
3793,
21846,
4045,
3906,
9334,
20253,
6565,
3815,
2592,
2800,
5024,
2110,
6370,
3906,
2089,
23306,
4824,
3430,
10752,
2174,
2028,
2350,
3291,
7669,
12151,
3430,
10752,
20423,
14988,
4539,
3430,
2817,
27... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1911.06049 | minimal polynomial power cycle ordinary representation symmetric alternating group denote alternating symmetric group degree respectively consider permutation whose nontrivial cycle length find minimal polynomial ordinary irreducible representation | 1 | [
101,
10124,
17505,
2373,
5402,
6623,
6630,
19490,
15122,
2177,
19090,
15122,
19490,
2177,
3014,
4414,
5136,
2566,
28120,
3370,
3005,
2512,
18886,
18660,
5402,
3091,
2424,
10124,
17505,
6623,
20868,
5596,
21104,
6630,
102
] | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
] |
2003.12208 | spectrerewind leaking secret past instruction transient execution attack utilize micro architectural covert channel leak secret accessible logical program execution commonly used micro architectural covert channel leave lasting footprint microarchitectural state example cache state change secret recovered transient execution completed paper present spectrerewind new approach create contention based covert channel transient execution attack approach covert channel established issuing necessary instruction logically transiently executed victim code unlike prior contention based covert channel require simultaneous multi threading smt spectrerewind support single hardware thread based covert channel making viable system attacker cannot utilize smt show contention floating point division unit commodity processor used create high performance kb low noise covert channel transient execution attack instead commonly used flushreload based cache covert channel implement meltdown attack utilizing proposed covert channel showing competitive performance compared stateof art cache based covert channel implementation also show covert channel work javascript engine chrome browser | 0 | [
101,
28699,
7913,
15603,
22254,
24325,
3595,
2627,
7899,
25354,
7781,
2886,
16462,
12702,
6549,
19813,
3149,
17271,
3595,
7801,
11177,
2565,
7781,
4141,
2109,
12702,
6549,
19813,
3149,
2681,
9879,
24319,
12702,
2906,
5428,
26557,
27431,
2140,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2107.03462 | study czernik ngc open cluster using ccd ubv photometric gaia edr data analysed open cluster czernik ngc using ccd ubv photometric gaia early data release edr photometric astrometric data structural parameter two cluster derived including physical size czernik ngc min calculated membership probability star based proper motion component released gaia edr identify member star cluster used membership probability taking account location impact binarity main sequence star used membership probability higher identify member star czernik ngc estimated colour excess metallicities separately using two colour diagram derive homogeneously determined parameter derived colour excess mag czernik mag ngc metallicities obtained first time cluster dex czernik dex ngc keeping reddening metallicity constant quantity fitted parsec model using colour magnitude diagram resulting estimated distance modulus age two cluster obtained distance modulus czernik mag ngc mag coincide age gyr myr respectively distance cluster calculated using gaia edr trigonometric parallax compared literature found good agreement distance obtained study literature present day mass function slope cluster comparable value salpeter czernik ngc | 10 | [
101,
2817,
1039,
6290,
8238,
27645,
2330,
9324,
2478,
10507,
2094,
1057,
2497,
2615,
6302,
12589,
11721,
2401,
3968,
2099,
2951,
20302,
23274,
2094,
2330,
9324,
1039,
6290,
8238,
27645,
2478,
10507,
2094,
1057,
2497,
2615,
6302,
12589,
1172... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2010.03129 | family constitutive model implemented plaxis simulate cemented mine backfill family constitutive model mine cemented backfill presented four formula density pressure dependency elastic modulus five formula density pressure dependency friction angle four formula age dependency elastic modulus effective cohesion incorporated isotropic hypoelasticity mohr coulomb perfect plasticity framework implemented plaxis user defined material model family includes standard mohr coulomb bolton leps barton hoek brown model trivial case nonlinear elasticity age dependency switched paper formulation model introduced basis numerical implementation outlined case history application cemented backfill sublevel stoping mine presented example | 2 | [
101,
2155,
9530,
16643,
8525,
6024,
2944,
7528,
20228,
8528,
2483,
26633,
11297,
2098,
3067,
2067,
8873,
3363,
2155,
9530,
16643,
8525,
6024,
2944,
3067,
11297,
2098,
2067,
8873,
3363,
3591,
2176,
5675,
4304,
3778,
24394,
21274,
16913,
1162... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1904.01595 | pushing energy cosmic frontier high energy astrophysical neutrino astrophysical neutrino recently discovered icecube neutrino telescope highest detected neutrino energy tev pev travel longest distance gpc size observable universe feature make naturally attractive probe fundamental particle physic property possibly tiny size energy scale unreachable mean decade icecube discovery saw many proposal particle physic study direction today proposal become reality spite prevalent astrophysical unknown showcase example studying fundamental neutrino physic scale including stringent test physic beyond standard model | 10 | [
101,
6183,
2943,
14448,
8880,
2152,
2943,
28625,
23302,
11265,
4904,
17815,
28625,
23302,
11265,
4904,
17815,
3728,
3603,
3256,
10841,
4783,
11265,
4904,
17815,
12772,
3284,
11156,
11265,
4904,
17815,
2943,
8915,
2615,
21877,
2615,
3604,
6493... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1902.06333 | formation native shortcut network two state protein folding network dynamic nd model proposed nd view protein folding partial ordering contact native state protein residue network prn nd process recreates prn keeping node fixed pdb position nd applies two fundamental folding principle loop entropy hydrophobic key folding interaction via two model parameter control probability edge creation probability node selection correlation experimental folding rate nd simulated folding rate strongest probability edge creation higher contact smaller sequence separation probability node selection higher node larger remaining degree within local neighborhood two condition also induce wider range simulated folding rate sign cooperative folding two state folder better correspondence experimental phi value calculated phi value calculated phi value based centrality node nd prns competitive phi value calculated transition state ensemble prns le phi value calculated local centrality measure introduced paper node prns two influence ordering prn contact also result route like formation native shortcut network hence two state like folding five nd variant tested best result macroscopic mesoscopic microscopic level produced single nd variant preliminary exploration mixed nd variant chain fold type suggests result could improved context sensitive model parameter nd model could extended multi state folder | 5 | [
101,
4195,
3128,
2460,
12690,
2897,
2048,
2110,
5250,
12745,
2897,
8790,
1050,
2094,
2944,
3818,
1050,
2094,
3193,
5250,
12745,
7704,
13063,
3967,
3128,
2110,
5250,
21755,
2897,
10975,
2078,
1050,
2094,
2832,
28667,
29313,
2015,
10975,
2078... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2103.05385 | naronet discovery tumor microenvironment element highly multiplexed image many effort made discover tumor specific microenvironment element tmes immunostained tissue section however identification yet unknown relevant tmes multiplex immunostained tissue remains challenge due number marker involved ten complexity spatial interaction present naronet us machine learning identify annotate known well novel tmes self supervised embeddings cell organized different level local cell phenotype cellular neighborhood us abundance tmes classify patient based biological clinical feature validate naronet using synthetic patient cohort adjustable incidence different tmes two cancer patient datasets synthetic real datasets naronet unsupervisedly identifies novel tmes relevant user defined classification task naronet requires patient level information render state art computational method accessible broad audience accelerating discovery biomarker signature | 0 | [
101,
6583,
20793,
2102,
5456,
13656,
12702,
2368,
21663,
2239,
3672,
5783,
3811,
3674,
19068,
3746,
2116,
3947,
2081,
7523,
13656,
3563,
12702,
2368,
21663,
2239,
3672,
5783,
1056,
7834,
10047,
23041,
28696,
21280,
8153,
2930,
2174,
8720,
2... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2107.00103 | limitation coupled cluster approximation highlyaccurate investigation rb reveal limitation several standard coupled cluster cc method perturbation theorybased noniterative approximate iterative treatment triple excitation applied thedetermination highly accurate potential energy curve pecs ionic dimer electronic ground state rb computation current interest understanding ofion atom interaction ultracold regime demonstrate coupled cluster methodslead unphysical long range barrier rb system barrier small clearly spoilsthe entire long range behavior pec effect also found system like li na calculation using new flexible framework obtaining leading perturbativetriples correction derived using analytic cc single double ccsd energy derivativeformulation demonstrate origin problem lie use amplitude obtained approximate cc single double triple ccsdt amplitude equation shown thatthe unphysical barrier related symmetry instability underlying hartree fock mean fieldsolution physically meaningful perturbative correction long range tail pec mayinstead obtained using symmetry broken reference determinant | 2 | [
101,
22718,
11211,
9324,
20167,
3811,
6305,
10841,
11657,
4812,
21144,
7487,
22718,
2195,
3115,
11211,
9324,
10507,
4118,
2566,
20689,
23757,
3399,
15058,
2094,
2512,
21646,
8082,
15796,
2009,
25284,
3949,
6420,
4654,
26243,
3370,
4162,
1996,... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2106.08534 | sharp convergence steady state allen cahn recent work found surprising breakdown symmetry conservation using standard numerical discretization high precision computed numerical solution corresponding nice initial data may converge completely incorrect steady state due gradual accumulation machine round error solved issue introducing new fourier filter technique solution certain band gap property investigate attracting basin steady state classify work possible bounded nontrivial steady state allen cahn equation characterize sharp dependence nontrivial steady state diffusion coefficient prove strict monotonicity associated energy particular establish certain self replicating property amongst hierarchy steady state give full classification energy profile develop new modulation theory prove sharp convergence steady state explicit rate profile | 1 | [
101,
4629,
19143,
6706,
2110,
5297,
6187,
7295,
3522,
2147,
2179,
11341,
12554,
14991,
5680,
2478,
3115,
15973,
5860,
13465,
3989,
2152,
11718,
24806,
15973,
5576,
7978,
3835,
3988,
2951,
2089,
28314,
3294,
16542,
6706,
2110,
2349,
16612,
2... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1902.05238 | atomic norm denoising complex exponential unknown waveform modulation non stationary blind super resolution extension traditional super resolution problem deal problem recovering fine detail coarse measurement non stationary blind super resolution problem appears many application including radar imaging single molecule microscopy computational photography etc growing interest solving non stationary blind super resolution task convex method due robustness noise strong theoretical guarantee motivated recent work atomic norm minimization blind inverse problem focus signal denoising problem non stationary blind super resolution particular use atomic norm regularized least square problem denoise sum complex exponential unknown waveform modulation quantify mean square error depends noise variance true signal parameter numerical experiment also implemented illustrate theoretical result | 1 | [
101,
9593,
13373,
7939,
10054,
2075,
3375,
27258,
4242,
4400,
14192,
25502,
2512,
17337,
6397,
3565,
5813,
5331,
3151,
3565,
5813,
3291,
3066,
3291,
13400,
2986,
6987,
20392,
10903,
2512,
17337,
6397,
3565,
5813,
3291,
3544,
2116,
4646,
216... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2007.13389 | quantum estimation acceleration temperature open quantum system open quantum system study evolution two level atom detector interacts given environment uniformly accelerated two level atom coupled massless scalar field minkowski vacuum evolves certain time find exists peak value quantum fisher information qfi acceleration indicates optimal precision estimation achieved choosing appropriate acceleration qfi different behavior different initial state parameter atom displaying periodicity however periodicity fade away evolution time mean initial state cannot affect later stable quantum state furthermore adding boundary observe peak value qfi increase atom close boundary show qfi protected boundary qfi fluctuates may exist two peak value certain moment expands detection range acceleration therefore enhance estimation precision acceleration choosing appropriate position acceleration periodicity qfi respect initial state parameter last longer time previous unbounded case indicates initial state protected boundary finally thermal bath boundary qfi temperature one peak value certain moment different qfi acceleration boundary periodicity also last longer time unbounded case show initial quantum state atom protected boundary two case | 13 | [
101,
8559,
24155,
16264,
4860,
2330,
8559,
2291,
2330,
8559,
2291,
2817,
6622,
2048,
2504,
13787,
19034,
11835,
2015,
2445,
4044,
27423,
14613,
2048,
2504,
13787,
11211,
3742,
3238,
26743,
2099,
2492,
8117,
15449,
11641,
19852,
2015,
3056,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1909.04405 | hierarchical planning ipc last year amount research hierarchical planning increased leading significant improvement performance planner however research diverging planner somewhat hard compare mostly caused fact standard set benchmark domain even common description language hierarchical planning problem consequence available planner support widely varying set feature almost none solve even parse problem developed another planner paper propose create new track ipc hierarchical planner compete competition result standardised description language broader support core feature language among planner set benchmark problem mean fairly objectively compare htn planner new challenge planner | 0 | [
101,
25835,
4041,
12997,
2278,
2197,
2095,
3815,
2470,
25835,
4041,
3445,
2877,
3278,
7620,
2836,
24555,
2174,
2470,
17856,
4726,
24555,
5399,
2524,
12826,
3262,
3303,
2755,
3115,
2275,
6847,
10665,
5884,
2130,
2691,
6412,
2653,
25835,
4041... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
2006.162 | near optimal sq lower bound agnostically learning halfspaces relus gaussian marginals study fundamental problem agnostically learning halfspaces relus gaussian marginals former problem given labeled example unknown distribution whose marginal distribution standard gaussian label arbitrary goal output hypothesis loss loss best fitting halfspace latter problem given labeled example unknown distribution whose marginal distribution standard gaussian label arbitrary goal output hypothesis square loss square loss best fitting relu prove statistical query sq lower bound problem sq lower bound provide strong evidence current upper bound task essentially best possible | 6 | [
101,
2379,
15502,
5490,
2896,
5391,
12943,
28199,
3973,
4083,
2431,
23058,
2015,
2128,
7393,
11721,
17854,
2937,
14785,
2015,
2817,
8050,
3291,
12943,
28199,
3973,
4083,
2431,
23058,
2015,
2128,
7393,
11721,
17854,
2937,
14785,
2015,
2280,
... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
1809.05659 | molding curved structure selective heating interest fabricate curved surface three dimension easily available homogeneous material form flat sheet aim obtain surface desired intrinsic riemannian metric get desired embedding translation rotation riemannian metric alone need uniquely determine paper demonstrate three generic method molding flat sheet thermo responsive plastic selective contraction induced targeted heating method involve cutting gluing property share origami first method inspired tailoring usual method making garment plain piece cloth unlike usual tailoring method produce desired embedding particular get desired intrinsic riemannian metric second method aim bring desired new riemannian metric via appropriate pattern local contraction without directly controlling embedding third method based triangulation seek induce desired local distance result getting desired embedding particular also give u target riemannian metric second third method also first method special case surface revolution algorithmic nature give theoretical account method followed illustrated example different shape physically molded method | 4 | [
101,
18282,
2075,
9203,
3252,
13228,
10808,
3037,
8313,
3686,
9203,
3302,
2093,
9812,
4089,
2800,
24854,
3430,
2433,
4257,
7123,
6614,
6855,
3302,
9059,
23807,
15544,
17545,
2937,
12046,
2131,
9059,
7861,
8270,
4667,
5449,
9963,
15544,
1754... | [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1... |
End of preview. Expand in Data Studio
No dataset card yet
- Downloads last month
- 6