Sinkhorn Normalization of Diffusion Kernels

ORID LJdacnMXkr · tags icml2026-repro paper-LJdacnMXkr

#StatusPageArtifactClaim excerpt
1VERIFIED 2/201-symmetric-smoothing-operator-positive-coefficienartifactTheorem 4.1 shows a symmetric smoothing operator with positive coefficients can …
2VERIFIED 2/202-gaussian-exponential-kernels-sinkhorn-normalizedartifactTheorem 4.2 proves that for Gaussian and exponential kernels, the Sinkhorn-norma…
3VERIFIED 2/203-symmetric-sinkhorn-algorithm-empirically-requireartifactThe symmetric Sinkhorn algorithm empirically requires only 5 to 10 iterations to…
4VERIFIED 2/204-normalized-operators-simultaneously-satisfy-symmartifactThe normalized operators simultaneously satisfy symmetry, mass conservation, ent…
5VERIFIED 2/205-method-demonstrated-point-clouds-sparseartifactThe method is demonstrated on point clouds, sparse voxel grids (jaw bone geometr…
6VERIFIED 2/206-shape-analysis-experiments-armadillo-meshartifactShape analysis experiments on the Armadillo mesh show spectral distributions rem…

Open logbook index · logbook.json