DMW relaxes and lower-bounds GW by transporting distributions of sampled distance matrices, with finite-sample bounds depending on intrinsic dimension and sliced/multi-scale variants for computation.
Marco Cuturi
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Distance-Matrix Wasserstein Statistics for Scalable Gromov--Wasserstein Learning
DMW relaxes and lower-bounds GW by transporting distributions of sampled distance matrices, with finite-sample bounds depending on intrinsic dimension and sliced/multi-scale variants for computation.