AI Paper Atlas

Research map / Learning theory and algorithms / Optimal transport

Fast Estimation of Wasserstein Distances via Regression on Sliced Wasserstein Distances

Khai Nguyen, Hai Nguyen, Nhat Ho

ICLR 2026 · Learning theory and algorithms › Optimal transport · approach: Sliced Wasserstein variants

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use RG regression on sliced Wasserstein distances instead of Wormhole (100 training samples)Wasserstein approximation · MNIST Point Cloud, ShapeNetV2, MERFISH, scRNA-seqR^2: MNIST 0.28 -> 0.93; MERFISH -3.6 -> 0.96Improved
unconstrained instead of constrained RG (ablation)RG-s · 4 datasets, 100 samplesR^2: MNIST 0.84 -> 0.93; scRNA-seq 1.00 -> 0.99Mixed
add lower+upper bound predictors (RG-se) rather than only SW and PWD (RG-s) (ablation)RG · ShapeNetV2 100 samplesR^2: 0.94 -> 0.95Improved
use optimization-based predictors (Max-SW, Min-SWGG) only: RG-o vs RG-s (ablation)RG · ShapeNetV2 kNN k=1accuracy (%): 82.1 -> 65.1Hurt
use RG-seo distances instead of single sliced metrics for kNNShapeNetV2 classification · ShapeNetV2 kNN k=5accuracy (%): 72.5 (EBSW, k=1) -> 83.5 (RG-seo, k=5)Improved
RG instead of Sinkhorn / Linear OTWasserstein approximation · ShapeNetV2R^2: Linear OT 0.918 -> RG-s 0.942; Sinkhorn -1.549Improved
RG-s instead of exact Wasserstein for inference200 samples (19,900 pairs) · ShapeNetV2inference time: 5,030.966s -> 21.408sImproved
replace Wasserstein calls in Wormhole with calibrated RG (RG-Wormhole)Wormhole · ModelNet40/ShapeNetV2training time and qualityImproved
train RG on intra-class pairs only, test on inter-class pairsRG · ShapeNetV2R^2: 0.902, 0.846, 0.932, 0.805 vs about 0.95Mixed
use higher dimension for Gaussian mixturesRG constrained weight · MoG simulationsweight on lower boundMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy