Online Sinkhorn: Optimal Transport distances from sample streams - Archive ouverte HAL Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2020

Online Sinkhorn: Optimal Transport distances from sample streams

Résumé

Optimal Transport (OT) distances are now routinely used as loss functions in ML tasks. Yet, computing OT distances between arbitrary (i.e. not necessarily discrete) probability distributions remains an open problem. This paper introduces a new online estimator of entropy-regularized OT distances between two such arbitrary distributions. It uses streams of samples from both distributions to iteratively enrich a non-parametric representation of the transportation plan. Compared to the classic Sinkhorn algorithm, our method leverages new samples at each iteration, which enables a consistent estimation of the true regularized OT distance. We provide a theoretical analysis of the convergence of the online Sinkhorn algorithm, showing a nearly-O(1/n) asymptotic sample complexity for the iterate sequence. We validate our method on synthetic 1D to 10D data and on real 3D shape data.
Fichier principal
Vignette du fichier
article.pdf (3.6 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-02495581 , version 1 (02-03-2020)
hal-02495581 , version 2 (24-06-2020)

Identifiants

Citer

Arthur Mensch, Gabriel Peyré. Online Sinkhorn: Optimal Transport distances from sample streams. 2020. ⟨hal-02495581v2⟩
74 Consultations
471 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More