Learning to Generate Wasserstein Barycenters - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Journal of Mathematical Imaging and Vision Année : 2022

Learning to Generate Wasserstein Barycenters

Julien Lacombe
Julie Digne
Nicolas Bonneel

Résumé

Optimal transport is a notoriously difficult problem to solve numerically, with current approaches often remaining intractable for very large scale applications such as those encountered in machine learning. Wasserstein barycenters -- the problem of finding measures in-between given input measures in the optimal transport sense -- is even more computationally demanding as it requires to solve an optimization problem involving optimal transport distances. By training a deep convolutional neural network, we improve by a factor of 60 the computational speed of Wasserstein barycenters over the fastest state-of-the-art approach on the GPU, resulting in milliseconds computational times on $512\times512$ regular grids. We show that our network, trained on Wasserstein barycenters of pairs of measures, generalizes well to the problem of finding Wasserstein barycenters of more than two measures. We demonstrate the efficiency of our approach for computing barycenters of sketches and transferring colors between multiple images.
Fichier principal
Vignette du fichier
2102.12178.pdf (20.86 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03374452 , version 1 (12-10-2021)

Identifiants

Citer

Julien Lacombe, Julie Digne, Nicolas Courty, Nicolas Bonneel. Learning to Generate Wasserstein Barycenters. Journal of Mathematical Imaging and Vision, 2022, 65, pp.354-370. ⟨10.1007/s10851-022-01121-y⟩. ⟨hal-03374452⟩
128 Consultations
56 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More