Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Transport optimal régularisé semi-déséquilibré pour la restauration d'images

Abstract : In this paper we consider the use of a penalty based on optimal transport in order to regularize inverse problems in imaging. The proposed approach is formulated in a variational setting and aims at promoting images whose patch distribution is close either to the one learned by a generative model, or to available uncorrupted patches. With the aid of numerical illustrations, we argue in favor of adopting an asymmetric form of unbalanced transport. We then provide details concerning the computation and the differentiation of the proposed penalty. Finally, we detail the application of our approach to a particular super-resolution setting: the image zoom completion problem.
Document type :
Preprints, Working Papers, ...
Complete list of metadata

https://hal.archives-ouvertes.fr/hal-03623275
Contributor : Cécile Louchet Connect in order to contact the contributor
Submitted on : Tuesday, March 29, 2022 - 3:52:15 PM
Last modification on : Thursday, March 31, 2022 - 4:08:57 AM

File

GRETSI2022_2(4).pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-03623275, version 1

Citation

Simon Mignon, Bruno Galerne, Moncef Hidane, Cécile Louchet, Julien Mille. Transport optimal régularisé semi-déséquilibré pour la restauration d'images. 2022. ⟨hal-03623275⟩

Share

Metrics

Record views

31

Files downloads

12