Compressive learning of deep regularization for denoising - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Compressive learning of deep regularization for denoising

Hui Shi
  • Fonction : Auteur
  • PersonId : 1089606
Yann Traonmilin
  • Fonction : Auteur
  • PersonId : 1077290
Jean François Aujol

Résumé

Solving ill-posed inverse problems can be done accurately if a regularizer well adapted to the nature of the data is available. Such regularizer can be systematically linked with the distribution of the data itself through the maximum a posteriori Bayesian framework. Recently, regularizers designed with the help of deep neural networks received impressive success. Such regularizers are typically learned from voluminous training data. To reduce the computational burden of this task, we propose to adapt the compressive learning framework to the learning of regularizers parametrized by deep neural networks (DNN). Our work shows the feasibility of batchless learning of regularizers from a compressed dataset. In order to achieve this, we propose an approximation of the compression operator that can be calculated explicitly for the task of learning a regularizer by DNN. We show that the proposed regularizer is capable of modeling complex regularity prior and can be used to solve the denoising inverse problem.
Fichier principal
Vignette du fichier
samplepaper.pdf (1.38 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03814336 , version 1 (13-10-2022)
hal-03814336 , version 2 (13-03-2023)

Identifiants

  • HAL Id : hal-03814336 , version 2

Citer

Hui Shi, Yann Traonmilin, Jean François Aujol. Compressive learning of deep regularization for denoising. International Conference on Scale Space and Variational Methods in Computer Vision (SSVM), May 2023, Cagliari, Italy. ⟨hal-03814336v2⟩

Collections

CNRS IMB ANR
120 Consultations
148 Téléchargements

Partager

Gmail Facebook X LinkedIn More