Batch-less stochastic gradient descent for compressive learning of deep regularization for image denoising - Archive ouverte HAL Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2023

Batch-less stochastic gradient descent for compressive learning of deep regularization for image denoising

Hui Shi
  • Fonction : Auteur
  • PersonId : 1089606
Yann Traonmilin
  • Fonction : Auteur
  • PersonId : 1077290
J-F Aujol

Résumé

We consider the problem of denoising with the help of prior information taken from a database of clean signals or images. Denoising with variational methods is very efficient if a regularizer well adapted to the nature of the data is available. Thanks to the maximum a posteriori Bayesian framework, such regularizer can be systematically linked with the distribution of the data. With deep neural networks (DNN), complex distributions can be recovered from a large training database. To reduce the computational burden of this task, we adapt the compressive learning framework to the learning of regularizers parametrized by DNN. We propose two variants of stochastic gradient descent (SGD) for the recovery of deep regularization parameters from a heavily compressed database. These algorithms outperform the initially proposed method that was limited to low-dimensional signals, each iteration using information from the whole database. They also benefit from classical SGD convergence guarantees. Thanks to these improvements we show that this method can be applied for patch based image denoising.}
Fichier principal
Vignette du fichier
deep_reg_sketch.pdf (2.53 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04222825 , version 1 (29-09-2023)

Identifiants

Citer

Hui Shi, Yann Traonmilin, J-F Aujol. Batch-less stochastic gradient descent for compressive learning of deep regularization for image denoising. 2023. ⟨hal-04222825⟩

Collections

CNRS IMB INSMI
49 Consultations
17 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More