Towards a unified view of unsupervised non-local methods for image denoising: the NL-Ridge approach - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Towards a unified view of unsupervised non-local methods for image denoising: the NL-Ridge approach

Résumé

We propose a unified view of unsupervised non-local methods for image denoising that linearly combine noisy image patches. The best methods, established in different modeling and estimation frameworks, are two-step algorithms. Leveraging Stein's unbiased risk estimate (SURE) for the first step and the "internal adaptation", a concept borrowed from deep learning theory, for the second one, we show that our NL-Ridge approach enables to reconcile several patch aggregation methods for image denoising. In the second step, our closed-form aggregation weights are computed through multivariate Ridge regressions. Experiments on artificially noisy images demonstrate that NL-Ridge may outperform well established state-of-the-art unsupervised denoisers such as BM3D and NL-Bayes, as well as recent unsupervised deep learning methods, while being simpler conceptually.
Fichier principal
Vignette du fichier
NL_Ridge_arxiv_template.pdf (977.42 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03926888 , version 1 (06-01-2023)

Identifiants

Citer

Sébastien Herbreteau, Charles Kervrann. Towards a unified view of unsupervised non-local methods for image denoising: the NL-Ridge approach. ICIP 2022 - 29th IEEE International Conference on Image Processing, Oct 2022, Bordeaux, France. pp.3376-3380, ⟨10.1109/ICIP46576.2022.9897992⟩. ⟨hal-03926888⟩
29 Consultations
37 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More