Plug-and-Play image restoration with Stochastic deNOising REgularization

Marien Renaud, Jean Prost, Arthur Leclaire, Nicolas Papadakis
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:42484-42520, 2024.

Abstract

Plug-and-Play (PnP) algorithms are a class of iterative algorithms that address image inverse problems by combining a physical model and a deep neural network for regularization. Even if they produce impressive image restoration results, these algorithms rely on a non-standard use of a denoiser on images that are less and less noisy along the iterations, which contrasts with recent algorithms based on Diffusion Models (DM), where the denoiser is applied only on re-noised images. We propose a new PnP framework, called Stochastic deNOising REgularization (SNORE), which applies the denoiser only on images with noise of the adequate level. It is based on an explicit stochastic regularization, which leads to a stochastic gradient descent algorithm to solve ill-posed inverse problems. A convergence analysis of this algorithm and its annealing extension is provided. Experimentally, we prove that SNORE is competitive with respect to state-of-the-art methods on deblurring and inpainting tasks, both quantitatively and qualitatively.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-renaud24a, title = {Plug-and-Play image restoration with Stochastic de{NO}ising {RE}gularization}, author = {Renaud, Marien and Prost, Jean and Leclaire, Arthur and Papadakis, Nicolas}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {42484--42520}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/renaud24a/renaud24a.pdf}, url = {https://proceedings.mlr.press/v235/renaud24a.html}, abstract = {Plug-and-Play (PnP) algorithms are a class of iterative algorithms that address image inverse problems by combining a physical model and a deep neural network for regularization. Even if they produce impressive image restoration results, these algorithms rely on a non-standard use of a denoiser on images that are less and less noisy along the iterations, which contrasts with recent algorithms based on Diffusion Models (DM), where the denoiser is applied only on re-noised images. We propose a new PnP framework, called Stochastic deNOising REgularization (SNORE), which applies the denoiser only on images with noise of the adequate level. It is based on an explicit stochastic regularization, which leads to a stochastic gradient descent algorithm to solve ill-posed inverse problems. A convergence analysis of this algorithm and its annealing extension is provided. Experimentally, we prove that SNORE is competitive with respect to state-of-the-art methods on deblurring and inpainting tasks, both quantitatively and qualitatively.} }
Endnote
%0 Conference Paper %T Plug-and-Play image restoration with Stochastic deNOising REgularization %A Marien Renaud %A Jean Prost %A Arthur Leclaire %A Nicolas Papadakis %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-renaud24a %I PMLR %P 42484--42520 %U https://proceedings.mlr.press/v235/renaud24a.html %V 235 %X Plug-and-Play (PnP) algorithms are a class of iterative algorithms that address image inverse problems by combining a physical model and a deep neural network for regularization. Even if they produce impressive image restoration results, these algorithms rely on a non-standard use of a denoiser on images that are less and less noisy along the iterations, which contrasts with recent algorithms based on Diffusion Models (DM), where the denoiser is applied only on re-noised images. We propose a new PnP framework, called Stochastic deNOising REgularization (SNORE), which applies the denoiser only on images with noise of the adequate level. It is based on an explicit stochastic regularization, which leads to a stochastic gradient descent algorithm to solve ill-posed inverse problems. A convergence analysis of this algorithm and its annealing extension is provided. Experimentally, we prove that SNORE is competitive with respect to state-of-the-art methods on deblurring and inpainting tasks, both quantitatively and qualitatively.
APA
Renaud, M., Prost, J., Leclaire, A. & Papadakis, N.. (2024). Plug-and-Play image restoration with Stochastic deNOising REgularization. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:42484-42520 Available from https://proceedings.mlr.press/v235/renaud24a.html.

Related Material