Provably Convergent Plug & Play Linearized ADMM, applied to Deblurring Spatially Varying Kernels

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Laroche, Charles, Almansa, Andrés, Coupeté, Eva, Tassano, Matias
Format: Preprint
Published: 2022
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929352744108032
author Laroche, Charles
Almansa, Andrés
Coupeté, Eva
Tassano, Matias
author_facet Laroche, Charles
Almansa, Andrés
Coupeté, Eva
Tassano, Matias
contents Plug & Play methods combine proximal algorithms with denoiser priors to solve inverse problems. These methods rely on the computability of the proximal operator of the data fidelity term. In this paper, we propose a Plug & Play framework based on linearized ADMM that allows us to bypass the computation of intractable proximal operators. We demonstrate the convergence of the algorithm and provide results on restoration tasks such as super-resolution and deblurring with non-uniform blur.
format Preprint
id arxiv_https___arxiv_org_abs_2210_10605
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Provably Convergent Plug & Play Linearized ADMM, applied to Deblurring Spatially Varying Kernels
Laroche, Charles
Almansa, Andrés
Coupeté, Eva
Tassano, Matias
Computer Vision and Pattern Recognition
Optimization and Control
Applications
Plug & Play methods combine proximal algorithms with denoiser priors to solve inverse problems. These methods rely on the computability of the proximal operator of the data fidelity term. In this paper, we propose a Plug & Play framework based on linearized ADMM that allows us to bypass the computation of intractable proximal operators. We demonstrate the convergence of the algorithm and provide results on restoration tasks such as super-resolution and deblurring with non-uniform blur.
title Provably Convergent Plug & Play Linearized ADMM, applied to Deblurring Spatially Varying Kernels
topic Computer Vision and Pattern Recognition
Optimization and Control
Applications
url https://arxiv.org/abs/2210.10605