Equivariant Denoisers for Image Restoration

TL;DR

This paper introduces a unified framework called Equivariant Regularization by Denoising (ERED) that incorporates equivariant denoisers into image restoration, leveraging invariance properties of image distributions for improved results.

Contribution

The paper proposes ERED, a novel framework that integrates equivariant denoisers into image restoration, addressing the lack of invariance in traditional deep architectures.

Findings

ERED converges under certain conditions.

Equivariant denoisers improve restoration quality.

The framework effectively leverages invariance properties.

Abstract

One key ingredient of image restoration is to define a realistic prior on clean images to complete the missing information in the observation. State-of-the-art restoration methods rely on a neural network to encode this prior. Moreover, typical image distributions are invariant to some set of transformations, such as rotations or flips. However, most deep architectures are not designed to represent an invariant image distribution. Recent works have proposed to overcome this difficulty by including equivariance properties within a Plug-and-Play paradigm. In this work, we propose a unified framework named Equivariant Regularization by Denoising (ERED) based on equivariant denoisers and stochastic optimization. We analyze the convergence of this algorithm and discuss its practical benefit.