Equivariant Denoisers for Image Restoration
Not on the map. Why?
Abstract (arXiv)
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.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 23 of its 32 references that are in the map. First posted on arXiv 6 Dec 2024.
Builds on
23 of 32 references.
References · 23
- 2020Plug-and-Play Image Restoration with Deep Denoiser Priorcited by 470
- 2016The Little Engine that Could: Regularization by Denoising (RED)cited by 307
- 2016Plug-and-Play ADMM for Image Restoration: Fixed Point Convergence and Applicationscited by 246
- 2014Understanding image representations by measuring their equivariance and equivalencecited by 243
- 2022Convergence of denoising diffusion models under the manifold hypothesiscited by 242
- 2019Plug-and-Play Methods Provably Converge with Properly Trained Denoiserscited by 198
- 2021Gradient Step Denoiser for convergent Plug-and-Playcited by 108
- 2021Equivariant Imaging: Learning Beyond the Range Spacecited by 83
- 2022Proximal Denoiser for Convergent Plug-and-Play Optimization with Nonconvex Regularizationcited by 80
- 2020Learning Maximally Monotone Operators for Image Recoverycited by 62
- 2019Block Coordinate Regularization by Denoisingcited by 33
- 2021Equivariant neural networks for inverse problemscited by 29
11 more
- 2023Equivariant plug-and-play image reconstructioncited by 26
- 2017Asymptotic Bias of Stochastic Gradient Searchcited by 19
- 2021As if by magic: self-supervised training of deep despeckling networks with MERLINcited by 18
- 2023Equivariant Bootstrapping for Uncertainty Quantification in Imaging Inverse Problemscited by 17
- 2023Rotation Equivariant Proximal Operator for Deep Unfolding Methods in Image Restorationcited by 17
- 2024Plug-and-Play image restoration with Stochastic deNOising REgularizationcited by 17
- 2024Stochastic Deep Restoration Priors for Imaging Inverse Problemscited by 14
- 2020A Fast Stochastic Plug-and-Play ADMM for Imaging Inverse Problemscited by 13
- 2023Normalization-Equivariant Neural Networks with Application to Image Denoisingcited by 11
- 2024Learning pseudo-contractive denoisers for inverse problemscited by 6
- 2024Empirical Bayesian image restoration by Langevin sampling with a denoising diffusion implicit priorcited by 6
Built on by
Citing papers per year: 2026: 1.
1 paper.