1: \begin{abstract}

2: One key ingredient of image restoration is to define a realistic prior on clean images to complete the missing information in the observation.

3: State-of-the-art restoration methods rely on a neural network to encode this prior. 

4: Moreover, typical image distributions are invariant to some set of transformations, such as rotations or flips.

5: However, most deep architectures are not designed to represent an invariant image distribution.

6: Recent works have proposed to overcome this difficulty by including equivariance properties within a Plug-and-Play paradigm.

7: 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.

8: 

9: \keywords{Image restoration, optimization, equivariance, plug-and-play.}

10: \end{abstract}

11: