A lifted Bregman strategy for training unfolded proximal neural network Gaussian denoisers

TL;DR

This paper introduces a novel lifted Bregman-based training method for unfolded proximal neural networks, enhancing training efficiency and robustness in image denoising tasks by leveraging a specialized optimization strategy.

Contribution

It proposes a new Bregman distance-based training formulation and a bespoke computational strategy for unfolded PNNs, improving training efficiency over traditional back-propagation.

Findings

Enhanced training efficiency demonstrated in numerical simulations.

Improved robustness of PNNs in image denoising tasks.

Effective application of Bregman distances in neural network training.

Abstract

Unfolded proximal neural networks (PNNs) form a family of methods that combines deep learning and proximal optimization approaches. They consist in designing a neural network for a specific task by unrolling a proximal algorithm for a fixed number of iterations, where linearities can be learned from prior training procedure. PNNs have shown to be more robust than traditional deep learning approaches while reaching at least as good performances, in particular in computational imaging. However, training PNNs still depends on the efficiency of available training algorithms. In this work, we propose a lifted training formulation based on Bregman distances for unfolded PNNs. Leveraging the deterministic mini-batch block-coordinate forward-backward method, we design a bespoke computational strategy beyond traditional back-propagation methods for solving the resulting learning problem efficiently. We assess the behaviour of the proposed training approach for PNNs through numerical simulations on image denoising, considering a denoising PNN whose structure is based on dual proximal-gradient iterations.