Batch-less stochastic gradient descent for compressive learning of deep regularization for image denoising

TL;DR

This paper introduces two stochastic gradient descent variants for efficiently learning deep regularizers from heavily compressed databases, significantly improving patch-based image denoising performance.

Contribution

It develops novel SGD algorithms tailored for compressive learning of deep regularizers, enabling scalable and effective image denoising from large datasets.

Findings

Algorithms outperform previous low-dimensional methods

Methods benefit from classical SGD convergence guarantees

Effective for patch-based image denoising

Abstract

We consider the problem of denoising with the help of prior information taken from a database of clean signals or images. Denoising with variational methods is very efficient if a regularizer well adapted to the nature of the data is available. Thanks to the maximum a posteriori Bayesian framework, such regularizer can be systematically linked with the distribution of the data. With deep neural networks (DNN), complex distributions can be recovered from a large training database.To reduce the computational burden of this task, we adapt the compressive learning framework to the learning of regularizers parametrized by DNN. We propose two variants of stochastic gradient descent (SGD) for the recovery of deep regularization parameters from a heavily compressed database. These algorithms outperform the initially proposed method that was limited to low-dimensional signals, each iteration using information from the whole database. They also benefit from classical SGD convergence guarantees. Thanks to these improvements we show that this method can be applied for patch based image denoising.}