Boltzmann Machines and Denoising Autoencoders for Image Denoising

TL;DR

This paper compares Boltzmann machines and denoising autoencoders for image denoising, showing that Boltzmann machines can perform better especially under high noise levels, with performance improved by increasing model depth.

Contribution

It demonstrates that Boltzmann machines are competitive with autoencoders for image denoising and explores the impact of model depth on denoising performance under various noise conditions.

Findings

Boltzmann machines can outperform autoencoders in high noise scenarios.

Adding hidden layers improves denoising performance.

Performance increases with model depth, especially at high noise levels.

Abstract

Image denoising based on a probabilistic model of local image patches has been employed by various researchers, and recently a deep (denoising) autoencoder has been proposed by Burger et al. [2012] and Xie et al. [2012] as a good model for this. In this paper, we propose that another popular family of models in the field of deep learning, called Boltzmann machines, can perform image denoising as well as, or in certain cases of high level of noise, better than denoising autoencoders. We empirically evaluate the two models on three different sets of images with different types and levels of noise. Throughout the experiments we also examine the effect of the depth of the models. The experiments confirmed our claim and revealed that the performance can be improved by adding more hidden layers, especially when the level of noise is high.