Exploiting Restricted Boltzmann Machines and Deep Belief Networks in Compressed Sensing

Abstract

This paper proposes a CS scheme that exploits the representational power of restricted Boltzmann machines and deep learning architectures to model the prior distribution of the sparsity pattern of signals belonging to the same class. The determined probability distribution is then used in a maximum a posteriori (MAP) approach for the reconstruction. The parameters of the prior distribution are learned from training data. The motivation behind this approach is to model the higher-order statistical dependencies between the coefficients of the sparse representation, with the final goal of improving the reconstruction. The performance of the proposed method is validated on the Berkeley Segmentation Dataset and the MNIST Database of handwritten digits.

arXiv commentAccepted for publication at IEEE Transactions on Signal Processing

References in corpus (9)

• Compressed Sensing of Block-Sparse Signals: Uncertainty Relations and Efficient Recovery
• Structured Compressed Sensing: From Theory to Applications
• Task-Driven Dictionary Learning
• Compressed Sensing and Redundant Dictionaries
• Exploiting Statistical Dependencies in Sparse Representations for Signal Recovery
• Cramer Rao-Type Bounds for Sparse Bayesian Learning
• Variational Bayesian algorithm for quantized compressed sensing
• Approximate Message Passing with Restricted Boltzmann Machine Priors
• Sparse approximation based on a random overcomplete basis

Cited by in corpus (1)

• Deep Learning Techniques for Compressive Sensing-Based Reconstruction and Inference -- A Ubiquitous Systems Perspective