Multi-dimensional sparse structured signal approximation using split Bregman iterations

TL;DR

This paper introduces a multi-dimensional split Bregman method for sparse structured signal approximation, effectively preserving signal priors and demonstrating superior performance over existing methods in empirical evaluations.

Contribution

It presents a novel multi-dimensional split Bregman optimization approach for sparse structured signal approximation, addressing prior structure preservation.

Findings

Outperforms state-of-the-art methods in empirical tests

Effectively preserves prior structure of multi-dimensional signals

Demonstrates robustness across various signal features

Abstract

The paper focuses on the sparse approximation of signals using overcomplete representations, such that it preserves the (prior) structure of multi-dimensional signals. The underlying optimization problem is tackled using a multi-dimensional split Bregman optimization approach. An extensive empirical evaluation shows how the proposed approach compares to the state of the art depending on the signal features.