Orthogonal Matching Pursuit with Replacement

TL;DR

This paper introduces OMPR, a novel sparse recovery algorithm that improves guarantees over traditional methods by allowing coordinate removal, and extends it with hashing for sub-linear performance, demonstrating robustness and speed in large-scale experiments.

Contribution

The paper proposes OMPR, a new greedy algorithm for compressed sensing with better theoretical guarantees and scalable variants using hashing, outperforming existing methods in large problems.

Findings

OMPR has the best known guarantees for sparse recovery under RIP.

OMPR-Hash achieves sub-linear complexity in dimensionality.

Proposed methods outperform existing algorithms in large-scale experiments.

Abstract

In this paper, we consider the problem of compressed sensing where the goal is to recover almost all the sparse vectors using a small number of fixed linear measurements. For this problem, we propose a novel partial hard-thresholding operator that leads to a general family of iterative algorithms. While one extreme of the family yields well known hard thresholding algorithms like ITI (Iterative Thresholding with Inversion) and HTP (Hard Thresholding Pursuit), the other end of the spectrum leads to a novel algorithm that we call Orthogonal Matching Pursuit with Replacement (OMPR). OMPR, like the classic greedy algorithm OMP, adds exactly one coordinate to the support at each iteration, based on the correlation with the current residual. However, unlike OMP, OMPR also removes one coordinate from the support. This simple change allows us to prove that OMPR has the best known guarantees for sparse recovery in terms of the Restricted Isometry Property (a condition on the measurement matrix). In contrast, OMP is known to have very weak performance guarantees under RIP. Given its simple structure, we are able to extend OMPR using locality sensitive hashing to get OMPR-Hash, the first provably sub-linear (in dimensionality) algorithm for sparse recovery. Our proof techniques are novel and flexible enough to also permit the tightest known analysis of popular iterative algorithms such as CoSaMP and Subspace Pursuit. We provide experimental results on large problems providing recovery for vectors of size up to million dimensions. We demonstrate that for large-scale problems our proposed methods are more robust and faster than existing methods.