A Geometrical Study of Matching Pursuit Parametrization

Abstract

This paper studies the effect of discretizing the parametrization of a dictionary used for Matching Pursuit decompositions of signals. Our approach relies on viewing the continuously parametrized dictionary as an embedded manifold in the signal space on which the tools of differential (Riemannian) geometry can be applied. The main contribution of this paper is twofold. First, we prove that if a discrete dictionary reaches a minimal density criterion, then the corresponding discrete MP (dMP) is equivalent in terms of convergence to a weakened hypothetical continuous MP. Interestingly, the corresponding weakness factor depends on a density measure of the discrete dictionary. Second, we show that the insertion of a simple geometric gradient ascent optimization on the atom dMP selection maintains the previous comparison but with a weakness factor at least two times closer to unity than without optimization. Finally, we present numerical experiments confirming our theoretical predictions for decomposition of signals and images on regular discretizations of dictionary parametrizations.

26 pages, 8 figures

References in corpus (1)

• Compressed Sensing and Redundant Dictionaries

Cited by in corpus (6)

• Newtonized Orthogonal Matching Pursuit: Frequency Estimation over the Continuum
• The Sliding Frank-Wolfe Algorithm and its Application to Super-Resolution Microscopy
• A Panorama on Multiscale Geometric Representations, Intertwining Spatial, Directional and Frequency Selectivity
• Compressive Parameter Estimation for Sparse Translation-Invariant Signals Using Polar Interpolation
• Image registration with sparse approximations in parametric dictionaries
• Factorization over interpolation: A fast continuous orthogonal matching pursuit