English

Factorization over interpolation: A fast continuous orthogonal matching pursuit

Signal Processing 2020-12-01 v2

Abstract

We propose a fast greedy algorithm to compute sparse representations of signals from continuous dictionaries that are factorizable, i.e., with atoms that can be separated as a product of sub-atoms. Existing algorithms strongly reduce the computational complexity of the sparse decomposition of signals in discrete factorizable dictionaries. On another flavour, existing greedy algorithms use interpolation strategies from a discretization of continuous dictionaries to perform off-the-grid decomposition. Our algorithm aims to combine the factorization and the interpolation concepts to enable low complexity computation of continuous sparse representation of signals. The efficiency of our algorithm is highlighted by simulations of its application to a radar system.

Keywords

Cite

@article{arxiv.2007.01060,
  title  = {Factorization over interpolation: A fast continuous orthogonal matching pursuit},
  author = {Gilles Monnoyer de Galland and Luc Vandendorpe and Laurent Jacques},
  journal= {arXiv preprint arXiv:2007.01060},
  year   = {2020}
}

Comments

in Proceedings of iTWIST'20, Paper-ID: 45, Nantes, France, December, 2-4, 2020

R2 v1 2026-06-23T16:47:56.436Z