Matching Pursuits with Random Sequential Subdictionaries

Abstract :

Matching pursuits are a class of greedy algorithms commonly used in signal processing, for solving the sparse approximation problem. They rely on an atom selection step that requires the calculation of numerous projections, which can be computationally costly for large dictionaries and burdens their competitiveness in coding applications. We propose using a non adaptive random sequence of subdictionaries in the decomposition process, thus parsing a large dictionary in a probabilistic fashion with no additional projection cost nor parameter estimation. A theoretical modeling based on order statistics is provided, along with experimental evidence showing that the novel algorithm can be efficiently used on sparse approximation problems. An application to audio signal compression with multiscale time-frequency dictionaries is presented, along with a discussion of the complexity and practical implementations.

Liste complète des métadonnées

Littérature citée [43 références]  Voir  Masquer  Télécharger

https://hal-imt.archives-ouvertes.fr/hal-00696181
Contributeur : Admin Télécom Paristech <>
Soumis le : vendredi 11 mai 2012 - 10:42:33
Dernière modification le : jeudi 11 janvier 2018 - 06:23:39
Document(s) archivé(s) le : dimanche 12 août 2012 - 02:22:02

Fichier

Moussallam2012_RSSMP.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Manuel Moussallam, L. Daudet, G. Richard. Matching Pursuits with Random Sequential Subdictionaries. Signal Processing, Elsevier, 2012, pp.2532-2544. 〈10.1016/j.sigpro.2012.03.019〉. 〈hal-00696181〉

Partager

Métriques

Consultations de la notice

269

Téléchargements de fichiers

560