A new method for kurtosis maximization and source separation

Abstract : This paper introduces a new method to maximize kurtosis-based contrast functions. Such contrast functions appear in the problem of blind source separation of convolutively mixed sources: the corresponding methods recover the sources one by one using a deflation approach. The proposed maximization algorithm is based on the particular nature of the criterion. The method is similar in spirit to a gradient ascent method, but differs in the fact that a "reference" contrast function is considered at each line search. The convergence of the method to a stationary point of the criterion can be proved. The theoretical result is illustrated by simulation
Type de document :
Communication dans un congrès
ICASSP 2010 : 35th International Conference on Acoustics, Speech, and Signal Processing , Mar 2010, Dallas, United States. IEEE, Proceedings ICASSP 2010 : 35th International Conference on Acoustics, Speech, and Signal Processing pp.2670 - 2673, 2010, 〈10.1109/ICASSP.2010.5496250 〉
Liste complète des métadonnées

Littérature citée [4 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01308032
Contributeur : Médiathèque Télécom Sudparis & Institut Mines-Télécom Business School <>
Soumis le : mercredi 27 avril 2016 - 10:40:21
Dernière modification le : jeudi 7 février 2019 - 16:18:35
Document(s) archivé(s) le : jeudi 28 juillet 2016 - 10:41:43

Fichier

icassp10.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Citation

Marc Castella, Eric Moreau. A new method for kurtosis maximization and source separation. ICASSP 2010 : 35th International Conference on Acoustics, Speech, and Signal Processing , Mar 2010, Dallas, United States. IEEE, Proceedings ICASSP 2010 : 35th International Conference on Acoustics, Speech, and Signal Processing pp.2670 - 2673, 2010, 〈10.1109/ICASSP.2010.5496250 〉. 〈hal-01308032〉

Partager

Métriques

Consultations de la notice

245

Téléchargements de fichiers

138