HAL : inria-00435807, version 2

arXiv : 0912.0171

Voir la fiche détaillée

BibTeX,EndNote,... BibTeX - EndNote - TEI - RefWorks -

Versions disponibles

v1 (01-12-2009)

v2 (14-12-2009)

Under-determined reverberant audio source separation using a full-rank spatial covariance model

Ngoc Q. K. Duong 1, Emmanuel Vincent 1, Remi Gribonval 1

(14/12/2009)

This article addresses the modeling of reverberant recording environments in the context of under-determined convolutive blind source separation. We model the contribution of each source to all mixture channels in the time-frequency domain as a zero-mean Gaussian random variable whose covariance encodes the spatial characteristics of the source. We then consider four specific covariance models, including a full-rank unconstrained model. We derive a family of iterative expectationmaximization (EM) algorithms to estimate the parameters of each model and propose suitable procedures to initialize the parameters and to align the order of the estimated sources across all frequency bins based on their estimated directions of arrival (DOA). Experimental results over reverberant synthetic mixtures and live recordings of speech data show the effectiveness of the proposed approach.

1 :	METISS (INRIA - IRISA)
	CNRS : UMR6074 – INRIA – INSA Rennes – Université de Rennes 1

Équipe de recherche (sauf Inria) : METISS

Domaine

:

Statistiques/Machine Learning Informatique/Traitement du signal et de l'image Sciences de l'ingénieur/Traitement du signal et de l'image

Mots-clés : Convolutive blind source separation – underdetermined mixtures – spatial covariance models – EM algorithm – permutation problem.

inria-00435807, version 2

http://hal.inria.fr/inria-00435807

oai:hal.inria.fr:inria-00435807

Contributeur : Ngoc Duong <>

Soumis le : Lundi 14 Décembre 2009, 13:34:46

Dernière modification le : Lundi 14 Décembre 2009, 14:14:30