Under-determined source separation: comparison of two approaches based on sparse decompositions

Sylvain Lesage 1 Sacha Krstulovic 1 Rémi Gribonval 1
1 METISS - Speech and sound data modeling and processing
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
Abstract : This paper focuses on under-determined source separation when the mixing parameters are known. The approach is based on a sparse decomposition of the mixture. In the proposed method, the mixture is decomposed with Matching Pursuit by introducing a new class of multi-channel dictionaries, where the atoms are given by a spatial direction and a waveform. The knowledge of the mixing matrix is directly integrated in the decomposition. Compared to the separation by multi-channel Matching Pursuit followed by a clustering, the new algorithm introduces less artifacts whereas the level of residual interferences is about the same. These two methods are compared to Bofill & Zibulevsky's separation algorithm and DUET method. We also study the effect of smoothing the decompositions and the importance of the quality of the estimation of the mixing matrix.
Liste complète des métadonnées

Cited literature [12 references]  Display  Hide  Download

https://hal.inria.fr/inria-00544929
Contributor : Rémi Gribonval <>
Submitted on : Tuesday, February 8, 2011 - 10:10:42 PM
Last modification on : Friday, November 16, 2018 - 1:23:47 AM
Document(s) archivé(s) le : Monday, May 9, 2011 - 2:51:25 AM

File

2006_ICA_LesageEtAl.pdf
Files produced by the author(s)

Identifiers

Citation

Sylvain Lesage, Sacha Krstulovic, Rémi Gribonval. Under-determined source separation: comparison of two approaches based on sparse decompositions. Proc. of the Int'l. Workshop on Independent Component Analysis and Blind Signal Separation (ICA 2006), Mar 2006, Charleston, South Carolina, United States. pp.633--640, ⟨10.1007/11679363_79⟩. ⟨inria-00544929⟩

Share

Metrics

Record views

270

Files downloads

172