Proposals for Performance Measurement in Source Separation

Rémi Gribonval 1 Laurent Benaroya 1 Emmanuel Vincent 2 Cédric Févotte 3
1 METISS - Speech and sound data modeling and processing
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
2 IRCAM - Institut de Recherche et Coordination Acoustique/Musique
3 IRCCyN - Institut de Recherche en Communications et en Cybernétique de Nantes
Abstract : In this paper, we address a few issues related to the evaluation of the performance of source separation algorithms. We propose several measures of distortion that take into account the gain indeterminacies of BSS algorithms. The total distortion includes interference from the other sources as well as noise and algorithmic artifacts, and we define performance criteria that measure separately these contributions. The criteria are valid even in the case of correlated sources. When the sources are estimated from a degenerate set of mixtures by applying a demixing matrix, we prove that there are upper bounds on the achievable Source to Interference Ratio. We propose these bounds as benchmarks to assess how well a (linear or nonlinear) BSS algorithm performs on a set of degenerate mixtures. We demonstrate on an example how to use these figures of merit to evaluate and compare the performance of BSS algorithms.
Type de document :
Communication dans un congrès
Cichocki, Andrzej and Murata, Noboru. 4th Int. Symp. on Independent Component Anal. and Blind Signal Separation (ICA2003), Apr 2003, Nara, Japan. pp.763--768, 2003, Proc. 4th Int. Symp. on Independent Component Anal. and Blind Signal Separation (ICA2003). 〈http://www.kecl.ntt.co.jp/icl/signal/ica2003/cdrom/data/0014.pdf〉
Liste complète des métadonnées

Littérature citée [12 références]  Voir  Masquer  Télécharger
https://hal.inria.fr/inria-00570123
Contributeur : Rémi Gribonval <>
Soumis le : samedi 26 février 2011 - 21:28:32
Dernière modification le : mercredi 19 décembre 2018 - 15:02:04
Document(s) archivé(s) le : vendredi 27 mai 2011 - 02:37:41

Fichier

0014.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : inria-00570123, version 1

Collections

INRIA | IRCAM | UNIV-NANTES | IRCCYN-ADTSI | IRCCYN | IRISA | UNAM | IRISA-D5 | UR1-UFR-ISTIC | UNIV-RENNES

Citation

Rémi Gribonval, Laurent Benaroya, Emmanuel Vincent, Cédric Févotte. Proposals for Performance Measurement in Source Separation. Cichocki, Andrzej and Murata, Noboru. 4th Int. Symp. on Independent Component Anal. and Blind Signal Separation (ICA2003), Apr 2003, Nara, Japan. pp.763--768, 2003, Proc. 4th Int. Symp. on Independent Component Anal. and Blind Signal Separation (ICA2003). 〈http://www.kecl.ntt.co.jp/icl/signal/ica2003/cdrom/data/0014.pdf〉. 〈inria-00570123〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

872

Téléchargements de fichiers

640

Contact


archive-ouverte@inria.fr