Skip to Main content Skip to Navigation
Conference papers

Construction d'estimateurs oracles pour la séparation de sources

Emmanuel Vincent 1 Rémi Gribonval 2
2 METISS - Speech and sound data modeling and processing
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
Abstract : Source separation of under-determined and/or convolutive mixtures is a difficult problem that has been addressed by many algorithms. In order to study their performance, we define oracle estimators that compute the maximal theoretical performance achievable for various classes of algorithms in an evaluation framework where the reference sources are available. We implement these estimators for two classes (stationary filtering separation algorithms and time-frequency masking separation algorithms) and we study their performance on a few audio mixture examples.
Complete list of metadatas

Cited literature [12 references]  Display  Hide  Download

https://hal.inria.fr/inria-00564757
Contributor : Rémi Gribonval <>
Submitted on : Wednesday, February 9, 2011 - 9:41:30 PM
Last modification on : Friday, July 10, 2020 - 4:23:10 PM
Long-term archiving on: : Tuesday, May 10, 2011 - 3:30:26 AM

File

gretsi05_vincent_gribonval.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00564757, version 1

Citation

Emmanuel Vincent, Rémi Gribonval. Construction d'estimateurs oracles pour la séparation de sources. XXe colloque GRETSI (traitement du signal et des images), 6-9 septembre 2005, Sep 2005, Louvain-la-Neuve, Belgique. ⟨inria-00564757⟩

Share

Metrics

Record views

564

Files downloads

181