On the use of latent mixing filters in audio source separation

Laurent Girin; Roland Badeau

doi:10.1007/978-3-319-53547-0_22

On the use of latent mixing filters in audio source separation

Laurent Girin
, Roland Badeau

LTHE (UMR 5564 CNRS/IRD/Université de Grenoble)
Université Paris-Saclay

Résultats de recherche: Le chapitre dans un livre, un rapport, une anthologie ou une collection › Contribution à une conférence › Revue par des pairs

3 Citations (Scopus)

Résumé

In this paper, we consider the underdetermined convolutive audio source separation (UCASS) problem. In the STFT domain, we consider both source signals and mixing filters as latent random variables, and we propose to estimate each source image, i.e. each individual source-filter product, by its posterior mean. Although, this is a quite straightforward application of the Bayesian estimation theory, to our knowledge, there exist no similar study in the UCASS context. In this paper, we discuss the interest of this estimator in this context and compare it with the conventional Wiener filter in a semi-oracle configuration.

langue originale	Anglais
titre	Latent Variable Analysis and Signal Separation - 13th International Conference, LVA/ICA 2017, Proceedings
rédacteurs en chef	Petr Tichavsky, Massoud Babaie-Zadeh, Olivier J.J. Michel, Nadege Thirion-Moreau
Editeur	Springer Verlag
Pages	225-235
Nombre de pages	11
ISBN (imprimé)	9783319535463
Les DOIs	https://doi.org/10.1007/978-3-319-53547-0_22
état	Publié - 1 janv. 2017
Modification externe	Oui
Evénement	13th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2017 - Grenoble, France Durée: 21 févr. 2017 → 23 févr. 2017

Série de publications

Nom	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10169 LNCS
ISSN (imprimé)	0302-9743
ISSN (Electronique)	1611-3349

Une conférence

Une conférence	13th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2017
Pays/Territoire	France
La ville	Grenoble
période	21/02/17 → 23/02/17

Accès au document

10.1007/978-3-319-53547-0_22

Contient cette citation

Girin, L., & Badeau, R. (2017). On the use of latent mixing filters in audio source separation. Dans P. Tichavsky, M. Babaie-Zadeh, O. J. J. Michel, & N. Thirion-Moreau (eds.), Latent Variable Analysis and Signal Separation - 13th International Conference, LVA/ICA 2017, Proceedings (p. 225-235). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol 10169 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-53547-0_22

Girin, Laurent ; Badeau, Roland. / On the use of latent mixing filters in audio source separation. Latent Variable Analysis and Signal Separation - 13th International Conference, LVA/ICA 2017, Proceedings. Editeur / Petr Tichavsky ; Massoud Babaie-Zadeh ; Olivier J.J. Michel ; Nadege Thirion-Moreau. Springer Verlag, 2017. p. 225-235 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{daec373eaeef4750ab1c1a738f53feb3,

title = "On the use of latent mixing filters in audio source separation",

abstract = "In this paper, we consider the underdetermined convolutive audio source separation (UCASS) problem. In the STFT domain, we consider both source signals and mixing filters as latent random variables, and we propose to estimate each source image, i.e. each individual source-filter product, by its posterior mean. Although, this is a quite straightforward application of the Bayesian estimation theory, to our knowledge, there exist no similar study in the UCASS context. In this paper, we discuss the interest of this estimator in this context and compare it with the conventional Wiener filter in a semi-oracle configuration.",

keywords = "Audio source separation, Filters, Latent mixing, MCMC, MMSE estimator, Sampling, Source image",

author = "Laurent Girin and Roland Badeau",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 13th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2017 ; Conference date: 21-02-2017 Through 23-02-2017",

year = "2017",

month = jan,

day = "1",

doi = "10.1007/978-3-319-53547-0\_22",

language = "English",

isbn = "9783319535463",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "225--235",

editor = "Petr Tichavsky and Massoud Babaie-Zadeh and Michel, \{Olivier J.J.\} and Nadege Thirion-Moreau",

booktitle = "Latent Variable Analysis and Signal Separation - 13th International Conference, LVA/ICA 2017, Proceedings",

}

Girin, L & Badeau, R 2017, On the use of latent mixing filters in audio source separation. Dans P Tichavsky, M Babaie-Zadeh, OJJ Michel & N Thirion-Moreau (eds), Latent Variable Analysis and Signal Separation - 13th International Conference, LVA/ICA 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), VOL. 10169 LNCS, Springer Verlag, p. 225-235, 13th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2017, Grenoble, France, 21/02/17. https://doi.org/10.1007/978-3-319-53547-0_22

On the use of latent mixing filters in audio source separation. / Girin, Laurent; Badeau, Roland.
Latent Variable Analysis and Signal Separation - 13th International Conference, LVA/ICA 2017, Proceedings. Ed. / Petr Tichavsky; Massoud Babaie-Zadeh; Olivier J.J. Michel; Nadege Thirion-Moreau. Springer Verlag, 2017. p. 225-235 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol 10169 LNCS).

Résultats de recherche: Le chapitre dans un livre, un rapport, une anthologie ou une collection › Contribution à une conférence › Revue par des pairs

TY - GEN

T1 - On the use of latent mixing filters in audio source separation

AU - Girin, Laurent

AU - Badeau, Roland

N1 - Publisher Copyright: © Springer International Publishing AG 2017.

PY - 2017/1/1

Y1 - 2017/1/1

N2 - In this paper, we consider the underdetermined convolutive audio source separation (UCASS) problem. In the STFT domain, we consider both source signals and mixing filters as latent random variables, and we propose to estimate each source image, i.e. each individual source-filter product, by its posterior mean. Although, this is a quite straightforward application of the Bayesian estimation theory, to our knowledge, there exist no similar study in the UCASS context. In this paper, we discuss the interest of this estimator in this context and compare it with the conventional Wiener filter in a semi-oracle configuration.

AB - In this paper, we consider the underdetermined convolutive audio source separation (UCASS) problem. In the STFT domain, we consider both source signals and mixing filters as latent random variables, and we propose to estimate each source image, i.e. each individual source-filter product, by its posterior mean. Although, this is a quite straightforward application of the Bayesian estimation theory, to our knowledge, there exist no similar study in the UCASS context. In this paper, we discuss the interest of this estimator in this context and compare it with the conventional Wiener filter in a semi-oracle configuration.

KW - Audio source separation

KW - Filters

KW - Latent mixing

KW - MCMC

KW - MMSE estimator

KW - Sampling

KW - Source image

U2 - 10.1007/978-3-319-53547-0_22

DO - 10.1007/978-3-319-53547-0_22

M3 - Conference contribution

AN - SCOPUS:85013484208

SN - 9783319535463

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 225

EP - 235

BT - Latent Variable Analysis and Signal Separation - 13th International Conference, LVA/ICA 2017, Proceedings

A2 - Tichavsky, Petr

A2 - Babaie-Zadeh, Massoud

A2 - Michel, Olivier J.J.

A2 - Thirion-Moreau, Nadege

PB - Springer Verlag

T2 - 13th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2017

Y2 - 21 February 2017 through 23 February 2017

ER -

Girin L, Badeau R. On the use of latent mixing filters in audio source separation. Dans Tichavsky P, Babaie-Zadeh M, Michel OJJ, Thirion-Moreau N, rédacteurs en chef, Latent Variable Analysis and Signal Separation - 13th International Conference, LVA/ICA 2017, Proceedings. Springer Verlag. 2017. p. 225-235. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-53547-0_22

On the use of latent mixing filters in audio source separation

Résumé

Série de publications

Une conférence

Accès au document

Empreinte digitale

Contient cette citation