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

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

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.

Original language	English
Title of host publication	Latent Variable Analysis and Signal Separation - 13th International Conference, LVA/ICA 2017, Proceedings
Editors	Petr Tichavsky, Massoud Babaie-Zadeh, Olivier J.J. Michel, Nadege Thirion-Moreau
Publisher	Springer Verlag
Pages	225-235
Number of pages	11
ISBN (Print)	9783319535463
DOIs	https://doi.org/10.1007/978-3-319-53547-0_22
Publication status	Published - 1 Jan 2017
Externally published	Yes
Event	13th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2017 - Grenoble, France Duration: 21 Feb 2017 → 23 Feb 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10169 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	13th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2017
Country/Territory	France
City	Grenoble
Period	21/02/17 → 23/02/17

Keywords

Audio source separation
Filters
Latent mixing
MCMC
MMSE estimator
Sampling
Source image

Access to Document

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

Cite this

Girin, L., & Badeau, R. (2017). On the use of latent mixing filters in audio source separation. In 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 (pp. 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. editor / Petr Tichavsky ; Massoud Babaie-Zadeh ; Olivier J.J. Michel ; Nadege Thirion-Moreau. Springer Verlag, 2017. pp. 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. in 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, pp. 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).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

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. In Tichavsky P, Babaie-Zadeh M, Michel OJJ, Thirion-Moreau N, editors, 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

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this