Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter

Ofer Schwartz; Sebastian Braun; Sharon Gannot; Emanuël A.P. Habets

doi:10.1007/978-3-319-53547-0_18

Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter

Ofer Schwartz, Sebastian Braun, Sharon Gannot, Emanuël A.P. Habets

Bar-Ilan University - The Alexander Kofkin Faculty of Engineering

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

2 Scopus citations

Abstract

The problem of source separation, dereverberation and noise reduction using a microphone array is addressed in this paper. The observed speech is modeled by two components, namely the early speech (including the direct path and some early reflections) and the late reverberation. The minimum mean square error (MMSE) estimator of the early speech components of the various speakers is derived, which jointly suppresses the noise and the overall reverberation from all speakers. The overall time-varying level of the reverberation is estimated using two different estimators, an estimator based on a temporal model and an estimator based on a spatial model. The experimental study consists of measured acoustic transfer functions (ATFs) and directional noise with various signal-to-noise ratio levels. The separation, dereverberation and noise reduction performance is examined in terms of perceptual evaluation of speech quality (PESQ) and signal-to-interference plus noise ratio improvement.

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	182-191
Number of pages	10
ISBN (Print)	9783319535463
DOIs	https://doi.org/10.1007/978-3-319-53547-0_18
State	Published - 2017
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

Bibliographical note

Publisher Copyright:
© Springer International Publishing AG 2017.

Access to Document

10.1007/978-3-319-53547-0_18

Cite this

Schwartz, O., Braun, S., Gannot, S., & Habets, E. A. P. (2017). Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter. 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. 182-191). (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_18

Schwartz, Ofer ; Braun, Sebastian ; Gannot, Sharon et al. / Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter. 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. 182-191 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{a56133ebfb2e4b60b38324cf1710e740,

title = "Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter",

abstract = "The problem of source separation, dereverberation and noise reduction using a microphone array is addressed in this paper. The observed speech is modeled by two components, namely the early speech (including the direct path and some early reflections) and the late reverberation. The minimum mean square error (MMSE) estimator of the early speech components of the various speakers is derived, which jointly suppresses the noise and the overall reverberation from all speakers. The overall time-varying level of the reverberation is estimated using two different estimators, an estimator based on a temporal model and an estimator based on a spatial model. The experimental study consists of measured acoustic transfer functions (ATFs) and directional noise with various signal-to-noise ratio levels. The separation, dereverberation and noise reduction performance is examined in terms of perceptual evaluation of speech quality (PESQ) and signal-to-interference plus noise ratio improvement.",

author = "Ofer Schwartz and Sebastian Braun and Sharon Gannot and Habets, {Emanu{\"e}l A.P.}",

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",

doi = "10.1007/978-3-319-53547-0_18",

language = "אנגלית",

isbn = "9783319535463",

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

publisher = "Springer Verlag",

pages = "182--191",

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",

address = "גרמניה",

}

Schwartz, O, Braun, S, Gannot, S & Habets, EAP 2017, Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter. 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. 182-191, 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_18

Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter. / Schwartz, Ofer; Braun, Sebastian; Gannot, Sharon et al.
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. 182-191 (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 - Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter

AU - Schwartz, Ofer

AU - Braun, Sebastian

AU - Gannot, Sharon

AU - Habets, Emanuël A.P.

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

PY - 2017

Y1 - 2017

N2 - The problem of source separation, dereverberation and noise reduction using a microphone array is addressed in this paper. The observed speech is modeled by two components, namely the early speech (including the direct path and some early reflections) and the late reverberation. The minimum mean square error (MMSE) estimator of the early speech components of the various speakers is derived, which jointly suppresses the noise and the overall reverberation from all speakers. The overall time-varying level of the reverberation is estimated using two different estimators, an estimator based on a temporal model and an estimator based on a spatial model. The experimental study consists of measured acoustic transfer functions (ATFs) and directional noise with various signal-to-noise ratio levels. The separation, dereverberation and noise reduction performance is examined in terms of perceptual evaluation of speech quality (PESQ) and signal-to-interference plus noise ratio improvement.

AB - The problem of source separation, dereverberation and noise reduction using a microphone array is addressed in this paper. The observed speech is modeled by two components, namely the early speech (including the direct path and some early reflections) and the late reverberation. The minimum mean square error (MMSE) estimator of the early speech components of the various speakers is derived, which jointly suppresses the noise and the overall reverberation from all speakers. The overall time-varying level of the reverberation is estimated using two different estimators, an estimator based on a temporal model and an estimator based on a spatial model. The experimental study consists of measured acoustic transfer functions (ATFs) and directional noise with various signal-to-noise ratio levels. The separation, dereverberation and noise reduction performance is examined in terms of perceptual evaluation of speech quality (PESQ) and signal-to-interference plus noise ratio improvement.

UR - http://www.scopus.com/inward/record.url?scp=85013407941&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-53547-0_18

DO - 10.1007/978-3-319-53547-0_18

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:85013407941

SN - 9783319535463

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

SP - 182

EP - 191

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 -

Schwartz O, Braun S, Gannot S, Habets EAP. Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter. 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. 182-191. (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_18

Source separation, dereverberation and noise reduction using LCMV beamformer and postfilter

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this