A variational em algorithm for the separation of moving sound sources

DIonyssos Kounades-Bastian; Laurent Girin; Xavier Alameda-Pineda; Sharon Gannot; Radu Horaud

doi:10.1109/WASPAA.2015.7336936

A variational em algorithm for the separation of moving sound sources

DIonyssos Kounades-Bastian, Laurent Girin, Xavier Alameda-Pineda, Sharon Gannot, Radu Horaud

Bar-Ilan University - The Alexander Kofkin Faculty of Engineering

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

14 Scopus citations

Abstract

This paper addresses the problem of separation of moving sound sources. We propose a probabilistic framework based on the complex Gaussian model combined with non-negative matrix factorization. The properties associated with moving sources are modeled using time-varying mixing filters described by a stochastic temporal process. We present a variational expectation-maximization (VEM) algorithm that employs a Kalman smoother to estimate the mixing filters. The sound sources are separated by means of Wiener filters, built from the estimators provided by the proposed VEM algorithm. Preliminary experiments with simulated data show that, while for static sources we obtain results comparable with the baseline method [1], in the case of moving source our method outperforms a piece-wise version of the baseline method.

Original language	English
Title of host publication	2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781479974504
DOIs	https://doi.org/10.1109/WASPAA.2015.7336936
State	Published - 24 Nov 2015
Event	IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015 - New Paltz, United States Duration: 18 Oct 2015 → 21 Oct 2015

Publication series

Name	2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015

Conference

Conference	IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015
Country/Territory	United States
City	New Paltz
Period	18/10/15 → 21/10/15

Bibliographical note

Publisher Copyright:
© 2015 IEEE.

Keywords

Audio-source separation
Kalman smoother
moving sources
time-varying mixing filters
variational EM

Access to Document

10.1109/WASPAA.2015.7336936

Cite this

Kounades-Bastian, DI., Girin, L., Alameda-Pineda, X., Gannot, S., & Horaud, R. (2015). A variational em algorithm for the separation of moving sound sources. In 2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015 Article 7336936 (2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/WASPAA.2015.7336936

Kounades-Bastian, DIonyssos ; Girin, Laurent ; Alameda-Pineda, Xavier et al. / A variational em algorithm for the separation of moving sound sources. 2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015. Institute of Electrical and Electronics Engineers Inc., 2015. (2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015).

@inproceedings{14ded67eab1746f49bc1d34ea710757d,

title = "A variational em algorithm for the separation of moving sound sources",

abstract = "This paper addresses the problem of separation of moving sound sources. We propose a probabilistic framework based on the complex Gaussian model combined with non-negative matrix factorization. The properties associated with moving sources are modeled using time-varying mixing filters described by a stochastic temporal process. We present a variational expectation-maximization (VEM) algorithm that employs a Kalman smoother to estimate the mixing filters. The sound sources are separated by means of Wiener filters, built from the estimators provided by the proposed VEM algorithm. Preliminary experiments with simulated data show that, while for static sources we obtain results comparable with the baseline method [1], in the case of moving source our method outperforms a piece-wise version of the baseline method.",

keywords = "Audio-source separation, Kalman smoother, moving sources, time-varying mixing filters, variational EM",

author = "DIonyssos Kounades-Bastian and Laurent Girin and Xavier Alameda-Pineda and Sharon Gannot and Radu Horaud",

note = "Publisher Copyright: {\textcopyright} 2015 IEEE.; IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015 ; Conference date: 18-10-2015 Through 21-10-2015",

year = "2015",

month = nov,

day = "24",

doi = "10.1109/WASPAA.2015.7336936",

language = "אנגלית",

series = "2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015",

address = "ארצות הברית",

}

Kounades-Bastian, DI, Girin, L, Alameda-Pineda, X, Gannot, S & Horaud, R 2015, A variational em algorithm for the separation of moving sound sources. in 2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015., 7336936, 2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015, Institute of Electrical and Electronics Engineers Inc., IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015, New Paltz, United States, 18/10/15. https://doi.org/10.1109/WASPAA.2015.7336936

A variational em algorithm for the separation of moving sound sources. / Kounades-Bastian, DIonyssos; Girin, Laurent; Alameda-Pineda, Xavier et al.
2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015. Institute of Electrical and Electronics Engineers Inc., 2015. 7336936 (2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015).

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

TY - GEN

T1 - A variational em algorithm for the separation of moving sound sources

AU - Kounades-Bastian, DIonyssos

AU - Girin, Laurent

AU - Alameda-Pineda, Xavier

AU - Gannot, Sharon

AU - Horaud, Radu

PY - 2015/11/24

Y1 - 2015/11/24

N2 - This paper addresses the problem of separation of moving sound sources. We propose a probabilistic framework based on the complex Gaussian model combined with non-negative matrix factorization. The properties associated with moving sources are modeled using time-varying mixing filters described by a stochastic temporal process. We present a variational expectation-maximization (VEM) algorithm that employs a Kalman smoother to estimate the mixing filters. The sound sources are separated by means of Wiener filters, built from the estimators provided by the proposed VEM algorithm. Preliminary experiments with simulated data show that, while for static sources we obtain results comparable with the baseline method [1], in the case of moving source our method outperforms a piece-wise version of the baseline method.

AB - This paper addresses the problem of separation of moving sound sources. We propose a probabilistic framework based on the complex Gaussian model combined with non-negative matrix factorization. The properties associated with moving sources are modeled using time-varying mixing filters described by a stochastic temporal process. We present a variational expectation-maximization (VEM) algorithm that employs a Kalman smoother to estimate the mixing filters. The sound sources are separated by means of Wiener filters, built from the estimators provided by the proposed VEM algorithm. Preliminary experiments with simulated data show that, while for static sources we obtain results comparable with the baseline method [1], in the case of moving source our method outperforms a piece-wise version of the baseline method.

KW - Audio-source separation

KW - Kalman smoother

KW - moving sources

KW - time-varying mixing filters

KW - variational EM

UR - http://www.scopus.com/inward/record.url?scp=84960921455&partnerID=8YFLogxK

U2 - 10.1109/WASPAA.2015.7336936

DO - 10.1109/WASPAA.2015.7336936

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

AN - SCOPUS:84960921455

T3 - 2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015

BT - 2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015

Y2 - 18 October 2015 through 21 October 2015

ER -

Kounades-Bastian DI, Girin L, Alameda-Pineda X, Gannot S, Horaud R. A variational em algorithm for the separation of moving sound sources. In 2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015. Institute of Electrical and Electronics Engineers Inc. 2015. 7336936. (2015 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2015). doi: 10.1109/WASPAA.2015.7336936

A variational em algorithm for the separation of moving sound sources

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this