Frequency domain blind source separation for many speech signals

Ryo Mukai; Hiroshi Sawada; Shoko Araki; Shoji Makino

doi:10.1007/978-3-540-30110-3_59

Frequency domain blind source separation for many speech signals

Ryo Mukai^*, Hiroshi Sawada, Shoko Araki, Shoji Makino

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Chapter

2 Citations (Scopus)

Abstract

This paper presents a method for solving the permutation problem of frequency domain blind source separation (BSS) when the number of source signals is large, and the potential source locations are omnidirectional. We propose a combination of small and large spacing sensor pairs with various axis directions in order to obtain proper geometric information for solving the permutation problem. Experimental results in a room (reverberation time T_R=130 ms) with eight microphones show that the proposed method can separate a mixture of six speech signals that come from various directions, even when two of them come from the same direction.

Original language	English
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Editors	Carlos G. Puntonet, Alberto Prieto
Publisher	Springer Verlag
Pages	461-469
Number of pages	9
ISBN (Electronic)	3540230564, 9783540230564
DOIs	https://doi.org/10.1007/978-3-540-30110-3_59
Publication status	Published - 2004
Externally published	Yes

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3195
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-540-30110-3_59

Cite this

Mukai, R., Sawada, H., Araki, S., & Makino, S. (2004). Frequency domain blind source separation for many speech signals. In C. G. Puntonet, & A. Prieto (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 461-469). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3195). Springer Verlag. https://doi.org/10.1007/978-3-540-30110-3_59

Frequency domain blind source separation for many speech signals. / Mukai, Ryo; Sawada, Hiroshi; Araki, Shoko et al.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). ed. / Carlos G. Puntonet; Alberto Prieto. Springer Verlag, 2004. p. 461-469 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3195).

Research output: Chapter in Book/Report/Conference proceeding › Chapter

Mukai, R, Sawada, H, Araki, S & Makino, S 2004, Frequency domain blind source separation for many speech signals. in CG Puntonet & A Prieto (eds), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3195, Springer Verlag, pp. 461-469. https://doi.org/10.1007/978-3-540-30110-3_59

Mukai R, Sawada H, Araki S, Makino S. Frequency domain blind source separation for many speech signals. In Puntonet CG, Prieto A, editors, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Springer Verlag. 2004. p. 461-469. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-540-30110-3_59

Mukai, Ryo ; Sawada, Hiroshi ; Araki, Shoko et al. / Frequency domain blind source separation for many speech signals. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). editor / Carlos G. Puntonet ; Alberto Prieto. Springer Verlag, 2004. pp. 461-469 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inbook{29c8a1a356a34ef5b37cf0402689ac6e,

title = "Frequency domain blind source separation for many speech signals",

abstract = "This paper presents a method for solving the permutation problem of frequency domain blind source separation (BSS) when the number of source signals is large, and the potential source locations are omnidirectional. We propose a combination of small and large spacing sensor pairs with various axis directions in order to obtain proper geometric information for solving the permutation problem. Experimental results in a room (reverberation time TR=130 ms) with eight microphones show that the proposed method can separate a mixture of six speech signals that come from various directions, even when two of them come from the same direction.",

author = "Ryo Mukai and Hiroshi Sawada and Shoko Araki and Shoji Makino",

year = "2004",

doi = "10.1007/978-3-540-30110-3_59",

language = "English",

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

publisher = "Springer Verlag",

pages = "461--469",

editor = "Puntonet, {Carlos G.} and Alberto Prieto",

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

}

TY - CHAP

T1 - Frequency domain blind source separation for many speech signals

AU - Mukai, Ryo

AU - Sawada, Hiroshi

AU - Araki, Shoko

AU - Makino, Shoji

PY - 2004

Y1 - 2004

N2 - This paper presents a method for solving the permutation problem of frequency domain blind source separation (BSS) when the number of source signals is large, and the potential source locations are omnidirectional. We propose a combination of small and large spacing sensor pairs with various axis directions in order to obtain proper geometric information for solving the permutation problem. Experimental results in a room (reverberation time TR=130 ms) with eight microphones show that the proposed method can separate a mixture of six speech signals that come from various directions, even when two of them come from the same direction.

AB - This paper presents a method for solving the permutation problem of frequency domain blind source separation (BSS) when the number of source signals is large, and the potential source locations are omnidirectional. We propose a combination of small and large spacing sensor pairs with various axis directions in order to obtain proper geometric information for solving the permutation problem. Experimental results in a room (reverberation time TR=130 ms) with eight microphones show that the proposed method can separate a mixture of six speech signals that come from various directions, even when two of them come from the same direction.

UR - http://www.scopus.com/inward/record.url?scp=35048903601&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=35048903601&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-30110-3_59

DO - 10.1007/978-3-540-30110-3_59

M3 - Chapter

AN - SCOPUS:35048903601

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

SP - 461

EP - 469

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

A2 - Puntonet, Carlos G.

A2 - Prieto, Alberto

PB - Springer Verlag

ER -

Frequency domain blind source separation for many speech signals

Abstract

Publication series

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this