On calculating the inverse of separation matrix in frequency-domain blind source separation

Hiroshi Sawada; Shoko Araki; Ryo Mukai; Shoji Makino

doi:10.1007/11679363_86

On calculating the inverse of separation matrix in frequency-domain blind source separation

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

^*Corresponding author for this work

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

3 Citations (Scopus)

Abstract

For blind source separation (BSS) of convolutive mixtures, the frequency-domain approach is efficient and practical, because the convolutive mixtures are modeled with instantaneous mixtures at each frequency bin and simple instantaneous independent component analysis (ICA) can be employed to separate the mixtures. However, the permutation and scaling ambiguities of ICA solutions need to be aligned to obtain proper time-domain separated signals. This paper discusses the idea that calculating the inverses of separation matrices obtained by ICA is very important as regards aligning these ambiguities. This paper also shows the relationship between the ICA-based method and the time-frequency masking method for BSS, which becomes clear by calculating the inverses.

Original language	English
Title of host publication	Independent Component Analysis and Blind Signal Separation - 6th International Conference, ICA 2006, Proceedings
Pages	691-699
Number of pages	9
DOIs	https://doi.org/10.1007/11679363_86
Publication status	Published - 2006
Externally published	Yes
Event	6th International Conference on Independent Component Analysis and Blind Signal Separation, ICA 2006 - Charleston, SC, United States Duration: 2006 Mar 5 → 2006 Mar 8

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3889 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	6th International Conference on Independent Component Analysis and Blind Signal Separation, ICA 2006
Country/Territory	United States
City	Charleston, SC
Period	06/3/5 → 06/3/8

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/11679363_86

Cite this

Sawada, H., Araki, S., Mukai, R., & Makino, S. (2006). On calculating the inverse of separation matrix in frequency-domain blind source separation. In Independent Component Analysis and Blind Signal Separation - 6th International Conference, ICA 2006, Proceedings (pp. 691-699). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3889 LNCS). https://doi.org/10.1007/11679363_86

On calculating the inverse of separation matrix in frequency-domain blind source separation. / Sawada, Hiroshi; Araki, Shoko; Mukai, Ryo et al.
Independent Component Analysis and Blind Signal Separation - 6th International Conference, ICA 2006, Proceedings. 2006. p. 691-699 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3889 LNCS).

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

Sawada, H, Araki, S, Mukai, R & Makino, S 2006, On calculating the inverse of separation matrix in frequency-domain blind source separation. in Independent Component Analysis and Blind Signal Separation - 6th International Conference, ICA 2006, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3889 LNCS, pp. 691-699, 6th International Conference on Independent Component Analysis and Blind Signal Separation, ICA 2006, Charleston, SC, United States, 06/3/5. https://doi.org/10.1007/11679363_86

Sawada H, Araki S, Mukai R, Makino S. On calculating the inverse of separation matrix in frequency-domain blind source separation. In Independent Component Analysis and Blind Signal Separation - 6th International Conference, ICA 2006, Proceedings. 2006. p. 691-699. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/11679363_86

Sawada, Hiroshi ; Araki, Shoko ; Mukai, Ryo et al. / On calculating the inverse of separation matrix in frequency-domain blind source separation. Independent Component Analysis and Blind Signal Separation - 6th International Conference, ICA 2006, Proceedings. 2006. pp. 691-699 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{5457b74964cb44399305e4ef2d1a6913,

title = "On calculating the inverse of separation matrix in frequency-domain blind source separation",

abstract = "For blind source separation (BSS) of convolutive mixtures, the frequency-domain approach is efficient and practical, because the convolutive mixtures are modeled with instantaneous mixtures at each frequency bin and simple instantaneous independent component analysis (ICA) can be employed to separate the mixtures. However, the permutation and scaling ambiguities of ICA solutions need to be aligned to obtain proper time-domain separated signals. This paper discusses the idea that calculating the inverses of separation matrices obtained by ICA is very important as regards aligning these ambiguities. This paper also shows the relationship between the ICA-based method and the time-frequency masking method for BSS, which becomes clear by calculating the inverses.",

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

year = "2006",

doi = "10.1007/11679363_86",

language = "English",

isbn = "3540326308",

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

pages = "691--699",

booktitle = "Independent Component Analysis and Blind Signal Separation - 6th International Conference, ICA 2006, Proceedings",

note = "6th International Conference on Independent Component Analysis and Blind Signal Separation, ICA 2006 ; Conference date: 05-03-2006 Through 08-03-2006",

}

TY - GEN

T1 - On calculating the inverse of separation matrix in frequency-domain blind source separation

AU - Sawada, Hiroshi

AU - Araki, Shoko

AU - Mukai, Ryo

AU - Makino, Shoji

PY - 2006

Y1 - 2006

N2 - For blind source separation (BSS) of convolutive mixtures, the frequency-domain approach is efficient and practical, because the convolutive mixtures are modeled with instantaneous mixtures at each frequency bin and simple instantaneous independent component analysis (ICA) can be employed to separate the mixtures. However, the permutation and scaling ambiguities of ICA solutions need to be aligned to obtain proper time-domain separated signals. This paper discusses the idea that calculating the inverses of separation matrices obtained by ICA is very important as regards aligning these ambiguities. This paper also shows the relationship between the ICA-based method and the time-frequency masking method for BSS, which becomes clear by calculating the inverses.

AB - For blind source separation (BSS) of convolutive mixtures, the frequency-domain approach is efficient and practical, because the convolutive mixtures are modeled with instantaneous mixtures at each frequency bin and simple instantaneous independent component analysis (ICA) can be employed to separate the mixtures. However, the permutation and scaling ambiguities of ICA solutions need to be aligned to obtain proper time-domain separated signals. This paper discusses the idea that calculating the inverses of separation matrices obtained by ICA is very important as regards aligning these ambiguities. This paper also shows the relationship between the ICA-based method and the time-frequency masking method for BSS, which becomes clear by calculating the inverses.

UR - http://www.scopus.com/inward/record.url?scp=33745725493&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33745725493&partnerID=8YFLogxK

U2 - 10.1007/11679363_86

DO - 10.1007/11679363_86

M3 - Conference contribution

AN - SCOPUS:33745725493

SN - 3540326308

SN - 9783540326304

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

SP - 691

EP - 699

BT - Independent Component Analysis and Blind Signal Separation - 6th International Conference, ICA 2006, Proceedings

T2 - 6th International Conference on Independent Component Analysis and Blind Signal Separation, ICA 2006

Y2 - 5 March 2006 through 8 March 2006

ER -

On calculating the inverse of separation matrix in frequency-domain blind source separation

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this