Frequency domain blind source separation of a reduced amount of data using frequency normalization

Enrique Robledo-Arnuncio; Hiroshi Sawada; Shoji Makino

Frequency domain blind source separation of a reduced amount of data using frequency normalization

Enrique Robledo-Arnuncio^*, Hiroshi Sawada, Shoji Makino

^*Corresponding author for this work

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

1 Citation (Scopus)

Abstract

The problem of blind source separation (BSS) from convolutive mixtures is often addressed using independent component analysis in the frequency domain. The separation performance with this approach degrades significantly when only a short amount of data is available, since the estimation of the separation system becomes inaccurate. In this paper we present a novel approach to the frequency domain BSS using frequency normalization. Under the conditions of almost sparse sources and of dominant direct path in the mixing systems, we show that the new approach provides better performance than the conventional one when the amount of available data is small.

Original language	English
Title of host publication	2006 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings
Pages	V837-V840
Publication status	Published - 2006
Externally published	Yes
Event	2006 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2006 - Toulouse, France Duration: 2006 May 14 → 2006 May 19

Publication series

Name	ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume	5
ISSN (Print)	1520-6149

Conference

Conference	2006 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2006
Country/Territory	France
City	Toulouse
Period	06/5/14 → 06/5/19

ASJC Scopus subject areas

Software
Signal Processing
Electrical and Electronic Engineering

Cite this

Robledo-Arnuncio, E., Sawada, H., & Makino, S. (2006). Frequency domain blind source separation of a reduced amount of data using frequency normalization. In 2006 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings (pp. V837-V840). [1661406] (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 5).

Frequency domain blind source separation of a reduced amount of data using frequency normalization. / Robledo-Arnuncio, Enrique; Sawada, Hiroshi; Makino, Shoji.
2006 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings. 2006. p. V837-V840 1661406 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 5).

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

Robledo-Arnuncio, E, Sawada, H & Makino, S 2006, Frequency domain blind source separation of a reduced amount of data using frequency normalization. in 2006 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings., 1661406, ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, vol. 5, pp. V837-V840, 2006 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2006, Toulouse, France, 06/5/14.

Robledo-Arnuncio, Enrique ; Sawada, Hiroshi ; Makino, Shoji. / Frequency domain blind source separation of a reduced amount of data using frequency normalization. 2006 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings. 2006. pp. V837-V840 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings).

@inproceedings{ac822652757c4d4facb0b833cf5fb9a9,

title = "Frequency domain blind source separation of a reduced amount of data using frequency normalization",

abstract = "The problem of blind source separation (BSS) from convolutive mixtures is often addressed using independent component analysis in the frequency domain. The separation performance with this approach degrades significantly when only a short amount of data is available, since the estimation of the separation system becomes inaccurate. In this paper we present a novel approach to the frequency domain BSS using frequency normalization. Under the conditions of almost sparse sources and of dominant direct path in the mixing systems, we show that the new approach provides better performance than the conventional one when the amount of available data is small.",

author = "Enrique Robledo-Arnuncio and Hiroshi Sawada and Shoji Makino",

year = "2006",

language = "English",

isbn = "142440469X",

series = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

pages = "V837--V840",

booktitle = "2006 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings",

note = "2006 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2006 ; Conference date: 14-05-2006 Through 19-05-2006",

}

TY - GEN

T1 - Frequency domain blind source separation of a reduced amount of data using frequency normalization

AU - Robledo-Arnuncio, Enrique

AU - Sawada, Hiroshi

AU - Makino, Shoji

PY - 2006

Y1 - 2006

N2 - The problem of blind source separation (BSS) from convolutive mixtures is often addressed using independent component analysis in the frequency domain. The separation performance with this approach degrades significantly when only a short amount of data is available, since the estimation of the separation system becomes inaccurate. In this paper we present a novel approach to the frequency domain BSS using frequency normalization. Under the conditions of almost sparse sources and of dominant direct path in the mixing systems, we show that the new approach provides better performance than the conventional one when the amount of available data is small.

AB - The problem of blind source separation (BSS) from convolutive mixtures is often addressed using independent component analysis in the frequency domain. The separation performance with this approach degrades significantly when only a short amount of data is available, since the estimation of the separation system becomes inaccurate. In this paper we present a novel approach to the frequency domain BSS using frequency normalization. Under the conditions of almost sparse sources and of dominant direct path in the mixing systems, we show that the new approach provides better performance than the conventional one when the amount of available data is small.

UR - http://www.scopus.com/inward/record.url?scp=33947708952&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33947708952&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33947708952

SN - 142440469X

SN - 9781424404698

T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

SP - V837-V840

BT - 2006 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings

T2 - 2006 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2006

Y2 - 14 May 2006 through 19 May 2006

ER -

Frequency domain blind source separation of a reduced amount of data using frequency normalization

Abstract

Publication series

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this