Analysis of broadband GEVD-based blind source separation

Soydan  Redif; Jennifer Pestana; Ian K. Proudler

doi:10.1109/ICASSP.2019.8683237

Analysis of broadband GEVD-based blind source separation

Soydan Redif, Jennifer Pestana, Ian K. Proudler

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

3 Citations (Scopus)

18 Downloads (Pure)

Abstract

One approach to blind source separation of instantaneously mixed, non-stationary sources involves using the generalized eigenvalue decomposition of two estimated covariance matrices. The assumption is that the source statistics change with time whilst the mixing matrix does not. A recent generalisation of this approach to convolutive mixtures was achieved by extending the generalized eigenvalue decomposition to polynomial matrices. In this paper, we present a further investigation into this broadband BSS technique. We derive some expressions for the conditions under which source separation is possible. The validity of our analysis is illustrated through some computer simulations.

Original language	English
Title of host publication	ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Place of Publication	Piscataway, NJ
Publisher	IEEE
Pages	8028-8032
Number of pages	5
ISBN (Electronic)	9781479981311
ISBN (Print)	9781479981311
DOIs	https://doi.org/10.1109/ICASSP.2019.8683237
Publication status	E-pub ahead of print - 17 Apr 2019

Keywords

non-stationary
broadband
blind signal separation
generalised eigenvalue decomposition

Access to Document

10.1109/ICASSP.2019.8683237

Redif-etal-ICASSP2019-Analysis-of-broadband-GEVD-based-blind-source-separationAccepted author manuscript, 459 KB

Cite this

@inproceedings{a965298c370e49ce8bd34f4f0dc9aeee,

title = "Analysis of broadband GEVD-based blind source separation",

abstract = "One approach to blind source separation of instantaneously mixed, non-stationary sources involves using the generalized eigenvalue decomposition of two estimated covariance matrices. The assumption is that the source statistics change with time whilst the mixing matrix does not. A recent generalisation of this approach to convolutive mixtures was achieved by extending the generalized eigenvalue decomposition to polynomial matrices. In this paper, we present a further investigation into this broadband BSS technique. We derive some expressions for the conditions under which source separation is possible. The validity of our analysis is illustrated through some computer simulations.",

keywords = "non-stationary, broadband, blind signal separation, generalised eigenvalue decomposition",

author = "Soydan Redif and Jennifer Pestana and Proudler, {Ian K.}",

note = "{\textcopyright} 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.",

year = "2019",

month = apr,

day = "17",

doi = "10.1109/ICASSP.2019.8683237",

language = "English",

isbn = "9781479981311",

pages = "8028--8032",

booktitle = "ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)",

publisher = "IEEE",

}

TY - GEN

T1 - Analysis of broadband GEVD-based blind source separation

AU - Redif, Soydan

AU - Pestana, Jennifer

AU - Proudler, Ian K.

N1 - © 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

PY - 2019/4/17

Y1 - 2019/4/17

N2 - One approach to blind source separation of instantaneously mixed, non-stationary sources involves using the generalized eigenvalue decomposition of two estimated covariance matrices. The assumption is that the source statistics change with time whilst the mixing matrix does not. A recent generalisation of this approach to convolutive mixtures was achieved by extending the generalized eigenvalue decomposition to polynomial matrices. In this paper, we present a further investigation into this broadband BSS technique. We derive some expressions for the conditions under which source separation is possible. The validity of our analysis is illustrated through some computer simulations.

AB - One approach to blind source separation of instantaneously mixed, non-stationary sources involves using the generalized eigenvalue decomposition of two estimated covariance matrices. The assumption is that the source statistics change with time whilst the mixing matrix does not. A recent generalisation of this approach to convolutive mixtures was achieved by extending the generalized eigenvalue decomposition to polynomial matrices. In this paper, we present a further investigation into this broadband BSS technique. We derive some expressions for the conditions under which source separation is possible. The validity of our analysis is illustrated through some computer simulations.

KW - non-stationary

KW - broadband

KW - blind signal separation

KW - generalised eigenvalue decomposition

UR - https://2019.ieeeicassp.org/

U2 - 10.1109/ICASSP.2019.8683237

DO - 10.1109/ICASSP.2019.8683237

M3 - Conference contribution book

SN - 9781479981311

SP - 8028

EP - 8032

BT - ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

PB - IEEE

CY - Piscataway, NJ

ER -

Analysis of broadband GEVD-based blind source separation

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this