Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Blind Signal Separation using Overcomplete Subband Representation

Grbic, Nedelko LU ; Tao, Xiao-Jiao ; Nordholm, Sven and Claesson, Ingvar LU (2001) In IEEE/ACM Transactions on Audio, Speech, and Language Processing 9(5). p.524-533
Abstract
This paper discusses a multirate filterbank-based extended infomax algorithm for real-world signal separation, i.e., convolved mixtures separation. Since convolution in the time domain corresponds to instantaneous mixing in the frequency domain, polyphase subband projection naturally becomes an efficient alternative to the Fourier transform based frequency domain approach. The online implementation proposed is featured by a simultaneous inverse channel identification in the frequency domain and signal filtering in the time domain. It is shown that an over-representation structure reduces aliasing between different bands and results in more accurate inverse channel estimates. Therefore, it provides better performance than the Fourier... (More)
This paper discusses a multirate filterbank-based extended infomax algorithm for real-world signal separation, i.e., convolved mixtures separation. Since convolution in the time domain corresponds to instantaneous mixing in the frequency domain, polyphase subband projection naturally becomes an efficient alternative to the Fourier transform based frequency domain approach. The online implementation proposed is featured by a simultaneous inverse channel identification in the frequency domain and signal filtering in the time domain. It is shown that an over-representation structure reduces aliasing between different bands and results in more accurate inverse channel estimates. Therefore, it provides better performance than the Fourier transform based structure in the measures of both separation and distortion. The performance limitation of the method is also evaluated in terms of the Wiener solution (Less)
Please use this url to cite or link to this publication:
author
; ; and
publishing date
type
Contribution to journal
publication status
published
subject
in
IEEE/ACM Transactions on Audio, Speech, and Language Processing
volume
9
issue
5
article number
5
pages
10 pages
publisher
Piscataway, NJ : Institute of Electrical and Electronics Engineers
external identifiers
  • scopus:0035397245
ISSN
2329-9290
DOI
10.1109/89.928917
language
English
LU publication?
no
id
d6e44ed3-1537-4a2b-96ff-7b592e9264cf
date added to LUP
2016-06-23 13:39:46
date last changed
2022-01-30 04:42:33
@article{d6e44ed3-1537-4a2b-96ff-7b592e9264cf,
  abstract     = {{This paper discusses a multirate filterbank-based extended infomax algorithm for real-world signal separation, i.e., convolved mixtures separation. Since convolution in the time domain corresponds to instantaneous mixing in the frequency domain, polyphase subband projection naturally becomes an efficient alternative to the Fourier transform based frequency domain approach. The online implementation proposed is featured by a simultaneous inverse channel identification in the frequency domain and signal filtering in the time domain. It is shown that an over-representation structure reduces aliasing between different bands and results in more accurate inverse channel estimates. Therefore, it provides better performance than the Fourier transform based structure in the measures of both separation and distortion. The performance limitation of the method is also evaluated in terms of the Wiener solution}},
  author       = {{Grbic, Nedelko and Tao, Xiao-Jiao and Nordholm, Sven and Claesson, Ingvar}},
  issn         = {{2329-9290}},
  language     = {{eng}},
  number       = {{5}},
  pages        = {{524--533}},
  publisher    = {{Piscataway, NJ : Institute of Electrical and Electronics Engineers}},
  series       = {{IEEE/ACM Transactions on Audio, Speech, and Language Processing}},
  title        = {{Blind Signal Separation using Overcomplete Subband Representation}},
  url          = {{http://dx.doi.org/10.1109/89.928917}},
  doi          = {{10.1109/89.928917}},
  volume       = {{9}},
  year         = {{2001}},
}