Presentation 2005/6/17
Studies on a high riverberation-resistant blind source deconvolution based on independent component analysis
Takeshi KOYA, Keiichi KANEDA, Hiromu GOTANDA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In order to solve the blind source deconvolution, a convolved mixing process in the time domain is often transformed into an instantaneous mixing model in the frequency domain. However, the model is only an approximated one and thus does not work effectively under highly reverberant environments. By deviding the impulse response properly, Serviere has precisely transformed the time-domain convolved mixture to a frequency-domain convolved mixture and has proposed a new FDICA approach available under high reverberation. In the approach, however, the permuation and scaling problems are unresolved as well as the distortion due to whitening. In the present report, an improved approach without such problems is proposed and is confirmed to be valid in a highly reverberant real environment.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) high riverberation-resistant BSD / frequency-domain ICA / segmentally convoluted mixing model / permutation / scaling / whitening distortion
Paper # NC2005-21
Date of Issue

Conference Information
Committee NC
Conference Date 2005/6/17(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Studies on a high riverberation-resistant blind source deconvolution based on independent component analysis
Sub Title (in English)
Keyword(1) high riverberation-resistant BSD
Keyword(2) frequency-domain ICA
Keyword(3) segmentally convoluted mixing model
Keyword(4) permutation
Keyword(5) scaling
Keyword(6) whitening distortion
1st Author's Name Takeshi KOYA
1st Author's Affiliation Graduate School of Advanced Technology, Kinki University()
2nd Author's Name Keiichi KANEDA
2nd Author's Affiliation Graduate School of Advanced Technology, Kinki University
3rd Author's Name Hiromu GOTANDA
3rd Author's Affiliation Graduate School of Advanced Technology, Kinki University
Date 2005/6/17
Paper # NC2005-21
Volume (vol) vol.105
Number (no) 131
Page pp.pp.-
#Pages 6
Date of Issue