Presentation 2007/6/8
Improvement of the Performance of Blind Signal Separation Using A Nonlinear Digital Filter
Masato KATOU, Hidekazu KUBOTA, Kaoru ARAKAWA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A new method for improving the performance of blind signal separation is proposed using an adaptive CS filter, which is a nonlinear digital filter proposed by authors before. The adaptive CS filter is effective for reducing small-amplitude additive random noise on speech signals, while preserving the signal waveform of both the abruptly changing parts and continuously changing ones of speech. In this paper, the adaptive CS filter is newly applied as a postprocessor of blind signal separation. When noise is mixed into the input speech signal for the blind signal separation, the performance of the separation is degraded. We show that the performance is improved by applying the output of the blind signal separation to the combination of spectrum subtraction method and the adaptive CS filter. The new system is verified to be effective in processing some actual mixed sound signals including noise.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) nonlinear digital filter / spectrum subtraction / an adaptation CS filter / blind signal separation / speech processing
Paper # SIS2007-15
Date of Issue

Conference Information
Committee SIS
Conference Date 2007/6/8(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 Smart Info-Media Systems (SIS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Improvement of the Performance of Blind Signal Separation Using A Nonlinear Digital Filter
Sub Title (in English)
Keyword(1) nonlinear digital filter
Keyword(2) spectrum subtraction
Keyword(3) an adaptation CS filter
Keyword(4) blind signal separation
Keyword(5) speech processing
1st Author's Name Masato KATOU
1st Author's Affiliation Dept. of Computer Science, Meiji University()
2nd Author's Name Hidekazu KUBOTA
2nd Author's Affiliation Dept. of Computer Science, Meiji University
3rd Author's Name Kaoru ARAKAWA
3rd Author's Affiliation Dept. of Computer Science, Meiji University
Date 2007/6/8
Paper # SIS2007-15
Volume (vol) vol.107
Number (no) 94
Page pp.pp.-
#Pages 4
Date of Issue