Presentation 1995/12/14
Evaluation of Sound Stream Segregation from the viewpoint of Speech Recognition
Hiroshi G. Okuno, Tomohiro Nakatani, Takeshi Kawabata,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper reports the results of the preliminary experiments for interfacing sound stream segregation systems to speech understanding system. Its main issue is spectrum distortion of input sound caused by three processing of sound stream segregation-harmonic structure extraction, head transfer function (HRTF), and grouping. The experiments with a discrete single codebook HMM-LR shows that spectrum distortion caused by harmonic structure extraction is negligible. Spectrum distortion caused by HRTF and grouping are considerable large and can be overcome by re-learning of the parameters of HMM-LR by training data applied by HRTF.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Computational Auditory Scene Analysis / Sound Stream Segregation / Speech Understanding / spectrum distortion
Paper # NLC95-51,SP95-86
Date of Issue

Conference Information
Committee NLC
Conference Date 1995/12/14(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 Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Evaluation of Sound Stream Segregation from the viewpoint of Speech Recognition
Sub Title (in English)
Keyword(1) Computational Auditory Scene Analysis
Keyword(2) Sound Stream Segregation
Keyword(3) Speech Understanding
Keyword(4) spectrum distortion
1st Author's Name Hiroshi G. Okuno
1st Author's Affiliation NTT Basic Research Laboratories()
2nd Author's Name Tomohiro Nakatani
2nd Author's Affiliation NTT Basic Research Laboratories
3rd Author's Name Takeshi Kawabata
3rd Author's Affiliation NTT Basic Research Laboratories
Date 1995/12/14
Paper # NLC95-51,SP95-86
Volume (vol) vol.95
Number (no) 428
Page pp.pp.-
#Pages 6
Date of Issue