Presentation 2006/1/6
A Study of Restoring Voiceless Fricatives on Laryngectomee Speech Using Machine Learning Techniques
Koji MURAKAMI, Minoru SASAKI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Laryngectomee people are able to communicate with other people adequately using esophageal or alaryngeal speech because of much training. However, it is impossible for them to utter several phonemes, such as voiceless palatal fricative and voiceless glottal fricative because they do not have the organs. This can prevent their communication because it is very difficult to distinguish those sounds from independent vowels. In this paper, we try to find the syllables, whether independent vowels or the syllables missing phonemes using transcriptions of spontaneous speech. We have obtained results with high accuracy by using contexts around the focused syllables.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Laryngectomee people / Voiceless palatal fricative / Voiceless glottal fricative / Machine learning / Support Vector Machine / phoneme restoring
Paper # TL2005-43,WIT2005-68
Date of Issue

Conference Information
Committee WIT
Conference Date 2006/1/6(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 Well-being Information Technology(WIT)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Study of Restoring Voiceless Fricatives on Laryngectomee Speech Using Machine Learning Techniques
Sub Title (in English)
Keyword(1) Laryngectomee people
Keyword(2) Voiceless palatal fricative
Keyword(3) Voiceless glottal fricative
Keyword(4) Machine learning
Keyword(5) Support Vector Machine
Keyword(6) phoneme restoring
1st Author's Name Koji MURAKAMI
1st Author's Affiliation Computer Science Department, New York University()
2nd Author's Name Minoru SASAKI
2nd Author's Affiliation Faculty of Engineering, Ibaraki University
Date 2006/1/6
Paper # TL2005-43,WIT2005-68
Volume (vol) vol.105
Number (no) 508
Page pp.pp.-
#Pages 6
Date of Issue