Presentation 2005/7/16
Recognition of Non-Verbal Speech Using Imitated Word Model
Shinya TAKAHASHI, Tsuyoshi MORIMOTO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes imitated word models that represent non-verbal sounds, especially cough sounds here, as phoneme sequences. In conventional speech recognition systems, non-verbal sounds, so-called human noises, are processed as burden noises that cause mis-recognition. Non-verbal sounds are, however, important information to know user's physical and psychological condition. In particular, coughing is one of the most important barometers of daily health check, so we propose an approach to detect the cough sounds from user utterances using the imitated word models constructed by clustering of phoneme sequences obtained in phoneme recognition. The experimental results show that this approach can improve the correct rates and the accuracies for words and coughs compared with the approach using HMM constructed from cough waveforms.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) non-verbal sound / onomatopoeia / cough recognition / imitated word model
Paper # NLC2005-21
Date of Issue

Conference Information
Committee NLC
Conference Date 2005/7/16(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) Recognition of Non-Verbal Speech Using Imitated Word Model
Sub Title (in English)
Keyword(1) non-verbal sound
Keyword(2) onomatopoeia
Keyword(3) cough recognition
Keyword(4) imitated word model
1st Author's Name Shinya TAKAHASHI
1st Author's Affiliation Dept. of EECS, Fukuoka University()
2nd Author's Name Tsuyoshi MORIMOTO
2nd Author's Affiliation Dept. of EECS, Fukuoka University
Date 2005/7/16
Paper # NLC2005-21
Volume (vol) vol.105
Number (no) 204
Page pp.pp.-
#Pages 5
Date of Issue