Presentation 2016-01-23
Analysis of 'current speaker selects next' behavior in speech rate converted conversations with voice feedback
Takahiro Hirose, Kosuke Kumagai, Hiroko Tokunaga, Hiroto Saito, Naoki Mukawa,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Speech rate conversion that slows utterances is one of the promising techniques for supporting impaired hearers such as the elderly and non-native language learners. When speech rate conversion systems are applied to real-time conversations, speakers have to wait until hearers finish listening to speakers’ whole utterances. This waiting time for speaker causes degradations of conversations, specifically at turn-takings. In order to solve these problems, we developed a novel speech rate conversation system that inserts speaker’s prolonged voice to that waiting pauses and lets the speaker to hear her/his feedback voice. Then, we apply the system to three-party video-mediated conversations and evaluate speakers’ gaze behaviors during the waiting time quantitatively. The results show that the number of averted gazes is decreased for inserted voice condition compared to without voice insertion condition. Case studies also show that the speaker who addresses next speaker gives fixed gaze to the person during the waiting time for inserted voice condition, while averted gazes are observed for without voice insertion condition. Thus, we can suggest that the system solves the troubles of turn-takings.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) speech rate conversion / voice feedback / turn taking / next speaker selection / three party conversation
Paper # HCS2015-76
Date of Issue 2016-01-15 (HCS)

Conference Information
Committee HCS
Conference Date 2016/1/22(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Yamato Kaigishitsu
Topics (in Japanese) (See Japanese page)
Topics (in English) Psychology and Life-stage of Communication, etc.
Chair Kinya Fujita(Tokyo Univ. of Agric. and Tech.)
Vice Chair Masafumi Matsuda(NTT) / Kyoko Ito(Osaka Univ.)
Secretary Masafumi Matsuda(Kanazawa Inst. of Tech.) / Kyoko Ito(Univ. of Tsukuba)
Assistant Masashi Komori(Osaka Electro-Comm. Univ.) / Takahiro Tanaka(Nagoya Univ.) / Haruka Yoshida(NEC)

Paper Information
Registration To Technical Committee on Human Communication Science
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Analysis of 'current speaker selects next' behavior in speech rate converted conversations with voice feedback
Sub Title (in English)
Keyword(1) speech rate conversion
Keyword(2) voice feedback
Keyword(3) turn taking
Keyword(4) next speaker selection
Keyword(5) three party conversation
1st Author's Name Takahiro Hirose
1st Author's Affiliation Tokyo Denki University(Tokyo Denki Univ.)
2nd Author's Name Kosuke Kumagai
2nd Author's Affiliation Tokyo Denki University(Tokyo Denki Univ.)
3rd Author's Name Hiroko Tokunaga
3rd Author's Affiliation Tokyo Denki University(Tokyo Denki Univ.)
4th Author's Name Hiroto Saito
4th Author's Affiliation Tokyo Denki University(Tokyo Denki Univ.)
5th Author's Name Naoki Mukawa
5th Author's Affiliation Tokyo Denki University(Tokyo Denki Univ.)
Date 2016-01-23
Paper # HCS2015-76
Volume (vol) vol.115
Number (no) HCS-418
Page pp.pp.101-106(HCS),
#Pages 6
Date of Issue 2016-01-15 (HCS)