Presentation 2016-10-27
Voice Activity Detection Using Throat Microphone and Lavalier Microphone for Multi-Party Conversations
Yoshihiro Otaka, Takashi Tsunakawa, Masafumi Nishida, Masafumi Nishimura,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) For analyzing multi-party conversations, accurate identification of the speaker and speech segment is important. For more accurate separation of the speech segments, multi-channel recording by using headset microphones is usually tried. Nevertheless, it has been difficult to detect the back-channel feedback and sympathize section, which tend to be superimposed on speech of the surrounding speaker. In this paper, we use throat microphone for stable recording of the target speaker without capturing the surrounding speech in the multi-party conversation. We propose a voice activity detection method using 2ch recording of throat microphone and lavalier microphone. The experimental results showed an improvement of voice activity detection performance by the proposed method. Besides, we developed on-line recording system working on smartphone to efficiently collect the Multi-party conversations. Furthermore, we collected multi-party conversations with elderly people and evaluated a voice activity detection performance of the proposed method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Voice Activity Detection / GMM / Throat Microphone / Multi-party Conversation
Paper # SP2016-43
Date of Issue 2016-10-20 (SP)

Conference Information
Committee SP
Conference Date 2016/10/27(1days)
Place (in Japanese) (See Japanese page)
Place (in English) Shizuoka University.
Topics (in Japanese) (See Japanese page)
Topics (in English) Learning, Recognition, Synthesis, Dialogue, etc.
Chair Kazunori Mano(Shibaura Inst. of Tech.)
Vice Chair Hiroki Mori(Utsunomiya Univ.)
Secretary Hiroki Mori(Kobe Univ.)
Assistant Taichi Asami(NTT) / Kei Hashimoto(Nagoya Inst. of Tech.)

Paper Information
Registration To Technical Committee on Speech
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Voice Activity Detection Using Throat Microphone and Lavalier Microphone for Multi-Party Conversations
Sub Title (in English)
Keyword(1) Voice Activity Detection
Keyword(2) GMM
Keyword(3) Throat Microphone
Keyword(4) Multi-party Conversation
1st Author's Name Yoshihiro Otaka
1st Author's Affiliation Shizuoka University(Shizuoka Univ.)
2nd Author's Name Takashi Tsunakawa
2nd Author's Affiliation Shizuoka University(Shizuoka Univ.)
3rd Author's Name Masafumi Nishida
3rd Author's Affiliation Shizuoka University(Shizuoka Univ.)
4th Author's Name Masafumi Nishimura
4th Author's Affiliation Shizuoka University(Shizuoka Univ.)
Date 2016-10-27
Paper # SP2016-43
Volume (vol) vol.116
Number (no) SP-279
Page pp.pp.15-20(SP),
#Pages 6
Date of Issue 2016-10-20 (SP)