Presentation | 2017-12-21 [Poster Presentation] Development of Speaker/Environment-Dependent Acoustic Model for Non-Audible Murmur Recognition Based on DNN Adaptation Seita Noda, Tomoki Hayashi, Tomoki Toda, Kazuya Takeda, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In this research, we aim to improve the performance of non-audible murmur (NAM) recognition towards the development of silent speech interfaces. First, we apply a deep learning technique to the acoustic modeling of NAM and develop speaker-dependent acoustic models using speaker adaptive training for a deep neural network (DNN). Moreover, to improve the recognition performance under noisy conditions, we also develop the noise-dependent acoustic models using noisy NAM data generated by superimposing noise signals recorded with NAM microphone. Experimental results show that our developed speaker- and noise-dependent acoustic models are effective for significantly improving NAM recognition performance under both clean and noisy conditions and achieving recognition performance under some noisy conditions comparable to that under the clean condition. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Non-Audible Murmur recognition / acoustic model / deep learning / speaker adaptation / noise adaptation |
Paper # | SP2017-56 |
Date of Issue | 2017-12-14 (SP) |
Conference Information | |
Committee | NLC / IPSJ-NL / SP / IPSJ-SLP |
---|---|
Conference Date | 2017/12/20(3days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Waseda Univ. Green Computing Systems Research Organization |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | The 4th Natural Language Processing Symposium & The 19th Spoken Language Symposium |
Chair | Hiroshi Kanayama(IBM) / Kentaro Inui(Tohoku Univ.) / Yoichi Yamashita(Ritsumeikan Univ.) / Nobuaki Minematsu(Univ. Tokyo) |
Vice Chair | Takeshi Sakaki(Hottolink) / Kazutaka Shimada(Kyushu Inst. of Tech.) / / Hiroki Mori(Utsunomiya Univ.) |
Secretary | Takeshi Sakaki(Ryukoku Univ.) / Kazutaka Shimada(NTT) / (Osaka Univ.) / Hiroki Mori(Tokyo Inst. of Tech.) / (Mixi Co. Ltd.) |
Assistant | Mitsuo Yoshida(Toyohashi Univ. of Tech.) / Takeshi Kobayakawa(NICT) / / Kei Hashimoto(Nagoya Inst. of Tech.) / Satoshi Kobashikawa(NTT) |
Paper Information | |
Registration To | Technical Committee on Natural Language Understanding and Models of Communication / Special Interest Group on Natural Language / Technical Committee on Speech / Special Interest Group on Spoken Language Processing |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | [Poster Presentation] Development of Speaker/Environment-Dependent Acoustic Model for Non-Audible Murmur Recognition Based on DNN Adaptation |
Sub Title (in English) | |
Keyword(1) | Non-Audible Murmur recognition |
Keyword(2) | acoustic model |
Keyword(3) | deep learning |
Keyword(4) | speaker adaptation |
Keyword(5) | noise adaptation |
1st Author's Name | Seita Noda |
1st Author's Affiliation | Nagoya University(Nagoya Univ.) |
2nd Author's Name | Tomoki Hayashi |
2nd Author's Affiliation | Nagoya University(Nagoya Univ.) |
3rd Author's Name | Tomoki Toda |
3rd Author's Affiliation | Nagoya University(Nagoya Univ.) |
4th Author's Name | Kazuya Takeda |
4th Author's Affiliation | Nagoya University(Nagoya Univ.) |
Date | 2017-12-21 |
Paper # | SP2017-56 |
Volume (vol) | vol.117 |
Number (no) | SP-368 |
Page | pp.pp.7-10(SP), |
#Pages | 4 |
Date of Issue | 2017-12-14 (SP) |