| Presentation | 2015-12-02 Distant-talking speech recognition by reverberation-aware denoising autoencoder Yuma Ueda, Longbiao Wang, Atsuhiko Kai, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In the distant-talking speech recognition, it is essential to deal with the noise and reverberation.Denoising autoencoder (DAE) is known to be effective as a method for removing these influences.However, conventional DAE is easily affected by mismatch between training data and test data because the performance of DAE depend on the environment or amount of data included in the training set.In this study, we also use reverberation features estimated by Multi Step Linear Prediction (MSLP) as additional to input of DAE.By explicitly considering the effects of reverberation, we solve the problems in conventional DAE-based system.We evaluate the proposed method by using the ``REVERB challenge'' (Reverberant Voice Enhancement and Recognition Benchmark) dataset.For SimData, the average Word Error Rate (WER) was reduced from 7.12% to 6.41%.For RealData, the average WER was reduced from 30.56% to 26.83%. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | speech recognition / dereverberation / denoising autoencoder / distant-talking speech |
| Paper # | SP2015-77 |
| Date of Issue | 2015-11-25 (SP) |
| Conference Information | |
| Committee | NLC / IPSJ-NL / SP / IPSJ-SLP |
|---|---|
| Conference Date | 2015/12/2(3days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Nagoya Inst of Tech. |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | The Second Natural Language Processing Symposium & The 17th Spoken Language Symposium |
| Chair | Koichi Takeuchi(Okayama Univ.) / Kentaro Inui(Tohoku Univ.) / Kazunori Mano(Shibaura Inst. of Tech.) / Koichi Shinoda(東工大) |
| Vice Chair | Hiroshi Kanayama(IBM) / Makoto Ichise(NTT DoCoMo) / / Norihide Kitaoka(Tokushima Univ.) |
| Secretary | Hiroshi Kanayama(Univ. of Tokyo/Hottolink) / Makoto Ichise(Ryukoku Univ.) / (Osaka Univ.) / Norihide Kitaoka(Tohoku Univ.) / (Mixi Co. Ltd.) |
| Assistant | Kazutaka Shimada(Kyushu Inst. of Tech.) / Ryuichiro Higashinaka(NTT) / / Takashi Nose(Tohoku Univ.) / Taichi Asami(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) | Distant-talking speech recognition by reverberation-aware denoising autoencoder |
| Sub Title (in English) | |
| Keyword(1) | speech recognition |
| Keyword(2) | dereverberation |
| Keyword(3) | denoising autoencoder |
| Keyword(4) | distant-talking speech |
| 1st Author's Name | Yuma Ueda |
| 1st Author's Affiliation | Shizuoka University(Shizuoka Univ.) |
| 2nd Author's Name | Longbiao Wang |
| 2nd Author's Affiliation | Nagaoka University of Technology(Nagaoka Univ.) |
| 3rd Author's Name | Atsuhiko Kai |
| 3rd Author's Affiliation | Shizuoka University(Shizuoka Univ.) |
| Date | 2015-12-02 |
| Paper # | SP2015-77 |
| Volume (vol) | vol.115 |
| Number (no) | SP-346 |
| Page | pp.pp.55-60(SP), |
| #Pages | 6 |
| Date of Issue | 2015-11-25 (SP) |