| Presentation | 2020-03-02 Japanese dialect speech classification using sequence-to-one neural networks Ryo Imaizumi, Ryo Masumura, Sayaka Shiota, Hitoshi Kiya, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | The language specific to a certain region is called a dialect, and the task of identifying which dialect the input speech is called dialect identification. Many of the speech recognition models are made up of standard languages, and it is known that the recognition performance is greatly reduced when using the model to recognize speech including dialects. One way to solve this problem is to use dialect information for learning and prepare a model that can identify both standard and dialects. Dialect identification to decide which recognizer to use for input speech is important. Also, if the accuracy of dialect identification is very high, improving the speech recognition system can be expected by optimizing the language model from the dialect-specific information, so improving the accuracy of the dialect identification model is an important task. In this study, we used a neural network as a model for dialect identification, and used a framework called End-to-End, which has been widely used in recent years. In the experiment, the performance of the discrimination model for dialects in six regions, Aomori, Hiroshima, Kumamoto, Nagoya, Sapporo, and Sendai, was investigated and analyzed using a sequential classification neural network with various parameters. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Japanese dialect identification / sequence-to-one neural network / LSTM / BLSTM |
| Paper # | EA2019-108,SIP2019-110,SP2019-57 |
| Date of Issue | 2020-02-24 (EA, SIP, SP) |
| Conference Information | |
| Committee | SP / EA / SIP |
|---|---|
| Conference Date | 2020/3/2(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Okinawa Industry Support Center |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Hisashi Kawai(NICT) / Kenichi Furuya(Oita Univ.) / Naoyuki Aikawa(TUS) |
| Vice Chair | Akinobu Ri(Nagoya Inst. of Tech.) / Suehiro Shimauchi(Kanazawa Inst. of Tech.) / Shigeto Takeoka(Shizuoka Inst. of Science and Tech.) / Kazunori Hayashi(Osaka City Univ) / Yukihiro Bandou(NTT) |
| Secretary | Akinobu Ri(Kyoto Univ.) / Suehiro Shimauchi(Waseda Univ.) / Shigeto Takeoka(NHK) / Kazunori Hayashi(Univ. of Tokyo) / Yukihiro Bandou(Hiroshima Univ.) |
| Assistant | Tomoki Koriyama(Univ. of Tokyo) / Yusuke Ijima(NTT) / Keisuke Imoto(Ritsumeikan Univ.) / Daisuke Morikawa(Toyama Pref Univ.) / Kenjiro Sugimoto(Waseda Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Speech / Technical Committee on Engineering Acoustics / Technical Committee on Signal Processing |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Japanese dialect speech classification using sequence-to-one neural networks |
| Sub Title (in English) | |
| Keyword(1) | Japanese dialect identification |
| Keyword(2) | sequence-to-one neural network |
| Keyword(3) | LSTM |
| Keyword(4) | BLSTM |
| 1st Author's Name | Ryo Imaizumi |
| 1st Author's Affiliation | Tokyo Metropolitan University(TMU) |
| 2nd Author's Name | Ryo Masumura |
| 2nd Author's Affiliation | Nippon Telegraph and Telephone Corporation(NTT) |
| 3rd Author's Name | Sayaka Shiota |
| 3rd Author's Affiliation | Tokyo Metropolitan University(TMU) |
| 4th Author's Name | Hitoshi Kiya |
| 4th Author's Affiliation | Tokyo Metropolitan University(TMU) |
| Date | 2020-03-02 |
| Paper # | EA2019-108,SIP2019-110,SP2019-57 |
| Volume (vol) | vol.119 |
| Number (no) | EA-439,SIP-440,SP-441 |
| Page | pp.pp.41-46(EA), pp.41-46(SIP), pp.41-46(SP), |
| #Pages | 6 |
| Date of Issue | 2020-02-24 (EA, SIP, SP) |