| Presentation | 2017-01-21 A Study on the Construction of Articulatory to Acoustic Mapping by Using Deep Neural Network Fumiaki Taguchi, Tokihiko Kaburagi, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | This paper presents a method for estimating time series of the acoustic property of the vocal tract expressed by line spectral frequencies from the movement trajectory of the articulatory organs observed with a magnetic sensor (EMA). Statistical methods have been used thus far for the articulatory-to-acoustic mapping because of the difficulty in modeling the relationship between both parameters physically. In this study, articulatory-to-acoustic mapping was constructed by using deep neural network, more specifically a multilayer perceptron, for which effectiveness has been confirmed in various problems. Experimental results showed that the cepstrum error in our estimation method was 3.39 dB when mngu0 dataset was used. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | articulatory movement / vocal tract spectrum / DNN / articulatory-to-acoustic mapping |
| Paper # | SP2016-73 |
| Date of Issue | 2017-01-14 (SP) |
| Conference Information | |
| Committee | SP |
|---|---|
| Conference Date | 2017/1/21(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | The University of Tokyo |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Synthesis, Generation, Prosody, 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) | A Study on the Construction of Articulatory to Acoustic Mapping by Using Deep Neural Network |
| Sub Title (in English) | |
| Keyword(1) | articulatory movement |
| Keyword(2) | vocal tract spectrum |
| Keyword(3) | DNN |
| Keyword(4) | articulatory-to-acoustic mapping |
| 1st Author's Name | Fumiaki Taguchi |
| 1st Author's Affiliation | Kyushu University(Kyushu Univ.) |
| 2nd Author's Name | Tokihiko Kaburagi |
| 2nd Author's Affiliation | Kyushu University(Kyushu Univ.) |
| Date | 2017-01-21 |
| Paper # | SP2016-73 |
| Volume (vol) | vol.116 |
| Number (no) | SP-414 |
| Page | pp.pp.53-57(SP), |
| #Pages | 5 |
| Date of Issue | 2017-01-14 (SP) |