| Presentation | 2013-01-31 A study on Speaker-Normalized Style Conversion for Arbitrary Speaker's Expressive Speech Synthesis Hiroki KANAGAWA, Takashi NOSE, Takao KOBAYASHI, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | This paper proposes a technique for improving naturalness of synthetic speech using a framework of speaker adaptive training in HMM-based style conversion. In the style conversion, speaker-independent linear transforms are estimated using neutral- and target-style speech data of multiple speakers, and estimated transforms are applied to a new speaker's neutral-style model. As a result, we can convert the style expressivity of the acoustic model to the target style without preparing any target-style speech of the speaker. When spectral and prosodic features of training speakers are significantly different from each other, naturalness of synthetic speech fioni the converted model decreases. The proposed technique attempt to alleviate this problem by normalizing speaker characteristics using an approach similar to speaker adaptive training. From the objective and subjective evaluation results, we show that the speaker normalization technique can provide more natural sounding speech. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | HMM-based speech synthesis / speaking style and emotional expression / expressive speech / style adaptation / linear regression |
| Paper # | SP2012-110 |
| Date of Issue |
| Conference Information | |
| Committee | SP |
|---|---|
| Conference Date | 2013/1/23(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Speech (SP) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | A study on Speaker-Normalized Style Conversion for Arbitrary Speaker's Expressive Speech Synthesis |
| Sub Title (in English) | |
| Keyword(1) | HMM-based speech synthesis |
| Keyword(2) | speaking style and emotional expression |
| Keyword(3) | expressive speech |
| Keyword(4) | style adaptation |
| Keyword(5) | linear regression |
| 1st Author's Name | Hiroki KANAGAWA |
| 1st Author's Affiliation | Interdisciplinary Graduate School of Science and Engineering, Tokyo Institute of Technology() |
| 2nd Author's Name | Takashi NOSE |
| 2nd Author's Affiliation | Interdisciplinary Graduate School of Science and Engineering, Tokyo Institute of Technology |
| 3rd Author's Name | Takao KOBAYASHI |
| 3rd Author's Affiliation | Interdisciplinary Graduate School of Science and Engineering, Tokyo Institute of Technology |
| Date | 2013-01-31 |
| Paper # | SP2012-110 |
| Volume (vol) | vol.112 |
| Number (no) | 422 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |