| 🛠 | This subject is work in progress. Please bear with us while improvements are being made and assume good faith until the edits are complete. For information on how to help, see the guidelines. More subjects categorized here. |
🛠 |
JSUT Corpus is a free database of Japanese audio data made for research purposes. The corpus is comprised of different corpuses, JSUT, which contains reading-style audio of various vocabulary,[1] JSUT-song, which contains singing audio of 27 different children's songs,[2] JSUT-vi, which contains recorded vocal imitations of various sounds and JSUT-book,[3] which contains narration data of 6 different books.[4] All voiced by a single voice provider.
JSUT-song was used to develop a voicebank, "JSUT", for the vocal synthesiser NEUTRINO and NNSVS.
Voicebanks[]
- JSUT-song
- JSUT Version 0.300 (NEUTRINO), June 12, 2020
- Version 0.310, July 10, 2020
- Version 0.400, September 18, 2020
- Version 0.411, November 11, 2020
- Version 0.412, December 9, 2020
- Version 0.420, December 24, 2020
- Version 0.422, January 29, 2021
- Version 0.430, March 29, 2021
- Version 0.501, October 1, 2021
- Version 0.510, October 7, 2021
- Version 0.540, February 10, 2022
- Version 0.550, March 25, 2022
- Electron Version 1.0.0, April 22, 2022
- Electron Version 1.1.0, June 8, 2022
References[]
External links[]
Articles[]
[]
| ||||||||||||||||||
| ||||||||||||||||||