JVS
A free Japanese multi-speaker voice corpus with 100 speakers, 30 hours of recordings in normal, whisper, and falsetto styles
Duration
30 hours
Languages
1
Sample Rate
24 kHz
Published
2019-08
Description
1100 speakers, 30 hours of speech data
2Includes three speaking styles: normal, whisper, and falsetto
322 hours of parallel normal speech
4Suitable for voice conversion and multi-speaker modeling research
Language Details
| Language | Duration |
|---|---|
| Japanese | 30 hours |
Publisher