JSUT
A Japanese reading corpus designed around everyday character readings, with recordings by a single speaker.
Visit official source (opens in a new tab)Research use
Use it to test Japanese text normalization and pronunciation in single-speaker synthesis.
Limitations
One voice and scripted utterances limit coverage of speaker diversity and conversational prosody.
License and access
- Dataset license
- JSUT corpus terms (opens in a new tab)
- Access status
- Direct access
The publisher lists usage terms for the recordings and for the labels. Review both before reusing any component.