LibriTTS-R
Audio quality enhanced version of LibriTTS using the Miipher speech restoration model, improving ~585 hours of English TTS data
Duration
585 hours
Languages
1
Sample Rate
24 kHz
Published
2023-05
Description
1Audio quality enhanced version based on the LibriTTS corpus
2Uses Google's Miipher speech restoration model to improve audio quality
3Maintains the same data structure and scale as LibriTTS (~585 hours)
424 kHz sampling rate, multi-speaker
5Designed for high-quality TTS research
Language Details
| Language | Duration |
|---|---|
| English | 585 hours |
Publisher