UniPASE
UniPASE is Xiaobin Rong and collaborators’ open audio-to-audio speech enhancement model for restoring degraded speech across multiple distortions and sampling rates with low linguistic hallucination.
- Type
- Audio
- Weights
- Open
- Downloads
- 725
- Likes
- 9
- Pipeline
- audio to audio
- License
- apache-2.0
Model hub
Xiaobin-Rong/unipase
- Repo
- Xiaobin-Rong/unipase
- Access
- Public
Technical tags and languages
Papers
1
About Xiaobin Rong
Independent AI researcher and open-source developer specializing in speech enhancement, speech representation, synthesis, and lightweight neural audio models.