Jinchuan Tian

Discover how preference alignment improves text-to-speech systems for better user experiences.

ESPnet-Codec enhances training and evaluation of neural codecs for audio and speech.

VERSA evaluates speech, audio, and music quality effectively.