Fetching the paper…

SoundCTM: Unifying Score-based and Consistency Models for Full-band Text-to-Sound Generation · Around