SegTune enables fine-grained control in song generation

Arxiv CS.SD (Sound/Audio ML)

SegTune introduces a method for structured control over temporally varying musical attributes like timbre and dynamics. The approach allows fine-grained editing beyond lyrics and global prompts.

Arxiv CS.SD (Sound/Audio ML)·Pengfei CaiJoanna WangHaorui ZhengXu LiZihao JiTeng MaZhongliang LiuChen ZhangPengfei Wan

8 days ago