An advanced singing voice synthesis system with high fidelity, expressiveness, controllability and flexibility based on DiffSinger: Singing Voice Synthesis via Shallow Diffusion Mechanism
Apache License 2.0
2.62k
stars
275
forks
source link
Support local retaking on pitch and variance parameters #87
retaking = re-generate parameter curves local retaking = keep curves where mask==0 and re-fill curves where mask==1