Text-to-Speech is awesome, but it requires text-audio pairs. So it's way too hard to extract data from animations or live streams. The good news is that voice conversion only needs audio data!
In this project, we achieve ACG voice convsersion using StarGANv2-VC, with non-parallel data!
- Google Drive
- Config: in this repository
- Colab: StarGANv2 (声音转换)
- Colab: StarGANv2 (语音、歌声声音转换)
- BILIBILI: 基于StarGANv2-VC的声音转换模型
QQ: 2235306122
BILIBILI: Francis-Komizu