issues
search
kookmin-sw
/
capstone-2024-08
아나운서 준비생을 위한 맞춤형 AI 스피치 연습 애플리케이션, Loro(로로)
https://kookmin-sw.github.io/capstone-2024-08
5
stars
4
forks
source link
[CD-162] speaker condition을 vits2 text encoder에 전달
#65
Closed
YunHaaaa
closed
3 months ago
YunHaaaa
commented
3 months ago
Feature Description
VAE에서 얻은 speaker condition을 vits2 text encoder에 전달
Tasks to Complete
[x] VAE inference
[x] vits2 text encoder로 전달
[x] config, param, code 수정
[x] Time step hidden representation과 Speech feature matrix와 attention
[x] speaker condition 전달
Additional Resources (Optional)
Feature Description
VAE에서 얻은 speaker condition을 vits2 text encoder에 전달
Tasks to Complete
Additional Resources (Optional)