search

georgeretsi / smirk

Official Pytorch Implementation of SMIRK: 3D Facial Expressions through Analysis-by-Neural-Synthesis (CVPR 2024)
https://georgeretsi.github.io/smirk/
MIT License
167 stars 16 forks source link

Any solution to improve the quality of lipsyncronized video? #21

Open airpdev opened 2 months ago

airpdev commented 2 months ago

Hello, I have generated lipsync video from video and audio using smirk and other models.

https://github.com/user-attachments/assets/acdeeeaf-116f-4699-a7e0-259036b013bd

Video as input:

https://github.com/user-attachments/assets/a4fe31a3-0f85-45ef-abed-3c700646fe5d

jeb0813 commented 1 week ago

Hi @airpdev , is the first video generated by smirk? I tried on input video you provided, but the output vid has much flickering. Did you add extra steps or other models?

Thank you!