v0.0.2
版本发布时间: 2023-04-08 18:49:41
OpenTalker/SadTalker最新发布版本:v0.0.2-rc(2023-06-05 22:44:03)
In v0.0.2, we finally solved the problems of paste back and got a more stable face with the following options using a newly trained model!
python inference.py --still --enhancer gfpgan --preprocess full
https://user-images.githubusercontent.com/4397546/230717108-6f6214ed-dbf7-4ff9-a7c0-45a5fd02f010.mp4
https://user-images.githubusercontent.com/4397546/230717142-3b2e8468-4b74-40fa-90d2-e884b6935416.mp4
https://user-images.githubusercontent.com/4397546/230717168-0e710491-6f49-481f-8b62-1b42cb058fd6.mp4
Have fun and thanks for all your support! ❤️
1、 auido2exp_00300-model.pth 32.69MB
2、 auido2pose_00140-model.pth 91.47MB
3、 BFM_Fitting.zip 385.33MB
4、 epoch_20.pth 275.48MB
5、 facevid2vid_00189-model.pth.tar 1.97GB
6、 hub.zip 105.17MB
7、 mapping_00109-model.pth.tar 148.56MB
8、 mapping_00229-model.pth.tar 148.32MB
9、 shape_predictor_68_face_landmarks.dat 95.08MB
10、 wav2lip.pth 415.62MB