search

OpenTalker / SadTalker

[CVPR 2023] SadTalker：Learning Realistic 3D Motion Coefficients for Stylized Audio-Driven Single Image Talking Face Animation
https://sadtalker.github.io/
Other
11.54k stars 2.16k forks source link

这个错误怎么搞 求大神指点 #880

Open Rainy-An opened 4 months ago

Rainy-An commented 4 months ago

2

hezilon commented 4 months ago

打开test_audio2coeff.py文件，将102-105行改成如下代码： savemat(os.path.join(coeff_savedir, '%s%s.mat'%(batch['pic_name'], batch['audio_name'])), {'coeff_3dmm': coeffs_pred_numpy}) return os.path.join(coeff_savedir, '%s%s.mat'%(batch['pic_name'], batch['audio_name']))

ermaccw commented 4 months ago

gradio版本降到3.46.1

Rainy-An commented 4 months ago

return os.path.join(coeff_savedir, '%s%s.mat'%(batch['pic_name'], batch['audio_name']))

搞定了兄弟 多谢