YuvalNirkin / fsgan

FSGAN - Official PyTorch Implementation
https://nirkin.com/fsgan
Creative Commons Zero v1.0 Universal
754 stars 147 forks source link

V1 face reenactment no sound #28

Closed lordlugo closed 4 years ago

lordlugo commented 4 years ago

Hi is it normal, it works great but there is no sound output

YuvalNirkin commented 4 years ago

Yes, I did not implement audio support. You can try adding the following lines at the end of the script:

import ffmpeg
ffmpeg.concat(ffmpeg.input(output_path), ffmpeg.input(target_path), v=1, a=1) \
    .output(output_with_audio_path, strict='-2').run(overwrite_output=True)
lordlugo commented 4 years ago

worked like a charm