search

instant-high / wav2lip-onnx-HQ

Full version of wav2lip-onnx including face alignment and face enhancement and more...
32 stars 7 forks source link

Hallo, ich würde gerne eine Frage zu einer der Flaggen klären. #2

Closed NeuroDonu closed 3 months ago

NeuroDonu commented 3 months ago

Ich habe das Notepad, das unter dem Link verfügbar ist, benutzt und dieses Flag gesehen: --final_audio in Ihrer Datei inference_onnxModel.py. Können Sie bitte erklären, wozu das gut sein soll?

instant-high commented 3 months ago

If you use --final_audio, the --audio is used only for inference while the final output video will be muxed using --final_audio.

Eg. use a (separated) vocal track for inference and full music track for final result.

Or sometimes a voice may be not suitable for wav2lip inference (rvc voice). In this case you can use original voice for inference and the cloned voice as final result.