search

TMElyralab / MuseTalk

MuseTalk: Real-Time High Quality Lip Synchorization with Latent Space Inpainting
Other
1.85k stars 223 forks source link

生成结果视频不连续，黑屏 #79

Closed missu263 closed 1 month ago

missu263 commented 1 month ago

生成结果在12秒时候，就没有画面了。原视频50多秒

使用素材 https://github.com/TMElyralab/MuseTalk/assets/15882147/91062cd4-dc8e-4c11-9357-20d801a93f03

音频：audio_path: "data/audio/sun.wav"

生成结果如下 https://github.com/TMElyralab/MuseTalk/assets/15882147/3b507621-565c-4bde-8711-4e2a9dbfa059

alexLIUMinhao commented 1 month ago

Hi，

Your footage has not been edited well, and it contains many non-human scenes, which may cause the face detection model to fail. Therefore, please prepare a clip that includes frontal human subjects for testing.