Dubbing Workflow

Thanks so much for the awesome work on this. In your paper, you shared some examples of video -> video, where the underlying video was kept the same and just the lips were changed. Is that possible with the currently released code or is that a separate pipeline? 