Based on SD, it supports cultural videos, graphic videos, and video videos, and can maintain consistent roles and unlimited video length.
MuseTalk will be released later to support real-time mouth synchronization.
Online experience:https://huggingface.co/spaces/AnchorFake/MuseVDemo
Github：https://github.com/TMElyralab/MuseV
Post the paper later (Ali looks at others, open source first and then publish the paper!)
Video: