@svpino: Huge leap in video generation! Look at the faces here. For the first time, we have a tool that doesn't change character…

X AI KOLs Following Models

Summary

BACH is introduced as a significant advancement in video generation, achieving unprecedented character consistency across scenes without face morphing or drift.

Huge leap in video generation! Look at the faces here. For the first time, we have a tool that doesn't change characters' faces between scenes. This is BACH. Characters with the same bone structure, same skin tone, same eyes across every single clip. Watch the close-ups. No morphing. No drift. Just one consistent character, shot after shot.
Original Article

Similar Articles

Avatar V: Scaling Video-Reference Avatar Video Generation

Hugging Face Daily Papers

Avatar V is a production-scale framework for generating behaviorally recognizable avatar videos conditioned on full video references, introducing sparse reference attention and motion representation streams to achieve state-of-the-art identity preservation and lip synchronization.