Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment

Add code
Jul 05, 2026

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: