HumanNet: Scaling Human-centric Video Learning to One Million Hours Paper • 2605.06747 • Published May 7 • 55
Rethinking Video Generation Model for the Embodied World Paper • 2601.15282 • Published Jan 21 • 46
Rethinking Video Generation Model for the Embodied World Paper • 2601.15282 • Published Jan 21 • 46
MHLA: Restoring Expressivity of Linear Attention via Token-Level Multi-Head Paper • 2601.07832 • Published Jan 12 • 53
MAGREF: Masked Guidance for Any-Reference Video Generation Paper • 2505.23742 • Published May 29, 2025 • 11