Divot | Lab Index

Diffusion-powered video tokenizer for comprehension and generation, unifying video understanding and synthesis. Published at CVPR 2025.

Paper

arXiv: 2412.04432

Venue: CVPR 2025

videotokenizergenerationresearch