Excited to present new work on using diffusion priors for video amodal segmentation and content completion!

with Deva Ramanan and @tarashakhurana.bsky.social

arXiv: arxiv.org/abs/2412.04623
project page: diffusion-vas.github.io