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