VOID: Video Object and Interaction Deletion with 5B CogVideoX Architecture by Netflix

VOID (Video Object and Interaction Deletion) by Netflix removes objects from videos along with all interactions they induce on the scene — not just secondary effects like shadows and reflections, but physical interactions like objects falling when a person is removed. Built on CogVideoX-Fun-V1.5-5b-InP, it uses a novel quadmask conditioning (4-value mask encoding primary object, overlap, affected regions, and background). Features a two-pass architecture with optical flow-warped refinement for temporal consistency.