Netflix has open-sourced VOID, an AI model that removes objects from videos along with their physical interactions. When a person holding an object is removed, the object falls naturally rather than floating. Built on Alibaba's CogVideoX, it uses a quadmask to understand what should physically change after removal. The code and weights are on Hugging Face. https://www.marktechpost.com/2026/04/04/netflix-ai-team-just-open-sourced-void-an-ai-model-that-erases-objects-from-videos-physics-and-all/ #AIagent #AI #GenAI #ComputerVision #Netflix