Netflix and INSAIT researchers have open-sourced VOID, an AI model that removes objects from video while preserving physical consistency — handling shadows, reflections, and even gravity. The release could reshape post-production workflows and democratize professional-grade video editing.
