Netflix Releases VOID AI Model for Video Object Removal

TLDR

Netflix has released VOID (Video Object and Interaction Deletion), a vision-language model that removes objects from video footage while realistically adjusting how remaining elements behave.

In a 25-person survey across multiple scenarios, 'VOID was preferred 64.8 percent of the time, with Runway coming in a distant second at 18.4 percent.' Netflix has made the model publicly available on Hugging Face.

Key Takeaways

• Netflix open-sources VOID, a vision-language model that removes objects from video while generating physically plausible replacements for remaining scene elements