Tag: vision-language models

1 article tagged vision-language models

6 Apr 2026 The Register

Netflix has open-sourced VOID, a vision-language model that removes objects from video and inpaints how remaining elements should behave afterwards.