Untitled entry

Summary from the TLDR AI newsletter

Researchers have created a new model called PVLFF that can look at 3D scenes and not only identify what's in them, but also distinguish between different instances of the same object—all by using text descriptions.