Morning Edition · Tuesday, August 4, 2026Published at 2:16 AM EDT · New York

Meta's Segmentation and Vision Models Move Into Assistive Robotics and National Laboratory Science

The University of Pittsburgh applies Segment Anything and DINO to assistive manipulation, while Lawrence Berkeley National Laboratory uses the same models in early Genesis Mission projects.

Two Meta write-ups describe the same pair of open perception models doing production work in very different settings. At the University of Pittsburgh, researchers use Segment Anything and DINO to give assistive robotic arms the ability to i…