Fetching the paper…

SAM2Act: Integrating Visual Foundation Model with A Memory Architecture for Robotic Manipulation · Around