Reposted by Haoquan Fang
Really interesting work from Nvidia, UW, and AI2
Its called SAM2Act—a multi-view robotic transformer-based policy that integrates a visual foundation model with a memory architecture for robotic manipulation.
Project page: sam2act.github.io
Open code: github.com/sam2act/SAM2...
Its called SAM2Act—a multi-view robotic transformer-based policy that integrates a visual foundation model with a memory architecture for robotic manipulation.
Project page: sam2act.github.io
Open code: github.com/sam2act/SAM2...
GitHub - sam2act/SAM2Act: Official Repository for the code of SAM2Act
Official Repository for the code of SAM2Act. Contribute to sam2act/SAM2Act development by creating an account on GitHub.
github.com
February 11, 2025 at 8:16 PM
Really interesting work from Nvidia, UW, and AI2
Its called SAM2Act—a multi-view robotic transformer-based policy that integrates a visual foundation model with a memory architecture for robotic manipulation.
Project page: sam2act.github.io
Open code: github.com/sam2act/SAM2...
Its called SAM2Act—a multi-view robotic transformer-based policy that integrates a visual foundation model with a memory architecture for robotic manipulation.
Project page: sam2act.github.io
Open code: github.com/sam2act/SAM2...