NTU-led survey reviews nearly 200 papers on affordance learning for embodied AI
jiqizhixin · x · 2026-10-18
A multi-institution team led by Nanyang Technological University (with HKUST(GZ), SJTU, UIUC, ETH Zurich and others) published a comprehensive survey, "From Passive Perception to Active Interaction," covering nearly 200 works on affordance learning for embodied AI.
- Core concept: affordances connect visual understanding to physical interaction — handles can be grasped, buttons pressed, empty tabletops used — and shift with the task, environment, and robot's body.
- Problem: existing surveys split the field between a computer-vision side (functional recognition, segmentation, localization) and a robotics side (cognitive foundations, developmental learning, RL).
- Contribution: as 3D perception, foundation models, and LLMs converge, the survey offers a unified view and traces the field's trajectory from passive perception to active interaction.
More from Embodied
- MIT's SANDO Drone Planner Avoids Every Moving Obstacle Across 12 Real Flights — Olivier__OG · 2026-10-18
- SimInfinity: open-source robotic simulation benchmark with 10K real-world tasks — joemeno · 2026-10-18
- Robi: a German/English voice companion demo built on Gemma, Tyger and Humaneness Voice TTS — realmrfakename · 2026-10-18
- Munich's Plim unveils wheeled home-care humanoid controllable from 9,000 km away — CyberRobooo · 2026-10-18
- Musk says Optimus and Cybercab will make up the vast majority of Tesla's value — XFreeze · 2026-10-18
- SAE weight editing steers robot behavior at inference time, IIIT Hyderabad paper lands IJCAI Safe Physical AI workshop — ponguru · 2026-10-18