Dex-One2Many learns dexterous manipulation from one human video, 95%+ success across 4 hands
jbhuang0604 · x · 2026-11-09
Researchers from SNU, University of Maryland and KAIST present Dex-One2Many (Oral at CoRL 2026 Scaling H2R Workshop): a system that learns dexterous manipulation from a single human demonstration video and generalizes to initial poses, goal poses, and grasps never shown in that video.
- Key insight: abstract the video into sequential scene graphs that guide RL, enabling efficient exploration while preserving broad generalizability — learning task structure beyond imitation.
- Results: across 5 tasks spanning direct manipulation (doll, can, stamp) and tool use (hammer, sweep), each learned from one human video, the policy reaches 94.6–96.8% mean success on four arm–hand embodiments by swapping only the robot model and its grasp set. Baselines merely imitate and only succeed where the scene matches the video.
More from Embodied
- Apple's HALO humanoid trains in 1000 simulated homes, hits 91% success in real houses — heyshrutimishra · 2026-11-10
- Boston Dynamics puts wheels on Atlas to stress-test its whole-body RL controller — CyberRobooo · 2026-11-09
- Guava distills GPT-5.4 into a 4B robot agent with just 2,000 simulated trajectories — jbhuang0604 · 2026-11-09
- Fly-by-Code: coding agents pilot drones via active visual and physical feedback loops — jbhuang0604 · 2026-11-09
- μ0 world model learns robot physics from video only, beats π0 by 5 points on RoboCasa365 — jbhuang0604 · 2026-11-09
- NVIDIA unveils GR00T 2 open-world robot action model, hitting 95% success with 5 demos — CyberRobooo · 2026-11-09