WING transfers egocentric video skills to robots, hitting 99.2% on LIBERO
hongkongust · hf · 2026-10-07
Researchers from HKUST(GZ) propose WING (World Action Learning via INteraction-Centric Spectral Latent Guidance) to transfer manipulation knowledge from human egocentric videos to robot policies.
- Problem: latent actions inferred from frame reconstruction are dominated by ego-camera motion, and human/robot temporal dynamics differ.
- Method: WING separates observer-induced motion from hand-object interaction, distills interaction-centric latent actions, then identifies shared low-frequency spectral components between egocentric latent actions and robot behavior to guide action generation.
- Results: 99.20% average success on LIBERO, 93.80% on RoboTwin 2.0, 57.7% on RoboCasa-GR1, plus strong performance on four real-world manipulation tasks.
More from Embodied
- YC-backed Vexo launches always-listening AI bracelet that keeps your context — ycombinator · 2026-10-07
- H-JEPA: Hierarchical world models lift visual AntMaze success from 18% to 73% — ylecun · 2026-10-07
- DiVeR preprint trains VLA verifiers only on decision-critical states for test-time scaling — SharonYixuanLi · 2026-10-07
- Million-mile trucker after Tesla Semi factory tour: this industry is going to be changed — elonmusk · 2026-10-07
- Attacca trains embodied agents for state-continuity, up to 7x on long-horizon tasks — nanyang-technological-university-singapore · 2026-10-07
- Boston Dynamics Names Former Amazon Executive Rohit Prasad as CEO — SumitGup · 2026-10-07