Contact-Aware Imitation Learning Through Contact Factorization
Overview
FACE framework enables robots to adapt to varying contact conditions while preserving task behavior, using normalized contact-relative coordinates and learned estimators for contact normal and friction scale. Evaluated on real-robot manipulation with unseen surface variations, showing robust generalization across contact conditions through controlled comparisons with prior approaches adapted to the setting. arXiv:2610.09533v1 Latest: FACE framework introduces contact-factorized imitation learning, separating task behavior from environment-dependent contact factors, allowing robots to generalize across unseen surface properties and geometries. arXiv:2610.09533v1
Generated from attributed reports · Updated 1 hours ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
No current evidence-backed claims. Missing information remains not reported.
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsContact-Aware Imitation Learning Through Contact Factorization
This research proposes FACE, a framework for imitation learning that enables robots to adapt to varying contact conditions while maintaining intended task behavior. By separating task behavior from environment-dependent contact factors, FACE uses normalized contact-relative coordinates and learned estimators for contact normal and friction scale, allowing for robust generalization across unseen surface properties and geometries.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).