Contact Imitation Interface: Transferring Dexterous Skills by Showing How Contact Unfolds
Overview
This research presents Contact Imitation Interface (CIMI), a framework for robot-free human demonstrations that preserves natural dexterity while generating contact-grounded observations and actions transferable to robot hands. The system combines shared contact geometry, fingertip RGB views, and contact-alignment-first retargeting in its demonstration interface, alongside a hierarchical policy with a contact-surface planner and wrist/finger controller in its learning interface. Evaluated on six hand-task pairs using the LEAP and Wuji platforms, CIMI improves mean final-stage success by at least 50 percentage points over baseline policies. Hardware designs and software are publicly available.
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 abstractsSignalContact Imitation Interface: Transferring Dexterous Skills by Showing How Contact Unfolds
This research presents Contact Imitation Interface (CIMI), a framework for robot-free human demonstrations that preserves natural dexterity while generating contact-grounded observations and actions transferable to robot hands. The system combines shared contact geometry, fingertip RGB views, and contact-alignment-first retargeting in its demonstration interface, alongside a hierarchical policy with a contact-surface planner and wrist/finger controller in its learning interface. Evaluated on six hand-task pairs using the LEAP and Wuji platforms, CIMI improves mean final-stage success by at least 50 percentage points over baseline policies. Hardware designs and software are publicly available.
Event coverage history
There is not enough continuous observation data to show a trend.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).