analiza aiqbench
The paper's core contribution is a learning pipeline: human demonstrations are retargeted to robot hands, then used to train a hierarchical policy that plans contact-surface trajectories. Its headline result, a 50-percentage-point gain in final-stage success over baselines, is reported on six hand-task pairs across two robot hands, so readers should treat it as promising but narrow evidence. The public release of code and hardware designs makes the claims checkable, which matters for anyone working on imitation learning for dexterous manipulation.
Contact Imitation Interface: Transferring Dexterous Skills by Showing How Contact Unfolds
arXiv:2610.41028v1 Announce Type: new Abstract: Robot-free human demonstrations offer a promising route to scaling dexterous manipulation data. The key challenge is to preserve the demonstrator's natural dexterity while producing contact-grounded observations and actions that transfer to robot hands. We present Contact Imitation Interface (CIMI), a framework with contact-grounded demonstration and learning interfaces. The demonstration interface combines shared contact geometry and fingertip RGB views with contact-alignment-first retargeting and reprojection of wider scene views. The learning interface introduces a hierarchical policy in which a contact-surface planner predicts relative contact-surface trajectories and a controller coordinates wrist and finger commands to improve contact alignment. CIMI combines contact-grounded observations, retargeting, and action representations, improving mean final-stage success by at least 50 percentage points over baseline policies across six hand-task pairs on LEAP and Wuji. Hardware designs and software are publicly available.
Abstract:Robot-free human demonstrations offer a promising route to scaling dexterous manipulation data. The key challenge is to preserve the demonstrator's natural dexterity while producing contact-grounded observations and actions that transfer to robot hands. We present Contact Imitation Interface (CIMI), a framework with contact-grounded demonstration and learning interfaces. The demonstration interface combines shared contact geometry and fingertip RGB views with contact-alignment-first retargeting and reprojection of wider scene views. The learning interface introduces a hierarchical policy in which a contact-surface planner predicts relative contact-surface trajectories and a controller coordinates wrist and finger commands to improve contact alignment. CIMI combines contact-grounded observations, retargeting, and action representations, improving mean final-stage success by at least 50 percentage points over baseline policies across six hand-task pairs on LEAP and Wuji. Hardware designs and software are publicly available.
15 pages (8-page main paper and 7-page supplementary material). Project website with code and hardware designs: this https URL
Robotics (cs.RO)
arXiv:2610.41028 [cs.RO]
(or arXiv:2610.41028v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2610.41028 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 243 cuvinte
Subiecte conexe
A Recipe for Efficient Sim-to-Real Transfer in Manipulation with Online Imitation-Pretrained World Models
arXiv · 22 sept. 2026
Beyond Trajectory Imitation: Strategy-Guided Policy Optimization for LLM Reasoning
arXiv · 15 oct. 2026
Coupled Calibration and Learning: Mitigating Teacher Bias in LLM Distillation without Target-Domain Reward Feedback
arXiv · 16 sept. 2026
Co-Evolving Harnesses and Models: On-Policy Correction Helps Weaker Models Catch Up Where Imitation Fails
arXiv · 9 sept. 2026