aiqbench analysis
The paper uses learned policies, specifically a residual RL policy over a frozen whole-body tracker, to teach a humanoid robot contact-rich manipulation from human videos. Its main idea is an object-centric contact plan that guides both pose reconstruction under occlusion and the reward for the policy. Reported gains are large in simulation, but they come from a single regrasp task and a few real-robot demos on one platform (Unitree G1), so readers should treat generalization as unproven.
ReContact: Reconstruct and Retarget Human Contacts from Video for Contact-Rich Humanoid Manipulation
arXiv:2611.02968v1 Announce Type: new Abstract: Contact transitions allow humanoid robots to redistribute object support across the whole body, enabling manipulation beyond hand-only grasps. Human videos provide a scalable source of such transitions for robots to learn from. However, body occlusion disrupts object pose tracking and corrupts retargeted motion. Moreover, accurately reproducing the object trajectory does not ensure that the robot applies stabilizing forces at the intended contacts, risking slips or drops in later phases. To address these challenges, we propose ReContact, which reconstructs and retargets human-object interactions through an object-centric contact plan. The plan specifies contact phases, contact pairs, anchors, pressing directions, and sticking or sliding modes. Contact priors stabilize object pose reconstruction through occlusion. A residual policy over a frozen whole-body tracker is rewarded for making contact at planned anchors with forces aligned to the specified directions. In simulation, on a thigh-supported regrasp task, ReContact achieves the planned two-palm bottom hold in 83.5 of episodes over three training seeds, versus 0.16% for the best adapted baseline. One policy trained on 247 references re-solved for new placements raises in-workspace tube-transfer success from 20.4% (source reference only) to 97.8%. On a real Unitree G1, policies execute seven contact-rich manipulation and two loco-manipulation tasks.
Abstract:Contact transitions allow humanoid robots to redistribute object support across the whole body, enabling manipulation beyond hand-only grasps. Human videos provide a scalable source of such transitions for robots to learn from. However, body occlusion disrupts object pose tracking and corrupts retargeted motion. Moreover, accurately reproducing the object trajectory does not ensure that the robot applies stabilizing forces at the intended contacts, risking slips or drops in later phases. To address these challenges, we propose ReContact, which reconstructs and retargets human-object interactions through an object-centric contact plan. The plan specifies contact phases, contact pairs, anchors, pressing directions, and sticking or sliding modes. Contact priors stabilize object pose reconstruction through occlusion. A residual policy over a frozen whole-body tracker is rewarded for making contact at planned anchors with forces aligned to the specified directions. In simulation, on a thigh-supported regrasp task, ReContact achieves the planned two-palm bottom hold in 83.5 of episodes over three training seeds, versus 0.16% for the best adapted baseline. One policy trained on 247 references re-solved for new placements raises in-workspace tube-transfer success from 20.4% (source reference only) to 97.8%. On a real Unitree G1, policies execute seven contact-rich manipulation and two loco-manipulation tasks.
Robotics (cs.RO)
arXiv:2611.02968 [cs.RO]
(or arXiv:2611.02968v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2611.02968 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.
Full text under the publisher's open licence · 290 words
Related coverage
Live Thread Reconstruction for Autonomous Manipulation
arXiv · Oct 23, 2026
Beyond Relative Geometry: Metric-Aware Geometry Perception for Robotics
arXiv · Aug 31, 2026
What Visual Representations Do World Action Models Need?
arXiv · Oct 12, 2026
TransDex: Pre-training Visuo-Tactile Policy with Point Cloud Reconstruction for Dexterous Manipulation of Transparent Objects
arXiv · Jul 23, 2026