GOTT is a reach‑acquire‑move framework that uses a single cross‑embodiment contact‑acquisition primitive to enable dexterous manipulation across different robot embodiments. The system first positions the hand near a task‑relevant contact region, then the shared closed‑loop primitive establishes stable contact, after which a pose‑conditioned controller tracks the desired object motion. Experiments in simulation and on real robots demonstrate that GOTT reliably establishes robust contact across diverse objects, arm‑hand platforms, and both seen and unseen hand morphologies, and consistently improves end‑to‑end task success compared to open‑loop grasp execution.
Recent work in humanoid whole-body control has found success with a simple recipe: retarget human motion to robot kinematic references, then train policies via reinforcement learning (RL) to track them. But how does this recipe transfer to dexterous manipulation?
arXiv:2607. 11874v1 Announce Type: cross Abstract: Recent work in humanoid whole-body control has found success with a simple recipe: retarget human motion to robot kinematic references, then train policies via reinforcement learning (RL) to track them.
arXiv:2609.16683v1 Announce Type: cross Abstract: Learning humanoid-object interaction requires coordinating whole-body balance, locomotion, and dexterous hand contact to control both robot and objec...
Learning humanoid-object interaction requires coordinating whole-body balance, locomotion, and dexterous hand contact to control both robot and object motion. Human demonstrations provide examples of...
arXiv:2609.24660v2 Announce Type: cross Abstract: Human demonstrations offer a scalable way to collect manipulation data, but their contacts may be unstable or infeasible when transferred to a robot...
arXiv:2607. 09218v1 Announce Type: cross Abstract: Whole-arm manipulation involves direct contact with the environment while the robot completes a task by distributing contact across multiple links as contacts form, slide, and break.
Retargeting human object interaction demonstrations to physics based simulation requires reproducing not only body motion but also the object motion and contacts that make manipulation succeed. However, position only hand trajectories do not specify the contact forces needed to manipulate objects, and directly tracking them can overconstrain contact rich finger behavior.
arXiv:2607. 00033v1 Announce Type: cross Abstract: Dexterous robot manipulation can benefit from the abundance of human demonstrations, but transferring such demonstrations to robot policies remains challenging.
UniCross proposes a unified framework that models the four canonical dexterous manipulation skills—grasping, relocation, in-hand rotation, and in-hand translation—within a single formulation sharing state, action spaces, and objective structure. This approach enables distillation of a single cross-skill policy that performs well across all skills, generalizes to unseen objects, remains robust to disturbances, and can chain skills into long-horizon manipulation without policy switching. The framework also transfers effectively across different hand morphologies, suggesting that these skills can be viewed as instantiations of a shared task formulation.
arXiv:2607. 09218v2 Announce Type: replace-cross Abstract: Whole-arm manipulation involves direct contact with the environment while the robot completes a task by distributing contact across multiple links as contacts form, slide, and break.
arXiv:2606. 10614v1 Announce Type: cross Abstract: Robotic foundation models pre-trained on human demonstration videos have shown promise, but a significant embodiment gap remains when the resulting policies are deployed on real robots.
arXiv:2608. 19759v1 Announce Type: cross Abstract: Multifingered grasping is a crucial robotic skill, but current deep-learning grasp planners often struggle to generalize to new objects because they are trained on limited, object-specific datasets.