aiqbench analysis
This research presents OTRetarget, a novel method for transferring human motion to humanoid robots while preserving complex interactions with the environment and manipulated objects. By utilizing entropic optimal transport and signed distance representations, the approach jointly optimizes robot kinematics and object trajectories without requiring scene rescaling. The study demonstrates significant improvements in interaction fidelity on simulation and physical G1 humanoid platforms, advancing the state of loco-manipulation capabilities.
OTRetarget: Joint Robot and Object Motion Retargeting via Optimal Transport
arXiv:2609.36602v1 Announce Type: new Abstract: Transferring human motion to humanoid robots requires adapting the demonstrated motion to the robot morphology while preserving interactions with the environment. This is particularly challenging for loco-manipulation tasks, where contacts with the ground and manipulated objects must remain consistent despite differences in body proportions. Yet, skeletal motion alone does not fully describe these interactions, and fixing object trajectories limits the adaptation to a new embodiment. In this paper, we introduce OTR ETARGET, a unified approach to jointly retarget robot and multi-object motion from human demonstrations. Our approach represents surface interactions through signed distances, closest surface points, and relative directions, and uses entropic optimal transport to transfer these quantities across human, robot, and object geometries. We incorporate the resulting interaction targets into a constrained inverse kinematics formulation that balances contact preservation with motion style and jointly optimizes robot and object poses at each frame. This formulation accommodates robot-object and object-object interactions without rescaling the scene or the demonstration. We validate the proposed approach on OMOMO, where it achieves a robot- object interaction Jaccard score of 87% and a depth error of 8.7 mm, compared with 28% and 29.3 mm for OmniRetarget. Finally, we demonstrate transfer to a physical G1 humanoid using whole-body policies trained with reinforcement learning on the retargeted references, across motions including two-handed box pick-and-place onto a table.
Abstract:Transferring human motion to humanoid robots requires adapting the demonstrated motion to the robot morphology while preserving interactions with the environment. This is particularly challenging for loco-manipulation tasks, where contacts with the ground and manipulated objects must remain consistent despite differences in body proportions. Yet, skeletal motion alone does not fully describe these interactions, and fixing object trajectories limits the adaptation to a new embodiment. In this paper, we introduce OTR ETARGET, a unified approach to jointly retarget robot and multi-object motion from human demonstrations. Our approach represents surface interactions through signed distances, closest surface points, and relative directions, and uses entropic optimal transport to transfer these quantities across human, robot, and object geometries. We incorporate the resulting interaction targets into a constrained inverse kinematics formulation that balances contact preservation with motion style and jointly optimizes robot and object poses at each frame. This formulation accommodates robot-object and object-object interactions without rescaling the scene or the demonstration. We validate the proposed approach on OMOMO, where it achieves a robot- object interaction Jaccard score of 87% and a depth error of 8.7 mm, compared with 28% and 29.3 mm for OmniRetarget. Finally, we demonstrate transfer to a physical G1 humanoid using whole-body policies trained with reinforcement learning on the retargeted references, across motions including two-handed box pick-and-place onto a table.
Robotics (cs.RO)
arXiv:2609.36602 [cs.RO]
(or arXiv:2609.36602v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.36602 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 · 312 words
Related coverage
Hand2Robot: Object-Centric 4D Hand Reconstruction from Human Demonstrations for Robotics
arXiv · Oct 14, 2026
Robust Nonprehensile Object Transport with Quadruped Robots
arXiv · Oct 7, 2026
DexTaG: Tactile-as-Guidance in Reinforcement Learning for Dexterous Manipulation
arXiv · Sep 29, 2026
From Video to Vision: Learning Generalizable Humanoid Carrying from Human Motion
arXiv · Oct 14, 2026