analiza aiqbench
This research introduces Pose Flow Matching (PFM-HR), a novel prior trained on unordered pose data to enhance reinforcement learning for humanoid robot tracking. By leveraging the new Pose Geometry Score, the method effectively guides policy exploration toward structured motion changes without requiring ordered clips or freezing priors across tasks. The approach demonstrates significant improvements in handling highly dynamic motions, advancing the field of robotics control through advanced generative modeling techniques.
PFM-HR: Pose Flow Matching for Humanoid Robots
arXiv:2608.03227v1 Announce Type: new Abstract: Motion priors improve reinforcement learning for physics-based humanoid tracking, but temporal priors require ordered motion clips, while pose priors provide limited guidance for policy-induced pose transitions. We present Pose Flow Matching for Humanoid Robots (PFM-HR), a reusable flow matching prior trained directly on large scale unordered pose data. PFM-HR introduces the Pose Geometry Score (PGS), which quantifies how joint coordinate changes during rollouts align with the local geometry of pose variation captured by the prior. Using PGS to modulate the tracking reward guides policy exploration toward structured pose changes while keeping the prior frozen across tracking tasks. Experiments demonstrate that PFM-HR improves both single motion and general motion tracking, especially for highly dynamic motions.
Abstract:Motion priors improve reinforcement learning for physics-based humanoid tracking, but temporal priors require ordered motion clips, while pose priors provide limited guidance for policy-induced pose transitions. We present Pose Flow Matching for Humanoid Robots (PFM-HR), a reusable flow matching prior trained directly on large scale unordered pose data. PFM-HR introduces the Pose Geometry Score (PGS), which quantifies how joint coordinate changes during rollouts align with the local geometry of pose variation captured by the prior. Using PGS to modulate the tracking reward guides policy exploration toward structured pose changes while keeping the prior frozen across tracking tasks. Experiments demonstrate that PFM-HR improves both single motion and general motion tracking, especially for highly dynamic motions.
Robotics (cs.RO)
arXiv:2608.03227 [cs.RO]
(or arXiv:2608.03227v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2608.03227 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 · 207 cuvinte
Subiecte conexe
BRACE: Adapting Whole-Body References for Force and Terrain Aware Humanoid Motion Tracking
arXiv · 7 oct. 2026
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · 13 oct. 2026
UniLAP: A Unified Latent Action Policy for Cross-Embodiment Manipulation
arXiv · 14 oct. 2026
RoPE-Flow: Monocular Vision-based Robot Pose Estimation via 2D Graph-Conditioned 3D Flow Matching
arXiv · 16 oct. 2026