aiqbench analysis
DynaRetarget introduces a sampling-based trajectory optimization framework that successfully converts imperfect kinematic motions into dynamically feasible humanoid trajectories for long-horizon tasks. By generalizing across diverse object properties, the method significantly boosts retargeting success rates and enables the generation of large-scale synthetic datasets to address data scarcity in loco-manipulation. This approach offers a robust solution for scaling humanoid robot training without relying on extensive real-world data collection.
DynaRetarget: Dynamically-Feasible Retargeting using Sampling-Based Trajectory Optimization
arXiv:2602.06827v3 Announce Type: replace Abstract: In this paper, we introduce DynaRetarget, a complete pipeline for retargeting human motions to humanoid control policies. The core component of DynaRetarget is a novel Sampling-Based Trajectory Optimization (SBTO) framework that refines imperfect kinematic trajectories into dynamically feasible motions. SBTO incrementally advances the optimization horizon, enabling optimization over the entire trajectory for long-horizon tasks. We validate DynaRetarget by successfully retargeting hundreds of humanoid-object demonstrations and achieving higher success rates than the state of the art. The framework also generalizes across varying object properties, such as mass, size, and geometry, using the same tracking objective. This ability to robustly retarget diverse demonstrations opens the door to generating large-scale synthetic datasets of humanoid loco-manipulation trajectories, addressing a major bottleneck in real-world data collection.
Abstract:In this paper, we introduce DynaRetarget, a complete pipeline for retargeting human motions to humanoid control policies. The core component of DynaRetarget is a novel Sampling-Based Trajectory Optimization (SBTO) framework that refines imperfect kinematic trajectories into dynamically feasible motions. SBTO incrementally advances the optimization horizon, enabling optimization over the entire trajectory for long-horizon tasks. We validate DynaRetarget by successfully retargeting hundreds of humanoid-object demonstrations and achieving higher success rates than the state of the art. The framework also generalizes across varying object properties, such as mass, size, and geometry, using the same tracking objective. This ability to robustly retarget diverse demonstrations opens the door to generating large-scale synthetic datasets of humanoid loco-manipulation trajectories, addressing a major bottleneck in real-world data collection.
Robotics (cs.RO)
arXiv:2602.06827 [cs.RO]
(or arXiv:2602.06827v3 [cs.RO] for this version)
Full text under the publisher's open licence · 131 words