SMR: Learning streaming motion retargeting for humanoids from human semantic
SMR distils physics experts that track robot-aware targets into a strictly causal Transformer, retargeting human motion to humanoids frame by frame. Reports better closed-loop results than kinematic baselines.SMR 先让物理专家在仿真中跟踪由人体动作转换来的机器人任务空间目标，再用其执行轨迹训练严格因果的 Transformer，实现逐帧流式的人形动作重定向；报告在相同全身控制器下闭环表现优于运动学与物理精修基线。
BibTeX
@misc{gu2026smr,
title = {{SMR: Learning streaming motion retargeting for humanoids from human semantic}},
author = {Chun Gu and Juntong Zhang and Jingyu Li and Li Zhang},
year = {2026},
eprint = {2610.29169},
archivePrefix = {arXiv},
primaryClass = {cs.RO},
url = {https://arxiv.org/abs/2610.29169},
}Locomotion & Humanoids运动控制与人形机器人Planning, Control & Safety规划、控制与安全Humanoid人形humanoidmotion-retargetingstreamingtransformerwhole-body-controlsim