SHOES: Sim-to-real Hardware-adaptation via On-policy Expansion of Skills for Legged Robots
SHOES adapts a frozen sim-trained quadruped policy to footwear by fine-tuning only small identity-initialized modulation modules on the robot, with a workspace-bounded sampler; Go2 tested with shoes.SHOES 冻结仿真训练的四足策略，仅在真机上微调以恒等初始化的轻量调制模块来适应穿鞋带来的形态变化，并用限定工作空间的采样器保证安全；在穿鞋的 Go2 上测试。
BibTeX
@misc{tsuruhara2026shoes,
title = {{SHOES: Sim-to-real Hardware-adaptation via On-policy Expansion of Skills for Legged Robots}},
author = {Kota Tsuruhara and Tsung-Yuan Chen and Kento Kawaharazuka},
year = {2026},
eprint = {2610.16457},
archivePrefix = {arXiv},
primaryClass = {cs.RO},
url = {https://arxiv.org/abs/2610.16457},
}