Adapting Humanoid Locomotion over Challenging Terrain via Two-Phase Training

Wenhao Cui, Shengtao Li, Huaxing Huang, Bangyu Qin, Tianchu Zhang, hanjincha hanjinchao, Liang Zheng, Ziyang Tang, Chenxu Hu, NING Yan, Jiahao Chen, Zheyuan Jiang
Proceedings of The 8th Conference on Robot Learning, PMLR 270:3439-3453, 2025.

Abstract

Humanoid robots are a key focus in robotics, with their capacity to navigate tough terrains being essential for many uses. While strides have been made, creating adaptable locomotion for complex environments is still tough. Recent progress in learning-based systems offers hope for robust legged locomotion, but challenges persist, such as tracking accuracy at high speeds and on uneven ground, and joint oscillations in actual robots. This paper proposes a novel training framework to address these challenges by employing a two-phase training paradigm with reinforcement learning. The proposed framework is further enhanced through the integration of command curriculum learning, refining the precision and adaptability of our approach. Additionally, we adapt DreamWaQ to our humanoid locomotion system and improve it to mitigate joint oscillations. Finally, we achieve the sim-to-real transfer of our method. A series of empirical results demonstrate the superior performance of our proposed method compared to state-of-the-art methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v270-cui25a, title = {Adapting Humanoid Locomotion over Challenging Terrain via Two-Phase Training}, author = {Cui, Wenhao and Li, Shengtao and Huang, Huaxing and Qin, Bangyu and Zhang, Tianchu and hanjinchao, hanjincha and Zheng, Liang and Tang, Ziyang and Hu, Chenxu and Yan, NING and Chen, Jiahao and Jiang, Zheyuan}, booktitle = {Proceedings of The 8th Conference on Robot Learning}, pages = {3439--3453}, year = {2025}, editor = {Agrawal, Pulkit and Kroemer, Oliver and Burgard, Wolfram}, volume = {270}, series = {Proceedings of Machine Learning Research}, month = {06--09 Nov}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v270/main/assets/cui25a/cui25a.pdf}, url = {https://proceedings.mlr.press/v270/cui25a.html}, abstract = {Humanoid robots are a key focus in robotics, with their capacity to navigate tough terrains being essential for many uses. While strides have been made, creating adaptable locomotion for complex environments is still tough. Recent progress in learning-based systems offers hope for robust legged locomotion, but challenges persist, such as tracking accuracy at high speeds and on uneven ground, and joint oscillations in actual robots. This paper proposes a novel training framework to address these challenges by employing a two-phase training paradigm with reinforcement learning. The proposed framework is further enhanced through the integration of command curriculum learning, refining the precision and adaptability of our approach. Additionally, we adapt DreamWaQ to our humanoid locomotion system and improve it to mitigate joint oscillations. Finally, we achieve the sim-to-real transfer of our method. A series of empirical results demonstrate the superior performance of our proposed method compared to state-of-the-art methods.} }
Endnote
%0 Conference Paper %T Adapting Humanoid Locomotion over Challenging Terrain via Two-Phase Training %A Wenhao Cui %A Shengtao Li %A Huaxing Huang %A Bangyu Qin %A Tianchu Zhang %A hanjincha hanjinchao %A Liang Zheng %A Ziyang Tang %A Chenxu Hu %A NING Yan %A Jiahao Chen %A Zheyuan Jiang %B Proceedings of The 8th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2025 %E Pulkit Agrawal %E Oliver Kroemer %E Wolfram Burgard %F pmlr-v270-cui25a %I PMLR %P 3439--3453 %U https://proceedings.mlr.press/v270/cui25a.html %V 270 %X Humanoid robots are a key focus in robotics, with their capacity to navigate tough terrains being essential for many uses. While strides have been made, creating adaptable locomotion for complex environments is still tough. Recent progress in learning-based systems offers hope for robust legged locomotion, but challenges persist, such as tracking accuracy at high speeds and on uneven ground, and joint oscillations in actual robots. This paper proposes a novel training framework to address these challenges by employing a two-phase training paradigm with reinforcement learning. The proposed framework is further enhanced through the integration of command curriculum learning, refining the precision and adaptability of our approach. Additionally, we adapt DreamWaQ to our humanoid locomotion system and improve it to mitigate joint oscillations. Finally, we achieve the sim-to-real transfer of our method. A series of empirical results demonstrate the superior performance of our proposed method compared to state-of-the-art methods.
APA
Cui, W., Li, S., Huang, H., Qin, B., Zhang, T., hanjinchao, h., Zheng, L., Tang, Z., Hu, C., Yan, N., Chen, J. & Jiang, Z.. (2025). Adapting Humanoid Locomotion over Challenging Terrain via Two-Phase Training. Proceedings of The 8th Conference on Robot Learning, in Proceedings of Machine Learning Research 270:3439-3453 Available from https://proceedings.mlr.press/v270/cui25a.html.

Related Material