SLR: Learning Quadruped Locomotion without Privileged Information

Shiyi Chen, Zeyu Wan, Shiyang Yan, Chun Zhang, Weiyi Zhang, Qiang Li, Debing Zhang, Fasih Ud Din Farrukh
Proceedings of The 8th Conference on Robot Learning, PMLR 270:3212-3224, 2025.

Abstract

Traditional reinforcement learning control for quadruped robots often relies on privileged information, demanding meticulous selection and precise estimation, thereby imposing constraints on the development process. This work proposes a Self-learning Latent Representation (SLR) method, which achieves high-performance control policy learning without the need for privileged information. To enhance the credibility of our proposed method’s evaluation, SLR is compared with open-source code repositories of state-of-the-art algorithms, retaining the original authors’ configuration parameters. Across four repositories, SLR consistently outperforms the reference results. Ultimately, the trained policy and encoder empower the quadruped robot to navigate steps, climb stairs, ascend rocks, and traverse various challenging terrains.

Cite this Paper

BibTeX
@InProceedings{pmlr-v270-chen25e, title = {SLR: Learning Quadruped Locomotion without Privileged Information}, author = {Chen, Shiyi and Wan, Zeyu and Yan, Shiyang and Zhang, Chun and Zhang, Weiyi and Li, Qiang and Zhang, Debing and Farrukh, Fasih Ud Din}, booktitle = {Proceedings of The 8th Conference on Robot Learning}, pages = {3212--3224}, 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/chen25e/chen25e.pdf}, url = {https://proceedings.mlr.press/v270/chen25e.html}, abstract = {Traditional reinforcement learning control for quadruped robots often relies on privileged information, demanding meticulous selection and precise estimation, thereby imposing constraints on the development process. This work proposes a Self-learning Latent Representation (SLR) method, which achieves high-performance control policy learning without the need for privileged information. To enhance the credibility of our proposed method’s evaluation, SLR is compared with open-source code repositories of state-of-the-art algorithms, retaining the original authors’ configuration parameters. Across four repositories, SLR consistently outperforms the reference results. Ultimately, the trained policy and encoder empower the quadruped robot to navigate steps, climb stairs, ascend rocks, and traverse various challenging terrains.} }
Endnote
%0 Conference Paper %T SLR: Learning Quadruped Locomotion without Privileged Information %A Shiyi Chen %A Zeyu Wan %A Shiyang Yan %A Chun Zhang %A Weiyi Zhang %A Qiang Li %A Debing Zhang %A Fasih Ud Din Farrukh %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-chen25e %I PMLR %P 3212--3224 %U https://proceedings.mlr.press/v270/chen25e.html %V 270 %X Traditional reinforcement learning control for quadruped robots often relies on privileged information, demanding meticulous selection and precise estimation, thereby imposing constraints on the development process. This work proposes a Self-learning Latent Representation (SLR) method, which achieves high-performance control policy learning without the need for privileged information. To enhance the credibility of our proposed method’s evaluation, SLR is compared with open-source code repositories of state-of-the-art algorithms, retaining the original authors’ configuration parameters. Across four repositories, SLR consistently outperforms the reference results. Ultimately, the trained policy and encoder empower the quadruped robot to navigate steps, climb stairs, ascend rocks, and traverse various challenging terrains.
APA
Chen, S., Wan, Z., Yan, S., Zhang, C., Zhang, W., Li, Q., Zhang, D. & Farrukh, F.U.D.. (2025). SLR: Learning Quadruped Locomotion without Privileged Information. Proceedings of The 8th Conference on Robot Learning, in Proceedings of Machine Learning Research 270:3212-3224 Available from https://proceedings.mlr.press/v270/chen25e.html.

Related Material