Contrastive Representation Learning for Robust Sim-to-Real Transfer of Adaptive Humanoid Locomotion
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918142122393600 |
|---|---|
| author | Lu, Yidan Yang, Rurui Kou, Qiran Chen, Mengting Fan, Tao Cui, Peter Dong, Yinzhao Lu, Peng |
| author_facet | Lu, Yidan Yang, Rurui Kou, Qiran Chen, Mengting Fan, Tao Cui, Peter Dong, Yinzhao Lu, Peng |
| contents | Reinforcement learning has produced remarkable advances in humanoid locomotion, yet a fundamental dilemma persists for real-world deployment: policies must choose between the robustness of reactive proprioceptive control or the proactivity of complex, fragile perception-driven systems. This paper resolves this dilemma by introducing a paradigm that imbues a purely proprioceptive policy with proactive capabilities, achieving the foresight of perception without its deployment-time costs. Our core contribution is a contrastive learning framework that compels the actor's latent state to encode privileged environmental information from simulation. Crucially, this ``distilled awareness" empowers an adaptive gait clock, allowing the policy to proactively adjust its rhythm based on an inferred understanding of the terrain. This synergy resolves the classic trade-off between rigid, clocked gaits and unstable clock-free policies. We validate our approach with zero-shot sim-to-real transfer to a full-sized humanoid, demonstrating highly robust locomotion over challenging terrains, including 30 cm high steps and 26.5° slopes, proving the effectiveness of our method. Website: https://lu-yidan.github.io/cra-loco. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_12858 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Contrastive Representation Learning for Robust Sim-to-Real Transfer of Adaptive Humanoid Locomotion Lu, Yidan Yang, Rurui Kou, Qiran Chen, Mengting Fan, Tao Cui, Peter Dong, Yinzhao Lu, Peng Robotics Reinforcement learning has produced remarkable advances in humanoid locomotion, yet a fundamental dilemma persists for real-world deployment: policies must choose between the robustness of reactive proprioceptive control or the proactivity of complex, fragile perception-driven systems. This paper resolves this dilemma by introducing a paradigm that imbues a purely proprioceptive policy with proactive capabilities, achieving the foresight of perception without its deployment-time costs. Our core contribution is a contrastive learning framework that compels the actor's latent state to encode privileged environmental information from simulation. Crucially, this ``distilled awareness" empowers an adaptive gait clock, allowing the policy to proactively adjust its rhythm based on an inferred understanding of the terrain. This synergy resolves the classic trade-off between rigid, clocked gaits and unstable clock-free policies. We validate our approach with zero-shot sim-to-real transfer to a full-sized humanoid, demonstrating highly robust locomotion over challenging terrains, including 30 cm high steps and 26.5° slopes, proving the effectiveness of our method. Website: https://lu-yidan.github.io/cra-loco. |
| title | Contrastive Representation Learning for Robust Sim-to-Real Transfer of Adaptive Humanoid Locomotion |
| topic | Robotics |
| url | https://arxiv.org/abs/2509.12858 |