TTT-Parkour: Rapid Test-Time Training for
Perceptive Robot Parkour

Shaoting Zhu*    Baijun Ye*    Jiaxuan Wang†    Jiakang Chen†    Ziwen Zhuang    Linzhan Mou    Runhan Huang    Hang Zhao
* Equal contribution  † Equal contribution  ‡ Corresponding author

Boston Terrain

Boxes Terrain

Wedge Terrain

Narrow Bridge Terrains

Stake Terrains

Trapezoid Terrains

Mixed Terrains

Abstract

Achieving highly dynamic humanoid parkour on unseen, complex terrains remains a challenge in robotics. Although general locomotion policies demonstrate capabilities across broad terrain distributions, they often struggle with arbitrary and highly challenging environments. To overcome this limitation, we propose a real-to-sim-to-real framework that leverages rapid test-time training (TTT) on novel terrains, significantly enhancing the robot's capability to traverse extremely difficult geometries. We adopt a two-stage end-to-end learning paradigm: a policy is first pre-trained on diverse procedurally generated terrains, followed by rapid fine-tuning on high-fidelity meshes reconstructed from real-world captures. Specifically, we develop a feed-forward, efficient, and high-fidelity geometry reconstruction pipeline using RGB-D inputs, ensuring both speed and quality during test-time training. We demonstrate that TTT-Parkour empowers humanoid robots to master complex obstacles, including wedges, stakes, boxes, trapezoids, and narrow beams. The whole pipeline of capturing, reconstructing, and test-time training requires less than 10 minutes on most tested terrains. Extensive experiments show that the policy after test-time training exhibits robust zero-shot sim-to-real transfer capability.

BibTeX

@misc{zhu2026tttparkourrapidtesttimetraining,
      title={TTT-Parkour: Rapid Test-Time Training for Perceptive Robot Parkour}, 
      author={Shaoting Zhu and Baijun Ye and Jiaxuan Wang and Jiakang Chen and Ziwen Zhuang and Linzhan Mou and Runhan Huang and Hang Zhao},
      year={2026},
      eprint={2602.02331},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2602.02331}, 
}