DAWN: Noise-Robust Quadruped Parkour via Depth-Denoising World Models

2026-09-24

AI Summary

This study addresses the reliance of existing vision-based quadruped robots on manual filtering for depth noise, which hinders reproducibility and limits performance. To overcome this, we propose DAWN, a framework that achieves filter-free robust perception via world models. Methodologically, DAWN leverages implicit denoising reconstruction and contrastive learning to align latent states with end-to-end reinforcement learning, eliminating the need for explicit noise modeling or manual tuning. This enables locomotion control directly from raw depth maps with zero inference overhead. Experimental results demonstrate that DAWN achieves zero-shot parkour on a Unitree Go1 robot, successfully traversing 18 cm steps, 70 cm gaps, and 45 cm platforms, thereby validating its robustness across complex terrains.

Abstract

Vision-based legged locomotion methods assume clean depth at training time and rely on hand-tuned post-processing filters at deployment. However, filter parameters are rarely disclosed, hindering reproducibility, and performance degrades substantially when depth noise is left unaddressed. Building noise robustness directly into the learning pipeline would eliminate this dependency. While such robustness has been explored for proprioceptive inputs, analogous approaches for depth perception remain largely absent in legged locomotion. We propose DAWN (Denoising and Alignment in World models for Noise-robustness), a noise-robust perception framework for legged locomotion, which builds noise robustness directly into a world model via two modifications: (1) feeding noisy depth to the encoder while keeping clean depth as the reconstruction target, forcing the model to implicitly denoise its input; and (2) applying contrastive learning to align the latent states of noisy and clean depth. Importantly, DAWN is not tied to a specific noise model, requiring no manual tuning to the noise distribution at deployment. Furthermore, it incurs no additional inference cost over existing world model-based methods. Without any manual filter calibration -- relying solely on the learned noise-robust representation -- DAWN achieves zero-shot quadruped parkour on a Unitree Go1: traversing stairs up to 18 cm, clearing gaps up to 70 cm, and mounting steps up to 45 cm from raw depth observations. Ablation studies show that denoising and contrastive alignment contribute at complementary levels -- reconstruction and representation, respectively -- and yield additive gains when combined. Videos and code are available at: https://dawn-parkour.github.io/

Problem

• legged locomotion
• depth noise robustness
• quadruped parkour
• world models
• visual perception

Innovation

• World Models
• Depth Denoising
• Contrastive Learning
• Noise-Robustness
• Quadruped Parkour