aiqbench analysis
This research advances humanoid robot locomotion by integrating cross-attention transformers with synthetic depth synthesis to bridge the sim-to-real gap in terrain perception. By leveraging pre-trained elevation maps as a blind backbone, the framework reduces hardware latency and training data requirements while maintaining robustness on complex terrains. The approach represents a significant methodological shift for embodied AI systems aiming to operate efficiently without relying on dense multi-sensor suites.
DPL: Depth-only Perceptive Humanoid Locomotion via Realistic Depth Synthesis and Cross-Attention Terrain Reconstruction
arXiv:2510.07152v3 Announce Type: replace Abstract: Recent advancements in legged robot perceptive locomotion have shown promising progress. However, terrain-aware humanoid locomotion remains largely constrained to two paradigms: depth image-based end-to-end learning and elevation map-based methods. The former suffers from limited training efficiency and a significant sim-to-real gap in depth perception, while the latter depends heavily on multiple vision sensors and localization systems, resulting in latency and reduced robustness. To overcome these challenges, we propose a novel framework that tightly integrates three key components: (1) Terrain-Aware Locomotion Policy with a Blind Backbone, which leverages pre-trained elevation map-based perception to guide reinforcement learning with minimal visual input; (2) Multi-Modality Cross-Attention Transformer, which reconstructs structured terrain representations from noisy depth images; (3) Realistic Depth Images Synthetic Method, which employs self-occlusion-aware ray casting and noise-aware modeling to synthesize realistic depth observations, achieving over 30\% reduction in terrain reconstruction error. This combination enables efficient policy training with limited data and hardware resources, while preserving critical terrain features essential for generalization. We validate our framework on a full-sized humanoid robot, demonstrating agile and adaptive locomotion across diverse and challenging terrains.
Abstract:Recent advancements in legged robot perceptive locomotion have shown promising progress. However, terrain-aware humanoid locomotion remains largely constrained to two paradigms: depth image-based end-to-end learning and elevation map-based methods. The former suffers from limited training efficiency and a significant sim-to-real gap in depth perception, while the latter depends heavily on multiple vision sensors and localization systems, resulting in latency and reduced robustness. To overcome these challenges, we propose a novel framework that tightly integrates three key components: (1) Terrain-Aware Locomotion Policy with a Blind Backbone, which leverages pre-trained elevation map-based perception to guide reinforcement learning with minimal visual input; (2) Multi-Modality Cross-Attention Transformer, which reconstructs structured terrain representations from noisy depth images; (3) Realistic Depth Images Synthetic Method, which employs self-occlusion-aware ray casting and noise-aware modeling to synthesize realistic depth observations, achieving over 30\% reduction in terrain reconstruction error. This combination enables efficient policy training with limited data and hardware resources, while preserving critical terrain features essential for generalization. We validate our framework on a full-sized humanoid robot, demonstrating agile and adaptive locomotion across diverse and challenging terrains.
8 pages, 8 figures. Published in IEEE Robotics and Automation Letters (RA-L), 2026
Robotics (cs.RO)
arXiv:2510.07152 [cs.RO]
(or arXiv:2510.07152v3 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2510.07152 Focus to learn more arXiv-issued DOI via DataCite
IEEE Robotics and Automation Letters, vol. 11, no. 8, pp. 9407-9414, 2026
https://doi.org/10.1109/LRA.2026.3703584 Focus to learn more DOI(s) linking to related resources
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 304 words
Related coverage
RPL: Learning Robust Humanoid Perceptive Locomotion on Challenging Terrains
arXiv · Aug 25, 2026
NVIDIA GTC Taipei at COMPUTEX: Live Updates on What’s Next in AI
NVIDIA · May 21, 2026
CAP: Continuously Adaptive Perception-Blind Humanoid Locomotion via Learned Denoising
arXiv · Sep 11, 2026
CReF: Cross-modal and Recurrent Fusion for Depth-conditioned Humanoid Locomotion
arXiv · Jul 28, 2026