# World-Model-Augmented Visual Locomotion for Humanoids on Foothold-Constrained Terrain

arXiv: https://arxiv.org/abs/2609.02542 · HTML: https://mlsanity.com/paper/2609.02542/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Humanoid and legged locomotion world models, one of 11 directions in World models.
- Builds on 8 of its 14 references that are in the map.
- First posted on arXiv 2 Sep 2026.

## Placed in

- World models / [Humanoid and legged locomotion world models](https://mlsanity.com/wm/humanoid-and-legged-locomotion-world-models/)

## Builds on (most cited first, 8 of 8)

- [Learning Humanoid Locomotion with Perceptive Internal Model](https://mlsanity.com/paper/2411.14386/) (2024)
- [Learning Humanoid Locomotion with World Model Reconstruction](https://mlsanity.com/paper/2502.16230/) (2025)
- [Hiking in the Wild: A Scalable Perceptive Parkour Framework for Humanoids](https://mlsanity.com/paper/2601.07718/) (2026)
- [RPL: Learning Robust Humanoid Perceptive Locomotion on Challenging Terrains](https://mlsanity.com/paper/2602.03002/) (2026)
- [Walk the PLANC: Physics-Guided RL for Agile Humanoid Locomotion on Constrained Footholds](https://mlsanity.com/paper/2601.06286/) (2026)
- [DreamPolicy: A Unified World-model Policy for Scalable Humanoid Locomotion](https://mlsanity.com/paper/2505.18780/) (2025)
- [START: Traversing Sparse Footholds with Terrain Reconstruction](https://mlsanity.com/paper/2512.13153/) (2025)
- [Perceptive Variable-Timing Footstep Planning for Humanoid Locomotion on Disconnected Footholds](https://mlsanity.com/paper/2603.07400/) (2026)

## Built on by (most cited first, 1 of 1)

- [LocoWM: High-Precision Locomotion through World-Model-Guided Residual Adaptation](https://mlsanity.com/paper/2609.39179/) (2026)

## Abstract (arXiv)

Foothold-constrained terrain is characterized by sparse, discontinuous, or geometrically restricted feasible foot contacts, as encountered on stepping stones, across gaps, and on narrow stair treads. On such terrain, a single misstep often leaves little room to recover, so policies that base foot-placement decisions primarily on the immediately visible terrain are prone to failure. We ask whether a learned predictive summary of near-future observations and rewards can provide the anticipatory information required in such settings. We present World-Model-Augmented Visual Locomotion (WM-LOCO), which jointly trains a recurrent world model and a PPO policy. Conditioned on proprioception and a single onboard depth image, the world model produces a predictive recurrent feature that guides the policy, without explicit foothold labels. In simulation, WM-LOCO succeeds on gaps and stepping stones where a matched baseline fails completely, and matches the baseline's success rate on stairs while improving stride efficiency and reducing pelvis acceleration. We deploy the same policy onboard a physical Unitree G1 humanoid using onboard proprioception and a single depth stream; it traverses all three terrain classes with an average success rate of 93.3%.