mlsanityData to 7 Oct 2026

Advancing Humanoid Locomotion: Mastering Challenging Terrains with Denoising World Model Learning

arXiv 2408.14472 ↗First posted 26 Aug 2024cs.RO cs.AI cs.SYcited by 38

On the map: World models → Humanoid and legged locomotion world models

Abstract (arXiv)

Humanoid robots, with their human-like skeletal structure, are especially suited for tasks in human-centric environments. However, this structure is accompanied by additional challenges in locomotion controller design, especially in complex real-world environments. As a result, existing humanoid robots are limited to relatively simple terrains, either with model-based control or model-free reinforcement learning. In this work, we introduce Denoising World Model Learning (DWL), an end-to-end reinforcement learning framework for humanoid locomotion control, which demonstrates the world's first humanoid robot to master real-world challenging terrains such as snowy and inclined land in the wild, up and down stairs, and extremely uneven terrains. All scenarios run the same learned neural network with zero-shot sim-to-real transfer, indicating the superior robustness and generalization capability of the proposed method.

Key facts. Cited by 38 papers in the map (data through 7 Oct 2026). Placed in Humanoid and legged locomotion world models, one of 11 directions in World models. Builds on 28 of its 43 references that are in the map. First posted on arXiv 26 Aug 2024.

Builds on

28 of 43 references.

References · 28

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2017Proximal Policy Optimization Algorithmscited by 14,441
  3. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,339
  4. 2017Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real Worldcited by 1,401
  5. 2018Learning Latent Dynamics for Planning from Pixelscited by 1,221
  6. 2021Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learningcited by 882
  7. 2020Learning Quadrupedal Locomotion over Challenging Terraincited by 494
  8. 2019Learning agile and dynamic motor skills for legged robotscited by 402
  9. 2021Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learningcited by 395
  10. 2019Learning by Cheatingcited by 361
  11. 2018Sim-to-Real: Learning Agile Locomotion For Quadruped Robotscited by 277
  12. 2017Gate-Variants of Gated Recurrent Unit (GRU) Neural Networkscited by 239
16 more
  1. 2017Asymmetric Actor Critic for Image-Based Robot Learningcited by 202
  2. 2022Legged Locomotion in Challenging Terrains using Egocentric Visioncited by 131
  3. 2018State Representation Learning for Control: An Overviewcited by 125
  4. 2018Neural probabilistic motor primitives for humanoid controlcited by 96
  5. 2021Reinforcement Learning for Robust Parameterized Locomotion Control of Bipedal Robotscited by 95
  6. 2018Masking: A New Perspective of Noisy Supervisioncited by 93
  7. 2021Blind Bipedal Stair Traversal via Sim-to-Real Reinforcement Learningcited by 73
  8. 2020Sim-to-Real Learning of All Common Bipedal Gaits via Periodic Reward Compositioncited by 71
  9. 2022Sparsity Winning Twice: Better Robust Generalization from More Efficient Trainingcited by 41
  10. 2022Adapting Rapid Motor Adaptation for Bipedal Robotscited by 39
  11. 2024Humanoid-Gym: Reinforcement Learning for Humanoid Robot with Zero-Shot Sim2Real Transfercited by 35
  12. 2021A Unified MPC Framework for Whole-Body Dynamic Locomotion and Manipulationcited by 33
  13. 2021The MIT Humanoid Robot: Design, Motion Planning, and Control For Acrobatic Behaviorscited by 28
  14. 2020Learning natural locomotion behaviors for humanoid robots using human knowledgecited by 9
  15. 2019Dynamic Walking with Compliance on a Cassie Bipedal Robotcited by 4
  16. 2022Multi-contact MPC for Dynamic Loco-manipulation on Humanoid Robotscited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 3; 2025: 25; 2026: 10.

38 papers.

Sort:

By area:

  1. 2025ASAP: Aligning Simulation and Real-World Physics for Learning Agile Humanoid Whole-Body Skillscited by 104
  2. 2025TWIST: Teleoperated Whole-Body Imitation Systemcited by 86
  3. 2025HOMIE: Humanoid Loco-Manipulation with Isomorphic Exoskeleton Cockpitcited by 65
  4. 2025A Comprehensive Survey on World Models for Embodied AIcited by 65
  5. 2025A Survey: Learning Embodied Intelligence from Physical Simulators and World Modelscited by 50
  6. 2024Generalizable Humanoid Manipulation with 3D Diffusion Policiescited by 49
  7. 2025Residual Off-Policy RL for Finetuning Behavior Cloning Policiescited by 42
  8. 2025BeamDojo: Learning Agile Humanoid Locomotion on Sparse Footholdscited by 32
  9. 2024Learning from Massive Human Videos for Universal Humanoid Pose Controlcited by 27
  10. 2024Learning Smooth Humanoid Locomotion through Lipschitz-Constrained Policiescited by 26
  11. 2025Learning Getting-Up Policies for Real-World Humanoid Robotscited by 21
  12. 2025HuB: Learning Extreme Humanoid Balancecited by 21
26 more
  1. 2025A Unified and General Humanoid Whole-Body Controller for Versatile Locomotioncited by 10
  2. 2025Learning Humanoid Locomotion with World Model Reconstructioncited by 10
  3. 2025Physical Plausibility-aware Trajectory Prediction via Locomotion Embodimentcited by 10
  4. 2025SkillBlender: Towards Versatile Humanoid Whole-Body Loco-Manipulation via Skill Blendingcited by 9
  5. 2025Humanoid Whole-Body Locomotion on Narrow Terrain via Dynamic Balance and Reinforcement Learningcited by 7
  6. 2025Towards Adaptable Humanoid Control via Adaptive Motion Trackingcited by 6
  7. 2025StyleLoco: Generative Adversarial Distillation for Natural Humanoid Robot Locomotioncited by 5
  8. 2025DreamPolicy: A Unified World-model Policy for Scalable Humanoid Locomotioncited by 5
  9. 2025Multi-Loco: Unifying Multi-Embodiment Legged Locomotion via Reinforcement Learning Augmented Diffusioncited by 5
  10. 2025From Generated Human Videos to Physically Plausible Robot Trajectoriescited by 5
  11. 2025Let Humanoids Hike! Integrative Skill Development on Complex Trailscited by 4
  12. 2025Coordinated Humanoid Manipulation with Choice Policiescited by 4
  13. 2026Overcoming Statistical Bias in Action-Controllable World Modelscited by 4
  14. 2026TTT-Parkour: Rapid Test-Time Training for Perceptive Robot Parkourcited by 3
  15. 2026HAIC: Humanoid Agile Object Interaction Control via Dynamics-Aware World Modelcited by 3
  16. 2026SplitAdapter: Load-Aware Humanoid Loco-Manipulation via Factorized Adaptationcited by 2
  17. 2026ARB4WM: An Adversarial Robustness Benchmark for World Models in Continuous Controlcited by 2
  18. 2025Model-based controller assisted domain randomization for transient vibration suppression of nonlinear powertrain system with parametric uncertaintycited by 1
  19. 2025MASH: Cooperative-Heterogeneous Multi-Agent Reinforcement Learning for Single Humanoid Robot Locomotioncited by 1
  20. 2026DAWN: Noise-Robust Quadruped Parkour via Depth-Denoising World Modelscited by 1
  21. 2025Teacher Motion Priors: Enhancing Robot Locomotion over Challenging Terraincited by 0
  22. 2025Architecture Is All You Need: Diversity-Enabled Sweet Spots for Robust Humanoid Locomotioncited by 0
  23. 2026Sim-to-Real of Humanoid Locomotion Policies via Joint Torque Space Perturbation Injectioncited by 0
  24. 2026Learning Task-Invariant Properties via Dreamer: Enabling Efficient Policy Transfer for Quadruped Robotscited by 0
  25. 2026GaitSpan: Growing Humanoid Locomotion from Walking to Runningcited by 0
  26. 2026MASkillBlender: Decentralized Whole-Body Coordination for Multi-Humanoid Loco-Manipulation via Skill Blendingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Learning Perceptive Humanoid Locomotion over Challenging Terrain
  2. 2026Learning Whole-Body Humanoid Locomotion via Motion Generation and Motion Tracking
  3. 2024BiRoDiff: Diffusion policies for bipedal robot locomotion on unseen terrains
  4. 2025GenTe: Generative Real-world Terrains for General Legged Robot Locomotion Control
  5. 2026World-Model-Augmented Visual Locomotion for Humanoids on Foothold-Constrained Terrain
  6. 2026CAP: Continuously Adaptive Perception-Blind Humanoid Locomotion via Learned Denoising
  7. 2025Learning Terrain-Specialized Policies for Adaptive Locomotion in Challenging Environments
  8. 2026Towards Bridging the Gap between Large-Scale Pretraining and Efficient Finetuning for Humanoid Control