mlsanityData to 14 Oct 2026

RuN: Residual Policy for Natural Humanoid Locomotion

arXiv 2509.20696 ↗First posted 25 Sep 2025cs.ROcited by 1

On the map: Deep reinforcement learning methods → RL for robot locomotion and manipulation

Abstract (arXiv)

Enabling humanoid robots to achieve natural and dynamic locomotion across a wide range of speeds, including smooth transitions from walking to running, presents a significant challenge. Existing deep reinforcement learning methods typically require the policy to directly track a reference motion, forcing a single policy to simultaneously learn motion imitation, velocity tracking, and stability maintenance. To address this, we introduce RuN, a novel decoupled residual learning framework. RuN decomposes the control task by pairing a pre-trained Conditional Motion Generator, which provides a kinematically natural motion prior, with a reinforcement learning policy that learns a lightweight residual correction to handle dynamical interactions. Experiments in simulation and reality on the Unitree G1 humanoid robot demonstrate that RuN achieves stable, natural gaits and smooth walk-run transitions across a broad velocity range (0-2.5 m/s), outperforming state-of-the-art methods in both training efficiency and final performance.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods. Builds on 22 of its 28 references that are in the map. First posted on arXiv 25 Sep 2025.

Builds on

22 of 28 references.

Closest ancestors · 2

Same direction, newest first

  1. 2024Whole-body Humanoid Robot Locomotion with Human Reference28 Feb
  2. 2024Reinforcement Learning for Versatile, Dynamic, and Robust Bipedal Locomotion Control30 Jan

Other references · 20

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2019AMASS: Archive of Motion Capture as Surface Shapescited by 1,236
  3. 2022Human Motion Diffusion Modelcited by 1,080
  4. 2015Scheduled Sampling for Sequence Prediction with Recurrent Neural Networkscited by 803
  5. 2015Generating Sentences from a Continuous Spacecited by 612
  6. 2023T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representationscited by 588
  7. 2022TEMOS: Generating diverse human motions from textual descriptionscited by 492
  8. 2020Action2Motion: Conditioned Generation of 3D Human Motionscited by 441
  9. 2022EDGE: Editable Dance Generation From Musiccited by 287
  10. 2017Asymmetric Actor Critic for Image-Based Robot Learningcited by 238
  11. 2023Real-World Humanoid Locomotion with Reinforcement Learningcited by 187
  12. 2021Character Controllers Using Motion VAEscited by 184
8 more
  1. 2022HUMANISE: Language-conditioned Human Motion Generation in 3D Scenescited by 158
  2. 2023Learning Agile Soccer Skills for a Bipedal Robot with Deep Reinforcement Learningcited by 133
  3. 2024Humanoid Parkour Learningcited by 96
  4. 2021Blind Bipedal Stair Traversal via Sim-to-Real Reinforcement Learningcited by 92
  5. 2024HOVER: Versatile Neural Whole-Body Controller for Humanoid Robotscited by 91
  6. 2024Humanoid-Gym: Reinforcement Learning for Humanoid Robot with Zero-Shot Sim2Real Transfercited by 53
  7. 2025Natural Humanoid Robot Locomotion with Generative Motion Priorcited by 7
  8. 2025Optimizing Bipedal Locomotion for The 100m Dash With Comparison to Human Runningcited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026SPRINT: Efficient Spectral Priors for Humanoid Athletic Sprintscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Chasing Stability: Humanoid Running via Control Lyapunov Function Guided Reinforcement Learning
  2. 2026Unified Walking, Running, and Recovery for Humanoids via State-Dependent Adversarial Motion Priors
  3. 2025RobotDancing: Residual-Action Reinforcement Learning Enables Robust Long-Horizon Humanoid Motion Tracking
  4. 2026DAMP: Humanoid Locomotion via Denoised Belief Learning and Adversarial Motion Priors ↗
  5. 2025Physics-Guided Residual Reinforcement Learning for Humanoid Narrow-Path Traversal
  6. 2024Achieving Stable High-Speed Locomotion for Humanoid Robots with Deep Reinforcement Learning
  7. 2025Gait-Conditioned Reinforcement Learning with Multi-Phase Curriculum for Humanoid Locomotion
  8. 2026Multi-Gait Learning for Humanoid Robots Using Reinforcement Learning with Selective Adversarial Motion Prior