Awesome-Legged-Robot-Learning

A collection of awesome Legged Robot Learning papers. From Reinforcement Learning and Sim-to-Real to Humanoids.

---

Acknowledgements

This repository is inspired by and references the following excellent paper lists. Special thanks to the authors for their contributions to the community:

• awesome-humanoid-robot-learning

• awesome-rl-for-legged-locomotion

---

Locomotion

• [arXiv 21.09] Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learning
• [arXiv 22.01] Learning Robust Perceptive Locomotion for Quadrupedal Robots in the Wild
• [arXiv 22.02] Concurrent Training of a Control Policy and a State Estimator for Dynamic and Robust Legged Locomotion
• [arXiv 23.01] DreamWaQ: Learning Robust Quadrupedal Locomotion With Implicit Terrain Imagination via Deep Reinforcement Learning
• [arXiv 23.04] Learning Robust and Agile Legged Locomotion Using Adversarial Motion Priors
• [arXiv 23.06] ANYmal Parkour: Learning Agile Navigation for Quadrupedal Robots
• [arXiv 23.09] Robot Parkour Learning
• [arXiv 23.09] Extreme Parkour with Legged Robots
• [arXiv 24.05] CTS: Concurrent Teacher-Student Reinforcement Learning for Legged Locomotion
• [arXiv 24.06] Humanoid Parkour Learning
• [arXiv 24.08] PIE: Parkour with Implicit-Explicit Learning Framework for Legged Robots
• [arXiv 25.01] Robotic World Model: A Neural Network Simulator for Robust Policy Optimization in Robotics
• [arXiv 25.06] Attention-Based Map Encoding for Learning Generalized Legged Locomotion
• [arXiv 25.06] Multi-Loco: Unifying Multi-Embodiment Legged Locomotion via Reinforcement Learning Augmented Diffusion
• [arXiv 25.09] LocoFormer: Generalist Locomotion via Long-Context Adaptation
• [arXiv 25.09] LIPM-Guided Reinforcement Learning for Stable and Perceptive Locomotion in Bipedal Robots

---

Loco-Manipulation and Whole-Body-Control

• [arXiv 24.02] Expressive Whole-Body Control for Humanoid Robots
• [arXiv 24.03] Learning Human-to-Humanoid Real-Time Whole-Body Teleoperation
• [arXiv 24.06] OmniH2O: Universal and Dexterous Human-to-Humanoid Whole-Body Teleoperation and Learning
• [arXiv 24.06] HumanPlus: Humanoid Shadowing and Imitation from Humans
• [arXiv 24.10] HOVER: Versatile Neural Whole-Body Controller for Humanoid Robots
• [arXiv 24.12] Learning Whole-Body Loco-Manipulation for Omni-Directional Task Space Pose Tracking with a Wheeled-Quadrupedal-Manipulator
• [arXiv 24.12] Mobile-TeleVision: Predictive Motion Priors for Humanoid Whole-Body Control
• [arXiv 24.12] ExBody2: Advanced Expressive Humanoid Whole-Body Control
• [arXiv 25.02] ASAP: Aligning Simulation and Real-World Physics for Learning Agile Humanoid Whole-Body Skills
• [arXiv 25.02] Embrace Collisions: Humanoid Shadowing for Deployable Contact-Agnostics Motions
• [arXiv 25.02] HugWBC: A Unified and General Humanoid Whole-Body Controller for Fine-Grained Locomotion
• [arXiv 25.02] HOMIE: Humanoid Loco-Manipulation with Isomorphic Exoskeleton Cockpit
• [arXiv 25.03] GR00T N1: An Open Foundation Model for Generalist Humanoid Robots
• [arXiv 25.04] LangWBC: Language-directed Humanoid Whole-Body Control via End-to-end Learning
• [arXiv 25.05] Learning coordinated badminton skills for legged manipulators
• [arXiv 25.05] TWIST: Teleoperated Whole-Body Imitation System
• [arXiv 25.06] KungfuBot: Physics-Based Humanoid Whole-Body Control for Learning Highly-Dynamic Skills
• [arXiv 25.06] General Motion Tracking for Humanoid Whole-Body Control
• [arXiv 25.06] LeVERB: Humanoid Whole-Body Control with Latent Vision-Language Instruction
• [arXiv 25.08] BeyondMimic: From Motion Tracking to Versatile Humanoid Control via Guided Diffusion
• [arXiv 25.09] KungfuBot 2: Learning Versatile Motion Skills for Humanoid Whole-Body Control
• [arXiv 25.10] Retargeting Matters: General Motion Retargeting for Humanoid Motion Tracking
• [arXiv 25.11] SONIC: Supersizing Motion Tracking for Natural Humanoid Whole-Body Control

---

Physics-Based Character Animation

• [arXiv 18.04] DeepMimic: Example-Guided Deep Reinforcement Learning of Physics-Based Character Skills
• [arXiv 21.04] AMP: Adversarial Motion Priors for Stylized Physics-Based Character Control
• [arXiv 23.05] Perpetual Humanoid Control for Real-time Simulated Avatars
• [arXiv 23.09] Unified Human-Scene Interaction via Prompted Chain-of-Contacts
• [arXiv 24.08] SkillMimic: Learning Basketball Interaction Skills from Demonstrations
• [arXiv 24.09] MaskedMimic: Unified Physics-Based Character Control Through Masked Motion Inpainting
• [arXiv 24.10] CLoSD: Closing the Loop between Simulation and Diffusion for multi-task character control
• [arXiv 25.03] TokenHSI: Unified Synthesis of Physical Human-Scene Interactions through Task Tokenization
• [arXiv 25.04] Zero-Shot Whole-Body Humanoid Control via Behavioral Foundation Models
• [arXiv 25.07] Feature-Based vs. GAN-Based Learning from Demonstrations: When and Why
• [arXiv 25.09] Learning to Ball: Composing Policies for Long-Horizon Basketball Moves