humanoid-robot-learning

A List for Humanoid Robot Learning.

---

Whole Body Control

• [2025.8] MASH: Cooperative-Heterogeneous Multi-Agent Reinforcement Learning for Single Humanoid Robot Locomotion [paper]
• [2025.8] Robot Trains Robot: Automatic Real-World Policy Adaptation and Learning for Humanoids [paper]
• [2025.8] Towards Immersive Human-X Interaction: A Real-Time Framework for Physically Plausible Motion Synthesis [paper]
• [2025.8] BeyondMimic: From Motion Tracking to Versatile Humanoid Control via Guided Diffusion [paper]
• [2025.8] Coordinated Humanoid Robot Locomotion with Symmetry Equivariant Reinforcement Learning Policy [paper]
• [2025.7] EMP: Executable Motion Prior for Humanoid Robot Standing Upper-body Motion Imitation [paper]
• [2025.7] Learning Robust Motion Skills via Critical Adversarial Attacks for Humanoid Robots[paper]
• [2025.7] Learning Humanoid Arm Motion via Centroidal Momentum Regularized Multi-Agent Reinforcement Learning [paper]
• [2025.7] ULC: A Unified and Fine-Grained Controller for Humanoid Loco-Manipulation [paper]
• [2025.6] RL from Physical Feedback: Aligning Large Motion Models with Humanoid Control [paper]
• [2025.6] KungfuBot: Physics-Based Humanoid Whole-Body Control for Learning Highly-Dynamic Skill [paper]
• [2025.6] Attention-Based Map Encoding for Learning Generalized Legged Locomotion [paper]
• [2025.6] SkillBlender: Towards Versatile Humanoid Whole-Body Loco-Manipulation via Skill Blending [paper]
• [2025.5] Unleashing Humanoid Reaching Potential via Real-world-Ready Skill Space [paper]
• [2025.5] FALCON: Learning Force-Adaptive Humanoid Loco-Manipulation [paper]
• [2025.5] AMO: Adaptive Motion Optimization for Hyper-Dexterous Humanoid Whole-Body Control [paper]
• [2025.5] SMAP : Self-supervised Motion Adaptation for Physically Plausible Humanoid Whole-body Control [paper]
• [2025.5] TD-GRPC: Temporal Difference Learning with Group Relative Policy Constraint for Humanoid Locomotion [paper]
• [2025.5] HuB: Learning Extreme Humanoid Balance [paper] [summary]
• [2025.4] HOMIE: Humanoid Loco-Manipulation with Isomorphic Exoskeleton Cockpit [paper]
• [2025.3] FLAM: Foundation Model-Based Body Stabilization for Humanoid Locomotion and Manipulation[paper]
• [2025.1] Humanoid Locomotion and Manipulation: Current Progress and Challenges in Control, Planning, and Learning [paper]
• [2024.6] HumanPlus: Humanoid Shadowing and Imitation from Humans [paper] [summary]

Locomotion

• [2025.5] Gait-Conditioned Reinforcement Learning with Multi-Phase Curriculum for Humanoid Locomotion [paper]
• [2025.3] NIL: No-data Imitation Learning by Leveraging Pre-trained Video Diffusion Models [paper]
• [2025.3] StyleLoco: Generative Adversarial Distillation for Natural Humanoid Robot Locomotion [paper]
• [2025.2] Embrace Collisions: Humanoid Shadowing for Deployable Contact-Agnostics Motions [paper]
• [2025.2] HiFAR: Multi-Stage Curriculum Learning for High-Dynamics Humanoid Fall Recovery [paper]
• [2025.2] Learning Humanoid Standing-up Control across Diverse Postures [paper]
• [2025.2] ASAP: Aligning Simulation and Real-World Physics for Learning Agile Humanoid Whole-Body Skills [paper]
• [2025.2] HiFAR: Multi-Stage Curriculum Learning for High-Dynamics Humanoid Fall Recovery [paper]
• [2024.8] VMP: Versatile Motion Priors for Robustly Tracking Motion on Physical Characters [paper]
• [2024.6] WoCoCo: Learning Whole-Body Humanoid Control with Sequential Contacts [paper]
• [2022.5] AMP: Adversarial Motion Priors for Stylized Physics-Based Character Control [paper] [note]

Planner

• [2025.8] BeyondMimic: From Motion Tracking to Versatile Humanoid Control via Guided Diffusion [paper]
• [2025.8] End-to-End Humanoid Robot Safe and Comfortable Locomotion Policy [paper]

Perceptive and Locomotion

• [2025.5] VideoMimic: Visual Imitation Enables Contextual Humanoid Control [paper]
• [2025.3] Distillation-PPO: A Novel Two-Stage Reinforcement Learning Framework for Humanoid Robot Perceptive Locomotion [paper]

Embodied AI

• [2025.3] Trinity: A Modular Humanoid Robot AI System [paper]

Manipulation

• [2024.10] Precise and Dexterous Robotic Manipulation via Human-in-the-Loop Reinforcement Learning [paper]

Quadrupedal

• [2023.3] DreamWaQ: Learning Robust Quadrupedal Locomotion With Implicit Terrain Imagination via Deep Reinforcement Learning [paper]
• [2022.2] Concurrent Training of a Control Policy and a State Estimator for Dynamic and Robust Legged Locomotion [paper]
• [2021.9] Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learning [paper]
• [2021.7] RMA: Rapid Motor Adaptation for Legged Robots [paper]
• [2020.10] Learning Quadrupedal Locomotion over Challenging Terrain [paper]
• [2020.4] Learning Agile Robotic Locomotion Skills by Imitating Animals [paper]

Open Source

• [2025.6] [PBHC] (https://github.com/TeleHuman/PBHC) -- TeleAI
• [2025.2] HOMIE -- OpenRobotLab
• [2024.12]: Unitree RL GYM -- Unitree
• [2024.9]: Engineai-Legged-Gym -- Engineai Robot
• [2024.3]: Humanoid-Gym -- Shanghai Qi Zhi Institute and RobotEra