Download
Humanoid-Gym is a reinforcement learning framework designed to train locomotion and control policies for humanoid robots using high-performance simulation environments. The system is built on top of NVIDIA Isaac Gym, which allows large-scale parallel simulation of robotic environments directly on GPU hardware. Its primary goal is to enable efficient training of humanoid robots in simulation while enabling policies to transfer effectively to real-world hardware without additional training. The framework emphasizes the concept of zero-shot sim-to-real transfer, meaning that behaviors learned in simulation can be deployed directly on physical robots with minimal adjustment. To improve reliability and generalization, the framework also includes sim-to-sim validation pipelines that test trained policies across different physics engines.
Features
- Reinforcement learning framework for humanoid robot locomotion training
- Built on NVIDIA Isaac Gym for GPU-accelerated physics simulation
- Zero-shot simulation-to-real transfer for deploying policies on physical robots
- Sim-to-sim validation pipelines for testing policy robustness
- Training scripts, reward functions, and robot configuration tools
- Designed for research in robotics, control systems, and embodied AI
Project Samples
