HybridRobotics/whole_body_tracking

Health v2 · maintenance only

Adoption not part of the score

2344 stars · 321 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

BeyondMimic's official motion tracking training code, a humanoid control framework built on Isaac Lab that trains sim-to-real-ready whole-body motion tracking policies. It supports training on LAFAN1 dataset motions without parameter tuning and uses guided diffusion-based controllers for steerable test-time control.

Use cases

• train humanoid motion tracking policies in simulation
• sim-to-real deployment of dynamic whole-body motions on humanoid robots
• reproduce BeyondMimic motion tracking results
• train tracking policies for LAFAN1 motions without tuning
• research diffusion-based controllers for humanoid control
• learn reinforcement learning for legged robot control with Isaac Lab

When to choose

• you need state-of-the-art humanoid motion tracking with sim-to-real transfer
• you work with Isaac Lab/Isaac Sim and Unitree humanoid robots
• you want to train LAFAN1 motions out of the box

When to avoid

• you need actual robot deployment code - use the companion motion_tracking_controller repo
• you don't have access to NVIDIA Isaac Sim/Isaac Lab or a GPU
• you need MuJoCo-native tooling - consider the mjlab alternative implementation

Facets

library · maturity active

machine-learning reinforcement-learning simulation robotics robotics machine-learning simulation deep-learning python humanoid-robotics motion-tracking sim-to-real reinforcement-learning isaac-lab diffusion-policy unitree linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="HybridRobotics/whole_body_tracking")