MS-Human-700: Whole-body Human Musculoskeletal Model

Project Page | Reinforcement Learning Environments (msgym)

Overview

This directory contains the MuJoCo XML and asset files for the MS-Human-700 model.

Related papers:

• MS-Human-700 (ICRA 2024)
• DynSyn (ICML 2024)
• MPC2 (ICLR 2025)
• QFlex (ICLR 2026)
• and more (on balance & fall, contact-rich & deformable, vision language model ...)

For reinforcement learning environments and training scripts, see msgym.

To visualize the models, drag-and-drop the MS-Human-700-*.xml files into MuJoCo's simulate viewer.

Models

Primary Model (Full Body)

File: MS-Human-700.xml

Full body human musculoskeletal model for whole-body locomotion tasks.

• Bodies: 90 (optimized to 80)
• Joints: 206 (constrained to 85 for control stability)
• Muscles: 700 actuators

Legs Locomotion Model

File: MS-Human-700-Locomotion.xml

Focusing on lower-body dynamics. This model isolates the legs for locomotion research while simplifying the upper limbs and torso.

• Bodies: 80
• Joints: 36
• Muscles: 100

Unimanual Manipulation Model

File: MS-Human-700-Manipulation.xml

Focusing on right arm and detailed right hand, designed for manipulation tasks.

• Bodies: 127
• Joints: 42
• Muscles: 81

Control Demos

DynSyn control results:

QFlex control results:

High-Fidelity Motion Tracking results:

Leveraging MuJoCo Warp for massively parallel GPU simulation enables the rapid and efficient training of control policies capable of high-precision motion tracking across diverse and dynamic trajectories.

The demos below illustrate these tracking capabilities of the MS-Human model:

• Overlap: The model and reference trajectory are rendered directly to visualize tracking accuracy.
• Separate: The model and reference trajectory are rendered with an offset to showcase motion details.

Running: Overlap	Running: Separate
Walking: Overlap	Walking: Separate