floyo logo
Pricing
Create with Alibaba Happy Horse model now! Try here 👉
floyo logo
Pricing
Create with Alibaba Happy Horse model now! Try here 👉
Last updated
2026-03-27

A custom node package for ComfyUI, ComfyUI-MotionCapture enables users to perform 3D human motion capture from video using GVHMR technology. It extracts SMPL parameters and skeletal motion through advanced pose estimation techniques, enhancing the capabilities of motion analysis in AI art workflows.

  • Enables 3D human motion capture by leveraging state-of-the-art pose estimation methods.
  • Extracts SMPL (Skinned Multi-Person Linear) parameters, providing detailed skeletal representations for animations.
  • Integrates seamlessly with ComfyUI, allowing for enhanced functionality within existing workflows.

Context

ComfyUI-MotionCapture is an innovative tool designed to facilitate 3D human motion capture directly from video inputs within the ComfyUI framework. Its primary goal is to simplify the extraction of motion data, making it accessible for artists and developers working with AI-generated content.

Key Features & Benefits

This tool stands out by utilizing GVHMR (Gravity-View Human Motion Recovery) to accurately capture and reconstruct human motion in three dimensions. The ability to extract SMPL parameters is particularly beneficial, as it allows for the creation of realistic animations and simulations based on real-world motion data.

Advanced Functionalities

ComfyUI-MotionCapture employs advanced pose estimation techniques that enhance the precision of motion capture. By utilizing models such as ViTPose and HMR2, it ensures high fidelity in motion tracking, which is essential for applications requiring realistic human movements.

Practical Benefits

By incorporating this tool into ComfyUI, users can significantly improve their workflow efficiency, gaining better control over motion data extraction. This leads to higher quality outputs in AI art projects, as artists can create more lifelike animations and interactions based on actual human movements.

Credits/Acknowledgments

The development of ComfyUI-MotionCapture is based on the GVHMR framework, which is credited to Zehong Qiu and colleagues. Additional components leveraged in this package include ViTPose, HMR2, and the SMPL and SMPL-X body models, each of which requires individual licensing agreements for use.

Inner Nodes

SaveSMPL