facebookresearch/momentum

A library for human kinematic motion and numerical optimization solvers to apply human motion

What it solves

Momentum provides foundational algorithms for human kinematic motion and numerical optimization solvers. It allows developers to apply human motion in various applications, such as body tracking from RGBD or monocular RGB images.

How it works

The library provides a set of tools for forward and inverse kinematics with interpretable parameterization, as well as differentiable optimization modules. It is available as both a C++ library and a Python package (pymomentum), with support for CPU and GPU (CUDA) acceleration via PyTorch.

Who it’s for

Researchers and developers working on human motion analysis, robotics, and computer vision tasks involving human kinematic tracking and optimization.

Highlights

  • Foundational algorithms for human kinematic motion.
  • Numerical optimization solvers for human motion applications.
  • Support for RGBD and monocular RGB body tracking solvers.
  • Available in C++ and Python with PyTorch integration for differentiable optimization.

Related

  • Project
    orocos/orocos_kinematics_dynamics

    A real-time kinematics and dynamics library for rigid body calculations, providing solvers for forward and inverse kinematics for robotic systems.

  • Project
    RussTedrake/underactuated

    A software library providing algorithms for the control of underactuated robots, supporting dynamic movements like walking, running, swimming, and flying.

  • Project
    Simple-Robotics/aligator

    A C++ and Python library for efficient constrained trajectory optimization in robotics, used for motion planning and model-predictive control.

  • Project
    adipandas/multi-object-tracker

    A Python library providing easy-to-use implementations of various multi-object tracking algorithms and OpenCV-based object detectors for video analysis.

  • Project
    stack-of-tasks/pinocchio

    A fast and flexible C++ library for rigid multi-body dynamics computations, providing analytical derivatives for robotics simulation, control, and learning.