MHR: Momentum Human Rig
Aaron Ferguson, Ahmed A. A. Osman, Berta Bescos, Carsten Stoll, Chris Twigg, Christoph Lassner, David Otte, Eric Vignola, Federica Bogo, Igor Santesteban, Javier Romero, Jenna Zarate, Jeongseok Lee, Jinhyung Park, Jinlong Yang, John Doublestein, Kishore Venkateshan, Kris Kitani, Ladislav Kavan, Marco Dal Farra, Matthew Hu, Matthew Cioffi
2025-11-20
Summary
This paper introduces a new computer model of the human body, called MHR, designed to create realistic and flexible animations of people.
What's the problem?
Creating believable human movement in virtual reality, augmented reality, and computer graphics is really hard. Existing models often look stiff or unnatural when characters are put into complex poses, and integrating them into existing software can be difficult. They need a way to represent both the overall shape of a person *and* how they bend and move, and do so in a way that’s easy for developers to use.
What's the solution?
The researchers built MHR by combining the best parts of two existing approaches. They used a system called ATLAS to separate the skeleton (bones) from the body’s shape, allowing for more realistic forms. Then, they added a modern ‘rigging’ system, inspired by a tool called Momentum, which makes the model bend and move naturally, even in extreme poses. This rigging system uses ‘correctives’ – small adjustments that automatically fix unnatural bends and distortions.
Why it matters?
This new model is important because it makes it easier to create realistic human characters for things like video games, virtual reality experiences, and even movies. Because it’s designed to work well with existing software, developers can more easily incorporate these realistic characters into their projects, improving the overall experience for users.
Abstract
We present MHR, a parametric human body model that combines the decoupled skeleton/shape paradigm of ATLAS with a flexible, modern rig and pose corrective system inspired by the Momentum library. Our model enables expressive, anatomically plausible human animation, supporting non-linear pose correctives, and is designed for robust integration in AR/VR and graphics pipelines.