Gravity-guided Contact Dynamics Estimation from 3D Human Motions

2026-06-06

AI Summary

This work addresses the limitations of existing motion-capture-based methods for estimating ground contact dynamics, which often rely solely on vertical distances between the body and the ground, struggle to model complex pressure distributions accurately, and are typically constrained to controlled environments. To overcome these challenges, we propose GraCE, a novel end-to-end framework that, for the first time, incorporates a gravity-guided mechanism to estimate full-body contact dynamics from 3D human motion. GraCE leverages human mass distribution, center-of-mass dynamics, and per-body-part contact probabilities to compute total external forces via centroidal momentum trajectories and subsequently distributes these forces across multiple contact points. Experiments on the GroundLink and MOYO datasets demonstrate that GraCE outperforms current state-of-the-art methods in both ground reaction force estimation and contact pressure distribution prediction.

Abstract

Ground contact forces acting on the human body, are crucial for biomechanics studies or sport performance analysis. Prior methods rely on force plates or pressure mats to collect ground contact dynamics, limiting their applicability to carefully controlled settings. A more scalable solution is to estimate the dynamics directly from motion capture data. Recent approaches only roughly estimate the ground contact dynamics from the vertical distance between the body and the ground plane, which cannot capture the complex pressure distribution of all contact points. To this end, we propose GraCE -- Gravity-guided Contact Dynamics Estimation, a novel full-body contact dynamics model for human motions using a realistic influence of body mass distribution and gravity. We use the human's center of gravity to estimate the ground contacts based on its relative distance to the human body. The applied force on each contact is estimated via the product of predicted contact probabilities and the total exterior force computed from the center of mass trajectory. We outperform related work on the GroundLink dataset for ground reaction force estimation, and on the MOYO dataset for detailed contact pressure prediction. The code is published upon acceptance.

Problem

• ground contact forces
• contact dynamics estimation
• motion capture
• pressure distribution
• biomechanics

Innovation

• gravity-guided
• contact dynamics estimation
• center of gravity
• ground reaction force
• 3D human motion