Beyond Motion Imitation: Is Human Motion Data Alone Sufficient to Explain Gait Control and Biomechanics?

TL;DR

This study shows that relying solely on motion data in imitation learning does not produce biomechanically accurate gait models, but incorporating foot-ground interaction data improves physical plausibility and realism.

Contribution

It demonstrates that adding kinetic information like ground reaction forces to imitation learning rewards enhances biomechanical accuracy beyond motion data alone.

Findings

Kinematic accuracy alone does not ensure biomechanical plausibility.

Including foot-ground contact data improves joint moment predictions.

Kinetics-based rewards lead to more realistic gait simulations.

Abstract

With the growing interest in motion imitation learning (IL) for human biomechanics and wearable robotics, this study investigates how additional foot-ground interaction measures, used as reward terms, affect human gait kinematics and kinetics estimation within a reinforcement learning-based IL framework. Results indicate that accurate reproduction of forward kinematics alone does not ensure biomechanically plausible joint kinetics. Adding foot-ground contacts and contact forces to the IL reward terms enables the prediction of joint moments in forward walking simulation, which are significantly closer to those computed by inverse dynamics. This finding highlights a fundamental limitation of motion-only IL approaches, which may prioritize kinematics matching over physical consistency. Incorporating kinetic constraints, particularly ground reaction force and center of pressure information, significantly enhances the realism of internal and external kinetics. These findings suggest that, when imitation learning is applied to human-related research domains such as biomechanics and wearable robot co-design, kinetics-based reward shaping is necessary to achieve physically consistent gait representations.