Estimation of Ground Reaction Forces in running via deep learning models: a comparative analysis
Loading...
Files
Accepted Version
Date
2025-09-04
Authors
Tedesco, Salvatore
Ahern, Sean Francis
O'Flynn, Brendan
Journal Title
Journal ISSN
Volume Title
Publisher
Elsevier B.V.
Published Version
Abstract
Ground Reaction Forces (GRFs) are fundamental to the analysis of running biomechanics and play a critical role in performance assessment, injury prevention, and rehabilitation [1]. While GRFs are traditionally measured using expensive and non-portable systems such as instrumented treadmills, there is growing interest in more accessible, scalable alternatives [2]. Inertial Measurement Units (IMUs), which record acceleration, angular velocity, and orientation, have demonstrated potential in this context [3-5]. Recent research has applied deep learning techniques to estimate GRFs from IMU-derived kinematic data [4-6]; however, a standardized benchmark for comparing different models using a publicly available dataset remains absent, limiting reproducibility and cross-study evaluation.
Description
Keywords
Ground Reaction Forces (GRFs) , Running biomechanics , Performance assessment , Injury prevention
Citation
Tedesco, S., Ahern, S. F. and O'Flynn, B. (2025) 'Estimation of Ground Reaction Forces in running via deep learning models: a comparative analysis', Gait & Posture, 121, pp.242-244. https://doi.org/10.1016/j.gaitpost.2025.07.260