Exploring the Use of Accelerometer for Assisting Weight Lifting Exercise

Nour Al Bogha, Sozo Inoue, Brahim Brenaissa,
Human Activity Recognition Challenge, Springer Nature
(Not Available)
(Not Available)
14 pages
One of the most important aspects of professional athleticism and rehabilitation is muscle activation. Although sport tracking technologies are advancing
rapidly, existing technologies and devices for detecting muscle performance are still
out of common users’ hands, due to their complexity, invasive nature and high cost.
However, sensor equipped devices are getting widely available, which opens the
door to alternative solutions using software. In this work we explore the possibility
of using accelerometer data for muscle performance in a weightlifting exercise. By
investigating a predefined common weightlifting exercise, we exploited a correlation between the acceleration and the lifted weights using a simple RBF prediction
method, in a real experiment where data obtained from three participants. The obtained results show that the weight could be predicted in acceptable accuracy for
each individual, based on acceleration data.

Data Files