Scientists Propose Gait-Based Biometric Identification Method for Seniors With Wearable Devices

A Chinese-Italian study published in Information Fusion proposed a gait-based biometric identification technique for seniors with wearable devices, reports Phys.org . The researchers described a gait template synthesis algorithm to ameliorate the problem of intra-subject gait fluctuation, with the addition of an arbitration-based score-level fusion algorithm. Two matching algorithms make preliminary decisions, and a third algorithm provides a final decision if the preliminary decisions have inconsistencies. The proposal's feasibility was confirmed using a public dataset containing acceleration signals from three inertial measurement units worn by users age 64 and older, which was shared by Osaka University eight years ago. The experimental outcomes demonstrated that the average recognition rate reached 96.7 percent, indicating the method was appropriate for robust gait-based identification of seniors.