The University of Manchester in the U.K. and the Universidad Autónoma de Madrid in Spain have collaborated on an artificial intelligence (AI)-based biometric verification system for identifying a person by their specific gait.
The researchers estimate the system can authenticate individuals with near-perfect accuracy, and only a 0.7% error rate.
The AI was trained to learn gait patterns using the world's biggest footstep database, SfootBD, which contains 20,000 footstep signals from 127 distinct individuals.
Manchester's Omar Costilla Reyes says every human's walking pattern features about 24 separate variables and movements, so "monitoring these movements can be used, like a fingerprint or retinal scan, to recognize and clearly identify or verify an individual."
The researchers tested their data by using a large number of "impostors" and a small population of users in three real-world security scenarios: airport security checkpoints, the workplace, and home settings.
From University of Manchester
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA
No entries found