Facial recognition technology being developed to work with drones

A new technology called FarSight, developed by researchers at Michigan State University, aims to enable accurate facial recognition from drones, extending the reach of AI-based surveillance. Funded by the Intelligence Advanced Research Projects Activity (IARPA), the system overcomes the typical challenges of drone surveillance, such as atmospheric distortion and low-resolution images.

FarSight uses a "whole-body biometric recognition" approach, combining multiple algorithms to identify a subject. One algorithm analyses a person's gait, while another creates a 3D reconstruction of their body. A third set of algorithms corrects for the distorting effects of atmospheric turbulence, allowing for a detailed mapping of facial features from a distance. The three biometric markers: gait, body shape, and facial features, are then combined into a single profile for matching. The entire process takes less than a third of a second.

Early tests of the experimental system have shown impressive results. The National Institute of Standards and Technology (NIST) tested FarSight on difficult, low-resolution aerial images and found that it outperformed all other facial recognition systems on the same test set. The team is now working to miniaturise the system to fit on a quadcopter.

£ - This article requires a subscription.

Training Announcement: Freevacy offers a range of independently recognised professional AI governance qualifications and AI Literacy short courses that enable specialist teams to implement robust oversight, benchmark AI governance maturity, and establish a responsible-by-design approach across the entire AI lifecycle. Find out more.