Researchers Develop Drone-Based Biometric System That Identifies People From Face, Gait & Body Images

WholeBodyIdentificationScan

A biometric identification system utilizing a combination of facial recognition, gait (the way a person walks) and the body’s shape has been developed. The system, known as FarSight, aims to enable the identification of humans from far distances, such as from a drone.

At the same time, drones are being weaponized and AI data centers are being constructed which will enable an electronic big brother control grid. While FarSight’s identification system has nothing to do with these other sectors of technological development, it doesn’t take much of an imagination to see how these systems could one day be combined to create a dystopian prison planet where humans are tracked, controlled and punished by AI-directed robotic overlords.

“Whole-body biometric recognition is an important area of research due to its vast applications in law enforcement, border security, and surveillance,” FarSight said in a white paper.

This multi-modal method “points toward a broader form of biometric surveillance in which people may be identifiable even when their faces are partially obscured, low resolution or unavailable,” Biometric Update said Monday.

“The research highlights how biometric identification may increasingly rely on multiple behavioral and physical characteristics rather than facial recognition alone, potentially expanding the reach of remote surveillance systems,” Biometric Update said.

While people generally wear clothing outdoors, obscuring the body’s shape and thus its identifiable features, FarSight detailed their method of deducing the shape of a person’s naked body in order to know who they are.

“Our method (3DInvarReID [33]) for encoding body features harnesses the power of Person Re-ID, with the primary aim to effectively capture static body features. We posit that the most reliable cue for body matching is the naked 3D body shape, despite the considerable challenges in reconstructing it from a 2D image. Taking cues from advancements in 3D feature learning, we introduce a pipeline to disentangle identity (naked body) from non-identity components (pose, clothing shape and texture) of 3D clothed humans. The core of our approach lies in a novel joint two-layer neural implicit function that disentangles these components in latent representations,” the white paper said.

While no admitted database of people’s naked bodies exists, air travelers often undergo a whole-body scan which creates an image of their nude bodies. Despite airport security personnel only being able to see a general outline of the passenger (with the newer millimeter wave machines) the machine itself generates an actual, identifiable nude image. Older X-ray based scanners showed screeners the entire nude image.

“There were two models available at the time: the millimetre wave scanner and the backscatter x-ray screener. The former uses non-ionising electromagnetic radiation that produces a 3D image of a person, while the latter relies on x-rays to emit low quantities of radiation to detect metallic and non-metallic items worn by a person,” Airport Technology said. “However, backscatter x-ray scanners lived a rather short life as they were quickly labelled too revealing – triggering privacy concerns – as well as potentially harmful due to the radiation they emitted. They were then phased out from most airports in 2013.”

With new systems being developed which can identify individuals from a wider range of bodily characteristics, a push to gather more categories of biometrics from people may arise.

“The paper itself frames whole-body recognition as useful for homeland security and forensic identification, but the same capabilities raise broader civil liberties and oversight questions about how remote biometric surveillance should be governed,” Biometric Update said. “As whole-body biometric recognition matures, lawmakers may face growing pressure to determine whether gait, posture and multimodal biometric systems should be regulated like facial recognition or treated as a separate category of surveillance technology.”


In case you missed it...

Special Report: Beware The Wrath Of The Fourth Turning

5 hours ago

Trending

User Registration

By clicking “Register” above, you agree to the Terms of Use & Privacy Policy.