Google Glass POV

Wearable Cameras Birth A New Biometric


Google Glass POV
Wearable cameras allow individuals to be identified based on their gait, new research suggests.

Wearable technology is a burgeoning category, and products like Google Glass and smart watches are just the beginning. As with mobile phones, on-board cameras are sure to be de rigueur.

But, as this article over at The Verge notes, those cameras will present new challenges (for privacy) and new opportunities (for security). Specifically: cameras coupled with your body seem to create new kinds of opportunities to uniquely identify you. One example: gait biometrics.

The Verge notes recent research published by Professor Shmuel Peleg and Yedid Hoshen of the Hebrew University of Jerusalem. Those researchers created a way to identify first-person filmmakers based on the signature wobble of their cameras. The identity of the user can be determined “quite reliably from a few seconds of video,” the researchers write in their paper.

[Interested in biometrics and wearables? Check out our article “Are Wearables the Future of Authentication?“]

“The idea of distinguishing one person from another comes from analyzing the person’s gait,” Professor Peleg told The Verge. “Everyone has a different body build, height, muscles, and skeleton, and so the body moves differently and uniquely.” Their paper, “Egocentric Video Biometrics” can be viewed here. (PDF)

The combination of cameras and wearables creates a whole new dimension for marketers and advertisers, as well as for law enforcement. Firms like CrowdOptic, for example, are using advanced data analytics to infer things like where an individual is focusing their attention based on data transmitted from cameras on mobile devices like Google Glass and mobile phones.

That information is a gold mine to retailers, who would love to be able to serve you product specific ads and discounts based on your gaze. But the same technology can also be used to isolate anomalous events (like a fist-fight or medical emergency) in large crowds, or to uniquely identify an individual.

Peleg and his colleagues see more practical applications in the short term. For one: wearable devices could be locked to their owner, preventing use by other individuals. Alternatively, the biometric quality of the wearer could be used to automatically sort videos shot by the same user.

Still: privacy-conscious users should be aware that wearing a camera will likely spoil any attempt to remain anonymous, the researchers caution.

Read more via The Verge.

Comments are closed.