Socially advantageous, a good grip could also keep a phone from spilling the beans

If the search for new sources of sustainable energy were as successful as the hunt for new biometric identifiers, the internal combustion engine would soon be an asterisk in the history of the industrial age.

Some Louisiana State University researchers say they have developed a method of reading a person’s grip as they hold a phone. Gripping-hand verification would prevent any information, including blank-screen notifications, from being seen.

This would create a new barrier to access, this one prior to a phone even being picked up by the wrong person. The phone is a brick to anyone not enrolled as an authorized user.

Biometric software would sense how a hand absorbs or reflects a notification’s sound and vibrations. Phone — Apple and Samsung models in experiments — microphones pick up the identifying signals reflecting off and transmitted through a person’s gripping hand. A convolutional neural network-based algorithm does the verifying.

Experiments with four phone models and 10 notification tones showed 95 percent accuracy.

Grip biometrics to date have focused on security measures for guns.

LSU scientists are working on another biometric security tool involving the back of a user’s hand. This is a comparatively slow-innovating field. For about 40 years, researchers have mostly focused on the palm, and even there, work has not advanced at the pace of, say, iris biometrics.

Holding a phone when using it for access to an event, for instance, typically bares the back of one’s hand. And identifier algorithm then scans the hand for an unobtrusive second verification factor.

More background on hand-geometry developments here.

Article Topics

accuracy  |  biometrics  |  grip biometrics  |  hand geometry  |  hand recognition  |  identity verification  |  mobile device  |  research and development  |  smartphones