April 8, 2014
LatentGesture Could Be The Next Step In Device Security
Enid Burns for redOrbit.com - Your Universe Online
Mobile security is important in our increasingly connected lives. Many security procedures are now gesture-based, password protected, and some even require fingerprints. Georgia Tech researchers are taking security to a new level, however, with a new system that actually learns a user's gestures in order to unlock a device such as a smartphone or tablet.
Imagine not having to punch in a number code, touch the screen in a designated pattern or even provide biometric details such as a fingerprint or palm print to unlock your device. The new system, called LatentGesture, is instead able to identify the device's user by the way he or she swipes the screen, checks boxes and pushes buttons on a touchscreen device.
Developed during a Georgia Tech lab study using Android devices, the system was found to have an accuracy rate of nearly 98 percent on smartphones and 97 percent on tablets.
“The system learns a person’s ‘touch signature,’ then constantly compares it to how the current user is interacting with the device,” said Polo Chau, a Georgia Tech College of Computing assistant professor who led the study, in a statement from the university.
The study consisted of 20 participants who were asked to fill out a form that included buttons to push, boxes to check and swipe slider bars. The system was able to learn the user's tendencies and then create a profile. In subsequent tests, LatentGesture was able to successfully match the owner and flag unauthorized users.
“Just like your fingerprint, everyone is unique when they use a touchscreen,” said Chau. “Some people slide the bar with one quick swipe. Others gradually move it across the screen. Everyone taps the screen with different pressures while checking boxes.”
LatentGesture is also able to be used with multiple users on a single device. Such a program would make it possible for children to access the family's tablet, and be restricted to appropriate content and activities.
“This feature could be used when a child uses her dad’s tablet,” said College of Computing sophomore Premkumar Saravanan. “The system would recognize her touch signature and allow her to use the device. But if she tried to buy an app, the system could prevent it.”
Gesture-based methods are increasingly being used to provide security for mobile devices. Some devices also use biometrics, which requires a particular identifier to unlock a device, such as the user's fingerprint. In the case of number codes or gesture-based security, the system can be defeated by observation.
“It’s pretty easy for someone to look over your shoulder while you’re unlocking your phone and see your password,” said Samuel Clarke, another College of Computing student on the research team. “This system ensures security even if someone takes your phone or tablet and starts using it.”