Google files patent application for facial expression recognition authentication system
Google has filed a patent application in regards to a facial recognition authentication system for mobile devices, which could see unique facial expressions used to perform an unlock on Android devices.
The patent application outlines a system of identifying facial landmarks as well as facial gestures, and using two images for authentication.
According to the application, “facial gesture includes at least one of a blink gesture, a wink gesture, an ocular movement, a smile gesture, a frown gesture, a tongue protrusion gesture, an open mouth gesture, an eyebrow movement, a forehead wrinkle gesture, and a nose wrinkle gesture.”
Google originally introduced its face unlock feature with the Galaxy Nexus and Android 4.0 Ice Cream Sandwich.
I’ve used this feature on my Android smartphone and tablet and though it is a convenient way to unlock your phone, and will be something I monitor closely in the future, I don’t believe the system is perfect yet. Lighting plays an important role in a successful unlock, though that isn’t altogether unexpected or unreasonable. Second, I have found that normal changes in my appearance like beard growth, eye glasses and even a bad hair day have seemed to prevent a successful unlock.
With the update to Jelly Bean, and following reports that the feature was easily spoofed with a photograph, Google introduced a “liveness check,” requiring the user to blink before unlocking. Google also allows users to “improve face matching” by taking picture in various conditions.
That being said, when enabling the face unlock option, the operating system tells you to keep a few things in mind, including “Face Unlock is less secure than a pattern, PIN or password,” and also that “someone who looks similar to you could unlock your phone.”
Reported previously in BiometricUpdate.com, Google recently officially stated that it will not add facial recognition features to its Google Glass consumer device.
authentication | facial recognition | Google | mobile | smartphone