At the University of Alaska Anchorage we have been studying how to use an eye tracker for user authentication. Our most recent project uses the image captured by the eye tracker to authenticate the user using iris recognition algorithms. Some benefits to this approach are:
1. Authentication is easy and unobtrusive. The user simply sits in view of the eye tracker.
2. Iris recognition algorithms can be fooled by static pictures of irises. This technique incorporates a strong liveness test by combining gaze tracking with the iris recognition.
3. Since authentication is unobtrusive it can be performed continuously in the background. This eliminates the problem where a user authenticates and leaves the computer unlocked and accessible while going to perform another task.
The EyeTech TM3 was an excellent choice for our project thanks to the API that provides access to the raw image data. The combination of performance and portability makes the TM3 one of our favorite and most frequently used eye trackers.
We have also used the TM3 to design a gaze-based system to quickly enter passwords using an on-screen keyboard. The user simply looks at the letters in his or her password and the system clusters the gaze points to determine the password symbols.