The voice recognition on my cell phone is helpful, but not particularly accurate. When I tell it to “call Mom,” it will sometimes dial my friend Tom or Mamma’s, the pizza place in my neighborhood. Still, this kind of hands-free operation is useful when my arms are full or I’m driving. Researchers at Dartmouth College are working to perfect a different kind of hands-free mode: eye tracking.
The scientists, led by Professor Andrew Campbell, have developed a system called EyePhone, which runs on a Nokia 900 smartphone. The program tracks the position of a user’s eye relative to the screen and enables the user to select an application by blinking.

Shown
here on a Nokia 900 smartphone, the EyePhone allows the user to click on an
application by blinking (source: Andrew Campbell).
Eye tracking is not an entirely new technology. Currently, it is employed by the military and by the disabled, who can use it to control devices like wheelchairs. Eye tracking on mobile devices, however, which have tiny screens and are often in motion, has proven difficult. One of the main problems with past research on eye-tracking phones was this changing environment. When both the user and the phone are moving, keeping track of gaze is tricky. EyePhone solves this problem by learning the user’s eyes and their movement under diverse conditions. For example, it tracks how each eye appears in different light and at varying distances.
Despite these advances, the EyePhone is still far from perfected. According to a paper published by the team, the application is 76 percent accurate while the user is standing still, but only 60 percent when the user is walking. Further advancements will need to be made before the application is available for commercial use.
Robert Jacob, professor of computer science at Tufts University who also works on eye tracking, sees the EyePhone as a great step forward. "One of the problems with the cell phone is that there's no place for the user interface. Eye tracking,” he says, “seems like a very clever idea."
Campbell and his team are interested in how the EyePhone might have applications beyond convenience. They imagine that the system could be used by drivers to track drowsiness and distraction. Simply by clipping the phone onto the dashboard, a driver could employ high-tech safety features even in an older, lower-end car.

