Researcher develops eye-tracker mouse

A researcher at Stanford University has created an application that allows users to interact with a computer simply by looking at the screen and tapping a single key.

Manu Kumar, a 32 year-old doctoral student at Stanford, developed the EyePoint software as part of the Gaze-enhanced User Interface Design (Guide) project.

The research aims to produce standard eye-tracking hardware that allows users to perform basic mouse operations with a combination of gaze and hotkeys.

If a user wants to select a hyperlink to a website, for example, they hold down a predefined button and the part of the screen being viewed will become magnified. The user then focuses on the desired link and releases the key to 'click' on the link.

"Eye-tracking technology was developed for disabled users, but we are trying to get it to a point where it becomes more useful for able-bodied users," said Kumar.

Combining the use of a hand and eyes may be a more natural way of doing things than relying on gaze alone, according to Kumar.

Robert Jacob, professor of computer science at Tufts University, said: "Kumar has got the right idea to let the eye augment the hand."

Kumar's system has no visible cursor or any other feedback system, which studies suggest normally distract users and diminish performance as operators attempt constantly to control the cursor's location.

Current eye-tracking hardware uses a high resolution camera and a collection of infrared LEDs to pick up the movement of the pupil and the reflection of the infrared light from the cornea.

This hardware is expensive, but as many of today's monitors come with cameras built in the technology could soon become cheap enough to be widely available.

Aside from cost Kumar has faced several other problems. Firstly, the eye is not very stable, which means that even the best eye-trackers are not perfect.

Even when focusing on a single point the pupil jitters. To help deal with this problem Kumar's software includes an algorithm to help smooth out small eye movements.

The system also requires a short set-up time to calibrate it with a new user's eye movements, which may prove to be a barrier to wide adoption.

Lastly, the system can fail to work properly if the user has problems such as a lazy eye, or needs to wear thick glasses or special contact lenses.

Kumar is confident that these problems can be overcome, however, and intends to expand the Guide program to switch applications, enter passwords and scroll through screens of information.