EYE TRACKING PHONES have moved a step closer to reality thanks to an improved iris recognition algorithm.
Researchers at Dartmouth College have developed an advanced iris recognition algorithm that they have cleverly christened the Eyephone. The algorithm runs on Nokia's 810 tablet, which runs Linux.
Although algorithms for facial tracking have been around for a while, on mobile phones the performance deteriorates due to arm movements, changing background light and composition. Eyephone advances on previous work by tracking the position of the eye relative to the screen as opposed to the direction of the person.
The algorithm then applies an "error box", the area where it can track the user's iris. According to the boffins, the system, though sounding somewhat Heath Robinson, has a 76 per cent success rate in daylight, though that figure drops to 60 per cent if the user is walking.
According to researchers in the area, Eyephone is a step in the right direction. However a lot of work still has to be done to improve reliability before phones will look into your eyes. µ
Larry Ellison pays tribute to an 'irreplaceable friend'
The way we found out may surprise you
Air to the throne
Wonder who will get 999.999.999.999