It used to be all about writing command lines. Then the mouse took over and computers performed on the basis of clicks, followed by touch screens, which put the hand into direct, tactile control of software and applications. But the future is on our doorstep, and it’s controlled by eye movement.
Yes you saw that right. It may still be a technology in its infancy, but Denmark-based Eye Tribe is giving us a glimpse of how pressing and swiping touch screens and mouse pads could one day be fully replaced by eye motion control.
Think eye activated login, gaming and user engagement analytics. Think eye controlled text messaging, medical eye diagnostics, drawing. It’s a pretty exciting concept.
Originally envisioned by four students from the IT University of Copenhagen, including current CEO and Co-founder Sune Alstrup Johansen, The Eye Tribe was founded during the European StartupBootcamp in 2011. The company’s goal was from the very start to make eye-tracking technology available for mass consumer markets at an affordable price. Using low-cost components and plenty of hard work, the present-day result is The Eye Tribe Tracker: a $99 standard USB 3.0 connectible piece of hardware that can detect the movement of the pupil with sub-millimetre precision.
The Eye Tribe technology relies on infrared illumination, which is the same method used in night-vision for example. The eye tracker’s sensors locate features of the eye and calculate an estimate of where the eye is looking based on the eye’s angle and position. It’s a math program of sorts that automatically analyses your eye movements with sufficient precision to understand what type of app response the gaze should convey (like slicing a watermelon in half in Fruit Ninja, or on a more serious note, a diagnosis of an eye impairment for example). Eye-tracking apps are basically divided between passive and active: passive includes stuff like advertising, performance analysis of design etc. whereas active would involve device control, a category where games, amongst stuff like texting, would fall under.
But how accurate is the detection exactly? According to Eye Tribe, the average accuracy goes in the 0.5 degree of visual angle. That means the software is capable of detecting and determining the one-screen gaze position roughly within the size of a fingertip (<10mm). The variety of interaction can be expanded by combining eye control with existing means of control, like gesture, touch, tilt and key press.
The current version of The Eye Tribe Tracker is intended for developers only. This is in hopes that these developers will get inspiration for new applications that could be conceptualised from such an interesting starting point as eye-movement based interaction. The Eye Tribe has worked hard to make the data necessary for development easily available, but to get developers started Eye Tribe will provide you with an SDK with implementations in C ++, C# and Java, including full source codes. The actual methodology behind the performance of the eye-tracking software would require a very lengthy post if it was to do the software any justice, so if you have any interest in discovering all about the juicy details, see here for more information (and dev. kits). In addition, here you’ll find the official developer forum.
Eye movement system control being somewhat of an interesting topic, it’s no wonder that The Eye Tribe’s journey has been one filled with attention and a few awards, including acknowledgments from Forbes as well as triumphs in the 2014 SXSW Interactive Accelerator program. The Danish start-up has by now raised $3.1 million in funding, as well as a grant of $2.1 million, and now includes 16 full time employees (whom you can see above).
The tracker is currently compatible with Windows 7, 8, Mac OS X and Android although more OS’s are sure to follow once new apps start sprouting more general awareness.