Eye Movements Could Be Next PC Password

Man peering over computer
A new system, still in early development, identifies people by the way they gaze at a computer screen or picture. In the future eye movements could supplement iris scanners, the system's creators say. (Image credit: pzAxe | Shutterstock.com)

No two people look at the world in the same way — literally. When looking at a picture, different people will move their eyes among points of interest in different sequences, researchers have found. Even if two people trace the same paths, the exact way they move their eyes differs. That's why Oleg Komogortsev, a computer scientist at Texas State University-San Marco, is looking to create a system that can identify people by the way they flicker their eyes while looking at a computer screen.

"We are seeing there are enough differences so we can talk about this as a biometric," Komogortsev told TechNewsDaily. A biometric is a measurement of something on the body — fingerprints, for instance — used to identify people. Computer scientists all over the world are studying biometrics for crime solving, for border security, and just as a high-tech way to sign into smartphones, tablets and other devices.

Komogortsev's research is in its earliest stages and needs years of work before it might show up at airports, high-security workplaces or even home computers. However, he thinks eye movements could be part of the next generation of a more established biometric, iris scans, which are already used in some airports and private companies, and in a countrywide ID effort in India

Previously, researchers showed that crooks can fool an iris scanner with printed contacts, or by holding up a high-quality printout of the correct person's eye in front of the scanner. Komogortsev hopes adding an eye-movement sensor could prevent this type of counterfeiting. "The strength of our method is it can work together with iris [scanning]," he said.

"They appear to be making progress in refining and perfecting the idea," Kevin Bowyer, an iris-scanning researcher at the University of Notre Dame, wrote to TechNewsDaily in an email. Bowyer reviewed papers for a recent conference in which Komogortsev presented his research, but was not involved in Komogortsev's work. 

If the Texas State University research goes well, Komogortsev's team could field test an eye-movement security machine in "the next year or two or three," Bowyer said.

Komogortsev's system records eye movements and analyzes two features. In one, the system measures "fixations," the times when people linger their gaze over a point on screen. In another, it measures "saccades," the swift movements the eye makes when it flies between points. Komogortsev’s system considers both the exact path that people's gazes take and the fixations and saccades they make along the way. [SEE ALSO: Eye Movements Control New Laptop Computer]

From those movements, the system calculates unique properties about people's eyes, including the force their eye muscles use and other properties about the fat and flesh around the eye and the eyeball itself, Komogortsev explained.

In research they recently presented, Komogortsev and his team recorded people's eyes as the subjects read part of a poem ("The Hunting of the Snark" by Lewis Carroll), looked at Rorschach inkblots and watched a black screen on which white dots suddenly appeared. All three images worked well. "If you collect enough eye-movement information, no matter the type of stimulus, it's pretty reliable," Komogortsev said.

Eye movements alone have an "equal error rate" of about 34 percent, he and his colleagues found. The equal error rate is a standard measure in security research that takes into account both false positives, letting someone through who doesn't belong, and false negatives, locking someone out who does belong. Smaller rates mean the system works more effectively, and rates for market-ready technologies are generally in the single digits.

The equal error rate of eye movements combined with low-cost iris scans is much better, at about 5 percent, Komogortsev found. The low-cost iris scans alone have an equal error rate of about 6 percent.

Further in the future, eye-movement scans could also help security officials determine if someone is ill or emotionally distressed, conditions that can affect eye movements according to some research, Komogortsev said. "Do we want to accept people into, let's say, some secure facility, if they are emotionally unstable?" Komogortsev asked rhetorically. If future iris scanners incorporate movement sensors, he said, "you are able not only to identify the person, but also to talk about his emotional state."

However, there's still plenty to do before people will check in with an eye-movement scanner at work or an airport. Komogortsev still needs to answer some basic questions, such as whether people's eye-movement patterns stay the same over time, or if they'll need to update their ID systems as they age

And like every other biometrics researcher, he'll eventually need to prove eye-movement analyses are cheaper, faster and more accurate than competing technology, Bowyer said.

Komogortsev and his colleagues presented their work in September at a biometrics conference hosted by the Institute of Electrical and Electronics Engineers.

This story is part of a series about exotic biometrics — unexpected ways that researchers are developing to identify people by their biological features. "It is important to keep track of the new/unusual/not-yet-much-studied things, because this is where the next big things come from," Kevin Bowyer, chair of the computer science and engineering department at the University of Notre Dame, told TechNewsDaily. "Of course, most exotic things never become big.  But history says that some will." 

Bowyer served as a reviewer for a biometrics conference held Sept. 24 through Sept. 26. He helped choose some of the research we examine in this series, which does not feature his own work.

His own area of expertise, iris scanning, was considered exotic 20 years ago, he added.

This story was provided by TechNewsDaily, sister site to LiveScience. You can follow TechNewsDaily staff writer Francie Diep on Twitter @franciediep. Follow TechNewsDaily on Twitter @TechNewsDaily, or on Facebook.

InnovationNewsDaily Staff Writer