The Eyes Have It
Update 11/4/14: The research of Oleg Komogortsev has been featured in a special report by the National Science Foundation.
Sight is the most important sense for most people. Seeing is believing. I saw it with my own eyes. Love at first sight. But are the eyes also a foolproof means of identification?
The iris of the eye is as unique as a fingerprint in identifying an individual, and iris scanning for identification is in wide use in certain parts of the world. But it is far from foolproof. Circumventing iris-scanning identification is easy to achieve.
Dr. Oleg Komogortsev, assistant professor of computer science at Texas State University, believes iris scanning is highly flawed.
“Because the eye is a visible area, somebody can take a picture of another person’s iris from a distance, print it on a high quality printer and present it to the iris scanning device, which will identify that photo as an authentic user,” Komogortsev says. “It's also possible to take a picture of the iris and put it on a contact lens, put the contact lens in the eye and present it to the scanner.”
He is working to make ocular biometrics more secure.
Komogortsev explains that biometrics identification was created to make life easier.
“Right now, we have to remember tons of passwords,” he says. “The biometrics technology idea is that you come to your computer and it knows that it's you. You don't have to do anything.”
He says that while the technology currently is used only for high-security government facilities in the United States, that’s not the case in other parts of the world.
“In the United States we use drivers licenses for identification, for example when we go to the bank,” he says. “In India, people were creating fake IDs and it was a big problem. So what they do now is iris scanning and fingerprints. Both of those technologies are spoofable, and in India, their use is widespread. That means almost 20 percent of the earth’s population is relying on flawed technology. ”
Komogortsev, whose research has earned him the prestigious Faculty Early Career Development (CAREER) award from the National Science Foundation, as well as grants from the National Institute of Standards, is working on ways to use eye tracking to create a more secure identification system than iris scanning. Part of his Ph.D. work at Kent State University involved developing a mathematical model of how the eye works. He discovered that because of the differences in muscular structure, one person’s eye movements are different from anyone else’s.
“With iris scanning, you just have to stare into the camera,” he says. “But for this eye-tracking system to work, you would have to move your eye to follow LED lights flashing in random order. The brain basically determines where we go with our eyes. In our preliminary research, we found that those traits create an algorithm that allows me to mathematically extract some of the properties related to the brain and be able to identify a person based on that.”
Because the muscular structure isn't visible, he adds, no one could take an image and duplicate it, eliminating the security problem with iris scanning. And, he says, eye tracking for identification should be easy to implement.
Komogortsev has filed for a patent for his eye-tracking system, and says that the iris-scanning device and the eye tracker use essentially the same hardware.
“So you can make a software upgrade on the existing iris recognition system to add to already existing system capabilities,” he says.
From Gamer to Scientist
Much of Komogortsev’s research originated with his fascination with computer gaming.
“I liked World of Warcraft very much and wanted to play it,” he says. “Beta accounts for this game were $15,000 in 2003. I didn't have $15,000 and even if I did I probably wouldn't have spent it on this.
“So I wrote to the company with a research proposal stating that if you give me the account I will research a way for people who cannot use their hands to play the game. I didn't even know if anyone would get the e-mail, but they did and were very interested. I did implement the project so it's available to anyone who wants to play. The research is published as well.”
Komogortsev and his students have created an application for the iPad that allows users to control the device with their eyes. They have also created their own computer game that is controlled with the eyes and presented to elementary, middle school and high school students to generate interest in STEM (science, technology, engineering and math).
He is collaborating with other departments at Texas State to use eye tracking in various types of research. He joined Drs. Natalie Ceballos, Reiko Graham and Maria Czyzewska in the psychology department to track participants’ eye movements as they were presented with images of different types of foods and correlate this with those participants’ body mass indices (BMIs), hunger levels, cravings and how often they dieted.
Komogortsev’s research has proven significant to many disciplines. Other Texas State researchers who use Komogortsev’s eye tracking technology include Dr. Dan Tamir in computer science, Dr. Sven Fuhrmann in geography, Dr. Pete Blair in criminal justice and Dr. Denise Gobert in physical therapy.
Collaborators in Komogortsev’s research are Dr. Cecilia Aragon from University of Washington and Dr. Larry Price, professor of psychometrics and statistics at Texas State. Postdoctoral researcher Dr. Alex Karpov and graduate student Corey Holland have provided key contributions, and Holland received the prestigious National Science Foundation Graduate Research Fellowship for his work with Komogortsev.
“Significant discoveries are constantly being made about human vision and related brain functions,” Komogortsev says. “However, there are still undiscovered mysteries of human vision that can be revealed and applied to making interaction with computers more fluid and secure. I dedicate my scientific life to that.”