It’s a common belief that eye movements can betray whether or not a person is lying. Previous studies have examined this concept, using human judges to determine whether or not eye movements could indicate dishonesty. Now, technology may back it up. Researchers at the University of Buffalo have created a machine that uses eye movement analysis to detect deceit. To put it in perspective, the average lie-detection accuracy for expert human investigators is around 65%.
The study was recently presented as part of the 2011 IEEE Conference on Automatic Face and Gesture Recognition. Ifeoma Nwogu, an assistant research professor, worked closely with colleague Mark G. Frank to develop the machine and software. Frank is a behavior scientist who focuses primarily on facial expressions and deception. He has previously worked with body heat and involuntary facial expressions to detect deception.
Their system is interesting. It uses a statistical technique to examine how people move their eyes during both regular conversation and situations where the researcher prompts a lie from the participant. Their preliminary test used 40 participants. Those who were caught in a lie had significant changes in their eye movements when shifting from standard conversation to the lie.
In this particular study, a computer measured the eye movements of recorded footage of the participants. It stands as an excellent foothold for future eye tracking research. Imagine integrating eye tracking technology with other biometrics already being used in lie detection, such as galvanic skin response and heart rate.
Here we have one more excellent opportunity in the eye tracking community.