Video Analysis Software To Spot Liars By Analyzing Eye Movements

A new futuristic software may change life as we know it if research by the University of Buffalo, State University of New York is anything to go by. Researchers have developed a video analysis system that can effectively spot lies by analyzing eye movements. The system's results are about 82.5 per cent conclusive. The technology, if successful, may greatly enhance lie recognition techniques, for instance the lie detector tests. The research comes in the wake of work by two Stanford professors who last year set out some interesting parameters for detection of CEO and CFO deception.

The Buffalo researchers based their work on determining if there are signal changes transmitted whenever someone is lying and if they can be detected. Speaking to the UB Reporter, Ifeoma Nwogu, a co-author of the study and professor at the Center for Unified Biometrics and Sensors, reiterated the success of the innovation affirming that indeed it could effectively detect lying. The work was inspired by studies done by Paul Ekman of University of California in San Francisco, School of Medicine.

Ekman is working on research regarding emotions as they relate to facial expressions. But critics have already raised controversy over the accuracy of the software. Controversy still reigns on traditional systems such as polygraphs, it remains to be seen what the face-detection system will offer. However, the University of Buffalo researchers are planning to widen their study to include body language. The researchers expect the creation of faster algorithms to allow for enhanced capacity to detect behavioral deviations in almost real time. Researchers have long known that certain behavioural signals can indicate someone is lying.