He moves his eyes, blinks, smiles. Now try to relax, says the machine. “We’ll see if you’re a good liar or a bad liar shortly.” Professor Hanein and his colleague Professor Dino Levy lead a team at Israel’s Tel Aviv University that has developed a new lie detection method.
They say they have identified two types of liars: those who involuntarily move their eyebrows when telling a lie, and those who cannot control a slight movement of their lips just at the corners, where they meet the cheeks.
Their software and algorithm can now detect 73% of lies and they intend to improve on that as they develop the system. “When you’re trying to hide a lie , one of the things you try to avoid is any kind of bodily reaction,” say the researchers, who say it’s “very difficult for anyone to hide a lie with this technology.”
An almost infallible software
Lie detection methods have probably been around as long as tall tales and tall tales were invented. One of the first documented examples comes from the year 1000 a. C. in China, where a suspect had to fill his mouth with dry rice. After a period of time, the pimples were checked and if they remained dry, the person was found guilty.
The 20th century saw the invention of the first lie detection machines, or polygraphs. The best known of these is the analog polygraph, which typically has three or four ink-filled needles that oscillate on a moving strip of paper. The suspect has sensors attached to his fingers, arms and body and the machine then measures breathing rate, pulse, blood pressure and perspiration as he answers a series of questions.
However, concerns remain about the accuracy of these machines and whether it is possible to fool them. Therefore, researchers and technology companies around the world are working to develop more high-tech polygraph systems.
increasingly sophisticated
At Erasmus University in Rotterdam, the Netherlands, researchers use a magnetic resonance imaging (MRI) machine to detect whether someone is lying or cheating. They do this by looking for color changes in brain scans in response to questions. “Essentially, we see [different] areas [of the brain] that are more strongly activated [light up on scans] when someone decides to cheat or be honest,” says one of the doctors involved in the project.
One high-tech lie detection system already in use is EyeDetect from the Utah-based firm Converus. A subject is asked to answer a series of true/false or yes/no questions. As they do so, the eye-tracking software watches and studies their responses. The result is then ready in five minutes and claims to be 86 to 88% accurate.
