Australian and German researchers have developed an artificial intelligence capable of analyzing certain personality traits, following only eye movement. This technology could make it possible to manufacture robots that can better interpret human signals.
It’s often said that the “the eyes are the windows of the soul”. This statement has now been given a whole new meaning. Indeed, an artificial intelligence (AI), designed by researchers from the University of South Australia in partnership with the University of Stuttgart, Flinders University and the Max Planck Institute of Computer Science (Germany), is able to read a person’s personality through the movement of his or her eyes! The results of this study were published in Frontiers in Human Neuroscience.
The researchers followed the eye movements of 42 participants, aged 22 on average, as they undertook daily tasks around a university campus. This technique was performed in order to have more natural results than if the participants had been in a laboratory. The “eye-tracking” method made it possible to know what the participants were looking at, for how long, how fast they were blinking and so on. Next, the scientists evaluated their personality traits using well-established questionnaires. The data collected was used to design an algorithm to find the personality traits of people based only on the movement of their eyes.
The results show that the artificial intelligence developed can reliably recognize four of the five central traits of personality, called in psychology the “Big Five”: neuroticism (ability to experience negative emotions such as anger), extraversion (ability to seek stimulation and companionship from others), friendliness (ability to be compassionate and cooperative) and conscientiousness (ability to follow obligations and to be organized). The AI can not, therefore, recognize the fifth personality trait that is openness, namely the ability, among other things, to appreciate art, to be curious and to have imagination.
This research provides opportunities to develop robots and computers that are more natural and capable to interpret human social cues. These results also demonstrate a considerable influence of the personality on the control of daily eye movements.