Created by researchers at Stanford University in Palo Alto, California, an artificial intelligence system is able to determine a person’s sexual orientation by analyzing a photograph of the face. According to the researchers, the results advance understanding about the origins of sexual orientation and the limits of human perception.
According to the study, published this week in the Journal of Personality and Social Psychology, the machine has 81% accuracy for men and 74% for women, but “accuracy of the algorithm increases to 91% and 83%, respectively, if five pictures of the person are given. ”
Scientists fed a deep neural network with 35,326 facial images. From them, the machine was able to distinguish by physical traits such as nose shape and gestural, gay and straight.
The photos were taken from a public relations site. The study raises questions about the biological origins of sexual orientation, the ethics of face recognition technologies, and the potential for artificial intelligence without violating people’s privacy.
The research indicated that some gestures and presentations are more amenable to homosexuality. Among men, gays would have narrower jaws than straight ones, noses, and larger forehead in relation as well. Women, jaws, and smaller foreheads are signs of homosexuality.