Michal Kosinsky and Yi Lun Wang, Stanford researchers who developed the AI, developed it based on neural networks with algorithms that can calculate visual possibilities.
The study, published in the Sept. 7 issue of Journal of Personality and Social Psychology, entitled "Deep neural networks are more accurate than humans at detecting sexual orientation from facial images".
This artificial intelligence by both can detect sexual orientation of thousands of images from various faces that appear on dating sites in America.
According to both, from data collected men and women who have same-sex sexual orientation does not have a distinctive face character. "For men, they usually have narrow jaws, long noses, and a bigger forehead than men with different sexual orientations," they wrote in the journal.
While women who have a homosexual orientation has a bigger jaw and a smaller forehead than a hetero female. In the journal, both explained, visuals are more informative than data about perceptions within a community group.
"It could be, sexual orientation is also caused by the composition of hormones that are formed before birth and ultimately have an impact on the physical."