A study from Stanford University has developed a computer algorithm would correctly distinguish between gay and straight men 80% of the time, and 74% of women. The study has generated curiosity over the biological origins of sexual orientation.

The research was published in the Journal of Personality and Social Psychology. It was based on a sample of more than 35,000 facial images that men and women publicly posted on a US dating website. Researchers, Michal Kosinski and Yilun Wang extracted features from the images using deep neural networks in order to analyze visuals from a large dataset.

Artificial Intelligence

The study found that effeminacy is commonly seen in gay men than in straight men.

Interestingly, the study was able to identify key trends as well. For instance, gay men have narrower jaws, longer noses and larger foreheads than straight men. In the same way, gay women have larger jaws and smaller foreheads than straight women.

“Faces contain much more information about sexual orientation than can be perceived and interpreted by the human brain”, the authors concluded from the study.

When the software reviewed five images per person, it could accurately identify sexual orientation 91 percent of the time with men and 83 percent with women. On the other hand, man judges were rather unsuccessful with only 61 percent accuracy for men, and 54 percent accuracy for women.

The paper published in the journal suggested sexual orientation originates from exposure to specific hormones in the womb. This means that people are born gay, and being queer is certainly not a choice. A lower score for females also suggests that their sexual orientation is more fluid.