Artificial Intelligence can now guess if you’re gay or straight from facial features
Critics have said such studies and software in the public domain can be misused by governments to prosecute or target the LGBT community.science Updated: Sep 12, 2017 17:03 IST
An Artificial Intelligence machine can guess the sexual orientation of a person by analysing facial features, a study conducted by Stanford University researchers Michal Kosinski and Yilun Wang showed.
The study -- to be published in the Journal of Personality and Social Psychology -- said the software could do this by noticing subtle differences in facial structures, The Economist reported this month.
When showed a single image of a person, the machine accurately distinguished between gay and heterosexual men in 81% of the cases. The accuracy went down to 74% in cases of women. The algorithm’s accuracy increased to 91% (men) and 83% (women) when the machine was shown five facial images per person.
In contrast, humans can judge the sexual orientation correctly 61% of the times for men and 54% for women.
The study suggests sexual orientation may be derived from biological traits such as hormones, raising questions about the individual ‘choice’ determining if a person is gay or straight.
It also found that gay men and women could have ‘gender-atypical’ features, expressions and ‘grooming styles’, The Guardian said, adding that it may mean homosexuals appeared to be more feminine.
One of the researchers, Dr Kosinski, told The Economist that similar AI systems can be used to determine other traits such as political leanings and IQ of people, sparking a debate about the possible misuse of such machines.
Researchers downloaded 130,741 images of more than 36,000 men and 170,360 images of over 38,000 women from a dating website. They then subjected them to facial-detection technology which left 35,326 pictures of 14,776 people.
The machine paid attention to the nose, eyes, eyebrows, cheeks, hairline and chin for determining male sexuality. Mouth corners, hair and neckline were more important features to determine’s women’s sexual orientation.
The study, however, did not include people of colour, transgender or biosexual people.
Opposition to the research
Two LGBT groups have called the study ‘reckless’ and warned of its negative impact.
Critics have also said such studies and software in the public domain can be misused by governments to prosecute or target the LGBT community, while images from social media websites can be used to determine sexuality without a person’s consent.
“It’s certainly unsettling. Like any new tool, if it gets into the wrong hands, it can be used for ill purposes,” The Guardian quoted Nick Rule, an associate professor of psychology at the University of Toronto, who has published research on the science of ‘gaydar’.
“If you can start profiling people based on their appearance, then identifying them and doing horrible things to them, that’s really bad.”
Others have pointed out that the differences between gay and straight men and women may not be due to facial features at all.
“These ‘subtle’ differences could be a consequence of gay and straight people choosing to portray themselves in systematically different ways, rather than differences in facial appearance itself,” Professor Benedict Jones, who runs a Face Research Lab at the University of Glasgow, told the BBC.