AI Research

New Study Uses Machine Learning to Predict Sexual Orientation

A new dissertation from University of Pretoria Information Technology master student John Leuner has revisited the thorny question of whether machine learning methods can effectively detect sexual orientation.

A new dissertation from University of Pretoria Information Technology master student John Leuner has revisited the thorny question of whether machine learning methods can effectively detect sexual orientation.

AI-empowered “gaydar” systems are one of the more controversial applications of machine learning. Previous research from Stanford University (Wang and Kosinski, 2017) was criticized in the AI community and beyond for perceived algorithm bias and the technique’s potential negative impact on individuals based on their sexual orientation.

One of the main criticisms of the Stanford study was that its dataset comprised mostly white faces. Google Software Architect Blaise Agüera y Arcas also questioned whether the algorithm might be affected by grooming patterns and presentation, etc. instead of facial traits.

Leuner first reproduced the Stanford study’s deep neural network (DNN) and facial morphology (FM) models on a new dataset and verified their efficacy (DNN accuracy male 68 percent, female 77 percent, FM male 62 percent, female 72 percent). He then created a new model based on highly blurred facial images to examine whether features such as brightness or predominant colours could contribute to sexual orientation prediction.

To diminish variables other than facial features, Leuner built a new dataset from dating websites and extracted features from facial images using VGGFace (a pre-trained deep learning neural network), classifying these features with a logistic regression model. For the FM model Leuner used Face++ to extract facial “landmarks”, and trained a logistic regression model with distances derived from these landmarks.

Leuner also examined whether skin colour itself was predictive of sexual orientation. He created a new ML model — a highly blurred image classifier — trained on input images that displayed no information on shape or size of facial features. Two types of blurred images were created for the regression model: 5×5 pixel images containing 25 colours, and 1 pixel images containing one colour.

Although the results suggest this classifier was capable of predicting sexual orientation at a rate of 63 percent for males and 72 percent for females, significant differences in image brightness could be a distractor.


Leuner’s model has thus far received a more positive response on social media sites such as Reddit, as its facial image training dataset draws from a wider variation of racial identities and nationalities.

image (13).png

Questions remain however on whether the prediction of sexual orientation is affected by biological features or image presentation; and whether “gaydar” classifiers in general could be misused.

In an opinion piece in The Conversation, Keele University Law School’s Alex Sharpe and Senthorun Raj argued that “predicting someone’s sexuality may sound innocuous, but in places that criminalise or police homosexuality and gender non-conformity, the consequences of prediction can be life threatening.”

The paper A Replication Study: Machine Learning Models Are Capable of Predicting Sexual Orientation From Facial Images is on Arxiv.

Author: Reina Qi Wan | Editor: Michael Sarazen

4 comments on “New Study Uses Machine Learning to Predict Sexual Orientation

  1. Albina D.

    I see you talking about dating. If you met a girl and want to try a French kiss, then I advise you to read this article There are plenty of tips in this article to help you avoid being disgraced in front of a man or girl. So if you want everything to go well, then follow the link above and read the article

  2. Marisa

    Hello, I am Maria and I want to present you TheYou – the biggest beauty and hair online shop in Bulgaria. This is my first guide for the customers. Some of them are simply looking for a very good make up product because they don’t have time or patience to go to the shops. That’s why Amazon provides a great service for their clients.  We are really lucky, because we can find almost everything for beauty in The You so we don’t have to walk around Sofia in order to purchase something we need at home.

  3. Kevin T


  4. Amanda Woodsons

    The idea of using machine learning algorithms to determine sexual orientation based on facial features and other characteristics raises significant ethical questions. It’s crucial to remember that sexual orientation is a deeply personal and private aspect of an individual’s identity, and attempting to deduce it through technology could lead to invasive consequences.

    On a positive note, I’m heartened by the growing awareness and acceptance of LGBTQ+ individuals and the challenges they face. It’s encouraging to see platforms like Calmerry offering LGBT counseling, you can explore it at as a resource for individuals seeking support. Mental health is a vital component of overall well-being, and having accessible counseling services tailored to specific needs is truly valuable.

    In the end, while technological advancements can bring about positive change, it’s essential to approach them with caution and a strong ethical framework. Respecting individuals’ privacy and rights should always be at the forefront of any innovation.

Leave a Reply

Your email address will not be published. Required fields are marked *