AI Can Guess Your Personality From a Selfie and Help in Making 'Best Matches' on Dating Websites

31,000 selfies uploaded by 12,000 Russian adults were used in the research

An emotionless selfie can be analyzed to determine an individual's personality using facial recognition technology, a study claims. An artificial neural network built by Russian researchers can access 128 factors of a person's face. It includes shaping of the face, facial width and height ratio, width of the mouth, and the height of the lips or eyes.

Data from these readings is used to categorize an individual based on five personality traits: conscientiousness, neuroticism, extraversion, agreeableness, and openness. The AI was accurate 58 per cent of the time when matched with questionnaires filled by volunteers.

The researchers said humans are less consistent in terms of calculating the personality than facial recognition method which has pure chance of getting it right 50 per cent of the time. The system was best at recognizing conscientiousness among "Big Five" traits and was found to be more accurate on women than on men.

31,000 Photographs Divided into 2 Groups

Representational image Reuters

The scientists used 31,000 selfies uploaded by 12,000 volunteers and divided them into two groups. One was used to train the AI system and the other to test the network. The researchers from Higher School of Economics University and Open University for the Humanities and Economics in Moscow claim that it could assist employers to identify the most suitable job applicants. It could also be helpful for customer service, online tutoring and improve selection of "best matches" for dating websites.

Co-author Evgeny Osin said a significant proportion of information about individuals rests on the face. "The aid of artificial intelligence in making partner choices could help individuals to achieve more satisfying interaction outcomes," he said. The study states new evidence supporting human personality is related to his face, the Daily Mail reported.


However, the researchers acknowledge that some of the variables like makeup, angle, light and facial expressions can affect the static facial features. It can make it difficult to draw fundamental conclusions from the findings. The data was limited to a single cultural and age group of Russian-speaking adults. "Future studies could replicate the effects using populations representing a more diverse variety of ethnic, cultural, and age groups," says the study.

Related topics : Artificial intelligence