Just selfies to know who you are

A new facial recognition technology by two Moscow Universities: Open University for the Humanities and Economics and the Higher School of Economics, can recognize a person’s personality by analyzing selfies without knowing anything else of a person, proving to be more capable than a human.

The software was developed by researchers and experts of BestFitMe, a company specialized in personnel research and selection.

A team of 12,000 volunteers filled a questionnaire and used the answers to build a database based on personality traits. The same volunteers then uploaded 31,000 selfies that were split into two groups: one used to train the A.I. system, and the other one to test the network.

Researchers then created an artificial neural network that assessed 128 different factors of a person’s face, such as the width of the mouth, the height of the lips and the eyes, etc… They used these data to categorize a person based on 5 personality traits: conscientiousness, neuroticism, extraversion, agreeableness, and openness.

The result was impressive: the A.I. was able to determine the psychological characteristics of a person starting from the analysis of the person’s selfies. The system was accurate the 58% of the time starting from pictures taken in uncontrolled conditions. The system was found to be more accurate on women than on men and was best at recognizing conscientiousness.

Researchers say the artificial neural network outperforms an average human who is asked to rate a person’s personality in person but without knowing anything else about the person.

This study presents new evidence confirming that human personality is related to individual facial appearance. It’s the strongest evidence that physiognomy really works.

>>>  Neuralink: brain and computers connected

The machine learning technique is described fully in Scientific Reports.

This software may also improve the selection of the ‘best matches’ for dating websites, for companies for promoting custom products and services. The system could also help bosses identify the most suitable job applicant.

Moreover, knowing personality reveals the kind of lives we are likely to lead, and it doesn’t need sophisticated 3D scanners, it can be easily implemented using a desktop computer.

The 5 personality traits

Conscientiousness: people who have conscientiousness traits are more likely to be organized and dependable. These people are self-disciplined and act dutifully, preferring planned as opposed to spontaneous behavior. They can sometimes be stubborn and obsessive.

Neuroticism: people with high levels of neuroticism are prone to psychological stress and get angry, anxious, and depressed easily.

Extroversion: these people tend to seek stimulation in the company of others and are energetic, positive, and assertive. They can sometimes be attention-seeking and domineering. Individuals with lower extroversion are reserved, and can be seen as aloof or self-absorbed.

Agreeableness: these individuals have a tendency to be compassionate and cooperative as opposed to antagonistic towards other people. Sometimes people who are highly agreeable are seen as naive or submissive.

Openness: this is about having an appreciation for emotion, adventure, and unusual ideas. People who are generally open have a higher degree of intellectual curiosity and creativity. They are also more unpredictable and likely to be involved in risky behavior such as drug-taking.

Artificial Intelligence is constantly amazing but also worrying, especially because it’s gradually making its way, not only through our privacy but also through the most intimate aspects of a person such as, in this case, personality; by categorizing it in a cold and artificial way to use it as a mean to classify people without worrying about if they are exactly as they have been described; removing nuances, which are and always will, the most relevant aspects of the people but not for the algorithms. And once you’re classified, it’s hard to debunk who they think you are. And the personalities, who will be excluded, for example by human resources because considered not suitable a priori? This will only increase the audience of the unsuitable ones without even needing a job interview.

>>>  What will cloud and streaming leave us?

Source dailymail.co.uk