“My light, mirror, tell me, but report the whole truth …” A gadget can soon become such a mirror for us. Moreover, he will report the truth not only about ourselves, but at the same time about our (potential) partner, client or colleague. Whoever you like. It will be enough to just upload the photo portrait of this person into a special program.
In principle, any of us is able to determine by the appearance of the interlocutor his personal qualities, character traits. For example, understand how friendly or emotionally stable he is. This assessment is far from always true, but still it is not random. Even better, a trained person, for example, a specialist in non -verbal communication, will cope with this task. It is enough for him to look at the photo.
And now, a group of Russian psychologists and specialists in artificial intelligence has created an algorithm that is able to recognize our personal features from a photograph more precisely than this will be done by a stranger or a specialist who saw us for the first time.
The team of Russian researchers taught several neural networks for this, which, interacting with each other and working in the complex, fulfill the same task as a person in the process of communication, namely: they evaluate the predisposition of personality in the photo to the characteristics of behavior, which are expressed in the form of fivebasic features (the so -called “big five”):
– extraversion (orientation towards the outside world, sociability, assertiveness),
– friendliness and readiness for compromises,
– conscientiousness and consciousness (including punctuality, reliability, readiness to follow the rules),
– emotional instability (high or low level of anxiety, self -esteem, stress resistance),
– openness to new experience (flexibility of the mind, imagination, independence of judgments, creativity).
“The neural network imitates the work of the human brain in some aspects and studies about the same as a person,” says Alexander Kachur, head of the scientific team, specialist in artificial intelligence. – when learning, it is given a set of input data and reference output (Dataset).
In our case, these were more than 30 thousand photographs from 12.5 thousand men and women sent through social networks, and the results of the tests that these volunteers passed at our request to determine the portrait of their personality and the severity of the features from the “Big Five”.
One part of these data served as a training sample for the neural network. The second is a control sample for evaluating its work. Gradually, the neural network learned to lay out each image on 128 invariant features and further calculate the laws of their connection with the character ”.
Control measurements showed that artificial intelligence correctly determines the level of severity of character traits in an average of 58% of cases, while in case of accidental guessing of a stranger this indicator 50%.
All the features except extraversion, the stranger guesses worse than the algorithm. The difference is small, but this is only the first stage of work. Researchers are sure: the more data the algorithm will compare, the more aiming the results will be in the future.
It has long been known about the connection between character and appearance – in particular, by the person. Physiognomy were fond of antiquity. But attempts to describe facial features and connect them with the character undertaken in the XIX and early XX centuries., were unsuccessful.
“The data that some features of the person are associated with character were gradually accumulated throughout the 20th century,” says one of the authors of the study, associate professor of the Department of Psychology of the Higher School of Economics Evgeny Osin. “But these ties were very weak, and it was impossible to rely on them to make judgments about the personality, as physiognomists did”.
The team of Russian researchers of the first suggested that these connections are nonlinear and are not based on individual features, but on their complex.
“Our neural network looks at the appearance as a person,” adds Alexander Kachur, “that is, it does not separate the nose or lips separately, but perceives the whole set of features at the same time, evaluating them in 128-dimensional space”.
The authors of the project suggest that this technology will allow us to better understand ourselves, make more reasonable and conscious decisions in various situations and it is better to interact with the outside world. Did you manage to open something new in itself by the researchers?
“Technology really helped us learn more about ourselves,” explains Alexander Kachur. – Among us, for example, there were introverts who did not know about it and forced to have an open, extrovert way of life. Having received the necessary recommendations from the program, we organized the workspace in a new way-we divided the office into two parts.
In the far
, quiet zone, introverts are located, there they can communicate less, plunging into formulas, algorithms and writing code. And we gave the other part to extroverts, life is seething there, it is noisy and fun there. I believe that as a result we began to interact more effectively in the team. “.
For business, the technology is also useful – understanding the buyer, companies can more accurately form their commodity offer and offer customers what they need, and not what they want to sell. Some experiments of the authors of the study have already confirmed this hypothesis.
In an experiment with the selection of a seller who can most effectively interact with the client, sales have really grown, since customers were more likely to offer what corresponded to their needs.
Other possible areas of application of the algorithm are applications for dating, diagnostics in employment, recommendation systems of online stores and so on.
