An Artificial intelligence technology able to recognize the facial features of the person attractive
Researchers have always been curious to see if a computer is able to recognize the facial features of the person we find attractive, and this is possible thanks to researchers at the University of Helsinki and the University of Copenhagen.
The researchers first collected the brain signals from the research participants and merged them with an interface between the brain and the computer with the aim of creating artificial matches, then users were shown faces created by GANs, and users were asked to focus on the face they were seeing, while the brain response was collected and monitored through its signals. Electrocardiogram with the help of EEG.
Through this, the brain-computer interface gained an opinion on what attracted the participant and based on the user's preference it produced a completely new face after taking indications of what it had learned from previous faces.
These images, which were designed in real-time, were presented to users to see if the model was successful and the result was surprising, as the new faces created by artificial intelligence matched the user's preferences with an accuracy of 80%.
Researcher Michael Sabbé, from the Department of Psychology at the University of Helsinki, explains that the study showed that we are able to produce images that correspond to personal preferences by linking an artificial neural network to brain responses.
Computer vision has so far succeeded in classifying images based on objective patterns, but by introducing brain responses into the design, it seemed possible to discover and create images based on psychological characteristics, just like the personal taste.