While it may seem like an odd evolution in dating apps like Tinder, researchers have developed an AI system that can find out which individual faces are attractive based on their brain waves and ict.
A joint team of psychologists and computer scientists at the University of Helsinki in Finland used electroencephalography (EEG) measurements to list the facial features that make people look attractive, and then submit the results to a machine learning system called the Generative Antigonist Neural Network (GAN). ). .
About 30 volunteers saw artificially produced, but very realistic faces, none of whom were real people, created from a dataset of about 200,000 popular images. Each participant was equipped with electrodes to measure their brain activity.
“They have nothing else but to look at the pictures,” Cognitive neuroscientist Michelle Spepp explains. “We measured their instantaneous brain response to images.”
The system learned their preferences and listed their unique EEG attraction identifiers. He created personalized faces based on the person’s preferences.
The system then developed second-generation faces produced by AI. The newly personalized faces were shown to 30 volunteers along with a random assortment of additional faces.
Participants rated 80% of 2nd generation positive faces as attractive, compared to just 20% for random faces.
“It is important to successfully assess attraction, because it is a serious psychological property of stimuli,” Speap said the machine learning system can successfully analyze and ict behind the attractive personal and subjective, with the possibilities of further exploring the human psyche with AI being endless.
“Potentially, we can guide the device towards identifying latent stereotypes or biases and better understanding individual differences.”
Also on Rt.com
Do you think your friends will be interested? Share this article!