Image via Shutterstock
What if, instead of sitting through a whole sequence of dating profiles to swipe to, you simply had to leaf through photos showing your “type”?
In a new paper titled Brain-computer interface for generating personally attractive images, researchers from the University of Helsinki and the University of Copenhagen detail that they have trained artificial intelligence to understand what 30 study participants personally deem to be attractive, and have taught it to generate portraits in line with their preferences.
Perceptions of attractiveness were measured by monitoring electrical activity in the brain, per Dazed. The scanning “worked a bit like the dating app Tinder,” Michiel Spapé, a senior researcher of the University of Helsinki, told Neuroscience News. “The participants [mentally] ‘swiped right’ when coming across an attractive face.”
Based on the volunteers’ responses, a generative neural network (GAN) would then simulate a new face mirroring preferred traits from photos of celebrities.
Researchers report that the AI was able to match individual perceptions of attractiveness with an 80-percent success rate.
While this experiment might suggest how widely ideas of beauty can differ due to cultural and psychological experiences, Spapé acknowledges that it could also expose unconscious biases and stereotypes.
“Potentially, we might gear the device towards identifying stereotypes or implicit bias and better understand individual differences,” Spapé shares.[via Dazed, Shutterstock]
TAXI Daily News