Researchers from the Stevens Institute of Technology in collaboration with Princeton University and
When two people first meet, theyinstantly evaluate each other and make hasty judgments: from age to intelligence, based solely on appearance. This first impression may not be accurate, but it is extremely powerful. It shapes our relationships and influences everything from hiring decisions to criminal convictions.
The authors of the new work asked thousands of peopleshare your first impressions of 1,000 generated faces ranked by criteria such as intelligence, religiosity, trustworthiness or sociability. The responses were then used to train a neural network to make similar judgments about people based solely on photographs of their faces.
Many conclusions of the algorithm are consistent withcommon cultural assumptions: for example, people who smile tend to be considered more trustworthy, and people with glasses are more intelligent. But there were also cases when it was not clear why the algorithm ascribes a particular character trait to a person.
The algorithm did not give us an answer as to why itrated this or that photograph in a certain way. But even so, it helps us understand how other people see us. This way we can rank, for example, photographs according to how we will be perceived.
Jordan W. Suchow, cognitive scientist and AI expert at Stevens School of Business
The algorithm was originally created to helppsychological researchers to select images of faces and use them in experiments on perception and social cognition. But AI can be applied in the real world. People on social media only share posts that make them look smart, confident, or attractive. An algorithm can be used to make this process easier.
Read more:
It has been hunted for centuries: what do we know about the planet Vulcan next to the Sun
Physicists have experimentally confirmed a new fundamental law for liquids
Astronomers have found the source of mysterious radio bursts that come from space