TEL AVIV — “Computer, computer on the desk, am I pretty or am I a mess?”A student at Tel Aviv University has developed software that may be able to answer that ages-old question by accomplishing esthetic judgment based on mathematical formulas.
“Until now, computers have been taught to identify basic facial characteristics — like is this a woman’s face or a man’s,” Amit Kagian, who developed the program for his masters thesis in computer science, told the Israeli newspaper Haaretz. “Our software allows the computer to complete a much more complex task of esthetic judgment, which humans cannot define exactly how they do it. Esthetic judgment is linked to sentiment and more abstract considerations, but now we have made the computer do it. This constitutes a substantial advance in the development of artificial intelligence.”
The results of Kagian’s research, which combines computer programming and psychological exploration, recently were published by the scientific journal Vision Research.
To accomplish his goals, Kagian first had 30 humans rate the attractiveness of several dozen people based on pictures of their faces. The “mug shots” then were mapped mathematically based on 98 characteristics like geometric shape of the face, hair color, eye color, facial symmetry and smoothness of skin. The subjects’ rankings of the images also were input.
Then came the test of the computer’s ability to learn and make judgments.
“We input new pictures of faces into the computer and it graded them based on the information it had,” Kagian explained.
The human subjects also were asked to grade the new pictures.
“The computer produced impressive results: The rankings were very similar to the rankings people gave,” Kagian told Haaretz.
In a nutshell, he said, the computer learned to identify the characteristics people found attractive in other people’s faces, even without the benefit of an explanation of why humans found the characteristics attractive. In both the human and computerized evaluations, average faces with no extreme characteristics ranked highest.
“The computer learned a mathematical function; however, it implicitly learned to prefer average faces,” Kagian told Haaretz.
The experiment only studied women’s faces.
