KEY POINTS:
A student has written software that identifies whether a woman is attractive or not.
The Haaretz Israel news site says computer science student Amit Kagian has developed the artificial intelligence application for his masters degree thesis at Tel Aviv University.
In its early stages, 30 human participants ranked the beauty of dozens of pictures, without explaining why they perceived some women as more beautiful than others.
The results were then processed and mathematically mapped, giving the computer 98 numbers representing various characteristics including the shape of the face, hair colour, and skin smoothness.
Once the software had established a core of information, it was able to learn from that information and then apply that knowledge to 'grade' other images of women.
He believes the study offers a huge leap in AI software development.
To test the computer's responses, a further group of humans were surveyed, and both people and machine came up with very similar rankings, interestingly preferring 'average' features.
"Until now, computers have been taught to identify basic facial characteristics - like is this a woman's face or a man's," Kagian told Haaretz.
"Our software allows the computer to complete a much more complex task of aesthetic judgment, which humans cannot define exactly how they do it," he said.
"Aesthetic judgment is linked to sentiment and more abstract considerations, but now we have made the computer do it."
Kagian's thesis was recently published in the science journal Vision Research.
The reason the software is currently only able to pick attractive women is that male 'beauty' is far more difficult to calculate, with a large number of variables.
- NZ HERALD STAFF
