AUSTRALIA — Although known for its rough and tumble outback as much as its innovative opera house, Australia has been working hard to earn a porn-unfriendly reputation. In addition to bans against certain kinds of content and the release of a porn filter that was immediately cracked by a minor, the country soon hopes to offer a method of determining a Web surfer’s age based literally upon sight. While some may scoff, researchers at Melbourne’s Deakin University think they’re on to something.
According to the head of School of Engineering and Information Technology professor Kate Smith-Miles and PhD student Xin Geng, their methods of using mathematics to determine age based on facial features is more effective than anything available elsewhere.
“In extensive experiments of over 2000 faces,” Smith-Miles explains, “our method outperformed the existing approaches” including human guesses after being shown tightly cropped portrait photos.
What Smith-Miles and her student are using is a technique dubbed “AGES,” which uses a combination of complex mathematical algorithms and an assortment of facial features. The professor says that unlike gender, expression, and identity, age estimation has rarely been investigated via facial variation, making their work potentially ground breaking – as well as potentially profitable.
If things go as Smith-Miles hopes, the technology being developed could one day be used in police investigations or as part of age verification on adult-only websites.
But that day is not yet now, as Smith-Miles explains. “In contrast to other facial variations, aging presents several unique characteristics which make age estimation a challenging task.”
And that’s probably even without Botox.
