gci

Let a Robot Tell You if You're Pretty ... for Science

Contact Author
Close
Fill out my online form.

Virtual mirrors and other AI applications are already making inroads in the beauty world, but a new platform takes the technology in a unique direction.

"Healthy people look more attractive despite their age and nationality," according the the founders of Beauty.AI, a new experiment that combines gerontology and machine learning.

Assuming that relative attractiveness, or "looking well," is tied to good health, the creators posit that machines can begin to judge human health based solely on photographic evidence.

Further reading: Print Your Perfect Foundation Shade

The premise of the project is simple: Users download the Beauty.AI app, photograph themselves (makeup, beards and glasses are not allowed), a "robot jury" calculates the attractiveness of submissions, and the most attractive men and women ("kings and queens") are appointed based on age category.

The creators' premise is that, with the right semantic learning, computers can learn to accurately identify wrinkles, facial symmetry, skin color, gender, age and ethnicity, and to judge attractiveness.

Assuming that relative attractiveness, or "looking well," is tied to good health, the creators posit that machines can begin to judge human health based solely on photographic evidence.

The creators add: "But evaluating beauty and health is not enough. The team’s challenge is to find effective ways to slow down ageing and help people look healthy and beautiful. There were many attempts to do this in the past, but this time it will have a global impact because of the advances in deep learning, symbolic learning and massive semantic analysis."