SoylentNews

takyon [soylentnews.org] writes:

A new contest allows the submission of selfies as well as algorithms for determining beauty [techcrunch.com]:

Robots are starting to appear everywhere: driving cars, cooking dinners and even as robotic pets [popsci.com]. But people don't usually give machine intelligence much credence when it comes to judging beauty. That may change with the launch of the world's first international beauty contest judged exclusively by a robot jury.

The contest, which requires participants to take selfies via a special app and submit them to the contest website [beauty.ai], is touting new sophisticated facial recognition algorithms that allow machines to judge beauty in new and improved ways. The contest intends to have robots analyze the many age-related changes on the human face and evaluate the impact on perception of these changes by people of various ages, races, ethnicities and nationalities.

Dr. Alex Zhavoronkov, a consultant on the competition and CEO of Insilco Medicine, a bioinformatics company focusing on aging research, says "Recent advances in Deep Learning have made machine recognition of beauty aspects far better than ever before." [...] Part of the AI beauty contest framework is not just for humans submitting selfies, but also programmers submitting their best algorithms for machine detection of beauty. Near the bottom of the contest website is a link [beauty.ai] for algorithm submissions that takes coders to a page saying, "Would you like to go down in history as one of the first data scientists who taught a machine to estimate human attractiveness?" In a way, this makes the contest a crowd-sourced event.

[EXTENDED COPY]

At least one of the organizers has an anti-aging or cosmetics motive:

"This contest will help build impartial feature-specific and general robots that will help us understand our faces. But my personal dream is to have this contest extended into anti-aging and general healthcare space," said Nastya Georgievskaya, robot tutor at Youth Laboratories, a company developing deep learning systems for facial analysis.

[...] "People may not care about how to extend their lifespans, but they definitely care about the way they look," Zhavoronkov wrote me. "Insilco Medicine used massive multi-omics data from academic and commercial partnerships to predict the likely geroprotectors that may have beneficial effects on human skin, and we need a way to test the efficacy of these interventions. We will be launching an application called RYNKL in the coming weeks if all goes well, which will allow users to take standardized selfies periodically to analyze the changes in 'wrinkleness' of their face in the context of their lifestyle, behavior, and other interventions."

Original Submission