Jan. 18, 2019

National Post interview by Marina Gavrilova on Facebook's "10 Year Challenge"

Biometric technologies lab co-founder Professor Marina Gavrilova gives an interview to National Post on calibrating facial recognition technology.

The full article discusses Facebook's "10 Year Challenge" and related privacy and security concerns. It appeared on January 17, 2019 in National Post:


Professor Gavrilova interview answers to Tristin Hopper, National Post journalist, that were included only partially in the article, are found below:

There is no evidence that Facebook created the challenge for anything other than fun, but would I please be able to ask if a vast quantity of images of people taken 10 years apart could potentially be useful in developing facial recognition technology?

It is not a secret that one of the most precious commodities in today’s competitive world is personal data.  Visual data is not an exception. Vast quantities of images can not only be used to train a neural network to recognize specific individuals, but taken 10 years apart to obtain an insight onto how aging processes work, which demographic group is more susceptible to effect of growing older, or  whether those effects are more prominent for men or women.  In my lab, we focus on multi-modal data analytics, which takes into an account a combination of various factors.

What other applications could you potentially see from such a collection of images?

Collection of facial images combined with other information (for instance, country of birth, employment, interests, hobbies, events, friends, gender etc), which is typically included in the social network user’s profile, can be used to deduct whether, for instance, certain profession or a hobby can have a significant effect on aging. In addition, emotions can be extracted from facial expressions and can correlate to the above factors, leading to other insights into intricate connections between human body and mind.”