Facial Recognition Grows More Sophisticated and Scary

Lots of things are scary on the Internet, from the Heartbleed vulnerability to the ability of hackers to flip switches on the power grid. Put facial recognition on that list.

Actually, it’s been on many people’s list for a long time. It now is growing to a level of sophistication that should get the attention of folks who haven’t focused on it yet. It isn’t quite there, but we’ve seen how this sort of thing plays out: A technical advance seems vague and nebulous one day and, seemingly overnight, the industry sees the potential benefits and makes it a reality.

This has some folks worried, including Dr. Joseph Atick, a pioneer of the technology, about whom Darlene Storm wrote at Computerworld:

Atick used the NameTag app as an example of face-matching technology being taken too far. The app offered Google Glass users “real-time facial recognition” by matching a stranger with everything about them that can be mined through social media such as their name, occupation and even real-time access to their Facebook, Twitter or Instagram accounts.

It is impossible to argue that the technology is not moving forward quickly. Grant Hatchimonji at CSO offers a very interesting post on where facial recognition is today. It actually turns out it is in different places, depending on whose face is being recognized. In cooperative environments – when folks are willing to look straight into a camera, as at a security checkpoint – the science is moving along well. In uncooperative situations – perhaps ferreting out a known cheater via an overhead camera at a casino – the challenges are understandably more difficult.

The scariness, touched with a bit of weirdness, even extends into the past and future. The Atlantic asks the question of whether computers can use images of children to identify them years later. This is not an unimportant issue, considering how many pictures of children are on Facebook and, in general, how deeply documented everyone’s lives are today. The answer seems to be that it may be possible, but that it becomes very difficult if the kid is younger than seven years of age due to the level of change a younger face undergoes moving forward.

This all sounds very futuristic. It is. But it also is current. The Guardian discusses some of the available commercial products. The usual names – Google and Facebook, to mention two – have or are near facial recognition-based products. The fear is well stated:

There are, naturally, problems, and most relate to privacy concerns. Although privacy is an issue with every form of data mining, at least online the majority of information absorbed by companies is anonymised. Facial recognition, of course, is precisely the opposite. And since facial recognition takes place in public spaces, it is not even necessary for the person surveilled actively to “opt in” to the service.

Facial recognition is another of the endless list of technologies that straddle the line between the good they can bring and the danger they present. Strong laws are needed, but somehow seem unlikely to be put in place.