How does Facial Recognition Technology work, and what does its future look like?
There’s an interesting psychological phenomenon which describes the uncanny ability for the human mind to respond to a stimulus by perceiving a familiar pattern where otherwise one wouldn’t exist.
Ever heard voices in the wind, or seen the man on the moon? Or, perhaps you’ve seen a picture of the Cydonia region of Mars, which contains a mountain that one could swear was once an ancient martian statue with a face uncannily like ours?
Sorry to say, none of those are real, but are a result of psychological process called pareidolia.
One explanation for this occurrence can be attributed to our ancestors, who were once prey to larger predators. In the dense grasslands and jungles of our early ancestors, being able to spot a pattern—like the snarling maw of a camouflaged jaguar or lion, for instance—could mean the distinction between life or death. Nowadays, however, this phenomenon tends to make us see funny faces in the front-ends of cars.
This ability to recognize faces is something innately human—everyone has a face—and this ability is now the newest feature of the burgeoning technological sphere designed to make our lives easier: Facial Recognition.
When Apple made Face ID—a facial recognition a feature on their iPhone X—the means of authenticating a user (as opposed to their previous Touch ID technology), what they actually did was normalize the technology in the market.
Now, it’s safe to say that, as the 21st century continues, we will start to see many more implementations of facial recognition technology cropping up in work and home lives, from the future of mobile technology and beyond.
What is Facial Recognition?
Facial Recognition Technology (FRT) is a type of biometric technology (much like fingerprints and the shape of the iris) that scans the unique facial features of a person, and uses those contours as an identification parameter.
As no human face is exactly the same, the likelihood of a similar enough face to fool a scan is about a million to one.
FRT can use a variety of different tools to distinguish unique features of a face:
- Thermal imaging, which measures the level of thermal radiation that is emitted from the face.
- 3D facial mapping, which takes multiple images of the face from different angles and meshes them together to form a 3D model.
- Analysis of the geometric characteristics of the face and unique facial features and distances between them.
Slight variations in the bone-structure, the shape and length of the nose, the position of the brows, the depth of the eye-sockets, are all characteristics known as ‘landmarks’ and are useful tools for identifying one person from another, which means that FRT is very useful in respect to authentication.
Uses for Facial Recognition Technology
The most obvious use of FRT is something we’ve already mentioned: security—particularly as a means of biometric authentication.
While in the early years of FRT, sensors could be easily duped by photographs, with the use of infrared and 3D modelling Facial Recognition is now far more accurate. The result of which means that digital identification is a reliable medium for keeping things secure.
But this functionality goes far beyond personal smart devices.
With a FRT system, employees could simply walk into the building without any sign-in process, as their identity would have already been verified and recorded as they entered the lobby.
This would also eliminate the problem of time fraud, where employees could get others to sign in for them before even arriving to work, because only that worker will have to be seen on the sensors to be read as present. Chinese companies have already implemented such technologies to great effect.
FRT is also a valuable tool to keep physical places safe: it is fully automated and has a far more accurate basis for spotting threats than human security teams that are limited by the human error factor. Using facial scanning to secure your business means that you can track both employees and visitors and ensure that no one that shouldn’t be there is in fact there—even from a distance.
At the same time, this same system could be used as a customer service tool.
If a customer chooses to buy-in, a FRT scanner at the door of your business could spot a returning customer in the process of making their way to your door, and thus notify your staff of their usual buying habits and personalize their customer experience accordingly.
You could literally walk into your favorite coffee shop, or potentially any other franchise location, and be greeted with, “Hi Dave, the usual today? Take a seat, it’s already in the works.”
Ethical Considerations of FRT
As with any new technology, there is always the potential for abuse by those that don’t believe in a thing called “ethics”.
Some of you reading this might have been suppressing involuntary shivers up your spine at the notion of FRT surveillance, having instant flashbacks to Orwell’s dystopian panopticon of ‘Big Brother’ in 1984.
Indeed, concerns have been raised about the potential abuse of the system, not only to monitor the populace—or particular religious or cultural minorities—but also the potential for something called ‘affect recognition’.
Affect recognition, basically describes the use of FRT technology, together with the aid of machine learning, to read a person’s emotional state in order to manipulate them into certain decisions. Something along the lines of: “Oh, Dave’s biometric data indicates that he is in a good mood, now would be a good time to advertise to him those expensive new shoes he was looking at a couple of weeks ago because he is likely to purchase them”.
Such practices are incredibly manipulative, and it probably needs no further discussion to question the ethics of such practices.
To that end, leading companies like Microsoft and Google have pledged to halt the distribution of FRT software until the time that there are stricter policies and regulations in place to ensure that such technology doesn’t inhibit the rights of the world’s citizens.
In any scenario, the use of FRT in a consumer services or business management extent should at the very least be opt-in and optional. If one wants that seamless work environment or coffee-buying experience, then one has to specifically give permission to allow it.
After all, our faces are each and everyone’s own personal property, and what they might or might not say is considered personal information according to POPI practices.
Ultimately, acknowledging ethical concerns surrounding any emerging technology is incredibly important. Regulation is key to ensuring fair-use, because, when it comes to the many benefits of Facial Recognition Technology, preventing anyone that wishes to take advantage of the system to do so is crucial.
As technological progress continues, as technology continues to evolve and refine itself, we’re starting to see the many different aspects of how technology can make our lives easier in many ways—from the smallest of conveniences to largest of paradigm shifts.
Digital Cabinet has used FRT technology successfully to deploy employee time management solutions, using facial recognition to record who has entered into the office and verify who has logged into our system. We have also used FRT as a means to automatically sign-in and register event attendees.
With such solutions in mind, it might not be too long at all before we see completely autonomous restaurants and services that are able to interact with us as well as any human would, that are able to give us customized personal service experiences tailored to our wants and tastes—all of which is written on our faces.
All we can hope for is that they’ll see only what is there, because, interestingly enough, AI experience pareidolia too.
You can find out more about Digital Cabinet at www.digitalcabinet.co.za