Chilling face of the future
As train companies say they’ll use facial recognition technology to replace tickets...
There are few sights more dispiriting than a long queue at ticket offices or machines when you’re rushing to catch a train. however, technology is now offering an ingenious solution.
A facial recognition system is being developed that uses two invisible, near-infrared lights flashing at high speed to help a single camera capture a 3-D image of a face in astonishing detail.
It will register the smallest details, down to tiny blemishes and wrinkles, and can recognise individuals even if they are wearing glasses or moving quickly along a platform. The image can then be checked against a customer database.
Such a system could potentially dispense with the need for tickets as passengers, their faces already scanned in a registration process, could pay by online account.
Developed by Bristol robotics Laboratory, the system has attracted funding from the railway Safety and Standards Board, and interest from Go-Ahead, Britain’s largest rail franchise.
It could be up and running within three years, available initially on special ‘fast-track’ lanes without ticket barriers, according to Prof Lyndon Smith of the laboratory.
The technology has wider applications, too: checking into a hotel, for example, to speed security checks at airports and accessing ATMs without a bank card or PIN.
‘The face is the key to doing everything you want to do in the modern world,’ says Prof Smith.
And recognising that face has never been easier. Facial recognition technology is one of the ‘next big things’ and, the way advocates describe it, sounds harmless and beneficial to the consumer in its range of applications.
Yet, others fear there are sinister aspects to the technology that strike at the heart of our right to privacy.
Last month, a UK-based technology company called Blippar unveiled the world’s first mobile phone app that uses facial recognition software to identify people instantly — from information that exists about them online such as Facebook or Twitter profiles and Instagram — simply by pointing a phone at them.
Blippar’s innovation is part of the trend for ‘augmented reality’ — which blends a digital experience with the real world.
It already provides a similar visual recognition service for objects, anything from famous paintings to products in a shop, providing accompanying information and even related adverts on screen.
Where people are concerned, however, it will be able to provide name, age, occupation, background, relationship status and income — anything about them that is on the internet — and display it on the phone’s screen.
If this sounds to you like hours and hours of endless entertainment, then you are just the sort of consumer Blippar is looking for.
If, on the other hand, you find it disconcerting that strangers might not only identify you, but discover your biographical information, then you’re not alone.
AND what use Blippar might make of that information is not clear. While the company is promising it won’t use facial scans to target individuals for advertising purposes, other internet companies have made similar pledges and more in the past — and quietly ignored them.
To some consumer rights advocates, facial recognition technology is arguably the last frontier of privacy in the digital age, and its growing use is causing alarm.
After all, nothing is more personal to us than our faces, and nothing is more immutable (excluding cosmetic surgery). If need be, we could disguise pretty much everything else about ourselves on the internet, but our face gives us away instantly.
One might have thought, then, that a company like Blippar would need official permission before digitally scanning a face and identifying it. Of course, the endlessly intrusive, fiercely predatory tech world doesn’t agree.
When the U.S. government con- ducted talks in 2015 with Silicon Valley and consumer groups to discuss how facial recognition should be regulated, the privacy advocates ended up walking out in protest.
They complained that not a single tech firm would accept that a company should only be able to scan the face of a member of the public and identify them with their permission.
In its defence, the tech fraternity insists that getting that sort of consent isn’t feasible. Do the killjoys seriously expect people to run after everyone they’ve snapped and get them to sign a legal waiver, they ask?
The debate is likely to get more intense. Facebook, Microsoft and Google have been pouring money into facial recognition because it will help ‘connectivity’ among their users.
And they have come a long way from the basic and least intrusive forms that have been around for years, such as the system on many digital cameras that detects a face in a shot so it can focus on it.
A little more sophisticated is the so-called ‘facial characterisation’ software which can identify a generic face — say, a white female in her 20s.
In Germany, this has been used in a campaign to market a particular brand of beer to women via outdoor digital advertising screens. The beer ad is triggered only when women walk by.
One step up is the software that is supposed to detect a specific face for verification purposes. Jaguar Land rover, for instance, is working on putting the technology in its cars so vehicles can recognise owners and unlock the doors for them.
Where the privacy alarm bells really start ringing is with the more elaborate systems, like Blippar, that recognise an unknown person from a digital database of faces. And here, facial recognition is already being used in far more places than you might think.
Shops, banks and casinos employ it to identify difficult customers, criminals and fraudsters, while Churchix is an Israeli system that alerts churches when it recognises known criminals at events.
And it’s not always simply to flag up troublemakers. A facial recognition company called FaceFirst offers retailers a service that, it boasts, can flag up big spenders entering a store.
Alvaro Bedoya, a law and privacy professor in Washington DC, warns that facial recognition will create a ‘world without anonymity ... you walk into a car dealership and the salesman knows your name and how much you earn. That’s not a world I want to live in.’
Facial recognition software generally comes in two forms: the first works rather like matching a fingerprint — measuring the distance between your eyes, the measurements of your nose, lips, ears and other facial features, and matching them against a data- base. however, to get a good match you need a well-lit, full frontal picture of a face. If it’s partially obscured or distorted by heavy make-up or even a smile, the system may not recognise it.
The newer and more powerful method involves creating algorithms, or complex formulas, that train computers to distinguish between human faces for themselves.
FACEBOOK’S Artificial Intelligence lab has announced it has an algorithm that can recognise, with 83 per cent accuracy, people in photographs even when it couldn’t see their faces.
It relied instead on characteristics such as body shape, hairdo, pose and clothing.
Meanwhile, a U.S. engineering professor claims to have developed technology that can scan someone’s iris from across a room and identify them with the precision of a fingerprint.
Privacy campaigners remain highly sceptical about tech companies’ promises to consider privacy rights — justifiably so. From the dawn of the web, internet companies have been obsessed with gathering as much data as possible about users in the hope it will one day be useful.
So when Blippar co-founder, Omar Tayeb, insists that privacy is ‘baked into’ its new face recognition feature, his claim is halfbaked at best.
For while users of the app can take themselves off it at any time, the identities of 70,000 public figures, from singers to scientists, have been programmed into the system without their permission. These celebrities have to ask Blippar to be taken off the system and must wait up to 24 hours for their request to be met.
Blippar boasts, however, that its ‘powerful tool’ will allow famous people to ‘monetise’ their faces — i.e. use them to generate profits — but it’s easier to imagine many of them will actually be appalled.
And we have already seen what can go wrong with face recognition technology. In russia, criminals have used it to scan social media and identify women who had appeared in adult films — then blackmail them.
experts say facial recognition isn’t yet at the stage that it can guarantee to put a name to any face that is snapped in public — the databases of faces aren’t large enough, nor are computers fast enough to access the potential billions of faces online.
Another problem with face recognition systems is throwing up too many possible matches for a face. But experts predict this will be overcome when computers can factor in other information that people disclose on social media, such as their location, physical height and build.
It’s only a matter of time, according to the tech world, before all of us will be identifiable — with accompanying biographical data appearing on the identifier’s screen — the moment we step out of our front door. A world without anonymity or privacy, indeed.
We can do our best to fight this unprecedented intrusion into our personal lives. But for those who cannot resist posting endless details of themselves online, it’s a little too late to be worrying about privacy.