Live Facial Recognition – coming to a street near you?

Live Facial Recognition uses CCTV feeding into powerful and fast computer systems

Did you know that just one image of your face captured by CCTV on the street without your knowledge can be used to match your name and details on Facebook or a dating site? Or perhaps to assess the probability that you might possibly be capable of criminal acts?

In fact, your face is probably already in some database – think passports and driving licences. Millions of people already have had their faces scanned in public places without their knowledge – both by private companies and police or a combination of both.

It’s a hot topic right now. Amazon’s Rekognition service is easy for software developers to integrate into any app or computer program. Facial recognition not regulated or restricted in any shape or form. That may change under Data Protection law but it may be hard to police.

Live Facial Recognition (LFR) – is it in widespread use?

This excellent article in Psychology Today magazine provides a good introduction to how facial recognition is being used around the world. China uses it to detect jaywalkers and prosecute traffic offenders, and it’s increasingly used in that country’s social ranking system. Police in Dubai have it built into special goggles that identify people they meet.

It’s being used all over the UK already by private companies and the police. The property management company responsible for King’s Cross station built it into the CCTV cameras there for “public safety”. That drew a concerned reaction from Information Commissioner Elizabeth Denham, whose office was set up to “uphold information rights in the public interest” and is responsible for enforcing Data Protection legislation. She said it was a threat to privacy.

Biometrics include your fingerprints, your eyes and even your gait

Have you ever recognised somebody who is walking down the street away from you just by their gait, height and build? We all recognise people we know the instant we see their faces.

Those that can be measured are called biometrics. Most can be scanned, identified and stored by clever Artificial Intelligence (AI) computer systems. It’s pretty easy to build up a digital record of a person’s characteristics once you have enough information about them.

Other measurable and usable biometrics include voice, iris and retinal patterns, earlobe and hand geometry, vein patterns and DNA. And of course facial recognition.

Biometrics are in everyday use for security, with our blessing. Many laptops and computing devices use fingerprint identification as an option for users to sign in. Door access systems use hand geometry and iris & retina identification to permit physical access. iPhones use facial recognition with their owners’ blessing to unlock access..

Behavioural biometrics

A second branch of identification capabilities looks at how people do things. They assess the interaction between a person and a device of some sort. For example, how they write their signature – the speed, angle of holding the pen and pressure. It could be their pattern of using keystrokes or scrolling when typing and using a device. Others include measuring finger movements, hand tremors, and so on.

Is facial recognition a new thing?

CCTV images have been used by police for some time and is acceptable as evidence in UK court proceedings. We, the public, seem to have accepted the trade-off between privacy and crime fighting and. But that’s for law enforcement on our streets.

Instant Live Facial Recognition by powerful computer systems is a very different new thing.

So what’s bad about Live Facial Recognition?

In the US, The Washington Post calls it a “convenience trap”, explaining that this is how privacy invasion of our private lives becomes accepted and the norm. The example quoted is faster boarding at airports, where passengers willingly trade speed for privacy. Seems it cuts 9 minutes off boarding time for the airlines too.

While we present ID in many situations, having that done by computer instead is the flaw that worries people. Too much can go wrong, from misidentification (it thinks you’re a terrorist suspect) to data theft. Current technology is not very good at identifying dark skinned people either.

Data protection fears

Facial recognition needs a “watch list” to match the image captured by the camera with a database of known faces. If shopping centres, supermarkets and bars are using facial recognition, where is the watch list held and who is on it?

Do our Data Protection laws prevent the images from being stored in a foreign country if the technology supplier is based there? Chinese technology company Huawei operating in Guyana is a case in point.

In summary

Facial recognition is new and mostly immature technology. In the right hands, used for the right objectives, it is very valuable in protecting us all.

The real downside is that it’s impossible to police. Digital data can be sold or stolen in seconds. Computer security is never going to be totally watertight.

Commercial interests will try to make money from it, of that we are sure.

So the jury’s out. The arguments for and against will rage on for years while the technology becomes all-pervasive. The law will always be playing catch-up with real life.



Amazon Rekognition

Big Brother Watch (May 2019). Face Off – Stop the police using authoritarian facial recognition cameras. Retrieved from

Chinoy S. (2018). We Built an ‘Unbelievable’ (but Legal) Facial Recognition Machine. New York Times. Retrieved from

Dearden, L. (Aug 2019). Facial recognition becoming ‘epidemic’ in British public spaces. The Independent. Retrieved from

Denham E., UK Information Commissioner (Aug 2019). Statement: Live facial recognition technology in King’s Cross. Information Commissioner’s Office. Retrieved from

Fowler, G. (Jun 2019). Don’t smile for surveillance: Why airport face scans are a privacy trap. The Washington Post. Retrieved from

Hutson M. (Jan 2018). All Over Your Face. Psychology Today. Retrieved from

Snow, J. (Jul 2018). Amazon’s Face Recognition Falsely Matched 28 Members of Congress With Mugshots. ACLU (American Civil Liberties Union). Retrieved from

Stabroek News. (Aug 2019). Facial recognition technology. Retrieved from

The Public Purse (Jun 2019). Big Data In Action – China’s Social Ranking System Under The Microscope Part 1. Retrieved from