In our sixth Snapshot Webinar, we looked at Super Recognisers, including the science and the practical issues of implementing a super recognisers’ programme in the police and the public and private sectors.
Super recogniser is the name given to a rare ability in some people to match facial images between different sources. While you might think that anyone can do this, it turns out most people are not very good at it, with an estimated 1-2% of the population able to be categorised as super recognisers.
In our webinar, we brought together three experts to explain the science and how it works:
Kenneth Long, business development manager for facial recognition at Digital Barriers and a former member of the Met Police super recognisers unit
Prof Josh Davis, Professor in Applied Psychology, University of Greenwich who has conducted extensive research into the psychology of facial recognition from CCTV images and the super recognisers talent
Chris Tritton, Senior Sergeant in the Queensland Police Service, Australia who set up the super recognisers unit in his force after discovering that he was a super recogniser himself
First, we heard from Prof Davis. He talked through how he started by researching the ability of people to recognise faces from CCTV images by testing 3,000 people at the Science Museum. The result was that most people are quite bad at it.
But as he was doing the research, he heard about so-called ‘super recognisers’, something he confesses he was quite sceptical about to begin with.
But he conducted research with the Met Police, agreeing to test 20 police officers who were making a disproportionate number of identifications from CCTV. He found significant abilities in some of these people.
Then four months later in August 2011, the London riots occurred. There was more than 200,000 hours of footage to review. Computerised systems were not very good at identifying suspects – the MPS system identified one – while one officer alone identified more than 180 suspects.
A couple of years later, with a grant from the European Union, he conducted a pilot test of 1,500 police officers to test short-term and long-term memory and simultaneous face matching. Following this, he has collaborated with police forces internationally including Australia.
More recently he has been looking at recognition through face masks and sunglasses, finding that super recognisers still scored quite highly even when looking at images of people with partially obscured faces.
Next, Kenneth Long talked about joining the Met Police super recognisers unit led by then-DCI Mike Neville.
At the Met, there were significant number of images that were unidentified, and he gave examples of identifications made by super recognisers that led to successful prosecutions of serial offenders.
He also talked about the use of super recognisers in the private sector – for instance, a casino that uses facial recognition to stop people who have been self-excluded from the casino for gambling addiction. Super recognisers were also used to stop troublemakers from gaining access to the National Television Awards and Brit Awards.
Kenneth said that people with innate super recognition skills are drawn to jobs where this can be put to use, such as policing or CCTV, but further training can enhance this skill.
And finally, we heard from Senior Sergeant Chris Tritton. In 2016, Chris took a test and discovered he was a super recogniser. In 2018 he was awarded a scholarship to go to the UK and review the super recognisers unit in the Met Police and to work with Josh Davis.
On his return to Queensland, he helped set up a mugshot database and a system to share CCTV images. He also tested 500 police officers for super recogniser abilities, and with Josh’s help identified a handful of super recognisers in his force.
Queensland uses image analytics to search the mugshot database and recommend potential matches. Sometimes the system can list hundreds of potential matches, and it is the super recogniser’s job to review the images to find an actual match (if it exists).
While a non-super recogniser would tend to pick the first image they were offered, regardless of whether it was the correct one or not, super recognisers are better at rejecting false matches and reviewing the images until they find the right one.
Watch the video learn more about how this talent is being used in policing and CCTV.
Further resources
As promised in the video, here are some links to further resources which you may find helpful.
www.SuperRecognisers.com – find out if you or members of your staff are super recognisers – take the free test that has been taken by more than five million people worldwide. They also pay £5 for selfie images for use in super recogniser research.
Super Recognisers International (SRI) – agency formed by Mike Neville who led the the super recognisers unit in the Metropolitan Police Service. Works with clients to supply super recognisers for live deployment, private investigations, CCTV reviews and other activities.
The Association of Super Recognisers (ASR) – association formed by Mike Neville to establish super recognition as a recognised branch of forensic science, encourage research into the basis of this ability and how to develop it, and set standards and accreditation for super recognisers.
'Super recogniser' PCSO has spotted 200 wanted suspects even in face masks | London Metro newspaper | 7 Feb 2021
Super recogniser page on Wikipedia with links to additional research and articles.
Comments