Global progressive policing
ADVERTISEMENT FEATURE:

Using human super recognisers to fight crime

Super Recognisers

Super Recognisers are humans with an amazing ability to remember faces and recall names. All law enforcement agencies have such people – they just need to be found, trained in tactics and used to identify more suspects or missing persons in a systematic manner. Former Met Police DCI Mike Neville, now CEO of Super Recognisers International Ltd, explains how this untapped resource can help turn the tide of low detection rates within police forces.

With the detection rate for crime in UK at an all-time low, Super Recognisers are a cheap and effective way to turn the tide and utilise the CCTV and images which are already in your force’s area. For Senior Investigating Officers, using Super Recognisers to view masses of CCTV will ensure that suspects are found more quickly. See our website for details of training: Courses, Training & Online Learning – With SRI (superrecognisersinternational.com)

Scientific research into Super Recognisers has been conducted for more than ten years, especially in UK, USA, Germany and Australia. There is little doubt, however, that the ability can assist law enforcement and other agencies. After reviewing the results of the Metropolitan Police’s Super Recogniser Unit in London, Dr David Robertson of York University stated “results…support the wider recruitment of Super Recognisers in other police forces and in agencies such as the passport office and border control in which accurate unfamiliar face matching is a vital to the nation’s security.”

Only around 1% of the population may have the innate skills required to be classed as a human Super Recogniser. These skills cannot simply be taught – people are born with the ability. Over seven million people around the world have now taken the initial online test run by Professor Josh Davis of the University of Greenwich.  Prof Davis said, “The effective deployment of London Police Super Recognisers has vastly increased suspect identifications…Many of these Super Recognisers perform exceptionally highly at empirical tests of familiar and unfamiliar face recognition, simultaneous face matching, and spotting face matches.”   These skills occur without differentiation between sex, race or geographic origin. And with many people still wearing masks, Super Recognisers can still make identifications, despite most of a face being covered.

Widely used in investigations

Super Recognisers have been used to find the Novichok suspects and find victims of the Hillsborough and Grenfell disasters. With the latter, the Super Recogniser was also able to state that seventeen people, who had claimed to live in the tower block, had never been recorded on CCTV visiting it in three months – enabling their prosecution for fraud. Super Recognisers have also been used to link offenders shown on images at different crime scenes, making them as effective as fingerprints and DNA for the first time. As a result, prolific burglars, thieves and robbers have been successfully prosecuted and given far longer sentences than if they had been prosecuted for one crime. Super Recognisers have also helped to bring killers to justice, as in the Alice Gross case and Ilford Crossbow murder, where they viewed hundreds of hours of footage and identified the suspects’ movements.

The TVP Super Recogniser Cadre are already being used to assist live investigations, and have recently deployed in support of Project Vigilant – Preventing Sexual Violence in the Night Time Economy. With the increased capacity, and a further course planned for later this year, the team are going from strength to strength!”
DCI Steve Jones

Officers from around the world have attended courses, with members of Thames Valley Police (TVP) and the City of London taking part in the training.  Following a course in July 2021, DCI Steve Jones noted that, “The TVP Super Recogniser Cadre are already being used to assist live investigations, and have recently deployed in support of Project Vigilant – Preventing Sexual Violence in the Night Time Economy. With the increased capacity, and a further course planned for later this year, the team are going from strength to strength!”

It could be argued that, with the development of computerised facial recognition, the use of human Super Recognisers may soon become obsolete. This would not seem to be the case. Research has been conducted by Professor Martin Innes of Cardiff University into the use of Automated Facial Recognition (AFR) by South Wales Police. The constabulary had featured in several press articles, where they had deployed AFR at large sporting events. Prof Innes noted “…it is more helpful to think of AFR in policing as ‘Assisted Facial Recognition’ rather than a fully ‘Automated Facial Recognition’ system. ‘Automated’ implies that the identification process is conducted solely by an algorithm, when in fact, the system serves as a decision-support tool to assist human operators in making identifications.”  

But it is not a competition between humans and computers. The US research agency, NIST (National Institute of Standards & Technology) reported that: “The data…shows that combining the latest algorithm and a super-recogniser…leads to near-perfect performance.”  Hence, Super Recognisers have a vital role to play in being the human element in the system. In UK this is also a legal requirement, as the Surveillance Camera Commissioner’s Code of Practice, which applies to all government CCTV, there must be “human intervention” before adverse action is taken against a person as a result of computerised facial recognition software. Super recognition is thus a force multiplier which adds significant value to our extant CCTV capability. 

Developing the use of Super Recognisers

The existence of Super Recognisers was first suggested by Prof. Richard Russell of Havard, when he tested four such people in 2009. When researching prosopagnosics (sufferers of face blindness, who cannot even recognise themselves in the mirror), he realised that Super Recognisers must exist at the other end of the human ability scale.   

Following the [London] riots, six officers were posted to New Scotland Yard to form the world’s first Super Recogniser Unit. This unit has achieved significant success in reducing volume crime and solving difficult cases where CCTV or still imagery has been involved.

The first operational use of Super Recognisers was by the Metropolitan Police. I was a Detective Chief Inspector (now retired) and had established a process of systematically circulating CCTV images of unidentified offenders and managed the identification process. I soon came to realise that a small group of officers were making a significant number of identifications and by tracking these cases through to court, he could see the success.

An academic partnership was formed with Prof Davis and initial testing began in July 2011. This joint working soon proved invaluable, as the need for Super Recognisers was soon realised the following the London Riots. The pubic disorder committed over those four days in August 2011, produced over 100,000 hours of CCTV and 4000 images of unidentified looters and robbers, as well as suspects for violent attacks on police and members of the public.

Following the riots, six officers were posted to New Scotland Yard to form the world’s first Super Recogniser Unit. This unit has achieved significant success in reducing volume crime and solving difficult cases where CCTV or still imagery has been involved. When first formed, their role was to attend major events and spot known thieves and other prolific criminals, but other time, it was found that greater success could be achieved by using the Super Recognisers to review the database of images of unidentified criminals, who had been caught on CCTV. Unique to police forces across the globe (strange as this may sound), the Met were able to recall images (often several years old) by crime type, location and other meta-data, including descriptions of suspects.

The first villain to be captured in this way was one Paulo Ronchi, who often wore a hat when breaking into property. By searching “white, male, burglar, hat”, the Super Recognisers reduced the number of images to be reviewed to several hundred and Ronchi was matched to fourteen offences. He was arrested and pleaded guilty. This was not unusual.  Whilst some 65% of defendants plead guilty in London courts, this increased to 92% for cases with multiple CCTV identifications. A great saving of police, court and victim time. And in sexual offences or violent cases, the victim does not have to suffer the trauma of giving evidence.

For many years, police forces have been linking suspects to multiple crime scenes by fingerprints and DNA – now imagery is being used in the same manner.

This tactic came to be known as ‘Linked Series Identification’ or “Face Snapping”, whereby a sequence of crimes involving the same person can be linked together via an overwhelming portfolio of undeniable imagery and evidence. For many years, police forces have been linking suspects to multiple crime scenes by fingerprints and DNA – now imagery is being used in the same manner. This tactic has proven particularly effective in addressing serial shoplifters and commercial burglars and robbers, who target a wide variety of stores and shopping malls across a wide geographic area. The latest British Retail Crime survey report indicates that losses from shoplifting have now reached over £700,000,000 accompanied by a significant increase in violence offered to sales staff. These offences occur in shops where there has been significant invest in CCTV – so images exist of the offenders.

Face snapping achieved its greatest success (in numerical terms) with the case of Austin Caberello. This prolific thief targeted high value shops in Covent Garden and other fashionable London districts. The Super Recognisers were able to match him to 43 offences.  He pleaded guilty and received a sentence of over three years imprisonment. This case resulted in a new tactic – “Domino.”  As it was known that Caberello was a very active offender, the Super Recognisers were supplied with mugshots of his known criminal associates. These faces were memorised and the unit scoured the database and solved another 60 offences.

The tactic of Face Snapping was enhanced by the use of pattern or logo recognition, which was developed for the advertising industry. Its original purpose was to establish the amount of time brand logos could be seen at sporting and other events shown on television. But in a world first, the software was used with the police image database and a burglar was found wearing the same “Everlast” branded T-shirt at two burglaries, a year apart.  His mug shot, wearing the same shirt (taken when arrested for an unrelated matter) was also found – giving a good quality facial image and a name.  The images were supplied to the Super Recogniser Unit, who matched the suspect to seven more crimes. Again, as with AFR, technology works best when used by the right operators.

Deploying Super Recognisers

Super Recognisers can be deployed within CCTV control rooms to spot prolific offenders or to target gang crime. In the Metropolitan Police, Super Recognisers are used in this manner for major events, such as the Notting Hill Carnival. They are particularly valuable at major sporting events such as high-risk soccer games where known ticket touts, previous troublemakers and those with banning orders against them can quickly be identified. At music festivals or pop concerts, known drugs dealers, thieves and pickpockets can rapidly be identified and targeted. Shopping malls and large retail outlets are obvious areas for operation.

Too often, police forces use untrained and poorly motivated staff to view CCTV. In this [Alice Gross] murder investigation, however, ten Super Recognisers were selected to review the footage. As a result of their findings, a river was searched for a second time and the heavily concealed body was found.

Super Recognisers can also be utilised for post-event CCTV analysis and this was particularly effective in the Alice Gross murder case. Too often, police forces use untrained and poorly motivated staff to view CCTV.  In this murder investigation, however, ten Super Recognisers were selected to review the footage.  As a result of their findings, a river was searched for a second time and the heavily concealed body was found. 

In addition to their facial matching skills, Super Recognisers are often adept at behavioural analysis, identifying the abnormal actions by individuals in crowds or in queues and, whilst Super Recognition skills cannot be taught, they can be trained as Behavioural Detection Officers. By using these attributes, they can prevent crimes and even identify likely extremists before they can carry out an atrocity. Where computerised AFR has given an alert of a potential suspect entering an area, Super Recognisers add the essential human element of confirmation of identity, which be an issue of critical life and death decision making. 

To further professionalise and develop the use of Super Recognition, the Association of Super Recognisers was launched in London by the chairman, Lord Lingfield.   The Association will ensure that standardised procedures are used across the globe and the members follow a Code of Conduct, based on that used by forensic scientists. It will also fund further research.  

Identifying Super Recognisers in your force

Super RecognisersMany officers will already know a Super Recogniser in their force – the beat officer or detective, who is always able to put a name to a face. But, often only local supervisors are aware of this vital investigative resource – the Force Headquarters has no idea of which Super Recognisers exist.

By identifying such officers and using them in a more systematic manner – ensuring they have access to all images of unidentified suspects and understand the tactics to utilise their natural ability – police forces will have another valuable weapon in the fight against crime.

Billions of pounds have been spent on CCTV, most of the population have mobile phones with high quality cameras and more and more households have Ring Door bells, with little tangible results in terms of crime detection. Indeed, over the last ten years, as the number of domestic CCTV systems has significantly increased, the burglary detection rate has halved. Super Recognisers can help to make this investment in technology a far more effective crime-fighting tool.

For more information about Super Recognisers, particularly how to develop Super Recognition in police forces and methods of enhancing their abilities, contact Mike Neville (retired Metropolitan Police DCI) at Super Recognisers International Ltd:

Web : www.superrecognisersinternational.com 

Email – [email protected]
Head Office : +44 (0) 330 133 4473 or  07415 634613


You must be registered and logged in to post a comment

Please LOG IN or REGISTER
Top