Figures released by Cambridgeshire Police show that 56,534 faces were scanned during operations in Peterborough city centre in May and June, as officers used Live Facial Recognition (LFR) technology for the first time in Cambridgeshire. The force has also revealed that Chief Constable Simon Megicks is among officers receiving training in the technology from the National Police Chiefs’ Council (NPCC).
The technology is designed to help identify wanted offenders, people subject to court warrants and, in some circumstances, vulnerable missing people. It does this by comparing faces captured on a live camera feed against a pre-prepared police watchlist. While civil liberties campaigners have raised concerns about privacy and surveillance, Cambridgeshire Police say the early results show that the technology can be used accurately and proportionately, with trained officers responsible for reviewing potential matches.
How the first deployments performed
The first deployment took place in Peterborough city centre on 16 May. During the six-hour operation, officers used a watchlist containing 579 individuals. The technology scanned 34,472 faces passing through the city centre and generated seven alerts, all of which were reviewed by specially trained officers and confirmed as genuine matches. Three resulted in further police action.
Cambridgeshire Police returned to Peterborough on 19 June for a second operation. On that occasion, the system compared 22,062 faces against a watchlist containing 577 people and generated three alerts. Two were confirmed as genuine matches, while one remained unconfirmed; the force did not classify it as a false alert. One resulted in further police action.
Across the two deployments, 56,534 faces were scanned and ten alerts generated. Nine were confirmed by officers as genuine matches, while one remained unconfirmed. The force recorded no alerts as false matches.
The force says the figures demonstrate that LFR is being used in a targeted way rather than to monitor members of the public indiscriminately. Only people included on the police watchlist are capable of triggering a match, according to Cambridgeshire Police, although the faces of other people passing the cameras are still scanned and compared with the watchlist.
In information released alongside the rollout, Cambridgeshire Police say: “We are using Live Facial Recognition (LFR) technology for the purpose of locating wanted offenders and safeguarding vulnerable people.
“We will be open and transparent in the development, deployment and use of LFR technology, recognising that we must retain the confidence of our communities as we use this new and important crime-fighting tool.”
How Live Facial Recognition works
Live Facial Recognition compares faces captured through a live camera feed against a police-generated watchlist in real time. The software analyses key facial features and converts them into a mathematical biometric template before comparing that information with individuals already included on the watchlist.
The watchlists are compiled specifically for each deployment. They can include suspects wanted for criminal offences, people wanted by the courts on arrest warrants and high-risk offenders subject to notification requirements imposed by the courts. Police can also use the technology when searching for vulnerable missing people or individuals linked to serious investigations.
A key safeguard highlighted by the force is that the technology does not independently decide whether someone has been identified. Instead, when the software flags a potential match, it produces an alert that must then be reviewed by a specially trained officer.
Police say: “All matches are reviewed by specially trained officers who will provide final confirmation on whether or not the match is correct. Only then will an officer engage with the individual and conduct a further assessment on the match before taking appropriate action.”
Privacy concerns and safeguards
Facial recognition technology has been the subject of continuing debate across the UK, with concerns focusing on privacy, data retention and the possibility of incorrect identifications. Cambridgeshire Police have sought to address those concerns by publishing details of how information collected during deployments is handled.
According to the constabulary, images and biometric data relating to people who do not trigger an alert are automatically deleted and cannot be retrieved. Where an alert is generated, associated data is deleted within 24 hours unless it is required for investigative purposes. CCTV footage generated during deployments is generally retained for up to 31 days before being deleted.
The force stresses that people who are not on the watchlist cannot be identified through the system, stating: “Unless you are on the watchlist you cannot be matched.” However, their faces are still scanned and compared against the watchlist before the biometric data are deleted.
Cambridgeshire Police say the technology will not be routinely deployed across towns and cities. Instead, LFR deployments will be intelligence-led and focused on specific policing purposes. The constabulary says the technology is “typically used at public events and crowded public spaces and is deployed to aid policing operations where there is intelligence that supports its use”.
Additional oversight has also been built into the process through agreements with the Police and Crime Commissioner and independent ethical scrutiny arrangements. The force says these measures are intended to ensure deployments remain lawful, proportionate and transparent.
Accuracy and facial recognition technology
Facial recognition technology has historically faced criticism over accuracy, including concerns about potential differences in performance across demographic groups. Cambridgeshire Police point to independent testing by the National Physical Laboratory, which assessed the facial recognition system and reported extremely low false-positive rates.
The testing found no statistically significant differences in the system’s performance by gender or ethnicity at the settings used by police. The force argues that the results from the Peterborough deployments provide further evidence that the technology is performing effectively in real-world conditions.
Cambridgeshire Police have pledged to publish deployment results and operate transparently as Live Facial Recognition becomes part of policing in Cambridgeshire.
This article is republished with kind permission of CambsNews. Read the original here.
More from East Anglia Bylines
