The surveillance trial scanned more than half a million faces in six months and generated one false positive
The British Transport Police scanned more than 500,000 faces during a six-month live facial recognition pilot at London railway stations. According to data obtained by civil liberties group Liberty through freedom of information requests, the system generated one alert, which proved to be a false positive. There were therefore no correct matches or arrests directly resulting from an alert.
The programme involved almost 100 officer hours. The cameras use NEC’s NeoFace M40 software and compare the faces of people passing through a designated area against a police watchlist. An officer assesses alerts before deciding whether to stop a person.
The BTP says it cannot identify people who are not on the watchlist and immediately deletes their biometric data. It also says deployments are guided by intelligence and focused on areas where it believes perpetrators of serious crimes may pass through. Privacy International challenged the effectiveness of this targeted approach, noting that the faces of innocent passers-by are still processed.
The BTP has extended the pilot until November and expanded its use from Network Rail stations to the London Underground.
