AI Case Study

South Wales police had a 92% false positive rate using facial recognition at the 2017 Champions League final

South Wales police trialled facial recognition at the 2017 UEFA Champions League final match at the Principality Stadium in Cardiff. However, the system had a 92% false-positive identification rate, with 2,297 out of the 2,470 people identified as potential matches being inaccurately matched.

Industry

Public And Social Sector

Security

Project Overview

"South Wales police began trialling the technology in June last year in an attempt to catch more criminals. The cameras scan faces in a crowd and compare them against a database of custody images." These databases seem to have been supplied from a variety of sources, including Interpol and the UEFA itself.

Reported Results

92% (2,297) of the 2,470 potential persons of interest identified by the facial recognition technology were found to be false positives. "The force blamed the high number of false positives at the football final on “poor quality images” supplied by agencies, including Uefa and Interpol, as well as the fact it was its first major deployment of the technology. South Wales police admitted that “no facial recognition system is 100% accurate”, but said the technology had led to more than 450 arrests since its introduction. It also said no one had been arrested after an incorrect match."
Use of the technology has also been criticised by civil rights groups.

Technology

Details undisclosed

Function

Operations

General Operations

Background

Facial recognition is being trialled by police to catch persons of interest at large crowded events, such as football matches and concerts.

Benefits

Data

Photos of individuals taken at events and compared with image datasets provided by organisations such as Interpol and the UEFA Champions League.