
How accurate police facial recognition is, according to the federal tests
Facial recognition accuracy is measured by the National Institute of Standards and Technology's FRVT program , and its 2019-2023 reports found the most…
Software coverage from Daily Detective News.

Facial recognition accuracy is measured by the National Institute of Standards and Technology's FRVT program , and its 2019-2023 reports found the most…

Police facial recognition can be highly accurate or badly wrong, depending on which algorithm is used and whose face it is analyzing.

Federal testing finds top facial-recognition algorithms highly accurate on balanced photo sets, and far less reliable across race, age, and sex — a gap behind a string of real wrongful arrests.

Probabilistic genotyping software turns complex DNA mixtures into likelihood ratios; since the PCAST report of 2016, courts have demanded validation studies first.

Body cameras turned discovery into a data problem measured in hours of video; platforms built for indexing and sharing now sit between prosecutors, defense and courts.

Automated redaction tools blur faces and mute audio before body camera video goes public; some states' deadlines, like California's 45-day rule, leave little slack.