Clearview AI facial recognition performs strongly in US gov NIST testing

Facial Recognition technology- woman

Consecutive National Institute of Standards and Technology tests confirmed reliability.

Consecutive tests by the US government’s National Institute of Standards and Technology (NIST) have shown that Clearview AI’s facial recognition platform performs reliably.

This is the second time in two months that the platform has performed with strong results.

In October, Clearview AI scored first place during a one-to-one (1:1) Facial Recognition Vendor Test (FRVT). This month, the platform was tested using the one-to-many (1:N) test and achieved the top ranking again. While using Rank-1 investigation mode – which is a method testing the effectiveness of this type of technology’s algorithm’s ability to accurately match one image from a gallery of millions, the company’s platform scored at the top in the United States and was in tenth place worldwide across every category. There were 328 algorithms tested in this process.

During the NIST testing, Clearview AI’s algorithm was shown to accurately match mugshot photos with a 99.85 percent accuracy rate in a photo sample of 12 million. Moreover, it was able to accurately match VISA borders images to a 99.86 percent accuracy rate in a 1.6 million photo sample. Clearview AI’s algorithm scored top place in the United States in the most difficult VISA kiosk test and second place worldwide.

Facial Recognition accuracy

The Clearview AI facial recognition accuracy is good news for a company with a controversial past.

Clearview AI has worked with a number of law enforcement departments and organizations around the world, including in the United States, Canada, Australia and others. That said, it faced tremendous criticism and controversy about a year ago when some countries and Europe determined that its data scraping processes were illegal. Moreover, even in the countries where the behavior was not banned, tech giants including Facebook and Venmo issued cessation demands against data scraping of their databases.

The facial recognition platform is designed to help law enforcement to be able to quickly and accurately investigate crimes, identify suspects and enhance public safety while offering victims justice. It has been used for thousands of cases ranging from locating abducted children and apprehending sex offenders, drug traffickers and violent criminals.

Leave a Comment


This site uses Akismet to reduce spam. Learn how your comment data is processed.