Idemia’s facial recognition ranks first on NIST’s new 1:N leaderboard
Idemia announced that its facial recognition algorithm submitted to the National Institute of Standards and Technology’s (NIST’s) latest Face Recognition Vendor Test (FRVT) 1:N test has topped the latest leaderboard for biometric accuracy.
The facial recognition algorithm achieved the top score against 75 tested systems, coming from 281 entrants in NIST’s FRVT, which results were updated on March 26.
The test evaluates the efficacy of facial recognition systems for various applications, including civil, law enforcement, and security applications.
Particularly, the FRVT takes into consideration the accuracy, speed, storage, and memory of facial recognition algorithms.
Idemia’s 1:N algorithms achieved the best accuracy scores in the test, with 99.65 percent correct matches out of 1.6 million face images in the border control systems category.
“IDEMIA has always advocated for the responsible and ethical use and development of biometric technologies,” commented Idemia Chief Technology Officer Jean-Christophe Fondeur. “The test results confirm IDEMIA’s long-standing expertise in facial recognition AI-based research and how advanced our technology is.”
According to the CTO, the company believes that regularly taking part in NIST tests is necessary to demonstrate its expertise and biometric accuracy.
“It’s very important to have government agencies check how our algorithms measure up against other algorithms based on large data volumes. We’re thrilled that our results consistently come out at the very top,” Fondeur concluded.
Idemia has been very active over month, partnering with Cybernetica on a biometric digital ID, joining a UK pilot with IDway biometrics for digital ID checks, and publishing a paper with key recommendations for Europe’s upcoming biometric Entry-Exit System.
Article Topics
accuracy | algorithms | biometric identification | biometric testing | biometrics | Face Recognition Vendor Test (FRVT) | facial recognition | IDEMIA | NIST | research and development
Comments