IDEMIA is celebrating another strong performance on the part of its face biometrics technology in testing run by the National Institute of Standards and Technology (NIST).
The company’s facial recognition algorithm for 1:N matching got the top ranking in the latest round of the Face Recognition Vendor Test, which featured 75 tested systems.
NIST breaks down algorithm performance across a number of metrics, and IDEMIA’s technology performed especially well in a couple of key categories. In 1:N matching using a mugshot database of 1.6 million images, IDEMIA’s solution had an error rate of just 0.0016. In testing with the same-sized database of visa border photos, its error rate was 0.0035 – or, in other words, an accuracy rate of 99.65 percent.
As NIST explains in its FRVT 1:N report, the error rate is the “False Negative Identification Rates (FNIR) for the case where a threshold is set to limit to the False Positive Identification Rate (FPIR) to 0.003.”
Commenting on the results in a statement, IDEMIA Chief Technology Officer Jean-Christophe Fondeur emphasized his company’s track record as a leading vendor in the biometrics industry, noting its regular participation in NIST testing.
“The test results confirm IDEMIA’s long-standing expertise in facial recognition AI-based research and how advanced our technology is,” he said. “We strive to demonstrate our leadership by regularly taking part in NIST tests. 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.”
IDEMIA’s technology also performed exceptionally well in recent testing by the Department of Homeland Security’s Science and Technology Directorate. IDEMIA’s facial recognition algorithm attained a 99.5 percent True Identification Rate for unmasked faces in the S&T’s Biometric Technology Rally, and a face acquisition rate exceeding 95 percent for masked individuals.
April 6, 2021 – by Alex Perala