Rank One fingerprint biometric algorithm shines in NIST PFT test
Rank One Computing says the biometric performance and accuracy of its first fingerprint algorithm submitted to the National Institute of Standards and Technology (NIST) Proprietary Fingerprint Template (PFT) Test places it among the top three performers in every category measured.
ROC scored the highest accuracy for cross-sensor IARPA Nail-to-Nail dataset matching, according to the company announcement. The company’s algorithm was also found to be the most interoperable with different sensors, and the most efficient, with search speeds 10,000 times faster than the closest participating vendor. The templates created by ROC’s algorithm were also the smallest, by 160 times compared to the next-smallest competitor templates in NIST testing.
The company emphasizes the importance for the domestic market of fingerprint algorithms developed in America. The test results show ROC’s value proposition for embedded and large-scale fingerprint biometrics deployments, it says.
“ROC is extremely proud to put the U.S. on top of these AI/ML capabilities,” says ROC CEO B. Scott Swann. For years there has been an over-dependence on foreign fingerprint technology. To my knowledge, the entire National Security screening apparatus has had no alternative to foreign fingerprint components to power these mission-critical systems. With this new algorithm, we are giving our customers the option to buy American while also deploying solutions that are more accurate and more efficient. As more agencies are migrating their enterprises to the cloud, these computing efficiencies translate to real cost savings.”
Over the past several weeks ROC has also added image quality validation and presentation attack detection features to its face biometrics SDK, and had its PAD algorithm confirmed for ISO standard compliance in testing by iBeta.