Neurotechnology, Innovatrics score high marks in NIST fingerprint biometrics test

Fingerprint biometrics accuracy continues to improve, as shown by December entries to NIST’s Friction Ridge Image and Features evaluation of fingerprint biometrics from Neurotechnology and Innovatrics.

They are two of only four biometrics developers to submit algorithms for Automated Biometric Identification Systems (ABISs) to NIST’s FRIF TE E1N since it was relaunched in September after a hiatus lasting more than a decade. Tech5 and Innovatrics made the initial submissions, showing off their respective gains in fingerprint biometrics accuracy, and Idemia Public Security followed

Neurotechnology entered the renewed evaluation on December 17, taking the top spot for accuracy in matching single prints and combinations of two fingers (Class A), with a false non-identification rate (FNIR) of 0.0002 at FPIR 0.001. The company also a near-perfect score in fingerprint recognition for Identification Flats (4-4-2), or Class B, and “top-tier performance” in matching plain 4-4-1-1 impressions (Class C), which Neurotechnology says demonstrates the technology’s suitability for demanding law enforcement scenarios.

Innovatrics slashed its error rate sixfold in the Class A evaluation from its September submission, to an FNIR of 0.0005 in the latest results. It also showed gains in median template creation time and the smallest size of templates among participants, with a single index finger averaging only 2.2 kB.

“This same evaluation was last conducted in 2012, and while many tenders have relied on that old evaluation when selecting technologies, these results shift the technology leaderboard and reflect the current state-of-the-art technology,” said Evaldas Borcovas, head of biometrics research for Neurotechnology. “Reaching the highest accuracy in the NIST FRIF TE E1N evaluation – and achieving zero-error rates in some experiments – reaffirms that our biometric technologies are suitable for the most demanding applications, such as law enforcement and national-scale identity programs.”

Neurotechnology notes in its announcement that the company’s fingerprint biometric algorithms have also posted bet-in-class results in prior evaluations including PFT III, MINEX, ELFT and SlapSeg III.

Article Topics

accuracy  |  biometric testing  |  biometrics  |  fingerprint biometrics  |  Friction Ridge Image and Features (FRIF)  |  Innovatrics  |  Neurotechnology  |  NIST