NIST relaunches 1-to-N fingerprint biometrics matching evaluations

The U.S. National Institute of Standards and Technology has relaunched its evaluation of performance capabilities of one-to-many fingerprint biometric identification algorithms after a 13 year hiatus.

NIST has renamed the Fingerprint Vendor Technology Evaluation (FpVTE) as the Friction Ridge Image and Features (FRIF) Technology Evaluation Exemplar One-to-Many (E1N).

The only submissions to the relaunched evaluation so far are from Tech5 and Innovatrics.

NIST uses the submitted algorithms to match biometrics from three datasets for the evaluation. Class A is made up of index fingers only, Class B consists of 4-4-2 or “identification flat” captures, and Class C includes all ten fingers.

Tech5 was found to have a false non-identification rate (FNIR) of 0.001 at a false positive identification rate (FPIR) of 0.001 or lower in matches against the 1.6 million images in Class A, and a Rank-1 FNIR of 0.0004. The best results so far for all pairings in Class C are Tech5’s. The company currently sits first for both operational thresholds and Rank-1 FNIR in plain-plain (0.0018 and 0.0042), plain-rolled (0.0006 and 0.0037) and rolled-rolled (0.0003 and 0.0034). The company’s algorithm also performed well in template creation speed and footprint.

At FPIR 0.001 or lower, Innovatrics scored FNIRs of 0.0160, 0.0120, 0.0022 and 0.0011 respectively for left slaps, right slaps, left and right slaps together and identification flats. At Rank-1 FNIR was 0.0092, 0.0027, 0.0010 and 0.0005.

The FpVTE, back in 2012, included submissions from “afis team,” NEC, id3, 3M Cogent, Sonda, Morpho, Neurotechnology, Tiger IT, Decatur Industries, Papillon, Innovatrics, BIO-key, Dermalog, SPEX, Aware, HiSign, ID Solutions and AA Technology.

“This evaluation was highly needed, because the previous similar 1:N fingerprint NIST testing was conducted 13 years ago under FpVTE,” says Tech5 CRO Ameya Bhagwat. “The end customers are looking at these tests as a reference when selecting large-scale ABIS (Automated Biometric Identification System) platforms for their projects in civil identity, foundational identity, elections, passport systems and the like. We at TECH5 are very proud that this fingerprint algorithm, that is used in the T5-OmniMatch ABIS platform, shows the best results.”

NIST notes that the evaluation is of template creation and search algorithms used with an automated biometric identification system (ABIS), and does not evaluate ABISs themselves.

Article Topics

biometric testing  |  biometrics  |  fingerprint biometrics  |  Friction Ridge Image and Features (FRIF)  |  Innovatrics  |  NIST  |  TECH5