Thales makes strong debut in NIST’s FRIF fingerprint biometrics benchmark

New entries to NIST’s benchmark for large-scale fingerprint biometric capture and comparison software from Thales and Innovatrics show significant gains in metrics that align with specific practical matching applications for forensic and civil biometrics.

NIST completed the evaluation of new algorithm submitted by Thales and Innovatrics to the Friction Ridge Image and Features (FRIF) Technology Evaluation (TE) Exemplar One-to-Many (E1N) on May 25 and June 3, respectively.

In FRIF TE E1N’s Class C 10-rolled-finger identification scenario, Thales had the lowest False Negative Identification Rank (FNIR) yet recorded at rank-1 when comparing rolled-to-rolled and plain-to-rolled fingerprints, at 0.0028 and 0.0027. For Class B identification flats, Thales’ algorithm did not make any errors at rank-1.

The Class C results show the algorithm’s utility in typical law enforcement and border management scenarios, while the Class B performance indicates its value for major government ID and public safety programs, according to a company announcement.

Notably, the submission is the first by Thales to NIST’s evaluation of ABIS core algorithms since it was rebranded from FpVTE to FRIF and relaunched in 2025.

Innovatrics’ latest submission made a major gain in the FRIF TE E1N with a dramatic reduction in the similarity score thresholds from the DET curve for Class A, to 0.916 at FPIR 0.001. Innovatrics’ previous submission slashed its FNIR rate at FPIR 0.001 sixfold, but left a threshold of 2,422,541, meaning that only impractically high similarity scores could be considered matches.

Article Topics

biometric testing  |  biometrics  |  fingerprint biometrics  |  Friction Ridge Image and Features (FRIF)  |  Innovatrics  |  NIST  |  Thales Digital Identity and Security