Armatura Algorithm ranks 11th Globally by the National Institute of Standards and Technology in USA

Release Time:2023-02-23   Views:856

Armatura announced today with its independently developed algorithm a ranking of 11th in the Face Recognition  Vendor Test (FRVT) report, released by the USA National Institute of Standards and Technology (NIST) on 2nd  February 2023.


Armatura Algorithm ranks 11th Globally by the National Institute of Standards and Technology in USA


The FVRT report covers 478 face recognition algorithms from over 300 different globally renowned face recognition  developers. Armatura’s algorithm – armatura_003 achieves top marks in both accuracy and efficiency, and is ranked  11th globally on the border benchmark test. The overall benchmarks show that Armatura is on top 3 in USA  following Clearview and Rank One and on top 25 worldwide. It clearly indicates from the performance rankings that,  Armatura is one of the global first tier developers (top 10% globally and top 3 of USA) to deliver top-tier accurate  and efficient face algorithms.


Armatura Algorithm ranks 11th Globally by the National Institute of Standards and Technology in USA


The participated FRVT algorithms are evaluated on different datasets taken from controlled scenarios from VISA,  MUGSHOT to BORDER and uncontrolled scenarios from Kiosk to Wild. The accuracy is measured on False  Non-Match rate (FNMR) across several different datasets. FNMR is the proportion of false rejected cases for a  specified False Match Rate (FMR) where FMR is the proportion of false accepted cases. FMR also refers to False  Acceptance Rate (FAR) while FNMR refers to False Rejection Rate (FRR). Noticeably, Amartura algorithm is  ranked 11th in BORDER category, and is ranked 19th in VisaBorder category (refer to table 1).


Armatura Algorithm ranks 11th Globally by the National Institute of Standards and Technology in USA


Each type benchmark in the metrics is compared with the TOP one from each category, it shows that the accuracy of  Armatura face algorithm is at the same level with global top algorithm from commercial application point of view.  Armatura’s face algorithm is compared with algorithms developed by other renowned companies globally, by  aggregating the accuracy from all benchmark ranks. By summing up all the ranks for an overall accuracy benchmark,  it shows that Armatura algorithm is ranked 26th out of 478 algorithms (refer to table 2).


Armatura Algorithm ranks 11th Globally by the National Institute of Standards and Technology in USA

Armatura Algorithm ranks 11th Globally by the National Institute of Standards and Technology in USA


“The result and the ranking is encouraging to the brand, and is a recognition of our efforts,” said David Li, Chief Technology Officer of Armatura. “All staffs are deeply honoured by the recognition, considering the reputation of NIST. Our innovations have proved that our competitiveness is no less than any renowned corporates in the top tier in the biometrics industry. We shall stay committed to offering people better lives with better biometrics.”


Armatura Algorithm ranks 11th Globally by the National Institute of Standards and Technology in USA


The FRVT report conducted by NIST is aimed at measurement of the performance of automated face recognition  technologies applied to a wide range of civil, law enforcement and homeland security applications including  verification of visa images, de-duplication of passports, recognition across photojournalism images, and  identification of child exploitation victims. The performance reports will include measurements of accuracy, speed,  storage and memory consumption, and resilience. NIST will report the dependence of performance on the properties  of the images and the subjects. In its initial form, FRVT has one assessment track, for face verification. For more information, please refer to: https://pages.nist.gov/frvt/html/frvt11.html#_Status_



Share with:

Back to List

This website uses cookies to store information on your device, cookies can enhance your user experience and help our website work normally.
For more information, please read our Cookie Policy and Privacy Policy.

Accept