Algorithm development at Cognitec continues to engineer the  optimal balance between speed and accuracy of face matching processes.  

The latest results of the U.S. NIST Face Recognition Vendor Test for identification tasks show the  Cognitec algorithm in the best position of all algorithms when relating the template generation  time to the false negative identification rate (miss rate) for mugshot databases. 

The identification test addresses the largest market for face recognition applications, including  detection of duplicates in image databases, and fraud detection during passport and driver’s  license applications. These tests apply a very high matching threshold, where only 0.3 % (3/1000)  of probes without a mate in the gallery produce a false hit—one of the most difficult face  recognition tasks.  

“We are proud to also see significant accuracy advances in comparison to the algorithm we  submitted to the last test in early 2021,” says Dr. Thorsten Thies, Director of Algorithm  Development. “For the test with 12 Million mugshot images, Cognitec achieved rank 24 of 165  algorithms with a match rate of 98.5 %. In the test with 1.6 Million mugshots, with a 99.4 % match  rate, we ranked 26 of 299 algorithms. These results show remarkable performance consistency, regardless of database size.” 

The September 2021 test evaluates matching algorithms from 85 vendors. Cognitec submitted a  new algorithm with a revised face finder, called cognitec-005 in the report.