Cubox tops NIST face biometrics test for visa kiosks, Trueface third in matching speed

The National Institute of Standards and Technology (NIST) has recently published the results of its most recent Face Recognition Vendor Test (FRVT) 1:N, aimed at assessing the effectiveness of algorithms in performing comparisons of a biometric template against various databases.

Cubox submitted the algorithm ranked third in the 1:1 tests’ ‘VisaBorder’ category, and followed that up by topping the ‘Visa Kiosk’ category in the NIST FRVT 1:N.

The NIST FRVT 1:N test is separated into two categories:  identification, focusing on biometric access control applications, and investigation, intended to test systems designed to identify criminals, missing people, and wanted offenders.

The Cubox algorithm scored the highest biometric accuracy in the September 21 visa kiosk photo test with a 0.06 percent false identification rate when searching datasets from 1.6 million visa border images among 344 algorithms from 240 global companies and institutions. The company also scored near the top of the face aging border recognition category.

“The achievements proved by the NIST FRVT 1:N/1:1 test can be the historical success of Korea, which has been considered as barren land of artificial intelligence,” commented Cubox CEO Nam Un-sung.

“As a global leader in artificial intelligence, Cubox will continue on aggressive investment and R&D and will take the lead in creating a digital infrastructure by incorporating artificial intelligence technology into our daily lives.”

The company already offers a series of face recognition-focused products in Korea, including ones for airport automation, subway payment and entrance, and the contactless access security market.

Trueface ranks third in template match speed category

According to the recent FRVT test results, Trueface executed a 1 to 1 biometric template comparison in as little as 186 nanoseconds, ranking third globally in the template match speed category.

“However, more often than not, we are performing 1 to N identity verification,” explained Cyrus Behroozi, senior computer vision software developer at Trueface.ai in a blog post.

“In these situations, we do not simply compare two biometric templates against each other, but instead, compare a biometric template against a database of potentially millions of biometric templates.”

In such cases, Behroozi explained, the template match comparison must be performed millions of times in order to determine if there was a single match in the database.

“Often, this search operation needs to be performed at least 30 times a second for a single 30 fps video camera. Once you add multiple camera feeds or extended video to the equation, it’s easy to see why match speed is so critical!”

According to the developer, if the Trueface algorithms were to be used to search every frame of a seven-minute video against a database of 1 million identities, Trueface would complete that search in 39 minutes. Based on the NIST test result, Behroozi said it would take Trueface’s closest American competitor 1 hour and 7 minutes to complete the search.

“In certain scenarios, 28 minutes of extra time can make all the difference,” he added.

Trueface was recently acquired by Pangiam to aid the company in providing customers with a safer, quicker, and more personalized travel experience.

Article Topics

accuracy  |  biometric identification  |  biometric matching  |  biometric testing  |  biometrics  |  CUBOX  |  face biometrics  |  facial recognition  |  NIST  |  research and development  |  Trueface