Aware has updated its facial recognition algorithm to improve its performance for people wearing masks. The new algorithm is able to match an image of a person wearing a mask to an unmasked image of that person in a database, which means that users will not need to re-register their faces to enjoy the benefits of masked recognition.
The feature is being introduced in the latest version of the mobile Knomi platform. Knomi is a biometric authentication solution that offers support for face and voice recognition, backing up both modalities with robust liveness detection capabilities. Aware released the on-device version of Knomi back in September, and then passed the second level of iBeta’s Presentation Attack Detection test later that same month.
According to Aware, Knomi version 2.13 does not sacrifice speed for accuracy, enabling fast mobile onboarding and passwordless authentication both during and after the pandemic.
“We’re proud that we’ve been able to deliver this important innovation to the biometric authentication market at such a difficult time for society,” said Aware President and CEO Bob Eckel. “Users need to be confident that facial images will match accurately, whether masks are on or off. With Knomi 2.13 our customers can be sure that their identity-matching and authentication procedures are solid and irrefutable.”
Eckel recently joined the Board of Directors for the International Biometrics + Identity Association (IBIA), while Aware itself has bolstered its biometric portfolio through the acquisition of the AFIX suite of products from Maxar Technologies. Aware is also one of the many companies that has added mask recognition capabilities in response to COVID-19, noting that the increased use of masks has created problems for many facial recognition algorithms. In that regard, Aware is following in the footsteps of companies like Cognitec, SAFR, and Invixium, all of which have all upgraded their facial recognition systems in the past few months.
December 8 2020 – by Eric Weiss