FRC Adapts Biometric Retail Surveillance Platform to Face Mask Challenge

The Face Recognition Company (FRC) has updated its technology to improve its performance during the pandemic. The upgraded solution is now able to recognize individuals based on partial information, which means that it is able to identify people who are wearing masks.

“Face coverings are likely to be part of our normal daily life for a long time to come – even after mandatory restrictions are lifted,” said FRC CEO Tim Noest. “This new capability allows our system to identify the person behind the mask and help security staff operate more effectively in this new normal.”

The FRC surveillance system is designed for retail outlets and other commercial environments, including sports venues and hotels. The system monitors the crowd to search for people on a watchlist of shoplifters and other persons of interest, and will send an alert to store staff if it spots anyone suspicious. The notifications are delivered through a mobile app.

The system can also be configured to enforce mask compliance. In those cases, it will send an alert if someone is not wearing a mask, allowing venue operators to step in quickly to reduce the risk of exposure to COVID-19.

“Wearing a mask is now compulsory in many retail and public transport environments, and we are providing businesses with an extra pair of eyes when it comes to ensuring that customers are following the rules,” added Noest.

FRC first unveiled its multi-camera surveillance system at the International Security Expo last December. It is now one of several companies that has started offering mask detection capabilities during the pandemic, joining a list that already includes CyberLink, Innovatrics, and ZKTeco.

The NIST has reported that facial recognition algorithms are not as accurate when asked to identify people wearing masks, although it did note that the algorithms it tested were built before the pandemic and were consequently not designed with masks in mind. FRC claims that its updated system has the same level of accuracy as its old one.

–

September 24, 2020 – by Eric Weiss