SAFR facial recognition update includes enhanced mask use features and face detection
A new version of SAFR facial recognition software from RealNetworks has been launched with upgrades to its COVID-19 response features and a new high-sensitivity face detection feature, the company has announced.
SAFR 3.0 provides the new highly sensitive face detector as a default setting, and the company says it detects 95.1 percent of faces and recognizes them biometrically at a 98.85 percent rate for face covered by PPE masks, including non-surgical masks with different patterns. The statistics are based on surveillance-type videos with faces in motion. Detection efficiency for multiple faces simultaneously in the frame has also been improved to ensure high-speed detection and facial recognition.
“With the 3.0 release, SAFR continues its track record of maintaining high-accuracy under challenging real-world conditions, including the new norms brought by this global pandemic,” said Brad Donaldson, VP Computer Vision, SAFR. “SAFR’s accuracy improvements will enable customers to deploy face-based contactless secure access without requiring removal of PPE and ensure persons of interest and registered VIPs don’t go unrecognized while wearing face masks.”
A mask detection dashboard is also built into the new version, allowing customers to track anonymized mask usage rates for different demographics or areas.
A representative of NTT Docomo, which sells SAFR to the Japanese market, states in the announcement that the new functionalities can help improve “the safety and security of people and organizations.”
A deployment of SAFR biometrics was recently announced at the San Diego – Tijuana terminal of Tijuana International Airport.