Seattle — April 6, 2021 — SAFR from RealNetworks, Inc. (NASDAQ: RNWK) has announced a new version of SAFR, the world’s foremost facial recognition solution for live video, offering accurate, fast, unbiased face recognition and additional computer vision features. Available now, SAFR version 3.4 introduces new passive liveness detection and anti-spoofing features, for both masked and unmasked faces, to enhance security for face biometric authentication solutions. The new version also includes SMS watchlist alarms.
Passive Liveness Detection
Now further hardened against spoofing, SAFR’s AI-powered liveness detection can quickly (within 0.3 seconds) and accurately (95.27% True Positive Rate) verify that a real, live person is in front of any standard RTSP or USB camera, and not a photo or video clip being presented. In version 3.4, the SAFR algorithm analyzes texture and context, based on the RGB visual spectrum field from a standard 2D camera – be it an IP camera embedded in an access control terminal, an ATM camera, or a USB or laptop camera used to authenticate the user.
Ideally suited for a broad range of applications such as touchless access control or authentication for applications like electronic wallet or online test-taking, SAFR’s passive liveness detection feature is very intuitive to operate since users are not required to do anything specific such as turn their faces, look left, or right, smile, or even take off their mask if they are wearing one.