Innovatrics’ SmartFace biometrics updated with passive liveness detection on live video
Innovatrics SmartFace passive liveness detection on live video
Innovatrics has released a new upgrade of its SmartFace biometric system to offer passive liveness detection and anti-spoofing capabilities on live video feeds.
The company made the announcement in a blog post, explaining that as spoofing attempts increase in both frequency and sophistication, liveness detection is becoming a necessary feature for facial recognition video systems.
According to the company, fraudsters attempting to spoof these systems often use a combination of an illegally obtained ID and a selfie printout.
But these attempts can also get more elaborate, with attackers deploying 3D masks, puppets, wax heads, sculptures, and even robotic heads with fake muscles to spoof facial recognition systems.
The new version of SmartFace was tested against several types of biometric spoofing attempts and was trained using hundreds of thousands of images of real faces and real attacks to make the liveness detection feature as accurate as possible.
After detecting a biometric spoofing attack, SmartFace will now send a notification to an administrator in real-time.
Innovatrics has been working on passive liveness detection for some time now, typically for authentication scenarios, claiming earlier this month the technology can deliver onboarding completion rates of 99 percent, as opposed to 63 percent for active liveness detection.
The company’s facial recognition platform is reportedly among the most accurate and efficient available commercially, with Innovatrics recently announcing SmartFace can process videos 25 times faster than it would take to watch them.
For more information about Innovatrics’ goals and future biometric plans, you can read this guest post by the company’s Head of Global Sales Peter Martis.
Article Topics
biometric liveness detection | biometrics | facial recognition | Innovatrics | passive facial liveness | real-time biometrics | spoof detection | video surveillance
Comments