The algorithm underwent 300 original Liveness tests that were performed with real faces, with a false positive error rate of 1 percent. All 1,000 attempted biometric spoof attacks were detected by the Oz Forensics’ technology, according to the announcement.
Oz Forensics’ biometric platform is designed particularly to prevent biometric spoofing and deepfake attacks using Liveness detection.
According to the company, the Oz Liveness algorithm does not require individuals to look directly into the camera for identification and has processing speeds of up to one second.
“Oz Liveness AI algorithms check shots from the video and track dozens of parameters, [including] presence of glare and reflections, micromotions, pulse, etc,” explained the company’s CEO, Artem Gerasimov.
“We trained the system on tens of thousands of attacks. We work closely with 3D mask manufacturers and are constantly looking for new samples and challenges,” he added.
The iBeta biometric testing laboratory is accredited by the National Institute of Standards and Technology (NIST) under the National Voluntary Laboratory Accreditation Program (NVLAP).
iBeta was the first laboratory accredited by the FIDO Alliance under the FIDO Alliance Biometric Component Certification Program.
BONAFiDEE unveils new liveness test tool
UK-based regulatory technology firm BONAFiDEE has released a liveness test tool to strengthen its Visual Biometric Verification check.
Visual Biometric Verification allows for comparison between a real-time image and an ID document to verify the identity of an individual.
The new update introduces the capability to record a short video, quote a unique phrase within a limited time period, to prove that the user is ‘present’ in front of the camera
“We believe that with the launch of our innovative ‘liveness test,’ we have been successful in creating a solution that protects both organizations and their customers against the emerging deepfake fraud vulnerability,” commented BONAFiDEE Founder, Francis Lang. “It strengthens BONAFiDEE’s already robust identification process, bringing your customer one step closer”.
Biometric experts discuss presentation attack instruments
The experts defined presentation attacks and described different standards and frameworks for biometric solutions to be able to effectively perform Presentation Attack Detection (PAD).
In particular, Dunstone and Pope explore the ISO/IEC 30107 standard, and how the document broadly defines the scope and requirements of PAD tools.