TAIPEI, Taiwan–(BUSINESS WIRE)–CyberLink Corp. (5203.TW), a pioneer in AI and facial recognition technologies, has achieved the highest ranking in the Video Convenience category of the most recent Face Analysis Technology Evaluation (FATE) tests performed by NIST (National Institute of Standards and Technology). The evaluation tested the performance of 82 passive, software-based presentation attack detection (PAD) algorithms and resulted in CyberLink’s FaceMe® excelling at the top of all three data sets within the Convenience category.
What is NIST FATE PAD Testing?
Under biometric face technology evaluations, NIST has two types of testing scopes: FRTE (Face Recognition Technology Evaluation) and FATE (Face Analysis Technology Evaluation). Testing that pertains to identity verification (who is in an image) falls under FRTE, which CyberLink’s FaceMe is already top ranked for in 1:1 and 1:N evaluations. Testing that involves the processing and analysis of images (what is in an image) falls under FATE. With NIST’s latest internal report, focuses on FATE Presentation Attack Detection (PAD), the performance of passive, software-based PAD for two use cases: impersonation and evasion.
What is Passive Presentation Attack Detection (PAD) and Why is Liveness Detection Testing Important?
A presentation attack occurs when a person tries to “trick” the facial recognition system’s camera so they may either evade a positive identification in a database or impersonate another individual. With passive PAD, a person is not required to do any more than present themselves to the camera, whereas active PAD requires a person to speak, turn their head, etc. to verify their liveness. Some common methods of presentation attacks within impersonation and evasion use cases include holding up a printed photo, presenting video from a smartphone or tablet, or even wearing elaborate silicone masks. Without liveness detection, in the form of PAD, a false positive will occur and the true identity bearer is subject to fraudulent activity. Committed to the prevention of fraud and enhancement of the customer experience, preventing impersonation is of paramount importance to CyberLink as demonstrated by the test results.
Understanding CyberLink’s Impressive PAD Test Results
NIST’s PAD test is separated into two categories: convenience and security. Under convenience, the True Acceptance Rate (TAR) is fixed to 99% and sorted by True Rejection Rate (TRR). In NIST’s FATE PAD video convenience category FaceMe® ranked first, based on the results of three different video tests in which FaceMe® achieved 100% True Rejection Rate, correctly stopping 100% of presentation attacks, when the True Acceptance Rate was set 99%.
With FaceMe’s already superior FRTE rankings, perfect iBeta’s PAD Level 2 test score, and now top NIST FATE PAD ranking, CyberLink’s dedication to excelling at not only facial recognition accuracy, but also liveness detection is clear. “Facial recognition does not work if it can be easily fooled,” said Jau Huang, Chairman and CEO of CyberLink. “CyberLink has invested heavily to ensure that FaceMe’s presentation attack detection is second to none, now solidified with NIST’s evaluation results.”
For more information on NIST’s Face Analysis Technology Evaluation (FATE) Presentation Attack Detection (PAD), please visit NIST’s official webpage: https://pages.nist.gov/frvt/html/frvt_pad.html
To learn more about FaceMe®, please visit: https://www.cyberlink.com/faceme
Founded in 1996, CyberLink Corp. (5203.TW) is the world leader in multimedia software and AI facial recognition technology. CyberLink addresses the demands of consumer, commercial and education markets through a wide range of solutions, covering digital content creation, multimedia playback, video conferencing, live casting, mobile applications and AI facial recognition. CyberLink has shipped several hundred million copies of its multimedia software and apps, including the award-winning PowerDirector, PhotoDirector, and PowerDVD. With years of research in the fields of artificial intelligence and facial recognition, CyberLink has developed the FaceMe® Facial Recognition Engine. Powered by deep learning algorithms, FaceMe® delivers the reliable, high-precision, and real-time facial recognition that is critical to AIoT applications such as smart retail, smart security, and surveillance, smart city and smart home. For more information about CyberLink, please visit the official website at www.cyberlink.com.
All companies and product names mentioned herein are for identification purposes only and are the sole property of their respective owners. Copyright © 2023 CyberLink Corp. All rights reserved.