The ability to identify someone’s face is a talent we all rely on, both for our social interactions, as well as for our own safety and protection. This is an innate ability, but we don’t all possess it at the same degree. According to research made by the University of York, humans can recognise 5,000 faces. A super-recogniser may be able to memorise and recall 10,000. But when asked to match a face and a picture, even the best-trained are accurate only 80% of the time.
As we look for secure and convenient ways to identify people, facial recognition is the least intrusive and most accessible form of biometric identification: contactless, fast and reliable. When asked to match a face and a picture, the top performing 100 Facial Recognition algorithms tested by the United States NIST agency in 2020 are accurate 95% of the time.
Facial recognition technologies are still making headway, and identity applications are constantly being improved. The industry’s R&D teams are blending their expertise in identity technology, biometrics and cybersecurity to design better performing solutions. Continuous improvements are being made regarding accuracy, speed, privacy best practices, security issues, standards and user experience.
This e-book explores the different aspects of this fascinating emerging technology.
