NEC Face Recognition engine provides highly accurate results even when there is a mix of mask wearers and non-wearers

NEC today announced the strengthening of its face recognition technology – recognized as the world’s most accurate, with the development of a new face recognition engine that provides high-precision certification even when masks of various colors and patterns are worn. The worldwide outbreak of COVID-19 has had a major impact on people’s lives and economic activities. As part of this, governments around the world are encouraging the use of face masks as a basic measure to combat infections under the New Normal. In conjunction with this, there is a need for face recognition technologies to adapt, as they are widely used for personal identification purposes in mobile devices, workplaces and a wide range of facilities.

See also  Xiaomi launches UV-Light Blocking Sunglasses

NEC recently developed a face recognition engine that is specially designed to recognize subjects wearing face masks. Face recognition operates by extracting feature points, such as the position, shape, and size of a subject’s eyes, nose, and mouth, and matching and identifying detected faces. This engine focuses on areas surrounding an individual’s eyes that are not covered by masks to extract and analyze characteristics. Once this engine detects the presence of a mask in a face image, the most appropriate facial recognition algorithm is used to extract and check the characteristics of each individual. This enables high-precision certification even when there is a mix of mask wearers and non-wearers.

See also  NEC develops AI technology that quickly and automatically updates critical systems without causing service interruption

NEC’s internal evaluation using this engine confirmed a high recognition accuracy, with a rate of over 99.9% in 1:1 verification when masks of various colors and patterns were worn. NEC aims to begin sales of the new engine with products that include NEC Enhanced Video Analytics, which supports integrated solutions that combine face recognition with various image analysis functions, and Bio-IDiom Services, which utilize a variety of face and other biometric information to provide multi-modal biometrics solutions.