Dusun Electron has developed a new access control product that provides identity confirmation using biometrics and a machine learning algorithm.The product incorporates alternate access methods including fingerprint, QR code and mobile phone apps. A facial recognition system captures facial images via digital camera and other sensors and verifies the identity of a person using a machine learning algorithm. Facial recognition has been perceived as the most natural and accurate way to confirm identity and prevent unauthorized access. It is also the most difficult to thwart.Dusun's facial recognition access control product offers an accurate recognition rate (99.9%), fast recognition time (0.5 seconds) and can detect faces up to 1 meter away. This product features a self-learning algorithm that can even compensate for face changes from aging. Dual digital cameras for 3D facial recognition along with an infrared radiation (IR) sensor for body recognition are provided to prevent identity spoofing using photos. This product can adapt to bright and dark lighting with its anti-backlight capability and also provides a video intercom for instances when person-to-person communication is required.”Access control and personnel identification are becoming very important to ensure that our living and work spaces are secure. Utilizing facial recognition as the basis of an access control system of IoT for granting physical access has many advantages over card reader-based access,” said Benny Chan, chief executive officer of Dusun. “Dusun is pleased to develop this highly accurate and efficient facial recognition access control product to add secure functionality and superior performance in an IoT access control offering.”