Kneron has released a real face recognition solution designed for smartphone unlocking applications. Even an entry-level smartphone, as long as it has a front camera with a million pixels or more, after installing Kneron’s real face recognition software, through the deep learning algorithm, it can not only accurately identify the facial features, but also judge whether it is a real person, and exclude the use of photos. , the risk of video unlocking. The unlocking speed is as fast as 0.2 seconds, and the error value is less than 1/10,000. One mobile phone can also recognize multiple faces. Kneron’s real face recognition solution has been adopted by mobile phone manufacturers and will be rolled out to the global market through its partner General Mobile (GMobi).
Kneron’s Real Face Recognition Solution Makes Unlocking Smartphones Easier
Liu Juncheng, founder and CEO of Kneron, said that traditional face recognition technology still has many security loopholes, such as photos and videos that can be cracked. If you want to further identify whether it is a real person, you often need to match high-end hardware specifications, such as infrared lenses. , high-resolution lenses, and even hardware design changes. Kneron provides an innovative real-person face recognition solution that enables entry-level mobile phones to use face unlocking without the need to update mobile phone design or add new hardware components. It only requires software updates to quickly achieve high accuracy and high accuracy. Secure real face recognition.
Wu Baiyi, CEO of GMobi, said that GMobi has many years of experience in professional mobile phone software services, assisting manufacturers around the world to update their mobile devices and provide a variety of advanced services. Currently, there are 250 million global end users. Kneron’s innovative real-person face recognition solution can help manufacturers quickly introduce artificial intelligence technology into mobile phones, bringing a new development to mobile phone face unlocking. GMobi hopes to promote this solution to the global market through the cooperation with Kneron, and make the real face unlock function of mobile phones popularized and become a function that everyone can use.
Kneron’s real face recognition solution combines sophisticated face detection and recognition technology and deep learning algorithms, unlocking speed as fast as 0.2 seconds, the error value is less than 1/10,000, and a mobile phone can recognize multiple faces. In order to determine whether it is a real person, three different technical solutions can be selected: optical sensing analysis, biological behavior identification, and background depth detection. Taking optical sensing analysis as an example, in addition to detecting facial features, it can also analyze the changes of light and shadow on the face to determine whether it is a three-dimensional face. Biological behavior identification is a combination of behavior detection, such as blinking, to determine whether it is a real person. In addition, if the front lens of the mobile phone already has an autofocus function, the background depth of field detection scheme can also be used to analyze the scene depth of field to eliminate the situation of unlocking with a flat image.