Nowadays, we treat facial recognition as a more secure method to confirm a person's identity on a device or at payment terminals. And facial recognition technology is increasingly used for everything from government surveillance to people's daily life.
However, a new test reported by Fortune triggers debate on the accuracy of facial recognition systems as an artificial intelligence company called Kneron utilized 3D mask successfully fool some facial recognition systems. The test covered airport, train station and convenience store where researchers could wear those high-quality 3D masks that could successfully make purchases from another person's account or complete self-boarding.
And this is not the first time facial recognition technology raises people to concern its accuracy and security. Earlier, the American Civil Liberties Union announced that the technology misidentified 26 California lawmakers, primarily people of color. Another report from the UK found that the police's facial recognition system has an 81 percent error rate.
However, lacking sufficient knowledge of facial recognition technology is the root reason why people doubt it. Facial recognition mainly contains capture, extract and compare and match four parts which will use a camera to collect face images or video streams. Then automatically detect and track faces in the images and carry out a series of related technologies for the detected faces. And face recognition algorithm has deep learning capability that enables it constantly optimized in the process of use and becomes more and more secure.
Furthermore, the face recognition algorithms and face recognition terminal are two critical elements to guarantee facial recognition accuracy. The different algorithms will own various face data and deep learning time, and face recognition cameras have monocular, binocular and 3D structural light three types all will influence the facial recognition accuracy.
Although face technology is not perfect now, the risk will be lower once adopt the superior and high-quality facial recognition terminal. Like the Telpo face terminals shown here which support monocular, binocular and 3D structural light camera and Amazon algorithm. Most importantly, the Telpo face terminal contains a liveness detection function that enhances face recognition accuracy and cuts down the opportunity to be fooled by a 3D mask.
Tag: 3D mask, facial recognition accuracy, security, face terminal, high quality, facial recognition system, algorithm
Brief Introduction: Telpo is a professional smart payment partner who focuses on the ODM service 20 years. It mainly provides the EFT-POS, cash registers, biometric devices, face recognition machines, self-service kiosks, and bus validators. Telpo has served for more than 1000 customers abroad, including government, banks, Telecom operators, police stations, Retail shops, and offices. Telpo adheres to R&D and innovation, aiming to provide more intelligent hardware for global partners.