Face recognition is an important part of biometric identification. In the late 1990 s, it entered the main application stage, mainly in the technology of America, Germany and Japan. The related technologies in China in recent years are summarized. Facial recognition is also a serious security risk, and it is attacked or maliciously attacked at our convenience. Therefore, for the application of face recognition technology and face database, the structure of face recognition model must be strictly controlled, and face data should be strictly protected by technology and legislation.
The application of state biometrics in face recognition technology is mainly composed of fingerprint, voiceprint, face, iris and other inherent physiological characteristics and behavioral characteristics (such as handwriting, voice, gait, etc.). Compared with other biometric technologies, face recognition has the characteristics of high accuracy, non-contact and high speed, and is increasingly applied to government, military, finance, e-commerce, security, entertainment and other fields.
The New Generation Identity Database (NGI) of the US Federal Bureau of Investigation (FBI) has added more than 6,543.8 billion personal fingerprint information records, including various biometric data, and collected more than half of American adults. The Australian Immigration and Border Protection Agency began to use a new entry system, that is, "free", set up an electronic scanning station, and used biometric technology to identify the faces, eyes, irises and fingerprints of immigrant tourists, replacing the tradition of quitting passports. In recent years, facial recognition technology in China has also developed rapidly. As the core technology of intelligent society, face recognition is prosperous and widely used, and various potential security risks are inevitable. For example, at the "March 15" party in 20 17, the host only took a selfie of the audience with the support of on-site technicians, and the "brush face login" authentication system was successful.
The attack means and risks faced by face recognition technology, such as the invasion, leakage, theft and illegal trading and utilization of human face data, are the main risks faced by face recognition technology and database. If these security risks can't be solved by technology, policies, laws and regulations, we have to highlight the bottleneck of the development of face recognition technology.
The attack of face recognition algorithm. The identity of a legitimate user needs to be a prerequisite for obtaining facial photos, videos and other data of a legitimate user. China's security team obtained more threats and attacks, and the complete face photos of different users bypassed the identification system. This attack means that deepening the application of image recognition leads to escape attack and data pollution attack. It does not depend on a specific deep learning model, and it is effective for current mainstream deep learning frameworks (such as Tensorflow and Caffe).