With the advent of artificial, face recognition technology in depth and big data-driven learning, show great potential for development, expanding its application scenarios, and gradually fall by the public security and other areas to the commercial areas of payment and verification。 However, face recognition is a double-edged sword, the technology continues to evolve, applications continue to promote the same time, it also brings data leakage, privacy violations and so suffer information security。 First, the issue of information security can not be ignored Face Recognition face recognition technology is not yet mature, it is difficult to eliminate the risk of miscarriage of justice。
First, the lack of interference can fully adapt to the environment of the human face preprocessing algorithm。 Especially in the net, shooting conditions personal image data is more complex, the greater the accuracy of face recognition algorithms for test。
Second, software vulnerabilities by deep learning framework, generate malicious samples affected by pollution and other training data to generate an error or omission on the ratio of template database。
Third, there is an artificial intelligence system calculation error in the decision support process, resulting in an error message。
Insufficient network security protection, data leakage risks exist。
First, the personal image data acquisition part, intelligent monitoring equipment currently there are serious security problems, can easily lead to data loss。
Second, part of the network transmission of data, even with the personal image data for multiple fingerprint biometrics, but also must be converted into binary code for network transmission, faced with the risk of theft。 Third, the data analysis part, Face Recognition artificial intelligence system modules still run on traditional information systems, any link vulnerable, are likely to be hacked。 Fourth, in the data storage areas, lack of personal protective capacity of image databases, data leakage risk of severe。
Neglect of personal image data management, privacy threats。
First, various social platforms, e-commerce, self-service, shooting and other commercial software widely used in the field of portrait collection, intelligent video camera capture personal data at any time without specific populations, large-scale enterprises to continuously collect, personal image data accumulated, for the purpose of profit phenomenon abound misuse of personal data。 Second, in recognition of the core algorithm has independent intellectual property rights of the very few on the market recognition of product quality varies greatly, not unified security technology, system security vulnerabilities abound, personal image data ready to face the danger of leaks。
Third, expand the application of face recognition in the business world, will allow personal data to be image authentication key biological information。 Along with the private sector Internet companies, financial institutions and other continuous access to vast amounts of personal image data, will be personal privacy, information security caused great test。
Fourth, citizen general lack of knowledge of information security and privacy protection awareness, once to unsafe websites, software upload personal image, personal privacy will be exposed。 Second, the face recognition technology and information security to promote the balanced development of face recognition Suggestions establish network and information security supervision system, strengthen network security。
One is to accelerate the development of face recognition technology standard application system。
Face recognition technology for development and security issues, encourages the Government to promote cooperation with leading enterprises face recognition error rate, the rate of correct recognition of technical standards, such as the identification rate study and formulate, develop sub-level and multi-level national security standards and industry safety standard。 Second, establish security assessment and auditing system for face recognition applications。
Approval procedures for the implementation of face recognition application and promotion of products, to ensure that products meet the safety requirements; the same time, guide enterprises to establish facial recognition data security risk management and prevention and control mechanism, the development of corporate compliance system。 Third, the establishment normalization personal image data management system。 Implement the "Network Security Act", increase penalties, and urge them to implement the main responsibility for security; at the same time, regulators should regulate according to the regulations, taking into account development issues, to avoid undue regulatory measures to curb technological innovation and market applications。 Face recognition technology to encourage innovation breakthroughs, help to promote industrial applications。 One is to promote the innovation and development of the core technology of face recognition。
Promote intelligent security popularize a variety of biometric authentication of the base platform, intelligent finance, and other key projects, R & D precision integrated image and video recognition, biometrics, code recognition and other technologies of face recognition products。
The second is to strengthen the network security technology research。 Face recognition promoting enterprise network security and business cooperation, expand the network security technology developed for face recognition。 Third is to promote market-oriented application of technological achievements。 Encourage innovative enterprises and scientific research institutions, brings together resources to play the role of innovation and entrepreneurship platform to promote the application and popularization of the results of face recognition technology。
Fourth, foster the development of new Face Recognition Format。 For security, finance, payment and other key market needs, ranging from core technology to build the whole industry chain business model intelligence applications。 Improve face recognition related laws and regulations, strengthen the protection of personal privacy。 One is to accelerate the development of personal information protection related laws and regulations, establish control over the information rights of personal information, the right to delete, such as the right to forget, improve personal information rights complaint mechanisms and remedies。
The second is to accelerate big data relevant legislation, promote the introduction of telecommunications and Internet industry data security guidance for the implementation of security principal responsibility for all aspects of the data lifecycle, sort large data government, corporate and personal data powers and responsibilities, to promote data market legal order。 Third is to develop biometric information relevant specifications and management measures to ensure the security of personal image data, standard use。
Fourth, accelerate ancillary provisions "Network Security Act," introduced, clear criteria for the classification level of the network, network security level, network security supervision and inspection of the body and other regulations。