摘要
在人脸识别为人们生活、工作带来便利的同时,也出现了技术滥用、侵犯隐私等问题,其中以公共场所的人脸识别应用问题最为突出。在对人脸识别技术特征、公共场所分类方法、人脸识别应用治理现状进行深入分析的基础上,提出从使用主体角度划分公共场所分类的新方法。并综合应用目的、底库规模、覆盖密度、管理水平和网络环境等5方面风险影响要素,结合层次分析赋权法,给出了“人脸识别应用综合风险值”的量化计算方法。进而使用风险评估矩阵方法,综合考虑“人脸识别应用综合风险值”和“公共场所分类”2大维度,将公共场所人脸识别应用风险等级划分为5级,并从事前评估、通用要求和分级要求等3个方面提出开展人脸识别应用治理的具体建议。
While face recognition brings convenience to people's life and work,there are also problems such as technology abuse and privacy invasion,among which the face recognition application problem in public places is the most serious.In this paper,on the basis of in-depth analysis of face recognition technology features,public place classification methods,face recognition application governance status,we put forward a new method of"public place classification"from the perspective of"user".This paper gives the quantitative calculation method of"comprehensive risk value of face recognition application"by integrating five aspects of risk factors,such as application purpose,base size,coverage density,management level and network environment,combined with hierarchical analysis weighting method.This paper uses the risk assessment matrix method to divide the"risk level of face recognition application in public places"into 5 levels by taking into account the two dimensions of"comprehensive risk value of face recognition application"and"classification of public places".It also puts forward specific suggestions on face recognition application governance from three aspects,such as pre-evaluation,general requirements and grading requirements.
出处
《信息技术与标准化》
2023年第9期101-107,共7页
Information Technology & Standardization
基金
上海市市场监督管理局科技项目“人脸识别技术在公共场所分类应用及关键技术标准研究”,项目编号:2022-62。
关键词
公共场所
人脸识别
分级分类
风险
public places
face recognition
classification
risk