Research on Face Image Encryption Based on Deep Learning

2018 4th International Conference on Environmental Science and Material Application

生态环境科学;材料科学

Qin, Yanyan^1 ; Zhang, Chennan^1 ; R., Liang ; M., Chen

School of Information Science and Technology, Hainan University, Haikou, China^1

关键词: Automatic extraction;    Data technologies;    Discriminative features;    Face recognition algorithms;    Face recognition technologies;    Feature extractor;    Privacy and security;    Security authentication;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/252/5/052007/pdf
DOI  :  10.1088/1755-1315/252/5/052007

来源: IOP

【 摘 要 】

With the development of artificial intelligence and big data technology, the requirements for information security are increasing, and the role of biometrics in network security and information security authentication has also increased. Face recognition technology has been widely applied in many Internet payment platforms. This paper proposes a face recognition algorithm based on improved deep network automatic extraction feature, which can extract the discriminative features of the target more accurately and encrypt the face image to ensure the privacy and security of face recognition. In this paper, an automatic deep feature extractor is generated by preprocessing and fine-tuning, and then the hyperchaotic image is encrypted. Several common face databases are used to test in this algorithm and this results show that the algorithm has more availability than the traditional and general deep learning methods in terms of performance.