计算机与现代化Issue(12)：36-40,47,6.DOI:10.3969/j.issn.1006-2475.2023.12.007
基于双向同态加密的深度联邦图片分类方法
Deep Federated Image Classification Method Based on Bilateral Homomorphic Encryption
摘要
Abstract
Concerning the privacy protection and data island problems of traditional machine learning paradigm,combined with deep learning,a deep federated image classification method based on bilateral homomorphic encryption called AFL algorithm is proposed.Firstly,AFL algorithm is a horizontal federated improvement of the VGG neural network.At the same time,a bi-directional Paillier homomorphic encryption mechanism based on the Paillier homomorphic encryption algorithm called Bi-HE mechanism is proposed,which can ensure the privacy and security of the federated system.Secondly,the AFL algorithm pro-poses an adaptive waiting strategy during model aggregation,which can effectively avoids the problem of low aggregation effi-ciency caused by communication blockage.Finally,the experiments using the CIFAR-10 data set have proved that the AFL algo-rithm has better generalization capabilities which can effectively solve the problems of privacy protection and data islands com-pared with the traditional VGG and DenseNet algorithms,and the AFL algorithm is better than the traditional federated learning model in efficiency.关键词
联邦学习/深度学习/人工智能/计算机视觉Key words
federated learning/deep learning/artificial intelligence/computer vision分类
信息技术与安全科学引用本文复制引用
梁天恺,黄康华,刘凯航,兰岚,曾碧..基于双向同态加密的深度联邦图片分类方法[J].计算机与现代化,2023,(12):36-40,47,6.基金项目
国家自然科学基金资助项目(62172111,61672169) （62172111,61672169）
广东省自然科学基金资助项目(2021A1515012233) （2021A1515012233）
