计算机工程与应用2024，Vol.60Issue(19)：259-267,9.DOI:10.3778/j.issn.1002-8331.2306-0323
遗忘学习前置的反后门学习方法研究
Research on Anti-Backdoor Learning Method Based on Preposed Unlearning
摘要
Abstract
The anti-backdoor learning(ABL)method can detect and suppress backdoor generation in real time during model training with poisoned datasets,and finally obtain a benign model.However,the ABL method suffers from the problem that the backdoor samples and benign samples cannot be effectively isolated and the efficiency of backdoor elimi-nation is not high.To this end,an anti-backdoor learning method based on preposed unlearning(ABL-PU)is proposed,which adds a purification operation to the training samples in the isolation phase to achieve the goal of effective isolation of benign samples,and adopts a paradigm of backdoor unlearning and model retraining in the elimination phase,and intro-duces unlearning coefficients to achieve efficient backdoor elimination.On the CIFAR-10 dataset,against the classical backdoor attack method BadNets,the anti-backdoor learning method based on preposed unlearning improves the benign accuracy rate by 1.21 percentage points and decreases the attack success rate by 1.38 percentage points compared with the anti-backdoor learning method(the baseline method).关键词
后门攻击/反后门学习/数据提纯/遗忘学习前置/遗忘系数Key words
backdoor attacks/anti-backdoor learning/data purification/preposed unlearning/unlearning coefficient分类
信息技术与安全科学引用本文复制引用
王晗旭,李欣,许文韬,斯彬洲..遗忘学习前置的反后门学习方法研究[J].计算机工程与应用,2024,60(19):259-267,9.基金项目
中国人民公安大学网络空间安全执法技术双一流创新研究专项(2023SYL07). （2023SYL07）
