首页|基于广义全变分低秩矩阵恢复的对抗样本防御

基于广义全变分低秩矩阵恢复的对抗样本防御

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
一阶全变分(TV)正则化的低秩矩阵分解在恢复图像结构上表现出优异性能.利用全变分在图像去噪方面的优异性能,提高深度神经网络鲁棒性.然而,尽管一阶全变分正则化可以提高模型鲁棒性,但其过度平滑降低了干净样本的准确率.本文提出一种新的低秩矩阵恢复模型,称为LRTGV,该模型将广义全变分(TGV)正则化引入到重加权低秩矩阵恢复模型.在所构建的模型中,TGV可以在不过度平滑的情况下更好地重建图像纹理信息.重加权核范数和L1范数可以增强全局结构信息.因此,本文所提出的LRTGV模型在破坏对抗噪声结构的同时能增强图像全局结构和局部纹理信息.为解决具有挑战性的最优模型问题,本文提出一种基于交替方向乘子法的算法.实验结果表明,该算法对黑盒攻击具有一定防御能力,并且在图像恢复方面优于现有低秩矩阵恢复方法.
Low-rank matrix recovery with total generalized variation for defending adversarial examples
Low-rank matrix decomposition with first-order total variation(TV)regularization exhibits excellent performance in exploration of image structure.Taking advantage of its excellent performance in image denoising,we apply it to improve the robustness of deep neural networks.However,although TV regularization can improve the robustness of the model,it reduces the accuracy of normal samples due to its over-smoothing.In our work,we develop a new low-rank matrix recovery model,called LRTGV,which incorporates total generalized variation(TGV)regularization into the reweighted low-rank matrix recovery model.In the proposed model,TGV is used to better reconstruct texture information without over-smoothing.The reweighted nuclear norm and L1-norm can enhance the global structure information.Thus,the proposed LRTGV can destroy the structure of adversarial noise while re-enhancing the global structure and local texture of the image.To solve the challenging optimal model issue,we propose an algorithm based on the alternating direction method of multipliers.Experimental results show that the proposed algorithm has a certain defense capability against black-box attacks,and outperforms state-of-the-art low-rank matrix recovery methods in image restoration.

Total generalized variationLow-rank matrixAlternating direction method of multipliersAdversarial example

李文、王恒友、霍连志、何强、陈琳琳、何志权、吴永贤

展开 >

北京建筑大学理学院,中国 北京市,100044

华南理工大学计算机科学与工程学院,中国 广州市,510006

北京建筑大学大数据建模与技术研究所,中国 北京市,100044

中国科学院空天信息研究所,中国 北京市,100094

广东省智能信息处理重点实验室,,中国 深圳市,518060

展开 >

广义全变分 低秩矩阵 交替方向乘子法 对抗样本

National Natural Science Foundation of ChinaOutstanding Youth Program of Beijing University of Civil Engineering and Architecture,ChinaShenzhen Stability Support General Project(Type A)

62072024JDJQ2022080520200826104014001

2024

10.1631/FITEE.2300017

信息与电子工程前沿(英文)
浙江大学

信息与电子工程前沿(英文)

CSTPCD
影响因子：0.371
ISSN：2095-9184
年,卷(期)：2024.25(3)
  • 37