当前位置: X-MOL 学术Int. J. Intell. Syst. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Fighting fire with fire: A spatial–frequency ensemble relation network with generative adversarial learning for adversarial image classification
International Journal of Intelligent Systems ( IF 7 ) Pub Date : 2021-01-26 , DOI: 10.1002/int.22372
Wenbo Zheng 1, 2 , Lan Yan 2, 3 , Chao Gou 4 , Fei‐Yue Wang 2
Affiliation  

Adversarial images generated by generative adversarial networks are not close to any existing benign images, and contain nonrobust features that have been identified as critical to the robustness of a machine learning model. Since adversarial images have an underlying distribution that differs from normal images, these kinds of images can offer valuable features for training a robust model. To deal with these special features, we focus on a novel machine learning task of adversarial images classification, where adversarial images can be used to investigate the problem of classifying adversarial images themselves. In the setting of this novel task, adversarial images are the ONLY kind of data used in training and testing, rather than not just a set of testing images as usual. To this end, we propose a novel spatial–frequency ensemble relation network with generative adversarial learning. First, we present a spatial–frequency ensemble representation learning to extract the feature of training images. Second, we design a meta‐learning‐based relation model to gain the relationship between images. Third, to achieve a robust model, we utilize generative adversarial learning and transform the relationship into a Jacobian matrix. Finally, we design a discriminator model that determines whether an adversarial image is from the matching category or not. Experimental results demonstrate that our approach achieves significantly higher performance compared with other state‐of‐the‐arts.

中文翻译:

用火扑救:具有生成性对抗学习的时频集成关系网络,用于对抗性图像分类

由生成对抗网络生成的对抗图像与任何现有的良性图像都不接近,并且包含已被识别为对机器学习模型的鲁棒性至关重要的非鲁棒特征。由于对抗性图像的基本分布与正常图像不同,因此这些类型的图像可提供有价值的功能,以训练健壮的模型。为了应对这些特殊功能,我们将重点放在对抗性图像分类的一种新颖的机器学习任务上,其中对抗性图像可用于研究对抗性图像本身进行分类的问题。在这项新颖任务的设置中,对抗性图像只是训练和测试中使用的唯一一种数据,而不仅仅是像往常一样不仅仅是一组测试图像。为此,我们提出了一种具有生成对抗性学习的新型时空集合关系网络。首先,我们提出一种空间频率合奏表示学习,以提取训练图像的特征。其次,我们设计了一个基于元学习的关系模型来获取图像之间的关系。第三,为了获得一个健壮的模型,我们利用生成式对抗学习并将该关系转换为雅可比矩阵。最后,我们设计了一个判别模型,该模型可以确定对抗图像是否来自匹配类别。实验结果表明,与其他最新技术相比,我们的方法可实现更高的性能。我们设计了一个基于元学习的关系模型来获取图像之间的关系。第三,为了获得一个健壮的模型,我们利用生成式对抗学习并将该关系转换为雅可比矩阵。最后,我们设计一个判别器模型,该模型确定对抗图像是否来自匹配类别。实验结果表明,与其他最新技术相比,我们的方法可实现更高的性能。我们设计了一个基于元学习的关系模型来获取图像之间的关系。第三,为了获得一个健壮的模型,我们利用生成式对抗学习并将该关系转换为雅可比矩阵。最后,我们设计了一个判别模型,该模型可以确定对抗图像是否来自匹配类别。实验结果表明,与其他最新技术相比,我们的方法可实现更高的性能。
更新日期:2021-03-31
down
wechat
bug