当前位置: X-MOL 学术Secur. Commun. Netw. › 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.)
Boosting Adversarial Attacks on Neural Networks with Better Optimizer
Security and Communication Networks Pub Date : 2021-06-07 , DOI: 10.1155/2021/9983309
Heng Yin 1 , Hengwei Zhang 1 , Jindong Wang 1 , Ruiyu Dou 1
Affiliation  

Convolutional neural networks have outperformed humans in image recognition tasks, but they remain vulnerable to attacks from adversarial examples. Since these data are crafted by adding imperceptible noise to normal images, their existence poses potential security threats to deep learning systems. Sophisticated adversarial examples with strong attack performance can also be used as a tool to evaluate the robustness of a model. However, the success rate of adversarial attacks can be further improved in black-box environments. Therefore, this study combines a modified Adam gradient descent algorithm with the iterative gradient-based attack method. The proposed Adam iterative fast gradient method is then used to improve the transferability of adversarial examples. Extensive experiments on ImageNet showed that the proposed method offers a higher attack success rate than existing iterative methods. By extending our method, we achieved a state-of-the-art attack success rate of 95.0% on defense models.

中文翻译:

使用更好的优化器增强对神经网络的对抗性攻击

卷积神经网络在图像识别任务中的表现优于人类,但它们仍然容易受到对抗样本的攻击。由于这些数据是通过向正常图像添加难以察觉的噪声来制作的,因此它们的存在对深度学习系统构成了潜在的安全威胁。具有强大攻击性能的复杂对抗样本也可以用作评估模型鲁棒性的工具。但是,在黑盒环境中可以进一步提高对抗性攻击的成功率。因此,本研究将改进的 Adam 梯度下降算法与基于迭代梯度的攻击方法相结合。然后使用提出的 Adam 迭代快速梯度方法来提高对抗性示例的可转移性。ImageNet 上的大量实验表明,所提出的方法比现有的迭代方法提供了更高的攻击成功率。通过扩展我们的方法,我们在防御模型上实现了 95.0% 的最先进的攻击成功率。
更新日期:2021-06-07
down
wechat
bug