用对比学习提升神经网络抗对抗攻击能力
Enhancing Adversarial Robustness of Deep Neural Networks Through Supervised Contrastive Learning
- 结合监督对比学习与边界约束损失，优化特征空间结构
- 在CIFAR-100上，ResNet-18模型对抗准确率显著提升
- 适合关注模型鲁棒性提升的研究者与工程应用
对抗攻击通过引入人眼无法察觉的扰动，导致卷积神经网络误分类，暴露出特征表示和决策边界的弱点。本文提出一种新框架，融合监督对比学习与基于间隔的对比损失，以增强对抗鲁棒性。监督对比学习通过聚合同类样本嵌入、分离异类样本嵌入，改善特征空间结构；基于间隔的对比损失受支持向量机启发，显式施加约束，构建具有清晰间隔的鲁棒决策边界。在使用ResNet-18作为主干网络的CIFAR-100数据集上，实验表明该方法在快速梯度符号法（FGSM）攻击下显著提升了对抗准确率。
原文摘要 · Abstract (English)
Adversarial attacks exploit the vulnerabilities of convolutional neural networks by introducing imperceptible perturbations that lead to misclassifications, exposing weaknesses in feature representations and decision boundaries. This paper presents a novel framework combining supervised contrastive learning and margin-based contrastive loss to enhance adversarial robustness. Supervised contrastive learning improves the structure of the feature space by clustering embeddings of samples within the same class and separating those from different classes. Margin-based contrastive loss, inspired by support vector machines, enforces explicit constraints to create robust decision boundaries with well-defined margins. Experiments on the CIFAR-100 dataset with a ResNet-18 backbone demonstrate robustness performance improvements in adversarial accuracy under Fast Gradient Sign Method attacks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。