Despite the fact that adversarial training has become the de facto method for improving the robustness of deep neural networks, it is well-known that vanilla adversarial training suffers from daunting robust overfitting, resulting in unsatisfactory robust generalization. A number of approaches have been proposed to address these drawbacks such as extra regularization, adversarial weights perturbation, and training with more data over the last few years. However, the robust generalization improvement is yet far from satisfactory. In this paper, we approach this challenge with a brand new perspective -- refining historical optimization trajectories. We propose a new method named \textbf{Weighted Optimization Trajectories (WOT)} that leverages the optimization trajectories of adversarial training in time. We have conducted extensive experiments to demonstrate the effectiveness of WOT under various state-of-the-art adversarial attacks. Our results show that WOT integrates seamlessly with the existing adversarial training methods and consistently overcomes the robust overfitting issue, resulting in better adversarial robustness. For example, WOT boosts the robust accuracy of AT-PGD under AA-$L_{\infty}$ attack by 1.53\% $\sim$ 6.11\% and meanwhile increases the clean accuracy by 0.55\%$\sim$5.47\% across SVHN, CIFAR-10, CIFAR-100, and Tiny-ImageNet datasets.

翻译：尽管对抗训练已成为提升深度神经网络鲁棒性的事实标准方法，但众所周知，原始对抗训练存在严重的鲁棒过拟合问题，导致鲁棒泛化性能不佳。近年来，研究者提出了多种方法来解决这些缺陷，例如额外正则化、对抗权重扰动以及使用更多数据进行训练等。然而，鲁棒泛化性能的提升仍远未令人满意。本文从一个全新视角——优化历史轨迹的精炼来应对这一挑战。我们提出一种名为\textbf{加权优化轨迹（WOT）}的新方法，该方法利用对抗训练在时间维度上的优化轨迹。我们进行了大量实验，以证明WOT在各种最先进对抗攻击下的有效性。结果表明，WOT能够无缝集成到现有对抗训练方法中，并持续克服鲁棒过拟合问题，从而获得更好的对抗鲁棒性。例如，在SVHN、CIFAR-10、CIFAR-100和Tiny-ImageNet数据集上，WOT将AT-PGD在AA-$L_{\infty}$攻击下的鲁棒准确率提升了1.53%$\sim$6.11%，同时将干净准确率提升了0.55%$\sim$5.47%。