论文标题
对抗性和自然扰动的一般鲁棒性
Adversarial and Natural Perturbations for General Robustness
论文作者
论文摘要
在本文中，我们旨在通过利用对抗性和自然扰动来探索神经网络分类器的一般鲁棒性。与以前的工作不同，这些工作主要集中于研究神经网络对对抗扰动的鲁棒性，我们还评估了它们在鲁棒化前后对自然扰动的鲁棒性。在标准化对抗和自然扰动之间的比较之后，我们证明，尽管对抗训练可以提高网络对对抗性扰动的性能，但除了干净的样品以外，自然扰动样品的性能下降。相比之下，诸如弹性变形，阻塞和波浪之类的自然扰动不仅可以改善针对自然扰动的性能，而且还可以改善对抗性扰动的性能。此外，它们不会在干净的图像上删除精度。
In this paper we aim to explore the general robustness of neural network classifiers by utilizing adversarial as well as natural perturbations. Different from previous works which mainly focus on studying the robustness of neural networks against adversarial perturbations, we also evaluate their robustness on natural perturbations before and after robustification. After standardizing the comparison between adversarial and natural perturbations, we demonstrate that although adversarial training improves the performance of the networks against adversarial perturbations, it leads to drop in the performance for naturally perturbed samples besides clean samples. In contrast, natural perturbations like elastic deformations, occlusions and wave does not only improve the performance against natural perturbations, but also lead to improvement in the performance for the adversarial perturbations. Additionally they do not drop the accuracy on the clean images.
