arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1810.02424cs.LGstat.ML
收藏

特征优先级和正则化提高了标准准确性和对抗鲁棒性

Feature Prioritization and Regularization Improve Standard Accuracy and Adversarial Robustness

Institute for Advanced Computer Studies(高级计算机研究所);University of Maryland(马里兰大学)

Chihuang Liu, Joseph JaJa

摘要:

对抗性训练已经成功地应用于以一定的代价构建健壮的模型。虽然模型的鲁棒性增加,但标准分类精度下降。这种现象被认为是一种内在的权衡。我们提出了一个模型,该模型采用非线性注意力模块和$L_2$特征正则化的特征优先级,以提高对抗性的鲁棒性和相对于对抗性训练的标准精度。注意力模块鼓励模型在很大程度上依赖于鲁棒特征,方法是为鲁棒特征分配更大的权重,同时抑制非鲁棒特征。正则化器鼓励模型为自然图像和对抗图像提取相似的特征,有效地忽略了增加的扰动。除了评估我们的模型的鲁棒性,我们提供了理由的注意力模块,并提出了一种新的实验策略,定量证明,我们的模型几乎是理想的突出数据特征对齐。额外的实验结果说明了我们的模型相对于最先进的方法的能力。

英文摘要:

Adversarial training has been successfully applied to build robust models at a certain cost. While the robustness of a model increases, the standard classification accuracy declines. This phenomenon is suggested to be an inherent trade-off. We propose a model that employs feature prioritization by a nonlinear attention module and $L_2$ feature regularization to improve the adversarial robustness and the standard accuracy relative to adversarial training. The attention module encourages the model to rely heavily on robust features by assigning larger weights to them while suppressing non-robust features. The regularizer encourages the model to extract similar features for the natural and adversarial images, effectively ignoring the added perturbation. In addition to evaluating the robustness of our model, we provide justification for the attention module and propose a novel experimental strategy that quantitatively demonstrates that our model is almost ideally aligned with salient data characteristics. Additional experimental results illustrate the power of our model relative to the state of the art methods.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑