arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.03141cs.LGcs.CVstat.ML
收藏

基于频率的鲁棒CNN解释

Towards Frequency-Based Explanation for Robust CNN

Carnegie Mellon University(卡内基梅隆大学)

Zifan Wang, Yilin Yang, Ankit Shrivastava, Varun Rawal, Zihao Ding

摘要:

目前的透明卷积神经网络(CNN)的解释技术主要集中在建立人类可理解的输入特征与模型预测之间的联系,忽略了输入的另一种表示,频率分量分解。在这项工作中,我们分析了输入数据集中频率分量的分布与模型从数据中学习的推理过程之间的联系。我们进一步提供了量化分析的不同频率分量对模型的预测的贡献。我们证明了模型对微小失真的脆弱性是模型依赖于高频特征,即对抗性(黑盒和白盒)攻击者的目标特征来进行预测的结果。我们进一步表明,如果模型在低频分量与真实标签之间建立了更强的关联,则模型更鲁棒,这就是为什么对抗训练的模型对微小失真更鲁棒的解释。

英文摘要:

Current explanation techniques towards a transparent Convolutional Neural Network (CNN) mainly focuses on building connections between the human-understandable input features with models' prediction, overlooking an alternative representation of the input, the frequency components decomposition. In this work, we present an analysis of the connection between the distribution of frequency components in the input dataset and the reasoning process the model learns from the data. We further provide quantification analysis about the contribution of different frequency components toward the model's prediction. We show that the vulnerability of the model against tiny distortions is a result of the model is relying on the high-frequency features, the target features of the adversarial (black and white-box) attackers, to make the prediction. We further show that if the model develops stronger association between the low-frequency component with true labels, the model is more robust, which is the explanation of why adversarially trained models are more robust against tiny distortions.

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。

↑