Hui Xie, Jirong Yi, Weiyu Xu, Raghu Mudumbai. An Information-Theoretic Explanation for the Adversarial Fragility of AI Classifiers. In IEEE International Symposium on Information Theory, ISIT 2019, Paris, France, July 7-12, 2019. pages 1977-1981, IEEE, 2019. [doi]
Abstract is missing.