This repository contains the code for the paper "Detecting Adversarial Image Examples in Deep Networks with Adaptive Noise Reduction". It will reproduce the results reported in the paper.
DeepDetector is a straightforward method for detecting adversarial image examples. The method can effectually detect adversarial examples crafted by Fast Gradient Sign Method, DeepFool method and attacks designed by Nicholas Carlini and David Wagner. Adversarial examples crafted by other attack techniques may also can be detected by this method.
Liang B, Li H, Su M, et al. Detecting Adversarial Image Examples in Deep Networks with Adaptive Noise Reduction[J]. arXiv preprint arXiv:1705.08378, 2017.