Defending against adversarial examples using perceptual image hashing

Abstract

Conventional deep neural networks (DNNs) have been shown to be vulnerable to images with adversarial perturbations, referred to as adversarial examples. In this study, we propose a method to protect neural networks against adversarial examples using perceptual image hashing. Because perceptual hashing is robust to adversarial perturbations, we combine hash sequences of input images with the parameters of a neural network in an image-hash processing network. Thus, outputs of the neural network are affected by image hashes, which render the model robust to adversarial examples to some extent. Thus, the proposed approach provides a defense against adversarial examples. The experiment was conducted on the CIFAR-10 dataset, and we used ResNet-18 as our target network. To verify our method, we tested our defense network using several common white-box attacks and black-box attacks. The results show that it achieved a significant improvement in the classification accuracy for adversarial examples.

Robust video hashing based on deep feature and quaternion generic Fourier descriptor for copy detection

“…where P i ¼ n i N is the probability of the gray value i, which can be obtained from the histogram of the image, n i is the number of pixels with gray value i, and N is the total number of pixels. For color image, its entropy 31 is defined as E Q -T A R G E T ; t e m p : i n t r a l i n k -; e 0 0 2 ; 1 1 4 ; 3 6 are the probabilities of the gray value i in the red, green, and blue components of the color image, respectively. As the calculation utilizes all color components, the color entropy can effectively reflect color information.…”

“…Hashing techniques are an effective approach of information retrieval. [3][4][5] In recent years, some researchers have proposed using hashing schemes 6,7 to detect copy versions of multimedia data for copyright protection. [8][9][10] Video hashing is an effective technique for video data processing.…”

Reconstructing images with attention generative adversarial network against adversarial attacks

“…Yang et al 59 Wu et al 60 proposed a method to protect neural networks against adversarial examples using perceptual image hashing. The method combines hash sequences of input images with the parameters of a neural network in an image-hash processing network, and the outputs of the neural network are affected by image hashes, which render the model robust to adversarial examples.…”