Attacks on state-of-the-art face recognition using attentional adversarial attack generative network

Abstract

With the broad use of face recognition, its weakness gradually emerges that it is able to be attacked. Therefore, it is very important to study how face recognition networks are subject to attacks. Generating adversarial examples is an effective attack method, which misleads the face recognition system through obfuscation attack (rejecting a genuine subject) or impersonation attack (matching to an impostor). In this paper, we introduce a novel GAN, Attentional Adversarial Attack Generative Network (A 3 GN), to generate adversarial examples that mislead the network to identify someone as the target person not misclassify inconspicuously. For capturing the geometric and context information of the target person, this work adds a conditional variational autoencoder and attention modules to learn the instance-level correspondences between faces. Unlike traditional two-player GAN, this work introduces a face recognition network as the third player to participate in the competition between generator and discriminator which allows the attacker to impersonate the target person better. The generated faces which are hard to arouse the notice of onlookers can evade recognition by state-of-the-art networks and most of them are recognized as the target person.

Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey

“…Adversarial patches for faces are also studied for their transferability in [245]. In another related example, Yang et al [246] proposed an Attentional Adversarial Attack Generative Network (A 3 GN ) for targeted fooling of face recognition models. It is claimed that their network is able to exploit geometric and context information of the target with the help of a conditional VAE and attention modules to achieve this feat.…”

Research on Multimodality Face Antispoofing Model Based on Adversarial Attacks

“…Bousnina et al [20] achieved high attack success rate by a black-box attack method in a CNN network based on transfer learning. Yang et al [21] introduced a new Attentional Adversarial Attack Generation Network (A 3 GN) as a way to generate the same adversarial examples as the original face images; unlike the traditional GAN, A 3 GN uses the face recognition network for the third player to participate in the competition between the generator and the discriminator. Yang et al [22] observed that ℓ0-norm has good sparsity but is hard to solve; for this reason, they proposed to use ℓq-norm to approximate ℓ0-norm and attack the face antispoofing task with the goal of minimizing the ℓq distance from the original image.…”

“…In previous works [18,19] on face antispoofing models against attacks, people used relatively basic single-modality convolutional networks to conduct experiments and achieved good attack results. However, the above research did not test the attack effect on Depth and IR images, such as [21]. Meanwhile, as the robustness models in this field continue to be researched in depth, whether existing models, especially multimodality models, can resist adversarial attacks remains to be experimented.…”

“…Although neural network-based face antispoofing models have achieved good performance in the face of presentation attacks, a small number of security researches on single-modality models [18][19][20][21] have shown that spoofing face will be incorrectly recognized by the model as the real face after adding small perturbations, which poses a new challenge to the security of the model. However, the above research is only conducted in the convolutional neural network with a relatively simple structure and did not involve multimodality detection models that have achieved excellent performance in recent years; meanwhile, the generation of perturbed images only involved RGB images and did not involve Depth and IR images.…”

Automatic Detection of Citrus Fruit and Leaves Diseases Using Deep Neural Network Model

“…Deep learning, on the other hand, can learn the hierarchical features of pathologies automatically, eliminating the need to manually design the morphological operations of feature extraction and classifier. The deep learning approach excels in several fields, including signal processing [10], pedestrian detection [11], face recognition [12], road crack detection [13], biomedical image analysis [14], and many others. Furthermore, deep learning techniques have produced promising results throughout the agricultural field, helping more farmers and food-producing workers, such as detection of plant disease [15], analysis of weeds [16], discovery of valuable seeds [17], insect detection [18], fruit processing [15], and so on, which has led to dealing with image analysis.…”