Minimizing Perceived Image Quality Loss Through Adversarial Attack\n Scoping

2019/04/23 by Kostiantyn Khabarlak, Khabarlak, Kostiantyn, Larysa Koriashkina +1
Computer Science · Engineering · Medicine · #68T10 #Advanced Neural Network Applications #Advanced X-ray and CT Imaging #Adversarial Robustness in Machine Learning #Anomaly Detection Techniques and Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Medical Imaging Techniques and Applications

paper · pdf · doi:10.48550/arxiv.1904.10390

openalex publication_date 2019/04/23 · openalex created_date 2022/07/29 · openalex updated_date 2026/07/28

Abstract

Neural networks are now actively being used for computer vision tasks in\nsecurity critical areas such as robotics, face recognition, autonomous vehicles\nyet their safety is under question after the discovery of adversarial attacks.\nIn this paper we develop simplified adversarial attack algorithms based on a\nscoping idea, which enables execution of fast adversarial attacks that minimize\nstructural image quality (SSIM) loss, allows performing efficient transfer\nattacks with low target inference network call count and opens a possibility of\nan attack using pen-only drawings on a paper for the MNIST handwritten digit\ndataset. The presented adversarial attack analysis and the idea of attack\nscoping can be easily expanded to different datasets, thus making the paper's\nresults applicable to a wide range of practical tasks.\n

Citations

• Intriguing properties of neural networks
• ImageNet classification with deep convolutional neural networks

Related

• Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey
• A Review of Adversarial Attacks in Computer Vision
• Attack Type Agnostic Perceptual Enhancement of Adversarial Images
• Scratch that! An Evolution-based Adversarial Attack against Neural Networks
• AdvDrop: Adversarial Attack to DNNs by Dropping Information