Scale-free and Task-agnostic Attack: Generating Photo-realistic Adversarial Patterns with Patch Quilting Generator

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gao, Xiangbo, Luo, Cheng, Lin, Qinliang, Xie, Weicheng, Liu, Minmin, Shen, Linlin, Kusumam, Keerthy, Song, Siyang
Format: Preprint
Published: 2022
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916440051810304
author Gao, Xiangbo
Luo, Cheng
Lin, Qinliang
Xie, Weicheng
Liu, Minmin
Shen, Linlin
Kusumam, Keerthy
Song, Siyang
author_facet Gao, Xiangbo
Luo, Cheng
Lin, Qinliang
Xie, Weicheng
Liu, Minmin
Shen, Linlin
Kusumam, Keerthy
Song, Siyang
contents \noindent Traditional L_p norm-restricted image attack algorithms suffer from poor transferability to black box scenarios and poor robustness to defense algorithms. Recent CNN generator-based attack approaches can synthesize unrestricted and semantically meaningful entities to the image, which is shown to be transferable and robust. However, such methods attack images by either synthesizing local adversarial entities, which are only suitable for attacking specific contents or performing global attacks, which are only applicable to a specific image scale. In this paper, we propose a novel Patch Quilting Generative Adversarial Networks (PQ-GAN) to learn the first scale-free CNN generator that can be applied to attack images with arbitrary scales for various computer vision tasks. The principal investigation on transferability of the generated adversarial examples, robustness to defense frameworks, and visual quality assessment show that the proposed PQG-based attack framework outperforms the other nine state-of-the-art adversarial attack approaches when attacking the neural networks trained on two standard evaluation datasets (i.e., ImageNet and CityScapes).
format Preprint
id arxiv_https___arxiv_org_abs_2208_06222
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Scale-free and Task-agnostic Attack: Generating Photo-realistic Adversarial Patterns with Patch Quilting Generator
Gao, Xiangbo
Luo, Cheng
Lin, Qinliang
Xie, Weicheng
Liu, Minmin
Shen, Linlin
Kusumam, Keerthy
Song, Siyang
Computer Vision and Pattern Recognition
Image and Video Processing
\noindent Traditional L_p norm-restricted image attack algorithms suffer from poor transferability to black box scenarios and poor robustness to defense algorithms. Recent CNN generator-based attack approaches can synthesize unrestricted and semantically meaningful entities to the image, which is shown to be transferable and robust. However, such methods attack images by either synthesizing local adversarial entities, which are only suitable for attacking specific contents or performing global attacks, which are only applicable to a specific image scale. In this paper, we propose a novel Patch Quilting Generative Adversarial Networks (PQ-GAN) to learn the first scale-free CNN generator that can be applied to attack images with arbitrary scales for various computer vision tasks. The principal investigation on transferability of the generated adversarial examples, robustness to defense frameworks, and visual quality assessment show that the proposed PQG-based attack framework outperforms the other nine state-of-the-art adversarial attack approaches when attacking the neural networks trained on two standard evaluation datasets (i.e., ImageNet and CityScapes).
title Scale-free and Task-agnostic Attack: Generating Photo-realistic Adversarial Patterns with Patch Quilting Generator
topic Computer Vision and Pattern Recognition
Image and Video Processing
url https://arxiv.org/abs/2208.06222