Research of patent technology development for adversarial attacks and defenses against intelligent models

doi:10.12267/j.issn.2096-5931.2026.006

Abstract

Abstract:

Given that intelligent models are vulnerable to adversarial sample attacks, which pose security risks such as misidentification and functional failure, their safe practical application is somewhat constrained. This paper focuses on patented technologies in the fields of adversarial attacks and defenses, reviews domestic and overseas related patent achievements, categorizes and summarizes key adversarial sample generation techniques and core defense methods against adversarial samples, and analyzes relevant patent layout characteristics and technological evolution trends, aiming to provide references for technological research, patent strategy, and security system development in related fields.

Key words: adversarial attack, adversarial sample, adversarial defense, patent, deep learning

CLC Number:

TP18

LIU Jie. Research of patent technology development for adversarial attacks and defenses against intelligent models[J]. Information and Communications Technology and Policy, 2026, 52(8): 37-44.

Add to citation manager EndNote|Ris|BibTeX

URL:

http://ictp.caict.ac.cn/EN/10.12267/j.issn.2096-5931.2026.006

http://ictp.caict.ac.cn/EN/Y2026/V52/I8/37

Figures/Tables 8

References 14

[1]	Szegedy C, Zaremba W, Sutskever I, et al. Intriguing properties of neural networks[PP/OL]. V1. arXiv ( 2013-12-21) [2026-06-06]. https://doi.org/10.48550/arXiv.1312.6199.
[2]	Guo Z L, Qian Y F, Li Y L, et al. Beyond vulnerabilities: a survey of adversarial attacks as both threats and defenses in computer vision systems[PP/OL]. arXiv (2025-08-03) [2026-06-06]. https://doi.org/10.48550/arXiv.2508.01845.
[3]	张家伟, 吴昊, 袁霖, 等. 对抗样本生成、防御与应用研究综述[J]. 中国图象图形学报, 2026, 31(1): 0062-0081.
[4]	Chen X Q, Gao X T, Zhao J J, et al. AdvDiffuser: natural adversarial example synthesis with diffusion models[C]//2023 IEEE/CVF International Conference on Computer Vision (ICCV). Paris: IEEE Press, 2023. DOI:10.1109/ICCV51070.2023.00421.
[5]	清华大学. 一种生成对抗图像的方法: CN201711487948.4[P]. 2018-07-06.
[6]	杭州电子科技大学. 一种基于贝叶斯优化的对抗样本生成方法: CN201910414533.7[P]. 2019-09-24.
[7]	国家计算机网络与信息安全管理中心. 基于生成对抗网络的恶意代码检测方法: CN202010524261.9[P]. 2020-10-27.
[8]	中国科学院深圳先进技术研究院. 无限制对抗样本生成方法、装置、电子设备及存储介质: CN202211549873.9[P]. 2023-05-23.
[9]	浙江工业大学. 一种具有防御对抗样本攻击功能的深度学习分类方法: CN201811138721.3[P]. 2019-03-12.
[10]	浙江大学. 基于图神经网络的隐私特征保护方法和系统: CN202311024849.8[P]. 2023-11-14.
[11]	International Business Machines Corporation. Generative adversarial network medical image generation for training of a classifier: US2019197358A1[P]. 2019-06-27.
[12]	灵犀视觉(深圳)科技有限公司.一种基于复杂场景的工业相机运行优化方法及系统: 202511073447.6[P]. 2025-11-04.
[13]	Microsoft Technology Licensing, LlC. Adversarial teacher-student learning for unsupervised domain adaptation: US2019287515A1[P]. 2019-09-19.
[14]	台建玮, 杨双宁, 王佳佳, 等. 大语言模型对抗性攻击与防御综述[J]. 计算机研究与发展, 2025, 62 (3): 563-588.

[1]	HUANG Xin, LIANG Liyan, ZHANG Qian. A review of decoding algorithms for MI-BCI [J]. Information and Communications Technology and Policy, 2026, 52(6): 9-17.
[2]	HU Tianyang, SUN Xiaoqiang, CHEN Longquan, ZHANG Dayuan. Research on traceability techniques for deep learning algorithms based on standard datasets [J]. Information and Communications Technology and Policy, 2026, 52(5): 50-57.
[3]	WANG Qiyao. Patent information analysis of artificial intelligence in the field of battery research [J]. Information and Communications Technology and Policy, 2025, 51(6): 60-65.
[4]	HE Linlin, ZHANG Qian. Patent situation analysis and development suggestions for quantum information technology [J]. Information and Communications Technology and Policy, 2024, 50(7): 30-38.
[5]	WANG Tianzhou, ZHANG Hua, LIU Qishen, HUANG Chao. The application of deep learning and large model technology in the real reconstruction of Beijing Central Axis [J]. Information and Communications Technology and Policy, 2024, 50(12): 82-88.
[6]	YU Xiaomeng, LI Jialu, LI Li, ZHANG Huili. Analysis of patent situation of solid-state batteries [J]. Information and Communications Technology and Policy, 2024, 50(11): 80-84.
[7]	LIU Jie. Patent analysis of artificial intelligence in the field of wind power generation [J]. Information and Communications Technology and Policy, 2024, 50(10): 8-14.
[8]	WANG Qiang, LI Jiahong. Research on innovative application scenarios of evolutionary deep learning in industrial park energy management [J]. Information and Communications Technology and Policy, 2024, 50(10): 91-96.
[9]	LI Jingwen, WANG Linglong, ZHAO Yangguang, CUI Weinan. Application and regulation of machine learning in auxiliary diagnosis and treatment of ophthalmic diseases [J]. Information and Communications Technology and Policy, 2023, 49(9): 87-91.
[10]	ZENG Han, XU Xiaoqing, QIAN Liuyihui, WU Juan. Application of deep learning in IP network optimization [J]. Information and Communications Technology and Policy, 2023, 49(7): 89-96.
[11]	YI Jiming. Comments on the US standards-essential patents policy [J]. Information and Communications Technology and Policy, 2023, 49(3): 1-9.
[12]	LIU Dan, HE Li. The progress of terahertz technology from the perspective of patent [J]. Information and Communications Technology and Policy, 2023, 49(3): 24-29.
[13]	QIN Le, BI Chunli. The regulatory path of NPE inthe US from the perspective of judicial practice [J]. Information and Communications Technology and Policy, 2023, 49(3): 48-53.
[14]	ZHOU Jie, WANG Chunjia, ZHANG Qian. Development trend analysis of implantable brain-computer interface electrode technology from patent perspective [J]. Information and Communications Technology and Policy, 2023, 49(3): 61-67.
[15]	YU Xiaomeng, YAN Kun. Patent analysis of drivetrain bearings for wind power [J]. Information and Communications Technology and Policy, 2023, 49(3): 68-73.