Adversarial Attacks on Scene Graph Generation
文献类型：期刊论文
| 作者 | Zhao, Mengnan1; Zhang, Lihe2; Wang, Wei3 ; Kong, Yuqiu1; Yin, Baocai1
|
| 刊名 | IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
 |
| 出版日期 | 2024 |
| 卷号 | 19页码:3210-3225 |
| 关键词 | Task analysis Object detection Windows Visualization Mirrors Predictive models Perturbation methods Scene graph generation adversarial attack bounding box relabeling two-step weighted attack |
| ISSN号 | 1556-6013 |
| DOI | 10.1109/TIFS.2024.3360880 |
| 通讯作者 | Zhang, Lihe(zhanglihe@dlut.edu.cn) |
| 英文摘要 | Scene graph generation (SGG) effectively improves semantic understanding of the visual world. However, the recent interest of researchers focuses on enhancing SGG in non-adversarial settings, which raises our curiosity about the adversarial robustness of SGG models. To bridge this gap, we perform adversarial attacks on two typical SGG tasks, Scene Graph Detection (SGDet) and Scene Graph Classification (SGCls). Specifically, we initially propose a bounding box relabeling method to reconstruct reasonable attack targets for SGCls. It solves the inconsistency between the specified bounding boxes and the scene graphs selected as attack targets. Subsequently, we introduce a two-step weighted attack by removing the predicted objects and relational triples that affect attack performance, which significantly increases the success rate of adversarial attacks on two SGG tasks. Extensive experiments demonstrate the effectiveness of our methods on five popular SGG models and four adversarial attacks. |
| WOS关键词 | NETWORK |
| 资助项目 | National Key Research and Development Program of China |
| WOS研究方向 | Computer Science ; Engineering |
| 语种 | 英语 |
| WOS记录号 | WOS:001174295900007 |
| 出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
| 资助机构 | National Key Research and Development Program of China |
| 源URL | [http://ir.ia.ac.cn/handle/173211/58376]  |
| 专题 | 多模态人工智能系统全国重点实验室 |
| 通讯作者 | Zhang, Lihe |
| 作者单位 | 1.Dalian Univ Technol, Sch Comp Sci & Technol, Dalian 116000, Peoples R China 2.Dalian Univ Technol, Sch Informat & Commun Engn, Dalian 116024, Peoples R China 3.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100080, Peoples R China |
| 推荐引用方式 GB/T 7714 | Zhao, Mengnan,Zhang, Lihe,Wang, Wei,et al. Adversarial Attacks on Scene Graph Generation[J]. IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY,2024,19:3210-3225. |
| APA | Zhao, Mengnan,Zhang, Lihe,Wang, Wei,Kong, Yuqiu,&Yin, Baocai.(2024).Adversarial Attacks on Scene Graph Generation.IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY,19,3210-3225. |
| MLA | Zhao, Mengnan,et al."Adversarial Attacks on Scene Graph Generation".IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY 19(2024):3210-3225. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。