Towards Robust Gan-Generated Image Detection: A Multi-View Completion Representation

被引:0
|
作者
Liu, Chi [1 ]
Zhu, Tianqing [1 ]
Shen, Sheng [2 ]
Zhou, Wanlei [3 ]
机构
[1] Univ Technol Sydney, Sch Comp Sci, Ultimo, Australia
[2] Univ Sydney, Sch Elect & Informat Engn, Camperdown, Australia
[3] City Univ Macau, Macau, Peoples R China
基金
澳大利亚研究理事会;
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
GAN-generated image detection now becomes the first line of defense against the malicious uses of machine-synthesized image manipulations such as deepfakes. Although some existing detectors work well in detecting clean, known GAN samples, their success is largely attributable to overfitting unstable features such as frequency artifacts, which will cause failures when facing unknown GANs or perturbation attacks. To overcome the issue, we propose a robust detection framework based on a novel multi-view image completion representation. The framework first learns various view-to-image tasks to model the diverse distributions of genuine images. Frequency-irrelevant features can be represented from the distributional discrepancies characterized by the completion models, which are stable, generalized, and robust for detecting unknown fake patterns. Then, a multi-view classification is devised with elaborated intra- and inter-view learning strategies to enhance view-specific feature representation and cross-view feature aggregation, respectively. We evaluated the generalization ability of our framework across six popular GANs at different resolutions and its robustness against a broad range of perturbation attacks. The results confirm our method's improved effectiveness, generalization, and robustness over various baselines.
引用
收藏
页码:464 / 472
页数:9
相关论文
共 50 条
  • [1] Generalized and robust model for GAN-generated image detection
    Raj, Surbhi
    Mathew, Jimson
    Mondal, Arijit
    PATTERN RECOGNITION LETTERS, 2024, 182 : 104 - 110
  • [2] Evasion on general GAN-generated image detection by disentangled representation
    Chan, Patrick P. K.
    Zhang, Chuanxin
    Chen, Haitao
    Deng, Jingwen
    Meng, Xiao
    Yeung, Daniel S.
    INFORMATION SCIENCES, 2024, 683
  • [3] Causal Representation Learning for GAN-Generated Face Image Quality Assessment
    Tian, Yu
    Wang, Shiqi
    Chen, Baoliang
    Kwong, Sam
    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, 2024, 34 (08) : 7589 - 7600
  • [4] Robust Multi-View Representation Learning
    Venkatesan, Sibi
    Miller, James K.
    Dubrawski, Artur
    THIRTY-FOURTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THE THIRTY-SECOND INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE AND THE TENTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2020, 34 : 13939 - 13940
  • [5] Learning on Gradients: Generalized Artifacts Representation for GAN-Generated Images Detection
    Tan, Chuangchuang
    Zhao, Yao
    Wei, Shikui
    Gu, Guanghua
    Wei, Yunchao
    2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2023, : 12105 - 12114
  • [6] Less is more: A minimalist approach to robust GAN-generated face detection
    Ghosh, Tanusree
    Naskar, Ruchira
    PATTERN RECOGNITION LETTERS, 2024, 179 : 185 - 191
  • [7] DEEP MULTI-VIEW ROBUST REPRESENTATION LEARNING
    Jiao, Zhenyu
    Xu, Chao
    2017 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 2017, : 2851 - 2855
  • [8] General GAN-generated Image Detection by Data Augmentation in Fingerprint Domain
    Wang, Huaming
    Fei, Jianwei
    Dai, Yunshu
    Leng, Lingyun
    Xia, Zhihua
    2023 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, ICME, 2023, : 1187 - 1192
  • [9] MetaViewer: Towards A Unified Multi-View Representation
    Wang, Ren
    Sun, Haoliang
    Ma, Yuling
    Xi, Xiaoming
    Yin, Yilong
    2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2023, : 11590 - 11599
  • [10] A robust ensemble model for Deepfake detection of GAN-generated images on social media
    Preeti Sharma
    Manoj Kumar
    Hitesh Kumar Sharma
    Discover Computing, 28 (1)