Bypassing Backdoor Detection Algorithms in Deep Learning

被引:72
|
作者
Tan, Te Juin Lester [1 ]
Shokri, Reza [1 ]
机构
[1] Natl Univ Singapore NUS, Dept Comp Sci, Singapore, Singapore
关键词
D O I
10.1109/EuroSP48549.2020.00019
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Deep learning models are vulnerable to various adversarial manipulations of their training data, parameters, and input sample. In particular, an adversary can modify the training data and model parameters to embed backdoors into the model, so the model behaves according to the adversary's objective if the input contains the backdoor features, referred to as the backdoor trigger (e.g., a stamp on an image). The poisoned model's behavior on clean data, however, remains unchanged. Many detection algorithms are designed to detect backdoors on input samples or model parameters, through the statistical difference between the latent representations of adversarial and clean input samples in the poisoned model. In this paper, we design an adversarial backdoor embedding algorithm that can bypass the existing detection algorithms including the state-of-the-art techniques. We design an adaptive adversarial training algorithm that optimizes the original loss function of the model, and also maximizes the indistinguishability of the hidden representations of poisoned data and clean data. This work calls for designing adversary-aware defense mechanisms for backdoor detection.
引用
收藏
页码:175 / 183
页数:9
相关论文
共 50 条
  • [41] Omni SCADA Intrusion Detection Using Deep Learning Algorithms
    Gao, Jun
    Gan, Luyun
    Buschendorf, Fabiola
    Zhang, Liao
    Liu, Hua
    Li, Peixue
    Dong, Xiaodai
    Lu, Tao
    IEEE INTERNET OF THINGS JOURNAL, 2021, 8 (02) : 951 - 961
  • [42] Smart Pothole Detection System using Deep Learning Algorithms
    Chougule, Savita
    Barhatte, Alka
    INTERNATIONAL JOURNAL OF INTELLIGENT TRANSPORTATION SYSTEMS RESEARCH, 2023, 21 (03) : 483 - 492
  • [43] Brain Hemorrhage Detection using Heatmaps and Deep Learning Algorithms
    Chevvuri, Swarna Tejaswi
    Kumar Reddy S, Venkata Rohit
    Nelluru, Sai Teja
    Yadlapalli, Priyanka
    International Conference on Innovative Data Communication Technologies and Application, ICIDCA 2023 - Proceedings, 2023, : 89 - 94
  • [44] Casting Discontinuity Detection Method Based on Deep Learning Algorithms
    Przepiora, Jan
    Ksiazek, Piotr
    2024 25TH INTERNATIONAL CARPATHIAN CONTROL CONFERENCE, ICCC 2024, 2024,
  • [45] Forest roads damage detection based on deep learning algorithms
    Heidari, Mohammad Javad
    Najafi, Akbar
    Borges, Jose G.
    SCANDINAVIAN JOURNAL OF FOREST RESEARCH, 2022, 37 (5-8) : 366 - 375
  • [46] A Comparative Analysis of Deep Learning Algorithms for Optical Drone Detection
    Shovon, Md Hedayetul Islam
    Gopalan, Rohit
    Campbell, Benjamin
    FIFTEENTH INTERNATIONAL CONFERENCE ON MACHINE VISION, ICMV 2022, 2023, 12701
  • [47] Damages Detection of Aeroengine Blades via Deep Learning Algorithms
    Li, Shuangbao
    Yu, Jingyi
    Wang, Hao
    IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, 2023, 72
  • [48] Deep Learning and Classification Algorithms for COVID-19 Detection
    Sidheeque, Mohammed
    Sumathy, P.
    Gafur, Abdul M.
    INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND APPLICATIONS, 2022, 13 (09) : 346 - 350
  • [49] Development and optimization of image fire detection on deep learning algorithms
    Yi Yang
    Mengyi Pan
    Pu Li
    Xuefeng Wang
    Yun-Ting Tsai
    Journal of Thermal Analysis and Calorimetry, 2023, 148 : 5089 - 5095
  • [50] Review of Small Object Detection Algorithms Based on Deep Learning
    Dong, Gang
    Xie, Weicheng
    Huang, Xiaolong
    Qiao, Yitian
    Mao, Qian
    Computer Engineering and Applications, 2023, 59 (11): : 16 - 27