Security Analysis of Camera-LiDAR Fusion Against Black-Box Attacks on Autonomous Vehicles

被引:0
作者
Hallyburton, R. Spencer [1 ]
Liu, Yupei [1 ]
Cao, Yulong [2 ]
Mao, Z. Morley [2 ]
Pajic, Miroslav [1 ]
机构
[1] Duke Univ, Durham, NC 27706 USA
[2] Univ Michigan, Ann Arbor, MI 48109 USA
来源
PROCEEDINGS OF THE 31ST USENIX SECURITY SYMPOSIUM | 2022年
关键词
D O I
暂无
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
To enable safe and reliable decision-making, autonomous vehicles (AVs) feed sensor data to perception algorithms to understand the environment. Sensor fusion with multi-frame tracking is becoming increasingly popular for detecting 3D objects. Thus, in this work, we perform an analysis of camera-LiDAR fusion, in the AV context, under LiDAR spoofing attacks. Recently, LiDAR-only perception was shown vulnerable to LiDAR spoofing attacks; however, we demonstrate these attacks are not capable of disrupting camera-LiDAR fusion. We then define a novel, context-aware attack: frustum attack, and show that out of 8 widely used perception algorithms - across 3 architectures of LiDAR-only and 3 architectures of camera-LiDAR fusion - all are significantly vulnerable to the frustum attack. In addition, we demonstrate that the frustum attack is stealthy to existing defenses against LiDAR spoofing as it preserves consistencies between camera and LiDAR semantics. Finally, we show that the frustum attack can be exercised consistently over time to form stealthy longitudinal attack sequences, compromising the tracking module and creating adverse outcomes on end-to-end AV control.
引用
收藏
页码:1903 / 1920
页数:18
相关论文
共 50 条
  • [21] Analysis of Sensor Attacks Against Autonomous Vehicles
    Jakobsen, Soren Bonning
    Knudsen, Kenneth Sylvest
    Andersen, Birger
    [J]. PROCEEDINGS OF THE 8TH INTERNATIONAL CONFERENCE ON INTERNET OF THINGS, BIG DATA AND SECURITY, IOTBDS 2023, 2023, : 131 - 139
  • [22] Ensemble adversarial black-box attacks against deep learning systems
    Hang, Jie
    Han, Keji
    Chen, Hui
    Li, Yun
    [J]. PATTERN RECOGNITION, 2020, 101
  • [23] Towards Lightweight Black-Box Attacks Against Deep Neural Networks
    Sun, Chenghao
    Zhang, Yonggang
    Wan Chaoqun
    Wang, Qizhou
    Li, Ya
    Liu, Tongliang
    Han, Bo
    Tian, Xinmei
    [J]. ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 35 (NEURIPS 2022), 2022,
  • [24] Instance Segmentation Frustum-PointPillars: A Lightweight Fusion Algorithm for Camera-LiDAR Perception in Autonomous Driving
    Wang, Yongsheng
    Han, Xiaobo
    Wei, Xiaoxu
    Luo, Jie
    [J]. MATHEMATICS, 2024, 12 (01)
  • [25] A Practical Black-Box Attack Against Autonomous Speech Recognition Model
    Fan, Wenshu
    Li, Hongwei
    Jiang, Wenbo
    Xu, Guowen
    Lu, Rongxing
    [J]. 2020 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM), 2020,
  • [26] Towards Interpretable Camera and LiDAR Data Fusion for Autonomous Ground Vehicles Localisation
    Tibebu, Haileleol
    De-Silva, Varuna
    Artaud, Corentin
    Pina, Rafael
    Shi, Xiyu
    [J]. SENSORS, 2022, 22 (20)
  • [27] Panoptic-FusionNet: Camera-LiDAR fusion-based point cloud panoptic segmentation for autonomous driving
    Song, Hamin
    Cho, Jieun
    Ha, Jinsu
    Park, Jaehyun
    Jo, Kichun
    [J]. EXPERT SYSTEMS WITH APPLICATIONS, 2024, 251
  • [28] Provably Efficient Black-Box Action Poisoning Attacks Against Reinforcement Learning
    Liu, Guanlin
    Lai, Lifeng
    [J]. ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 34 (NEURIPS 2021), 2021, 34
  • [29] Adversarial Black-Box Attacks Against Network Intrusion Detection Systems: A Survey
    Alatwi, Huda Ali
    Aldweesh, Amjad
    [J]. 2021 IEEE WORLD AI IOT CONGRESS (AIIOT), 2021, : 34 - 40
  • [30] PRADA: Practical Black-box Adversarial Attacks against Neural Ranking Models
    Wu, Chen
    Zhang, Ruqing
    Guo, Jiafeng
    De Rijke, Maarten
    Fan, Yixing
    Cheng, Xueqi
    [J]. ACM TRANSACTIONS ON INFORMATION SYSTEMS, 2023, 41 (04)