Explainable AI for Defending Recognition Models in Connected Autonomous Vehicles

This report explores Explainable Artificial Intelligence (XAI) methodologies aimed at enhancing the robustness, transparency, and cybersecurity of recognition models within Connected Autonomous Vehicles (CAVs). Two primary approaches are investigated: first, using Occlusion Sensitivity to analyze feature dependencies and vulnerabilities in intrusion detection systems, and second, employing explainable machine learning approaches to detect adversarial attacks targeting deep neural networks. Key findings highlight significant model reliance on limited critical features and underline the robustness and efficiency advantages of Random Forest classifiers. Despite certain limitations, the explainable machine learning method demonstrated potential in identifying adversarial perturbations. This research aligns closely with the strategic goals of the USDOT and CYBER-CARE initiative, emphasizing transparency, interpretability, and resilience in intelligent transportation systems.

• Record URL:
  • https://uh.edu/cybercare/research/projects/c23c07/
• Supplemental Notes:
  • This material is based on work supported by the U.S. Department of Transportation, OST-R, University Transportation Center Program, the USDOT Tier 1 UTC Transportation Cybersecurity Center for Advanced Research and Education (CYBER-CARE).

Language

• English

Project

• Status: Active
• Funding: $100,000.00
• Contract Numbers:

  69A3552348332

• Sponsor Organizations:

  Office of the Assistant Secretary for Research and Technology

  University Transportation Centers Program
  Department of Transportation
  Washington, DC  United States  20590
• Managing Organizations:

  Transportation Cybersecurity Center for Advanced Research and Education (CYBER-CARE)

  University of Houston
  Houston, TX  United States 
• Project Managers:

  Zhang, Yunpeng

  Kline, Robin

• Performing Organizations:

  Embry-Riddle Aeronautical University

  600 S. Clyde Morris Boulevard
  Daytona Beach, Fl  United States  32114
• Principal Investigators:

  Liu, Yongxin

• Start Date: 20240101
• Expected Completion Date: 20260101
• Actual Completion Date: 0
• USDOT Program: University Transportation Centers Program

Subject/Index Terms

• TRT Terms: Artificial intelligence; Autonomous vehicles; Computer security; Connected vehicles; Machine learning; Risk assessment
• Subject Areas: Data and Information Technology; Highways; Planning and Forecasting; Security and Emergencies; Vehicles and Equipment;

Filing Info

• Accession Number: 01953802
• Record Type: Research project
• Source Agency: Transportation Cybersecurity Center for Advanced Research and Education (CYBER-CARE)
• Contract Numbers: 69A3552348332
• Files: UTC, RIP, STATEDOT
• Created Date: Apr 29 2025 3:25PM