Resilient Autonomous Vehicle Perception Under Adversarial Settings

This project, titled Resilient Autonomous Vehicle Perception under Adversarial Settings, addresses critical challenges in the safety and reliability of autonomous vehicles (AVs) operating in real-world environments. Modern AV systems depend heavily on deep learning-based perception modules for tasks such as object detection, automated lane centering, and traffic sign recognition. However, these systems remain vulnerable to adversarial attacks, such as environmental modifications designed to mislead AV sensors and compromise decision-making. This research aims to develop robust model-end defenses by employing adversarial training and integrating Vision-Language Models (VLMs), ensuring AV perception systems are resilient to both known (white-box) and unknown (black-box) adversarial scenarios. This transformative research aligns with the broader goal of ensuring safer and more secure transportation systems as AV adoption accelerates. This project directly supports the U.S. Department of Transportation’s (US DOT) priorities by advancing the safety, security, and reliability of the transportation system. It aligns with the US DOT Office of Research Development and Technology (RD&T) Strategic Plan goals, particularly in enhancing resilience and security for autonomous vehicle technologies. The research engages in breakthrough, transformative approaches, such as: Developing advanced adversarial training techniques to strengthen deep learning models used in AV perception systems; Incorporating Vision-Language Models to provide a multimodal, context-aware understanding of the AV environment; Addressing safety-critical issues that improve public confidence and regulatory compliance, fostering the widespread adoption of AV technologies; This project emphasizes innovative methodologies to mitigate adversarial threats, ensuring the long-term safety and sustainability of AV deployment.

• Record URL:
  • https://www.clemson.edu/cecas/tracr/research/projects/#resilientautonomousvehicleperceptionunderadversarialsettings

Language

• English

Project

• Status: Active
• Funding: $276,314.00
• Contract Numbers:

  69A3552344812

  69A3552348317

• Sponsor Organizations:

  Office of the Assistant Secretary for Research and Technology

  University Transportation Centers Program
  Department of Transportation
  Washington, DC  United States  20590

  Clemson University

  216 Lowry Hall
  Clemson, SC, SC  United States  29634

  Benedict College

  1600 Harden Street
  Columbia, South Carolina  United States  29204
• Managing Organizations:

  National Center for Transportation Cybersecurity and Resiliency (TraCR)

  Clemson University
  Clemson, SC  United States 

  Clemson University

  216 Lowry Hall
  Clemson, SC, SC  United States  29634
• Project Managers:

  Chowdhury, Mashrur

• Performing Organizations:

  Clemson University

  216 Lowry Hall
  Clemson, SC, SC  United States  29634

  Benedict College

  1600 Harden Street
  Columbia, South Carolina  United States  29204
• Principal Investigators:

  Li, Bing

  Pesé, Mert

  Iyangar, Balaji

• Start Date: 20250101
• Expected Completion Date: 20251231
• Actual Completion Date: 0
• USDOT Program: University Transportation Centers

Subject/Index Terms

• TRT Terms: Autonomous vehicle guidance; Autonomous vehicles; Computer security; Connected vehicles; Deep learning; Traffic safety; Vehicle safety
• Subject Areas: Highways; Safety and Human Factors; Security and Emergencies; Vehicles and Equipment;

Filing Info

• Accession Number: 01950155
• Record Type: Research project
• Source Agency: National Center for Transportation Cybersecurity and Resiliency (TraCR)
• Contract Numbers: 69A3552344812, 69A3552348317
• Files: UTC, RIP
• Created Date: Mar 27 2025 2:53PM