mlsanityData to 14 Oct 2026

PaniCar: Securing the Perception of Advanced Driving Assistance Systems Against Emergency Vehicle Lighting

arXiv 2505.05183 ↗First posted 8 May 2025cs.CV cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

The safety of autonomous cars has come under scrutiny in recent years, especially after 16 documented incidents involving Teslas (with autopilot engaged) crashing into parked emergency vehicles (police cars, ambulances, and firetrucks). While previous studies have revealed that strong light sources often introduce flare artifacts in the captured image, which degrade the image quality, the impact of flare on object detection performance remains unclear. In this research, we unveil PaniCar, a digital phenomenon that causes an object detector's confidence score to fluctuate below detection thresholds when exposed to activated emergency vehicle lighting. This vulnerability poses a significant safety risk, and can cause autonomous vehicles to fail to detect objects near emergency vehicles. In addition, this vulnerability could be exploited by adversaries to compromise the security of advanced driving assistance systems (ADASs). We assess seven commercial ADASs (Tesla Model 3, "manufacturer C", HP, Pelsee, AZDOME, Imagebon, Rexing), four object detectors (YOLO, SSD, RetinaNet, Faster R-CNN), and 14 patterns of emergency vehicle lighting to understand the influence of various technical and environmental factors. We also evaluate four SOTA flare removal methods and show that their performance and latency are insufficient for real-time driving constraints. To mitigate this risk, we propose Caracetamol, a robust framework designed to enhance the resilience of object detectors against the effects of activated emergency vehicle lighting. Our evaluation shows that on YOLOv3 and Faster RCNN, Caracetamol improves the models' average confidence of car detection by 0.20, the lower confidence bound by 0.33, and reduces the fluctuation range by 0.33. In addition, Caracetamol is capable of processing frames at a rate of between 30-50 FPS, enabling real-time ADAS car detection.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 32 of its 71 references that are in the map. First posted on arXiv 8 May 2025.

Builds on

32 of 71 references.

References · 32

  1. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
  2. 2015SSD: Single Shot MultiBox Detectorcited by 3,130
  3. 2019MMDetection: Open MMLab Detection Toolbox and Benchmarkcited by 1,385
  4. 2018BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learningcited by 1,288
  5. 2016Simple Online and Realtime Trackingcited by 868
  6. 2021ByteTrack: Multi-Object Tracking by Associating Every Detection Boxcited by 823
  7. 2024YOLOv9: Learning What You Want to Learn Using Programmable Gradient Informationcited by 425
  8. 2022Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Trackingcited by 346
  9. 2018Physical Adversarial Examples for Object Detectorscited by 163
  10. 2018ShapeShifter: Robust Physical Adversarial Attack on Faster R-CNN Object Detectorcited by 159
  11. 2019Adversarial Sensor Attack on LiDAR-based Perception in Autonomous Drivingcited by 152
  12. 2021Invisible for both Camera and LiDAR: Security of Multi-Sensor Fusion based Perception in Autonomous Driving Under Physical-World Attackscited by 101
20 more
  1. 2020Towards Robust LiDAR-based Perception in Autonomous Driving: General Black-box Adversarial Sensor Attack and Countermeasurescited by 95
  2. 2020Dirty Road Can Attack: Security of Deep Learning based Automated Lane Centering under Physical-World Attackcited by 64
  3. 2020The Translucent Patch: A Physical and Universal Attack on Object Detectorscited by 62
  4. 2020SLAP: Improving Physical Adversarial Examples with Short-Lived Adversarial Perturbationscited by 50
  5. 2020How to Train Neural Networks for Flare Removalcited by 50
  6. 2022Flare7K: A Phenomenological Nighttime Flare Removal Datasetcited by 49
  7. 2022You Can't See Me: Physical Removal Attacks on LiDAR-based Autonomous Vehicles Driving Frameworkscited by 32
  8. 2019Fooling a Real Car with Adversarial Traffic Signscited by 29
  9. 2020Drift with Devil: Security of Multi-Sensor Fusion based Localization in High-Level Autonomous Driving under GPS Spoofing (Extended Version)cited by 25
  10. 2022SoK: On the Semantic AI Security in Autonomous Drivingcited by 22
  11. 2023Improving Lens Flare Removal with General Purpose Pipeline and Multiple Light Sources Recoverycited by 21
  12. 2022Rolling Colors: Adversarial Laser Exploits against Traffic Light Recognitioncited by 15
  13. 2024Invisible Reflections: Leveraging Infrared Laser Reflections to Target Traffic Sign Perceptioncited by 14
  14. 2017Another Flip in the Wall of Rowhammer Defensescited by 12
  15. 2015DRAMA: Exploiting DRAM Addressing for Cross-CPU Attackscited by 10
  16. 2023LiDAR Spoofing Meets the New-Gen: Capability Improvements, Broken Assumptions, and New Attack Strategiescited by 9
  17. 2019MobilBye: Attacking ADAS with Camera Spoofingcited by 8
  18. 2022Robust Perception Architecture Design for Automotive Cyber-Physical Systemscited by 4
  19. 2023MadRadar: A Black-Box Physical Layer Attack Framework on mmWave Automotive FMCW Radarscited by 3
  20. 2022bAdvertisement: Attacking Advanced Driver-Assistance Systems Using Print Advertisementscited by 1

Built on by

Not cited yet.