Model-Agnostic Defense for Lane Detection against Adversarial Attack

Abstract

Susceptibility of neural networks to adversarial attack prompts serious safety concerns for lane detection efforts, a domain where such models have been widely applied. Recent work on adversarial road patches have successfully induced perception of lane lines with arbitrary form, presenting an avenue for rogue control of vehicle behavior. In this paper, we propose a modular lane verification system that can catch such threats before the autonomous driving system is misled while remaining agnostic to the particular lane detection model. Our experiments show that implementing the system with a simple convolutional neural network (CNN) can defend against a wide gamut of attacks on lane detection models. With a 10% impact to inference time, we can detect 96% of bounded non-adaptive attacks, 90% of bounded adaptive attacks, and 98% of patch attacks while preserving accurate identification at least 95% of true lanes, indicating that our proposed verification system is effective at mitigating lane detection security risks with minimal overhead.

arXiv comment6 pages, 6 figures, 3 tables. Part of AutoSec 2021 proceedings

References in corpus (8)

• Focal Loss for Dense Object Detection
• On the Convergence and Robustness of Adversarial Training
• Houdini: Fooling Deep Structured Prediction Models
• Learning Lightweight Lane Detection CNNs by Self Attention Distillation
• Adversarial Examples for Semantic Segmentation and Object Detection
• Key Points Estimation and Point Instance Segmentation Approach for Lane Detection
• Improving Back-Propagation by Adding an Adversarial Gradient
• Towards Lightweight Lane Detection by Optimizing Spatial Embedding