End-to-end Uncertainty-based Mitigation of Adversarial Attacks to Automated Lane Centering

TL;DR

This paper presents an end-to-end method to mitigate adversarial attacks in autonomous lane centering systems by quantifying perception uncertainty and adjusting planning and control, significantly improving robustness.

Contribution

It introduces a comprehensive approach that addresses adversarial vulnerabilities across perception, planning, and control modules in autonomous driving.

Findings

Achieves 55% to 90% improvement in attack mitigation

Effectively quantifies perception uncertainty under adversarial conditions

Demonstrates robustness in both dataset and simulation environments

Abstract

In the development of advanced driver-assistance systems (ADAS) and autonomous vehicles, machine learning techniques that are based on deep neural networks (DNNs) have been widely used for vehicle perception. These techniques offer significant improvement on average perception accuracy over traditional methods, however, have been shown to be susceptible to adversarial attacks, where small perturbations in the input may cause significant errors in the perception results and lead to system failure. Most prior works addressing such adversarial attacks focus only on the sensing and perception modules. In this work, we propose an end-to-end approach that addresses the impact of adversarial attacks throughout perception, planning, and control modules. In particular, we choose a target ADAS application, the automated lane centering system in OpenPilot, quantify the perception uncertainty under adversarial attacks, and design a robust planning and control module accordingly based on the uncertainty analysis. We evaluate our proposed approach using both the public dataset and production-grade autonomous driving simulator. The experiment results demonstrate that our approach can effectively mitigate the impact of adversarial attacks and can achieve 55% to 90% improvement over the original OpenPilot.