Security of Deep Learning based Lane Keeping System under Physical-World Adversarial Attack
This addresses a critical safety problem for autonomous and assisted driving systems, with incremental novelty in applying adversarial attacks to a specific domain.
The paper tackled the security of deep learning-based lane keeping systems by designing a physical-world adversarial attack using dirty road patches, and demonstrated that the attack can cause a state-of-the-art system to drive off lane boundaries within 1.3 seconds.
Lane-Keeping Assistance System (LKAS) is convenient and widely available today, but also extremely security and safety critical. In this work, we design and implement the first systematic approach to attack real-world DNN-based LKASes. We identify dirty road patches as a novel and domain-specific threat model for practicality and stealthiness. We formulate the attack as an optimization problem, and address the challenge from the inter-dependencies among attacks on consecutive camera frames. We evaluate our approach on a state-of-the-art LKAS and our preliminary results show that our attack can successfully cause it to drive off lane boundaries within as short as 1.3 seconds.