Generating Socially Acceptable Perturbations for Efficient Evaluation of Autonomous Vehicles
This addresses safety evaluation for autonomous vehicles, but it is incremental as it builds on existing adversarial attack methods.
The paper tackles the problem of evaluating autonomous vehicle decision-making by generating socially acceptable perturbations that cause crashes primarily due to the AV's actions, and results show that policies safe in naturalistic environments have many crashes in perturbed settings.
Deep reinforcement learning methods have been widely used in recent years for autonomous vehicle's decision-making. A key issue is that deep neural networks can be fragile to adversarial attacks or other unseen inputs. In this paper, we address the latter issue: we focus on generating socially acceptable perturbations (SAP), so that the autonomous vehicle (AV agent), instead of the challenging vehicle (attacker), is primarily responsible for the crash. In our process, one attacker is added to the environment and trained by deep reinforcement learning to generate the desired perturbation. The reward is designed so that the attacker aims to fail the AV agent in a socially acceptable way. After training the attacker, the agent policy is evaluated in both the original naturalistic environment and the environment with one attacker. The results show that the agent policy which is safe in the naturalistic environment has many crashes in the perturbed environment.