IEEE Transactions on Intelligent Vehicles · 2023 · 23 citations · 55 references
ComputingAutonomous SystemsComputer ScienceArtificial IntelligenceAutonomous VehiclesMachine LearningImitation LearningReinforcement LearningTraffic EngineeringTraffic SafetyAdversarial Imitation LearningAutonomous Vehicle SafetyDriving Behavior ModelingGenerative Adversarial Imitation LearningPolicy OptimizationProximal Policy OptimizationScenario GenerationSimulation Environment Design
Evaluating the decision-making system is indispensable in developing autonomous vehicles, while realistic and challenging safety-critical test scenarios play a crucial role. Obtaining these scenarios is non-trivial, thanks to the long-tailed distribution, sparsity, and rarity in real-world data sets. To tackle this problem, in this paper, we introduce a natural adversarial scenario generation solution using naturalistic human driving priors and reinforcement learning techniques. By doing this, we can obtain large-scale test scenarios that are both diverse and realistic. Specifically, we build a simulation environment that mimics natural traffic interaction scenarios. Informed by this environment, we implement a two-stage procedure. The first stage incorporates conventional rule-based models, e.g., IDM (Intelligent Driver Model) and MOBIL (Minimizing Overall Braking Induced by Lane changes) model, to coarsely and discretely capture and calibrate key control parameters from the real-world dataset. Next, we leverage GAIL (Generative Adversarial Imitation Learning) to represent driver behaviors continuously. The derived GAIL can be further used to design a PPO (Proximal Policy Optimization)-based actor-critic network framework to fine-tune the reward function, and then optimize our natural adversarial scenario generation solution. Extensive experiments have been conducted in two popular datasets, NGSIM and INTERACTION. Essential traffic parameters were measured in comparison with the baseline model, e.g., the collision rate, accelerations, steering, and the number of lane changes. Our findings demonstrate that the proposed model can generate realistic safety-critical test scenarios covering both naturalness and adversariality with an advanced 44% efficiency gain over the baseline model, which can be a cornerstone for the development of autonomous vehicles.
55
Deep Residual Learning for Image Recognition
Kaiming He, Xiangyu Zhang, Shaoqing Ren et al. · 2016 · 222.1K citations · Full text
Generative adversarial networks
Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza et al. · Communications of the ACM · 2020 · 13.6K citations · Full text
Robust Physical-World Attacks on Deep Learning Visual Classification
Kevin Eykholt, Ivan Evtimov, Earlence Fernandes et al. · 2018 · 2.1K citations