Abstract: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 optimizes our natural adversarial scenario generation solution. Extensive experiments have been conducted in the NGSIM dataset including the trajectory of 3,000 vehicles. 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, which can be a cornerstone for the development of autonomous vehicles.

Task-Driven Controllable Scenario Generation Framework Based on AOG

Genetic Algorithm-Based Challenging Scenarios Generation for Autonomous Vehicle Testing

A Real-time Critical-scenario-generation Framework for Testing Autonomous Driving System

Critical Scenario Generation for Realistic Testing of Autonomous Driving System Via Multi-vehicle Cooperative Control

Scenario Generation for Autonomous Vehicles with Deep-Learning-Based Heterogeneous Driver Models: Implementation and Verification

Life-long Learning and Testing for Automated Vehicles via Adaptive Scenario Sampling as A Continuous Optimization Process

A Dynamic Test Scenario Generation Method for Autonomous Vehicles Based on Conditional Generative Adversarial Imitation Learning

Game-Theoretic Adversarial Interaction-Based Critical Scenario Generation for Autonomous Vehicles

Simulation Testing Scenario Generation for Comfort Evaluation of Automated Vehicles

An Open-Source Scenario Architect for Autonomous Vehicles

scenoRITA: Generating Less-Redundant, Safety-Critical and Motion Sickness-Inducing Scenarios for Autonomous Vehicles

Evolving Testing Scenario Generation Method and Intelligence Evaluation Framework for Automated Vehicles

Bridging Data-Driven and Knowledge-Driven Approaches for Safety-Critical Scenario Generation in Automated Vehicle Validation

Autonomous Vehicles Testing Considering Utility-Based Operable Tasks

Adversarial Safety-Critical Scenario Generation using Naturalistic Human Driving Priors

Testing Scenario Library Generation for Connected and Automated Vehicles, Part I: Methodology

A Scenario-Based Development Framework for Autonomous Driving

A Survey on Safety-Critical Driving Scenario Generation -- A Methodological Perspective

A Preliminary Attempt of Deploying Aigc Inautodriving Scenario Generation: Via LLM.

Multi-Vehicle Interaction Scenarios Generation with Interpretable Traffic Primitives and Gaussian Process Regression

Automated Scenario Generation and Iterative Regression Testing Method for Autonomous Driving Systems