GraphSCENE: On-Demand Critical Scenario Generation for Autonomous Vehicles in Simulation
This work addresses the problem of time-consuming scenario creation for autonomous vehicle validation, offering an on-demand generation tool that could accelerate testing processes, though it appears incremental as it builds on existing graph neural network and simulation techniques.
The paper tackles the challenge of manually creating diverse and safety-critical traffic scenarios for autonomous vehicle testing in simulation by introducing a method that generates dynamic temporal scene graphs on-demand based on user-defined preferences. The result is a model that consistently outperforms baselines in accurately generating links for requested scenarios, which are then rendered in simulation to demonstrate effectiveness as testing environments.
Testing and validating Autonomous Vehicle (AV) performance in safety-critical and diverse scenarios is crucial before real-world deployment. However, manually creating such scenarios in simulation remains a significant and time-consuming challenge. This work introduces a novel method that generates dynamic temporal scene graphs corresponding to diverse traffic scenarios, on-demand, tailored to user-defined preferences, such as AV actions, sets of dynamic agents, and criticality levels. A temporal Graph Neural Network (GNN) model learns to predict relationships between ego-vehicle, agents, and static structures, guided by real-world spatiotemporal interaction patterns and constrained by an ontology that restricts predictions to semantically valid links. Our model consistently outperforms the baselines in accurately generating links corresponding to the requested scenarios. We render the predicted scenarios in simulation to further demonstrate their effectiveness as testing environments for AV agents.