BeamNG.AI Test Scenarios

该数据集是为了评估生物启发测试生成器在引发深度神经网络车道保持系统多样化故障揭示测试场景中的有效性而生成的模拟场景。它包含了基于莱文斯坦距离的检测故障和故障多样性指标。该实验设置具有特定的配置，其任务是测试机器学习模型在自动驾驶系统中的应用。

This dataset is a simulated scenario constructed to evaluate the effectiveness of bio-inspired test generators in inducing diverse fault-revealing test scenarios for deep neural network-based lane keeping systems. It incorporates fault detection metrics based on Levenshtein distance alongside fault diversity metrics. The experimental setup features specific configurations, designed to test the deployment of machine learning models in autonomous driving systems.