冲突解决数据集

冲突解决数据集是由代尔夫特理工大学开发的高质量数据集，源自Argoverse-2运动预测数据。该数据集包含5000多个自动驾驶车辆与人类道路用户冲突解决案例，以及16000多个无自动驾驶车辆的冲突解决案例。数据集通过特定的数据处理流程，选择感兴趣的场景，纠正数据错误，并增强原始数据，确保运动数据的平滑和一致性。此开放数据集包含多样化和平衡的冲突解决机制，适用于比较自动驾驶车辆参与和非参与场景下的冲突解决行为。数据集的应用领域包括评估自动驾驶车辆对城市交通的安全和效率影响，为研究人员提供宝贵的资源，以深入了解自动驾驶车辆的可能影响。

The Conflict Resolution Dataset is a high-quality dataset developed by Delft University of Technology, sourced from the Argoverse-2 motion forecasting dataset. This dataset contains over 5,000 conflict resolution cases involving autonomous vehicles and human road users, as well as more than 16,000 conflict resolution cases without autonomous vehicle participation. Through a dedicated data processing pipeline, the dataset selects scenes of interest, corrects data errors, and enhances the original data to ensure the smoothness and consistency of motion data. This open dataset features diverse and balanced conflict resolution mechanisms, and is suitable for comparing conflict resolution behaviors in scenarios with and without autonomous vehicle involvement. Its application areas include evaluating the safety and efficiency impacts of autonomous vehicles on urban traffic, providing valuable resources for researchers to gain in-depth insights into the potential impacts of autonomous vehicles.