CARLA Leaderboard 2.0 数据集
收藏资源简介:
CARLA Leaderboard 2.0 数据集是由蒂宾根大学创建的,用于支持端到端自动驾驶系统的训练和评估。该数据集包含531,000帧数据,涵盖了RGB图像、LiDAR点云以及用于训练的标签,如路径检查点、专家目标速度和辅助标签。数据集的创建过程利用了PDM-Lite规划器,该规划器能够解决CARLA Leaderboard 2.0中的复杂场景。数据集的应用领域主要集中在自动驾驶系统的训练和评估,旨在解决复杂城市环境中的驾驶问题,提升模型在高速、变道和障碍物处理等场景中的表现。
The CARLA Leaderboard 2.0 dataset was developed by the University of Tübingen to support the training and evaluation of end-to-end autonomous driving systems. It contains 531,000 frames of data, including RGB images, LiDAR point clouds, and training labels such as path checkpoints, expert target speeds, and auxiliary labels. The dataset was constructed using the PDM-Lite planner, which can handle complex scenarios within the CARLA Leaderboard 2.0 benchmark. Its primary application lies in the training and evaluation of autonomous driving systems, with the goal of addressing driving challenges in complex urban environments and enhancing model performance in scenarios including high-speed driving, lane changing, and obstacle handling.




