Lane-Keeping Task Dataset

该数据集是通过车道保持任务生成的，使用高保真模拟环境让车辆沿着车道中心行驶。数据集包含了来自摄像头的图像、基于车辆定位的标量奖励以及每一集结束的指示器。这些数据被用于训练强化学习中的潜在动态模型。该数据集规模较大，包含了来自模拟车辆动态的高维数据，其任务是进行车道保持控制。

This dataset is generated for the lane-keeping task, which uses a high-fidelity simulation environment to guide vehicles to drive along the center of the driving lane. The dataset includes camera-captured images, scalar rewards based on vehicle localization, and episode termination indicators. This data is employed to train latent dynamic models in the field of reinforcement learning. As a large-scale dataset, it contains high-dimensional data derived from simulated vehicle dynamics, with the core task being lane-keeping control.