CoCAtt

该数据集名为CoCAtt，是一个关于驾驶员注意力的数据集，它包含了每帧的标注信息，描述了驾驶员的分心状态和意图。该数据集通过不同分辨率的眼动追踪设备，在手动驾驶和自动驾驶模式下捕捉注意力数据。此外，该数据集还包含了手动驾驶和自动驾驶模式下的注意力数据，探讨了驾驶员的分心状态和意图等问题，为驾驶员注意力建模提供了宝贵的见解。在自主级别、眼动追踪设备分辨率以及驾驶场景的多样性方面，它是目前最大且最多样化的驾驶员注意力数据集。该数据集的任务是预测驾驶员的注意力。

This dataset, named CoCAtt, centers on driver attention research. It includes per-frame annotation information describing driver distraction states and driving intentions. It collects attention data under both manual driving and autonomous driving modes via eye-tracking devices with varying resolutions, and additionally explores issues such as driver distraction states and intentions through this data, providing valuable insights for driver attention modeling. It is currently the largest and most diverse driver attention dataset available, covering aspects including automation levels, eye-tracking device resolutions, and the diversity of driving scenarios. The core task of this dataset is to predict driver attention.