Emirates Multi-Task (EMT) dataset

该数据集是首次公开的、在阿拉伯海湾地区收集的自动驾驶数据集，它捕捉到了该地区独特的道路拓扑、交通拥堵情况，以及行人物品穿着和天气条件的多样性。此外，该数据集还包含了可在 github.com/AV-Lab/emt-dataset 上获取的预处理脚本和评估模型。其规模宏大，包含超过30,000帧画面和570,000个标注的边界框，覆盖了大约150公里的驾驶路线。该数据集适用于多代理跟踪、轨迹预测和意图预测等任务。

This is the first publicly available autonomous driving dataset collected in the Arabian Gulf region. It captures the unique road topologies, traffic congestion scenarios, and the diversity of pedestrian attire, personal belongings, and weather conditions in this area. Additionally, this dataset provides preprocessing scripts and evaluation models accessible at github.com/AV-Lab/emt-dataset. It has a substantial scale, containing over 30,000 image frames and 570,000 annotated bounding boxes, covering approximately 150 kilometers of driving routes. This dataset is applicable to tasks such as multi-agent tracking, trajectory prediction, and intention prediction.