ANNA

ANNA数据集是由军事科学与技术学院创建的，专注于异构交通环境下的自动驾驶车辆研究。该数据集包含1800张从孟加拉国城市街道拍摄的图像，涵盖多种交通情况、天气和光照条件。创建过程涉及使用智能手机摄像头收集视频，提取图像并进行手动标注，包括8个类别的车辆和行人。ANNA数据集旨在解决自动驾驶车辆在复杂交通环境中的感知问题，特别是针对孟加拉国特有的交通情况，如罕见的车辆类型和拥挤的道路条件。

The ANNA dataset was developed by the Military College of Science and Technology, with a focus on research into autonomous vehicles in heterogeneous traffic environments. This dataset contains 1,800 images captured from urban streets in Bangladesh, covering a wide range of traffic scenarios, weather conditions, and lighting settings. Its creation process involved collecting video footage using smartphone cameras, extracting frames to obtain the images, and conducting manual annotations for 8 categories of vehicles and pedestrians. The ANNA dataset is designed to address the perception challenges of autonomous vehicles in complex traffic environments, specifically targeting traffic conditions unique to Bangladesh, such as rare vehicle types and congested road scenarios.