VLD
收藏资源简介:
VLD数据集是由中国科学院自动化研究所的研究人员构建的,用于评估最后一公里无人机配送系统。该数据集基于CARLA模拟器,包含了300个配送任务,分布在22个不同的建筑物中。数据集涵盖了各种场景和任务,包括各种类型的建筑物和目标物体,如工具、容器、家居用品、食品、家具、海报、玩具和装饰品等。此外,数据集还考虑了任务的难度水平和目标楼层数,以确保其多样性。VLD数据集的构建旨在填补现有视觉-语言导航基准的空白,为研究者在最后一公里无人机配送系统领域的研究和评估提供支持。
The VLD dataset was constructed by researchers from the Institute of Automation, Chinese Academy of Sciences, for evaluating last-mile UAV delivery systems. Based on the CARLA simulator, this dataset contains 300 delivery tasks distributed across 22 distinct buildings. It covers a wide range of scenarios and tasks, including various types of buildings and target objects such as tools, containers, household items, food, furniture, posters, toys, decorations and more. In addition, the dataset takes into account task difficulty levels and the number of target floors to ensure its diversity. The construction of the VLD dataset aims to fill the gap in existing visual-language navigation benchmarks, providing support for researchers to carry out research and evaluation in the field of last-mile UAV delivery systems.




