AID Remote Sensing Dataset

该数据集名为AID，包含了30种不同航拍场景类型的10,000张图像。针对神经学习和融合，该数据集分别支持5折和2折交叉验证，总共产生了300个特征映射（FMs）。同时，数据集还提供了复现结果的代码。规模上，数据集包含了10,000张图像和30个类别。任务方面，主要是物体检测和土地分类。

This dataset, named AID, comprises 10,000 images spanning 30 distinct aerial scene categories. It supports 5-fold and 2-fold cross-validation for neural learning and fusion tasks respectively, generating a total of 300 feature maps (FMs). Additionally, the dataset provides code for reproducing reported experimental results. In terms of scale, the dataset contains 10,000 images and 30 categories. Its core downstream tasks include object detection and land classification.