sharkntn/ai_tod

该数据集包含2194个训练样本，每个样本由三部分组成：一张普通图像（image）、对应的文本描述（image_caption）和一张卫星图像（satilte）。数据集总大小约为2.44GB，旨在提供图像与卫星图像之间的配对数据，可能用于计算机视觉任务，如图像生成、多模态分析或遥感应用。This dataset contains 2,194 training examples, each consisting of three components: a regular image (image), a corresponding textual description (image_caption), and a satellite image (satilte). The total dataset size is approximately 2.44 GB, and it is designed to provide paired data between images and satellite images, potentially for computer vision tasks such as image generation, multimodal analysis, or remote sensing applications.