Korea-MES/DriveLM-MLT-Full

这是一个视觉问答数据集，包含场景ID、帧ID、问题ID、场景描述、类别、问题、答案、上下文、标签、聚类、层级、上下文关联、关键物体及其数量、可用摄像头列表、图片路径、多语言标签等字段。数据集分为训练集，共有309555个示例，数据集大小为675595039.8920109字节。

This is a Visual Question Answering dataset containing fields such as scene ID, frame ID, question ID, scene description, category, question, answer, context, tag, cluster, layer, context relationship, key objects and their count, available camera list, image paths, multilingual labels, etc. The dataset is split into a training set with a total of 309555 examples, and the dataset size is 675595039.8920109 bytes.