SinD

该数据集是一个用于视觉故事讲述任务的顺序视觉至语言数据集。它主要用于评估视觉故事讲述中的句子顺序预测。每个故事包含五张图片及其对应描述，旨在进行句子顺序预测的任务。

This dataset is a sequential vision-to-language dataset for visual storytelling tasks. It is primarily employed to evaluate sentence order prediction within the context of visual storytelling. Each story consists of five images along with their corresponding descriptions, and is specifically constructed for the sentence order prediction task.