SNLI-VE
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/necla-ml/snli-ve
下载链接
链接失效反馈官方服务：
资源简介：
该数据集融合了视觉和文本信息，旨在执行视觉蕴含任务。在实验中，该数据集被用来展示添加谓词特征和利用图形结构所带来的益处。其所涉及的任务是视觉蕴含。
This dataset integrates visual and textual information, with the goal of tackling visual entailment tasks. In experiments, it was employed to demonstrate the advantages of incorporating predicate features and leveraging graph structures. The core task involved in this dataset is visual entailment.
