SWOW vision baseline dataset
收藏arXiv2025-09-30 收录
下载链接：
https://smallworldofwords.org/en/project/explore
下载链接
链接失效反馈官方服务：
资源简介：
该数据集名为SWOW视觉基准数据集，包含了1,000个实例，这些实例是通过将SWOW数据集中的干扰项与关联项相结合而创建的。此外，SWOW数据集遵循创意共享署名-非商业性使用-禁止演绎3.0未本地化许可协议。该数据集的规模为1,000个实例，其任务是视觉与语言关联。
This dataset is named the SWOW Visual Benchmark Dataset. It includes 1,000 instances, which are constructed by combining distractors and related items from the original SWOW dataset. Additionally, the SWOW Visual Benchmark Dataset is released under the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License. With 1,000 instances overall, its core task is vision-language association.
