GEM
收藏数据链接：
官方服务：
资源简介：
GEM是一个大规模的视觉-语言基准，包含用于图像-语言任务的GEM-I和用于视频-语言任务的GEM-V。它是目前覆盖图像-语言和视频-语言任务的最大视觉-语言数据集，并且标签化支持多种语言。
GEM is a large-scale vision-language benchmark comprising GEM-I for image-language tasks and GEM-V for video-language tasks. It currently stands as the largest vision-language dataset covering both image-language and video-language tasks, with its annotations supporting multiple languages.
创建时间：
2021-06-18
