XTD10 Dataset

XTD10 is a dataset for cross-lingual image retrieval and tagging consisting of the MSCOCO2014 caption test dataset annotated in 7 languages that were collected using a crowdsourcing platform.

XTD10是一个针对跨语言图像检索与标记的数据集，由7种语言标注的MSCOCO2014标题测试数据集构成，这些数据通过众包平台收集而来。