The milestone improvements brought about by deep representation learning and
pre-training techniques have led to large performance gains across downstream
NLP, IR and Vision tasks. Multimodal modeling techniques aim to leverage large
high-quality visio-linguistic datasets for learning complementary information
(across image and text modalities). In this paper, we introduce the
Wikipedia-based Image Text (WIT) Dataset
(this https URL) to better facilitate
multimodal, multilingual learning. WIT is composed of a curated set of 37.6
million entity rich image-text examples with 11.5 million unique images across
108 Wikipedia languages. Its size enables WIT to be used as a pretraining
dataset for multimodal models, as we show when applied to downstream tasks such
as image-text retrieval. WIT has four main and unique advantages. First, WIT is
the largest multimodal dataset by the number of image-text examples by 3x (at
the time of writing). Second, WIT is massively multilingual (first of its kind)
with coverage over 100+ languages (each of which has at least 12K examples) and
provides cross-lingual texts for many images. Third, WIT represents a more
diverse set of concepts and real world entities relative to what previous
datasets cover. Lastly, WIT provides a very challenging real-world test set, as
we empirically illustrate using an image-text retrieval task as an example.

本文介绍了 Wikipedia-based Image Text（WIT）数据集，该数据集是由不同语言构成的 37.6 百万个实体丰富的图像文本示例组成，可用于多模态模型的预训练，如应用于图像文本检索等下游任务。WIT 数据集有四个主要的优势，规模大、多语种、覆盖的概念和实体比以前的数据集更加多样化，并提供一个非常具有挑战性的真实世界的测试集。