Multi30k dataset for multimodal machine translation

A multimodal machine translation dataset, Multi30k, which has short, simple and repetitive sentences.

BibTex: