Multi30K

Introduced by Elliott et al. in Multi30K: Multilingual English-German Image Descriptions

Multi30K is a large-scale multilingual multimodal dataset for interdisciplinary machine learning research. It extends the Flickr30K dataset with German translations created by professional translators over a subset of the English descriptions, and descriptions crowdsourced independently of the original English descriptions. The dataset was introduced to stimulate multilingual multimodal research.

Homepage

Benchmarks

Add a new result Link an existing benchmark

Task	Dataset Variant	Best Model
Multimodal Machine Translation	Multi30K	ERNIE-UniX2
Translation deu-eng	multi30k_test_2016_flickr	opus-mt-tc-base-gmw-gmw
Translation eng-deu	multi30k_test_2016_flickr	opus-mt-tc-base-gmw-gmw
Translation deu-eng	multi30k_test_2017_flickr	opus-mt-tc-base-gmw-gmw
Translation eng-deu	multi30k_test_2017_flickr	opus-mt-tc-base-gmw-gmw
Translation deu-eng	multi30k_test_2017_mscoco	opus-mt-tc-base-gmw-gmw
Translation eng-deu	multi30k_test_2017_mscoco	opus-mt-tc-base-gmw-gmw
Translation deu-eng	multi30k_test_2018_flickr	opus-mt-tc-base-gmw-gmw
Translation eng-deu	multi30k_test_2018_flickr	opus-mt-tc-base-gmw-gmw