</> Code with Papers by Syntology

Datasets › Multi30K

Multi30K

Introduced by Desmond Elliott et al. in Multi30K: Multilingual English-German Image Descriptions2 May 2016 archive 2025-07-28

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.

HomepageBenchmarksPapers JSON

Benchmarks archive 2025-07-28

All 2 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

First row (archive order)PaperCode
Multimodal Machine Translation Multi30K ERNIE-UniX2 BLEU (EN-DE) 49.3 ERNIE-UniX2: A Unified Cross-lingual Cross-modal... — 15 Compare
Real-time Instance Segmentation multi30k_test_2017_mscoco RTMDet-Ins-s mask AP 38.7 RTMDet: An Empirical Study of Designing Real-Time Object... open-mmlab/mmdetection +13 1 Compare

Papers archive 2025-07-28

15 shown of 15 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 139. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

DateSamples run Syntology
RTMDet: An Empirical Study of Designing Real-Time Object Detectors 14 1 14 Dec 2022 ran 3 of 20 samples (17 unverified)
ERNIE-UniX2: A Unified Cross-lingual Cross-modal Framework for Understanding and Generation 0 1 9 Nov 2022 not harvested
Distill the Image to Nowhere: Inversion Knowledge Distillation for Multimodal Machine Translation 1 1 10 Oct 2022 ran 4 of 5 samples (1 unverified; 5 pointer-only for licence)
Gumbel-Attention for Multi-modal Machine Translation 0 1 16 Mar 2021 not harvested
Generative Imagination Elevates Machine Translation 0 1 21 Sep 2020 not harvested
Dynamic Context-guided Capsule Network for Multimodal Machine Translation 1 1 4 Sep 2020 not harvested
Multimodal Transformer for Multimodal Machine Translation 1 1 1 Jul 2020 not harvested
Self-Knowledge Distillation with Progressive Refinement of Targets 1 1 22 Jun 2020 ran 4 of 5 samples (1 unverified)
Multimodal Machine Translation through Visuals and Speech 0 1 28 Nov 2019 not harvested
Distilling Translations with Visual Awareness 1 2 18 Jun 2019 not harvested
Latent Variable Model for Multi-modal Translation 1 1 1 Nov 2018 not harvested
A Visual Attention Grounding Neural Model for Multimodal Machine Translation 1 1 24 Aug 2018 not harvested
Attention Is All You Need 595 1 12 Jun 2017 ran 600 of 946 samples (346 unverified; 451 pointer-only for licence)
Doubly-Attentive Decoder for Multi-modal Neural Machine Translation 0 1 4 Feb 2017 not harvested
Incorporating Global Visual Features into Attention-Based Neural Machine Translation 0 1 23 Jan 2017 not harvested

Dataset loaders archive 2025-07-28

No loader listed in the archive.

Tasks archive 2025-07-28

Translation deu-engTranslation eng-deuReal-time Instance SegmentationMultimodal Machine TranslationTranslation eng-fraTranslation ces-engTranslation fra-engTranslation ces-deuTranslation ces-fraTranslation deu-fraTranslation fra-deu

License archive 2025-07-28

No licence recorded in the archive. Absence here is not a statement about the dataset's terms.

Modalities archive 2025-07-28

No modality tagged.

Languages archive 2025-07-28

No language tagged.

Variants archive 2025-07-28

  • multi30k_test_2018_flickr eng-deu
  • multi30k_test_2018_flickr deu-eng
  • multi30k_test_2018_flickr
  • multi30k_test_2017_mscoco
  • multi30k_test_2017_flickr
  • multi30k_test_2016_flickr
  • multi30k_task2_test_2016
  • Multi30K

8 variant names, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections