SOTAVerified

Multimodal Machine Translation

Multimodal machine translation is the task of doing machine translation with multiple data sources - for example, translating "a bird is flying over water" + an image of a bird over water to German text.

( Image credit: Findings of the Third Shared Task on Multimodal Machine Translation )

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 51–75 of 108 papers

TitleStatusHype
MSVD-Turkish: A Comprehensive Multimodal Dataset for Integrated Vision and Language Research in Turkish—0
TMU Japanese-English Multimodal Machine Translation System for WAT 2020—0
Generative Imagination Elevates Machine Translation—0
Dynamic Context-guided Capsule Network for Multimodal Machine TranslationCode1
Multimodal Transformer for Multimodal Machine TranslationCode1
Self-Knowledge Distillation with Progressive Refinement of TargetsCode1
M3P: Learning Universal Representations via Multitask Multilingual Multimodal Pre-trainingCode1
A Visually-Grounded Parallel Corpus with Phrase-to-Region Linking—0
Investigating the Decoders of Maximum Likelihood Sequence Models: A Look-ahead Approach—0
Multimodal Machine Translation through Visuals and Speech—0
Adaptive Fusion Techniques for Multimodal Data—0
Understanding the Effect of Textual Adversaries in Multimodal Machine Translation—0
Transformer-based Cascaded Multimodal Speech Translation—0
On Leveraging the Visual Modality for Neural Machine Translation—0
Probing Representations Learned by Multimodal Recurrent and Transformer Models—0
Multilingual Multimodal Machine Translation for Dravidian Languages utilizing Phonetic Transcription—0
Hindi Visual Genome: A Dataset for Multimodal English-to-Hindi Machine Translation—0
Distilling Translations with Visual AwarenessCode0
Grounded Word Sense Translation—0
Debiasing Word Embeddings Improves Multimodal Machine Translation—0
Multimodal Machine Translation with Embedding PredictionCode0
Probing the Need for Visual Context in Multimodal Machine Translation—0
Latent Variable Model for Multi-modal TranslationCode0
UMONS Submission for WMT18 Multimodal Translation TaskCode0
Input Combination Strategies for Multi-Source Transformer Decoder—0
Show:102550
← PrevPage 3 of 5Next →
All datasetsMulti30KHindi Visual Genome (Challenge Set)Hindi Visual Genome (Test Set)

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1delMeteor (EN-FR)74.6—Unverified
2ERNIE-UniX2BLEU (EN-DE)49.3—Unverified
3IKD-MMTBLEU (EN-DE)41.28—Unverified
4DCCNBLEU (EN-DE)39.7—Unverified
5CaglayanBLEU (EN-DE)39.4—Unverified
6Gumbel-Attention MMTBLEU (EN-DE)39.2—Unverified
7Multimodal TransformerBLEU (EN-DE)38.7—Unverified
8ImagiTBLEU (EN-DE)38.4—Unverified
9del+objBLEU (EN-DE)38—Unverified
10VMMTFBLEU (EN-DE)37.6—Unverified
#ModelMetricClaimedVerifiedStatus
1ViTABLEU (EN-HI)51.6—Unverified
#ModelMetricClaimedVerifiedStatus
1ViTABLEU (EN-HI)44.6—Unverified