Matches in SemOpenAlex for { <https://semopenalex.org/work/W3021584868> ?p ?o ?g. }
- W3021584868 abstract "Unsupervised machine translation (MT) has recently achieved impressive results with monolingual corpora only. However, it is still challenging to associate source-target sentences in the latent space. As people speak different languages biologically share similar visual systems, the potential of achieving better alignment through visual content is promising yet under-explored in unsupervised multimodal MT (MMT). In this paper, we investigate how to utilize visual content for disambiguation and promoting latent space alignment in unsupervised MMT. Our model employs multimodal back-translation and features pseudo visual pivoting in which we learn a shared multilingual visual-semantic embedding space and incorporate visually-pivoted captioning as additional weak supervision. The experimental results on the widely used Multi30K dataset show that the proposed model significantly improves over the state-of-the-art methods and generalizes well when the images are not available at the testing time." @default.
- W3021584868 created "2020-05-13" @default.
- W3021584868 creator A5034967388 @default.
- W3021584868 creator A5062546146 @default.
- W3021584868 creator A5063149046 @default.
- W3021584868 creator A5085263462 @default.
- W3021584868 date "2020-05-06" @default.
- W3021584868 modified "2023-09-23" @default.
- W3021584868 title "Unsupervised Multimodal Neural Machine Translation with Pseudo Visual Pivoting" @default.
- W3021584868 cites W1522301498 @default.
- W3021584868 cites W1523385540 @default.
- W3021584868 cites W1527575280 @default.
- W3021584868 cites W1753482797 @default.
- W3021584868 cites W1861492603 @default.
- W3021584868 cites W2124807415 @default.
- W3021584868 cites W2130942839 @default.
- W3021584868 cites W2194775991 @default.
- W3021584868 cites W2509282593 @default.
- W3021584868 cites W2513263213 @default.
- W3021584868 cites W2525778437 @default.
- W3021584868 cites W2546938941 @default.
- W3021584868 cites W2548228487 @default.
- W3021584868 cites W2550821151 @default.
- W3021584868 cites W2581101319 @default.
- W3021584868 cites W2610245951 @default.
- W3021584868 cites W2613718673 @default.
- W3021584868 cites W2745461083 @default.
- W3021584868 cites W2793424345 @default.
- W3021584868 cites W2808084195 @default.
- W3021584868 cites W2888070626 @default.
- W3021584868 cites W2891555348 @default.
- W3021584868 cites W2944815030 @default.
- W3021584868 cites W2950162424 @default.
- W3021584868 cites W2950848235 @default.
- W3021584868 cites W2962784628 @default.
- W3021584868 cites W2962793481 @default.
- W3021584868 cites W2962824887 @default.
- W3021584868 cites W2962968835 @default.
- W3021584868 cites W2963084599 @default.
- W3021584868 cites W2963206679 @default.
- W3021584868 cites W2963216553 @default.
- W3021584868 cites W2963341956 @default.
- W3021584868 cites W2963360627 @default.
- W3021584868 cites W2963386218 @default.
- W3021584868 cites W2963403868 @default.
- W3021584868 cites W2963407669 @default.
- W3021584868 cites W2963496089 @default.
- W3021584868 cites W2963602293 @default.
- W3021584868 cites W2963988211 @default.
- W3021584868 cites W2964007535 @default.
- W3021584868 cites W2964192290 @default.
- W3021584868 cites W2964308564 @default.
- W3021584868 cites W2970632400 @default.
- W3021584868 cites W2980216782 @default.
- W3021584868 cites W2981473723 @default.
- W3021584868 doi "https://doi.org/10.48550/arxiv.2005.03119" @default.
- W3021584868 hasPublicationYear "2020" @default.
- W3021584868 type Work @default.
- W3021584868 sameAs 3021584868 @default.
- W3021584868 citedByCount "3" @default.
- W3021584868 countsByYear W30215848682020 @default.
- W3021584868 countsByYear W30215848682021 @default.
- W3021584868 crossrefType "posted-content" @default.
- W3021584868 hasAuthorship W3021584868A5034967388 @default.
- W3021584868 hasAuthorship W3021584868A5062546146 @default.
- W3021584868 hasAuthorship W3021584868A5063149046 @default.
- W3021584868 hasAuthorship W3021584868A5085263462 @default.
- W3021584868 hasBestOaLocation W30215848681 @default.
- W3021584868 hasConcept C104317684 @default.
- W3021584868 hasConcept C105580179 @default.
- W3021584868 hasConcept C111919701 @default.
- W3021584868 hasConcept C115961682 @default.
- W3021584868 hasConcept C119857082 @default.
- W3021584868 hasConcept C149364088 @default.
- W3021584868 hasConcept C153180895 @default.
- W3021584868 hasConcept C154945302 @default.
- W3021584868 hasConcept C157657479 @default.
- W3021584868 hasConcept C169760540 @default.
- W3021584868 hasConcept C185592680 @default.
- W3021584868 hasConcept C203005215 @default.
- W3021584868 hasConcept C204321447 @default.
- W3021584868 hasConcept C207363949 @default.
- W3021584868 hasConcept C26760741 @default.
- W3021584868 hasConcept C2778572836 @default.
- W3021584868 hasConcept C41008148 @default.
- W3021584868 hasConcept C41608201 @default.
- W3021584868 hasConcept C55493867 @default.
- W3021584868 hasConcept C8038995 @default.
- W3021584868 hasConcept C86803240 @default.
- W3021584868 hasConceptScore W3021584868C104317684 @default.
- W3021584868 hasConceptScore W3021584868C105580179 @default.
- W3021584868 hasConceptScore W3021584868C111919701 @default.
- W3021584868 hasConceptScore W3021584868C115961682 @default.
- W3021584868 hasConceptScore W3021584868C119857082 @default.
- W3021584868 hasConceptScore W3021584868C149364088 @default.
- W3021584868 hasConceptScore W3021584868C153180895 @default.
- W3021584868 hasConceptScore W3021584868C154945302 @default.
- W3021584868 hasConceptScore W3021584868C157657479 @default.
- W3021584868 hasConceptScore W3021584868C169760540 @default.
- W3021584868 hasConceptScore W3021584868C185592680 @default.