Matches in SemOpenAlex for { <https://semopenalex.org/work/W3213653482> ?p ?o ?g. }
- W3213653482 endingPage "3168" @default.
- W3213653482 startingPage "3153" @default.
- W3213653482 abstract "Neural Machine Translation models are sensitive to noise in the input texts, such as misspelled words and ungrammatical constructions. Existing robustness techniques generally fail when faced with unseen types of noise and their performance degrades on clean texts. In this paper, we focus on three types of realistic noise that are commonly generated by humans and introduce the idea of visual context to improve translation robustness for noisy texts. In addition, we describe a novel error correction training regime that can be used as an auxiliary task to further improve translation robustness. Experiments on English-French and English-German translation show that both multimodal and error correction components improve model robustness to noisy texts, while still retaining translation quality on clean texts." @default.
- W3213653482 created "2021-11-22" @default.
- W3213653482 creator A5053217291 @default.
- W3213653482 creator A5059143011 @default.
- W3213653482 creator A5062209641 @default.
- W3213653482 date "2021-11-01" @default.
- W3213653482 modified "2023-09-24" @default.
- W3213653482 title "Visual Cues and Error Correction for Translation Robustness" @default.
- W3213653482 cites W2098297786 @default.
- W3213653482 cites W2133459682 @default.
- W3213653482 cites W2153013403 @default.
- W3213653482 cites W2157331557 @default.
- W3213653482 cites W2470324779 @default.
- W3213653482 cites W2509282593 @default.
- W3213653482 cites W2512381898 @default.
- W3213653482 cites W2553522418 @default.
- W3213653482 cites W2581101319 @default.
- W3213653482 cites W2593341061 @default.
- W3213653482 cites W2741494657 @default.
- W3213653482 cites W2745461083 @default.
- W3213653482 cites W2811010710 @default.
- W3213653482 cites W2889903020 @default.
- W3213653482 cites W2903343986 @default.
- W3213653482 cites W2916835973 @default.
- W3213653482 cites W2950207430 @default.
- W3213653482 cites W2950886580 @default.
- W3213653482 cites W2962929176 @default.
- W3213653482 cites W2963062785 @default.
- W3213653482 cites W2963331233 @default.
- W3213653482 cites W2963360627 @default.
- W3213653482 cites W2963403868 @default.
- W3213653482 cites W2963407669 @default.
- W3213653482 cites W2963661177 @default.
- W3213653482 cites W2963823140 @default.
- W3213653482 cites W2964048171 @default.
- W3213653482 cites W2964121744 @default.
- W3213653482 cites W2964247056 @default.
- W3213653482 cites W2965350935 @default.
- W3213653482 cites W2970558573 @default.
- W3213653482 cites W2971134989 @default.
- W3213653482 cites W2979636403 @default.
- W3213653482 cites W2984051011 @default.
- W3213653482 cites W2994772196 @default.
- W3213653482 cites W3034773362 @default.
- W3213653482 cites W3035507081 @default.
- W3213653482 cites W3035577668 @default.
- W3213653482 cites W3098970007 @default.
- W3213653482 cites W3104234009 @default.
- W3213653482 cites W3118509208 @default.
- W3213653482 cites W3119067070 @default.
- W3213653482 cites W3120734549 @default.
- W3213653482 cites W3154987757 @default.
- W3213653482 cites W3156401640 @default.
- W3213653482 cites W2982671023 @default.
- W3213653482 hasPublicationYear "2021" @default.
- W3213653482 type Work @default.
- W3213653482 sameAs 3213653482 @default.
- W3213653482 citedByCount "0" @default.
- W3213653482 crossrefType "proceedings-article" @default.
- W3213653482 hasAuthorship W3213653482A5053217291 @default.
- W3213653482 hasAuthorship W3213653482A5059143011 @default.
- W3213653482 hasAuthorship W3213653482A5062209641 @default.
- W3213653482 hasConcept C103088060 @default.
- W3213653482 hasConcept C104317684 @default.
- W3213653482 hasConcept C11413529 @default.
- W3213653482 hasConcept C119857082 @default.
- W3213653482 hasConcept C153180895 @default.
- W3213653482 hasConcept C154945302 @default.
- W3213653482 hasConcept C185592680 @default.
- W3213653482 hasConcept C203005215 @default.
- W3213653482 hasConcept C204321447 @default.
- W3213653482 hasConcept C28490314 @default.
- W3213653482 hasConcept C41008148 @default.
- W3213653482 hasConcept C55493867 @default.
- W3213653482 hasConcept C63479239 @default.
- W3213653482 hasConceptScore W3213653482C103088060 @default.
- W3213653482 hasConceptScore W3213653482C104317684 @default.
- W3213653482 hasConceptScore W3213653482C11413529 @default.
- W3213653482 hasConceptScore W3213653482C119857082 @default.
- W3213653482 hasConceptScore W3213653482C153180895 @default.
- W3213653482 hasConceptScore W3213653482C154945302 @default.
- W3213653482 hasConceptScore W3213653482C185592680 @default.
- W3213653482 hasConceptScore W3213653482C203005215 @default.
- W3213653482 hasConceptScore W3213653482C204321447 @default.
- W3213653482 hasConceptScore W3213653482C28490314 @default.
- W3213653482 hasConceptScore W3213653482C41008148 @default.
- W3213653482 hasConceptScore W3213653482C55493867 @default.
- W3213653482 hasConceptScore W3213653482C63479239 @default.
- W3213653482 hasLocation W32136534821 @default.
- W3213653482 hasOpenAccess W3213653482 @default.
- W3213653482 hasPrimaryLocation W32136534821 @default.
- W3213653482 hasRelatedWork W118339868 @default.
- W3213653482 hasRelatedWork W2250458204 @default.
- W3213653482 hasRelatedWork W2898989428 @default.
- W3213653482 hasRelatedWork W2932851732 @default.
- W3213653482 hasRelatedWork W2949788262 @default.
- W3213653482 hasRelatedWork W2965116050 @default.
- W3213653482 hasRelatedWork W2982162582 @default.