
Visual information can enhance multimodal machine translation, but its effectiveness depends on the alignment and coherence between textual and visual content. Supplementary textual information can often substitute for visual information in the translation process.


coremsg

evaluating-the-necessity-and-impact-of-visual-information-in-multimodal-machine-translation-using-real-world-datasets


Evaluating the Necessity and Impact of Visual Information in Multimodal Machine Translation using Real-World Datasets
