ViTA: Visual-Linguistic Translation by Aligning Object Tags
This work addresses the problem of integrating visual data into translation systems for researchers in multimodal AI, but it is incremental as it builds on existing pipelines and focuses on a specific language pair.
The paper tackles the lack of quality datasets in Multimodal Machine Translation (MMT) by proposing ViTA, a system that enhances textual input with visual information via object tag extraction for English-to-Hindi translation, achieving BLEU scores of 44.6 and 51.6 on test and challenge sets.
Multimodal Machine Translation (MMT) enriches the source text with visual information for translation. It has gained popularity in recent years, and several pipelines have been proposed in the same direction. Yet, the task lacks quality datasets to illustrate the contribution of visual modality in the translation systems. In this paper, we propose our system under the team name Volta for the Multimodal Translation Task of WAT 2021 from English to Hindi. We also participate in the textual-only subtask of the same language pair for which we use mBART, a pretrained multilingual sequence-to-sequence model. For multimodal translation, we propose to enhance the textual input by bringing the visual information to a textual domain by extracting object tags from the image. We also explore the robustness of our system by systematically degrading the source text. Finally, we achieve a BLEU score of 44.6 and 51.6 on the test set and challenge set of the multimodal task.