m3P: Towards Multimodal Multilingual Translation with Multimodal Prompt
This work addresses translation quality issues in massively multilingual systems, particularly for low-resource scenarios, though it appears incremental as it builds on existing multimodal and multilingual translation approaches.
The paper tackles the problem of translation quality degradation in multilingual systems due to language differences by introducing visual context as a universal representation to facilitate translation, resulting in m3P outperforming previous text-only and multimodal methods by a large margin.
Multilingual translation supports multiple translation directions by projecting all languages in a shared space, but the translation quality is undermined by the difference between languages in the text-only modality, especially when the number of languages is large. To bridge this gap, we introduce visual context as the universal language-independent representation to facilitate multilingual translation. In this paper, we propose a framework to leverage the multimodal prompt to guide the Multimodal Multilingual neural Machine Translation (m3P), which aligns the representations of different languages with the same meaning and generates the conditional vision-language memory for translation. We construct a multilingual multimodal instruction dataset (InstrMulti102) to support 102 languages. Our method aims to minimize the representation distance of different languages by regarding the image as a central language. Experimental results show that m3P outperforms previous text-only baselines and multilingual multimodal methods by a large margin. Furthermore, the probing experiments validate the effectiveness of our method in enhancing translation under the low-resource and massively multilingual scenario.