A Survey of Vision-Language Pre-training from the Lens of Multimodal Machine Translation

TL;DR

This survey reviews vision-language pre-training models, focusing on their architectures and datasets, and discusses their potential and challenges in advancing multimodal machine translation.

Contribution

It provides a comprehensive overview of current vision-language pre-training methods specifically from the perspective of multimodal machine translation, highlighting gaps and future directions.

Findings

Pre-trained models like CLIP have improved image captioning and VQA.

There is limited research on applying these models to multimodal machine translation.

The paper identifies key architectures and datasets used in the field.

Abstract

Large language models such as BERT and the GPT series started a paradigm shift that calls for building general-purpose models via pre-training on large datasets, followed by fine-tuning on task-specific datasets. There is now a plethora of large pre-trained models for Natural Language Processing and Computer Vision. Recently, we have seen rapid developments in the joint Vision-Language space as well, where pre-trained models such as CLIP (Radford et al., 2021) have demonstrated improvements in downstream tasks like image captioning and visual question answering. However, surprisingly there is comparatively little work on exploring these models for the task of multimodal machine translation, where the goal is to leverage image/video modality in text-to-text translation. To fill this gap, this paper surveys the landscape of language-and-vision pre-training from the lens of multimodal machine translation. We summarize the common architectures, pre-training objectives, and datasets from literature and conjecture what further is needed to make progress on multimodal machine translation.