Generalization algorithm of multimodal pre-training model based on graph-text self-supervised training

TL;DR

This paper proposes a self-supervised multimodal pre-training algorithm that enhances neural machine translation by effectively utilizing visual information without requiring manual image annotation, leading to improved translation performance.

Contribution

It introduces a novel self-supervised training method that leverages web-sourced images to improve multimodal neural machine translation, overcoming data scarcity issues.

Findings

BLEU score increased by 0.5 on the global voice dataset

Effective use of web-sourced images improves translation accuracy

Overcomes visual data limitations in multimodal NMT

Abstract

Recently, a large number of studies have shown that the introduction of visual information can effectively improve the effect of neural machine translation (NMT). Its effectiveness largely depends on the availability of a large number of bilingual parallel sentence pairs and manual image annotation. The lack of images and the effectiveness of images have been difficult to solve. In this paper, a multimodal pre-training generalization algorithm for self-supervised training is proposed, which overcomes the lack of visual information and inaccuracy, and thus extends the applicability of images on NMT. Specifically, we will search for many pictures from the existing sentences through the search engine, and then through the relationship between visual information and text, do the self-supervised training task of graphics and text to obtain more effective visual information for text. We show that when the filtered information is used as multimodal machine translation for fine-tuning, the effect of translation in the global voice dataset is 0.5 BLEU higher than the baseline.