CLIPTrans: Transferring Visual Knowledge with Pre-trained Models for Multimodal Machine Translation

TL;DR

CLIPTrans leverages pre-trained multilingual and multimodal models to improve multimodal machine translation, achieving state-of-the-art results without complex new modules by aligning embedding spaces through a lightweight mapping.

Contribution

It introduces a simple adaptation method that aligns pre-trained models for effective multimodal translation, especially in low-resource language scenarios.

Findings

Achieves an average of +2.67 BLEU over benchmarks.

Effectively aligns multilingual and multimodal embeddings.

Demonstrates strong generalization in low-resource settings.

Abstract

There has been a growing interest in developing multimodal machine translation (MMT) systems that enhance neural machine translation (NMT) with visual knowledge. This problem setup involves using images as auxiliary information during training, and more recently, eliminating their use during inference. Towards this end, previous works face a challenge in training powerful MMT models from scratch due to the scarcity of annotated multilingual vision-language data, especially for low-resource languages. Simultaneously, there has been an influx of multilingual pre-trained models for NMT and multimodal pre-trained models for vision-language tasks, primarily in English, which have shown exceptional generalisation ability. However, these are not directly applicable to MMT since they do not provide aligned multimodal multilingual features for generative tasks. To alleviate this issue, instead of designing complex modules for MMT, we propose CLIPTrans, which simply adapts the independently pre-trained multimodal M-CLIP and the multilingual mBART. In order to align their embedding spaces, mBART is conditioned on the M-CLIP features by a prefix sequence generated through a lightweight mapping network. We train this in a two-stage pipeline which warms up the model with image captioning before the actual translation task. Through experiments, we demonstrate the merits of this framework and consequently push forward the state-of-the-art across standard benchmarks by an average of +2.67 BLEU. The code can be found at www.github.com/devaansh100/CLIPTrans.