Multimodal Compact Bilinear Pooling for Multimodal Neural Machine Translation

TL;DR

This paper explores the use of Multimodal Compact Bilinear pooling to improve multimodal neural machine translation by effectively combining textual and visual features, leading to better translation quality.

Contribution

It introduces the application of Multimodal Compact Bilinear pooling to multimodal neural machine translation, demonstrating its advantages over traditional combination methods.

Findings

Improved translation performance with bilinear pooling.

Effective integration of image and text features.

Enhanced attention mechanism for multimodal data.

Abstract

In state-of-the-art Neural Machine Translation, an attention mechanism is used during decoding to enhance the translation. At every step, the decoder uses this mechanism to focus on different parts of the source sentence to gather the most useful information before outputting its target word. Recently, the effectiveness of the attention mechanism has also been explored for multimodal tasks, where it becomes possible to focus both on sentence parts and image regions. Approaches to pool two modalities usually include element-wise product, sum or concatenation. In this paper, we evaluate the more advanced Multimodal Compact Bilinear pooling method, which takes the outer product of two vectors to combine the attention features for the two modalities. This has been previously investigated for visual question answering. We try out this approach for multimodal image caption translation and show improvements compared to basic combination methods.