Informative Image Captioning with External Sources of Information

TL;DR

This paper introduces a multimodal Transformer-based model that combines image features and external fine-grained entity labels to generate more informative and fluent image captions.

Contribution

It proposes a novel multi-encoder Transformer model that integrates external entity labels with image features for improved caption informativeness.

Findings

Captions include more fine-grained entity mentions.

Model achieves better informativeness without sacrificing fluency.

Controlled appearance of entity labels in captions.

Abstract

An image caption should fluently present the essential information in a given image, including informative, fine-grained entity mentions and the manner in which these entities interact. However, current captioning models are usually trained to generate captions that only contain common object names, thus falling short on an important "informativeness" dimension. We present a mechanism for integrating image information together with fine-grained labels (assumed to be generated by some upstream models) into a caption that describes the image in a fluent and informative manner. We introduce a multimodal, multi-encoder model based on Transformer that ingests both image features and multiple sources of entity labels. We demonstrate that we can learn to control the appearance of these entity labels in the output, resulting in captions that are both fluent and informative.