Clue: Cross-modal Coherence Modeling for Caption Generation

TL;DR

This paper introduces a coherence-based approach to image captioning, leveraging discourse relations to improve caption consistency and relevance, and presents a new annotation protocol and inference task for better modeling.

Contribution

It proposes a novel coherence relation prediction task for image captioning and demonstrates how coherence annotations enhance caption quality and control.

Findings

Significant improvement in caption consistency and quality.

Effective use of coherence annotations for relation classification.

Development of a new annotation protocol for image-caption coherence.

Abstract

We use coherence relations inspired by computational models of discourse to study the information needs and goals of image captioning. Using an annotation protocol specifically devised for capturing image--caption coherence relations, we annotate 10,000 instances from publicly-available image--caption pairs. We introduce a new task for learning inferences in imagery and text, coherence relation prediction, and show that these coherence annotations can be exploited to learn relation classifiers as an intermediary step, and also train coherence-aware, controllable image captioning models. The results show a dramatic improvement in the consistency and quality of the generated captions with respect to information needs specified via coherence relations.