Revising Image-Text Retrieval via Multi-Modal Entailment

TL;DR

This paper introduces a multi-modal entailment classifier to improve image-text retrieval datasets by identifying captions that are truly entailed by images, leading to better training and evaluation accuracy.

Contribution

The paper proposes a novel multi-modal entailment classifier and dataset revision method to address many-to-many caption-image matching issues in retrieval datasets.

Findings

Entailment classifier achieves about 78% accuracy.

Revised datasets improve retrieval model performance.

Enhanced dataset quality reduces confusion during training.

Abstract

An outstanding image-text retrieval model depends on high-quality labeled data. While the builders of existing image-text retrieval datasets strive to ensure that the caption matches the linked image, they cannot prevent a caption from fitting other images. We observe that such a many-to-many matching phenomenon is quite common in the widely-used retrieval datasets, where one caption can describe up to 178 images. These large matching-lost data not only confuse the model in training but also weaken the evaluation accuracy. Inspired by visual and textual entailment tasks, we propose a multi-modal entailment classifier to determine whether a sentence is entailed by an image plus its linked captions. Subsequently, we revise the image-text retrieval datasets by adding these entailed captions as additional weak labels of an image and develop a universal variable learning rate strategy to teach a retrieval model to distinguish the entailed captions from other negative samples. In experiments, we manually annotate an entailment-corrected image-text retrieval dataset for evaluation. The results demonstrate that the proposed entailment classifier achieves about 78% accuracy and consistently improves the performance of image-text retrieval baselines.