AsCL: An Asymmetry-sensitive Contrastive Learning Method for Image-Text Retrieval with Cross-Modal Fusion

TL;DR

This paper introduces AsCL, a contrastive learning approach that addresses asymmetry in image-text retrieval by enhancing fine-grained differentiation and cross-modal fusion, leading to improved retrieval accuracy.

Contribution

The paper proposes a novel asymmetry-sensitive contrastive learning method combined with hierarchical cross-modal fusion for better image-text retrieval.

Findings

Outperforms existing methods on MSCOCO and Flickr30K datasets.

Effectively distinguishes fine-grained semantic differences across modalities.

Enhances concept alignment through multimodal attention mechanism.

Abstract

The image-text retrieval task aims to retrieve relevant information from a given image or text. The main challenge is to unify multimodal representation and distinguish fine-grained differences across modalities, thereby finding similar contents and filtering irrelevant contents. However, existing methods mainly focus on unified semantic representation and concept alignment for multi-modalities, while the fine-grained differences across modalities have rarely been studied before, making it difficult to solve the information asymmetry problem. In this paper, we propose a novel asymmetry-sensitive contrastive learning method. By generating corresponding positive and negative samples for different asymmetry types, our method can simultaneously ensure fine-grained semantic differentiation and unified semantic representation between multi-modalities. Additionally, a hierarchical cross-modal fusion method is proposed, which integrates global and local-level features through a multimodal attention mechanism to achieve concept alignment. Extensive experiments performed on MSCOCO and Flickr30K, demonstrate the effectiveness and superiority of our proposed method.