Remote Sensing Cross-Modal Text-Image Retrieval Based on Global and Local Information

TL;DR

This paper introduces a novel cross-modal remote sensing text-image retrieval framework that effectively combines global and local features, improving retrieval accuracy by dynamic feature fusion and enhanced local representations.

Contribution

The paper proposes GaLR, a new RSCTIR framework with a multi-level dynamic fusion module and enhanced local feature extraction using DREA, achieving state-of-the-art results.

Findings

GaLR outperforms existing methods on public datasets.

The DREA module improves local feature quality.

The multivariate rerank algorithm enhances retrieval precision.

Abstract

Cross-modal remote sensing text-image retrieval (RSCTIR) has recently become an urgent research hotspot due to its ability of enabling fast and flexible information extraction on remote sensing (RS) images. However, current RSCTIR methods mainly focus on global features of RS images, which leads to the neglect of local features that reflect target relationships and saliency. In this article, we first propose a novel RSCTIR framework based on global and local information (GaLR), and design a multi-level information dynamic fusion (MIDF) module to efficaciously integrate features of different levels. MIDF leverages local information to correct global information, utilizes global information to supplement local information, and uses the dynamic addition of the two to generate prominent visual representation. To alleviate the pressure of the redundant targets on the graph convolution network (GCN) and to improve the model s attention on salient instances during modeling local features, the de-noised representation matrix and the enhanced adjacency matrix (DREA) are devised to assist GCN in producing superior local representations. DREA not only filters out redundant features with high similarity, but also obtains more powerful local features by enhancing the features of prominent objects. Finally, to make full use of the information in the similarity matrix during inference, we come up with a plug-and-play multivariate rerank (MR) algorithm. The algorithm utilizes the k nearest neighbors of the retrieval results to perform a reverse search, and improves the performance by combining multiple components of bidirectional retrieval. Extensive experiments on public datasets strongly demonstrate the state-of-the-art performance of GaLR methods on the RSCTIR task. The code of GaLR method, MR algorithm, and corresponding files have been made available at https://github.com/xiaoyuan1996/GaLR .