Exploring the Cross-Temporal Interaction: Feature Exchange and Enhancement for Remote Sensing Change Detection

Abstract

Change detection (CD) aims to identify surface changes from bitemporal remote sensing (RS) images, which is a crucial and challenging topic in RS. In recent years, RS images CD has achieved significant advancements through the use of convolutional neural networks. However, existing deep learning-based CD methods still face some challenges, such as blurry boundaries, pseudochanges caused by fluctuations in imaging conditions, and complexity of change objects in the scene. In this study, we present a novel perspective for exploring the cross-temporal interaction in the feature fusion stage and propose a bitemporal feature exchange and enhancement network (ExNet). Specifically, we argue that the heterogeneity of bitemporal RS images leads to the failure of the CD model in some cases. Therefore, we attempt to achieve feature alignment in two aspects by feature exchange. On one hand, we exchange the statistical information of bitemporal features, facilitating the transfer of their style. On the other hand, bitemporal features are composed of content correlation embeddings and domain correlation embeddings (DCEs). We design a dynamic low-pass filter (DLF) to partially extract and exchange DCEs in bitemporal features to achieve the feature distribution alignment. Moreover, a frequency separation enhancement module is proposed, which transforms the fused features into the frequency domain and enhances the corresponding frequency representation. Comprehensive experimental results on three popular CD datasets demonstrate the effectiveness and efficiency of ExNet compared with state-of-the-art methods.

Keywords

• Change detection (CD)
• dynamic low-pass filter (DLF)
• feature alignment
• frequency domain enhancement