ADHR-CDNet: Attentive Differential High-Resolution Change Detection Network for Remote Sensing Images.

With the development of deep learning, change detection technology has gained great progress. However, how to effectively extract multiscale substantive changed features and accurately detect small changed objects as well as the accurate details is still a challenge. To solve the problem, we propose attentive differential high-resolution change detection network (ADHR-CDNet) for remote sensing images. In ADHR-CDNet, a novel high-resolution backbone with a differential pyramid module (DPM) is proposed to extract multilevel and multiscale substantive changed features. The backbone structure with four interconnected subnetwork branches of different resolution is helpful to extract multilevel and multiscale features. DPM is capable of distinguishing between substantive changes and pseudochanges induced by illumination, shadow, seasonal variation, and so on. Then, a novel multiscale spatial feature attention module (MSSAM) is presented to effectively fuse the spatial detail information of different scale features produced by our backbone to generate finer prediction. We conduct quantitative and qualitative experiments on three public change detection datasets: the Lebedev, LEVIR building change detection dataset (LEVIR-CD), and Wuhan University (WHU) Building datasets. The proposed ADHR-CDNet reaches F1-score of 97.2% (improved 3.1%) on the Lebedev dataset, 91.4% (improved 1.6%) on the LEVIR-CD dataset, and 90.9% (improved 1.2%) on the WHU Building dataset. The experimental results demonstrate that our method performs much better than the state-of-the-art (SOTA) methods. The visualization comparison results show that our method can effectively detect small changed objects and significantly improve the details of detected changed objects.