Remote Sensing Building Change Detection With Global High-Frequency Cues Guidance and Result-Aware Alignment

Change detection (CD) is an important task in remote sensing (RS) imagery which has seen tremendous advances in recent years with the progress of deep neural networks. Although convolutional neural network (CNN) and transformer-based methods work sufficiently well, the building specificity is usually less considered to contribute to the performance improvement. In this letter, we propose a global and specific high-frequency cues guidance network for building CD, called GSHF-Net. More specifically, a high-frequency cues extraction module is introduced to provide global building-specific edge cues, which consists of a high-frequency cues extraction block (HFEB) and an attention-guidance block (AGB). Subsequently, we design a result-aware alignment to mitigate the effect of building slight offset in co- registered images on the introduction of high-frequency information to aid edge recovery. Extensive experiments were performed on LEVIR-CD and WHU-CD, and our GSHF-Net reached the state-of-the-art results of
$F 1~90.18$
and 89.53, respectively.