CS-HSNet: A Cross-Siamese Change Detection Network Based on Hierarchical-Split Attention
Qingtian Ke, Peng Zhang
Abstract
Change detection methods for optical remote sensing images play an important role in environmental resource management. Although recent methods based on deep learning demonstrate incredible ability by constructing networks: 1)Extracting bi-temporal features in a separate manner; 2)Fusing bi-temporal images before forwarding them into the single-level network. Both severely neglect the effect of spatial-temporal feature correlation between bi-temporal images. In addition, most existing methods represent multi-scale feature pairs in a layer-wise manner like ResNet, failing to consider the inner multi-level structure. In this work, we propose a new siamese change detection feature encoder backbone named Cross-Siamese Res2Net(CSRes2Net), by establishing crossed and hierarchical residual-like connections within one single residual block. The CSRes2Net represents dual features in a fine-grained manner and fully leads to the flow of bi-temporal features. In addition, recent learning-based methods designed some spatial-temporal relation modules to capture the pixel-level pairwise relationship and channel dependency based on self-attention mechanism, but they only consider spatial and channel dimension corrections separately with excessive parameters. So we propose a lightweight Cross Spatial-Channel Triplet Attention (CSCTA) module to capture cross-dimensional long-range relationship between triplet combinations: channel with height, channel with width, channel with channel. Finally, we propose a hierarchical-split block for generating multi-scale feature representations in a coarse-to-fine fashion. The experiments results on LEVIR-CD and SeasonVarying Change Detection Data Set outperform most state-ofthe-art models