HEANet: Hierarchical-Feature Enhanced Attention Network for Remote Sensing Change Detection

Change detection enables the detection of changes in objects from multi-temporal images. Recently, deep learning plays an important role in the field of change detection. Current methods perform multi-stage feature extraction from the input images to obtain high-level and low-level features, but ignoring the relationship between high-level features and low-level features. To deal with the above problem, this paper proposes a hierarchical-feature enhanced attention Network (HEANet), which integrates a hierarchical-feature enhanced attention (HEA) module for strengthening the association of hierarchical-feature and an adaptive scale enhancement (ASE) module for better feature representation. Extensive experiments show that our method achieves state-of-the-art performance compared to other methods on SYSU dataset.