BASNet: A Boundary-Aware Siamese Network for Accurate Remote Sensing Change Detection

Change detection in remote sensing images is one of the most crucial topics in the computer vision community. Most recent change detection pipelines focus on introducing attention mechanism to enhance the discriminative ability of network, but their crude model architectures lead to inaccurate predictions and irregular boundaries. In this letter, we present a boundary-aware Siamese network (BASNet) for accurate remote sensing change detection. Based on the encoder-decoder architecture, we first propose a novel multi-scale paired fusion module (MPCM) to effectively fuse the same-level feature pairs from the Siamese encoding stream. In addition, we design a location guidance module (LGM) to accurately locate the changed regions. Based on the observation that hierarchical features show different level information, we propose a multi-level feature aggregation module (MFAM) to merge the bottom-up features. Finally, we introduce a hybrid loss that fuses Structural Similarity loss and Binary Cross Entropy loss to focus on the structural integrity and boundary quality of changed regions. Experimental results on two public datasets demonstrate that our proposed method significantly improves the performance and outperforms other state-of-the-art methods.