2025
IEEE Transactions on Geoscience and Remote Sensing
MDANet: A Mamba-Driven Domain Adaptation Network for Cropland Change Detection
Paper ↗Remote Sensing
PROFILE
For the task of building change detection in remote sensing images, SAR and multispectral image data are fused, and a deep learning model is adopted. The two temporal phases of the two real data modalities are used for training, with change detection labels providing supervision. Appropriate change detection methods are ultimately employed to mine the change characteristics of building elements across multi-temporal image sequences, thereby building a highly robust building change detection system.
Keywords: Remote Sensing Images, Building Elements, Change Detection, Deep Learning