Hierarchical Feature Alignment-based Progressive Addition Network for Multimodal Change Detection

Tongfei Liu, Yan Pu, Tao Lei, Jianjian Xu, Maoguo Gong, Lifeng He, Asoke K. Nandi

Abstract: Highlights•A novel HFA-PANet is proposed for multimodal change detection.•A hierarchical feature alignment module is devised to mitigate modality differences.•A progressive addition module is built to aggregate difference features.•A hierarchical domain alignment loss is introduced to train the network.