融合ConvLSTM和多注意力机制网络的高光谱图像分类

唐婷, 潘新^*, 罗小玲, 郜晓晶

Fusion of ConvLSTM and Multi-Attention Mechanism Network for Hyperspectral Image Classification

TANG Ting, PAN Xin^*, LUO Xiao-ling, GAO Xiao-jing

使用0.5%训练样本的UP数据集的各个方法分类图 (a): 伪彩图; (b): 对应标签; (c): SVM (83.75%); (d): CDCNN (82.53%);(e): FDSSC (93.56%); (f): SSRN (91.94%); (g): DBDA (90.79%); (h): 本方法 (95.60%)