CNN-Transformer結(jié)合對(duì)比學(xué)習(xí)的高光譜與LiDAR數(shù)據(jù)協(xié)同分類

摘要： 針對(duì)高光譜圖像（hyperspectral images, HSI）與LiDAR數(shù)據(jù)多模態(tài)分類任務(wù)中的跨模態(tài)信息表達(dá)和特征對(duì)齊等問(wèn)題，提出一種基于對(duì)比學(xué)習(xí)CNN-Transformer高光譜和LiDAR數(shù)據(jù)協(xié)同分類網(wǎng)絡(luò)（Contrastive Learning based CNN-Transformer Network, CLCT-Net）。CLCT-Net通過(guò)由ConvNeX...