Dual Branch Masked Transformer for Hyperspectral Image Classification

Kuo Li;Yushi Chen;Lingbo Huang
{"title":"Dual Branch Masked Transformer for Hyperspectral Image Classification","authors":"Kuo Li;Yushi Chen;Lingbo Huang","doi":"10.1109/LGRS.2024.3490534","DOIUrl":null,"url":null,"abstract":"Transformer has been widely used in hyperspectral image (HSI) classification tasks because of its ability to capture long-range dependencies. However, most Transformer-based classification methods lack the extraction of local information or do not combine spatial and spectral information well, resulting in insufficient extraction of features. To address these issues, in this study, a dual-branch masked Transformer (Dual-MTr) model is proposed. Masked Transformer (MTr) is used to pretrain vision transformer (ViT) by reconstruction of both masked spatial image and spectral spectrum, which embeds the local bias by the process of recovering from localized patches to the global original input. Different tokenization methods are used for different types of input data. Patch embedding with overlapping regions is used for 2-D spatial data and group embedding is used for 1-D spectral data. Supervised learning has been added to the pretraining process to enhance strong discriminability. Then, the dual-branch structure is proposed to combine the spatial and spectral features. To strengthen the connection between the two branches better, Kullback-Leibler (KL) divergence is used to measure the differences between the classification results of the two branches, and the loss resulting from the computed differences is incorporated into the training process. Experimental results from two hyperspectral datasets demonstrate the effectiveness of the proposed method compared to other methods.","PeriodicalId":91017,"journal":{"name":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","volume":"21 ","pages":"1-5"},"PeriodicalIF":0.0000,"publicationDate":"2024-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","FirstCategoryId":"1085","ListUrlMain":"https://ieeexplore.ieee.org/document/10745157/","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Transformer has been widely used in hyperspectral image (HSI) classification tasks because of its ability to capture long-range dependencies. However, most Transformer-based classification methods lack the extraction of local information or do not combine spatial and spectral information well, resulting in insufficient extraction of features. To address these issues, in this study, a dual-branch masked Transformer (Dual-MTr) model is proposed. Masked Transformer (MTr) is used to pretrain vision transformer (ViT) by reconstruction of both masked spatial image and spectral spectrum, which embeds the local bias by the process of recovering from localized patches to the global original input. Different tokenization methods are used for different types of input data. Patch embedding with overlapping regions is used for 2-D spatial data and group embedding is used for 1-D spectral data. Supervised learning has been added to the pretraining process to enhance strong discriminability. Then, the dual-branch structure is proposed to combine the spatial and spectral features. To strengthen the connection between the two branches better, Kullback-Leibler (KL) divergence is used to measure the differences between the classification results of the two branches, and the loss resulting from the computed differences is incorporated into the training process. Experimental results from two hyperspectral datasets demonstrate the effectiveness of the proposed method compared to other methods.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
用于高光谱图像分类的双分支屏蔽变换器
变换器因其捕捉长距离依赖关系的能力而被广泛应用于高光谱图像(HSI)分类任务中。然而,大多数基于变换器的分类方法缺乏对局部信息的提取,或者不能很好地结合空间和光谱信息,从而导致特征提取不充分。为了解决这些问题,本研究提出了双分支掩蔽变换器(Dual-MTr)模型。屏蔽变换器(MTr)是通过重建屏蔽空间图像和光谱来预训练视觉变换器(ViT)的,它通过从局部斑块恢复到全局原始输入的过程来嵌入局部偏差。不同类型的输入数据采用不同的标记化方法。二维空间数据使用重叠区域的补丁嵌入,一维光谱数据使用组嵌入。在预训练过程中加入了监督学习,以增强识别能力。然后,提出了双分支结构来结合空间和频谱特征。为了更好地加强两个分支之间的联系,使用库尔贝-莱布勒(KL)发散来测量两个分支分类结果之间的差异,并将计算差异产生的损失纳入训练过程。两个高光谱数据集的实验结果表明,与其他方法相比,建议的方法非常有效。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
Deeper and Broader Multimodal Fusion: Cascaded Forest-of-Experts for Land Cover Classification Impact of Targeted Sounding Observations From FY-4B GIIRS on Two Super Typhoon Forecasts in 2024 Structural Representation-Guided GAN for Remote Sensing Image Cloud Removal Multispectral Airborne LiDAR Point Cloud Classification With Maximum Entropy Hierarchical Pooling A Satellite Selection Algorithm for GNSS-R InSAR Elevation Deformation Retrieval
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1