Hyperspectral Image Classification Based on Two-Branch Spectral–Spatial-Feature Attention Network

Although most of deep-learning-based hyperspectral image (HSI) classification methods achieve great performance, there still remains a challenge to utilize small-size training samples to remarkably enhance the classification accuracy. To tackle this challenge, a novel two-branch spectral–spatial-fea...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Hanjie Wu, Dan Li, Yujian Wang, Xiaojun Li, Fanqiang Kong, Qiang Wang
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	hyperspectral image classification spectral–spatial-feature extraction attention mechanism 2DCNN 3DCNN Science Q
Acceso en línea:	https://doaj.org/article/233a5449993e40928ca3248a48b09e23
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:233a5449993e40928ca3248a48b09e23
record_format	dspace
spelling	oai:doaj.org-article:233a5449993e40928ca3248a48b09e232021-11-11T18:52:01ZHyperspectral Image Classification Based on Two-Branch Spectral–Spatial-Feature Attention Network10.3390/rs132142622072-4292https://doaj.org/article/233a5449993e40928ca3248a48b09e232021-10-01T00:00:00Zhttps://www.mdpi.com/2072-4292/13/21/4262https://doaj.org/toc/2072-4292Although most of deep-learning-based hyperspectral image (HSI) classification methods achieve great performance, there still remains a challenge to utilize small-size training samples to remarkably enhance the classification accuracy. To tackle this challenge, a novel two-branch spectral–spatial-feature attention network (TSSFAN) for HSI classification is proposed in this paper. Firstly, two inputs with different spectral dimensions and spatial sizes are constructed, which can not only reduce the redundancy of the original dataset but also accurately explore the spectral and spatial features. Then, we design two parallel 3DCNN branches with attention modules, in which one focuses on extracting spectral features and adaptively learning the more discriminative spectral channels, and the other focuses on exploring spatial features and adaptively learning the more discriminative spatial structures. Next, the feature attention module is constructed to automatically adjust the weights of different features based on their contributions for classification to remarkably improve the classification performance. Finally, we design the hybrid architecture of 3D–2DCNN to acquire the final classification result, which can significantly decrease the sophistication of the network. Experimental results on three HSI datasets indicate that our presented TSSFAN method outperforms several of the most advanced classification methods.Hanjie WuDan LiYujian WangXiaojun LiFanqiang KongQiang WangMDPI AGarticlehyperspectral image classificationspectral–spatial-feature extractionattention mechanism2DCNN3DCNNScienceQENRemote Sensing, Vol 13, Iss 4262, p 4262 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	hyperspectral image classification spectral–spatial-feature extraction attention mechanism 2DCNN 3DCNN Science Q
spellingShingle	hyperspectral image classification spectral–spatial-feature extraction attention mechanism 2DCNN 3DCNN Science Q Hanjie Wu Dan Li Yujian Wang Xiaojun Li Fanqiang Kong Qiang Wang Hyperspectral Image Classification Based on Two-Branch Spectral–Spatial-Feature Attention Network
description	Although most of deep-learning-based hyperspectral image (HSI) classification methods achieve great performance, there still remains a challenge to utilize small-size training samples to remarkably enhance the classification accuracy. To tackle this challenge, a novel two-branch spectral–spatial-feature attention network (TSSFAN) for HSI classification is proposed in this paper. Firstly, two inputs with different spectral dimensions and spatial sizes are constructed, which can not only reduce the redundancy of the original dataset but also accurately explore the spectral and spatial features. Then, we design two parallel 3DCNN branches with attention modules, in which one focuses on extracting spectral features and adaptively learning the more discriminative spectral channels, and the other focuses on exploring spatial features and adaptively learning the more discriminative spatial structures. Next, the feature attention module is constructed to automatically adjust the weights of different features based on their contributions for classification to remarkably improve the classification performance. Finally, we design the hybrid architecture of 3D–2DCNN to acquire the final classification result, which can significantly decrease the sophistication of the network. Experimental results on three HSI datasets indicate that our presented TSSFAN method outperforms several of the most advanced classification methods.
format	article
author	Hanjie Wu Dan Li Yujian Wang Xiaojun Li Fanqiang Kong Qiang Wang
author_facet	Hanjie Wu Dan Li Yujian Wang Xiaojun Li Fanqiang Kong Qiang Wang
author_sort	Hanjie Wu
title	Hyperspectral Image Classification Based on Two-Branch Spectral–Spatial-Feature Attention Network
title_short	Hyperspectral Image Classification Based on Two-Branch Spectral–Spatial-Feature Attention Network
title_full	Hyperspectral Image Classification Based on Two-Branch Spectral–Spatial-Feature Attention Network
title_fullStr	Hyperspectral Image Classification Based on Two-Branch Spectral–Spatial-Feature Attention Network
title_full_unstemmed	Hyperspectral Image Classification Based on Two-Branch Spectral–Spatial-Feature Attention Network
title_sort	hyperspectral image classification based on two-branch spectral–spatial-feature attention network
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/233a5449993e40928ca3248a48b09e23
work_keys_str_mv	AT hanjiewu hyperspectralimageclassificationbasedontwobranchspectralspatialfeatureattentionnetwork AT danli hyperspectralimageclassificationbasedontwobranchspectralspatialfeatureattentionnetwork AT yujianwang hyperspectralimageclassificationbasedontwobranchspectralspatialfeatureattentionnetwork AT xiaojunli hyperspectralimageclassificationbasedontwobranchspectralspatialfeatureattentionnetwork AT fanqiangkong hyperspectralimageclassificationbasedontwobranchspectralspatialfeatureattentionnetwork AT qiangwang hyperspectralimageclassificationbasedontwobranchspectralspatialfeatureattentionnetwork
_version_	1718431691342610432

Hyperspectral Image Classification Based on Two-Branch Spectral–Spatial-Feature Attention Network

Ejemplares similares