Deep Spectral Spatial Inverted Residual Network for Hyperspectral Image Classification
Convolutional neural networks (CNNs) have been widely used in hyperspectral image classification in recent years. The training of CNNs relies on a large amount of labeled sample data. However, the number of labeled samples of hyperspectral data is relatively small. Moreover, for hyperspectral images...
Guardado en:
Autores principales: | , , , |
---|---|
Formato: | article |
Lenguaje: | EN |
Publicado: |
MDPI AG
2021
|
Materias: | |
Acceso en línea: | https://doaj.org/article/bb856118ed184d3b857026df1cefcc80 |
Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
id |
oai:doaj.org-article:bb856118ed184d3b857026df1cefcc80 |
---|---|
record_format |
dspace |
spelling |
oai:doaj.org-article:bb856118ed184d3b857026df1cefcc802021-11-11T18:58:30ZDeep Spectral Spatial Inverted Residual Network for Hyperspectral Image Classification10.3390/rs132144722072-4292https://doaj.org/article/bb856118ed184d3b857026df1cefcc802021-11-01T00:00:00Zhttps://www.mdpi.com/2072-4292/13/21/4472https://doaj.org/toc/2072-4292Convolutional neural networks (CNNs) have been widely used in hyperspectral image classification in recent years. The training of CNNs relies on a large amount of labeled sample data. However, the number of labeled samples of hyperspectral data is relatively small. Moreover, for hyperspectral images, fully extracting spectral and spatial feature information is the key to achieve high classification performance. To solve the above issues, a deep spectral spatial inverted residuals network (DSSIRNet) is proposed. In this network, a data block random erasing strategy is introduced to alleviate the problem of limited labeled samples by data augmentation of small spatial blocks. In addition, a deep inverted residuals (DIR) module for spectral spatial feature extraction is proposed, which locks the effective features of each layer while avoiding network degradation. Furthermore, a global 3D attention module is proposed, which can realize the fine extraction of spectral and spatial global context information under the condition of the same number of input and output feature maps. Experiments are carried out on four commonly used hyperspectral datasets. A large number of experimental results show that compared with some state-of-the-art classification methods, the proposed method can provide higher classification accuracy for hyperspectral images.Tianyu ZhangCuiping ShiDiling LiaoLiguo WangMDPI AGarticlehyperspectral imageclassificationdeep feature extractioninverted residual convolution networkglobal 3D attentionScienceQENRemote Sensing, Vol 13, Iss 4472, p 4472 (2021) |
institution |
DOAJ |
collection |
DOAJ |
language |
EN |
topic |
hyperspectral image classification deep feature extraction inverted residual convolution network global 3D attention Science Q |
spellingShingle |
hyperspectral image classification deep feature extraction inverted residual convolution network global 3D attention Science Q Tianyu Zhang Cuiping Shi Diling Liao Liguo Wang Deep Spectral Spatial Inverted Residual Network for Hyperspectral Image Classification |
description |
Convolutional neural networks (CNNs) have been widely used in hyperspectral image classification in recent years. The training of CNNs relies on a large amount of labeled sample data. However, the number of labeled samples of hyperspectral data is relatively small. Moreover, for hyperspectral images, fully extracting spectral and spatial feature information is the key to achieve high classification performance. To solve the above issues, a deep spectral spatial inverted residuals network (DSSIRNet) is proposed. In this network, a data block random erasing strategy is introduced to alleviate the problem of limited labeled samples by data augmentation of small spatial blocks. In addition, a deep inverted residuals (DIR) module for spectral spatial feature extraction is proposed, which locks the effective features of each layer while avoiding network degradation. Furthermore, a global 3D attention module is proposed, which can realize the fine extraction of spectral and spatial global context information under the condition of the same number of input and output feature maps. Experiments are carried out on four commonly used hyperspectral datasets. A large number of experimental results show that compared with some state-of-the-art classification methods, the proposed method can provide higher classification accuracy for hyperspectral images. |
format |
article |
author |
Tianyu Zhang Cuiping Shi Diling Liao Liguo Wang |
author_facet |
Tianyu Zhang Cuiping Shi Diling Liao Liguo Wang |
author_sort |
Tianyu Zhang |
title |
Deep Spectral Spatial Inverted Residual Network for Hyperspectral Image Classification |
title_short |
Deep Spectral Spatial Inverted Residual Network for Hyperspectral Image Classification |
title_full |
Deep Spectral Spatial Inverted Residual Network for Hyperspectral Image Classification |
title_fullStr |
Deep Spectral Spatial Inverted Residual Network for Hyperspectral Image Classification |
title_full_unstemmed |
Deep Spectral Spatial Inverted Residual Network for Hyperspectral Image Classification |
title_sort |
deep spectral spatial inverted residual network for hyperspectral image classification |
publisher |
MDPI AG |
publishDate |
2021 |
url |
https://doaj.org/article/bb856118ed184d3b857026df1cefcc80 |
work_keys_str_mv |
AT tianyuzhang deepspectralspatialinvertedresidualnetworkforhyperspectralimageclassification AT cuipingshi deepspectralspatialinvertedresidualnetworkforhyperspectralimageclassification AT dilingliao deepspectralspatialinvertedresidualnetworkforhyperspectralimageclassification AT liguowang deepspectralspatialinvertedresidualnetworkforhyperspectralimageclassification |
_version_ |
1718431668551811072 |