Spectral-Spatial Attention Networks for Hyperspectral Image Classification
Many deep learning models, such as convolutional neural network (CNN) and recurrent neural network (RNN), have been successfully applied to extracting deep features for hyperspectral tasks. Hyperspectral image classification allows distinguishing the characterization of land covers by utilizing thei...
Main Authors: | , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2019-04-01
|
Series: | Remote Sensing |
Subjects: | |
Online Access: | https://www.mdpi.com/2072-4292/11/8/963 |
id |
doaj-6b7a1f3cc82446b58cceccddebc7ef27 |
---|---|
record_format |
Article |
spelling |
doaj-6b7a1f3cc82446b58cceccddebc7ef272020-11-25T00:52:24ZengMDPI AGRemote Sensing2072-42922019-04-0111896310.3390/rs11080963rs11080963Spectral-Spatial Attention Networks for Hyperspectral Image ClassificationXiaoguang Mei0Erting Pan1Yong Ma2Xiaobing Dai3Jun Huang4Fan Fan5Qinglei Du6Hong Zheng7Jiayi Ma8Electronic Information School, Wuhan University, Wuhan 430072, ChinaElectronic Information School, Wuhan University, Wuhan 430072, ChinaElectronic Information School, Wuhan University, Wuhan 430072, ChinaElectronic Information School, Wuhan University, Wuhan 430072, ChinaElectronic Information School, Wuhan University, Wuhan 430072, ChinaElectronic Information School, Wuhan University, Wuhan 430072, ChinaElectronic Information School, Wuhan University, Wuhan 430072, ChinaElectronic Information School, Wuhan University, Wuhan 430072, ChinaElectronic Information School, Wuhan University, Wuhan 430072, ChinaMany deep learning models, such as convolutional neural network (CNN) and recurrent neural network (RNN), have been successfully applied to extracting deep features for hyperspectral tasks. Hyperspectral image classification allows distinguishing the characterization of land covers by utilizing their abundant information. Motivated by the attention mechanism of the human visual system, in this study, we propose a spectral-spatial attention network for hyperspectral image classification. In our method, RNN with attention can learn inner spectral correlations within a continuous spectrum, while CNN with attention is designed to focus on saliency features and spatial relevance between neighboring pixels in the spatial dimension. Experimental results demonstrate that our method can fully utilize the spectral and spatial information to obtain competitive performance.https://www.mdpi.com/2072-4292/11/8/963hyperspectal image classificationattention mechanismRNNCNN |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Xiaoguang Mei Erting Pan Yong Ma Xiaobing Dai Jun Huang Fan Fan Qinglei Du Hong Zheng Jiayi Ma |
spellingShingle |
Xiaoguang Mei Erting Pan Yong Ma Xiaobing Dai Jun Huang Fan Fan Qinglei Du Hong Zheng Jiayi Ma Spectral-Spatial Attention Networks for Hyperspectral Image Classification Remote Sensing hyperspectal image classification attention mechanism RNN CNN |
author_facet |
Xiaoguang Mei Erting Pan Yong Ma Xiaobing Dai Jun Huang Fan Fan Qinglei Du Hong Zheng Jiayi Ma |
author_sort |
Xiaoguang Mei |
title |
Spectral-Spatial Attention Networks for Hyperspectral Image Classification |
title_short |
Spectral-Spatial Attention Networks for Hyperspectral Image Classification |
title_full |
Spectral-Spatial Attention Networks for Hyperspectral Image Classification |
title_fullStr |
Spectral-Spatial Attention Networks for Hyperspectral Image Classification |
title_full_unstemmed |
Spectral-Spatial Attention Networks for Hyperspectral Image Classification |
title_sort |
spectral-spatial attention networks for hyperspectral image classification |
publisher |
MDPI AG |
series |
Remote Sensing |
issn |
2072-4292 |
publishDate |
2019-04-01 |
description |
Many deep learning models, such as convolutional neural network (CNN) and recurrent neural network (RNN), have been successfully applied to extracting deep features for hyperspectral tasks. Hyperspectral image classification allows distinguishing the characterization of land covers by utilizing their abundant information. Motivated by the attention mechanism of the human visual system, in this study, we propose a spectral-spatial attention network for hyperspectral image classification. In our method, RNN with attention can learn inner spectral correlations within a continuous spectrum, while CNN with attention is designed to focus on saliency features and spatial relevance between neighboring pixels in the spatial dimension. Experimental results demonstrate that our method can fully utilize the spectral and spatial information to obtain competitive performance. |
topic |
hyperspectal image classification attention mechanism RNN CNN |
url |
https://www.mdpi.com/2072-4292/11/8/963 |
work_keys_str_mv |
AT xiaoguangmei spectralspatialattentionnetworksforhyperspectralimageclassification AT ertingpan spectralspatialattentionnetworksforhyperspectralimageclassification AT yongma spectralspatialattentionnetworksforhyperspectralimageclassification AT xiaobingdai spectralspatialattentionnetworksforhyperspectralimageclassification AT junhuang spectralspatialattentionnetworksforhyperspectralimageclassification AT fanfan spectralspatialattentionnetworksforhyperspectralimageclassification AT qingleidu spectralspatialattentionnetworksforhyperspectralimageclassification AT hongzheng spectralspatialattentionnetworksforhyperspectralimageclassification AT jiayima spectralspatialattentionnetworksforhyperspectralimageclassification |
_version_ |
1725242566934265856 |