Attention-Embedded Triple-Fusion Branch CNN for Hyperspectral Image Classification
Hyperspectral imaging (HSI) is widely used in various fields owing to its rich spectral information. Nonetheless, the high dimensionality of HSI and the limited number of labeled data remain significant obstacles to HSI classification technology. To alleviate the above problems, we propose an attent...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-04-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/8/2150 |
| _version_ | 1797603618804727808 |
|---|---|
| author | Erlei Zhang Jiayi Zhang Jiaxin Bai Jiarong Bian Shaoyi Fang Tao Zhan Mingchen Feng |
| author_facet | Erlei Zhang Jiayi Zhang Jiaxin Bai Jiarong Bian Shaoyi Fang Tao Zhan Mingchen Feng |
| author_sort | Erlei Zhang |
| collection | DOAJ |
| description | Hyperspectral imaging (HSI) is widely used in various fields owing to its rich spectral information. Nonetheless, the high dimensionality of HSI and the limited number of labeled data remain significant obstacles to HSI classification technology. To alleviate the above problems, we propose an attention-embedded triple-branch fusion convolutional neural network (AETF-Net) for an HSI classification. The network consists of a spectral attention branch, a spatial attention branch, and a multi-attention fusion branch (MAFB). The spectral branch introduces the cross-channel attention to alleviate the band redundancy problem in high dimensions, while the spatial branch preserves the location information of features and eliminates interfering image elements by a bi-directional spatial attention module. These pre-extracted spectral and spatial attention features are then embedded into a novel MAFB with large kernel decomposition technique. The proposed AETF-Net achieves multi-attention features reuse and extracts more representative and discriminative features. Experimental results on three well-known datasets demonstrate the superiority of the method AETF-Net. |
| first_indexed | 2024-03-11T04:34:32Z |
| format | Article |
| id | doaj.art-f112ddc34c4d45afbf669002e4aa8365 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-11T04:34:32Z |
| publishDate | 2023-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-f112ddc34c4d45afbf669002e4aa83652023-11-17T21:12:46ZengMDPI AGRemote Sensing2072-42922023-04-01158215010.3390/rs15082150Attention-Embedded Triple-Fusion Branch CNN for Hyperspectral Image ClassificationErlei Zhang0Jiayi Zhang1Jiaxin Bai2Jiarong Bian3Shaoyi Fang4Tao Zhan5Mingchen Feng6School of Information Engineering, Northwest A&F University, Xi’an 712100, ChinaSchool of Information Engineering, Northwest A&F University, Xi’an 712100, ChinaSchool of Information Engineering, Northwest A&F University, Xi’an 712100, ChinaSchool of Information Engineering, Northwest A&F University, Xi’an 712100, ChinaSchool of Information Engineering, Northwest A&F University, Xi’an 712100, ChinaSchool of Information Engineering, Northwest A&F University, Xi’an 712100, ChinaSchool of Information Engineering, Northwest A&F University, Xi’an 712100, ChinaHyperspectral imaging (HSI) is widely used in various fields owing to its rich spectral information. Nonetheless, the high dimensionality of HSI and the limited number of labeled data remain significant obstacles to HSI classification technology. To alleviate the above problems, we propose an attention-embedded triple-branch fusion convolutional neural network (AETF-Net) for an HSI classification. The network consists of a spectral attention branch, a spatial attention branch, and a multi-attention fusion branch (MAFB). The spectral branch introduces the cross-channel attention to alleviate the band redundancy problem in high dimensions, while the spatial branch preserves the location information of features and eliminates interfering image elements by a bi-directional spatial attention module. These pre-extracted spectral and spatial attention features are then embedded into a novel MAFB with large kernel decomposition technique. The proposed AETF-Net achieves multi-attention features reuse and extracts more representative and discriminative features. Experimental results on three well-known datasets demonstrate the superiority of the method AETF-Net.https://www.mdpi.com/2072-4292/15/8/2150hyperspectral image classificationattention mechanismfeature fusiondeep learning |
| spellingShingle | Erlei Zhang Jiayi Zhang Jiaxin Bai Jiarong Bian Shaoyi Fang Tao Zhan Mingchen Feng Attention-Embedded Triple-Fusion Branch CNN for Hyperspectral Image Classification Remote Sensing hyperspectral image classification attention mechanism feature fusion deep learning |
| title | Attention-Embedded Triple-Fusion Branch CNN for Hyperspectral Image Classification |
| title_full | Attention-Embedded Triple-Fusion Branch CNN for Hyperspectral Image Classification |
| title_fullStr | Attention-Embedded Triple-Fusion Branch CNN for Hyperspectral Image Classification |
| title_full_unstemmed | Attention-Embedded Triple-Fusion Branch CNN for Hyperspectral Image Classification |
| title_short | Attention-Embedded Triple-Fusion Branch CNN for Hyperspectral Image Classification |
| title_sort | attention embedded triple fusion branch cnn for hyperspectral image classification |
| topic | hyperspectral image classification attention mechanism feature fusion deep learning |
| url | https://www.mdpi.com/2072-4292/15/8/2150 |
| work_keys_str_mv | AT erleizhang attentionembeddedtriplefusionbranchcnnforhyperspectralimageclassification AT jiayizhang attentionembeddedtriplefusionbranchcnnforhyperspectralimageclassification AT jiaxinbai attentionembeddedtriplefusionbranchcnnforhyperspectralimageclassification AT jiarongbian attentionembeddedtriplefusionbranchcnnforhyperspectralimageclassification AT shaoyifang attentionembeddedtriplefusionbranchcnnforhyperspectralimageclassification AT taozhan attentionembeddedtriplefusionbranchcnnforhyperspectralimageclassification AT mingchenfeng attentionembeddedtriplefusionbranchcnnforhyperspectralimageclassification |