Hierarchical Self-Attention Embedded Neural Network With Dense Connection for Remote-Sensing Image Semantic Segmentation
Semantic segmentation of remote-sensing imagery strives to assign a pixel-wise semantic label. Since encoder-decoder networks have demonstrated tremendous success in natural image semantic segmentation, the adoption and extension of this kind of method are transferring such superior performance for...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9535467/ |
| _version_ | 1818602162911969280 |
|---|---|
| author | Chunhua Li Xin Li Runliang Xia Tao Li Xin Lyu Yao Tong Liancheng Zhao Xinyuan Wang |
| author_facet | Chunhua Li Xin Li Runliang Xia Tao Li Xin Lyu Yao Tong Liancheng Zhao Xinyuan Wang |
| author_sort | Chunhua Li |
| collection | DOAJ |
| description | Semantic segmentation of remote-sensing imagery strives to assign a pixel-wise semantic label. Since encoder-decoder networks have demonstrated tremendous success in natural image semantic segmentation, the adoption and extension of this kind of method are transferring such superior performance for the problems in remote-sensing. Facing the high-altitude angle of imaging and complex and diverse ground objects of remote-sensing data, it is necessary to strengthen the features’ distinguishability by enhancing the network’s capability. Nevertheless, the existing methods suffer from the structural stereotype, leveraging the short-range and long-range contextual information insufficiently. Attempting to address the problems mentioned above, a hierarchical self-attention embedded neural network with dense connection for remote sensing image semantic segmentation (HSDCN) is proposed. In the encoder stage, multiple self-attention modules (SAM) are embedded to model pixel-wise and channel-wise relationships at various scales hierarchically, making the representations more refined and discriminative. Then the dense connections are used to fuse the heterogeneous features. Thus, the network could produce logical and reasonable clues for labeling pixels. The extensive experiments are conducted on ISPRS Vaihingen and Potsdam benchmarks. And the results reveal significant improvements in comparison with other state-of-the-art methods. |
| first_indexed | 2024-12-16T13:02:54Z |
| format | Article |
| id | doaj.art-a0712185f53e4cb190291d0ad974b501 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-16T13:02:54Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-a0712185f53e4cb190291d0ad974b5012022-12-21T22:30:47ZengIEEEIEEE Access2169-35362021-01-01912662312663410.1109/ACCESS.2021.31118999535467Hierarchical Self-Attention Embedded Neural Network With Dense Connection for Remote-Sensing Image Semantic SegmentationChunhua Li0https://orcid.org/0000-0002-2103-8396Xin Li1https://orcid.org/0000-0003-0576-3181Runliang Xia2https://orcid.org/0000-0001-6957-8546Tao Li3https://orcid.org/0000-0001-5538-1865Xin Lyu4https://orcid.org/0000-0003-1862-2070Yao Tong5https://orcid.org/0000-0002-7885-3922Liancheng Zhao6https://orcid.org/0000-0001-8139-2664Xinyuan Wang7https://orcid.org/0000-0002-3100-7120College of Computer and Information, Hohai University, Nanjing, ChinaCollege of Computer and Information, Hohai University, Nanjing, ChinaInformation Engineering Center, Yellow River Institute of Hydraulic Research, Zhengzhou, ChinaInformation Engineering Center, Yellow River Institute of Hydraulic Research, Zhengzhou, ChinaCollege of Computer and Information, Hohai University, Nanjing, ChinaSchool of Information Engineering, Zhengzhou University, Zhengzhou, ChinaCollege of Computer and Information, Hohai University, Nanjing, ChinaCollege of Computer and Information, Hohai University, Nanjing, ChinaSemantic segmentation of remote-sensing imagery strives to assign a pixel-wise semantic label. Since encoder-decoder networks have demonstrated tremendous success in natural image semantic segmentation, the adoption and extension of this kind of method are transferring such superior performance for the problems in remote-sensing. Facing the high-altitude angle of imaging and complex and diverse ground objects of remote-sensing data, it is necessary to strengthen the features’ distinguishability by enhancing the network’s capability. Nevertheless, the existing methods suffer from the structural stereotype, leveraging the short-range and long-range contextual information insufficiently. Attempting to address the problems mentioned above, a hierarchical self-attention embedded neural network with dense connection for remote sensing image semantic segmentation (HSDCN) is proposed. In the encoder stage, multiple self-attention modules (SAM) are embedded to model pixel-wise and channel-wise relationships at various scales hierarchically, making the representations more refined and discriminative. Then the dense connections are used to fuse the heterogeneous features. Thus, the network could produce logical and reasonable clues for labeling pixels. The extensive experiments are conducted on ISPRS Vaihingen and Potsdam benchmarks. And the results reveal significant improvements in comparison with other state-of-the-art methods.https://ieeexplore.ieee.org/document/9535467/Semantic segmentationremote-sensing imageryself-attentiondense connectionISPRS benchmarks |
| spellingShingle | Chunhua Li Xin Li Runliang Xia Tao Li Xin Lyu Yao Tong Liancheng Zhao Xinyuan Wang Hierarchical Self-Attention Embedded Neural Network With Dense Connection for Remote-Sensing Image Semantic Segmentation IEEE Access Semantic segmentation remote-sensing imagery self-attention dense connection ISPRS benchmarks |
| title | Hierarchical Self-Attention Embedded Neural Network With Dense Connection for Remote-Sensing Image Semantic Segmentation |
| title_full | Hierarchical Self-Attention Embedded Neural Network With Dense Connection for Remote-Sensing Image Semantic Segmentation |
| title_fullStr | Hierarchical Self-Attention Embedded Neural Network With Dense Connection for Remote-Sensing Image Semantic Segmentation |
| title_full_unstemmed | Hierarchical Self-Attention Embedded Neural Network With Dense Connection for Remote-Sensing Image Semantic Segmentation |
| title_short | Hierarchical Self-Attention Embedded Neural Network With Dense Connection for Remote-Sensing Image Semantic Segmentation |
| title_sort | hierarchical self attention embedded neural network with dense connection for remote sensing image semantic segmentation |
| topic | Semantic segmentation remote-sensing imagery self-attention dense connection ISPRS benchmarks |
| url | https://ieeexplore.ieee.org/document/9535467/ |
| work_keys_str_mv | AT chunhuali hierarchicalselfattentionembeddedneuralnetworkwithdenseconnectionforremotesensingimagesemanticsegmentation AT xinli hierarchicalselfattentionembeddedneuralnetworkwithdenseconnectionforremotesensingimagesemanticsegmentation AT runliangxia hierarchicalselfattentionembeddedneuralnetworkwithdenseconnectionforremotesensingimagesemanticsegmentation AT taoli hierarchicalselfattentionembeddedneuralnetworkwithdenseconnectionforremotesensingimagesemanticsegmentation AT xinlyu hierarchicalselfattentionembeddedneuralnetworkwithdenseconnectionforremotesensingimagesemanticsegmentation AT yaotong hierarchicalselfattentionembeddedneuralnetworkwithdenseconnectionforremotesensingimagesemanticsegmentation AT lianchengzhao hierarchicalselfattentionembeddedneuralnetworkwithdenseconnectionforremotesensingimagesemanticsegmentation AT xinyuanwang hierarchicalselfattentionembeddedneuralnetworkwithdenseconnectionforremotesensingimagesemanticsegmentation |