Hybridizing Cross-Level Contextual and Attentive Representations for Remote Sensing Imagery Semantic Segmentation
Semantic segmentation of remote sensing imagery is a fundamental task in intelligent interpretation. Since deep convolutional neural networks (DCNNs) performed considerable insight in learning implicit representations from data, numerous works in recent years have transferred the DCNN-based model to...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-07-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/15/2986 |
| _version_ | 1797525159953825792 |
|---|---|
| author | Xin Li Feng Xu Runliang Xia Xin Lyu Hongmin Gao Yao Tong |
| author_facet | Xin Li Feng Xu Runliang Xia Xin Lyu Hongmin Gao Yao Tong |
| author_sort | Xin Li |
| collection | DOAJ |
| description | Semantic segmentation of remote sensing imagery is a fundamental task in intelligent interpretation. Since deep convolutional neural networks (DCNNs) performed considerable insight in learning implicit representations from data, numerous works in recent years have transferred the DCNN-based model to remote sensing data analysis. However, the wide-range observation areas, complex and diverse objects and illumination and imaging angle influence the pixels easily confused, leading to undesirable results. Therefore, a remote sensing imagery semantic segmentation neural network, named HCANet, is proposed to generate representative and discriminative representations for dense predictions. HCANet hybridizes cross-level contextual and attentive representations to emphasize the distinguishability of learned features. First of all, a cross-level contextual representation module (CCRM) is devised to exploit and harness the superpixel contextual information. Moreover, a hybrid representation enhancement module (HREM) is designed to fuse cross-level contextual and self-attentive representations flexibly. Furthermore, the decoder incorporates DUpsampling operation to boost the efficiency losslessly. The extensive experiments are implemented on the Vaihingen and Potsdam benchmarks. In addition, the results indicate that HCANet achieves excellent performance on overall accuracy and mean intersection over union. In addition, the ablation study further verifies the superiority of CCRM. |
| first_indexed | 2024-03-10T09:08:49Z |
| format | Article |
| id | doaj.art-10cd51ad0db245b4b07455b3e06a4df7 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T09:08:49Z |
| publishDate | 2021-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-10cd51ad0db245b4b07455b3e06a4df72023-11-22T06:07:10ZengMDPI AGRemote Sensing2072-42922021-07-011315298610.3390/rs13152986Hybridizing Cross-Level Contextual and Attentive Representations for Remote Sensing Imagery Semantic SegmentationXin Li0Feng Xu1Runliang Xia2Xin Lyu3Hongmin Gao4Yao Tong5College of Computer and Information, Hohai University, Nanjing 211100, ChinaCollege of Computer and Information, Hohai University, Nanjing 211100, ChinaInformation Engineering Center, Yellow River Institute of Hydraulic Research, Zhengzhou 450003, ChinaCollege of Computer and Information, Hohai University, Nanjing 211100, ChinaCollege of Computer and Information, Hohai University, Nanjing 211100, ChinaSchool of Information Engineering, Zhengzhou University, Zhengzhou 450001, ChinaSemantic segmentation of remote sensing imagery is a fundamental task in intelligent interpretation. Since deep convolutional neural networks (DCNNs) performed considerable insight in learning implicit representations from data, numerous works in recent years have transferred the DCNN-based model to remote sensing data analysis. However, the wide-range observation areas, complex and diverse objects and illumination and imaging angle influence the pixels easily confused, leading to undesirable results. Therefore, a remote sensing imagery semantic segmentation neural network, named HCANet, is proposed to generate representative and discriminative representations for dense predictions. HCANet hybridizes cross-level contextual and attentive representations to emphasize the distinguishability of learned features. First of all, a cross-level contextual representation module (CCRM) is devised to exploit and harness the superpixel contextual information. Moreover, a hybrid representation enhancement module (HREM) is designed to fuse cross-level contextual and self-attentive representations flexibly. Furthermore, the decoder incorporates DUpsampling operation to boost the efficiency losslessly. The extensive experiments are implemented on the Vaihingen and Potsdam benchmarks. In addition, the results indicate that HCANet achieves excellent performance on overall accuracy and mean intersection over union. In addition, the ablation study further verifies the superiority of CCRM.https://www.mdpi.com/2072-4292/13/15/2986semantic segmentationremote sensing imagerycross-level contextual informationrepresentation enhancement |
| spellingShingle | Xin Li Feng Xu Runliang Xia Xin Lyu Hongmin Gao Yao Tong Hybridizing Cross-Level Contextual and Attentive Representations for Remote Sensing Imagery Semantic Segmentation Remote Sensing semantic segmentation remote sensing imagery cross-level contextual information representation enhancement |
| title | Hybridizing Cross-Level Contextual and Attentive Representations for Remote Sensing Imagery Semantic Segmentation |
| title_full | Hybridizing Cross-Level Contextual and Attentive Representations for Remote Sensing Imagery Semantic Segmentation |
| title_fullStr | Hybridizing Cross-Level Contextual and Attentive Representations for Remote Sensing Imagery Semantic Segmentation |
| title_full_unstemmed | Hybridizing Cross-Level Contextual and Attentive Representations for Remote Sensing Imagery Semantic Segmentation |
| title_short | Hybridizing Cross-Level Contextual and Attentive Representations for Remote Sensing Imagery Semantic Segmentation |
| title_sort | hybridizing cross level contextual and attentive representations for remote sensing imagery semantic segmentation |
| topic | semantic segmentation remote sensing imagery cross-level contextual information representation enhancement |
| url | https://www.mdpi.com/2072-4292/13/15/2986 |
| work_keys_str_mv | AT xinli hybridizingcrosslevelcontextualandattentiverepresentationsforremotesensingimagerysemanticsegmentation AT fengxu hybridizingcrosslevelcontextualandattentiverepresentationsforremotesensingimagerysemanticsegmentation AT runliangxia hybridizingcrosslevelcontextualandattentiverepresentationsforremotesensingimagerysemanticsegmentation AT xinlyu hybridizingcrosslevelcontextualandattentiverepresentationsforremotesensingimagerysemanticsegmentation AT hongmingao hybridizingcrosslevelcontextualandattentiverepresentationsforremotesensingimagerysemanticsegmentation AT yaotong hybridizingcrosslevelcontextualandattentiverepresentationsforremotesensingimagerysemanticsegmentation |