STLNet: Symmetric Transformer Learning Network for Remote Sensing Image Change Detection
Change detection (CD) signifies a pivotal domain within remote sensing image processing. The transformer has been introduced in the field of CD for its global perception capabilities. However, existing transformer-based methodologies serve primarily as mere direct feature extractors, rendering the a...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10804568/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841550835701514240 |
|---|---|
| author | Liye Mei Andong Huang Zhaoyi Ye Yaxiaer Yalikun Ying Wang Chuan Xu Wei Yang Xinghua Li |
| author_facet | Liye Mei Andong Huang Zhaoyi Ye Yaxiaer Yalikun Ying Wang Chuan Xu Wei Yang Xinghua Li |
| author_sort | Liye Mei |
| collection | DOAJ |
| description | Change detection (CD) signifies a pivotal domain within remote sensing image processing. The transformer has been introduced in the field of CD for its global perception capabilities. However, existing transformer-based methodologies serve primarily as mere direct feature extractors, rendering the attention mechanism within the decoder underutilized. To address this, we propose a symmetric transformer learning network (STLNet) specifically tailored for remote sensing image CD tasks, constructed entirely using transformers. The STLNet is designed to leverage the intrinsic capability of transformers to model extensive long-range dependencies effectively. This approach significantly bolsters the extraction of distinctive global-level features, thereby facilitating the accurate delineation of CD regions. Initially, we utilize an adaptive multigrain encoder to extract feature information from bitemporal images, thereby honing the focus on changing targets and providing deeper and more comprehensive information. Subsequently, we adopt an effective decoder architecture comprised of transformer structures, namely, local gather decoder (LGD). The LGD employs a multilevel semantic feature integration from the encoder to augment feature representation and interdependencies, crucial for detailing small changed areas effectively via a hierarchical attentional fusion block. Ultimately, the detection of changes is based on the rich semantic information provided by the LGD, enabling us to achieve enhanced precision in our remote sensing CD efforts. Results show that the proposed STLNet achieved F1 scores of 92.32% on the LEVIR-CD dataset, 90.01% on the WHU-CD dataset, and 82.15% on the SYSU-CD dataset, surpassing mainstream CD methods. |
| format | Article |
| id | doaj-art-477eb3054b6643afa9fd0f9e86ad14f2 |
| institution | Kabale University |
| issn | 1939-1404 2151-1535 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj-art-477eb3054b6643afa9fd0f9e86ad14f22025-01-10T00:00:19ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-01182655266710.1109/JSTARS.2024.351930510804568STLNet: Symmetric Transformer Learning Network for Remote Sensing Image Change DetectionLiye Mei0https://orcid.org/0000-0002-2555-9199Andong Huang1https://orcid.org/0009-0007-0644-7080Zhaoyi Ye2https://orcid.org/0000-0003-2193-9124Yaxiaer Yalikun3https://orcid.org/0000-0003-0569-6740Ying Wang4https://orcid.org/0009-0007-3655-9702Chuan Xu5https://orcid.org/0000-0002-7099-7833Wei Yang6https://orcid.org/0000-0002-2014-8120Xinghua Li7https://orcid.org/0000-0002-2094-6480School of Computer Science, Hubei University of Technology, Wuhan, ChinaSchool of Computer Science, Hubei University of Technology, Wuhan, ChinaSchool of Computer Science, Hubei University of Technology, Wuhan, ChinaDivision of Materials Science, Nara Institute of Science and Technology, Ikoma, JapanSchool of Information Science and Engineering, Wuchang Shouyi University, Wuhan, ChinaSchool of Computer Science, Hubei University of Technology, Wuhan, ChinaSchool of Information Science and Engineering, Wuchang Shouyi University, Wuhan, ChinaSchool of Remote Sensing Information Engineering, Wuhan University, Wuhan, ChinaChange detection (CD) signifies a pivotal domain within remote sensing image processing. The transformer has been introduced in the field of CD for its global perception capabilities. However, existing transformer-based methodologies serve primarily as mere direct feature extractors, rendering the attention mechanism within the decoder underutilized. To address this, we propose a symmetric transformer learning network (STLNet) specifically tailored for remote sensing image CD tasks, constructed entirely using transformers. The STLNet is designed to leverage the intrinsic capability of transformers to model extensive long-range dependencies effectively. This approach significantly bolsters the extraction of distinctive global-level features, thereby facilitating the accurate delineation of CD regions. Initially, we utilize an adaptive multigrain encoder to extract feature information from bitemporal images, thereby honing the focus on changing targets and providing deeper and more comprehensive information. Subsequently, we adopt an effective decoder architecture comprised of transformer structures, namely, local gather decoder (LGD). The LGD employs a multilevel semantic feature integration from the encoder to augment feature representation and interdependencies, crucial for detailing small changed areas effectively via a hierarchical attentional fusion block. Ultimately, the detection of changes is based on the rich semantic information provided by the LGD, enabling us to achieve enhanced precision in our remote sensing CD efforts. Results show that the proposed STLNet achieved F1 scores of 92.32% on the LEVIR-CD dataset, 90.01% on the WHU-CD dataset, and 82.15% on the SYSU-CD dataset, surpassing mainstream CD methods.https://ieeexplore.ieee.org/document/10804568/Change detection (CD)local gather decoder (LGD)remote sensing (RS)rich semantic informationsymmetric transformer |
| spellingShingle | Liye Mei Andong Huang Zhaoyi Ye Yaxiaer Yalikun Ying Wang Chuan Xu Wei Yang Xinghua Li STLNet: Symmetric Transformer Learning Network for Remote Sensing Image Change Detection IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Change detection (CD) local gather decoder (LGD) remote sensing (RS) rich semantic information symmetric transformer |
| title | STLNet: Symmetric Transformer Learning Network for Remote Sensing Image Change Detection |
| title_full | STLNet: Symmetric Transformer Learning Network for Remote Sensing Image Change Detection |
| title_fullStr | STLNet: Symmetric Transformer Learning Network for Remote Sensing Image Change Detection |
| title_full_unstemmed | STLNet: Symmetric Transformer Learning Network for Remote Sensing Image Change Detection |
| title_short | STLNet: Symmetric Transformer Learning Network for Remote Sensing Image Change Detection |
| title_sort | stlnet symmetric transformer learning network for remote sensing image change detection |
| topic | Change detection (CD) local gather decoder (LGD) remote sensing (RS) rich semantic information symmetric transformer |
| url | https://ieeexplore.ieee.org/document/10804568/ |
| work_keys_str_mv | AT liyemei stlnetsymmetrictransformerlearningnetworkforremotesensingimagechangedetection AT andonghuang stlnetsymmetrictransformerlearningnetworkforremotesensingimagechangedetection AT zhaoyiye stlnetsymmetrictransformerlearningnetworkforremotesensingimagechangedetection AT yaxiaeryalikun stlnetsymmetrictransformerlearningnetworkforremotesensingimagechangedetection AT yingwang stlnetsymmetrictransformerlearningnetworkforremotesensingimagechangedetection AT chuanxu stlnetsymmetrictransformerlearningnetworkforremotesensingimagechangedetection AT weiyang stlnetsymmetrictransformerlearningnetworkforremotesensingimagechangedetection AT xinghuali stlnetsymmetrictransformerlearningnetworkforremotesensingimagechangedetection |