FTA-Net: Frequency-Temporal-Aware Network for Remote Sensing Change Detection
Change detection (CD) aims to explore surface changes in coaligned image pairs. However, many existing networks primarily focus on learning deep features, without considering the impact of attention and fusion strategies on detection performance. Therefore, a new frequency-temporal-aware network (FT...
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/10824909/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832592973832912896 |
|---|---|
| author | Taojun Zhu Zikai Zhao Min Xia Junqing Huang Liguo Weng Kai Hu Haifeng Lin Wenyu Zhao |
| author_facet | Taojun Zhu Zikai Zhao Min Xia Junqing Huang Liguo Weng Kai Hu Haifeng Lin Wenyu Zhao |
| author_sort | Taojun Zhu |
| collection | DOAJ |
| description | Change detection (CD) aims to explore surface changes in coaligned image pairs. However, many existing networks primarily focus on learning deep features, without considering the impact of attention and fusion strategies on detection performance. Therefore, a new frequency-temporal-aware network (FTA-Net) is proposed, it recognizes changes by means of a frequency-domain temporal fusion module and supervised attention to multilevel time-difference features, while reducing the model size. The frequency temporal fusion module is designed to introduce the frequency attention mechanism into the fusion process. First, it has a two-branch Transformer-INN feature extractor using a Lite-Transformer that utilizes remote attention for low-frequency global features, and a invertible neural network that focuses on extracting high-frequency local information. The semantic information and details of the object in both high-frequency and low-frequency feature maps are further strengthened by fusing the high-frequency local features and low-frequency global representations. Then, a stepwise modification detection module is proposed to better extract temporal difference information from bitemporal features. In addition, a supervised learning module is constructed to reweight features to efficiently aggregate multilevel features from high-level to low-level. FTA-Net outperforms state-of-the-art methods on three challenging CD datasets, and it have fewer parameters (4.93M) and lower computational cost (6.71 G). |
| format | Article |
| id | doaj-art-ff7b012ce8864769bb375291fe027bba |
| 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-ff7b012ce8864769bb375291fe027bba2025-01-21T00:00:13ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-01183448346010.1109/JSTARS.2025.352559510824909FTA-Net: Frequency-Temporal-Aware Network for Remote Sensing Change DetectionTaojun Zhu0Zikai Zhao1Min Xia2https://orcid.org/0000-0003-4681-9129Junqing Huang3https://orcid.org/0009-0000-4425-1168Liguo Weng4https://orcid.org/0000-0003-3734-3114Kai Hu5https://orcid.org/0000-0001-7181-9935Haifeng Lin6https://orcid.org/0000-0002-3835-6075Wenyu Zhao7Collaborative Innovation Center on Atmospheric Environment and Equipment Technology, Nanjing University of Information Science and Technology, Nanjing, ChinaCollaborative Innovation Center on Atmospheric Environment and Equipment Technology, Nanjing University of Information Science and Technology, Nanjing, ChinaCollaborative Innovation Center on Atmospheric Environment and Equipment Technology, Nanjing University of Information Science and Technology, Nanjing, ChinaFaculty of Applied Sciences, Macao Polytechnic University, Macau, ChinaCollaborative Innovation Center on Atmospheric Environment and Equipment Technology, Nanjing University of Information Science and Technology, Nanjing, ChinaCollaborative Innovation Center on Atmospheric Environment and Equipment Technology, Nanjing University of Information Science and Technology, Nanjing, ChinaCollege of Information Science and Technology, Nanjing Forestry University, Nanjing, ChinaDepartment of Computer Science, University of Reading, Reading, U.K.Change detection (CD) aims to explore surface changes in coaligned image pairs. However, many existing networks primarily focus on learning deep features, without considering the impact of attention and fusion strategies on detection performance. Therefore, a new frequency-temporal-aware network (FTA-Net) is proposed, it recognizes changes by means of a frequency-domain temporal fusion module and supervised attention to multilevel time-difference features, while reducing the model size. The frequency temporal fusion module is designed to introduce the frequency attention mechanism into the fusion process. First, it has a two-branch Transformer-INN feature extractor using a Lite-Transformer that utilizes remote attention for low-frequency global features, and a invertible neural network that focuses on extracting high-frequency local information. The semantic information and details of the object in both high-frequency and low-frequency feature maps are further strengthened by fusing the high-frequency local features and low-frequency global representations. Then, a stepwise modification detection module is proposed to better extract temporal difference information from bitemporal features. In addition, a supervised learning module is constructed to reweight features to efficiently aggregate multilevel features from high-level to low-level. FTA-Net outperforms state-of-the-art methods on three challenging CD datasets, and it have fewer parameters (4.93M) and lower computational cost (6.71 G).https://ieeexplore.ieee.org/document/10824909/Change detection (CD)deep learingfeature fusionremote sensing |
| spellingShingle | Taojun Zhu Zikai Zhao Min Xia Junqing Huang Liguo Weng Kai Hu Haifeng Lin Wenyu Zhao FTA-Net: Frequency-Temporal-Aware Network for Remote Sensing Change Detection IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Change detection (CD) deep learing feature fusion remote sensing |
| title | FTA-Net: Frequency-Temporal-Aware Network for Remote Sensing Change Detection |
| title_full | FTA-Net: Frequency-Temporal-Aware Network for Remote Sensing Change Detection |
| title_fullStr | FTA-Net: Frequency-Temporal-Aware Network for Remote Sensing Change Detection |
| title_full_unstemmed | FTA-Net: Frequency-Temporal-Aware Network for Remote Sensing Change Detection |
| title_short | FTA-Net: Frequency-Temporal-Aware Network for Remote Sensing Change Detection |
| title_sort | fta net frequency temporal aware network for remote sensing change detection |
| topic | Change detection (CD) deep learing feature fusion remote sensing |
| url | https://ieeexplore.ieee.org/document/10824909/ |
| work_keys_str_mv | AT taojunzhu ftanetfrequencytemporalawarenetworkforremotesensingchangedetection AT zikaizhao ftanetfrequencytemporalawarenetworkforremotesensingchangedetection AT minxia ftanetfrequencytemporalawarenetworkforremotesensingchangedetection AT junqinghuang ftanetfrequencytemporalawarenetworkforremotesensingchangedetection AT liguoweng ftanetfrequencytemporalawarenetworkforremotesensingchangedetection AT kaihu ftanetfrequencytemporalawarenetworkforremotesensingchangedetection AT haifenglin ftanetfrequencytemporalawarenetworkforremotesensingchangedetection AT wenyuzhao ftanetfrequencytemporalawarenetworkforremotesensingchangedetection |