Restricted Label-Based Self-Supervised Learning Using SAR and Multispectral Imagery for Local Climate Zone Classification
Deep learning techniques have garnered significant attention in remote sensing scene classification. However, obtaining a large volume of labeled data for supervised learning (SL) remains challenging. Additionally, SL methods frequently struggle with limited generalization ability. To address these...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-04-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/17/8/1335 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850144085391704064 |
|---|---|
| author | Amjad Nawaz Wei Yang Hongcheng Zeng Yamin Wang Jie Chen |
| author_facet | Amjad Nawaz Wei Yang Hongcheng Zeng Yamin Wang Jie Chen |
| author_sort | Amjad Nawaz |
| collection | DOAJ |
| description | Deep learning techniques have garnered significant attention in remote sensing scene classification. However, obtaining a large volume of labeled data for supervised learning (SL) remains challenging. Additionally, SL methods frequently struggle with limited generalization ability. To address these limitations, self-supervised multi-mode representation learning (SSMMRL) is introduced for local climate zone classification (LCZC). Unlike conventional supervised learning methods, SSMMRL utilizes a novel encoder architecture that exclusively processes augmented positive samples (PSs), eliminating the need for negative samples. An attention-guided fusion mechanism is integrated, using positive samples as a form of regularization. The novel encoder captures informative representations from the unannotated So2Sat-LCZ42 dataset, which are then leveraged to enhance performance in a challenging few-shot classification task with limited labeled samples. Co-registered Synthetic Aperture Radar (SAR) and Multispectral (MS) images are used for evaluation and training. This approach enables the model to exploit extensive unlabeled data, enhancing performance on downstream tasks. Experimental evaluations on the So2Sat-LCZ42 benchmark dataset show the efficacy of the SSMMRL method. Our method for LCZC outperforms state-of-the-art (SOTA) approaches. |
| format | Article |
| id | doaj-art-c86a29d5e48b4a3ca53a5eb04842801c |
| institution | OA Journals |
| issn | 2072-4292 |
| language | English |
| publishDate | 2025-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj-art-c86a29d5e48b4a3ca53a5eb04842801c2025-08-20T02:28:28ZengMDPI AGRemote Sensing2072-42922025-04-01178133510.3390/rs17081335Restricted Label-Based Self-Supervised Learning Using SAR and Multispectral Imagery for Local Climate Zone ClassificationAmjad Nawaz0Wei Yang1Hongcheng Zeng2Yamin Wang3Jie Chen4School of Electronics and Information Engineering, Beihang University, Beijing 100191, ChinaSchool of Electronics and Information Engineering, Beihang University, Beijing 100191, ChinaSchool of Electronics and Information Engineering, Beihang University, Beijing 100191, ChinaSchool of Electronics and Information Engineering, Beihang University, Beijing 100191, ChinaSchool of Electronics and Information Engineering, Beihang University, Beijing 100191, ChinaDeep learning techniques have garnered significant attention in remote sensing scene classification. However, obtaining a large volume of labeled data for supervised learning (SL) remains challenging. Additionally, SL methods frequently struggle with limited generalization ability. To address these limitations, self-supervised multi-mode representation learning (SSMMRL) is introduced for local climate zone classification (LCZC). Unlike conventional supervised learning methods, SSMMRL utilizes a novel encoder architecture that exclusively processes augmented positive samples (PSs), eliminating the need for negative samples. An attention-guided fusion mechanism is integrated, using positive samples as a form of regularization. The novel encoder captures informative representations from the unannotated So2Sat-LCZ42 dataset, which are then leveraged to enhance performance in a challenging few-shot classification task with limited labeled samples. Co-registered Synthetic Aperture Radar (SAR) and Multispectral (MS) images are used for evaluation and training. This approach enables the model to exploit extensive unlabeled data, enhancing performance on downstream tasks. Experimental evaluations on the So2Sat-LCZ42 benchmark dataset show the efficacy of the SSMMRL method. Our method for LCZC outperforms state-of-the-art (SOTA) approaches.https://www.mdpi.com/2072-4292/17/8/1335self supervised learning (SSL)synthetic aperture radar (SAR)multispectral (MS) |
| spellingShingle | Amjad Nawaz Wei Yang Hongcheng Zeng Yamin Wang Jie Chen Restricted Label-Based Self-Supervised Learning Using SAR and Multispectral Imagery for Local Climate Zone Classification Remote Sensing self supervised learning (SSL) synthetic aperture radar (SAR) multispectral (MS) |
| title | Restricted Label-Based Self-Supervised Learning Using SAR and Multispectral Imagery for Local Climate Zone Classification |
| title_full | Restricted Label-Based Self-Supervised Learning Using SAR and Multispectral Imagery for Local Climate Zone Classification |
| title_fullStr | Restricted Label-Based Self-Supervised Learning Using SAR and Multispectral Imagery for Local Climate Zone Classification |
| title_full_unstemmed | Restricted Label-Based Self-Supervised Learning Using SAR and Multispectral Imagery for Local Climate Zone Classification |
| title_short | Restricted Label-Based Self-Supervised Learning Using SAR and Multispectral Imagery for Local Climate Zone Classification |
| title_sort | restricted label based self supervised learning using sar and multispectral imagery for local climate zone classification |
| topic | self supervised learning (SSL) synthetic aperture radar (SAR) multispectral (MS) |
| url | https://www.mdpi.com/2072-4292/17/8/1335 |
| work_keys_str_mv | AT amjadnawaz restrictedlabelbasedselfsupervisedlearningusingsarandmultispectralimageryforlocalclimatezoneclassification AT weiyang restrictedlabelbasedselfsupervisedlearningusingsarandmultispectralimageryforlocalclimatezoneclassification AT hongchengzeng restrictedlabelbasedselfsupervisedlearningusingsarandmultispectralimageryforlocalclimatezoneclassification AT yaminwang restrictedlabelbasedselfsupervisedlearningusingsarandmultispectralimageryforlocalclimatezoneclassification AT jiechen restrictedlabelbasedselfsupervisedlearningusingsarandmultispectralimageryforlocalclimatezoneclassification |