Residual channel attention based sample adaptation few-shot learning for hyperspectral image classification
Abstract Few-shot learning (FSL) uses prior knowledge and supervised experience to effectively classify hyperspectral images (HSIs), thereby reducing the cost of large numbers of labeled samples. However, existing few-shot methods ignore the correlation between cross-domain feature channels, and the...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-11-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-024-77747-2 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846172030145134592 |
|---|---|
| author | Yuefeng Zhao Jingqi Sun Nannan Hu Chengmin Zai Yanwei Han |
| author_facet | Yuefeng Zhao Jingqi Sun Nannan Hu Chengmin Zai Yanwei Han |
| author_sort | Yuefeng Zhao |
| collection | DOAJ |
| description | Abstract Few-shot learning (FSL) uses prior knowledge and supervised experience to effectively classify hyperspectral images (HSIs), thereby reducing the cost of large numbers of labeled samples. However, existing few-shot methods ignore the correlation between cross-domain feature channels, and the feature representation ability is insufficient. To address above issue, this paper proposes a novel Residual Channel Attention Based Sample Adaptation Few-Shot Learning for Hyperspectral Image Classification(RCASA-FSL) for hyperspectral image classification (HSIC), which can capture and enhance cross-domain dependencies through multi-layer residual connection and random-based feature recalibration. Specifically, a Deep Residual Feature Channel Attention Mechanism (DRFCAM) is designed to obtain cross-domain dependencies by residual concatenation, and further the residual structure is stacked for mining depth discrimination information. Furthermore, a new Random-based Feature Recalibration Module (RFRM) is proposed to reassign the feature weights via random matrix, which fully explore feature weight relationships to guide the sample adaptation process. Besides, we design a joint loss function with combining the FSL loss and domain adaptive loss for further optimization model. Experiments conducted on several standard hyperspectral datasets demonstrate that the proposed RCASA-FSL is superior to other FSL techniques in both quantitative and qualitative aspects. |
| format | Article |
| id | doaj-art-5460b2e98ff94c49b26c1f823ffaf80e |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-5460b2e98ff94c49b26c1f823ffaf80e2024-11-10T12:22:00ZengNature PortfolioScientific Reports2045-23222024-11-0114111710.1038/s41598-024-77747-2Residual channel attention based sample adaptation few-shot learning for hyperspectral image classificationYuefeng Zhao0Jingqi Sun1Nannan Hu2Chengmin Zai3Yanwei Han4Shandong Provincial Engineering and Technical Center of Light Manipulation, Shandong Provincial Key Laboratory of Optics and Photonic Devices, School of Physics and Electronics, Shandong Normal UniversityShandong Provincial Engineering and Technical Center of Light Manipulation, Shandong Provincial Key Laboratory of Optics and Photonic Devices, School of Physics and Electronics, Shandong Normal UniversityShandong Provincial Engineering and Technical Center of Light Manipulation, Shandong Provincial Key Laboratory of Optics and Photonic Devices, School of Physics and Electronics, Shandong Normal UniversityShandong Provincial Engineering and Technical Center of Light Manipulation, Shandong Provincial Key Laboratory of Optics and Photonic Devices, School of Physics and Electronics, Shandong Normal UniversityShandong Provincial Engineering and Technical Center of Light Manipulation, Shandong Provincial Key Laboratory of Optics and Photonic Devices, School of Physics and Electronics, Shandong Normal UniversityAbstract Few-shot learning (FSL) uses prior knowledge and supervised experience to effectively classify hyperspectral images (HSIs), thereby reducing the cost of large numbers of labeled samples. However, existing few-shot methods ignore the correlation between cross-domain feature channels, and the feature representation ability is insufficient. To address above issue, this paper proposes a novel Residual Channel Attention Based Sample Adaptation Few-Shot Learning for Hyperspectral Image Classification(RCASA-FSL) for hyperspectral image classification (HSIC), which can capture and enhance cross-domain dependencies through multi-layer residual connection and random-based feature recalibration. Specifically, a Deep Residual Feature Channel Attention Mechanism (DRFCAM) is designed to obtain cross-domain dependencies by residual concatenation, and further the residual structure is stacked for mining depth discrimination information. Furthermore, a new Random-based Feature Recalibration Module (RFRM) is proposed to reassign the feature weights via random matrix, which fully explore feature weight relationships to guide the sample adaptation process. Besides, we design a joint loss function with combining the FSL loss and domain adaptive loss for further optimization model. Experiments conducted on several standard hyperspectral datasets demonstrate that the proposed RCASA-FSL is superior to other FSL techniques in both quantitative and qualitative aspects.https://doi.org/10.1038/s41598-024-77747-2 |
| spellingShingle | Yuefeng Zhao Jingqi Sun Nannan Hu Chengmin Zai Yanwei Han Residual channel attention based sample adaptation few-shot learning for hyperspectral image classification Scientific Reports |
| title | Residual channel attention based sample adaptation few-shot learning for hyperspectral image classification |
| title_full | Residual channel attention based sample adaptation few-shot learning for hyperspectral image classification |
| title_fullStr | Residual channel attention based sample adaptation few-shot learning for hyperspectral image classification |
| title_full_unstemmed | Residual channel attention based sample adaptation few-shot learning for hyperspectral image classification |
| title_short | Residual channel attention based sample adaptation few-shot learning for hyperspectral image classification |
| title_sort | residual channel attention based sample adaptation few shot learning for hyperspectral image classification |
| url | https://doi.org/10.1038/s41598-024-77747-2 |
| work_keys_str_mv | AT yuefengzhao residualchannelattentionbasedsampleadaptationfewshotlearningforhyperspectralimageclassification AT jingqisun residualchannelattentionbasedsampleadaptationfewshotlearningforhyperspectralimageclassification AT nannanhu residualchannelattentionbasedsampleadaptationfewshotlearningforhyperspectralimageclassification AT chengminzai residualchannelattentionbasedsampleadaptationfewshotlearningforhyperspectralimageclassification AT yanweihan residualchannelattentionbasedsampleadaptationfewshotlearningforhyperspectralimageclassification |