Regularised transfer learning for hyperspectral image classification
This study presents a transfer learning method for addressing the insufficient sample problem in hyperspectral image classification. In order to find common feature representation for both the source domain and target domain, we introduce a regularisation based on Bregman divergence into the objecti...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2019-03-01
|
| Series: | IET Computer Vision |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/iet-cvi.2018.5145 |
| _version_ | 1797684258899230720 |
|---|---|
| author | Qian Shi Yipeng Zhang Xiaoping Liu Kefei Zhao |
| author_facet | Qian Shi Yipeng Zhang Xiaoping Liu Kefei Zhao |
| author_sort | Qian Shi |
| collection | DOAJ |
| description | This study presents a transfer learning method for addressing the insufficient sample problem in hyperspectral image classification. In order to find common feature representation for both the source domain and target domain, we introduce a regularisation based on Bregman divergence into the objective function of the subspace learning algorithm, which can minimise the Bregman divergence between the distribution of training samples in the source domain and the test samples in the target domain. Hyperspectral image with biased sampling is used to evaluate the effectiveness of the proposed method. The results show that the proposed method can achieve a higher classification accuracy than traditional subspace learning methods under the condition of biased sampling. |
| first_indexed | 2024-03-12T00:28:02Z |
| format | Article |
| id | doaj.art-43048482486f475ab3c44d22c026df19 |
| institution | Directory Open Access Journal |
| issn | 1751-9632 1751-9640 |
| language | English |
| last_indexed | 2024-03-12T00:28:02Z |
| publishDate | 2019-03-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Computer Vision |
| spelling | doaj.art-43048482486f475ab3c44d22c026df192023-09-15T10:31:50ZengWileyIET Computer Vision1751-96321751-96402019-03-0113218819310.1049/iet-cvi.2018.5145Regularised transfer learning for hyperspectral image classificationQian Shi0Yipeng Zhang1Xiaoping Liu2Kefei Zhao3School of Geography and Planning, Sun Yat‐sen UniversityGuangzhouGuangdongPeople's Republic of ChinaSchool of Computer Engineering, Syracuse UniversitySyracuseNew YorkUSASchool of Geography and Planning, Sun Yat‐sen UniversityGuangzhouGuangdongPeople's Republic of ChinaSchool of Management, Guangdong University of TechnologyGuangzhouGuangdongPeople's Republic of ChinaThis study presents a transfer learning method for addressing the insufficient sample problem in hyperspectral image classification. In order to find common feature representation for both the source domain and target domain, we introduce a regularisation based on Bregman divergence into the objective function of the subspace learning algorithm, which can minimise the Bregman divergence between the distribution of training samples in the source domain and the test samples in the target domain. Hyperspectral image with biased sampling is used to evaluate the effectiveness of the proposed method. The results show that the proposed method can achieve a higher classification accuracy than traditional subspace learning methods under the condition of biased sampling.https://doi.org/10.1049/iet-cvi.2018.5145hyperspectral image classificationsource domaintarget domainregularisationBregman divergencesubspace learning algorithm |
| spellingShingle | Qian Shi Yipeng Zhang Xiaoping Liu Kefei Zhao Regularised transfer learning for hyperspectral image classification IET Computer Vision hyperspectral image classification source domain target domain regularisation Bregman divergence subspace learning algorithm |
| title | Regularised transfer learning for hyperspectral image classification |
| title_full | Regularised transfer learning for hyperspectral image classification |
| title_fullStr | Regularised transfer learning for hyperspectral image classification |
| title_full_unstemmed | Regularised transfer learning for hyperspectral image classification |
| title_short | Regularised transfer learning for hyperspectral image classification |
| title_sort | regularised transfer learning for hyperspectral image classification |
| topic | hyperspectral image classification source domain target domain regularisation Bregman divergence subspace learning algorithm |
| url | https://doi.org/10.1049/iet-cvi.2018.5145 |
| work_keys_str_mv | AT qianshi regularisedtransferlearningforhyperspectralimageclassification AT yipengzhang regularisedtransferlearningforhyperspectralimageclassification AT xiaopingliu regularisedtransferlearningforhyperspectralimageclassification AT kefeizhao regularisedtransferlearningforhyperspectralimageclassification |