A New Method of Image Classification Based on Domain Adaptation
Deep neural networks can learn powerful representations from massive amounts of labeled data; however, their performance is unsatisfactory in the case of large samples and small labels. Transfer learning can bridge between a source domain with rich sample data and a target domain with only a few or...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/4/1315 |
| _version_ | 1797476743596998656 |
|---|---|
| author | Fangwen Zhao Weifeng Liu Chenglin Wen |
| author_facet | Fangwen Zhao Weifeng Liu Chenglin Wen |
| author_sort | Fangwen Zhao |
| collection | DOAJ |
| description | Deep neural networks can learn powerful representations from massive amounts of labeled data; however, their performance is unsatisfactory in the case of large samples and small labels. Transfer learning can bridge between a source domain with rich sample data and a target domain with only a few or zero labeled samples and, thus, complete the transfer of knowledge by aligning the distribution between domains through methods, such as domain adaptation. Previous domain adaptation methods mostly align the features in the feature space of all categories on a global scale. Recently, the method of locally aligning the sub-categories by introducing label information achieved better results. Based on this, we present a deep fuzzy domain adaptation (DFDA) that assigns different weights to samples of the same category in the source and target domains, which enhances the domain adaptive capabilities. Our experiments demonstrate that DFDA can achieve remarkable results on standard domain adaptation datasets. |
| first_indexed | 2024-03-09T21:07:00Z |
| format | Article |
| id | doaj.art-45013ec4dcda435eb3553cc16a05f69f |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T21:07:00Z |
| publishDate | 2022-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-45013ec4dcda435eb3553cc16a05f69f2023-11-23T21:57:37ZengMDPI AGSensors1424-82202022-02-01224131510.3390/s22041315A New Method of Image Classification Based on Domain AdaptationFangwen Zhao0Weifeng Liu1Chenglin Wen2School of Electrical and Control Engineering, Shaanxi University of Science and Technology, Xi’an 710021, ChinaSchool of Electrical and Control Engineering, Shaanxi University of Science and Technology, Xi’an 710021, ChinaSchool of Automation, Guangdong University of Petrochemical Technology, Maoming 525000, ChinaDeep neural networks can learn powerful representations from massive amounts of labeled data; however, their performance is unsatisfactory in the case of large samples and small labels. Transfer learning can bridge between a source domain with rich sample data and a target domain with only a few or zero labeled samples and, thus, complete the transfer of knowledge by aligning the distribution between domains through methods, such as domain adaptation. Previous domain adaptation methods mostly align the features in the feature space of all categories on a global scale. Recently, the method of locally aligning the sub-categories by introducing label information achieved better results. Based on this, we present a deep fuzzy domain adaptation (DFDA) that assigns different weights to samples of the same category in the source and target domains, which enhances the domain adaptive capabilities. Our experiments demonstrate that DFDA can achieve remarkable results on standard domain adaptation datasets.https://www.mdpi.com/1424-8220/22/4/1315domain adaptationunsupervised learningmaximum mean discrepancy |
| spellingShingle | Fangwen Zhao Weifeng Liu Chenglin Wen A New Method of Image Classification Based on Domain Adaptation Sensors domain adaptation unsupervised learning maximum mean discrepancy |
| title | A New Method of Image Classification Based on Domain Adaptation |
| title_full | A New Method of Image Classification Based on Domain Adaptation |
| title_fullStr | A New Method of Image Classification Based on Domain Adaptation |
| title_full_unstemmed | A New Method of Image Classification Based on Domain Adaptation |
| title_short | A New Method of Image Classification Based on Domain Adaptation |
| title_sort | new method of image classification based on domain adaptation |
| topic | domain adaptation unsupervised learning maximum mean discrepancy |
| url | https://www.mdpi.com/1424-8220/22/4/1315 |
| work_keys_str_mv | AT fangwenzhao anewmethodofimageclassificationbasedondomainadaptation AT weifengliu anewmethodofimageclassificationbasedondomainadaptation AT chenglinwen anewmethodofimageclassificationbasedondomainadaptation AT fangwenzhao newmethodofimageclassificationbasedondomainadaptation AT weifengliu newmethodofimageclassificationbasedondomainadaptation AT chenglinwen newmethodofimageclassificationbasedondomainadaptation |