Improved triplet loss for domain adaptation
Abstract A technique known as domain adaptation is utilised to address classification challenges in an unlabelled target domain by leveraging labelled source domains. Previous domain adaptation approaches have predominantly focussed on global domain adaptation, neglecting class‐level information and...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2024-02-01
|
| Series: | IET Computer Vision |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/cvi2.12226 |
| _version_ | 1797320638981996544 |
|---|---|
| author | Xiaoshun Wang Yunhan Li Xiangliang Zhang |
| author_facet | Xiaoshun Wang Yunhan Li Xiangliang Zhang |
| author_sort | Xiaoshun Wang |
| collection | DOAJ |
| description | Abstract A technique known as domain adaptation is utilised to address classification challenges in an unlabelled target domain by leveraging labelled source domains. Previous domain adaptation approaches have predominantly focussed on global domain adaptation, neglecting class‐level information and resulting in suboptimal transfer performance. In recent years, a considerable number of researchers have explored class‐level domain adaptation, aiming to precisely align the distribution of diverse domains. Nevertheless, existing research on class‐level alignment tends to align domain features either on or in proximity to classification boundaries, which introduces ambiguous samples that can impact classification accuracy. In this study, the authors propose a novel strategy called class guided constraints (CGC) to tackle this issue. Specifically, CGC is employed to preserve the compactness within classes and separability between classes of domain features prior to class‐level alignment. Furthermore, the authors incorporate CGC in conjunction with similarity guided constraint. Comprehensive evaluations conducted on four public datasets demonstrate that our approach outperforms numerous state‐of‐the‐art domain adaptation methods significantly and achieves greater improvements compared to the baseline approach. |
| first_indexed | 2024-03-08T04:45:55Z |
| format | Article |
| id | doaj.art-8863f6978e06425bb2edfcb9207a9352 |
| institution | Directory Open Access Journal |
| issn | 1751-9632 1751-9640 |
| language | English |
| last_indexed | 2024-03-08T04:45:55Z |
| publishDate | 2024-02-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Computer Vision |
| spelling | doaj.art-8863f6978e06425bb2edfcb9207a93522024-02-08T10:33:59ZengWileyIET Computer Vision1751-96321751-96402024-02-01181849610.1049/cvi2.12226Improved triplet loss for domain adaptationXiaoshun Wang0Yunhan Li1Xiangliang Zhang2Huzhou Key Laboratory of Green Energy Materials and Battery Cascade Utilization School of Intelligent Manufacturing Huzhou College Huzhou ChinaHuzhou Key Laboratory of Green Energy Materials and Battery Cascade Utilization School of Intelligent Manufacturing Huzhou College Huzhou ChinaHuzhou Key Laboratory of Green Energy Materials and Battery Cascade Utilization School of Intelligent Manufacturing Huzhou College Huzhou ChinaAbstract A technique known as domain adaptation is utilised to address classification challenges in an unlabelled target domain by leveraging labelled source domains. Previous domain adaptation approaches have predominantly focussed on global domain adaptation, neglecting class‐level information and resulting in suboptimal transfer performance. In recent years, a considerable number of researchers have explored class‐level domain adaptation, aiming to precisely align the distribution of diverse domains. Nevertheless, existing research on class‐level alignment tends to align domain features either on or in proximity to classification boundaries, which introduces ambiguous samples that can impact classification accuracy. In this study, the authors propose a novel strategy called class guided constraints (CGC) to tackle this issue. Specifically, CGC is employed to preserve the compactness within classes and separability between classes of domain features prior to class‐level alignment. Furthermore, the authors incorporate CGC in conjunction with similarity guided constraint. Comprehensive evaluations conducted on four public datasets demonstrate that our approach outperforms numerous state‐of‐the‐art domain adaptation methods significantly and achieves greater improvements compared to the baseline approach.https://doi.org/10.1049/cvi2.12226computer visionimage classification |
| spellingShingle | Xiaoshun Wang Yunhan Li Xiangliang Zhang Improved triplet loss for domain adaptation IET Computer Vision computer vision image classification |
| title | Improved triplet loss for domain adaptation |
| title_full | Improved triplet loss for domain adaptation |
| title_fullStr | Improved triplet loss for domain adaptation |
| title_full_unstemmed | Improved triplet loss for domain adaptation |
| title_short | Improved triplet loss for domain adaptation |
| title_sort | improved triplet loss for domain adaptation |
| topic | computer vision image classification |
| url | https://doi.org/10.1049/cvi2.12226 |
| work_keys_str_mv | AT xiaoshunwang improvedtripletlossfordomainadaptation AT yunhanli improvedtripletlossfordomainadaptation AT xiangliangzhang improvedtripletlossfordomainadaptation |