A weakly supervised learning pipeline for profiled fibre inspection
Abstract Automatic profiled fibre recognition and analysis can accelerate quality inspection and contributes to the upgrade of the textile industry. However, these tasks often require significant manual effort to generate instance‐level annotations for fully supervised training. In this paper, the a...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2024-02-01
|
| Series: | IET Image Processing |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/ipr2.12984 |
| _version_ | 1827351208891777024 |
|---|---|
| author | Zhao Chen Yahui Xiu Yuxin Zheng Xinxin Wang Qian Wang Danqi Guo Yan Wan |
| author_facet | Zhao Chen Yahui Xiu Yuxin Zheng Xinxin Wang Qian Wang Danqi Guo Yan Wan |
| author_sort | Zhao Chen |
| collection | DOAJ |
| description | Abstract Automatic profiled fibre recognition and analysis can accelerate quality inspection and contributes to the upgrade of the textile industry. However, these tasks often require significant manual effort to generate instance‐level annotations for fully supervised training. In this paper, the authors propose a weakly supervised pipeline for profiled fibre inspection using electron‐microscopic (EM) images with only image‐level annotations. It automatically identifies fibre instances and estimates shape factors to facilitate fibre quality inspection. As the core of the pipeline, the weakly supervised network (WesNet) is designed to localize hundreds of crowded fibre samples by raw patch generation and fibre sample sifting. Particularly, the composite similarity measurement integrates different patch‐wise similarities, enabling the network to distinguish fibre from background robustly. For quality inspection, the pipeline further analyzes the fibre instances, utilizing several efficient techniques to estimate the shape factors. Experiments on the real fibre electron‐microscopic images demonstrate the efficacy and efficiency of the pipeline. Results show that WesNet outperforms several supervised and weakly supervised methods, including two state‐of‐the‐art weakly supervised networks. |
| first_indexed | 2024-03-08T01:57:50Z |
| format | Article |
| id | doaj.art-5fc6045432264760a097390f93ec3246 |
| institution | Directory Open Access Journal |
| issn | 1751-9659 1751-9667 |
| language | English |
| last_indexed | 2024-03-08T01:57:50Z |
| publishDate | 2024-02-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Image Processing |
| spelling | doaj.art-5fc6045432264760a097390f93ec32462024-02-14T07:53:24ZengWileyIET Image Processing1751-96591751-96672024-02-0118377278410.1049/ipr2.12984A weakly supervised learning pipeline for profiled fibre inspectionZhao Chen0Yahui Xiu1Yuxin Zheng2Xinxin Wang3Qian Wang4Danqi Guo5Yan Wan6School of Computer Science and Technology Donghua University Shanghai ChinaSchool of Computer Science and Technology Donghua University Shanghai ChinaSchool of Computer Science and Technology Donghua University Shanghai ChinaSchool of Computer Science and Technology Donghua University Shanghai ChinaSchool of Computer Science and Technology Donghua University Shanghai ChinaSchool of Computer Science and Technology Donghua University Shanghai ChinaSchool of Computer Science and Technology Donghua University Shanghai ChinaAbstract Automatic profiled fibre recognition and analysis can accelerate quality inspection and contributes to the upgrade of the textile industry. However, these tasks often require significant manual effort to generate instance‐level annotations for fully supervised training. In this paper, the authors propose a weakly supervised pipeline for profiled fibre inspection using electron‐microscopic (EM) images with only image‐level annotations. It automatically identifies fibre instances and estimates shape factors to facilitate fibre quality inspection. As the core of the pipeline, the weakly supervised network (WesNet) is designed to localize hundreds of crowded fibre samples by raw patch generation and fibre sample sifting. Particularly, the composite similarity measurement integrates different patch‐wise similarities, enabling the network to distinguish fibre from background robustly. For quality inspection, the pipeline further analyzes the fibre instances, utilizing several efficient techniques to estimate the shape factors. Experiments on the real fibre electron‐microscopic images demonstrate the efficacy and efficiency of the pipeline. Results show that WesNet outperforms several supervised and weakly supervised methods, including two state‐of‐the‐art weakly supervised networks.https://doi.org/10.1049/ipr2.12984composite similarity measurementprofiled fibre recognitionshape factor estimationweakly supervised learning |
| spellingShingle | Zhao Chen Yahui Xiu Yuxin Zheng Xinxin Wang Qian Wang Danqi Guo Yan Wan A weakly supervised learning pipeline for profiled fibre inspection IET Image Processing composite similarity measurement profiled fibre recognition shape factor estimation weakly supervised learning |
| title | A weakly supervised learning pipeline for profiled fibre inspection |
| title_full | A weakly supervised learning pipeline for profiled fibre inspection |
| title_fullStr | A weakly supervised learning pipeline for profiled fibre inspection |
| title_full_unstemmed | A weakly supervised learning pipeline for profiled fibre inspection |
| title_short | A weakly supervised learning pipeline for profiled fibre inspection |
| title_sort | weakly supervised learning pipeline for profiled fibre inspection |
| topic | composite similarity measurement profiled fibre recognition shape factor estimation weakly supervised learning |
| url | https://doi.org/10.1049/ipr2.12984 |
| work_keys_str_mv | AT zhaochen aweaklysupervisedlearningpipelineforprofiledfibreinspection AT yahuixiu aweaklysupervisedlearningpipelineforprofiledfibreinspection AT yuxinzheng aweaklysupervisedlearningpipelineforprofiledfibreinspection AT xinxinwang aweaklysupervisedlearningpipelineforprofiledfibreinspection AT qianwang aweaklysupervisedlearningpipelineforprofiledfibreinspection AT danqiguo aweaklysupervisedlearningpipelineforprofiledfibreinspection AT yanwan aweaklysupervisedlearningpipelineforprofiledfibreinspection AT zhaochen weaklysupervisedlearningpipelineforprofiledfibreinspection AT yahuixiu weaklysupervisedlearningpipelineforprofiledfibreinspection AT yuxinzheng weaklysupervisedlearningpipelineforprofiledfibreinspection AT xinxinwang weaklysupervisedlearningpipelineforprofiledfibreinspection AT qianwang weaklysupervisedlearningpipelineforprofiledfibreinspection AT danqiguo weaklysupervisedlearningpipelineforprofiledfibreinspection AT yanwan weaklysupervisedlearningpipelineforprofiledfibreinspection |