Lightweight and efficient dual-path fusion network for iris segmentation
Abstract In order to tackle limitations of current iris segmentation methods based on deep learning, such as an enormous amount of parameters, intensive computation and excessive storage space, a lightweight and efficient iris segmentation network is proposed in this article. Based on the classical...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-08-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-39743-w |
| _version_ | 1797452853554446336 |
|---|---|
| author | Songze Lei Aokui Shan Bo Liu Yanxiao Zhao Wei Xiang |
| author_facet | Songze Lei Aokui Shan Bo Liu Yanxiao Zhao Wei Xiang |
| author_sort | Songze Lei |
| collection | DOAJ |
| description | Abstract In order to tackle limitations of current iris segmentation methods based on deep learning, such as an enormous amount of parameters, intensive computation and excessive storage space, a lightweight and efficient iris segmentation network is proposed in this article. Based on the classical semantic segmentation network U-net, the proposed approach designs a dual-path fusion network model to integrate deep semantic information and rich shallow context information at multiple levels. Our model uses the depth-wise separable convolution for feature extraction and introduces a novel attention mechanism, which strengthens the capability of extracting significant features as well as the segmentation capability of the network. Experiments on four public datasets reveal that the proposed approach can raise the MIoU and F1 scores by 15% and 9% on average compared with traditional methods, respectively, and 1.5% and 2.5% on average compared with the classical semantic segmentation method U-net and other relevant methods. Compared with the U-net, the proposed approach reduces about 80%, 90% and 99% in terms of computation, parameters and storage, respectively, and the average run time up to 0.02 s. Our approach not only exhibits a good performance, but also is simpler in terms of computation, parameters and storage compared with existing classical semantic segmentation methods. |
| first_indexed | 2024-03-09T15:13:43Z |
| format | Article |
| id | doaj.art-76bac48902b649dcbf1cd3652935cb7c |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-03-09T15:13:43Z |
| publishDate | 2023-08-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-76bac48902b649dcbf1cd3652935cb7c2023-11-26T13:11:43ZengNature PortfolioScientific Reports2045-23222023-08-0113111310.1038/s41598-023-39743-wLightweight and efficient dual-path fusion network for iris segmentationSongze Lei0Aokui Shan1Bo Liu2Yanxiao Zhao3Wei Xiang4School of Computer Science and Engineering, Xi’an Technological UniversitySchool of Computer Science and Engineering, Xi’an Technological UniversitySchool of Computer Science and Engineering, Xi’an Technological UniversitySchool of Information Communication Engineering, Beijing Information Science and Technology UniversitySchool of Computing Engineering and Mathematical Sciences, La Trobe UniversityAbstract In order to tackle limitations of current iris segmentation methods based on deep learning, such as an enormous amount of parameters, intensive computation and excessive storage space, a lightweight and efficient iris segmentation network is proposed in this article. Based on the classical semantic segmentation network U-net, the proposed approach designs a dual-path fusion network model to integrate deep semantic information and rich shallow context information at multiple levels. Our model uses the depth-wise separable convolution for feature extraction and introduces a novel attention mechanism, which strengthens the capability of extracting significant features as well as the segmentation capability of the network. Experiments on four public datasets reveal that the proposed approach can raise the MIoU and F1 scores by 15% and 9% on average compared with traditional methods, respectively, and 1.5% and 2.5% on average compared with the classical semantic segmentation method U-net and other relevant methods. Compared with the U-net, the proposed approach reduces about 80%, 90% and 99% in terms of computation, parameters and storage, respectively, and the average run time up to 0.02 s. Our approach not only exhibits a good performance, but also is simpler in terms of computation, parameters and storage compared with existing classical semantic segmentation methods.https://doi.org/10.1038/s41598-023-39743-w |
| spellingShingle | Songze Lei Aokui Shan Bo Liu Yanxiao Zhao Wei Xiang Lightweight and efficient dual-path fusion network for iris segmentation Scientific Reports |
| title | Lightweight and efficient dual-path fusion network for iris segmentation |
| title_full | Lightweight and efficient dual-path fusion network for iris segmentation |
| title_fullStr | Lightweight and efficient dual-path fusion network for iris segmentation |
| title_full_unstemmed | Lightweight and efficient dual-path fusion network for iris segmentation |
| title_short | Lightweight and efficient dual-path fusion network for iris segmentation |
| title_sort | lightweight and efficient dual path fusion network for iris segmentation |
| url | https://doi.org/10.1038/s41598-023-39743-w |
| work_keys_str_mv | AT songzelei lightweightandefficientdualpathfusionnetworkforirissegmentation AT aokuishan lightweightandefficientdualpathfusionnetworkforirissegmentation AT boliu lightweightandefficientdualpathfusionnetworkforirissegmentation AT yanxiaozhao lightweightandefficientdualpathfusionnetworkforirissegmentation AT weixiang lightweightandefficientdualpathfusionnetworkforirissegmentation |