From coarse to fine: Two-stage deep residual attention generative adversarial network for repair of iris textures obscured by eyelids and eyelashes
Summary: We propose a two-stage deep residual attention generative adversarial network (TSDRA-GAN) for inpainting iris textures obscured by eyelids. This two-stage generation approach ensures that the semantic and texture information of the generated images is preserved. In the second stage of the f...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-07-01
|
| Series: | iScience |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2589004223012464 |
| _version_ | 1797774563768008704 |
|---|---|
| author | Ying Chen Yugang Zeng Liang Xu Shubin Guo Ali Asghar Heidari Huiling Chen Yudong Zhang |
| author_facet | Ying Chen Yugang Zeng Liang Xu Shubin Guo Ali Asghar Heidari Huiling Chen Yudong Zhang |
| author_sort | Ying Chen |
| collection | DOAJ |
| description | Summary: We propose a two-stage deep residual attention generative adversarial network (TSDRA-GAN) for inpainting iris textures obscured by eyelids. This two-stage generation approach ensures that the semantic and texture information of the generated images is preserved. In the second stage of the fine network, a modified residual block (MRB) is used to further extract features and mitigate the performance degradation caused by the deepening of the network, thus following the concept of using a residual structure as a component of the encoder. In addition, for the skip connection part of this phase, we propose a dual-attention computing connection (DACC) to computationally fuse the features of the encoder and decoder in both directions to achieve more effective information fusion for iris inpainting tasks. Under completely fair and equal experimental conditions, it is shown that the method presented in this paper can effectively restore original iris images and improve recognition accuracy. |
| first_indexed | 2024-03-12T22:22:00Z |
| format | Article |
| id | doaj.art-a271b260ef3049209fee2fc5dbb537ce |
| institution | Directory Open Access Journal |
| issn | 2589-0042 |
| language | English |
| last_indexed | 2024-03-12T22:22:00Z |
| publishDate | 2023-07-01 |
| publisher | Elsevier |
| record_format | Article |
| series | iScience |
| spelling | doaj.art-a271b260ef3049209fee2fc5dbb537ce2023-07-23T04:55:48ZengElsevieriScience2589-00422023-07-01267107169From coarse to fine: Two-stage deep residual attention generative adversarial network for repair of iris textures obscured by eyelids and eyelashesYing Chen0Yugang Zeng1Liang Xu2Shubin Guo3Ali Asghar Heidari4Huiling Chen5Yudong Zhang6School of Software, Nanchang Hangkong University, Nanchang, Jiangxi 330063, ChinaSchool of Software, Nanchang Hangkong University, Nanchang, Jiangxi 330063, China; Corresponding authorSchool of Software, Nanchang Hangkong University, Nanchang, Jiangxi 330063, ChinaSchool of Software, Nanchang Hangkong University, Nanchang, Jiangxi 330063, ChinaSchool of Surveying and Geospatial Engineering, College of Engineering, University of Tehran, Tehran, IranKey Laboratory of Intelligent Informatics for Safety & Emergency of Zhejiang Province, Wenzhou University, Wenzhou 325035, PR China; Corresponding authorSchool of Computing and Mathematical Sciences, University of Leicester, LE1 7RH Leicester, UK; Corresponding authorSummary: We propose a two-stage deep residual attention generative adversarial network (TSDRA-GAN) for inpainting iris textures obscured by eyelids. This two-stage generation approach ensures that the semantic and texture information of the generated images is preserved. In the second stage of the fine network, a modified residual block (MRB) is used to further extract features and mitigate the performance degradation caused by the deepening of the network, thus following the concept of using a residual structure as a component of the encoder. In addition, for the skip connection part of this phase, we propose a dual-attention computing connection (DACC) to computationally fuse the features of the encoder and decoder in both directions to achieve more effective information fusion for iris inpainting tasks. Under completely fair and equal experimental conditions, it is shown that the method presented in this paper can effectively restore original iris images and improve recognition accuracy.http://www.sciencedirect.com/science/article/pii/S2589004223012464Health sciencesMedicineOptometry |
| spellingShingle | Ying Chen Yugang Zeng Liang Xu Shubin Guo Ali Asghar Heidari Huiling Chen Yudong Zhang From coarse to fine: Two-stage deep residual attention generative adversarial network for repair of iris textures obscured by eyelids and eyelashes iScience Health sciences Medicine Optometry |
| title | From coarse to fine: Two-stage deep residual attention generative adversarial network for repair of iris textures obscured by eyelids and eyelashes |
| title_full | From coarse to fine: Two-stage deep residual attention generative adversarial network for repair of iris textures obscured by eyelids and eyelashes |
| title_fullStr | From coarse to fine: Two-stage deep residual attention generative adversarial network for repair of iris textures obscured by eyelids and eyelashes |
| title_full_unstemmed | From coarse to fine: Two-stage deep residual attention generative adversarial network for repair of iris textures obscured by eyelids and eyelashes |
| title_short | From coarse to fine: Two-stage deep residual attention generative adversarial network for repair of iris textures obscured by eyelids and eyelashes |
| title_sort | from coarse to fine two stage deep residual attention generative adversarial network for repair of iris textures obscured by eyelids and eyelashes |
| topic | Health sciences Medicine Optometry |
| url | http://www.sciencedirect.com/science/article/pii/S2589004223012464 |
| work_keys_str_mv | AT yingchen fromcoarsetofinetwostagedeepresidualattentiongenerativeadversarialnetworkforrepairofiristexturesobscuredbyeyelidsandeyelashes AT yugangzeng fromcoarsetofinetwostagedeepresidualattentiongenerativeadversarialnetworkforrepairofiristexturesobscuredbyeyelidsandeyelashes AT liangxu fromcoarsetofinetwostagedeepresidualattentiongenerativeadversarialnetworkforrepairofiristexturesobscuredbyeyelidsandeyelashes AT shubinguo fromcoarsetofinetwostagedeepresidualattentiongenerativeadversarialnetworkforrepairofiristexturesobscuredbyeyelidsandeyelashes AT aliasgharheidari fromcoarsetofinetwostagedeepresidualattentiongenerativeadversarialnetworkforrepairofiristexturesobscuredbyeyelidsandeyelashes AT huilingchen fromcoarsetofinetwostagedeepresidualattentiongenerativeadversarialnetworkforrepairofiristexturesobscuredbyeyelidsandeyelashes AT yudongzhang fromcoarsetofinetwostagedeepresidualattentiongenerativeadversarialnetworkforrepairofiristexturesobscuredbyeyelidsandeyelashes |