Efficient Face Region Occlusion Repair Based on T-GANs
In the image restoration task, the generative adversarial network (GAN) demonstrates excellent performance. However, there remain significant challenges concerning the task of generative face region inpainting. Traditional model approaches are ineffective in maintaining global consistency among faci...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/12/10/2162 |
| _version_ | 1797600296064516096 |
|---|---|
| author | Qiaoyue Man Young-Im Cho |
| author_facet | Qiaoyue Man Young-Im Cho |
| author_sort | Qiaoyue Man |
| collection | DOAJ |
| description | In the image restoration task, the generative adversarial network (GAN) demonstrates excellent performance. However, there remain significant challenges concerning the task of generative face region inpainting. Traditional model approaches are ineffective in maintaining global consistency among facial components and recovering fine facial details. To address this challenge, this study proposes a facial restoration generation network combined a transformer module and GAN to accurately detect the missing feature parts of the face and perform effective and fine-grained restoration generation. We validate the proposed model using different image quality evaluation methods and several open-source face datasets and experimentally demonstrate that our model outperforms other current state-of-the-art network models in terms of generated image quality and the coherent naturalness of facial features in face image restoration generation tasks. |
| first_indexed | 2024-03-11T03:47:27Z |
| format | Article |
| id | doaj.art-350533a8dc274c9b8be720c76210c526 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-11T03:47:27Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-350533a8dc274c9b8be720c76210c5262023-11-18T01:08:28ZengMDPI AGElectronics2079-92922023-05-011210216210.3390/electronics12102162Efficient Face Region Occlusion Repair Based on T-GANsQiaoyue Man0Young-Im Cho1Department of Computer Engineering, Gachon University, Seongnam-si 13120, Republic of KoreaDepartment of Computer Engineering, Gachon University, Seongnam-si 13120, Republic of KoreaIn the image restoration task, the generative adversarial network (GAN) demonstrates excellent performance. However, there remain significant challenges concerning the task of generative face region inpainting. Traditional model approaches are ineffective in maintaining global consistency among facial components and recovering fine facial details. To address this challenge, this study proposes a facial restoration generation network combined a transformer module and GAN to accurately detect the missing feature parts of the face and perform effective and fine-grained restoration generation. We validate the proposed model using different image quality evaluation methods and several open-source face datasets and experimentally demonstrate that our model outperforms other current state-of-the-art network models in terms of generated image quality and the coherent naturalness of facial features in face image restoration generation tasks.https://www.mdpi.com/2079-9292/12/10/2162face detectionconvolutional neural networkgenerative adversarial networksimage fusion |
| spellingShingle | Qiaoyue Man Young-Im Cho Efficient Face Region Occlusion Repair Based on T-GANs Electronics face detection convolutional neural network generative adversarial networks image fusion |
| title | Efficient Face Region Occlusion Repair Based on T-GANs |
| title_full | Efficient Face Region Occlusion Repair Based on T-GANs |
| title_fullStr | Efficient Face Region Occlusion Repair Based on T-GANs |
| title_full_unstemmed | Efficient Face Region Occlusion Repair Based on T-GANs |
| title_short | Efficient Face Region Occlusion Repair Based on T-GANs |
| title_sort | efficient face region occlusion repair based on t gans |
| topic | face detection convolutional neural network generative adversarial networks image fusion |
| url | https://www.mdpi.com/2079-9292/12/10/2162 |
| work_keys_str_mv | AT qiaoyueman efficientfaceregionocclusionrepairbasedontgans AT youngimcho efficientfaceregionocclusionrepairbasedontgans |