Face deblurring based on regularized structure and enhanced texture information
Abstract Image deblurring is an essential problem in computer vision. Due to highly structured and special facial components (e.g. eyes), most general image deblurring methods and face deblurring methods failed to yield more explicit structure and facial details, resulting in too smooth, uncoordinat...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2023-10-01
|
| Series: | Complex & Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s40747-023-01234-w |
| _version_ | 1797233167788146688 |
|---|---|
| author | Canghong Shi Xian Zhang Xiaojie Li Imran Mumtaz Jiancheng Lv |
| author_facet | Canghong Shi Xian Zhang Xiaojie Li Imran Mumtaz Jiancheng Lv |
| author_sort | Canghong Shi |
| collection | DOAJ |
| description | Abstract Image deblurring is an essential problem in computer vision. Due to highly structured and special facial components (e.g. eyes), most general image deblurring methods and face deblurring methods failed to yield more explicit structure and facial details, resulting in too smooth, uncoordinated and distorted face structure. Considering the unique face texture and sufficient facial details, we present an effective face deblurring network by exploiting more regularized structure and enhanced texture information (RSETNet). We first incorporate the face parsing network with fine-tuning to obtain more accurate face structure, and we present the feature adaptive denormalization (FAD) to regularize the facial structure as a condition of auxiliary to generate more harmonious and undistorted face structure. Meanwhile, to improve the generated facial texture information, we propose a new Laplace depth-wise separable convolution (LDConv) and multi-patch discriminator. Compared with existing methods, our face deblurring method could restore face structure more accurately and with more facial details. Experiments on two public face datasets have demonstrated the effectiveness of our proposed methods in terms of qualitative and quantitative indicators. |
| first_indexed | 2024-04-24T16:11:52Z |
| format | Article |
| id | doaj.art-2e2a0e4383b546409e581e3f13f15079 |
| institution | Directory Open Access Journal |
| issn | 2199-4536 2198-6053 |
| language | English |
| last_indexed | 2024-04-24T16:11:52Z |
| publishDate | 2023-10-01 |
| publisher | Springer |
| record_format | Article |
| series | Complex & Intelligent Systems |
| spelling | doaj.art-2e2a0e4383b546409e581e3f13f150792024-03-31T11:39:25ZengSpringerComplex & Intelligent Systems2199-45362198-60532023-10-011021769178610.1007/s40747-023-01234-wFace deblurring based on regularized structure and enhanced texture informationCanghong Shi0Xian Zhang1Xiaojie Li2Imran Mumtaz3Jiancheng Lv4Xihua UniversitySchool of Computer Science, Chengdu University of Information TechnologySchool of Computer Science, Chengdu University of Information TechnologyUniversity of AgricultureSichuan UniversityAbstract Image deblurring is an essential problem in computer vision. Due to highly structured and special facial components (e.g. eyes), most general image deblurring methods and face deblurring methods failed to yield more explicit structure and facial details, resulting in too smooth, uncoordinated and distorted face structure. Considering the unique face texture and sufficient facial details, we present an effective face deblurring network by exploiting more regularized structure and enhanced texture information (RSETNet). We first incorporate the face parsing network with fine-tuning to obtain more accurate face structure, and we present the feature adaptive denormalization (FAD) to regularize the facial structure as a condition of auxiliary to generate more harmonious and undistorted face structure. Meanwhile, to improve the generated facial texture information, we propose a new Laplace depth-wise separable convolution (LDConv) and multi-patch discriminator. Compared with existing methods, our face deblurring method could restore face structure more accurately and with more facial details. Experiments on two public face datasets have demonstrated the effectiveness of our proposed methods in terms of qualitative and quantitative indicators.https://doi.org/10.1007/s40747-023-01234-wFace deblurringLaplace depth-wise separable convolutionFacial adaptive denormalizationEnhanced texture information |
| spellingShingle | Canghong Shi Xian Zhang Xiaojie Li Imran Mumtaz Jiancheng Lv Face deblurring based on regularized structure and enhanced texture information Complex & Intelligent Systems Face deblurring Laplace depth-wise separable convolution Facial adaptive denormalization Enhanced texture information |
| title | Face deblurring based on regularized structure and enhanced texture information |
| title_full | Face deblurring based on regularized structure and enhanced texture information |
| title_fullStr | Face deblurring based on regularized structure and enhanced texture information |
| title_full_unstemmed | Face deblurring based on regularized structure and enhanced texture information |
| title_short | Face deblurring based on regularized structure and enhanced texture information |
| title_sort | face deblurring based on regularized structure and enhanced texture information |
| topic | Face deblurring Laplace depth-wise separable convolution Facial adaptive denormalization Enhanced texture information |
| url | https://doi.org/10.1007/s40747-023-01234-w |
| work_keys_str_mv | AT canghongshi facedeblurringbasedonregularizedstructureandenhancedtextureinformation AT xianzhang facedeblurringbasedonregularizedstructureandenhancedtextureinformation AT xiaojieli facedeblurringbasedonregularizedstructureandenhancedtextureinformation AT imranmumtaz facedeblurringbasedonregularizedstructureandenhancedtextureinformation AT jianchenglv facedeblurringbasedonregularizedstructureandenhancedtextureinformation |