Real‐world image deblurring using data synthesis and feature complementary network
Abstract Many learning‐based approaches to image deblurring have received increasing attention in recent years. However, the models trained on existing synthetic datasets do not generalize well to real‐world blur, resulting in undesirable artifacts and residual blur. This work attempts to address th...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2024-04-01
|
| Series: | IET Image Processing |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/ipr2.13029 |
| _version_ | 1797216887867703296 |
|---|---|
| author | Hao Wei Chenyang Ge Xin Qiao Pengchao Deng |
| author_facet | Hao Wei Chenyang Ge Xin Qiao Pengchao Deng |
| author_sort | Hao Wei |
| collection | DOAJ |
| description | Abstract Many learning‐based approaches to image deblurring have received increasing attention in recent years. However, the models trained on existing synthetic datasets do not generalize well to real‐world blur, resulting in undesirable artifacts and residual blur. This work attempts to address this problem from two aspects: training data synthesis and network architecture. To narrow the domain gap between synthetic and real domains, a realistic blur synthesis pipeline to generate high‐quality blurred data is proposed. Since the blur is non‐uniform and has different scales and degrees, a parallel feature complementary module to fully exploit the local and non‐local information, which improves the feature representation and helps the network to perceive the non‐uniform blur, is developed. In addition, a spatial Fourier reconstruction block to facilitate correct detail recovery in the spatial and Fourier domains is introduced. Based on these two designs, an effective encoder–decoder network for deblurring is designed. Extensive experiments demonstrate the validity and superiority of the proposed blur synthesis method and deblurring network. In particular, the proposed deblurring network can achieve superior or comparable performance to Restormer, while saving 70% of network parameters and 53% of floating point operations (FLOPs). |
| first_indexed | 2024-04-24T11:53:06Z |
| format | Article |
| id | doaj.art-3f0d861697424288967050fe4b5cb104 |
| institution | Directory Open Access Journal |
| issn | 1751-9659 1751-9667 |
| language | English |
| last_indexed | 2024-04-24T11:53:06Z |
| publishDate | 2024-04-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Image Processing |
| spelling | doaj.art-3f0d861697424288967050fe4b5cb1042024-04-09T06:07:10ZengWileyIET Image Processing1751-96591751-96672024-04-011851328134310.1049/ipr2.13029Real‐world image deblurring using data synthesis and feature complementary networkHao Wei0Chenyang Ge1Xin Qiao2Pengchao Deng3National Key Laboratory of Human‐Machine Hybrid Augmented Intelligence, National Engineering Research Center of Visual Information and Applications, Institute of Artificial Intelligence and Robotics Xi'an Jiaotong University Xi'an ChinaNational Key Laboratory of Human‐Machine Hybrid Augmented Intelligence, National Engineering Research Center of Visual Information and Applications, Institute of Artificial Intelligence and Robotics Xi'an Jiaotong University Xi'an ChinaNational Key Laboratory of Human‐Machine Hybrid Augmented Intelligence, National Engineering Research Center of Visual Information and Applications, Institute of Artificial Intelligence and Robotics Xi'an Jiaotong University Xi'an ChinaNational Key Laboratory of Human‐Machine Hybrid Augmented Intelligence, National Engineering Research Center of Visual Information and Applications, Institute of Artificial Intelligence and Robotics Xi'an Jiaotong University Xi'an ChinaAbstract Many learning‐based approaches to image deblurring have received increasing attention in recent years. However, the models trained on existing synthetic datasets do not generalize well to real‐world blur, resulting in undesirable artifacts and residual blur. This work attempts to address this problem from two aspects: training data synthesis and network architecture. To narrow the domain gap between synthetic and real domains, a realistic blur synthesis pipeline to generate high‐quality blurred data is proposed. Since the blur is non‐uniform and has different scales and degrees, a parallel feature complementary module to fully exploit the local and non‐local information, which improves the feature representation and helps the network to perceive the non‐uniform blur, is developed. In addition, a spatial Fourier reconstruction block to facilitate correct detail recovery in the spatial and Fourier domains is introduced. Based on these two designs, an effective encoder–decoder network for deblurring is designed. Extensive experiments demonstrate the validity and superiority of the proposed blur synthesis method and deblurring network. In particular, the proposed deblurring network can achieve superior or comparable performance to Restormer, while saving 70% of network parameters and 53% of floating point operations (FLOPs).https://doi.org/10.1049/ipr2.13029blurred data synthesiscomputer visiondeburring |
| spellingShingle | Hao Wei Chenyang Ge Xin Qiao Pengchao Deng Real‐world image deblurring using data synthesis and feature complementary network IET Image Processing blurred data synthesis computer vision deburring |
| title | Real‐world image deblurring using data synthesis and feature complementary network |
| title_full | Real‐world image deblurring using data synthesis and feature complementary network |
| title_fullStr | Real‐world image deblurring using data synthesis and feature complementary network |
| title_full_unstemmed | Real‐world image deblurring using data synthesis and feature complementary network |
| title_short | Real‐world image deblurring using data synthesis and feature complementary network |
| title_sort | real world image deblurring using data synthesis and feature complementary network |
| topic | blurred data synthesis computer vision deburring |
| url | https://doi.org/10.1049/ipr2.13029 |
| work_keys_str_mv | AT haowei realworldimagedeblurringusingdatasynthesisandfeaturecomplementarynetwork AT chenyangge realworldimagedeblurringusingdatasynthesisandfeaturecomplementarynetwork AT xinqiao realworldimagedeblurringusingdatasynthesisandfeaturecomplementarynetwork AT pengchaodeng realworldimagedeblurringusingdatasynthesisandfeaturecomplementarynetwork |