Feature Attention Parallel Aggregation Network for Single Image Haze Removal
Images captured in hazy weather often suffer from color distortion and texture blur due to turbid media suspended in the atmosphere. In this paper, we propose a Feature Attention Parallel Aggregation Network (FAPANet) to restore a clear image directly from the corresponding hazy input. It adopts the...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9698206/ |
| _version_ | 1818323716815192064 |
|---|---|
| author | Cuili Li Yufeng He Xu Li |
| author_facet | Cuili Li Yufeng He Xu Li |
| author_sort | Cuili Li |
| collection | DOAJ |
| description | Images captured in hazy weather often suffer from color distortion and texture blur due to turbid media suspended in the atmosphere. In this paper, we propose a Feature Attention Parallel Aggregation Network (FAPANet) to restore a clear image directly from the corresponding hazy input. It adopts the encoder-decoder structure while incorporating residual learning and attention mechanism. FAPANet consists of two key modules: a novel feature attention aggregation module (FAAM) and an adaptive feature fusion module (AFFM). FAAM recalibrates features by integrating channel attention and pixel attention in parallel to stimulate useful information and suppress redundant features. The shallow and deep layers of neural networks tend to characterize the low-level and high-level semantic features of images, respectively, so we introduce AFFM to fuse these two features adaptively. Meanwhile, a joint loss function, composed of L1 loss, perceptual loss, and structural similarity (SSIM) loss, is employed in the training stage for better results with more vivid colors and richer details. Comprehensive experiments on both synthetic and real-world images demonstrate the impressive performance of the proposed approach. |
| first_indexed | 2024-12-13T11:17:07Z |
| format | Article |
| id | doaj.art-055fb02de2b44ff3b135989a373063f7 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-13T11:17:07Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-055fb02de2b44ff3b135989a373063f72022-12-21T23:48:35ZengIEEEIEEE Access2169-35362022-01-0110153221533510.1109/ACCESS.2022.31481679698206Feature Attention Parallel Aggregation Network for Single Image Haze RemovalCuili Li0https://orcid.org/0000-0001-5836-3216Yufeng He1https://orcid.org/0000-0002-3325-7756Xu Li2https://orcid.org/0000-0002-8154-1188College of Information Engineering, Tarim University, Aral, ChinaCollege of Information Engineering, Tarim University, Aral, ChinaCollege of Information Engineering, Tarim University, Aral, ChinaImages captured in hazy weather often suffer from color distortion and texture blur due to turbid media suspended in the atmosphere. In this paper, we propose a Feature Attention Parallel Aggregation Network (FAPANet) to restore a clear image directly from the corresponding hazy input. It adopts the encoder-decoder structure while incorporating residual learning and attention mechanism. FAPANet consists of two key modules: a novel feature attention aggregation module (FAAM) and an adaptive feature fusion module (AFFM). FAAM recalibrates features by integrating channel attention and pixel attention in parallel to stimulate useful information and suppress redundant features. The shallow and deep layers of neural networks tend to characterize the low-level and high-level semantic features of images, respectively, so we introduce AFFM to fuse these two features adaptively. Meanwhile, a joint loss function, composed of L1 loss, perceptual loss, and structural similarity (SSIM) loss, is employed in the training stage for better results with more vivid colors and richer details. Comprehensive experiments on both synthetic and real-world images demonstrate the impressive performance of the proposed approach.https://ieeexplore.ieee.org/document/9698206/Attention mechanismdeep learningfeature fusionimage restorationsingle image dehazing |
| spellingShingle | Cuili Li Yufeng He Xu Li Feature Attention Parallel Aggregation Network for Single Image Haze Removal IEEE Access Attention mechanism deep learning feature fusion image restoration single image dehazing |
| title | Feature Attention Parallel Aggregation Network for Single Image Haze Removal |
| title_full | Feature Attention Parallel Aggregation Network for Single Image Haze Removal |
| title_fullStr | Feature Attention Parallel Aggregation Network for Single Image Haze Removal |
| title_full_unstemmed | Feature Attention Parallel Aggregation Network for Single Image Haze Removal |
| title_short | Feature Attention Parallel Aggregation Network for Single Image Haze Removal |
| title_sort | feature attention parallel aggregation network for single image haze removal |
| topic | Attention mechanism deep learning feature fusion image restoration single image dehazing |
| url | https://ieeexplore.ieee.org/document/9698206/ |
| work_keys_str_mv | AT cuilili featureattentionparallelaggregationnetworkforsingleimagehazeremoval AT yufenghe featureattentionparallelaggregationnetworkforsingleimagehazeremoval AT xuli featureattentionparallelaggregationnetworkforsingleimagehazeremoval |