Physical-model guided self-distillation network for single image dehazing
MotivationImage dehazing, as a key prerequisite of high-level computer vision tasks, has gained extensive attention in recent years. Traditional model-based methods acquire dehazed images via the atmospheric scattering model, which dehazed favorably but often causes artifacts due to the error of par...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-12-01
|
| Series: | Frontiers in Neurorobotics |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fnbot.2022.1036465/full |
| _version_ | 1798014387803389952 |
|---|---|
| author | Yunwei Lan Zhigao Cui Yanzhao Su Nian Wang Aihua Li Deshuai Han |
| author_facet | Yunwei Lan Zhigao Cui Yanzhao Su Nian Wang Aihua Li Deshuai Han |
| author_sort | Yunwei Lan |
| collection | DOAJ |
| description | MotivationImage dehazing, as a key prerequisite of high-level computer vision tasks, has gained extensive attention in recent years. Traditional model-based methods acquire dehazed images via the atmospheric scattering model, which dehazed favorably but often causes artifacts due to the error of parameter estimation. By contrast, recent model-free methods directly restore dehazed images by building an end-to-end network, which achieves better color fidelity. To improve the dehazing effect, we combine the complementary merits of these two categories and propose a physical-model guided self-distillation network for single image dehazing named PMGSDN.Proposed methodFirst, we propose a novel attention guided feature extraction block (AGFEB) and build a deep feature extraction network by it. Second, we propose three early-exit branches and embed the dark channel prior information to the network to merge the merits of model-based methods and model-free methods, and then we adopt self-distillation to transfer the features from the deeper layers (perform as teacher) to shallow early-exit branches (perform as student) to improve the dehazing effect.ResultsFor I-HAZE and O-HAZE datasets, better than the other methods, the proposed method achieves the best values of PSNR and SSIM being 17.41dB, 0.813, 18.48dB, and 0.802. Moreover, for real-world images, the proposed method also obtains high quality dehazed results.ConclusionExperimental results on both synthetic and real-world images demonstrate that the proposed PMGSDN can effectively dehaze images, resulting in dehazed results with clear textures and good color fidelity. |
| first_indexed | 2024-04-11T15:17:45Z |
| format | Article |
| id | doaj.art-0b0ee0db2ab04c0fa20a4a262a1e59d3 |
| institution | Directory Open Access Journal |
| issn | 1662-5218 |
| language | English |
| last_indexed | 2024-04-11T15:17:45Z |
| publishDate | 2022-12-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neurorobotics |
| spelling | doaj.art-0b0ee0db2ab04c0fa20a4a262a1e59d32022-12-22T04:16:27ZengFrontiers Media S.A.Frontiers in Neurorobotics1662-52182022-12-011610.3389/fnbot.2022.10364651036465Physical-model guided self-distillation network for single image dehazingYunwei LanZhigao CuiYanzhao SuNian WangAihua LiDeshuai HanMotivationImage dehazing, as a key prerequisite of high-level computer vision tasks, has gained extensive attention in recent years. Traditional model-based methods acquire dehazed images via the atmospheric scattering model, which dehazed favorably but often causes artifacts due to the error of parameter estimation. By contrast, recent model-free methods directly restore dehazed images by building an end-to-end network, which achieves better color fidelity. To improve the dehazing effect, we combine the complementary merits of these two categories and propose a physical-model guided self-distillation network for single image dehazing named PMGSDN.Proposed methodFirst, we propose a novel attention guided feature extraction block (AGFEB) and build a deep feature extraction network by it. Second, we propose three early-exit branches and embed the dark channel prior information to the network to merge the merits of model-based methods and model-free methods, and then we adopt self-distillation to transfer the features from the deeper layers (perform as teacher) to shallow early-exit branches (perform as student) to improve the dehazing effect.ResultsFor I-HAZE and O-HAZE datasets, better than the other methods, the proposed method achieves the best values of PSNR and SSIM being 17.41dB, 0.813, 18.48dB, and 0.802. Moreover, for real-world images, the proposed method also obtains high quality dehazed results.ConclusionExperimental results on both synthetic and real-world images demonstrate that the proposed PMGSDN can effectively dehaze images, resulting in dehazed results with clear textures and good color fidelity.https://www.frontiersin.org/articles/10.3389/fnbot.2022.1036465/fullimage dehazingknowledge distillationattention mechanismdeep learningcomputer vision |
| spellingShingle | Yunwei Lan Zhigao Cui Yanzhao Su Nian Wang Aihua Li Deshuai Han Physical-model guided self-distillation network for single image dehazing Frontiers in Neurorobotics image dehazing knowledge distillation attention mechanism deep learning computer vision |
| title | Physical-model guided self-distillation network for single image dehazing |
| title_full | Physical-model guided self-distillation network for single image dehazing |
| title_fullStr | Physical-model guided self-distillation network for single image dehazing |
| title_full_unstemmed | Physical-model guided self-distillation network for single image dehazing |
| title_short | Physical-model guided self-distillation network for single image dehazing |
| title_sort | physical model guided self distillation network for single image dehazing |
| topic | image dehazing knowledge distillation attention mechanism deep learning computer vision |
| url | https://www.frontiersin.org/articles/10.3389/fnbot.2022.1036465/full |
| work_keys_str_mv | AT yunweilan physicalmodelguidedselfdistillationnetworkforsingleimagedehazing AT zhigaocui physicalmodelguidedselfdistillationnetworkforsingleimagedehazing AT yanzhaosu physicalmodelguidedselfdistillationnetworkforsingleimagedehazing AT nianwang physicalmodelguidedselfdistillationnetworkforsingleimagedehazing AT aihuali physicalmodelguidedselfdistillationnetworkforsingleimagedehazing AT deshuaihan physicalmodelguidedselfdistillationnetworkforsingleimagedehazing |