{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,10]],"date-time":"2026-02-10T17:00:01Z","timestamp":1770742801126,"version":"3.49.0"},"reference-count":38,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T00:00:00Z","timestamp":1669852800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Neurorobot."],"abstract":"<jats:sec><jats:title>Motivation<\/jats:title><jats:p>Image 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 <jats:italic>via<\/jats:italic> 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.<\/jats:p><\/jats:sec><jats:sec><jats:title>Proposed method<\/jats:title><jats:p>First, 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.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>For 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.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusion<\/jats:title><jats:p>Experimental 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.<\/jats:p><\/jats:sec>","DOI":"10.3389\/fnbot.2022.1036465","type":"journal-article","created":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T17:55:07Z","timestamp":1669917307000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":8,"title":["Physical-model guided self-distillation network for single image dehazing"],"prefix":"10.3389","volume":"16","author":[{"given":"Yunwei","family":"Lan","sequence":"first","affiliation":[]},{"given":"Zhigao","family":"Cui","sequence":"additional","affiliation":[]},{"given":"Yanzhao","family":"Su","sequence":"additional","affiliation":[]},{"given":"Nian","family":"Wang","sequence":"additional","affiliation":[]},{"given":"Aihua","family":"Li","sequence":"additional","affiliation":[]},{"given":"Deshuai","family":"Han","sequence":"additional","affiliation":[]}],"member":"1965","published-online":{"date-parts":[[2022,12,1]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"620","DOI":"10.1007\/978-3-030-01449-0_52","article-title":"\u201cI-HAZE: a dehazing benchmark with real hazy and haze-free indoor images,\u201d","author":"Ancuti","year":"2018","journal-title":"Advanced Concepts for Intelligent Vision Systems"},{"key":"B2","doi-asserted-by":"crossref","first-page":"867","DOI":"10.1109\/CVPRW.2018.00119","article-title":"\u201cO-HAZE: a dehazing benchmark with real hazy and haze-free outdoor images,\u201d","volume-title":"2018 IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)","author":"Ancuti","year":"2018"},{"key":"B3","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2016.185","article-title":"\u201cNon-local image dehazing,\u201d","volume-title":"2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Berman","year":"2016"},{"key":"B4","doi-asserted-by":"publisher","first-page":"5187","DOI":"10.1109\/TIP.2016.2598681","article-title":"DehazeNet: an end-to-end system for single image haze removal","volume":"25","author":"Cai","year":"2016","journal-title":"IEEE Trans. Image Process."},{"key":"B5","doi-asserted-by":"crossref","first-page":"1375","DOI":"10.1109\/WACV.2019.00151","article-title":"\u201cGated context aggregation network for image dehazing and deraining,\u201d","volume-title":"2019 IEEE Winter Conference on Applications of Computer Vision (WACV)","author":"Chen","year":"2019"},{"key":"B6","doi-asserted-by":"publisher","first-page":"4367","DOI":"10.1007\/s10489-020-02116-1","article-title":"Image super, resolution reconstruction based on feature map attention mechanism","volume":"51","author":"Chen","year":"","journal-title":"Appl. Intell."},{"key":"B7","doi-asserted-by":"publisher","first-page":"30839","DOI":"10.1007\/s11042-020-09969-1","article-title":"The face image super-resolution algorithm based on combined representation learning","volume":"80","author":"Chen","year":"","journal-title":"Multimed. Tools Appl."},{"key":"B8","doi-asserted-by":"publisher","DOI":"10.1007\/s12652-020-02778-2","article-title":"Research on image inpainting algorithm of improved total variation minimization method","author":"Chen","year":"","journal-title":"J. Ambient Intell. Humaniz. Comput"},{"key":"B9","doi-asserted-by":"crossref","first-page":"7176","DOI":"10.1109\/CVPR46437.2021.00710","article-title":"\u201cPSD: principled synthetic-to-real dehazing guided by physical priors,\u201d","volume-title":"2021 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Chen","year":"2021"},{"key":"B10","doi-asserted-by":"crossref","first-page":"2154","DOI":"10.1109\/CVPR42600.2020.00223","article-title":"\u201cMulti-scale boosted dehazing network with dense feature fusion,\u201d","volume-title":"2020 IEEE\/CVF conference on computer vision and pattern recognition (CVPR)","author":"Dong","year":"2020"},{"key":"B11","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/2651362","article-title":"Dehazing using color-lines","volume":"34","author":"Fattal","year":"2014","journal-title":"ACM Trans. Graph."},{"key":"B12","doi-asserted-by":"crossref","first-page":"203","DOI":"10.1109\/CVPRW53098.2021.00029","article-title":"\u201cDW-GAN-A discrete wavelet transform GAN for nonhomogeneous dehazing,\u201d","volume-title":"2021 IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)","author":"Fu","year":"2021"},{"key":"B13","doi-asserted-by":"publisher","first-page":"2341","DOI":"10.1109\/TPAMI.2010.168","article-title":"Single image haze removal using dark channel prior","volume":"33","author":"He","year":"2011","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"B14","doi-asserted-by":"publisher","year":"2015","DOI":"10.48550\/arXiv.1503.02531"},{"key":"B15","doi-asserted-by":"crossref","first-page":"3459","DOI":"10.1109\/CVPR42600.2020.00352","article-title":"\u201cDistilling image dehazing with heterogeneous task imitation,\u201d","volume-title":"2020 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Hong","year":"2020"},{"key":"B16","doi-asserted-by":"crossref","first-page":"10659","DOI":"10.1109\/CVPR46437.2021.01052","article-title":"\u201cRefine myself by teaching myself: feature refinement via self-knowledge distillation,\u201d","volume-title":"2021 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Ji","year":"2021"},{"key":"B17","doi-asserted-by":"crossref","first-page":"4780","DOI":"10.1109\/ICCV.2017.511","article-title":"\u201cAOD-Net: all-in-one dehazing network,\u201d","volume-title":"2017 IEEE International Conference on Computer Vision (ICCV)","author":"Li","year":"2017"},{"key":"B18","doi-asserted-by":"publisher","first-page":"492","DOI":"10.1109\/TIP.2018.2867951","article-title":"Benchmarking single-image dehazing and beyond","volume":"28","author":"Li","year":"2019","journal-title":"IEEE Trans. Image Process."},{"key":"B19","doi-asserted-by":"crossref","first-page":"11720","DOI":"10.1109\/ICCV48922.2021.01153","article-title":"\u201cOnline knowledge distillation for efficient pose estimation,\u201d","volume-title":"2021 IEEE\/CVF International Conference on Computer Vision (ICCV)","author":"Li","year":"2021"},{"key":"B20","doi-asserted-by":"crossref","first-page":"1732","DOI":"10.1109\/CVPRW50498.2020.00223","article-title":"\u201cTrident dehazing network,\u201d","volume-title":"2020 IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)","author":"Liu","year":"2020"},{"key":"B21","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV.2019.00741","article-title":"\u201cGridDehazeNet: attention-based multi-scale network for image dehazing,\u201d","volume-title":"2019 IEEE\/CVF International Conference on Computer Vision (ICCV)","author":"Liu","year":"2019"},{"key":"B22","doi-asserted-by":"crossref","first-page":"2599","DOI":"10.1109\/CVPR.2019.00271","article-title":"\u201cStructured knowledge distillation for semantic segmentation,\u201d","volume-title":"2019 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Liu","year":"2019"},{"key":"B23","doi-asserted-by":"publisher","first-page":"4695","DOI":"10.1109\/TIP.2012.2214050","article-title":"No-reference image quality assessment in the spatial domain","volume":"21","author":"Mittal","year":"2012","journal-title":"IEEE Trans. Image Process."},{"key":"B24","doi-asserted-by":"publisher","first-page":"209","DOI":"10.1109\/LSP.2012.2227726","article-title":"Making a \u201ccompletely blind\u201d image quality analyzer","volume":"20","author":"Mittal","year":"2013","journal-title":"IEEE Signal Process. Lett."},{"key":"B25","doi-asserted-by":"publisher","first-page":"11908","DOI":"10.1609\/aaai.v34i07.6865","article-title":"FFA-Net: feature fusion attention network for single image dehazing","volume":"34","author":"Qin","year":"2020","journal-title":"Proc. AAAI Conf. Artif. Intell."},{"key":"B26","doi-asserted-by":"crossref","first-page":"8152","DOI":"10.1109\/CVPR.2019.00835","article-title":"\u201cEnhanced Pix2pix dehazing network,\u201d","volume-title":"2019 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Qu","year":"2019"},{"key":"B27","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-319-46475-6_10","article-title":"\u201cSingle image dehazing via multi-scale convolutional neural networks,\u201d","volume-title":"European Conference on Computer Vision","author":"Ren","year":"2016"},{"key":"B28","doi-asserted-by":"crossref","first-page":"2805","DOI":"10.1109\/CVPR42600.2020.00288","article-title":"\u201cDomain adaptation for image dehazing,\u201d","volume-title":"2020 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Shao","year":"2020"},{"key":"B29","doi-asserted-by":"publisher","first-page":"600","DOI":"10.1109\/TIP.2003.819861","article-title":"Image quality assessment: from error visibility to structural similarity","volume":"13","author":"Wang","year":"2004","journal-title":"IEEE Trans. Image Process."},{"key":"B30","doi-asserted-by":"publisher","first-page":"6008","DOI":"10.1016\/j.jksuci.2022.02.004","article-title":"Improved anti-occlusion object tracking algorithm using Unscented Rauch-Tung-Striebel smoother and kernel correlation filter","volume":"34","author":"Xia","year":"2022","journal-title":"J. King Saud Univ - Comput. Inf. Sci."},{"key":"B31","doi-asserted-by":"publisher","first-page":"3194","DOI":"10.1109\/CVPR.2018.00337","article-title":"\u201cDensely connected pyramid dehazing network,\u201d","author":"Zhang","year":"2018","journal-title":"2018 IEEE\/CVF Conference on Computer Vision and Pattern Recognition"},{"key":"B32","doi-asserted-by":"publisher","first-page":"72","DOI":"10.1109\/TIP.2019.2922837","article-title":"FAMED-Net: a fast and accurate multi-scale end-to-end dehazing network","volume":"29","author":"Zhang","year":"2020","journal-title":"IEEE Trans. Image Process."},{"key":"B33","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TPAMI.2021.3067100","article-title":"Self-distillation: towards, efficient, and compact neural networks","volume":"44","author":"Zhang","year":"2021","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"B34","doi-asserted-by":"crossref","first-page":"3205","DOI":"10.1109\/ICIP.2017.8296874","article-title":"\u201cHazeRD: an outdoor scene dataset and benchmark for single image dehazing,\u201d","volume-title":"2017 IEEE International Conference on Image Processing (ICIP)","author":"Zhang","year":"2017"},{"key":"B35","doi-asserted-by":"publisher","first-page":"4240","DOI":"10.1049\/iet-ipr.2020.0928","article-title":"Image dehazing method via a cycle generative adversarial network","volume":"14","author":"Zhao","year":"2020","journal-title":"IET Image Process."},{"key":"B36","doi-asserted-by":"publisher","first-page":"3391","DOI":"10.1109\/TIP.2021.3060873","article-title":"RefineDNet: a weakly supervised refinement framework for single image dehazing","volume":"30","author":"Zhao","year":"2021","journal-title":"IEEE Trans. Image Process."},{"key":"B37","doi-asserted-by":"crossref","first-page":"3451","DOI":"10.1109\/WACV51458.2022.00351","article-title":"\u201cSelf-guidance: improve deep neural network generalization via knowledge distillation,\u201d","volume-title":"2022 IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV)","author":"Zheng","year":"2022"},{"key":"B38","doi-asserted-by":"publisher","first-page":"3522","DOI":"10.1109\/TIP.2015.2446191","article-title":"Fast single image haze removal algorithm using color attenuation prior","volume":"24","author":"Zhu","year":"2015","journal-title":"IEEE Trans. Image Process."}],"container-title":["Frontiers in Neurorobotics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2022.1036465\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T17:55:18Z","timestamp":1669917318000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2022.1036465\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,12,1]]},"references-count":38,"alternative-id":["10.3389\/fnbot.2022.1036465"],"URL":"https:\/\/doi.org\/10.3389\/fnbot.2022.1036465","relation":{},"ISSN":["1662-5218"],"issn-type":[{"value":"1662-5218","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,12,1]]},"article-number":"1036465"}}