{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/cycle-dehaze-enhanced-cyclegan-for-single","title":"Cycle-Dehaze: Enhanced CycleGAN for Single Image Dehazing","arxiv_id":"1805.05308","date":"2018-05-14","proceeding":null,"authors":["Deniz Engin","Anıl Genç","Hazim Kemal Ekenel"],"abstract":"In this paper, we present an end-to-end network, called Cycle-Dehaze, for\nsingle image dehazing problem, which does not require pairs of hazy and\ncorresponding ground truth images for training. That is, we train the network\nby feeding clean and hazy images in an unpaired manner. Moreover, the proposed\napproach does not rely on estimation of the atmospheric scattering model\nparameters. Our method enhances CycleGAN formulation by combining\ncycle-consistency and perceptual losses in order to improve the quality of\ntextural information recovery and generate visually better haze-free images.\nTypically, deep learning models for dehazing take low resolution images as\ninput and produce low resolution outputs. However, in the NTIRE 2018 challenge\non single image dehazing, high resolution images were provided. Therefore, we\napply bicubic downscaling. After obtaining low-resolution outputs from the\nnetwork, we utilize the Laplacian pyramid to upscale the output images to the\noriginal resolution. We conduct experiments on NYU-Depth, I-HAZE, and O-HAZE\ndatasets. Extensive experiments demonstrate that the proposed approach improves\nCycleGAN method both quantitatively and qualitatively.","url_abs":"http://arxiv.org/abs/1805.05308v1","url_pdf":"http://arxiv.org/pdf/1805.05308v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"cycle-dehaze-enhanced-cyclegan-for-single","repo_url":"https://github.com/engindeniz/Cycle-Dehaze","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"cycle-dehaze-enhanced-cyclegan-for-single","repo_url":"https://github.com/niranjangavade98/CycleDehaze-Pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"cycle-dehaze-enhanced-cyclegan-for-single","repo_url":"https://github.com/code-implementation1/Code9/tree/main/CycleGAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"image-dehazing","task_name":"Image Dehazing"},{"task_slug":"single-image-dehazing","task_name":"Single Image Dehazing"}],"methods":[{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"cycle-consistency-loss","method_name":"Cycle Consistency Loss"},{"method_slug":"gan-least-squares-loss","method_name":"GAN Least Squares Loss"},{"method_slug":"instance-normalization","method_name":"Instance Normalization"},{"method_slug":"patchgan","method_name":"PatchGAN"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-block","method_name":"Residual Block"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-dehazing-on-o-haze","task":"Image Dehazing","dataset":"O-Haze","model":"Cycle-Dehaze","rank_in_archive_order":7,"of":7,"metrics":{"PSNR":"19.62","SSIM":"0.67"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1805.05308","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}