{"url":"/sota/image-dehazing-on-sots-outdoor","task":{"name":"Image Dehazing","url":"/task/image-dehazing","note":null},"dataset":{"name":"SOTS Outdoor","url":"/dataset/reside"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"<span style=\"color:grey; opacity: 0.6\">( Image credit: [Densely Connected Pyramid Dehazing Network](https://github.com/hezhangsprinter/DCPDN) )</span>","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["PSNR","SSIM"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"PSNR":"higher","SSIM":"higher"}},"counts":{"rows":31,"rows_with_code":29,"rows_with_paper_page":31,"rows_dated":31,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"ChaIR","metrics":{"PSNR":"40.73","SSIM":"0.997"},"uses_additional_data":false,"paper_date":"2023-12-20","paper":"/paper/exploring-the-potential-of-channel","paper_url":"https://www.sciencedirect.com/science/article/abs/pii/S0950705123009061","paper_title":"Exploring the potential of channel interactions for image restoration","code":"https://github.com/c-yn/ChaIR","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"FSNet","metrics":{"PSNR":"40.40","SSIM":"0.997"},"uses_additional_data":false,"paper_date":"2023-11-06","paper":"/paper/image-restoration-via-frequency-selection","paper_url":"https://ieeexplore.ieee.org/document/10310164","paper_title":"Image Restoration via Frequency Selection","code":"https://github.com/c-yn/FSNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"SFNet","metrics":{"PSNR":"40.05","SSIM":"0.996"},"uses_additional_data":false,"paper_date":"2023-04-13","paper":"/paper/selective-frequency-network-for-image","paper_url":"https://openreview.net/forum?id=tyZ1ChGZIKO","paper_title":"Selective Frequency Network for Image Restoration","code":"https://github.com/c-yn/SFNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"Instruct-IPT","metrics":{"PSNR":"39.95","SSIM":"0.992"},"uses_additional_data":false,"paper_date":"2024-06-30","paper":"/paper/instruct-ipt-all-in-one-image-processing-1","paper_url":"https://arxiv.org/abs/2407.00676v2","paper_title":"Instruct-IPT: All-in-One Image Processing Transformer via Weight Modulation","code":"https://github.com/huawei-noah/Pretrained-IPT","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"ConvIR","metrics":{"PSNR":"39.42","SSIM":"0.996"},"uses_additional_data":false,"paper_date":"2024-06-25","paper":"/paper/revitalizing-convolutional-network-for-image","paper_url":"https://ieeexplore.ieee.org/abstract/document/10571568","paper_title":"Revitalizing Convolutional Network for Image Restoration","code":"https://github.com/c-yn/ConvIR","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"IRNeXt","metrics":{"PSNR":"39.18","SSIM":"0.996"},"uses_additional_data":false,"paper_date":"2023-04-24","paper":"/paper/irnext-rethinking-convolutional-network","paper_url":"https://openreview.net/forum?id=MZkbgahv4a","paper_title":"IRNeXt: Rethinking Convolutional Network Design for Image Restoration","code":"https://github.com/c-yn/IRNeXt","n_code_links":1,"syntology":null},{"rank_in_archive_order":7,"model":"CasDyF-Net","metrics":{"PSNR":"38.86","SSIM":"0.995"},"uses_additional_data":false,"paper_date":"2024-09-13","paper":"/paper/casdyf-net-image-dehazing-via-cascaded","paper_url":"https://arxiv.org/abs/2409.08510v1","paper_title":"CasDyF-Net: Image Dehazing via Cascaded Dynamic Filters","code":"https://github.com/dauing/casdyf-net","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"DSANet","metrics":{"PSNR":"38.39","SSIM":"0.995"},"uses_additional_data":false,"paper_date":"2024-03-01","paper":"/paper/dual-domain-strip-attention-for-image","paper_url":"https://www.sciencedirect.com/science/article/pii/S0893608023006974","paper_title":"Dual-domain strip attention for image restoration","code":"https://github.com/c-yn/DSANet","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"SANet","metrics":{"PSNR":"38.01","SSIM":"0.995"},"uses_additional_data":false,"paper_date":"2023-08-01","paper":"/paper/strip-attention-for-image-restoration","paper_url":"https://www.ijcai.org/proceedings/2023/72","paper_title":"Strip Attention for Image Restoration","code":"https://github.com/c-yn/SANet","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"SAD-Net","metrics":{"PSNR":"37.78","SSIM":"0.990"},"uses_additional_data":false,"paper_date":"2025-04-07","paper":"/paper/sad-net-a-full-spectral-self-attention-detail","paper_url":"https://www.nature.com/articles/s41598-025-92061-1","paper_title":"SAD-Net: a full spectral self-attention detail enhancement network for single image dehazing","code":"https://github.com/niuqj/SAD-Net","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"FocalNet","metrics":{"PSNR":"37.71","SSIM":"0.995"},"uses_additional_data":false,"paper_date":"2023-01-01","paper":"/paper/focal-network-for-image-restoration","paper_url":"http://openaccess.thecvf.com//content/ICCV2023/html/Cui_Focal_Network_for_Image_Restoration_ICCV_2023_paper.html","paper_title":"Focal Network for Image Restoration","code":"https://github.com/c-yn/focalnet","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"OKNet","metrics":{"PSNR":"37.68","SSIM":"0.995"},"uses_additional_data":false,"paper_date":"2024-03-24","paper":"/paper/omni-kernel-network-for-image-restoration","paper_url":"https://ojs.aaai.org/index.php/AAAI/article/view/27907","paper_title":"Omni-Kernel Network for Image Restoration","code":"https://github.com/c-yn/OKNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"MaIR","metrics":{"PSNR":"36.96","SSIM":"0.991"},"uses_additional_data":false,"paper_date":"2024-12-28","paper":"/paper/mair-a-locality-and-continuity-preserving","paper_url":"https://arxiv.org/abs/2412.20066v1","paper_title":"MaIR: A Locality- and Continuity-Preserving Mamba for Image Restoration","code":"https://github.com/XLearning-SCU/2025-CVPR-MaIR","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"C2PNet","metrics":{"PSNR":"36.68","SSIM":"0.99"},"uses_additional_data":false,"paper_date":"2023-03-24","paper":"/paper/curricular-contrastive-regularization-for","paper_url":"https://arxiv.org/abs/2303.14218v2","paper_title":"Curricular Contrastive Regularization for Physics-aware Single Image Dehazing","code":"https://github.com/yuzheng9/c2pnet","n_code_links":1,"syntology":{"n_ran":5,"n_unverified":0,"n_samples":5,"n_pointer_only_licence":5}},{"rank_in_archive_order":15,"model":"gUNet-D","metrics":{"PSNR":"36.64","SSIM":"0.986"},"uses_additional_data":false,"paper_date":"2022-09-23","paper":"/paper/rethinking-performance-gains-in-image","paper_url":"https://arxiv.org/abs/2209.11448v1","paper_title":"Rethinking Performance Gains in Image Dehazing Networks","code":"https://github.com/idkiro/gunet","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"DEA-Net-CR","metrics":{"PSNR":"36.59","SSIM":"0.9897"},"uses_additional_data":false,"paper_date":"2023-01-12","paper":"/paper/dea-net-single-image-dehazing-based-on-detail","paper_url":"https://arxiv.org/abs/2301.04805v1","paper_title":"DEA-Net: Single image dehazing based on detail-enhanced convolution and content-guided attention","code":"https://github.com/cecret3350/DEA-Net","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"MixDehazeNet","metrics":{"PSNR":"36.50","SSIM":"0.986"},"uses_additional_data":false,"paper_date":"2023-05-28","paper":"/paper/mixdehazenet-mix-structure-block-for-image","paper_url":"https://arxiv.org/abs/2305.17654v1","paper_title":"MixDehazeNet : Mix Structure Block For Image Dehazing Network","code":"https://github.com/ameryxiong/mixdehazenet","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"OneRestore","metrics":{"PSNR":"35.58","SSIM":"0.9814"},"uses_additional_data":false,"paper_date":"2024-07-05","paper":"/paper/onerestore-a-universal-restoration-framework","paper_url":"https://arxiv.org/abs/2407.04621v4","paper_title":"OneRestore: A Universal Restoration Framework for Composite Degradation","code":"https://github.com/gy65896/onerestore","n_code_links":1,"syntology":{"n_ran":7,"n_unverified":6,"n_samples":13,"n_pointer_only_licence":13}},{"rank_in_archive_order":19,"model":"DehazeFormer-B","metrics":{"PSNR":"34.95","SSIM":"0.984"},"uses_additional_data":false,"paper_date":"2022-04-08","paper":"/paper/vision-transformers-for-single-image-dehazing","paper_url":"https://arxiv.org/abs/2204.03883v1","paper_title":"Vision Transformers for Single Image Dehazing","code":"https://github.com/IDKiro/DehazeFormer","n_code_links":1,"syntology":{"n_ran":5,"n_unverified":5,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":20,"model":"MAXIM-2S","metrics":{"PSNR":"34.19"},"uses_additional_data":false,"paper_date":"2022-01-09","paper":"/paper/maxim-multi-axis-mlp-for-image-processing","paper_url":"https://arxiv.org/abs/2201.02973v2","paper_title":"MAXIM: Multi-Axis MLP for Image Processing","code":"https://github.com/google-research/maxim","n_code_links":3,"syntology":{"n_ran":31,"n_unverified":15,"n_samples":46,"n_pointer_only_licence":0}},{"rank_in_archive_order":21,"model":"FFA-Net","metrics":{"PSNR":"33.57","SSIM":"0.9804"},"uses_additional_data":false,"paper_date":"2019-11-18","paper":"/paper/ffa-net-feature-fusion-attention-network-for","paper_url":"https://arxiv.org/abs/1911.07559v2","paper_title":"FFA-Net: Feature Fusion Attention Network for Single Image Dehazing","code":"https://github.com/zhilin007/FFA-Net","n_code_links":4,"syntology":{"n_ran":0,"n_unverified":11,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":22,"model":"U2-Former","metrics":{"PSNR":"31.10","SSIM":"0.976"},"uses_additional_data":false,"paper_date":"2021-12-04","paper":"/paper/u2-former-a-nested-u-shaped-transformer-for","paper_url":"https://arxiv.org/abs/2112.02279v2","paper_title":"U2-Former: A Nested U-shaped Transformer for Image Restoration","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":23,"model":"GridDehazeNet","metrics":{"PSNR":"30.86","SSIM":"0.982"},"uses_additional_data":false,"paper_date":"2019-08-08","paper":"/paper/griddehazenet-attention-based-multi-scale","paper_url":"https://arxiv.org/abs/1908.03245v1","paper_title":"GridDehazeNet: Attention-Based Multi-Scale Network for Image Dehazing","code":"https://github.com/proteus1991/GridDehazeNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"GMAN","metrics":{"PSNR":"28.19","SSIM":"0.9638"},"uses_additional_data":false,"paper_date":"2018-10-05","paper":"/paper/generic-model-agnostic-convolutional-neural","paper_url":"https://arxiv.org/abs/1810.02862v2","paper_title":"Generic Model-Agnostic Convolutional Neural Network for Single Image Dehazing","code":"https://github.com/Seanforfun/GMAN_Net_Haze_Removal","n_code_links":3,"syntology":null},{"rank_in_archive_order":25,"model":"Uformer","metrics":{"PSNR":"26.52","SSIM":"0.945"},"uses_additional_data":false,"paper_date":"2021-06-06","paper":"/paper/uformer-a-general-u-shaped-transformer-for","paper_url":"https://arxiv.org/abs/2106.03106v2","paper_title":"Uformer: A General U-Shaped Transformer for Image Restoration","code":"https://github.com/ZhendongWang6/Uformer","n_code_links":4,"syntology":{"n_ran":6,"n_unverified":0,"n_samples":6,"n_pointer_only_licence":2}},{"rank_in_archive_order":26,"model":"EMRA-Net","metrics":{"PSNR":"25.81","SSIM":"0.9409"},"uses_additional_data":false,"paper_date":"2021-06-23","paper":"/paper/an-ensemble-multi-scale-residual-attention","paper_url":"https://link.springer.com/article/10.1007/s11042-021-11081-x","paper_title":"An ensemble multi-scale residual attention network (EMRA-net) for image Dehazing","code":"https://github.com/Maverick-3/EMRA-Net","n_code_links":1,"syntology":null},{"rank_in_archive_order":27,"model":"AOD-Net","metrics":{"PSNR":"24.14","SSIM":"0.920"},"uses_additional_data":false,"paper_date":"2017-10-01","paper":"/paper/aod-net-all-in-one-dehazing-network","paper_url":"http://openaccess.thecvf.com/content_iccv_2017/html/Li_AOD-Net_All-In-One_Dehazing_ICCV_2017_paper.html","paper_title":"AOD-Net: All-In-One Dehazing Network","code":"https://github.com/kritiksoman/GIMP-ML","n_code_links":1,"syntology":null},{"rank_in_archive_order":28,"model":"Deep DCP","metrics":{"PSNR":"24.08","SSIM":"0.933"},"uses_additional_data":false,"paper_date":"2018-12-06","paper":"/paper/unsupervised-single-image-dehazing-using-dark","paper_url":"https://arxiv.org/abs/1812.07051v2","paper_title":"Unsupervised Single Image Dehazing Using Dark Channel Prior Loss","code":"https://github.com/AlonaGolts/Deep_Energy","n_code_links":1,"syntology":null},{"rank_in_archive_order":29,"model":"Deep Energy (Network)","metrics":{"PSNR":"24.07","SSIM":"0.933"},"uses_additional_data":false,"paper_date":"2018-05-31","paper":"/paper/deep-energy-using-energy-functions-for","paper_url":"https://arxiv.org/abs/1805.12355v2","paper_title":"Deep-Energy: Unsupervised Training of Deep Neural Networks","code":"https://github.com/AlonaGolts/Deep_Energy","n_code_links":1,"syntology":null},{"rank_in_archive_order":30,"model":"EPDN","metrics":{"PSNR":"22.57","SSIM":"0.8630"},"uses_additional_data":false,"paper_date":"2019-06-01","paper":"/paper/enhanced-pix2pix-dehazing-network","paper_url":"http://openaccess.thecvf.com/content_CVPR_2019/html/Qu_Enhanced_Pix2pix_Dehazing_Network_CVPR_2019_paper.html","paper_title":"Enhanced Pix2pix Dehazing Network","code":"https://github.com/ErinChen1/EPDN","n_code_links":1,"syntology":null},{"rank_in_archive_order":31,"model":"GFN","metrics":{"PSNR":"22.30","SSIM":"0.880"},"uses_additional_data":false,"paper_date":"2018-03-31","paper":"/paper/gated-fusion-network-for-single-image","paper_url":"http://arxiv.org/abs/1804.00213v1","paper_title":"Gated Fusion Network for Single Image Dehazing","code":null,"n_code_links":0,"syntology":null}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 7,081 of the 9,623 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9623,"papers_checked":7081,"papers_extracted_not_yet_verified":217,"boards_without_verdict":27,"papers_not_yet_extracted":2325},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-25T09:33:49+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":6,"rows_with_any_sample_ran":5,"distinct_papers_with_graph_line":6,"distinct_papers_with_any_sample_ran":5,"samples_over_distinct_papers":{"n_ran":54,"n_unverified":37,"n_samples":91,"n_pointer_only_licence":20,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":54,"n_unverified":37,"n_samples":91,"n_pointer_only_licence":20,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}