{"url":"/sota/image-super-resolution-on-set5-3x-upscaling","task":{"name":"Image Super-Resolution","url":"/task/image-super-resolution","note":null},"dataset":{"name":"Set5 - 3x upscaling","url":"/dataset/set5"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"**Image Super-Resolution** is a machine learning task where the goal is to increase the resolution of an image, often by a factor of 4x or more, while maintaining its content and details as much as possible. The end result is a high-resolution version of the original image. This task can be used for various applications such as improving image quality, enhancing visual detail, and increasing the accuracy of computer vision algorithms.","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":32,"rows_with_code":27,"rows_with_paper_page":32,"rows_dated":32,"rows_using_additional_data":2},"rows":[{"rank_in_archive_order":1,"model":"HMA†","metrics":{"PSNR":"35.35","SSIM":"0.9347"},"uses_additional_data":true,"paper_date":"2024-05-08","paper":"/paper/hmanet-hybrid-multi-axis-aggregation-network","paper_url":"https://arxiv.org/abs/2405.05001v1","paper_title":"HMANet: Hybrid Multi-Axis Aggregation Network for Image Super-Resolution","code":"https://github.com/korouuuuu/hma","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"DRCT-L","metrics":{"PSNR":"35.32","SSIM":"0.9348"},"uses_additional_data":false,"paper_date":"2024-03-31","paper":"/paper/drct-saving-image-super-resolution-away-from","paper_url":"https://arxiv.org/abs/2404.00722v5","paper_title":"DRCT: Saving Image Super-resolution away from Information Bottleneck","code":"https://github.com/ming053l/drct","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":0,"n_samples":4,"n_pointer_only_licence":3}},{"rank_in_archive_order":3,"model":"HAT-L","metrics":{"PSNR":"35.28","SSIM":"0.9345"},"uses_additional_data":false,"paper_date":"2022-05-09","paper":"/paper/activating-more-pixels-in-image-super","paper_url":"https://arxiv.org/abs/2205.04437v3","paper_title":"Activating More Pixels in Image Super-Resolution Transformer","code":"https://github.com/xpixelgroup/hat","n_code_links":2,"syntology":null},{"rank_in_archive_order":4,"model":"HAT_FIR","metrics":{"PSNR":"35.21"},"uses_additional_data":false,"paper_date":"2022-08-24","paper":"/paper/swinfir-revisiting-the-swinir-with-fast","paper_url":"https://arxiv.org/abs/2208.11247v3","paper_title":"SwinFIR: Revisiting the SwinIR with Fast Fourier Convolution and Improved Training for Image Super-Resolution","code":"https://github.com/Zdafeng/SwinFIR","n_code_links":2,"syntology":null},{"rank_in_archive_order":5,"model":"Hi-IR-L","metrics":{"PSNR":"35.2","SSIM":"0.938"},"uses_additional_data":false,"paper_date":"2024-11-27","paper":"/paper/hierarchical-information-flow-for-generalized","paper_url":"https://arxiv.org/abs/2411.18588v1","paper_title":"Hierarchical Information Flow for Generalized Efficient Image Restoration","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"CPAT+","metrics":{"PSNR":"35.19","SSIM":"0.9335"},"uses_additional_data":false,"paper_date":"2024-07-23","paper":"/paper/channel-partitioned-windowed-attention-and","paper_url":"https://arxiv.org/abs/2407.16232v2","paper_title":"Channel-Partitioned Windowed Attention And Frequency Learning for Single Image Super-Resolution","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":7,"model":"DRCT","metrics":{"PSNR":"35.18","SSIM":"0.9338"},"uses_additional_data":false,"paper_date":"2024-03-31","paper":"/paper/drct-saving-image-super-resolution-away-from","paper_url":"https://arxiv.org/abs/2404.00722v5","paper_title":"DRCT: Saving Image Super-resolution away from Information Bottleneck","code":"https://github.com/ming053l/drct","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":0,"n_samples":4,"n_pointer_only_licence":3}},{"rank_in_archive_order":8,"model":"HAT","metrics":{"PSNR":"35.16","SSIM":"0.9335"},"uses_additional_data":false,"paper_date":"2022-05-09","paper":"/paper/activating-more-pixels-in-image-super","paper_url":"https://arxiv.org/abs/2205.04437v3","paper_title":"Activating More Pixels in Image Super-Resolution Transformer","code":"https://github.com/xpixelgroup/hat","n_code_links":2,"syntology":null},{"rank_in_archive_order":9,"model":"CPAT","metrics":{"PSNR":"35.16","SSIM":"0.9334"},"uses_additional_data":false,"paper_date":"2024-07-23","paper":"/paper/channel-partitioned-windowed-attention-and","paper_url":"https://arxiv.org/abs/2407.16232v2","paper_title":"Channel-Partitioned Windowed Attention And Frequency Learning for Single Image Super-Resolution","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":10,"model":"SwinFIR","metrics":{"PSNR":"35.15","SSIM":"0.9330"},"uses_additional_data":false,"paper_date":"2022-08-24","paper":"/paper/swinfir-revisiting-the-swinir-with-fast","paper_url":"https://arxiv.org/abs/2208.11247v3","paper_title":"SwinFIR: Revisiting the SwinIR with Fast Fourier Convolution and Improved Training for Image Super-Resolution","code":"https://github.com/Zdafeng/SwinFIR","n_code_links":2,"syntology":null},{"rank_in_archive_order":11,"model":"EDT-B","metrics":{"PSNR":"35.13","SSIM":"0.9328"},"uses_additional_data":true,"paper_date":"2021-12-19","paper":"/paper/on-efficient-transformer-and-image-pre","paper_url":"https://arxiv.org/abs/2112.10175v2","paper_title":"On Efficient Transformer-Based Image Pre-training for Low-Level Vision","code":"https://github.com/fenglinglwb/edt","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"LTE","metrics":{"PSNR":"34.89"},"uses_additional_data":false,"paper_date":"2021-11-17","paper":"/paper/local-texture-estimator-for-implicit","paper_url":"https://arxiv.org/abs/2111.08918v6","paper_title":"Local Texture Estimator for Implicit Representation Function","code":"https://github.com/jaewon-lee-b/lte","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":3,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":13,"model":"DRLN+","metrics":{"PSNR":"34.86","SSIM":"0.9307"},"uses_additional_data":false,"paper_date":"2019-06-28","paper":"/paper/densely-residual-laplacian-super-resolution","paper_url":"https://arxiv.org/abs/1906.12021v2","paper_title":"Densely Residual Laplacian Super-Resolution","code":"https://github.com/saeed-anwar/DRLN","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":1,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":14,"model":"HAN+","metrics":{"PSNR":"34.85","SSIM":"0.9300"},"uses_additional_data":false,"paper_date":"2020-08-20","paper":"/paper/single-image-super-resolution-via-a-holistic","paper_url":"https://arxiv.org/abs/2008.08767v1","paper_title":"Single Image Super-Resolution via a Holistic Attention Network","code":"https://github.com/wwlCape/HAN","n_code_links":2,"syntology":null},{"rank_in_archive_order":15,"model":"CSNLN","metrics":{"PSNR":"34.74","SSIM":"0.9300"},"uses_additional_data":false,"paper_date":"2020-06-02","paper":"/paper/image-super-resolution-with-cross-scale-non","paper_url":"https://arxiv.org/abs/2006.01424v1","paper_title":"Image Super-Resolution with Cross-Scale Non-Local Attention and Exhaustive Self-Exemplars Mining","code":"https://github.com/SHI-Labs/Cross-Scale-Non-Local-Attention","n_code_links":3,"syntology":null},{"rank_in_archive_order":16,"model":"FACD","metrics":{"PSNR":"34.729"},"uses_additional_data":false,"paper_date":"2022-11-29","paper":"/paper/feature-based-adaptive-contrastive","paper_url":"https://arxiv.org/abs/2211.15951v2","paper_title":"Feature-domain Adaptive Contrastive Distillation for Efficient Single Image Super-Resolution","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":17,"model":"ML-CrAIST","metrics":{"PSNR":"34.7","SSIM":"0.9302"},"uses_additional_data":false,"paper_date":"2024-08-19","paper":"/paper/ml-craist-multi-scale-low-high-frequency","paper_url":"https://arxiv.org/abs/2408.09940v1","paper_title":"ML-CrAIST: Multi-scale Low-high Frequency Information-based Cross black Attention with Image Super-resolving Transformer","code":"https://github.com/alik033/ml-craist","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"SRFBN","metrics":{"PSNR":"34.70"},"uses_additional_data":false,"paper_date":"2019-03-23","paper":"/paper/feedback-network-for-image-super-resolution","paper_url":"https://arxiv.org/abs/1903.09814v2","paper_title":"Feedback Network for Image Super-Resolution","code":"https://github.com/Paper99/SRFBN_CVPR19","n_code_links":4,"syntology":{"n_ran":2,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":19,"model":"SwinOIR","metrics":{"PSNR":"34.69","SSIM":"0.9296"},"uses_additional_data":false,"paper_date":"2023-03-16","paper":"/paper/a-framework-for-real-time-object-detection","paper_url":"https://arxiv.org/abs/2303.09190v2","paper_title":"Resolution Enhancement Processing on Low Quality Images Using Swin Transformer Based on Interval Dense Connection Strategy","code":"https://github.com/rubbbbbbbbby/swinoir","n_code_links":2,"syntology":null},{"rank_in_archive_order":20,"model":"PMRN+","metrics":{"PSNR":"34.65","SSIM":"0.9289"},"uses_additional_data":false,"paper_date":"2020-07-19","paper":"/paper/progressive-multi-scale-residual-network-for","paper_url":"https://arxiv.org/abs/2007.09552v4","paper_title":"Sequential Hierarchical Learning with Distribution Transformation for Image Super-Resolution","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":21,"model":"ML-CrAIST-Li","metrics":{"PSNR":"34.58","SSIM":"0.9294"},"uses_additional_data":false,"paper_date":"2024-08-19","paper":"/paper/ml-craist-multi-scale-low-high-frequency","paper_url":"https://arxiv.org/abs/2408.09940v1","paper_title":"ML-CrAIST: Multi-scale Low-high Frequency Information-based Cross black Attention with Image Super-resolving Transformer","code":"https://github.com/alik033/ml-craist","n_code_links":1,"syntology":null},{"rank_in_archive_order":22,"model":"IMDN","metrics":{"PSNR":"34.36"},"uses_additional_data":false,"paper_date":"2019-09-26","paper":"/paper/lightweight-image-super-resolution-with-1","paper_url":"https://arxiv.org/abs/1909.11856v1","paper_title":"Lightweight Image Super-Resolution with Information Multi-distillation Network","code":"https://github.com/Zheng222/IMDN","n_code_links":4,"syntology":{"n_ran":2,"n_unverified":1,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":23,"model":"AdaFM-Net","metrics":{"PSNR":"34.34"},"uses_additional_data":false,"paper_date":"2019-04-17","paper":"/paper/modulating-image-restoration-with-continual","paper_url":"https://arxiv.org/abs/1904.08118v3","paper_title":"Modulating Image Restoration with Continual Levels via Adaptive Feature Modification Layers","code":"https://github.com/hejingwenhejingwen/AdaFM","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"MWCNN","metrics":{"PSNR":"34.17"},"uses_additional_data":false,"paper_date":"2018-05-18","paper":"/paper/multi-level-wavelet-cnn-for-image-restoration","paper_url":"http://arxiv.org/abs/1805.07071v2","paper_title":"Multi-level Wavelet-CNN for Image Restoration","code":"https://github.com/lpj0/MWCNN","n_code_links":5,"syntology":null},{"rank_in_archive_order":25,"model":"LFFN-S","metrics":{"PSNR":"34.04","SSIM":"0.9233"},"uses_additional_data":false,"paper_date":"2019-02-15","paper":"/paper/lightweight-feature-fusion-network-for-single","paper_url":"http://arxiv.org/abs/1902.05694v2","paper_title":"Lightweight Feature Fusion Network for Single Image Super-Resolution","code":"https://github.com/qibao77/LFFN-master","n_code_links":2,"syntology":null},{"rank_in_archive_order":26,"model":"LCSCNet","metrics":{"PSNR":"33.99"},"uses_additional_data":false,"paper_date":"2019-09-09","paper":"/paper/lcscnet-linear-compressing-based-skip","paper_url":"https://arxiv.org/abs/1909.03573v1","paper_title":"LCSCNet: Linear Compressing Based Skip-Connecting Network for Image Super-Resolution","code":"https://github.com/XuechenZhang123/LCSC","n_code_links":1,"syntology":null},{"rank_in_archive_order":27,"model":"N3Net","metrics":{"PSNR":"33.84"},"uses_additional_data":false,"paper_date":"2018-10-30","paper":"/paper/neural-nearest-neighbors-networks","paper_url":"http://arxiv.org/abs/1810.12575v1","paper_title":"Neural Nearest Neighbors Networks","code":"https://github.com/visinf/n3net","n_code_links":2,"syntology":null},{"rank_in_archive_order":28,"model":"RED30","metrics":{"PSNR":"33.82","SSIM":"0.923"},"uses_additional_data":false,"paper_date":"2016-06-29","paper":"/paper/image-restoration-using-convolutional-auto","paper_url":"http://arxiv.org/abs/1606.08921v3","paper_title":"Image Restoration Using Convolutional Auto-encoders with Symmetric Skip Connections","code":"https://github.com/titu1994/Image-Super-Resolution","n_code_links":17,"syntology":null},{"rank_in_archive_order":29,"model":"DnCNN-3","metrics":{"PSNR":"33.75"},"uses_additional_data":false,"paper_date":"2016-08-13","paper":"/paper/beyond-a-gaussian-denoiser-residual-learning","paper_url":"http://arxiv.org/abs/1608.03981v1","paper_title":"Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising","code":"https://github.com/cszn/DnCNN","n_code_links":22,"syntology":{"n_ran":1,"n_unverified":5,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":30,"model":"RC-Net","metrics":{"PSNR":"33.43"},"uses_additional_data":false,"paper_date":"2019-10-19","paper":"/paper/image-restoration-using-deep-regulated","paper_url":"https://arxiv.org/abs/1910.08853v1","paper_title":"Image Restoration Using Deep Regulated Convolutional Networks","code":"https://github.com/cswin/RC-Nets","n_code_links":1,"syntology":null},{"rank_in_archive_order":31,"model":"Deep CNN Denoiser","metrics":{"PSNR":"31.26"},"uses_additional_data":false,"paper_date":"2017-04-11","paper":"/paper/learning-deep-cnn-denoiser-prior-for-image","paper_url":"http://arxiv.org/abs/1704.03264v1","paper_title":"Learning Deep CNN Denoiser Prior for Image Restoration","code":"https://github.com/cszn/ircnn","n_code_links":2,"syntology":null},{"rank_in_archive_order":32,"model":"HyperRes","metrics":{"PSNR":"29.77","SSIM":"0.84"},"uses_additional_data":false,"paper_date":"2022-06-13","paper":"/paper/one-size-fits-all-hypernetwork-for-tunable","paper_url":"https://arxiv.org/abs/2206.05970v3","paper_title":"Hypernetwork-Based Adaptive Image Restoration","code":"https://github.com/ifryed/HyperRes","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}}],"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":8,"rows_with_any_sample_ran":8,"distinct_papers_with_graph_line":7,"distinct_papers_with_any_sample_ran":7,"samples_over_distinct_papers":{"n_ran":19,"n_unverified":15,"n_samples":34,"n_pointer_only_licence":4,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":23,"n_unverified":15,"n_samples":38,"n_pointer_only_licence":7,"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"}}}