{"url":"/task/image-restoration","name":"Image Restoration","slug":"image-restoration","description_markdown":"**Image Restoration** is a family of inverse problems for obtaining a high quality image from a corrupted input image. Corruption may occur due to the image-capture process (e.g., noise, lens blur), post-processing (e.g., JPEG compression), or photography in non-ideal conditions (e.g., haze, motion blur).\r\n\r\n\r\n<span class=\"description-source\">Source: [Blind Image Restoration without Prior Knowledge ](https://arxiv.org/abs/2003.01764)</span>","categories":[{"name":"Computer Vision","url":"/area/computer-vision"}],"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","slug_source":"archive_url"},"counts":{"papers_tagged":1459,"papers_with_code":666,"benchmarks":2,"benchmark_tables_in_archive":3,"benchmark_tables_shown":2,"benchmark_tables_withheld_as_spam":1,"benchmark_definition":"a leaderboard table with at least one row; benchmark_tables_shown also counts the zero-row tables; benchmark_tables_in_archive adds the tables withheld as spam","datasets":18,"subtasks":13,"parent_tasks":1},"benchmarks":[{"leaderboard":"/sota/image-restoration-on-cdd-11","slug":"image-restoration-on-cdd-11","dataset":"CDD-11","dataset_url":"/dataset/cdd-11","rows_in_archive":14,"metrics":["Average PSNR (dB)","SSIM"],"first_row_in_archive_order":{"model":"OneRestore","paper_title":"OneRestore: A Universal Restoration Framework for Composite Degradation","paper_url":"/paper/onerestore-a-universal-restoration-framework","paper_date":"2024-07-05","arxiv_id":"2407.04621","code_links":[{"title":"gy65896/onerestore","url":"https://github.com/gy65896/onerestore"}],"syntology":{"n":13,"n_ran":7,"n_unverified":6,"n_pointer_only":13}}},{"leaderboard":"/sota/image-restoration-on-uhdm","slug":"image-restoration-on-uhdm","dataset":"UHDM","dataset_url":"/dataset/uhdm","rows_in_archive":2,"metrics":["PSNR"],"first_row_in_archive_order":{"model":"ESDNet-L","paper_title":"Towards Efficient and Scale-Robust Ultra-High-Definition Image Demoireing","paper_url":"/paper/towards-efficient-and-scale-robust-ultra-high","paper_date":"2022-07-20","arxiv_id":"2207.09935","code_links":[{"title":"CVMI-Lab/UHDM","url":"https://github.com/CVMI-Lab/UHDM"}],"syntology":null}}],"datasets":[{"url":"/dataset/sidd","name":"SIDD","full_name":"Smartphone Image Denoising Dataset","num_papers_in_archive":245},{"url":"/dataset/cbsd68","name":"CBSD68","full_name":"Color BSD68","num_papers_in_archive":142},{"url":"/dataset/raindrop","name":"Raindrop","full_name":"","num_papers_in_archive":120},{"url":"/dataset/hide","name":"HIDE","full_name":"","num_papers_in_archive":95},{"url":"/dataset/pirm","name":"PIRM","full_name":"Perceptual Image Restoration and Manipulation","num_papers_in_archive":32},{"url":"/dataset/tinyperson","name":"TinyPerson","full_name":"","num_papers_in_archive":25},{"url":"/dataset/docunet","name":"DocUNet","full_name":"Document Image Unwarping via a Stacked U-Net","num_papers_in_archive":23},{"url":"/dataset/cdd-11","name":"CDD-11","full_name":"Composite Degradation Dataset 11","num_papers_in_archive":21},{"url":"/dataset/real-rain-dataset","name":"Real Rain Dataset","full_name":"","num_papers_in_archive":11},{"url":"/dataset/uhdm","name":"UHDM","full_name":"","num_papers_in_archive":2},{"url":"/dataset/awmm-100k","name":"AWMM-100k","full_name":"","num_papers_in_archive":1},{"url":"/dataset/fingerprint-inpainting-and-denoising","name":"Fingerprint inpainting and denoising","full_name":"","num_papers_in_archive":1},{"url":"/dataset/hac","name":"HAC","full_name":"Hybrid Adverse Conditions","num_papers_in_archive":1},{"url":"/dataset/hri","name":"HRI","full_name":"High-resolution Rainy Image","num_papers_in_archive":1},{"url":"/dataset/l1bsr","name":"L1BSR","full_name":"L1BSR dataset","num_papers_in_archive":1},{"url":"/dataset/raw-natural-image-noise-dataset","name":"RawNIND","full_name":"Raw Natural Image Noise Dataset","num_papers_in_archive":1},{"url":"/dataset/sidl-smartphone-images-with-dirty-lenses","name":"SIDL: A Real-World Dataset for Restoring Smartphone Images with Dirty Lenses","full_name":"","num_papers_in_archive":1},{"url":"/dataset/ug-2","name":"UG^2","full_name":"","num_papers_in_archive":1}],"subtasks":[{"url":"/task/5-degradation-blind-all-in-one-image","name":"5-Degradation Blind All-in-One Image Restoration"},{"url":"/task/blind-super-resolution","name":"Blind Super-Resolution"},{"url":"/task/demosaicking","name":"Demosaicking"},{"url":"/task/flare-removal","name":"Flare Removal"},{"url":"/task/highlight-removal","name":"highlight removal"},{"url":"/task/image-compression-artifact-reduction","name":"Image Compression Artifact Reduction"},{"url":"/task/jpeg-artifact-correction","name":"JPEG Artifact Correction"},{"url":"/task/jpeg-artifact-removal","name":"JPEG Artifact Removal"},{"url":"/task/single-image-desnowing","name":"Single Image Desnowing"},{"url":"/task/spectral-reconstruction","name":"Spectral Reconstruction"},{"url":"/task/spectral-super-resolution","name":"Spectral Super-Resolution"},{"url":"/task/underwater-image-restoration","name":"Underwater Image Restoration"},{"url":"/task/unified-image-restoration","name":"Unified Image Restoration"}],"parent_tasks":[{"url":"/task/10-shot-image-generation","name":"10-shot image generation"}],"papers":{"order":"repositories listed in the archive (desc), then date (desc); the archive holds no stars","population":"papers tagged with this task that list at least one repository in the archive","shown":30,"of":666,"tagged_in_all":1459,"items":[{"url":"/paper/noise2noise-learning-image-restoration","title":"Noise2Noise: Learning Image Restoration without Clean Data","date":"2018-03-12","arxiv_id":"1803.04189","repositories_listed":21,"syntology":{"n":7,"n_ran":1,"n_unverified":6,"n_pointer_only":1}},{"url":"/paper/image-restoration-using-convolutional-auto","title":"Image Restoration Using Convolutional Auto-encoders with Symmetric Skip Connections","date":"2016-06-29","arxiv_id":"1606.08921","repositories_listed":17,"syntology":null},{"url":"/paper/deep-image-prior","title":"Deep Image Prior","date":"2017-11-29","arxiv_id":"1711.10925","repositories_listed":14,"syntology":{"n":6,"n_ran":5,"n_unverified":1,"n_pointer_only":6}},{"url":"/paper/simple-baselines-for-image-restoration","title":"Simple Baselines for Image Restoration","date":"2022-04-10","arxiv_id":"2204.04676","repositories_listed":13,"syntology":{"n":30,"n_ran":23,"n_unverified":7,"n_pointer_only":17}},{"url":"/paper/restormer-efficient-transformer-for-high","title":"Restormer: Efficient Transformer for High-Resolution Image Restoration","date":"2021-11-18","arxiv_id":"2111.09881","repositories_listed":13,"syntology":{"n":4,"n_ran":4,"n_unverified":0,"n_pointer_only":4}},{"url":"/paper/learning-enriched-features-for-real-image","title":"Learning Enriched Features for Real Image Restoration and Enhancement","date":"2020-03-15","arxiv_id":"2003.06792","repositories_listed":12,"syntology":{"n":20,"n_ran":3,"n_unverified":17,"n_pointer_only":3}},{"url":"/paper/swinir-image-restoration-using-swin","title":"SwinIR: Image Restoration Using Swin Transformer","date":"2021-08-23","arxiv_id":"2108.10257","repositories_listed":9,"syntology":{"n":45,"n_ran":30,"n_unverified":15,"n_pointer_only":5}},{"url":"/paper/multi-stage-progressive-image-restoration","title":"Multi-Stage Progressive Image Restoration","date":"2021-02-04","arxiv_id":"2102.02808","repositories_listed":8,"syntology":{"n":26,"n_ran":18,"n_unverified":8,"n_pointer_only":25}},{"url":"/paper/old-photo-restoration-via-deep-latent-space","title":"Old Photo Restoration via Deep Latent Space Translation","date":"2020-09-14","arxiv_id":"2009.07047","repositories_listed":8,"syntology":null},{"url":"/paper/cycleisp-real-image-restoration-via-improved","title":"CycleISP: Real Image Restoration via Improved Data Synthesis","date":"2020-03-17","arxiv_id":"2003.07761","repositories_listed":8,"syntology":null},{"url":"/paper/enlightengan-deep-light-enhancement-without","title":"EnlightenGAN: Deep Light Enhancement without Paired Supervision","date":"2019-06-17","arxiv_id":"1906.06972","repositories_listed":8,"syntology":null},{"url":"/paper/the-2018-pirm-challenge-on-perceptual-image","title":"The 2018 PIRM Challenge on Perceptual Image Super-resolution","date":"2018-09-20","arxiv_id":"1809.07517","repositories_listed":8,"syntology":null},{"url":"/paper/bringing-old-photos-back-to-life","title":"Bringing Old Photos Back to Life","date":"2020-04-20","arxiv_id":"2004.09484","repositories_listed":7,"syntology":{"n":31,"n_ran":2,"n_unverified":29,"n_pointer_only":20}},{"url":"/paper/deblurgan-v2-deblurring-orders-of-magnitude","title":"DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Better","date":"2019-08-10","arxiv_id":"1908.03826","repositories_listed":6,"syntology":null},{"url":"/paper/swin2sr-swinv2-transformer-for-compressed","title":"Swin2SR: SwinV2 Transformer for Compressed Image Super-Resolution and Restoration","date":"2022-09-22","arxiv_id":"2209.11345","repositories_listed":5,"syntology":null},{"url":"/paper/nafssr-stereo-image-super-resolution-using","title":"NAFSSR: Stereo Image Super-Resolution Using NAFNet","date":"2022-04-19","arxiv_id":"2204.08714","repositories_listed":5,"syntology":null},{"url":"/paper/sar2sar-a-self-supervised-despeckling","title":"SAR2SAR: a semi-supervised despeckling algorithm for SAR images","date":"2020-06-26","arxiv_id":"2006.15037","repositories_listed":5,"syntology":null},{"url":"/paper/hifacegan-face-renovation-via-collaborative","title":"HiFaceGAN: Face Renovation via Collaborative Suppression and Replenishment","date":"2020-05-11","arxiv_id":"2005.05005","repositories_listed":5,"syntology":null},{"url":"/paper/multi-level-wavelet-cnn-for-image-restoration","title":"Multi-level Wavelet-CNN for Image Restoration","date":"2018-05-18","arxiv_id":"1805.07071","repositories_listed":5,"syntology":null},{"url":"/paper/zero-shot-image-restoration-using-denoising","title":"Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model","date":"2022-12-01","arxiv_id":"2212.00490","repositories_listed":4,"syntology":{"n":23,"n_ran":12,"n_unverified":11,"n_pointer_only":0}},{"url":"/paper/mask-guided-spectral-wise-transformer-for","title":"Mask-guided Spectral-wise Transformer for Efficient Hyperspectral Image Reconstruction","date":"2021-11-15","arxiv_id":"2111.07910","repositories_listed":4,"syntology":{"n":3,"n_ran":3,"n_unverified":0,"n_pointer_only":0}},{"url":"/paper/uformer-a-general-u-shaped-transformer-for","title":"Uformer: A General U-Shaped Transformer for Image Restoration","date":"2021-06-06","arxiv_id":"2106.03106","repositories_listed":4,"syntology":{"n":6,"n_ran":6,"n_unverified":0,"n_pointer_only":2}},{"url":"/paper/plug-and-play-image-restoration-with-deep","title":"Plug-and-Play Image Restoration with Deep Denoiser Prior","date":"2020-08-31","arxiv_id":"2008.13751","repositories_listed":4,"syntology":{"n":28,"n_ran":3,"n_unverified":25,"n_pointer_only":0}},{"url":"/paper/deep-learning-based-channel-estimation","title":"Deep Learning-Based Channel Estimation","date":"2018-10-13","arxiv_id":"1810.05893","repositories_listed":4,"syntology":{"n":2,"n_ran":2,"n_unverified":0,"n_pointer_only":2}},{"url":"/paper/recurrent-inference-machines-for-solving","title":"Recurrent Inference Machines for Solving Inverse Problems","date":"2017-06-13","arxiv_id":"1706.04008","repositories_listed":4,"syntology":null},{"url":"/paper/self-rectifying-diffusion-sampling-with","title":"Self-Rectifying Diffusion Sampling with Perturbed-Attention Guidance","date":"2024-03-26","arxiv_id":"2403.17377","repositories_listed":3,"syntology":{"n":16,"n_ran":13,"n_unverified":3,"n_pointer_only":0}},{"url":"/paper/denoising-diffusion-models-for-plug-and-play","title":"Denoising Diffusion Models for Plug-and-Play Image Restoration","date":"2023-05-15","arxiv_id":"2305.08995","repositories_listed":3,"syntology":{"n":16,"n_ran":13,"n_unverified":3,"n_pointer_only":1}},{"url":"/paper/cross-aggregation-transformer-for-image","title":"Cross Aggregation Transformer for Image Restoration","date":"2022-11-24","arxiv_id":"2211.13654","repositories_listed":3,"syntology":{"n":11,"n_ran":7,"n_unverified":4,"n_pointer_only":0}},{"url":"/paper/masked-frequency-modeling-for-self-supervised","title":"Masked Frequency Modeling for Self-Supervised Visual Pre-Training","date":"2022-06-15","arxiv_id":"2206.07706","repositories_listed":3,"syntology":{"n":5,"n_ran":4,"n_unverified":1,"n_pointer_only":2}},{"url":"/paper/mst-multi-stage-spectral-wise-transformer-for","title":"MST++: Multi-stage Spectral-wise Transformer for Efficient Spectral Reconstruction","date":"2022-04-17","arxiv_id":"2204.07908","repositories_listed":3,"syntology":{"n":3,"n_ran":3,"n_unverified":0,"n_pointer_only":3}}],"syntology_records":18,"syntology_note":"a paper without a record is not a recorded non-run: it may lack an arXiv id or simply be absent from the graph layer"},"description_links":{"kept":0,"unwrapped_to_text":0,"bare_urls_linked":0,"relative_images_dropped":0,"rule":"internal links are kept only when the target slug exists in the catalog"},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per-sample execution status on synthesized fixtures ('ran N of M samples'); not a correctness claim and not a ranking signal.","status_vocabulary":{"ran_honours":"ran, honoured the contract we drafted","ran_violates":"ran, violated the contract we drafted","ran_draft_wrong":"ran; our contract draft was wrong, not the code","ran_fixture":"ran; our fixture could not drive it","ran":"ran on a synthesized input","unverified":"unverified (harvested, no recorded run)"}},"not_shown":{"libraries":"the archive has no per-task library table","trend_sparklines":"the Trend column of the benchmarks table was a rendered image; it is not in the archive","social_and_latest_sorts":"stars and social signals are not in the archive"}}