{"url":"/dataset/polyu-dataset","name":"PolyU","full_name":null,"description_markdown":"PolyU Dataset is a large dataset of real-world noisy images with reasonably obtained corresponding “ground truth” images. The basic idea is to capture the same and unchanged scene for many (e.g., 500) times and compute their mean image, which can be roughly taken as the “ground truth” image for the real-world noisy images. The rational of this strategy is that for each pixel, the noise is generated randomly larger or smaller than 0. Sampling the same pixel many times and computing the average value will approximate the truth pixel value and alleviate significantly the noise.\r\n\r\nSource: [Real-world Noisy Image Denoising: A New Benchmark](https://arxiv.org/pdf/1804.02603.pdf)","description_withheld":null,"homepage":"https://github.com/csjunxu/PolyU-Real-World-Noisy-Images-Dataset","introduced_date":"2018-04-07","introduced_date_note":null,"introduced_by":{"paper":"/paper/real-world-noisy-image-denoising-a-new","title":"Real-world Noisy Image Denoising: A New Benchmark","first_author":"Jun Xu","url":null},"license":null,"modalities":[{"name":"Images","url":"/datasets/modality/images"}],"tasks":[{"name":"Denoising","url":"/task/denoising","datasets_with_task":"/datasets/task/denoising"},{"name":"Image Enhancement","url":"/task/image-enhancement","datasets_with_task":"/datasets/task/image-enhancement"},{"name":"Image Denoising","url":"/task/image-denoising","datasets_with_task":"/datasets/task/image-denoising"}],"languages":[],"variants":["PolyU"],"data_loaders":[{"repo":"https://github.com/csjunxu/PolyU-Real-World-Noisy-Images-Dataset","url":"https://github.com/csjunxu/PolyU-Real-World-Noisy-Images-Dataset","frameworks":[]}],"num_papers_in_archive":34,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/image-denoising-on-polyu-dataset","task":"Image Denoising","dataset_variant":"PolyU","rows":1,"metrics":["PSNR","SSIM"],"first_row_in_archive_order":{"model":"PNGAN","paper":"/paper/learning-to-generate-realistic-noisy-images-2","metrics":{"PSNR":"40.55","SSIM":"0.983"},"code_links":[{"title":"caiyuanhao1998/PNGAN","url":"https://github.com/caiyuanhao1998/PNGAN"},{"title":"GarrickZ2/Image-Denoising","url":"https://github.com/GarrickZ2/Image-Denoising"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/learning-to-generate-realistic-noisy-images-2","title":"Learning to Generate Realistic Noisy Images via Pixel-level Noise-aware Adversarial Training","date":"2022-04-06","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":18,"samples_ran":14,"samples_unverified":4,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":1,"samples_harvested":18,"samples_ran":14,"samples_unverified":4,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}