{"url":"/dataset/aim-500","name":"AIM-500","full_name":"Automatic Image Matting-500","description_markdown":"AIM-500 is the first natural image matting test set, contains 500 high-resolution real-world natural images from three types of images (salient opaque foregrounds, salient transparent/meticulous foregrounds, non-salient foregrounds), and multiple categories. The amount of each category is shown in the following table.\r\n\r\n| Portrait | Animal | Transparent | Plant | Furniture | Toy | Fruit |\r\n| :----:| :----: |  :----: |  :----: |  :----: |  :----: |  :----: | \r\n| 100 | 200 | 34 | 75 | 45 | 36 | 10 |","description_withheld":null,"homepage":"https://github.com/JizhiziLi/AIM","introduced_date":"2021-07-15","introduced_date_note":null,"introduced_by":{"paper":"/paper/deep-automatic-natural-image-matting","title":"Deep Automatic Natural Image Matting","first_author":"Jizhizi Li","url":null},"license":{"name":"MIT","url":null},"modalities":[],"tasks":[{"name":"Image Matting","url":"/task/image-matting","datasets_with_task":"/datasets/task/image-matting"}],"languages":[],"variants":["AIM-500"],"data_loaders":[{"repo":"https://github.com/JizhiziLi/AIM","url":"https://github.com/JizhiziLi/AIM","frameworks":["tf","pytorch"]}],"num_papers_in_archive":23,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/image-matting-on-aim-500","task":"Image Matting","dataset_variant":"AIM-500","rows":6,"metrics":["SAD","MSE","MAD","Conn.","Grad."],"first_row_in_archive_order":{"model":"DiffMatte","paper":"/paper/diffusion-for-natural-image-matting","metrics":{"Conn.":"15.98","Grad.":"15.68","MAD":"0.0098","MSE":"0.0033","SAD":"16.31"},"code_links":[{"title":"yihanhu-2022/diffmatte","url":"https://github.com/yihanhu-2022/diffmatte"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/diffusion-for-natural-image-matting","title":"Diffusion for Natural Image Matting","date":"2023-12-10","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":7,"samples_ran":6,"samples_unverified":1,"pointer_only_for_licence":7,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/deep-automatic-natural-image-matting","title":"Deep Automatic Natural Image Matting","date":"2021-07-15","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":2,"samples_unverified":1,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/end-to-end-animal-image-matting","title":"Bridging Composite and Real: Towards End-to-end Deep Image Matting","date":"2020-10-30","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":3,"samples_unverified":0,"pointer_only_for_licence":1,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/a-late-fusion-cnn-for-digital-matting","title":"A Late Fusion CNN for Digital Matting","date":"2019-06-01","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/u2-net-a-bayesian-u-net-model-with-epistemic","title":"U2-Net: A Bayesian U-Net model with epistemic uncertainty feedback for photoreceptor layer segmentation in pathological OCT scans","date":"2019-01-23","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/semantic-human-matting","title":"Semantic Human Matting","date":"2018-09-05","rows_on_this_dataset":1,"code_links":2,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":3,"samples_harvested":13,"samples_ran":11,"samples_unverified":2,"pointer_only_for_licence":8,"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."}