{"url":"/dataset/aflw-19","name":"AFLW-19","full_name":"The 19 landmark variant of AFLW.","description_markdown":"The [original AFLW](https://www.tugraz.at/institute/icg/research/team-bischof/lrs/downloads/aflw/) provides at most 21 points for each face, but excluding coordinates for invisible landmarks, causing difficulties for training most of the existing baseline approaches. To make fair comparisons, the authors manually annotate the coordinates of these invisible landmarks to enable training of those baseline approaches. The new annotation does not include\r\ntwo ear points because it is very difficult to decide the location of invisible ears. This causes the point number of AFLW-19 to be 19.\r\n\r\nThe original AFLW does not provide train-test partition. AFLW-19 adopts a partition with 20,000 images for training and 4,386 images for testing (AFLW-Full). In addition,  a frontal subset  (AFLW-Frontal) is proposed where all landmarks are visible (totally 1,165 images).\r\n\r\nThe new 19-point annotation file is available at the [project page](http://mmlab.ie.cuhk.edu.hk/projects/compositional.html).","description_withheld":null,"homepage":"http://mmlab.ie.cuhk.edu.hk/projects/compositional.html","introduced_date":"2016-06-01","introduced_date_note":null,"introduced_by":{"paper":"/paper/unconstrained-face-alignment-via-cascaded","title":"Unconstrained Face Alignment via Cascaded Compositional Learning","first_author":"Shizhan Zhu","url":null},"license":null,"modalities":[{"name":"Images","url":"/datasets/modality/images"}],"tasks":[{"name":"Face Alignment","url":"/task/face-alignment","datasets_with_task":"/datasets/task/face-alignment"}],"languages":[],"variants":["AFLW-19"],"data_loaders":[],"num_papers_in_archive":24,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/face-alignment-on-aflw-19","task":"Face Alignment","dataset_variant":"AFLW-19","rows":23,"metrics":["NME_diag (%, Full)","NME_diag (%, Frontal)","NME_box (%, Full)","AUC_box@0.07 (%, Full)"],"first_row_in_archive_order":{"model":"FiFA","paper":"/paper/fiducial-focus-augmentation-for-facial","metrics":{"AUC_box@0.07 (%, Full)":"81.8","NME_box (%, Full)":"1.31","NME_diag (%, Frontal)":"0.80","NME_diag (%, Full)":"0.92"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/fiducial-focus-augmentation-for-facial","title":"Fiducial Focus Augmentation for Facial Landmark Detection","date":"2024-02-23","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/exploring-stylegan-latent-space-for-face","title":"Exploring StyleGAN Latent Space for Face Alignment with Limited Training Data","date":"2022-09-16","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/towards-accurate-facial-landmark-detection-1","title":"Towards Accurate Facial Landmark Detection via Cascaded Transformers","date":"2022-08-23","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/general-facial-representation-learning-in-a","title":"General Facial Representation Learning in a Visual-Linguistic Manner","date":"2021-12-06","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":5,"samples_ran":1,"samples_unverified":4,"pointer_only_for_licence":4,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/subpixel-heatmap-regression-for-facial","title":"Subpixel Heatmap Regression for Facial Landmark Localization","date":"2021-11-03","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/pre-training-strategies-and-datasets-for","title":"Pre-training strategies and datasets for facial representation learning","date":"2021-03-30","rows_on_this_dataset":1,"code_links":2,"syntology":null},{"paper":"/paper/robust-face-alignment-by-multi-order-high","title":"Robust Face Alignment by Multi-order High-precision Hourglass Network","date":"2020-10-17","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/atf-towards-robust-face-alignment-via","title":"ATF: Towards Robust Face Alignment via Leveraging Similarity and Diversity across Different Datasets","date":"2020-10-12","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/anchorface-an-anchor-based-facial-landmark","title":"AnchorFace: An Anchor-based Facial Landmark Detector Across Large Poses","date":"2020-07-07","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/luvli-face-alignment-estimating-landmarks","title":"LUVLi Face Alignment: Estimating Landmarks' Location, Uncertainty, and Visibility Likelihood","date":"2020-04-06","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/pixel-in-pixel-net-towards-efficient-facial","title":"Pixel-in-Pixel Net: Towards Efficient Facial Landmark Detection in the Wild","date":"2020-03-08","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":17,"samples_ran":2,"samples_unverified":15,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/3fabrec-fast-few-shot-face-alignment-by","title":"3FabRec: Fast Few-shot Face alignment by Reconstruction","date":"2019-11-24","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/face-alignment-with-kernel-density-deep","title":"Face Alignment With Kernel Density Deep Neural Network","date":"2019-10-01","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/high-resolution-representations-for-labeling","title":"High-Resolution Representations for Labeling Pixels and Regions","date":"2019-04-09","rows_on_this_dataset":1,"code_links":39,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":18,"samples_ran":3,"samples_unverified":15,"pointer_only_for_licence":5,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/laplace-landmark-localization","title":"Laplace Landmark Localization","date":"2019-03-27","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/direct-shape-regression-networks-for-end-to","title":"Direct Shape Regression Networks for End-to-End Face Alignment","date":"2018-06-01","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/look-at-boundary-a-boundary-aware-face","title":"Look at Boundary: A Boundary-Aware Face Alignment Algorithm","date":"2018-05-26","rows_on_this_dataset":2,"code_links":2,"syntology":null},{"paper":"/paper/style-aggregated-network-for-facial-landmark","title":"Style Aggregated Network for Facial Landmark Detection","date":"2018-03-12","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/wing-loss-for-robust-facial-landmark","title":"Wing Loss for Robust Facial Landmark Localisation with Convolutional Neural Networks","date":"2017-11-17","rows_on_this_dataset":1,"code_links":6,"syntology":null},{"paper":"/paper/dynamic-attention-controlled-cascaded-shape","title":"Dynamic Attention-controlled Cascaded Shape Regression Exploiting Training Data Augmentation and Fuzzy-set Sample Weighting","date":"2016-11-16","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/unconstrained-face-alignment-via-cascaded","title":"Unconstrained Face Alignment via Cascaded Compositional Learning","date":"2016-06-01","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/face-alignment-by-coarse-to-fine-shape-1","title":"Face alignment by coarse-to-fine shape searching","date":"2015-06-07","rows_on_this_dataset":1,"code_links":1,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":3,"samples_harvested":40,"samples_ran":6,"samples_unverified":34,"pointer_only_for_licence":9,"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."}