{"url":"/sota/face-alignment-on-aflw-19","task":{"name":"Face Alignment","url":"/task/face-alignment","note":null},"dataset":{"name":"AFLW-19","url":"/dataset/aflw-19"},"category":"Computer Vision","categories":["Computer Vision","Miscellaneous","Music"],"category_note":null,"description":"Face alignment is the task of identifying the geometric structure of faces in digital images, and attempting to obtain a canonical alignment of the face based on translation, scale, and rotation.\r\n\r\n<span style=\"color:grey; opacity: 0.6\">( Image credit: [3DDFA_V2](https://github.com/cleardusk/3DDFA_V2) )</span>","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":["NME_diag (%, Full)","NME_diag (%, Frontal)","NME_box (%, Full)","AUC_box@0.07 (%, Full)"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"NME_diag (%, Full)":null,"NME_diag (%, Frontal)":null,"NME_box (%, Full)":null,"AUC_box@0.07 (%, Full)":"higher"}},"counts":{"rows":23,"rows_with_code":13,"rows_with_paper_page":23,"rows_dated":23,"rows_using_additional_data":6},"rows":[{"rank_in_archive_order":1,"model":"FiFA","metrics":{"AUC_box@0.07 (%, Full)":"81.8","NME_box (%, Full)":"1.31","NME_diag (%, Frontal)":"0.80","NME_diag (%, Full)":"0.92"},"uses_additional_data":false,"paper_date":"2024-02-23","paper":"/paper/fiducial-focus-augmentation-for-facial","paper_url":"https://arxiv.org/abs/2402.15044v1","paper_title":"Fiducial Focus Augmentation for Facial Landmark Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":2,"model":"FaRL-B (epoch 16)","metrics":{"AUC_box@0.07 (%, Full)":"81.3","NME_box (%, Full)":"1.334","NME_diag (%, Frontal)":"0.821","NME_diag (%, Full)":"0.943"},"uses_additional_data":true,"paper_date":"2021-12-06","paper":"/paper/general-facial-representation-learning-in-a","paper_url":"https://arxiv.org/abs/2112.03109v3","paper_title":"General Facial Representation Learning in a Visual-Linguistic Manner","code":"https://github.com/FacePerceiver/FaRL","n_code_links":2,"syntology":{"n_ran":1,"n_unverified":4,"n_samples":5,"n_pointer_only_licence":4}},{"rank_in_archive_order":3,"model":"FASE","metrics":{"AUC_box@0.07 (%, Full)":"79.1","NME_box (%, Full)":"1.45","NME_diag (%, Frontal)":"0.90","NME_diag (%, Full)":"1.02"},"uses_additional_data":false,"paper_date":"2022-09-16","paper":"/paper/exploring-stylegan-latent-space-for-face","paper_url":"https://hal.archives-ouvertes.fr/INRIA/hal-03778322v1","paper_title":"Exploring StyleGAN Latent Space for Face Alignment with Limited Training Data","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":4,"model":"LAB (w/ B)","metrics":{"NME_diag (%, Frontal)":"1.14","NME_diag (%, Full)":"1.25"},"uses_additional_data":true,"paper_date":"2018-05-26","paper":"/paper/look-at-boundary-a-boundary-aware-face","paper_url":"http://arxiv.org/abs/1805.10483v1","paper_title":"Look at Boundary: A Boundary-Aware Face Alignment Algorithm","code":"https://github.com/open-mmlab/mmpose","n_code_links":2,"syntology":null},{"rank_in_archive_order":5,"model":"SHR-FAN","metrics":{"AUC_box@0.07 (%, Full)":"70.0","NME_box (%, Full)":"2.14","NME_diag (%, Frontal)":"1.12","NME_diag (%, Full)":"1.31"},"uses_additional_data":true,"paper_date":"2021-11-03","paper":"/paper/subpixel-heatmap-regression-for-facial","paper_url":"https://arxiv.org/abs/2111.02360v1","paper_title":"Subpixel Heatmap Regression for Facial Landmark Localization","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"DTLD+","metrics":{"NME_diag (%, Full)":"1.37"},"uses_additional_data":false,"paper_date":"2022-08-23","paper":"/paper/towards-accurate-facial-landmark-detection-1","paper_url":"https://arxiv.org/abs/2208.10808v1","paper_title":"Towards Accurate Facial Landmark Detection via Cascaded Transformers","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":7,"model":"MHHN","metrics":{"NME_diag (%, Frontal)":"1.19","NME_diag (%, Full)":"1.38"},"uses_additional_data":false,"paper_date":"2020-10-17","paper":"/paper/robust-face-alignment-by-multi-order-high","paper_url":"https://arxiv.org/abs/2010.08722v1","paper_title":"Robust Face Alignment by Multi-order High-precision Hourglass Network","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":8,"model":"LUVLi","metrics":{"AUC_box@0.07 (%, Full)":"68.0","NME_box (%, Full)":"2.28","NME_diag (%, Frontal)":"1.19","NME_diag (%, Full)":"1.39"},"uses_additional_data":false,"paper_date":"2020-04-06","paper":"/paper/luvli-face-alignment-estimating-landmarks","paper_url":"https://arxiv.org/abs/2004.02980v1","paper_title":"LUVLi Face Alignment: Estimating Landmarks' Location, Uncertainty, and Visibility Likelihood","code":"https://github.com/abhi1kumar/LUVLi","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"PIPNet (ResNet-101)","metrics":{"NME_diag (%, Full)":"1.42"},"uses_additional_data":false,"paper_date":"2020-03-08","paper":"/paper/pixel-in-pixel-net-towards-efficient-facial","paper_url":"https://arxiv.org/abs/2003.03771v3","paper_title":"Pixel-in-Pixel Net: Towards Efficient Facial Landmark Detection in the Wild","code":"https://github.com/jhb86253817/PIPNet","n_code_links":2,"syntology":{"n_ran":2,"n_unverified":15,"n_samples":17,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"VGG-F","metrics":{"NME_diag (%, Full)":"1.55"},"uses_additional_data":true,"paper_date":"2021-03-30","paper":"/paper/pre-training-strategies-and-datasets-for","paper_url":"https://arxiv.org/abs/2103.16554v2","paper_title":"Pre-training strategies and datasets for facial representation learning","code":"https://github.com/tomas-gajarsky/facetorch","n_code_links":2,"syntology":null},{"rank_in_archive_order":11,"model":"ATF","metrics":{"NME_diag (%, Full)":"1.55"},"uses_additional_data":true,"paper_date":"2020-10-12","paper":"/paper/atf-towards-robust-face-alignment-via","paper_url":"https://dl.acm.org/doi/10.1145/3394171.3414037","paper_title":"ATF: Towards Robust Face Alignment via Leveraging Similarity and Diversity across Different Datasets","code":"https://github.com/starhiking/ATF","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"AnchorFace","metrics":{"NME_diag (%, Frontal)":"1.38","NME_diag (%, Full)":"1.56"},"uses_additional_data":false,"paper_date":"2020-07-07","paper":"/paper/anchorface-an-anchor-based-facial-landmark","paper_url":"https://arxiv.org/abs/2007.03221v3","paper_title":"AnchorFace: An Anchor-based Facial Landmark Detector Across Large Poses","code":"https://github.com/nothingelse92/AnchorFace","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"HR-Net","metrics":{"NME_diag (%, Frontal)":"1.46","NME_diag (%, Full)":"1.57"},"uses_additional_data":false,"paper_date":"2019-04-09","paper":"/paper/high-resolution-representations-for-labeling","paper_url":"http://arxiv.org/abs/1904.04514v1","paper_title":"High-Resolution Representations for Labeling Pixels and Regions","code":"https://github.com/PaddlePaddle/PaddleDetection","n_code_links":39,"syntology":{"n_ran":3,"n_unverified":15,"n_samples":18,"n_pointer_only_licence":5}},{"rank_in_archive_order":14,"model":"Wing","metrics":{"AUC_box@0.07 (%, Full)":"53.5","NME_box (%, Full)":"3.56","NME_diag (%, Full)":"1.65"},"uses_additional_data":false,"paper_date":"2017-11-17","paper":"/paper/wing-loss-for-robust-facial-landmark","paper_url":"http://arxiv.org/abs/1711.06753v5","paper_title":"Wing Loss for Robust Facial Landmark Localisation with Convolutional Neural Networks","code":"https://github.com/open-mmlab/mmpose","n_code_links":6,"syntology":null},{"rank_in_archive_order":15,"model":"LAB (w/o B)","metrics":{"NME_diag (%, Frontal)":"1.62","NME_diag (%, Full)":"1.85"},"uses_additional_data":false,"paper_date":"2018-05-26","paper":"/paper/look-at-boundary-a-boundary-aware-face","paper_url":"http://arxiv.org/abs/1805.10483v1","paper_title":"Look at Boundary: A Boundary-Aware Face Alignment Algorithm","code":"https://github.com/open-mmlab/mmpose","n_code_links":2,"syntology":null},{"rank_in_archive_order":16,"model":"DSRN","metrics":{"NME_diag (%, Full)":"1.86"},"uses_additional_data":false,"paper_date":"2018-06-01","paper":"/paper/direct-shape-regression-networks-for-end-to","paper_url":"http://openaccess.thecvf.com/content_cvpr_2018/html/Miao_Direct_Shape_Regression_CVPR_2018_paper.html","paper_title":"Direct Shape Regression Networks for End-to-End Face Alignment","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":17,"model":"SAN","metrics":{"AUC_box@0.07 (%, Full)":"54.0","NME_box (%, Full)":"4.04","NME_diag (%, Frontal)":"1.85","NME_diag (%, Full)":"1.91"},"uses_additional_data":false,"paper_date":"2018-03-12","paper":"/paper/style-aggregated-network-for-facial-landmark","paper_url":"http://arxiv.org/abs/1803.04108v4","paper_title":"Style Aggregated Network for Facial Landmark Detection","code":"https://github.com/D-X-Y/SAN","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"DAC-CSR","metrics":{"NME_diag (%, Frontal)":"1.81","NME_diag (%, Full)":"2.21"},"uses_additional_data":false,"paper_date":"2016-11-16","paper":"/paper/dynamic-attention-controlled-cascaded-shape","paper_url":"http://arxiv.org/abs/1611.05396v2","paper_title":"Dynamic Attention-controlled Cascaded Shape Regression Exploiting Training Data Augmentation and Fuzzy-set Sample Weighting","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":19,"model":"CCL","metrics":{"NME_diag (%, Frontal)":"2.17","NME_diag (%, Full)":"2.72"},"uses_additional_data":false,"paper_date":"2016-06-01","paper":"/paper/unconstrained-face-alignment-via-cascaded","paper_url":"http://openaccess.thecvf.com/content_cvpr_2016/html/Zhu_Unconstrained_Face_Alignment_CVPR_2016_paper.html","paper_title":"Unconstrained Face Alignment via Cascaded Compositional Learning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":20,"model":"CFSS","metrics":{"NME_diag (%, Frontal)":"2.68","NME_diag (%, Full)":"3.92"},"uses_additional_data":false,"paper_date":"2015-06-07","paper":"/paper/face-alignment-by-coarse-to-fine-shape-1","paper_url":"https://www.cv-foundation.org/openaccess/content_cvpr_2015/html/Zhu_Face_Alignment_by_2015_CVPR_paper.html","paper_title":"Face alignment by coarse-to-fine shape searching","code":"https://github.com/zhusz/CVPR15-CFSS","n_code_links":1,"syntology":null},{"rank_in_archive_order":21,"model":"3FabRec","metrics":{"NME_box (%, Full)":"1.84"},"uses_additional_data":true,"paper_date":"2019-11-24","paper":"/paper/3fabrec-fast-few-shot-face-alignment-by","paper_url":"https://arxiv.org/abs/1911.10448v2","paper_title":"3FabRec: Fast Few-shot Face alignment by Reconstruction","code":"https://github.com/browatbn2/3FabRec","n_code_links":1,"syntology":null},{"rank_in_archive_order":22,"model":"LLL","metrics":{"NME_box (%, Full)":"1.97"},"uses_additional_data":false,"paper_date":"2019-03-27","paper":"/paper/laplace-landmark-localization","paper_url":"https://arxiv.org/abs/1903.11633v2","paper_title":"Laplace Landmark Localization","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":23,"model":"KDN","metrics":{"AUC_box@0.07 (%, Full)":"60.3","NME_box (%, Full)":"2.80"},"uses_additional_data":false,"paper_date":"2019-10-01","paper":"/paper/face-alignment-with-kernel-density-deep","paper_url":"http://openaccess.thecvf.com/content_ICCV_2019/html/Chen_Face_Alignment_With_Kernel_Density_Deep_Neural_Network_ICCV_2019_paper.html","paper_title":"Face Alignment With Kernel Density Deep Neural Network","code":null,"n_code_links":0,"syntology":null}],"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 6,264 of the 9,581 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":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+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":3,"rows_with_any_sample_ran":3,"distinct_papers_with_graph_line":3,"distinct_papers_with_any_sample_ran":3,"samples_over_distinct_papers":{"n_ran":6,"n_unverified":34,"n_samples":40,"n_pointer_only_licence":9,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":6,"n_unverified":34,"n_samples":40,"n_pointer_only_licence":9,"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"}}}