{"url":"/sota/pedestrian-detection-on-citypersons","task":{"name":"Pedestrian Detection","url":"/task/pedestrian-detection","note":null},"dataset":{"name":"CityPersons","url":"/dataset/citypersons"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"Pedestrian detection is the task of detecting pedestrians from a camera.\r\n\r\nFurther state-of-the-art results (e.g. on the KITTI dataset) can be found at [3D Object Detection](https://paperswithcode.com/task/object-detection).\r\n\r\n<span style=\"color:grey; opacity: 0.6\">( Image credit: [High-level Semantic Feature Detection: A New Perspective for Pedestrian Detection](https://github.com/liuwei16/CSP) )</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":["Reasonable MR^-2","Heavy MR^-2","Partial MR^-2","Bare MR^-2","Small MR^-2","Medium MR^-2","Large MR^-2","Test Time"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Reasonable MR^-2":null,"Heavy MR^-2":null,"Partial MR^-2":null,"Bare MR^-2":null,"Small MR^-2":null,"Medium MR^-2":null,"Large MR^-2":null,"Test Time":"lower"}},"counts":{"rows":22,"rows_with_code":15,"rows_with_paper_page":22,"rows_dated":22,"rows_using_additional_data":7},"rows":[{"rank_in_archive_order":1,"model":"DIW Loss","metrics":{"Heavy MR^-2":"28.37","Reasonable MR^-2":"6.23","Small MR^-2":"7.36"},"uses_additional_data":true,"paper_date":"2022-12-08","paper":"/paper/increasing-pedestrian-detection-performance","paper_url":"https://dl.acm.org/doi/pdf/10.1145/3568160.3570225","paper_title":"Increasing pedestrian detection performance through weighting of detection impairing factors","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":2,"model":"LSFM (Additional Data)","metrics":{"Heavy MR^-2":"24.73","Reasonable MR^-2":"6.38","Small MR^-2":"7.90","Test Time":"0.18"},"uses_additional_data":true,"paper_date":"2023-01-01","paper":"/paper/localized-semantic-feature-mixers-for","paper_url":"http://openaccess.thecvf.com//content/CVPR2023/html/Khan_Localized_Semantic_Feature_Mixers_for_Efficient_Pedestrian_Detection_in_Autonomous_CVPR_2023_paper.html","paper_title":"Localized Semantic Feature Mixers for Efficient Pedestrian Detection in Autonomous Driving","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":3,"model":"Pedestron","metrics":{"Bare MR^-2":"6.2","Heavy MR^-2":"33.9","Large MR^-2":"4.3","Medium MR^-2":"3.0","Partial MR^-2":"5.7","Reasonable MR^-2":"7.5","Small MR^-2":"8.0"},"uses_additional_data":true,"paper_date":"2020-03-19","paper":"/paper/pedestrian-detection-the-elephant-in-the-room","paper_url":"https://arxiv.org/abs/2003.08799v7","paper_title":"Generalizable Pedestrian Detection: The Elephant In The Room","code":"https://github.com/hasanirtiza/Pedestron","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"F2DNet (extra data)","metrics":{"Heavy MR^-2":"26.23","Reasonable MR^-2":"7.8","Small MR^-2":"9.43","Test Time":"0.44s/img"},"uses_additional_data":true,"paper_date":"2022-03-04","paper":"/paper/f2dnet-fast-focal-detection-network-for","paper_url":"https://arxiv.org/abs/2203.02331v2","paper_title":"F2DNet: Fast Focal Detection Network for Pedestrian Detection","code":"https://github.com/hasanirtiza/Pedestron","n_code_links":2,"syntology":null},{"rank_in_archive_order":5,"model":"LSFM","metrics":{"Heavy MR^-2":"31.9","Reasonable MR^-2":"8.5","Small MR^-2":"8.8","Test Time":"0.18"},"uses_additional_data":false,"paper_date":"2023-01-01","paper":"/paper/localized-semantic-feature-mixers-for","paper_url":"http://openaccess.thecvf.com//content/CVPR2023/html/Khan_Localized_Semantic_Feature_Mixers_for_Efficient_Pedestrian_Detection_in_Autonomous_CVPR_2023_paper.html","paper_title":"Localized Semantic Feature Mixers for Efficient Pedestrian Detection in Autonomous Driving","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"F2DNet","metrics":{"Heavy MR^-2":"32.6","Reasonable MR^-2":"8.7","Small MR^-2":"11.3","Test Time":"0.44s/img"},"uses_additional_data":false,"paper_date":"2022-03-04","paper":"/paper/f2dnet-fast-focal-detection-network-for","paper_url":"https://arxiv.org/abs/2203.02331v2","paper_title":"F2DNet: Fast Focal Detection Network for Pedestrian Detection","code":"https://github.com/hasanirtiza/Pedestron","n_code_links":2,"syntology":null},{"rank_in_archive_order":7,"model":"ACSP","metrics":{"Bare MR^-2":"5.6","Heavy MR^-2":"46.3","Partial MR^-2":"8.7","Reasonable MR^-2":"9.3"},"uses_additional_data":false,"paper_date":"2020-02-20","paper":"/paper/adapted-center-and-scale-prediction-more","paper_url":"https://arxiv.org/abs/2002.09053v2","paper_title":"Adapted Center and Scale Prediction: More Stable and More Accurate","code":"https://github.com/WangWenhao0716/Adapted-Center-and-Scale-Prediction","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"VLPD","metrics":{"Bare MR^-2":"6.1","Heavy MR^-2":"43.1","Partial MR^-2":"8.8","Reasonable MR^-2":"9.4","Small MR^-2":"10.9"},"uses_additional_data":false,"paper_date":"2023-04-06","paper":"/paper/vlpd-context-aware-pedestrian-detection-via","paper_url":"https://arxiv.org/abs/2304.03135v1","paper_title":"VLPD: Context-Aware Pedestrian Detection via Vision-Language Semantic Self-Supervision","code":"https://github.com/lmy98129/vlpd","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":3,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":9,"model":"SOLIDER","metrics":{"Heavy MR^-2":"39.4","Reasonable MR^-2":"9.7"},"uses_additional_data":true,"paper_date":"2023-03-30","paper":"/paper/beyond-appearance-a-semantic-controllable","paper_url":"https://arxiv.org/abs/2303.17602v1","paper_title":"Beyond Appearance: a Semantic Controllable Self-Supervised Learning Framework for Human-Centric Visual Tasks","code":"https://github.com/modelscope/modelscope","n_code_links":4,"syntology":{"n_ran":1,"n_unverified":1,"n_samples":2,"n_pointer_only_licence":1}},{"rank_in_archive_order":10,"model":"NMS-Loss","metrics":{"Reasonable MR^-2":"10.08"},"uses_additional_data":false,"paper_date":"2021-06-04","paper":"/paper/nms-loss-learning-with-non-maximum","paper_url":"https://arxiv.org/abs/2106.02426v1","paper_title":"NMS-Loss: Learning with Non-Maximum Suppression for Crowded Pedestrian Detection","code":"https://github.com/IcebergKnight/NMS-Loss","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"Beta R-CNN","metrics":{"Bare MR^-2":"6.4","Heavy MR^-2":"47.1","Partial MR^-2":"10.3","Reasonable MR^-2":"10.6"},"uses_additional_data":false,"paper_date":"2022-10-23","paper":"/paper/beta-r-cnn-looking-into-pedestrian-detection-1","paper_url":"https://arxiv.org/abs/2210.12758v1","paper_title":"Beta R-CNN: Looking into Pedestrian Detection from Another Perspective","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":12,"model":"FRCNN+FPN-Res50+refined feature map+Crowdhuman","metrics":{"Reasonable MR^-2":"10.67"},"uses_additional_data":true,"paper_date":"2018-04-30","paper":"/paper/crowdhuman-a-benchmark-for-detecting-human-in","paper_url":"http://arxiv.org/abs/1805.00123v1","paper_title":"CrowdHuman: A Benchmark for Detecting Human in a Crowd","code":"https://github.com/aibeedetect/bfjdet","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"NOH-NMS","metrics":{"Bare MR^-2":"6.6","Heavy MR^-2":"53.0","Partial MR^-2":"11.2","Reasonable MR^-2":"10.8"},"uses_additional_data":false,"paper_date":"2020-07-27","paper":"/paper/noh-nms-improving-pedestrian-detection-by","paper_url":"https://arxiv.org/abs/2007.13376v1","paper_title":"NOH-NMS: Improving Pedestrian Detection by Nearby Objects Hallucination","code":"https://github.com/TencentYoutuResearch/PedestrianDetection-NohNMS","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"CSP (with offset) + ResNet-50","metrics":{"Bare MR^-2":"7.3","Heavy MR^-2":"49.3","Large MR^-2":"6.5","Medium MR^-2":"3.7","Partial MR^-2":"10.4","Reasonable MR^-2":"11.0","Small MR^-2":"16.0","Test Time":"0.33s/img"},"uses_additional_data":false,"paper_date":"2019-04-05","paper":"/paper/high-level-semantic-feature-detectiona-new","paper_url":"https://arxiv.org/abs/1904.02948v4","paper_title":"Center and Scale Prediction: Anchor-free Approach for Pedestrian and Face Detection","code":"https://github.com/liuwei16/CSP","n_code_links":2,"syntology":null},{"rank_in_archive_order":15,"model":"ALFNet","metrics":{"Bare MR^-2":"8.4","Heavy MR^-2":"51.9","Large MR^-2":"6.6","Medium MR^-2":"5.7","Partial MR^-2":"11.4","Reasonable MR^-2":"12.0","Small MR^-2":"19.0","Test Time":"0.27"},"uses_additional_data":false,"paper_date":"2018-09-01","paper":"/paper/learning-efficient-single-stage-pedestrian","paper_url":"http://openaccess.thecvf.com/content_ECCV_2018/html/Wei_Liu_Learning_Efficient_Single-stage_ECCV_2018_paper.html","paper_title":"Learning Efficient Single-stage Pedestrian Detectors by Asymptotic Localization Fitting","code":"https://github.com/VideoObjectSearch/ALFNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"OR-CNN","metrics":{"Bare MR^-2":"6.7","Heavy MR^-2":"55.7","Partial MR^-2":"15.3","Reasonable MR^-2":"12.8"},"uses_additional_data":false,"paper_date":"2018-07-23","paper":"/paper/occlusion-aware-r-cnn-detecting-pedestrians","paper_url":"http://arxiv.org/abs/1807.08407v1","paper_title":"Occlusion-aware R-CNN: Detecting Pedestrians in a Crowd","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":17,"model":"RepLoss","metrics":{"Bare MR^-2":"7.6","Heavy MR^-2":"56.9","Partial MR^-2":"16.8","Reasonable MR^-2":"13.2"},"uses_additional_data":false,"paper_date":"2017-11-21","paper":"/paper/repulsion-loss-detecting-pedestrians-in-a","paper_url":"http://arxiv.org/abs/1711.07752v2","paper_title":"Repulsion Loss: Detecting Pedestrians in a Crowd","code":"https://github.com/bailvwangzi/repulsion_loss_ssd","n_code_links":2,"syntology":null},{"rank_in_archive_order":18,"model":"TLL+MRF","metrics":{"Bare MR^-2":"9.2","Heavy MR^-2":"52.0","Partial MR^-2":"15.9","Reasonable MR^-2":"14.4"},"uses_additional_data":false,"paper_date":"2018-07-04","paper":"/paper/small-scale-pedestrian-detection-based-on","paper_url":"http://arxiv.org/abs/1807.01438v1","paper_title":"Small-scale Pedestrian Detection Based on Somatic Topology Localization and Temporal Feature Aggregation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":19,"model":"FRCNN+Seg","metrics":{"Large MR^-2":"8.0","Medium MR^-2":"6.7","Reasonable MR^-2":"14.8","Small MR^-2":"22.6"},"uses_additional_data":false,"paper_date":"2017-02-19","paper":"/paper/citypersons-a-diverse-dataset-for-pedestrian","paper_url":"http://arxiv.org/abs/1702.05693v1","paper_title":"CityPersons: A Diverse Dataset for Pedestrian Detection","code":"https://github.com/aibeedetect/bfjdet","n_code_links":2,"syntology":null},{"rank_in_archive_order":20,"model":"FRCNN","metrics":{"Large MR^-2":"7.9","Medium MR^-2":"7.2","Reasonable MR^-2":"15.4","Small MR^-2":"25.6"},"uses_additional_data":false,"paper_date":"2017-02-19","paper":"/paper/citypersons-a-diverse-dataset-for-pedestrian","paper_url":"http://arxiv.org/abs/1702.05693v1","paper_title":"CityPersons: A Diverse Dataset for Pedestrian Detection","code":"https://github.com/aibeedetect/bfjdet","n_code_links":2,"syntology":null},{"rank_in_archive_order":21,"model":"TLL","metrics":{"Bare MR^-2":"10.0","Heavy MR^-2":"53.6","Partial MR^-2":"17.2","Reasonable MR^-2":"15.5"},"uses_additional_data":false,"paper_date":"2018-07-04","paper":"/paper/small-scale-pedestrian-detection-based-on","paper_url":"http://arxiv.org/abs/1807.01438v1","paper_title":"Small-scale Pedestrian Detection Based on Somatic Topology Localization and Temporal Feature Aggregation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":22,"model":"ACSP + EuroCity Persons","metrics":{"Bare MR^-2":"4.9","Heavy MR^-2":"42.5","Partial MR^-2":"6.9"},"uses_additional_data":true,"paper_date":"2020-02-20","paper":"/paper/adapted-center-and-scale-prediction-more","paper_url":"https://arxiv.org/abs/2002.09053v2","paper_title":"Adapted Center and Scale Prediction: More Stable and More Accurate","code":"https://github.com/WangWenhao0716/Adapted-Center-and-Scale-Prediction","n_code_links":1,"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":2,"rows_with_any_sample_ran":2,"distinct_papers_with_graph_line":2,"distinct_papers_with_any_sample_ran":2,"samples_over_distinct_papers":{"n_ran":2,"n_unverified":4,"n_samples":6,"n_pointer_only_licence":1,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":2,"n_unverified":4,"n_samples":6,"n_pointer_only_licence":1,"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"}}}