{"url":"/sota/human-object-interaction-detection-on-v-coco","task":{"name":"Human-Object Interaction Detection","url":"/task/human-object-interaction-detection","note":null},"dataset":{"name":"V-COCO","url":"/dataset/v-coco"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"Human-Object Interaction (HOI) detection is a task of identifying \"a set of interactions\" in an image, which involves the i) localization of the subject (i.e., humans) and target (i.e., objects) of interaction, and ii) the classification of the interaction labels.","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":["AP(S1)","AP(S2)","Time Per Frame(ms)","MAP"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"AP(S1)":"higher","AP(S2)":"higher","Time Per Frame(ms)":"lower","MAP":"higher"}},"counts":{"rows":34,"rows_with_code":34,"rows_with_paper_page":34,"rows_dated":33,"rows_using_additional_data":1},"rows":[{"rank_in_archive_order":1,"model":"RLIPv2","metrics":{"AP(S1)":"72.1","AP(S2)":"74.1"},"uses_additional_data":false,"paper_date":"2023-08-18","paper":"/paper/rlipv2-fast-scaling-of-relational-language","paper_url":"https://arxiv.org/abs/2308.09351v1","paper_title":"RLIPv2: Fast Scaling of Relational Language-Image Pre-training","code":"https://github.com/jacobyuan7/rlipv2","n_code_links":3,"syntology":{"n_ran":22,"n_unverified":8,"n_samples":30,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"MUREN","metrics":{"AP(S1)":"68.8","AP(S2)":"71.0"},"uses_additional_data":false,"paper_date":"2023-04-11","paper":"/paper/relational-context-learning-for-human-object","paper_url":"https://arxiv.org/abs/2304.04997v1","paper_title":"Relational Context Learning for Human-Object Interaction Detection","code":"https://github.com/OreoChocolate/MUREN","n_code_links":1,"syntology":{"n_ran":10,"n_unverified":2,"n_samples":12,"n_pointer_only_licence":12}},{"rank_in_archive_order":3,"model":"STIP","metrics":{"AP(S1)":"66.0","AP(S2)":"70.7","Time Per Frame(ms)":"74"},"uses_additional_data":false,"paper_date":"2022-06-13","paper":"/paper/exploring-structure-aware-transformer-over-1","paper_url":"https://arxiv.org/abs/2206.06291v1","paper_title":"Exploring Structure-aware Transformer over Interaction Proposals for Human-Object Interaction Detection","code":"https://github.com/zyong812/stip","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"DiffHOI","metrics":{"AP(S1)":"65.7","AP(S2)":"68.2"},"uses_additional_data":false,"paper_date":"2023-05-20","paper":"/paper/boosting-human-object-interaction-detection","paper_url":"https://arxiv.org/abs/2305.12252v1","paper_title":"Boosting Human-Object Interaction Detection with Text-to-Image Diffusion Model","code":"https://github.com/IDEA-Research/DiffHOI","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"OCN  (ResNet101)","metrics":{"AP(S1)":"65.3","AP(S2)":"67.1"},"uses_additional_data":false,"paper_date":"2022-02-01","paper":"/paper/detecting-human-object-interactions-with-2","paper_url":"https://arxiv.org/abs/2202.00259v1","paper_title":"Detecting Human-Object Interactions with Object-Guided Cross-Modal Calibrated Semantics","code":"https://github.com/jacobyuan7/ocn-hoi-benchmark","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":6,"model":"OCN  (ResNet50)","metrics":{"AP(S1)":"64.2","AP(S2)":"66.3","Time Per Frame(ms)":"43"},"uses_additional_data":false,"paper_date":"2022-02-01","paper":"/paper/detecting-human-object-interactions-with-2","paper_url":"https://arxiv.org/abs/2202.00259v1","paper_title":"Detecting Human-Object Interactions with Object-Guided Cross-Modal Calibrated Semantics","code":"https://github.com/jacobyuan7/ocn-hoi-benchmark","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"CDN (ResNet101)","metrics":{"AP(S1)":"63.91","AP(S2)":"65.89"},"uses_additional_data":false,"paper_date":"2021-08-11","paper":"/paper/mining-the-benefits-of-two-stage-and-one","paper_url":"https://arxiv.org/abs/2108.05077v2","paper_title":"Mining the Benefits of Two-stage and One-stage HOI Detection","code":"https://github.com/YueLiao/CDN","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":1,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"HOICLIP","metrics":{"AP(S1)":"63.50","AP(S2)":"64.81"},"uses_additional_data":false,"paper_date":"2023-03-28","paper":"/paper/hoiclip-efficient-knowledge-transfer-for-hoi","paper_url":"https://arxiv.org/abs/2303.15786v3","paper_title":"HOICLIP: Efficient Knowledge Transfer for HOI Detection with Vision-Language Models","code":"https://github.com/artanic30/hoiclip","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"Body Part Interactiveness","metrics":{"AP(S1)":"63.0","AP(S2)":"65.1"},"uses_additional_data":false,"paper_date":"2022-07-28","paper":"/paper/mining-cross-person-cues-for-body-part","paper_url":"https://arxiv.org/abs/2207.14192v2","paper_title":"Mining Cross-Person Cues for Body-Part Interactiveness Learning in HOI Detection","code":"https://github.com/enlighten0707/body-part-map-for-interactiveness","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"UPT-R101-DC5","metrics":{"AP(S1)":"61.3","AP(S2)":"67.1","Time Per Frame(ms)":"131"},"uses_additional_data":false,"paper_date":"2021-12-03","paper":"/paper/efficient-two-stage-detection-of-human-object","paper_url":"https://arxiv.org/abs/2112.01838v2","paper_title":"Efficient Two-Stage Detection of Human-Object Interactions with a Novel Unary-Pairwise Transformer","code":"https://github.com/fredzzhang/upt","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":2,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":11,"model":"UPT-R101","metrics":{"AP(S1)":"60.7","AP(S2)":"66.2","Time Per Frame(ms)":"64"},"uses_additional_data":false,"paper_date":"2021-12-03","paper":"/paper/efficient-two-stage-detection-of-human-object","paper_url":"https://arxiv.org/abs/2112.01838v2","paper_title":"Efficient Two-Stage Detection of Human-Object Interactions with a Novel Unary-Pairwise Transformer","code":"https://github.com/fredzzhang/upt","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":2,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":12,"model":"UPT-R50","metrics":{"AP(S1)":"59.0","AP(S2)":"64.5","Time Per Frame(ms)":"43"},"uses_additional_data":false,"paper_date":"2021-12-03","paper":"/paper/efficient-two-stage-detection-of-human-object","paper_url":"https://arxiv.org/abs/2112.01838v2","paper_title":"Efficient Two-Stage Detection of Human-Object Interactions with a Novel Unary-Pairwise Transformer","code":"https://github.com/fredzzhang/upt","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":2,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":13,"model":"QPIC (ResNet50)","metrics":{"AP(S1)":"58.8","AP(S2)":"61.0","Time Per Frame(ms)":"46"},"uses_additional_data":false,"paper_date":"2021-03-09","paper":"/paper/qpic-query-based-pairwise-human-object","paper_url":"https://arxiv.org/abs/2103.05399v1","paper_title":"QPIC: Query-Based Pairwise Human-Object Interaction Detection with Image-Wide Contextual Information","code":"https://github.com/hitachi-rd-cv/qpic","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":14,"model":"QPIC (ResNet101)","metrics":{"AP(S1)":"58.3","AP(S2)":"60.7","Time Per Frame(ms)":"63"},"uses_additional_data":false,"paper_date":"2021-03-09","paper":"/paper/qpic-query-based-pairwise-human-object","paper_url":"https://arxiv.org/abs/2103.05399v1","paper_title":"QPIC: Query-Based Pairwise Human-Object Interaction Detection with Image-Wide Contextual Information","code":"https://github.com/hitachi-rd-cv/qpic","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":15,"model":"DIRV","metrics":{"AP(S1)":"56.1","Time Per Frame(ms)":"68"},"uses_additional_data":false,"paper_date":"2020-10-02","paper":"/paper/dirv-dense-interaction-region-voting-for-end","paper_url":"https://arxiv.org/abs/2010.01005v2","paper_title":"DIRV: Dense Interaction Region Voting for End-to-End Human-Object Interaction Detection","code":"https://github.com/MVIG-SJTU/DIRV","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":16,"model":"HOTR","metrics":{"AP(S1)":"55.2","AP(S2)":"64.4"},"uses_additional_data":false,"paper_date":"2021-04-28","paper":"/paper/hotr-end-to-end-human-object-interaction","paper_url":"https://arxiv.org/abs/2104.13682v1","paper_title":"HOTR: End-to-End Human-Object Interaction Detection with Transformers","code":"https://github.com/kakaobrain/HOTR","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":17,"model":"GGNet","metrics":{"AP(S1)":"54.7"},"uses_additional_data":false,"paper_date":"2021-04-12","paper":"/paper/glance-and-gaze-inferring-action-aware-points","paper_url":"https://arxiv.org/abs/2104.05269v1","paper_title":"Glance and Gaze: Inferring Action-aware Points for One-Stage Human-Object Interaction Detection","code":"https://github.com/SherlockHolmes221/GGNet","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":5,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":18,"model":"SCG","metrics":{"AP(S1)":"54.2","AP(S2)":"60.9","Time Per Frame(ms)":"500"},"uses_additional_data":true,"paper_date":"2020-12-11","paper":"/paper/spatio-attentive-graphs-for-human-object","paper_url":"https://arxiv.org/abs/2012.06060v3","paper_title":"Spatially Conditioned Graphs for Detecting Human-Object Interactions","code":"https://github.com/fredzzhang/spatially-conditioned-graphs","n_code_links":2,"syntology":null},{"rank_in_archive_order":19,"model":"PD-Net","metrics":{"AP(S1)":"53.34"},"uses_additional_data":false,"paper_date":"2020-08-07","paper":"/paper/polysemy-deciphering-network-for-robust-human","paper_url":"https://arxiv.org/abs/2008.02918v3","paper_title":"Polysemy Deciphering Network for Robust Human-Object Interaction Detection","code":"https://github.com/MuchHair/PD-Net","n_code_links":2,"syntology":null},{"rank_in_archive_order":20,"model":"IDN","metrics":{"AP(S1)":"53.3","AP(S2)":"60.3"},"uses_additional_data":false,"paper_date":"2020-10-30","paper":"/paper/hoi-analysis-integrating-and-decomposing","paper_url":"https://arxiv.org/abs/2010.16219v2","paper_title":"HOI Analysis: Integrating and Decomposing Human-Object Interaction","code":"https://github.com/DirtyHarryLYL/HAKE-Action-Torch","n_code_links":2,"syntology":null},{"rank_in_archive_order":21,"model":"ACP","metrics":{"AP(S1)":"53.23"},"uses_additional_data":false,"paper_date":null,"paper":"/paper/detecting-human-object-interactions-with-1","paper_url":"https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/3850_ECCV_2020_paper.php","paper_title":"Detecting Human-Object Interactions with Action Co-occurrence Priors","code":"https://github.com/Dong-JinKim/ActionCooccurrencePriors","n_code_links":1,"syntology":null},{"rank_in_archive_order":22,"model":"SGCN4HOI","metrics":{"AP(S1)":"53.1","AP(S2)":"57.9"},"uses_additional_data":false,"paper_date":"2022-07-11","paper":"/paper/a-skeleton-aware-graph-convolutional-network","paper_url":"https://arxiv.org/abs/2207.05733v1","paper_title":"A Skeleton-aware Graph Convolutional Network for Human-Object Interaction Detection","code":"https://github.com/zhumanli/sgcn4hoi","n_code_links":1,"syntology":null},{"rank_in_archive_order":23,"model":"PMN","metrics":{"AP(S1)":"51.8"},"uses_additional_data":false,"paper_date":"2020-08-05","paper":"/paper/pose-based-modular-network-for-human-object","paper_url":"https://arxiv.org/abs/2008.02042v1","paper_title":"Pose-based Modular Network for Human-Object Interaction Detection","code":"https://github.com/birlrobotics/PMN","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"VSGNet","metrics":{"AP(S1)":"51.76","AP(S2)":"57.0","Time Per Frame(ms)":"312"},"uses_additional_data":false,"paper_date":"2020-03-11","paper":"/paper/vsgnet-spatial-attention-network-for","paper_url":"https://arxiv.org/abs/2003.05541v1","paper_title":"VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions","code":"https://github.com/ASMIftekhar/VSGNet","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":8,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":25,"model":"PPDM","metrics":{"AP(S1)":"51.1","Time Per Frame(ms)":"71"},"uses_additional_data":false,"paper_date":"2019-12-30","paper":"/paper/ppdm-parallel-point-detection-and-matching","paper_url":"https://arxiv.org/abs/1912.12898v3","paper_title":"PPDM: Parallel Point Detection and Matching for Real-time Human-Object Interaction Detection","code":"https://github.com/YueLiao/PPDM","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":26,"model":"PaStaNet","metrics":{"AP(S1)":"51.0","AP(S2)":"57.5"},"uses_additional_data":false,"paper_date":"2020-04-02","paper":"/paper/pastanet-toward-human-activity-knowledge","paper_url":"https://arxiv.org/abs/2004.00945v2","paper_title":"PaStaNet: Toward Human Activity Knowledge Engine","code":"https://github.com/DirtyHarryLYL/HAKE-Action-Torch","n_code_links":2,"syntology":null},{"rank_in_archive_order":27,"model":"DRG","metrics":{"AP(S1)":"51.0"},"uses_additional_data":false,"paper_date":"2020-08-26","paper":"/paper/drg-dual-relation-graph-for-human-object","paper_url":"https://arxiv.org/abs/2008.11714v1","paper_title":"DRG: Dual Relation Graph for Human-Object Interaction Detection","code":"https://github.com/vt-vl-lab/DRG","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":1,"n_samples":3,"n_pointer_only_licence":2}},{"rank_in_archive_order":28,"model":"TIN (TIPAMI)","metrics":{"AP(S1)":"49.1"},"uses_additional_data":false,"paper_date":"2021-01-25","paper":"/paper/transferable-interactiveness-knowledge-for","paper_url":"https://arxiv.org/abs/2101.10292v3","paper_title":"Transferable Interactiveness Knowledge for Human-Object Interaction Detection","code":"https://github.com/DirtyHarryLYL/Transferable-Interactiveness-Network","n_code_links":1,"syntology":null},{"rank_in_archive_order":29,"model":"TIN (Interactiveness)","metrics":{"AP(S1)":"49.0","Time Per Frame(ms)":"513"},"uses_additional_data":false,"paper_date":"2018-11-20","paper":"/paper/transferable-interactiveness-prior-for-human","paper_url":"https://arxiv.org/abs/1811.08264v4","paper_title":"Transferable Interactiveness Knowledge for Human-Object Interaction Detection","code":"https://github.com/DirtyHarryLYL/Transferable-Interactiveness-Network","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":6,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":30,"model":"TIN (CVPR)","metrics":{"AP(S1)":"48.7"},"uses_additional_data":false,"paper_date":"2018-11-20","paper":"/paper/transferable-interactiveness-prior-for-human","paper_url":"https://arxiv.org/abs/1811.08264v4","paper_title":"Transferable Interactiveness Knowledge for Human-Object Interaction Detection","code":"https://github.com/DirtyHarryLYL/Transferable-Interactiveness-Network","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":6,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":31,"model":"iCAN","metrics":{"AP(S1)":"44.7"},"uses_additional_data":false,"paper_date":"2018-08-30","paper":"/paper/ican-instance-centric-attention-network-for","paper_url":"http://arxiv.org/abs/1808.10437v1","paper_title":"iCAN: Instance-Centric Attention Network for Human-Object Interaction Detection","code":"https://github.com/vt-vl-lab/iCAN","n_code_links":4,"syntology":{"n_ran":0,"n_unverified":5,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":32,"model":"GPNN","metrics":{"AP(S1)":"44.0"},"uses_additional_data":false,"paper_date":"2018-08-23","paper":"/paper/learning-human-object-interactions-by-graph","paper_url":"http://arxiv.org/abs/1808.07962v1","paper_title":"Learning Human-Object Interactions by Graph Parsing Neural Networks","code":"https://github.com/SiyuanQi/gpnn","n_code_links":1,"syntology":null},{"rank_in_archive_order":33,"model":"QPIC + CPC","metrics":{"MAP":"63.1"},"uses_additional_data":false,"paper_date":"2022-04-11","paper":"/paper/consistency-learning-via-decoding-path","paper_url":"https://arxiv.org/abs/2204.04836v1","paper_title":"Consistency Learning via Decoding Path Augmentation for Transformers in Human Object Interaction Detection","code":"https://github.com/mlvlab/cpchoi","n_code_links":1,"syntology":null},{"rank_in_archive_order":34,"model":"HOTR + CPC","metrics":{"MAP":"61.6"},"uses_additional_data":false,"paper_date":"2022-04-11","paper":"/paper/consistency-learning-via-decoding-path","paper_url":"https://arxiv.org/abs/2204.04836v1","paper_title":"Consistency Learning via Decoding Path Augmentation for Transformers in Human Object Interaction Detection","code":"https://github.com/mlvlab/cpchoi","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":20,"rows_with_any_sample_ran":13,"distinct_papers_with_graph_line":15,"distinct_papers_with_any_sample_ran":10,"samples_over_distinct_papers":{"n_ran":50,"n_unverified":46,"n_samples":96,"n_pointer_only_licence":17,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":56,"n_unverified":62,"n_samples":118,"n_pointer_only_licence":17,"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"}}}