{"url":"/sota/video-object-detection-on-imagenet-vid","task":{"name":"Video Object Detection","url":"/task/video-object-detection","note":null},"dataset":{"name":"ImageNet VID","url":"/dataset/imagenet-vid"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"Video object detection is the task of detecting objects from a video as opposed to images.\r\n\r\n<span style=\"color:grey; opacity: 0.6\">( Image credit: [Learning Motion Priors for Efficient Video Object Detection](https://arxiv.org/pdf/1911.05253v1.pdf) )</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":["MAP "],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"MAP ":"higher"}},"counts":{"rows":33,"rows_with_code":26,"rows_with_paper_page":33,"rows_dated":31,"rows_using_additional_data":3},"rows":[{"rank_in_archive_order":1,"model":"YOLOV++","metrics":{"MAP ":"93.2"},"uses_additional_data":true,"paper_date":"2024-07-29","paper":"/paper/practical-video-object-detection-via-feature","paper_url":"https://arxiv.org/abs/2407.19650v1","paper_title":"Practical Video Object Detection via Feature Selection and Aggregation","code":"https://github.com/yuhengsss/yolov","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"DiffusionVID (Swin-B)","metrics":{"MAP ":"92.5"},"uses_additional_data":false,"paper_date":"2023-10-30","paper":"/paper/diffusionvid-denoising-object-boxes-with","paper_url":"https://doi.org/10.1109/ACCESS.2023.3328341","paper_title":"DiffusionVID: Denoising Object Boxes with Spatio-temporal Conditioning for Video Object Detection","code":"https://github.com/sdroh1027/DiffusionVID","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"Ours (Def. DETR + SwinB)","metrics":{"MAP ":"91.3"},"uses_additional_data":false,"paper_date":"2023-08-09","paper":"/paper/objects-do-not-disappear-video-object","paper_url":"https://arxiv.org/abs/2308.04770v1","paper_title":"Objects do not disappear: Video object detection by single-frame object location anticipation","code":"https://github.com/l-kid/video-object-detection-by-location-anticipation","n_code_links":2,"syntology":null},{"rank_in_archive_order":4,"model":"VSTAM","metrics":{"MAP ":"91.1"},"uses_additional_data":false,"paper_date":"2022-06-17","paper":"/paper/video-sparse-transformer-with-attention","paper_url":"https://ieeexplore.ieee.org/document/9798833","paper_title":"Video Sparse Transformer With Attention-Guided Memory for Video Object Detection","code":"https://github.com/Malik1998/VSTAM","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"TGBFormer (Swin B)","metrics":{"MAP ":"90.3"},"uses_additional_data":false,"paper_date":"2025-03-18","paper":"/paper/tgbformer-transformer-graphformer-blender","paper_url":"https://arxiv.org/abs/2503.13903v2","paper_title":"TGBFormer: Transformer-GraphFormer Blender Network for Video Object Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"TransVOD (Swin Base)","metrics":{"MAP ":"90.1"},"uses_additional_data":true,"paper_date":"2022-01-13","paper":"/paper/transvod-end-to-end-video-object-detection","paper_url":"https://arxiv.org/abs/2201.05047v4","paper_title":"TransVOD: End-to-End Video Object Detection with Spatial-Temporal Transformers","code":"https://github.com/SJTU-LuHe/TransVOD","n_code_links":3,"syntology":{"n_ran":5,"n_unverified":1,"n_samples":6,"n_pointer_only_licence":3}},{"rank_in_archive_order":7,"model":"PTSEFormer (ResNet-101)","metrics":{"MAP ":"88.1"},"uses_additional_data":false,"paper_date":"2022-09-06","paper":"/paper/ptseformer-progressive-temporal-spatial","paper_url":"https://arxiv.org/abs/2209.02242v1","paper_title":"PTSEFormer: Progressive Temporal-Spatial Enhanced TransFormer Towards Video Object Detection","code":"https://github.com/hon-wong/ptseformer","n_code_links":1,"syntology":{"n_ran":14,"n_unverified":7,"n_samples":21,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"Ours (Def. DETR + R101)","metrics":{"MAP ":"87.9"},"uses_additional_data":false,"paper_date":"2023-08-09","paper":"/paper/objects-do-not-disappear-video-object","paper_url":"https://arxiv.org/abs/2308.04770v1","paper_title":"Objects do not disappear: Video object detection by single-frame object location anticipation","code":"https://github.com/l-kid/video-object-detection-by-location-anticipation","n_code_links":2,"syntology":null},{"rank_in_archive_order":9,"model":"YOLOV","metrics":{"MAP ":"87.5"},"uses_additional_data":false,"paper_date":"2022-08-20","paper":"/paper/yolov-making-still-image-object-detectors","paper_url":"https://arxiv.org/abs/2208.09686v2","paper_title":"YOLOV: Making Still Image Object Detectors Great at Video Object Detection","code":"https://github.com/yuhengsss/yolov","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"Ours (Faster RCNN + R101)","metrics":{"MAP ":"87.2"},"uses_additional_data":false,"paper_date":"2023-08-09","paper":"/paper/objects-do-not-disappear-video-object","paper_url":"https://arxiv.org/abs/2308.04770v1","paper_title":"Objects do not disappear: Video object detection by single-frame object location anticipation","code":"https://github.com/l-kid/video-object-detection-by-location-anticipation","n_code_links":2,"syntology":null},{"rank_in_archive_order":11,"model":"DiffusionVID (ResNet-101)","metrics":{"MAP ":"87.1"},"uses_additional_data":false,"paper_date":"2023-10-30","paper":"/paper/diffusionvid-denoising-object-boxes-with","paper_url":"https://doi.org/10.1109/ACCESS.2023.3328341","paper_title":"DiffusionVID: Denoising Object Boxes with Spatio-temporal Conditioning for Video Object Detection","code":"https://github.com/sdroh1027/DiffusionVID","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"DAFA-F (ResNeXt-101)","metrics":{"MAP ":"85.9"},"uses_additional_data":false,"paper_date":"2022-09-01","paper":"/paper/dafa-diversity-aware-feature-aggregation-for","paper_url":"https://doi.org/10.1109/ACCESS.2022.3203399","paper_title":"DAFA: Diversity-Aware Feature Aggregation for Attention-Based Video Object Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":13,"model":"ClipVID","metrics":{"MAP ":"85.8"},"uses_additional_data":false,"paper_date":"2023-08-15","paper":"/paper/identity-consistent-aggregation-for-video","paper_url":"https://arxiv.org/abs/2308.07737v1","paper_title":"Identity-Consistent Aggregation for Video Object Detection","code":"https://github.com/bladewaltz1/clipvid","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"HVRNet (ResNeXt101-32x4d)","metrics":{"MAP ":"85.5"},"uses_additional_data":false,"paper_date":null,"paper":"/paper/mining-inter-video-proposal-relations-for","paper_url":"https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/3764_ECCV_2020_paper.php","paper_title":"Mining Inter-Video Proposal Relations for Video Object Detection","code":"https://github.com/youthHan/HVRNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":15,"model":"MEGA (ResNeXt101)","metrics":{"MAP ":"85.4"},"uses_additional_data":false,"paper_date":"2020-03-26","paper":"/paper/memory-enhanced-global-local-aggregation-for","paper_url":"https://arxiv.org/abs/2003.12063v1","paper_title":"Memory Enhanced Global-Local Aggregation for Video Object Detection","code":"https://github.com/Scalsol/mega.pytorch","n_code_links":2,"syntology":null},{"rank_in_archive_order":16,"model":"BoxMask(ResNeXt101)","metrics":{"MAP ":"84.8"},"uses_additional_data":false,"paper_date":"2022-10-12","paper":"/paper/boxmask-revisiting-bounding-box-supervision","paper_url":"https://arxiv.org/abs/2210.06008v1","paper_title":"BoxMask: Revisiting Bounding Box Supervision for Video Object Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":17,"model":"DAFA-F (ResNet-101)","metrics":{"MAP ":"84.5"},"uses_additional_data":false,"paper_date":"2022-09-01","paper":"/paper/dafa-diversity-aware-feature-aggregation-for","paper_url":"https://doi.org/10.1109/ACCESS.2022.3203399","paper_title":"DAFA: Diversity-Aware Feature Aggregation for Attention-Based Video Object Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":18,"model":"Temporal ROI Align (ResNeXt101)","metrics":{"MAP ":"84.3"},"uses_additional_data":false,"paper_date":"2021-09-08","paper":"/paper/temporal-roi-align-for-video-object","paper_url":"https://arxiv.org/abs/2109.03495v2","paper_title":"Temporal RoI Align for Video Object Recognition","code":"https://github.com/open-mmlab/mmtracking","n_code_links":1,"syntology":null},{"rank_in_archive_order":19,"model":"SELSA (ResNeXt-101)","metrics":{"MAP ":"84.3"},"uses_additional_data":false,"paper_date":"2019-07-15","paper":"/paper/sequence-level-semantics-aggregation-for","paper_url":"https://arxiv.org/abs/1907.06390v2","paper_title":"Sequence Level Semantics Aggregation for Video Object Detection","code":"https://github.com/open-mmlab/mmtracking","n_code_links":2,"syntology":null},{"rank_in_archive_order":20,"model":"REPP + SELSA (ResNet-101)","metrics":{"MAP ":"84.2"},"uses_additional_data":false,"paper_date":"2020-10-01","paper":"/paper/robust-and-efficient-post-processing-for","paper_url":"https://arxiv.org/abs/2009.11050","paper_title":"Robust and Efficient Post-Processing for Video Object Detection (REPP)","code":"https://github.com/AlbertoSabater/Robust-and-efficient-post-processing-for-video-object-detection","n_code_links":1,"syntology":null},{"rank_in_archive_order":21,"model":"HVRNet (ResNest101)","metrics":{"MAP ":"83.8"},"uses_additional_data":false,"paper_date":null,"paper":"/paper/mining-inter-video-proposal-relations-for","paper_url":"https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/3764_ECCV_2020_paper.php","paper_title":"Mining Inter-Video Proposal Relations for Video Object Detection","code":"https://github.com/youthHan/HVRNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":22,"model":"Tracklet-Conditioned Detection+DCNv2+FGFA","metrics":{"MAP ":"83.5"},"uses_additional_data":false,"paper_date":"2018-11-27","paper":"/paper/integrated-object-detection-and-tracking-with","paper_url":"http://arxiv.org/abs/1811.11167v1","paper_title":"Integrated Object Detection and Tracking with Tracklet-Conditioned Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":23,"model":"SELSA (ResNet-101)","metrics":{"MAP ":"82.69"},"uses_additional_data":false,"paper_date":"2019-07-15","paper":"/paper/sequence-level-semantics-aggregation-for","paper_url":"https://arxiv.org/abs/1907.06390v2","paper_title":"Sequence Level Semantics Aggregation for Video Object Detection","code":"https://github.com/open-mmlab/mmtracking","n_code_links":2,"syntology":null},{"rank_in_archive_order":24,"model":"SLTnet FPN-X101","metrics":{"MAP ":"82.4"},"uses_additional_data":false,"paper_date":"2021-04-18","paper":"/paper/short-term-anchor-linking-and-long-term-self","paper_url":"https://www.sciencedirect.com/science/article/abs/pii/S0262885621000846","paper_title":"Short-term anchor linking and long-term self-guided attention for video object detection","code":"https://github.com/daniel-cores/SLTnet","n_code_links":1,"syntology":null},{"rank_in_archive_order":25,"model":"LSTS (ResNet-101)","metrics":{"MAP ":"81.7"},"uses_additional_data":false,"paper_date":"2019-11-13","paper":"/paper/learning-motion-priors-for-efficient-video","paper_url":"https://arxiv.org/abs/1911.05253v2","paper_title":"Learning Where to Focus for Efficient Video Object Detection","code":"https://github.com/jiangzhengkai/LSTS","n_code_links":1,"syntology":null},{"rank_in_archive_order":26,"model":"BoxMask (ResNet-50)","metrics":{"MAP ":"80.7"},"uses_additional_data":false,"paper_date":"2022-10-12","paper":"/paper/boxmask-revisiting-bounding-box-supervision","paper_url":"https://arxiv.org/abs/2210.06008v1","paper_title":"BoxMask: Revisiting Bounding Box Supervision for Video Object Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":27,"model":"SparseVOD (ResNet-50)","metrics":{"MAP ":"80.3"},"uses_additional_data":false,"paper_date":"2022-10-05","paper":"/paper/spatio-temporal-learnable-proposals-for-end","paper_url":"https://arxiv.org/abs/2210.02368v2","paper_title":"Spatio-Temporal Learnable Proposals for End-to-End Video Object Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":28,"model":"REPP + FGFA","metrics":{"MAP ":"80.1"},"uses_additional_data":false,"paper_date":"2020-10-01","paper":"/paper/robust-and-efficient-post-processing-for","paper_url":"https://arxiv.org/abs/2009.11050","paper_title":"Robust and Efficient Post-Processing for Video Object Detection (REPP)","code":"https://github.com/AlbertoSabater/Robust-and-efficient-post-processing-for-video-object-detection","n_code_links":1,"syntology":null},{"rank_in_archive_order":29,"model":"FGFA + Seq-NMS","metrics":{"MAP ":"80.1"},"uses_additional_data":false,"paper_date":"2017-03-29","paper":"/paper/flow-guided-feature-aggregation-for-video","paper_url":"http://arxiv.org/abs/1703.10025v2","paper_title":"Flow-Guided Feature Aggregation for Video Object Detection","code":"https://github.com/open-mmlab/mmtracking","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":30,"model":"Online TSM","metrics":{"MAP ":"76.3"},"uses_additional_data":false,"paper_date":"2018-11-20","paper":"/paper/temporal-shift-module-for-efficient-video","paper_url":"https://arxiv.org/abs/1811.08383v3","paper_title":"TSM: Temporal Shift Module for Efficient Video Understanding","code":"https://github.com/open-mmlab/mmaction2","n_code_links":13,"syntology":{"n_ran":6,"n_unverified":10,"n_samples":16,"n_pointer_only_licence":4}},{"rank_in_archive_order":31,"model":"REPP + YOLOv3","metrics":{"MAP ":"75.1"},"uses_additional_data":false,"paper_date":"2020-10-01","paper":"/paper/robust-and-efficient-post-processing-for","paper_url":"https://arxiv.org/abs/2009.11050","paper_title":"Robust and Efficient Post-Processing for Video Object Detection (REPP)","code":"https://github.com/AlbertoSabater/Robust-and-efficient-post-processing-for-video-object-detection","n_code_links":1,"syntology":null},{"rank_in_archive_order":32,"model":"YOLOv3","metrics":{"MAP ":"68.6"},"uses_additional_data":false,"paper_date":"2020-10-01","paper":"/paper/robust-and-efficient-post-processing-for","paper_url":"https://arxiv.org/abs/2009.11050","paper_title":"Robust and Efficient Post-Processing for Video Object Detection (REPP)","code":"https://github.com/AlbertoSabater/Robust-and-efficient-post-processing-for-video-object-detection","n_code_links":1,"syntology":null},{"rank_in_archive_order":33,"model":"Looking Fast and Slow","metrics":{"MAP ":"63.9"},"uses_additional_data":true,"paper_date":"2019-03-25","paper":"/paper/looking-fast-and-slow-memory-guided-mobile","paper_url":"http://arxiv.org/abs/1903.10172v1","paper_title":"Looking Fast and Slow: Memory-Guided Mobile Video Object Detection","code":"https://github.com/tensorflow/models/tree/master/research/lstm_object_detection","n_code_links":2,"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 6,885 of the 9,623 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":9623,"papers_checked":6885,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2737},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-25T09:33:49+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":4,"rows_with_any_sample_ran":3,"distinct_papers_with_graph_line":4,"distinct_papers_with_any_sample_ran":3,"samples_over_distinct_papers":{"n_ran":25,"n_unverified":21,"n_samples":46,"n_pointer_only_licence":10,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":25,"n_unverified":21,"n_samples":46,"n_pointer_only_licence":10,"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"}}}