{"url":"/sota/open-vocabulary-object-detection-on-lvis-v1-0","task":{"name":"Open Vocabulary Object Detection","url":"/task/open-vocabulary-object-detection","note":null},"dataset":{"name":"LVIS v1.0","url":"/dataset/lvis"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"Open-vocabulary detection (OVD) aims to generalize beyond the limited number of base classes labeled during the training phase. The goal is to detect novel classes defined by an unbounded\r\n(open) vocabulary at inference.","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 novel-LVIS base training","AP novel-Unrestricted open-vocabulary training"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"AP novel-LVIS base training":"higher","AP novel-Unrestricted open-vocabulary training":"higher"}},"counts":{"rows":28,"rows_with_code":27,"rows_with_paper_page":28,"rows_dated":28,"rows_using_additional_data":8},"rows":[{"rank_in_archive_order":1,"model":"LaMI-DETR","metrics":{"AP novel-LVIS base training":"43.4"},"uses_additional_data":false,"paper_date":"2024-07-16","paper":"/paper/lami-detr-open-vocabulary-detection-with","paper_url":"https://arxiv.org/abs/2407.11335v2","paper_title":"LaMI-DETR: Open-Vocabulary Detection with Language Model Instruction","code":"https://github.com/eternaldolphin/lami-detr","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":1,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"DITO","metrics":{"AP novel-LVIS base training":"40.4","AP novel-Unrestricted open-vocabulary training":"45.8"},"uses_additional_data":false,"paper_date":"2023-09-29","paper":"/paper/detection-oriented-image-text-pretraining-for","paper_url":"https://arxiv.org/abs/2310.00161v2","paper_title":"Region-centric Image-Language Pretraining for Open-Vocabulary Detection","code":"https://github.com/google-research/google-research/tree/master/fvlm/dito","n_code_links":2,"syntology":null},{"rank_in_archive_order":3,"model":"OV-DQUO(ViT-L/14)","metrics":{"AP novel-LVIS base training":"39.3"},"uses_additional_data":false,"paper_date":"2024-05-28","paper":"/paper/ov-dquo-open-vocabulary-detr-with-denoising","paper_url":"https://arxiv.org/abs/2405.17913v2","paper_title":"OV-DQUO: Open-Vocabulary DETR with Denoising Text Query Training and Open-World Unknown Objects Supervision","code":"https://github.com/xiaomoguhz/ov-dquo","n_code_links":1,"syntology":{"n_ran":9,"n_unverified":2,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"CoDet (EVA02-L)","metrics":{"AP novel-LVIS base training":"37.0"},"uses_additional_data":true,"paper_date":"2023-10-25","paper":"/paper/codet-co-occurrence-guided-region-word-1","paper_url":"https://arxiv.org/abs/2310.16667v1","paper_title":"CoDet: Co-Occurrence Guided Region-Word Alignment for Open-Vocabulary Object Detection","code":"https://github.com/cvmi-lab/codet","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":2,"n_samples":8,"n_pointer_only_licence":8}},{"rank_in_archive_order":5,"model":"CLIPSelf","metrics":{"AP novel-LVIS base training":"34.9"},"uses_additional_data":false,"paper_date":"2023-10-02","paper":"/paper/clipself-vision-transformer-distills-itself","paper_url":"https://arxiv.org/abs/2310.01403v2","paper_title":"CLIPSelf: Vision Transformer Distills Itself for Open-Vocabulary Dense Prediction","code":"https://github.com/wusize/clipself","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":4,"n_samples":6,"n_pointer_only_licence":6}},{"rank_in_archive_order":6,"model":"OVMR","metrics":{"AP novel-LVIS base training":"34.4"},"uses_additional_data":true,"paper_date":"2024-06-07","paper":"/paper/ovmr-open-vocabulary-recognition-with-multi","paper_url":"https://arxiv.org/abs/2406.04675v1","paper_title":"OVMR: Open-Vocabulary Recognition with Multi-Modal References","code":"https://github.com/zehong-ma/ovmr","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":3,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"DE-ViT","metrics":{"AP novel-LVIS base training":"34.3"},"uses_additional_data":false,"paper_date":"2023-09-22","paper":"/paper/detect-every-thing-with-few-examples","paper_url":"https://arxiv.org/abs/2309.12969v4","paper_title":"Detect Everything with Few Examples","code":"https://github.com/mlzxy/devit","n_code_links":1,"syntology":{"n_ran":15,"n_unverified":5,"n_samples":20,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"CFM-ViT","metrics":{"AP novel-LVIS base training":"33.9"},"uses_additional_data":false,"paper_date":"2023-09-02","paper":"/paper/contrastive-feature-masking-open-vocabulary","paper_url":"https://arxiv.org/abs/2309.00775v1","paper_title":"Contrastive Feature Masking Open-Vocabulary Vision Transformer","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":9,"model":"CLIM (RN50x64)","metrics":{"AP novel-LVIS base training":"32.3"},"uses_additional_data":false,"paper_date":"2023-12-18","paper":"/paper/clim-contrastive-language-image-mosaic-for","paper_url":"https://arxiv.org/abs/2312.11376v2","paper_title":"CLIM: Contrastive Language-Image Mosaic for Region Representation","code":"https://github.com/wusize/clim","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":2,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":10,"model":"RO-ViT","metrics":{"AP novel-LVIS base training":"32.1"},"uses_additional_data":false,"paper_date":"2023-05-11","paper":"/paper/region-aware-pretraining-for-open-vocabulary","paper_url":"https://arxiv.org/abs/2305.07011v4","paper_title":"Region-Aware Pretraining for Open-Vocabulary Object Detection with Vision Transformers","code":"https://github.com/google-research/google-research/tree/master/fvlm/rovit","n_code_links":2,"syntology":{"n_ran":5,"n_unverified":3,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":11,"model":"Prova (Swin-Base)","metrics":{"AP novel-LVIS base training":"31.5"},"uses_additional_data":true,"paper_date":"2024-12-23","paper":"/paper/comprehensive-multi-modal-prototypes-are","paper_url":"https://arxiv.org/abs/2412.17800v1","paper_title":"Comprehensive Multi-Modal Prototypes are Simple and Effective Classifiers for Vast-Vocabulary Object Detection","code":"https://github.com/row11n/prova","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"RTGen","metrics":{"AP novel-LVIS base training":"30.2"},"uses_additional_data":true,"paper_date":"2024-05-30","paper":"/paper/rtgen-generating-region-text-pairs-for-open","paper_url":"https://arxiv.org/abs/2405.19854v1","paper_title":"RTGen: Generating Region-Text Pairs for Open-Vocabulary Object Detection","code":"https://github.com/seermer/RTGen","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"OV-DQUO(ViT-B/16)","metrics":{"AP novel-LVIS base training":"29.7"},"uses_additional_data":false,"paper_date":"2024-05-28","paper":"/paper/ov-dquo-open-vocabulary-detr-with-denoising","paper_url":"https://arxiv.org/abs/2405.17913v2","paper_title":"OV-DQUO: Open-Vocabulary DETR with Denoising Text Query Training and Open-World Unknown Objects Supervision","code":"https://github.com/xiaomoguhz/ov-dquo","n_code_links":1,"syntology":{"n_ran":9,"n_unverified":2,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":14,"model":"ViLD-ensemble w/ ALIGN (Eb7-FPN)","metrics":{"AP novel-LVIS base training":"26.3","AP novel-Unrestricted open-vocabulary training":"27.0"},"uses_additional_data":false,"paper_date":"2021-04-28","paper":"/paper/zero-shot-detection-via-vision-and-language","paper_url":"https://arxiv.org/abs/2104.13921v3","paper_title":"Open-vocabulary Object Detection via Vision and Language Knowledge Distillation","code":"https://github.com/tensorflow/tpu","n_code_links":4,"syntology":null},{"rank_in_archive_order":15,"model":"OWL-ViT (CLIP-L/14)","metrics":{"AP novel-LVIS base training":"25.6","AP novel-Unrestricted open-vocabulary training":"31.2"},"uses_additional_data":true,"paper_date":"2022-05-12","paper":"/paper/simple-open-vocabulary-object-detection-with","paper_url":"https://arxiv.org/abs/2205.06230v2","paper_title":"Simple Open-Vocabulary Object Detection with Vision Transformers","code":"https://github.com/google-research/scenic/tree/main/scenic/projects/owl_vit","n_code_links":2,"syntology":null},{"rank_in_archive_order":16,"model":"POMP","metrics":{"AP novel-LVIS base training":"25.2"},"uses_additional_data":false,"paper_date":"2023-04-10","paper":"/paper/prompt-pre-training-with-twenty-thousand-1","paper_url":"https://arxiv.org/abs/2304.04704v2","paper_title":"Prompt Pre-Training with Twenty-Thousand Classes for Open-Vocabulary Visual Recognition","code":"https://github.com/amazon-science/prompt-pretraining","n_code_links":1,"syntology":{"n_ran":5,"n_unverified":2,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":17,"model":"BARON","metrics":{"AP novel-LVIS base training":"22.6"},"uses_additional_data":false,"paper_date":"2023-02-27","paper":"/paper/aligning-bag-of-regions-for-open-vocabulary","paper_url":"https://arxiv.org/abs/2302.13996v1","paper_title":"Aligning Bag of Regions for Open-Vocabulary Object Detection","code":"https://github.com/wusize/ovdet","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"MEDet","metrics":{"AP novel-LVIS base training":"22.4"},"uses_additional_data":false,"paper_date":"2022-06-22","paper":"/paper/open-vocabulary-object-detection-with","paper_url":"https://arxiv.org/abs/2206.11134v4","paper_title":"Open Vocabulary Object Detection with Proposal Mining and Prediction Equalization","code":"https://github.com/peixianchen/medet","n_code_links":2,"syntology":{"n_ran":7,"n_unverified":2,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":19,"model":"Region-CLIP (RN50x4-C4)","metrics":{"AP novel-LVIS base training":"22.0"},"uses_additional_data":false,"paper_date":"2021-12-16","paper":"/paper/regionclip-region-based-language-image","paper_url":"https://arxiv.org/abs/2112.09106v1","paper_title":"RegionCLIP: Region-based Language-Image Pretraining","code":"https://github.com/microsoft/regionclip","n_code_links":1,"syntology":null},{"rank_in_archive_order":20,"model":"RALF","metrics":{"AP novel-LVIS base training":"21.9"},"uses_additional_data":true,"paper_date":"2024-04-08","paper":"/paper/retrieval-augmented-open-vocabulary-object","paper_url":"https://arxiv.org/abs/2404.05687v1","paper_title":"Retrieval-Augmented Open-Vocabulary Object Detection","code":"https://github.com/mlvlab/RALF","n_code_links":1,"syntology":null},{"rank_in_archive_order":21,"model":"OADP","metrics":{"AP novel-LVIS base training":"21.7"},"uses_additional_data":false,"paper_date":"2023-03-10","paper":"/paper/object-aware-distillation-pyramid-for-open","paper_url":"https://arxiv.org/abs/2303.05892v1","paper_title":"Object-Aware Distillation Pyramid for Open-Vocabulary Object Detection","code":"https://github.com/lutingwang/oadp","n_code_links":1,"syntology":null},{"rank_in_archive_order":22,"model":"X-Paste","metrics":{"AP novel-LVIS base training":"21.4","AP novel-Unrestricted open-vocabulary training":"22.8"},"uses_additional_data":false,"paper_date":"2022-12-07","paper":"/paper/x-paste-revisit-copy-paste-at-scale-with-clip","paper_url":"https://arxiv.org/abs/2212.03863v2","paper_title":"X-Paste: Revisiting Scalable Copy-Paste for Instance Segmentation using CLIP and StableDiffusion","code":"https://github.com/aim-uofa/DiverGen","n_code_links":2,"syntology":{"n_ran":1,"n_unverified":2,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":23,"model":"Object-Centric-OVD","metrics":{"AP novel-LVIS base training":"21.1"},"uses_additional_data":true,"paper_date":"2022-07-07","paper":"/paper/bridging-the-gap-between-object-and-image","paper_url":"https://arxiv.org/abs/2207.03482v3","paper_title":"Bridging the Gap between Object and Image-level Representations for Open-Vocabulary Detection","code":"https://github.com/mmaaz60/mvits_for_class_agnostic_od","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":2,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":24,"model":"ViLD-ensemble (R152-FPN)","metrics":{"AP novel-LVIS base training":"18.7","AP novel-Unrestricted open-vocabulary training":"19.8"},"uses_additional_data":false,"paper_date":"2021-04-28","paper":"/paper/zero-shot-detection-via-vision-and-language","paper_url":"https://arxiv.org/abs/2104.13921v3","paper_title":"Open-vocabulary Object Detection via Vision and Language Knowledge Distillation","code":"https://github.com/tensorflow/tpu","n_code_links":4,"syntology":null},{"rank_in_archive_order":25,"model":"Detic","metrics":{"AP novel-LVIS base training":"17.8"},"uses_additional_data":true,"paper_date":"2022-01-07","paper":"/paper/detecting-twenty-thousand-classes-using-image","paper_url":"https://arxiv.org/abs/2201.02605v3","paper_title":"Detecting Twenty-thousand Classes using Image-level Supervision","code":"https://github.com/facebookresearch/Detic","n_code_links":1,"syntology":null},{"rank_in_archive_order":26,"model":"Region-CLIP (RN50-C4)","metrics":{"AP novel-LVIS base training":"17.1"},"uses_additional_data":false,"paper_date":"2021-12-16","paper":"/paper/regionclip-region-based-language-image","paper_url":"https://arxiv.org/abs/2112.09106v1","paper_title":"RegionCLIP: Region-based Language-Image Pretraining","code":"https://github.com/microsoft/regionclip","n_code_links":1,"syntology":null},{"rank_in_archive_order":27,"model":"ViLD-ensemble (R50-FPN)","metrics":{"AP novel-LVIS base training":"16.6","AP novel-Unrestricted open-vocabulary training":"16.7"},"uses_additional_data":false,"paper_date":"2021-04-28","paper":"/paper/zero-shot-detection-via-vision-and-language","paper_url":"https://arxiv.org/abs/2104.13921v3","paper_title":"Open-vocabulary Object Detection via Vision and Language Knowledge Distillation","code":"https://github.com/tensorflow/tpu","n_code_links":4,"syntology":null},{"rank_in_archive_order":28,"model":"ViLD (R50-FPN)","metrics":{"AP novel-LVIS base training":"16.1","AP novel-Unrestricted open-vocabulary training":"16.3"},"uses_additional_data":false,"paper_date":"2021-04-28","paper":"/paper/zero-shot-detection-via-vision-and-language","paper_url":"https://arxiv.org/abs/2104.13921v3","paper_title":"Open-vocabulary Object Detection via Vision and Language Knowledge Distillation","code":"https://github.com/tensorflow/tpu","n_code_links":4,"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 6,795 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":6795,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2785},"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":13,"rows_with_any_sample_ran":13,"distinct_papers_with_graph_line":12,"distinct_papers_with_any_sample_ran":12,"samples_over_distinct_papers":{"n_ran":60,"n_unverified":30,"n_samples":90,"n_pointer_only_licence":20,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":69,"n_unverified":32,"n_samples":101,"n_pointer_only_licence":20,"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"}}}