{"url":"/sota/dense-video-captioning-on-activitynet","task":{"name":"Dense Video Captioning","url":"/task/dense-video-captioning","note":null},"dataset":{"name":"ActivityNet Captions","url":"/dataset/activitynet-captions"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"Most natural videos contain numerous events. For example, in a video of a “man playing a piano”, the video might also contain “another man dancing” or “a crowd clapping”. The task of dense video captioning involves both detecting and describing events in a video.","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":["METEOR","BLEU-3","BLEU-4","CIDEr","SODA","DIV-1","DIV-2","RE-4","BLEU4","F1","Precision","Recall"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"METEOR":null,"BLEU-3":"higher","BLEU-4":"higher","CIDEr":null,"SODA":null,"DIV-1":null,"DIV-2":null,"RE-4":null,"BLEU4":null,"F1":"higher","Precision":"higher","Recall":"higher"}},"counts":{"rows":12,"rows_with_code":10,"rows_with_paper_page":12,"rows_dated":12,"rows_using_additional_data":1},"rows":[{"rank_in_archive_order":1,"model":"Vid2Seq","metrics":{"CIDEr":"28","METEOR":"17"},"uses_additional_data":true,"paper_date":"2023-02-27","paper":"/paper/vid2seq-large-scale-pretraining-of-a-visual","paper_url":"https://arxiv.org/abs/2302.14115v2","paper_title":"Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense Video Captioning","code":"https://github.com/google-research/scenic/tree/main/scenic/projects/vid2seq","n_code_links":3,"syntology":null},{"rank_in_archive_order":2,"model":"ADV-INF + Global","metrics":{"BLEU-4":"9.45","CIDEr":"19.40","DIV-1":"0.60","DIV-2":"0.78","METEOR":"16.36","RE-4":"0.05"},"uses_additional_data":false,"paper_date":"2021-07-18","paper":"/paper/global-object-proposals-for-improving-multi","paper_url":"https://ieeexplore.ieee.org/document/9533883","paper_title":"Global Object Proposals for Improving Multi-Sentence Video Descriptions","code":"https://github.com/cskanani/global_object_proposals","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"Bi-directional+intra captioning","metrics":{"METEOR":"11.28"},"uses_additional_data":false,"paper_date":"2020-06-14","paper":"/paper/team-ruc-aim3-technical-report-at-activitynet","paper_url":"https://arxiv.org/abs/2006.07896v1","paper_title":"Team RUC_AIM3 Technical Report at Activitynet 2020 Task 2: Exploring Sequential Events Detection for Dense Video Captioning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":4,"model":"GVL","metrics":{"CIDEr":"33.33","METEOR":"10.03","SODA":"7.11"},"uses_additional_data":false,"paper_date":"2023-03-11","paper":"/paper/learning-grounded-vision-language","paper_url":"https://arxiv.org/abs/2303.06378v2","paper_title":"Learning Grounded Vision-Language Representation for Versatile Understanding in Untrimmed Videos","code":"https://github.com/zjr2000/gvl","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":11,"n_samples":12,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"TSRM-CMG-HRNN+SCST","metrics":{"METEOR":"9.71"},"uses_additional_data":false,"paper_date":"2020-06-21","paper":"/paper/dense-captioning-events-in-videos-sysu","paper_url":"https://arxiv.org/abs/2006.11693v2","paper_title":"Dense-Captioning Events in Videos: SYSU Submission to ActivityNet Challenge 2020","code":"https://github.com/ttengwang/dense-video-captioning-pytorch","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"PDVC (TSP features, no SCST)","metrics":{"BLEU-4":"2.17","CIDEr":"31.14","METEOR":"9.03","SODA":"6.05"},"uses_additional_data":false,"paper_date":"2021-08-17","paper":"/paper/end-to-end-dense-video-captioning-with","paper_url":"https://arxiv.org/abs/2108.07781v2","paper_title":"End-to-End Dense Video Captioning with Parallel Decoding","code":"https://github.com/ttengwang/pdvc","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":4,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"TSP","metrics":{"BLEU-3":"4.16","BLEU-4":"2.02","METEOR":"8.75"},"uses_additional_data":false,"paper_date":"2020-11-23","paper":"/paper/tsp-temporally-sensitive-pretraining-of-video","paper_url":"https://arxiv.org/abs/2011.11479v3","paper_title":"TSP: Temporally-Sensitive Pretraining of Video Encoders for Localization Tasks","code":"https://github.com/HumamAlwassel/TSP","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"CM²","metrics":{"BLEU4":"2.38","CIDEr":"33.01","F1":"55.21","METEOR":"8.55","Precision":"56.81","Recall":"53.71","SODA":"6.18"},"uses_additional_data":false,"paper_date":"2024-04-11","paper":"/paper/do-you-remember-dense-video-captioning-with","paper_url":"https://arxiv.org/abs/2404.07610v1","paper_title":"Do You Remember? Dense Video Captioning with Cross-Modal Memory Retrieval","code":"https://github.com/ailab-kyunghee/cm2_dvc","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":1,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":9,"model":"BMT","metrics":{"BLEU-3":"3.84","BLEU-4":"1.88","METEOR":"8.44"},"uses_additional_data":false,"paper_date":"2020-05-17","paper":"/paper/a-better-use-of-audio-visual-cues-dense-video","paper_url":"https://arxiv.org/abs/2005.08271v2","paper_title":"A Better Use of Audio-Visual Cues: Dense Video Captioning with Bi-modal Transformer","code":"https://github.com/v-iashin/video_features","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":7,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"iPerceive (Chadha et al., 2020)","metrics":{"BLEU-3":"2.93","BLEU-4":"1.29","METEOR":"7.87"},"uses_additional_data":false,"paper_date":"2020-11-16","paper":"/paper/iperceive-applying-common-sense-reasoning-to-1","paper_url":"https://arxiv.org/abs/2011.07735v1","paper_title":"iPerceive: Applying Common-Sense Reasoning to Multi-Modal Dense Video Captioning and Video Question Answering","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":11,"model":"MDVC","metrics":{"BLEU-3":"2.6","BLEU-4":"1.07","METEOR":"7.31"},"uses_additional_data":false,"paper_date":"2020-03-17","paper":"/paper/multi-modal-dense-video-captioning","paper_url":"https://arxiv.org/abs/2003.07758v2","paper_title":"Multi-modal Dense Video Captioning","code":"https://github.com/v-iashin/video_features","n_code_links":4,"syntology":null},{"rank_in_archive_order":12,"model":"VTimeLLM","metrics":{"CIDEr":"27.6","SODA":"5.8"},"uses_additional_data":false,"paper_date":"2023-11-30","paper":"/paper/vtimellm-empower-llm-to-grasp-video-moments","paper_url":"https://arxiv.org/abs/2311.18445v1","paper_title":"VTimeLLM: Empower LLM to Grasp Video Moments","code":"https://github.com/huangb23/vtimellm","n_code_links":1,"syntology":{"n_ran":5,"n_unverified":6,"n_samples":11,"n_pointer_only_licence":11}}],"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":5,"rows_with_any_sample_ran":4,"distinct_papers_with_graph_line":5,"distinct_papers_with_any_sample_ran":4,"samples_over_distinct_papers":{"n_ran":11,"n_unverified":29,"n_samples":40,"n_pointer_only_licence":11,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":11,"n_unverified":29,"n_samples":40,"n_pointer_only_licence":11,"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"}}}