{"url":"/sota/scene-text-detection-on-icdar-2013","task":{"name":"Scene Text Detection","url":"/task/scene-text-detection","note":null},"dataset":{"name":"ICDAR 2013","url":"/dataset/icdar-2013"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"**Scene Text Detection** is a computer vision task that involves automatically identifying and localizing text within natural images or videos. The goal of scene text detection is to develop algorithms that can robustly detect and and label text with bounding boxes in uncontrolled and complex environments, such as street signs, billboards, or license plates.\r\n\r\n\r\n<span class=\"description-source\">Source: [ContourNet: Taking a Further Step toward Accurate Arbitrary-shaped Scene Text Detection ](https://arxiv.org/abs/2004.04940)</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":["F-Measure","Precision","Recall","H-Mean"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"F-Measure":null,"Precision":"higher","Recall":"higher","H-Mean":null}},"counts":{"rows":16,"rows_with_code":13,"rows_with_paper_page":16,"rows_dated":16,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"TextFuseNet (ResNeXt-101)","metrics":{"F-Measure":"94.61%","Precision":"97.27","Recall":"92.09"},"uses_additional_data":false,"paper_date":"2020-05-17","paper":"/paper/textfusenet-scene-text-detection-with-richer","paper_url":"https://doi.org/10.24963/ijcai.2020/72","paper_title":"TextFuseNet: Scene Text Detection with Richer Fused Features","code":"https://github.com/ying09/TextFuseNet","n_code_links":6,"syntology":null},{"rank_in_archive_order":2,"model":"SPCNET","metrics":{"F-Measure":"92.1%","Precision":"93.8","Recall":"90.5"},"uses_additional_data":false,"paper_date":"2018-11-21","paper":"/paper/scene-text-detection-with-supervised-pyramid","paper_url":"http://arxiv.org/abs/1811.08605v1","paper_title":"Scene Text Detection with Supervised Pyramid Context Network","code":"https://github.com/AirBernard/Scene-Text-Detection-with-SPCNET","n_code_links":2,"syntology":null},{"rank_in_archive_order":3,"model":"Mask TextSpotter","metrics":{"F-Measure":"91.7%","Precision":"95","Recall":"88.6"},"uses_additional_data":false,"paper_date":"2018-07-06","paper":"/paper/mask-textspotter-an-end-to-end-trainable","paper_url":"http://arxiv.org/abs/1807.02242v2","paper_title":"Mask TextSpotter: An End-to-End Trainable Neural Network for Spotting Text with Arbitrary Shapes","code":"https://github.com/lvpengyuan/masktextspotter.caffe2","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"WordSup (VGG16-synth-icdar)","metrics":{"F-Measure":"90.34%","Precision":"93.34","Recall":"87.53"},"uses_additional_data":false,"paper_date":"2017-08-22","paper":"/paper/wordsup-exploiting-word-annotations-for","paper_url":"http://arxiv.org/abs/1708.06720v1","paper_title":"WordSup: Exploiting Word Annotations for Character based Text Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":5,"model":"STN-OCR","metrics":{"F-Measure":"90.3%"},"uses_additional_data":false,"paper_date":"2017-07-27","paper":"/paper/stn-ocr-a-single-neural-network-for-text","paper_url":"http://arxiv.org/abs/1707.08831v1","paper_title":"STN-OCR: A single Neural Network for Text Detection and Text Recognition","code":"https://github.com/Bartzi/stn-ocr","n_code_links":3,"syntology":null},{"rank_in_archive_order":6,"model":"PixelLink+VGG16 2s MS","metrics":{"F-Measure":"88.1%","Precision":"88.6","Recall":"87.5"},"uses_additional_data":false,"paper_date":"2018-01-04","paper":"/paper/pixellink-detecting-scene-text-via-instance","paper_url":"http://arxiv.org/abs/1801.01315v1","paper_title":"PixelLink: Detecting Scene Text via Instance Segmentation","code":"https://github.com/ZJULearning/pixel_link","n_code_links":5,"syntology":null},{"rank_in_archive_order":7,"model":"Corner Localization (multi-scale)","metrics":{"F-Measure":"88%","Precision":"92","Recall":"84.4"},"uses_additional_data":false,"paper_date":"2018-02-25","paper":"/paper/multi-oriented-scene-text-detection-via","paper_url":"http://arxiv.org/abs/1802.08948v2","paper_title":"Multi-Oriented Scene Text Detection via Corner Localization and Region Segmentation","code":"https://github.com/lvpengyuan/corner","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"TextBoxes++_MS","metrics":{"F-Measure":"88%%","Precision":"91","Recall":"84"},"uses_additional_data":false,"paper_date":"2018-01-09","paper":"/paper/textboxes-a-single-shot-oriented-scene-text","paper_url":"http://arxiv.org/abs/1801.02765v3","paper_title":"TextBoxes++: A Single-Shot Oriented Scene Text Detector","code":"https://github.com/MhLiao/TextBoxes_plusplus","n_code_links":3,"syntology":null},{"rank_in_archive_order":9,"model":"Corner-based Region Proposals","metrics":{"F-Measure":"87.6%%","Precision":"91.9","Recall":"83.9"},"uses_additional_data":false,"paper_date":"2018-04-08","paper":"/paper/detecting-multi-oriented-text-with-corner","paper_url":"https://arxiv.org/abs/1804.02690v2","paper_title":"Detecting Multi-Oriented Text with Corner-based Region Proposals","code":"https://github.com/xhzdeng/crpn","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"SSTD","metrics":{"F-Measure":"87%","Precision":"88","Recall":"86"},"uses_additional_data":false,"paper_date":"2017-09-01","paper":"/paper/single-shot-text-detector-with-regional","paper_url":"http://arxiv.org/abs/1709.00138v1","paper_title":"Single Shot Text Detector with Regional Attention","code":"https://github.com/BestSonny/SSTD","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"SegLink","metrics":{"F-Measure":"85.3%","Precision":"87.7","Recall":"83"},"uses_additional_data":false,"paper_date":"2017-03-19","paper":"/paper/detecting-oriented-text-in-natural-images-by","paper_url":"http://arxiv.org/abs/1703.06520v3","paper_title":"Detecting Oriented Text in Natural Images by Linking Segments","code":"https://github.com/dengdan/seglink","n_code_links":6,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":12,"model":"Gupta et al.","metrics":{"F-Measure":"83.0%","Precision":"92.0","Recall":"75.5"},"uses_additional_data":false,"paper_date":"2016-04-22","paper":"/paper/synthetic-data-for-text-localisation-in","paper_url":"http://arxiv.org/abs/1604.06646v1","paper_title":"Synthetic Data for Text Localisation in Natural Images","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":6,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":13,"model":"USM (COCO TS + ICDAR–2013)","metrics":{"F-Measure":"80.40%"},"uses_additional_data":false,"paper_date":"2019-09-29","paper":"/paper/unsharp-masking-layer-injecting-prior","paper_url":"https://doi.org/10.1007/978-3-030-30508-6_1","paper_title":"Unsharp Masking Layer: Injecting Prior Knowledge in Convolutional Networks for Image Classification","code":"https://github.com/maeotaku/pytorch_usm","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"Neumann et al. *","metrics":{"F-Measure":"77.1%","Precision":"81.8","Recall":"72.4"},"uses_additional_data":false,"paper_date":"2015-04-14","paper":"/paper/efficient-scene-text-localization-and","paper_url":"http://arxiv.org/abs/1504.03522v1","paper_title":"Efficient Scene Text Localization and Recognition with Local Character Refinement","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":15,"model":"Jaderberg et al.","metrics":{"F-Measure":"76.8%","Precision":"88.5","Recall":"67.8"},"uses_additional_data":false,"paper_date":"2014-12-04","paper":"/paper/reading-text-in-the-wild-with-convolutional","paper_url":"http://arxiv.org/abs/1412.1842v1","paper_title":"Reading Text in the Wild with Convolutional Neural Networks","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":16,"model":"CRAFT","metrics":{"H-Mean":"95.2","Precision":"97.4","Recall":"93.1"},"uses_additional_data":false,"paper_date":"2019-04-03","paper":"/paper/character-region-awareness-for-text-detection","paper_url":"http://arxiv.org/abs/1904.01941v1","paper_title":"Character Region Awareness for Text Detection","code":"https://github.com/JaidedAI/EasyOCR","n_code_links":18,"syntology":{"n_ran":6,"n_unverified":34,"n_samples":40,"n_pointer_only_licence":4}}],"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":3,"rows_with_any_sample_ran":2,"distinct_papers_with_graph_line":3,"distinct_papers_with_any_sample_ran":2,"samples_over_distinct_papers":{"n_ran":7,"n_unverified":40,"n_samples":47,"n_pointer_only_licence":5,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":7,"n_unverified":40,"n_samples":47,"n_pointer_only_licence":5,"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"}}}