{"url":"/sota/scene-text-recognition-on-icdar2013","task":{"name":"Scene Text Recognition","url":"/task/scene-text-recognition","note":null},"dataset":{"name":"ICDAR2013","url":"/dataset/icdar-2013"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"See [Scene Text Detection](https://paperswithcode.com/task/scene-text-detection) for leaderboards in this task.","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":["Accuracy"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Accuracy":"higher"}},"counts":{"rows":38,"rows_with_code":35,"rows_with_paper_page":38,"rows_dated":38,"rows_using_additional_data":9},"rows":[{"rank_in_archive_order":1,"model":"CLIP4STR-L*","metrics":{"Accuracy":"99.42"},"uses_additional_data":true,"paper_date":"2023-12-29","paper":"/paper/an-empirical-study-of-scaling-law-for-ocr","paper_url":"https://arxiv.org/abs/2401.00028v3","paper_title":"An Empirical Study of Scaling Law for OCR","code":"https://github.com/large-ocr-model/large-ocr-model.github.io","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"DTrOCR 105M","metrics":{"Accuracy":"99.4"},"uses_additional_data":false,"paper_date":"2023-08-30","paper":"/paper/dtrocr-decoder-only-transformer-for-optical","paper_url":"https://arxiv.org/abs/2308.15996v1","paper_title":"DTrOCR: Decoder-only Transformer for Optical Character Recognition","code":"https://github.com/arvindrajan92/DTrOCR","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"CLIP4STR-L (DataComp-1B)","metrics":{"Accuracy":"99.0"},"uses_additional_data":false,"paper_date":"2023-05-23","paper":"/paper/clip4str-a-simple-baseline-for-scene-text-1","paper_url":"https://arxiv.org/abs/2305.14014v4","paper_title":"CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model","code":"https://github.com/VamosC/CLIP4STR","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":4,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"MGP-STR","metrics":{"Accuracy":"98.5"},"uses_additional_data":true,"paper_date":"2022-09-08","paper":"/paper/multi-granularity-prediction-for-scene-text","paper_url":"https://arxiv.org/abs/2209.03592v2","paper_title":"Multi-Granularity Prediction for Scene Text Recognition","code":"https://github.com/alibabaresearch/advancedliteratemachinery","n_code_links":3,"syntology":null},{"rank_in_archive_order":5,"model":"CLIP4STR-L","metrics":{"Accuracy":"98.5"},"uses_additional_data":true,"paper_date":"2023-05-23","paper":"/paper/clip4str-a-simple-baseline-for-scene-text-1","paper_url":"https://arxiv.org/abs/2305.14014v4","paper_title":"CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model","code":"https://github.com/VamosC/CLIP4STR","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":4,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":6,"model":"PARSeq","metrics":{"Accuracy":"98.4±0.2"},"uses_additional_data":true,"paper_date":"2022-07-14","paper":"/paper/scene-text-recognition-with-permuted","paper_url":"https://arxiv.org/abs/2207.06966v1","paper_title":"Scene Text Recognition with Permuted Autoregressive Sequence Models","code":"https://github.com/topdu/openocr","n_code_links":2,"syntology":{"n_ran":5,"n_unverified":3,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"CCD-ViT-Base(ARD_2.8M)","metrics":{"Accuracy":"98.3"},"uses_additional_data":true,"paper_date":"2022-11-01","paper":"/paper/self-supervised-character-to-character","paper_url":"https://arxiv.org/abs/2211.00288v4","paper_title":"Self-supervised Character-to-Character Distillation for Text Recognition","code":"https://github.com/tongkunguan/ccd","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"CCD-ViT-Small(ARD_2.8M)","metrics":{"Accuracy":"98.3"},"uses_additional_data":true,"paper_date":"2022-11-01","paper":"/paper/self-supervised-character-to-character","paper_url":"https://arxiv.org/abs/2211.00288v4","paper_title":"Self-supervised Character-to-Character Distillation for Text Recognition","code":"https://github.com/tongkunguan/ccd","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"CLIP4STR-B","metrics":{"Accuracy":"98.3"},"uses_additional_data":true,"paper_date":"2023-05-23","paper":"/paper/clip4str-a-simple-baseline-for-scene-text-1","paper_url":"https://arxiv.org/abs/2305.14014v4","paper_title":"CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model","code":"https://github.com/VamosC/CLIP4STR","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":4,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"MATRN","metrics":{"Accuracy":"97.9"},"uses_additional_data":false,"paper_date":"2021-11-30","paper":"/paper/multi-modal-text-recognition-networks","paper_url":"https://arxiv.org/abs/2111.15263v3","paper_title":"Multi-modal Text Recognition Networks: Interactive Enhancements between Visual and Semantic Features","code":"https://github.com/topdu/openocr","n_code_links":3,"syntology":null},{"rank_in_archive_order":11,"model":"S-GTR","metrics":{"Accuracy":"97.8"},"uses_additional_data":true,"paper_date":"2021-12-24","paper":"/paper/visual-semantics-allow-for-textual-reasoning-1","paper_url":"https://arxiv.org/abs/2112.12916v1","paper_title":"Visual Semantics Allow for Textual Reasoning Better in Scene Text Recognition","code":"https://github.com/adeline-cs/GTR","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"SIGA_T","metrics":{"Accuracy":"97.8"},"uses_additional_data":false,"paper_date":"2022-03-07","paper":"/paper/a-glyph-driven-topology-enhancement-network","paper_url":"https://arxiv.org/abs/2203.03382v4","paper_title":"Self-supervised Implicit Glyph Attention for Text Recognition","code":"https://github.com/tongkunguan/siga","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"DPAN","metrics":{"Accuracy":"97.7"},"uses_additional_data":false,"paper_date":"2021-08-01","paper":"/paper/look-back-again-dual-parallel-attention","paper_url":"https://dl.acm.org/doi/10.1145/3460426.3463674","paper_title":"Look Back Again: Dual Parallel Attention Network for Accurate and Robust Scene Text Recognition","code":"https://github.com/siddagra/DPAN-look-back-Again-Dual-Parallel-Attention-Network-for-Accurate-and-Robust-Scene-Text-Recognition","n_code_links":2,"syntology":null},{"rank_in_archive_order":14,"model":"CDistNet (Ours)","metrics":{"Accuracy":"97.67"},"uses_additional_data":false,"paper_date":"2021-11-22","paper":"/paper/cdistnet-perceiving-multi-domain-character","paper_url":"https://arxiv.org/abs/2111.11011v5","paper_title":"CDistNet: Perceiving Multi-Domain Character Distance for Robust Text Recognition","code":"https://github.com/topdu/openocr","n_code_links":3,"syntology":null},{"rank_in_archive_order":15,"model":"CCD-ViT-Tiny(ARD_2.8M)","metrics":{"Accuracy":"97.5"},"uses_additional_data":true,"paper_date":"2022-11-01","paper":"/paper/self-supervised-character-to-character","paper_url":"https://arxiv.org/abs/2211.00288v4","paper_title":"Self-supervised Character-to-Character Distillation for Text Recognition","code":"https://github.com/tongkunguan/ccd","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"SVTR-L (Large)","metrics":{"Accuracy":"97.2"},"uses_additional_data":false,"paper_date":"2022-04-30","paper":"/paper/svtr-scene-text-recognition-with-a-single","paper_url":"https://arxiv.org/abs/2205.00159v2","paper_title":"SVTR: Scene Text Recognition with a Single Visual Model","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":4,"syntology":null},{"rank_in_archive_order":17,"model":"SVTR-B (Base)","metrics":{"Accuracy":"97.1"},"uses_additional_data":false,"paper_date":"2022-04-30","paper":"/paper/svtr-scene-text-recognition-with-a-single","paper_url":"https://arxiv.org/abs/2205.00159v2","paper_title":"SVTR: Scene Text Recognition with a Single Visual Model","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":4,"syntology":null},{"rank_in_archive_order":18,"model":"DiffusionSTR","metrics":{"Accuracy":"97.1"},"uses_additional_data":false,"paper_date":"2023-06-29","paper":"/paper/diffusionstr-diffusion-model-for-scene-text","paper_url":"https://arxiv.org/abs/2306.16707v1","paper_title":"DiffusionSTR: Diffusion Model for Scene Text Recognition","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":19,"model":"Yet Another Text Recognizer","metrics":{"Accuracy":"96.8"},"uses_additional_data":false,"paper_date":"2021-07-29","paper":"/paper/why-you-should-try-the-real-data-for-the","paper_url":"https://arxiv.org/abs/2107.13938v1","paper_title":"Why You Should Try the Real Data for the Scene Text Recognition","code":"https://github.com/openvinotoolkit/training_extensions","n_code_links":1,"syntology":null},{"rank_in_archive_order":20,"model":"SVTR-T (Tiny)","metrics":{"Accuracy":"96.3"},"uses_additional_data":false,"paper_date":"2022-04-30","paper":"/paper/svtr-scene-text-recognition-with-a-single","paper_url":"https://arxiv.org/abs/2205.00159v2","paper_title":"SVTR: Scene Text Recognition with a Single Visual Model","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":4,"syntology":null},{"rank_in_archive_order":21,"model":"SVTR-S (Small)","metrics":{"Accuracy":"95.7"},"uses_additional_data":false,"paper_date":"2022-04-30","paper":"/paper/svtr-scene-text-recognition-with-a-single","paper_url":"https://arxiv.org/abs/2205.00159v2","paper_title":"SVTR: Scene Text Recognition with a Single Visual Model","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":4,"syntology":null},{"rank_in_archive_order":22,"model":"SRN","metrics":{"Accuracy":"95.5"},"uses_additional_data":false,"paper_date":"2020-03-27","paper":"/paper/towards-accurate-scene-text-recognition-with","paper_url":"https://arxiv.org/abs/2003.12294v1","paper_title":"Towards Accurate Scene Text Recognition with Semantic Reasoning Networks","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":3,"syntology":null},{"rank_in_archive_order":23,"model":"RCEED","metrics":{"Accuracy":"94.7"},"uses_additional_data":false,"paper_date":"2021-06-13","paper":"/paper/representation-and-correlation-enhanced","paper_url":"https://arxiv.org/abs/2106.06960v2","paper_title":"Representation and Correlation Enhanced Encoder-Decoder Framework for Scene Text Recognition","code":"https://github.com/Mona9955/RCEED-ICDAR2021","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"SATRN","metrics":{"Accuracy":"94.1"},"uses_additional_data":false,"paper_date":"2019-10-10","paper":"/paper/on-recognizing-texts-of-arbitrary-shapes-with","paper_url":"https://arxiv.org/abs/1910.04396v1","paper_title":"On Recognizing Texts of Arbitrary Shapes with 2D Self-Attention","code":"https://github.com/Media-Smart/vedastr","n_code_links":2,"syntology":null},{"rank_in_archive_order":25,"model":"DAN","metrics":{"Accuracy":"93.9"},"uses_additional_data":false,"paper_date":"2019-12-21","paper":"/paper/decoupled-attention-network-for-text","paper_url":"https://arxiv.org/abs/1912.10205v1","paper_title":"Decoupled Attention Network for Text Recognition","code":"https://github.com/topdu/openocr","n_code_links":5,"syntology":null},{"rank_in_archive_order":26,"model":"CSTR","metrics":{"Accuracy":"93.2"},"uses_additional_data":false,"paper_date":"2021-02-22","paper":"/paper/cstr-a-classification-perspective-on-scene","paper_url":"https://arxiv.org/abs/2102.10884v3","paper_title":"Revisiting Classification Perspective on Scene Text Recognition","code":"https://github.com/Media-Smart/vedastr","n_code_links":1,"syntology":null},{"rank_in_archive_order":27,"model":"TextScanner","metrics":{"Accuracy":"92.9"},"uses_additional_data":false,"paper_date":"2019-12-28","paper":"/paper/textscanner-reading-characters-in-order-for","paper_url":"https://arxiv.org/abs/1912.12422v2","paper_title":"TextScanner: Reading Characters in Order for Robust Scene Text Recognition","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":28,"model":"SEED","metrics":{"Accuracy":"92.8"},"uses_additional_data":false,"paper_date":"2020-05-22","paper":"/paper/seed-semantics-enhanced-encoder-decoder","paper_url":"https://arxiv.org/abs/2005.10977v1","paper_title":"SEED: Semantics Enhanced Encoder-Decoder Framework for Scene Text Recognition","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":3,"syntology":null},{"rank_in_archive_order":29,"model":"SAFL","metrics":{"Accuracy":"92.8"},"uses_additional_data":false,"paper_date":"2022-01-01","paper":"/paper/safl-a-self-attention-scene-text-recognizer-1","paper_url":"https://arxiv.org/abs/2201.00132v1","paper_title":"SAFL: A Self-Attention Scene Text Recognizer with Focal Loss","code":"https://github.com/ICMLA-SAFL/SAFL_pytorch","n_code_links":1,"syntology":null},{"rank_in_archive_order":30,"model":"ViTSTR","metrics":{"Accuracy":"92.4"},"uses_additional_data":false,"paper_date":"2021-05-18","paper":"/paper/vision-transformer-for-fast-and-efficient","paper_url":"https://arxiv.org/abs/2105.08582v1","paper_title":"Vision Transformer for Fast and Efficient Scene Text Recognition","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":3,"syntology":null},{"rank_in_archive_order":31,"model":"Baek et al.","metrics":{"Accuracy":"92.3"},"uses_additional_data":false,"paper_date":"2019-04-03","paper":"/paper/what-is-wrong-with-scene-text-recognition","paper_url":"https://arxiv.org/abs/1904.01906v4","paper_title":"What Is Wrong With Scene Text Recognition Model Comparisons? Dataset and Model Analysis","code":"https://github.com/clovaai/deep-text-recognition-benchmark","n_code_links":13,"syntology":{"n_ran":13,"n_unverified":6,"n_samples":19,"n_pointer_only_licence":0}},{"rank_in_archive_order":32,"model":"ASTER","metrics":{"Accuracy":"91.8"},"uses_additional_data":false,"paper_date":"2018-06-25","paper":"/paper/aster-an-attentional-scene-text-recognizer","paper_url":"http://122.205.5.5:8071/UpLoadFiles/Papers/ASTER_PAMI18.pdf","paper_title":"ASTER: An Attentional Scene Text Recognizer with Flexible Rectification","code":"https://github.com/bgshih/aster","n_code_links":4,"syntology":null},{"rank_in_archive_order":33,"model":"CA-FCN","metrics":{"Accuracy":"91.5"},"uses_additional_data":false,"paper_date":"2018-09-18","paper":"/paper/scene-text-recognition-from-two-dimensional","paper_url":"http://arxiv.org/abs/1809.06508v2","paper_title":"Scene Text Recognition from Two-Dimensional Perspective","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":34,"model":"SAR","metrics":{"Accuracy":"91.0"},"uses_additional_data":false,"paper_date":"2018-11-02","paper":"/paper/show-attend-and-read-a-simple-and-strong","paper_url":"http://arxiv.org/abs/1811.00751v2","paper_title":"Show, Attend and Read: A Simple and Strong Baseline for Irregular Text Recognition","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":8,"syntology":{"n_ran":0,"n_unverified":8,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":35,"model":"STAR-Net","metrics":{"Accuracy":"89.1"},"uses_additional_data":false,"paper_date":"2016-09-20","paper":"/paper/star-net-a-spatial-attention-residue-network","paper_url":"http://www.bmva.org/bmvc/2016/papers/paper043/paper043.pdf","paper_title":"Star-net: A spatial attention residue network for scene text recognition.","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":1,"syntology":null},{"rank_in_archive_order":36,"model":"RARE","metrics":{"Accuracy":"88.6"},"uses_additional_data":false,"paper_date":"2016-03-12","paper":"/paper/robust-scene-text-recognition-with-automatic","paper_url":"http://arxiv.org/abs/1603.03915v2","paper_title":"Robust Scene Text Recognition with Automatic Rectification","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":5,"syntology":null},{"rank_in_archive_order":37,"model":"CRNN","metrics":{"Accuracy":"86.7"},"uses_additional_data":false,"paper_date":"2015-07-21","paper":"/paper/an-end-to-end-trainable-neural-network-for","paper_url":"http://arxiv.org/abs/1507.05717v1","paper_title":"An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition","code":"https://github.com/PaddlePaddle/PaddleOCR","n_code_links":85,"syntology":{"n_ran":19,"n_unverified":62,"n_samples":81,"n_pointer_only_licence":17}},{"rank_in_archive_order":38,"model":"CHAR","metrics":{"Accuracy":"79.5"},"uses_additional_data":false,"paper_date":"2014-06-09","paper":"/paper/synthetic-data-and-artificial-neural-networks","paper_url":"http://arxiv.org/abs/1406.2227v4","paper_title":"Synthetic Data and Artificial Neural Networks for Natural Scene Text Recognition","code":"https://github.com/bupt-ai-cz/Meta-SelfLearning","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 7,081 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":7081,"papers_extracted_not_yet_verified":217,"boards_without_verdict":27,"papers_not_yet_extracted":2325},"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":7,"rows_with_any_sample_ran":6,"distinct_papers_with_graph_line":5,"distinct_papers_with_any_sample_ran":4,"samples_over_distinct_papers":{"n_ran":43,"n_unverified":83,"n_samples":126,"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":55,"n_unverified":91,"n_samples":146,"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"}}}