{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/multi-digit-number-recognition-from-street","title":"Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks","arxiv_id":"1312.6082","date":"2013-12-20","proceeding":null,"authors":["Ian J. Goodfellow","Yaroslav Bulatov","Julian Ibarz","Sacha Arnoud","Vinay Shet"],"abstract":"Recognizing arbitrary multi-character text in unconstrained natural\nphotographs is a hard problem. In this paper, we address an equally hard\nsub-problem in this domain viz. recognizing arbitrary multi-digit numbers from\nStreet View imagery. Traditional approaches to solve this problem typically\nseparate out the localization, segmentation, and recognition steps. In this\npaper we propose a unified approach that integrates these three steps via the\nuse of a deep convolutional neural network that operates directly on the image\npixels. We employ the DistBelief implementation of deep neural networks in\norder to train large, distributed neural networks on high quality images. We\nfind that the performance of this approach increases with the depth of the\nconvolutional network, with the best performance occurring in the deepest\narchitecture we trained, with eleven hidden layers. We evaluate this approach\non the publicly available SVHN dataset and achieve over $96\\%$ accuracy in\nrecognizing complete street numbers. We show that on a per-digit recognition\ntask, we improve upon the state-of-the-art, achieving $97.84\\%$ accuracy. We\nalso evaluate this approach on an even more challenging dataset generated from\nStreet View imagery containing several tens of millions of street number\nannotations and achieve over $90\\%$ accuracy. To further explore the\napplicability of the proposed system to broader text recognition tasks, we\napply it to synthetic distorted text from reCAPTCHA. reCAPTCHA is one of the\nmost secure reverse turing tests that uses distorted text to distinguish humans\nfrom bots. We report a $99.8\\%$ accuracy on the hardest category of reCAPTCHA.\nOur evaluations on both tasks indicate that at specific operating thresholds,\nthe performance of the proposed system is comparable to, and in some cases\nexceeds, that of human operators.","url_abs":"http://arxiv.org/abs/1312.6082v4","url_pdf":"http://arxiv.org/pdf/1312.6082v4.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/Factotum8/test_task_street_view_house_numbers","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/GurpartapSS/anpr_CNN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/JennyVanessa/Paddle-SVHN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"paddle","reach":{"status":"ok"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/JoeyMcNatt/Deep-Learning-Final-Project-Maloney-McNatt-Satriale","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/KathanKP/Multi-digit-number-sequence-prediciton-using-deep-neural-networks","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/Pek20180909/SVHN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/ahmad0790/svhn-multi-digit-address-recognition","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/ap3885/street_view_imagery","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/apoorvasrinivasan26/street_view_imagery","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/beeps82/SVHN_CNN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/caoquanjie/ConditionalLearnToPayAttention","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/lixilinx/MCMIL","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/mimikaan/Attention-Model","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/ms5898/e4040-2019fall-project","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/nikola310/svhn_classification","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/tianyu-tristan/Visual-Attention-Model","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"multi-digit-number-recognition-from-street","repo_url":"https://github.com/titulacion2021/Image-Classification-ResNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-classification-on-svhn","task":"Image Classification","dataset":"SVHN","model":"DCNN","rank_in_archive_order":29,"of":62,"metrics":{"Percentage error":"2.2"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1312.6082","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1312.6082"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ahmad0790/svhn-multi-digit-address-recognition","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Factotum8/test_task_street_view_house_numbers","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/caoquanjie/ConditionalLearnToPayAttention","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/JoeyMcNatt/Deep-Learning-Final-Project-Maloney-McNatt-Satriale","reach":{"status":"ok","spdx":"GPL-3.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/titulacion2021/Image-Classification-ResNet","reach":{"status":"ok","spdx":"NOASSERTION"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ap3885/street_view_imagery","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/nikola310/svhn_classification","reach":{"status":"ok","spdx":"GPL-3.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/GurpartapSS/anpr_CNN","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/tianyu-tristan/Visual-Attention-Model","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/apoorvasrinivasan26/street_view_imagery","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/KathanKP/Multi-digit-number-sequence-prediciton-using-deep-neural-networks","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/lixilinx/MCMIL","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ms5898/e4040-2019fall-project","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/beeps82/SVHN_CNN","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Pek20180909/SVHN","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/mimikaan/Attention-Model","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/JennyVanessa/Paddle-SVHN","reach":{"status":"ok"}}],"summary":{"ran_draft_wrong":1,"ran_fixture":1,"unverified":4},"by_repo_kind":{"listed":{"samples":6,"ran":2,"repositories":3}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":3,"samples":[{"code_sha256_prefix":"bd9deb72fd95c9ae","entry":"get_lr_metric","repo":"beeps82/SVHN_CNN","repo_kind":"listed","path":"final_runv3.py","file_url":"https://github.com/beeps82/SVHN_CNN/blob/HEAD/final_runv3.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"bd9deb72fd95c9ae"}},{"code_sha256_prefix":"c3f32e94cdacfaf7","entry":"new_accuracy","repo":"beeps82/SVHN_CNN","repo_kind":"listed","path":"final_runv3.py","file_url":"https://github.com/beeps82/SVHN_CNN/blob/HEAD/final_runv3.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"c3f32e94cdacfaf7"}},{"code_sha256_prefix":"c1f2fbdbe5cfa21f","entry":"gauss","repo":"Factotum8/test_task_street_view_house_numbers","repo_kind":"listed","path":"CNN_eval.py","file_url":"https://github.com/Factotum8/test_task_street_view_house_numbers/blob/HEAD/CNN_eval.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c1f2fbdbe5cfa21f"}},{"code_sha256_prefix":"cb97dd4d6528d2bd","entry":"gaussian_filter","repo":"Factotum8/test_task_street_view_house_numbers","repo_kind":"listed","path":"CNN_eval.py","file_url":"https://github.com/Factotum8/test_task_street_view_house_numbers/blob/HEAD/CNN_eval.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cb97dd4d6528d2bd"}},{"code_sha256_prefix":"efb7bd6cbcb8bf60","entry":"gaussian_filter_","repo":"Factotum8/test_task_street_view_house_numbers","repo_kind":"listed","path":"CNN_eval.py","file_url":"https://github.com/Factotum8/test_task_street_view_house_numbers/blob/HEAD/CNN_eval.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"efb7bd6cbcb8bf60"}},{"code_sha256_prefix":"8fb0900a352c2e0b","entry":"infer_best_match","repo":"ahmad0790/svhn-multi-digit-address-recognition","repo_kind":"listed","path":"predictions.py","file_url":"https://github.com/ahmad0790/svhn-multi-digit-address-recognition/blob/HEAD/predictions.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8fb0900a352c2e0b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}