{"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/stackmix-and-blot-augmentations-for","title":"StackMix and Blot Augmentations for Handwritten Text Recognition","arxiv_id":"2108.11667","date":"2021-08-26","proceeding":null,"authors":["Alex Shonenkov","Denis Karachev","Maxim Novopoltsev","Mark Potanin","Denis Dimitrov"],"abstract":"This paper proposes a handwritten text recognition(HTR) system that outperforms current state-of-the-artmethods. The comparison was carried out on three of themost frequently used in HTR task datasets, namely Ben-tham, IAM, and Saint Gall. In addition, the results on tworecently presented datasets, Peter the Greats manuscriptsand HKR Dataset, are provided.The paper describes the architecture of the neural net-work and two ways of increasing the volume of train-ing data: augmentation that simulates strikethrough text(HandWritten Blots) and a new text generation method(StackMix), which proved to be very effective in HTR tasks.StackMix can also be applied to the standalone task of gen-erating handwritten text based on printed text.","url_abs":"https://arxiv.org/abs/2108.11667v1","url_pdf":"https://arxiv.org/pdf/2108.11667v1.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":"stackmix-and-blot-augmentations-for","repo_url":"https://github.com/sberbank-ai/StackMix-OCR","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"data-augmentation","task_name":"Data Augmentation"},{"task_slug":"htr","task_name":"HTR"},{"task_slug":"handwritten-text-recognition","task_name":"Handwritten Text Recognition"},{"task_slug":"text-generation","task_name":"Text Generation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/handwritten-text-recognition-on-bentham","task":"Handwritten Text Recognition","dataset":"Bentham","model":"StackMix+Blots","rank_in_archive_order":1,"of":1,"metrics":{"CER":"1.73"},"uses_additional_data":false},{"leaderboard":"/sota/handwritten-text-recognition-on-digital-peter","task":"Handwritten Text Recognition","dataset":"Digital Peter","model":"StackMix+Blots","rank_in_archive_order":1,"of":1,"metrics":{"CER":"2.5"},"uses_additional_data":false},{"leaderboard":"/sota/handwritten-text-recognition-on-hkr","task":"Handwritten Text Recognition","dataset":"HKR","model":"StackMix+Blots","rank_in_archive_order":1,"of":1,"metrics":{"CER":"3.49"},"uses_additional_data":false},{"leaderboard":"/sota/handwritten-text-recognition-on-iam-b","task":"Handwritten Text Recognition","dataset":"IAM-B","model":"StackMix+Blots","rank_in_archive_order":1,"of":1,"metrics":{"CER":"3.77"},"uses_additional_data":false},{"leaderboard":"/sota/handwritten-text-recognition-on-iam-d","task":"Handwritten Text Recognition","dataset":"IAM-D","model":"StackMix+Blots","rank_in_archive_order":1,"of":1,"metrics":{"CER":"3.01"},"uses_additional_data":false},{"leaderboard":"/sota/handwritten-text-recognition-on-saint-gall","task":"Handwritten Text Recognition","dataset":"Saint Gall","model":"StackMix+Blots","rank_in_archive_order":1,"of":1,"metrics":{"CER":"3.65"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}