{"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/textproposals-a-text-specific-selective","title":"TextProposals: a Text-specific Selective Search Algorithm for Word Spotting in the Wild","arxiv_id":"1604.02619","date":"2016-04-10","proceeding":null,"authors":["Lluis Gomez-Bigorda","Dimosthenis Karatzas"],"abstract":"Motivated by the success of powerful while expensive techniques to recognize\nwords in a holistic way, object proposals techniques emerge as an alternative\nto the traditional text detectors. In this paper we introduce a novel object\nproposals method that is specifically designed for text. We rely on a\nsimilarity based region grouping algorithm that generates a hierarchy of word\nhypotheses. Over the nodes of this hierarchy it is possible to apply a holistic\nword recognition method in an efficient way.\n  Our experiments demonstrate that the presented method is superior in its\nability of producing good quality word proposals when compared with\nclass-independent algorithms. We show impressive recall rates with a few\nthousand proposals in different standard benchmarks, including focused or\nincidental text datasets, and multi-language scenarios. Moreover, the\ncombination of our object proposals with existing whole-word recognizers shows\ncompetitive performance in end-to-end word spotting, and, in some benchmarks,\noutperforms previously published results. Concretely, in the challenging\nICDAR2015 Incidental Text dataset, we overcome in more than 10 percent f-score\nthe best-performing method in the last ICDAR Robust Reading Competition. Source\ncode of the complete end-to-end system is available at\nhttps://github.com/lluisgomez/TextProposals","url_abs":"http://arxiv.org/abs/1604.02619v3","url_pdf":"http://arxiv.org/pdf/1604.02619v3.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":"textproposals-a-text-specific-selective","repo_url":"https://github.com/lluisgomez/TextProposals","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"object","task_name":"Object"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1604.02619","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}