eprintid: 10116299 rev_number: 22 eprint_status: archive userid: 608 dir: disk0/10/11/62/99 datestamp: 2020-12-02 17:03:24 lastmod: 2022-01-27 15:07:39 status_changed: 2020-12-02 17:03:24 type: article metadata_visibility: show creators_name: Deng, G creators_name: Ming, Y creators_name: Xue, J-H title: RFRN: A recurrent feature refinement network for accurate and efficient scene text detection ispublished: pub divisions: UCL divisions: B04 divisions: C06 divisions: F61 keywords: Scene text detection; Recurrent segmentation; Feature pyramid network; Feature refinement note: This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions. abstract: Scene text detection plays a vital role for scene text understanding, but arbitrary-shaped text detection remains a significant challenge. To extract discriminative features, most recent state-of-the-art methods adopt heavy networks, resulting in parameter redundancy and inference inefficiency. For accurate and efficient scene text detection, in this paper we propose a novel recurrent feature refinement network (RFRN). RFRN, as a recurrent segmentation framework, contains a recurrent path augmentation that refines the previous feature maps as inner states, which not only helps improve the segmentation quality, but also fully facilitates the reuse of parameters and low computational cost. During testing, RFRN discards redundant prediction procedures for efficient inference, and achieves a good balance between speed and accuracy of inference. We conduct experiments on four challenging scene text benchmarks, CTW1500, Total-Text, ICDAR2015 and ICDAR2017-MLT, which include curved texts and multi-oriented texts with complex background. The results show that the proposed RFRN achieves competitive performance on detection accuracy while maintaining computational efficiency. date: 2021-09-17 date_type: published publisher: Elsevier BV official_url: https://doi.org/10.1016/j.neucom.2020.10.099 oa_status: green full_text_type: other language: eng primo: open primo_central: open_green verified: verified_manual elements_id: 1832738 doi: 10.1016/j.neucom.2020.10.099 lyricists_name: Xue, Jinghao lyricists_id: JXUEX60 actors_name: Thomas, Chloe actors_id: CTHOM59 actors_role: owner full_text_status: public publication: Neurocomputing volume: 453 pagerange: 465-481 citation: Deng, G; Ming, Y; Xue, J-H; (2021) RFRN: A recurrent feature refinement network for accurate and efficient scene text detection. Neurocomputing , 453 pp. 465-481. 10.1016/j.neucom.2020.10.099 <https://doi.org/10.1016/j.neucom.2020.10.099>. Green open access document_url: https://discovery.ucl.ac.uk/id/eprint/10116299/7/Xue_1-s2.0-S0925231220317124-main.pdf