HFENet: Hybrid Feature Enhancement Network for Detecting Texts in Scenes and Traffic Panels

Text detection in complex scene images is a challenging task for intelligent transportation. Existing scene text detection methods often adopt multi-scale feature learning strategies to extract informative feature representations for covering objects of various sizes. However, the sampling operation inherent in multi-scale feature generation can easily impair high-frequency details (e.g., textures and boundaries), which are critical for text detection. In this work, the authors propose an innovative Hybrid Feature Enhancement Network (dubbed HFENet) to explicitly improve the quality of high-frequency information for detecting texts in scenes and traffic panels. To be concrete, they propose a simple yet effective self-guided feature enhancement module (SFEM) for globally lifting feature representations to highly discriminative and high-frequency abundant ones. Notably, their SFEM is pluggable and will be removed after training without introducing extra computational costs. In addition, due to the challenge and importance of accurately predicting boundaries for text detection, they propose a novel boundary enhancement module (BEM) to explicitly strengthen local feature representations in the guidance of boundary annotation for accurate localization. Extensive experiments on multiple publicly available datasets (i.e., MSRA-TD500, CTW1500, Total-Text, Traffic Guide Panel Dataset, Chinese Road Plate Dataset, and ASAYAR_TXT) verify the state-of-the-art performance of their method.

• Record URL:
  • https://doi.org/10.1109/TITS.2023.3305686
• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/oclc/41297384
• Supplemental Notes:
  • Copyright © 2023, IEEE.
• Authors:
  • Liang, Min
  • Zhu, Xiaobin
  • Zhou, Hongyang
  • Qin, Jingyan
  • Yin, Xu-Cheng
• Publication Date: 2023-12

Language

• English

Media Info

• Media Type: Web
• Features: References;
• Pagination: pp 14200-14212
• Serial:
  • IEEE Transactions on Intelligent Transportation Systems
  • Volume: 24
  • Issue Number: 12
  • Publisher: Institute of Electrical and Electronics Engineers (IEEE)
  • ISSN: 1524-9050
  • Serial URL: http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=6979

Subject/Index Terms

• TRT Terms: Detection and identification technologies; Image analysis; Intelligent transportation systems; Traffic signs
• Subject Areas: Data and Information Technology; Highways; Operations and Traffic Management;

Filing Info

• Accession Number: 01906454
• Record Type: Publication
• Files: TRIS
• Created Date: Jan 30 2024 9:25AM