Multi global context-aware transformer for ship name recognition in IoT

IF 1.5 4区 计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC
Yunting Xian, Lu Lu, Xuanrui Qiu, Jing Xian
{"title":"Multi global context-aware transformer for ship name recognition in IoT","authors":"Yunting Xian,&nbsp;Lu Lu,&nbsp;Xuanrui Qiu,&nbsp;Jing Xian","doi":"10.1049/cmu2.12773","DOIUrl":null,"url":null,"abstract":"<p>Scene text recognition has gained increasing attention in recent years, as it can connect products without an open interface in IoT. The non-local network is particularly popular in text recognition, as it can aggregate the temporal message of the input. However, existing text recognition methods based on RNN encoder-decoder structures encounter the problem of attention drift, especially in complex ship name recognition scenarios, because the features extracted by these methods are extremely similar. To address this problem, this paper proposes a novel text recognition approach named Multi Global Context-aware Transformer (MG-Cat). The proposed approach has two main properties: (1) a Global Context block that captures the global relationships among pixels inside the encoder, and (2) multiple global context-aware attention modules stacked in the encoder process. This way, the MG-Cat approach can learn a more robust intermediate feature representation in the text recognition pipeline. Moreover, the paper collected a new ship name dataset to evaluate the proposed approach. Extensive experiments were conducted on the collected dataset to verify the effectiveness of the proposed approach. The experimental results show the generalization ability of our squeeze-and-excitation global context attention module.</p>","PeriodicalId":55001,"journal":{"name":"IET Communications","volume":"19 1","pages":""},"PeriodicalIF":1.5000,"publicationDate":"2024-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1049/cmu2.12773","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IET Communications","FirstCategoryId":"94","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1049/cmu2.12773","RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0

Abstract

Scene text recognition has gained increasing attention in recent years, as it can connect products without an open interface in IoT. The non-local network is particularly popular in text recognition, as it can aggregate the temporal message of the input. However, existing text recognition methods based on RNN encoder-decoder structures encounter the problem of attention drift, especially in complex ship name recognition scenarios, because the features extracted by these methods are extremely similar. To address this problem, this paper proposes a novel text recognition approach named Multi Global Context-aware Transformer (MG-Cat). The proposed approach has two main properties: (1) a Global Context block that captures the global relationships among pixels inside the encoder, and (2) multiple global context-aware attention modules stacked in the encoder process. This way, the MG-Cat approach can learn a more robust intermediate feature representation in the text recognition pipeline. Moreover, the paper collected a new ship name dataset to evaluate the proposed approach. Extensive experiments were conducted on the collected dataset to verify the effectiveness of the proposed approach. The experimental results show the generalization ability of our squeeze-and-excitation global context attention module.

Abstract Image

用于物联网中船舶名称识别的多全局上下文感知转换器
场景文本识别近年来越来越受到关注,因为它可以在物联网中无需开放接口即可连接产品。非局部网络在文本识别中特别受欢迎,因为它可以聚合输入的时间信息。然而,现有的基于RNN编码器-解码器结构的文本识别方法遇到了注意力漂移的问题,特别是在复杂的船舶名称识别场景中,因为这些方法提取的特征非常相似。为了解决这一问题,本文提出了一种新的文本识别方法——多全局上下文感知转换器(MG-Cat)。该方法有两个主要特性:(1)全局上下文块捕获编码器内像素之间的全局关系;(2)多个全局上下文感知注意力模块堆叠在编码器过程中。这样,MG-Cat方法可以在文本识别管道中学习更健壮的中间特征表示。此外,本文还收集了一个新的船舶名称数据集来评估所提出的方法。在收集的数据集上进行了大量实验,以验证所提出方法的有效性。实验结果表明,我们的压缩激励全局上下文注意模块具有良好的泛化能力。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
IET Communications
IET Communications 工程技术-工程:电子与电气
CiteScore
4.30
自引率
6.20%
发文量
220
审稿时长
5.9 months
期刊介绍: IET Communications covers the fundamental and generic research for a better understanding of communication technologies to harness the signals for better performing communication systems using various wired and/or wireless media. This Journal is particularly interested in research papers reporting novel solutions to the dominating problems of noise, interference, timing and errors for reduction systems deficiencies such as wasting scarce resources such as spectra, energy and bandwidth. Topics include, but are not limited to: Coding and Communication Theory; Modulation and Signal Design; Wired, Wireless and Optical Communication; Communication System Special Issues. Current Call for Papers: Cognitive and AI-enabled Wireless and Mobile - https://digital-library.theiet.org/files/IET_COM_CFP_CAWM.pdf UAV-Enabled Mobile Edge Computing - https://digital-library.theiet.org/files/IET_COM_CFP_UAV.pdf
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信