基于残差注意网络的端到端手写文本识别算法
DOI:
作者:
作者单位:

1.南京晓庄学院;2.De Montfort University

作者简介:

通讯作者:

中图分类号:

TP181

基金项目:

国家自然科学基金项目(面上项目,重点项目,重大项目)


An End-to-end Handwritten Full Page Text Recognition Method Using Convolutional Neural Networks
Author:
Affiliation:

1.Nanjing Xiaozhuang University;2.De Montfort University

Fund Project:

The National Natural Science Foundation of China (General Program, Key Program, Major Research Plan)

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    中文手写文本识别是模式识别领域中的研究热点问题之一,其字符类别数量多、书写风格差异大和训练数据集标记难等问题.针对上述问题,提出了无切分无循环的残差注意网络结构用于端到端手写文本识别,以ResNet-26为主体结构,使用深度可分离卷积提取有意义特征,残差注意门控模块提升文本图像中的关键区域的重要性.其次,采用批量双线性插值模型对输入表征进行拉伸-挤压,实现二维文本表征到一维文本行表征的文本行上采样.最后,以连接时序分类作为识别模型的损失函数,实现高层次抽取表征与字符序列标记的对应关系.在CASIA-HWDB2.x和ICDAR2013两个数据集上进行了实验研究,结果表明,该方法在没有任何字符或文本行的位置信息时能够有效地实现端到端手写文本识别且优于现有的方法.

    Abstract:

    Handwritten Chinese text recognition which involves thousands of character categories, variant writing styles and monotonous data collection process is a long-standing focus in the field of pattern recognition research. In response to these issues, we propose a residual attention networks of segmentation-free and recurrent-free for end-to-end handwritten text recognition with ResNet-26 as the main architecture, using depthwise separable convolution to extract the representation features. The residual attention gate block enhances the important of the key areas of input text image. Secondly, the text-lines up-sampling of batch bilinear interpolation is used to implement the mapping from two dimension text representation to one dimension text line representation. Finally, connectionist temporal classification as the loss function is employed to realize the corresponding relationship between the high-level extraction representation features and the character sequence labels. An experimental study was carried out on two datasets of CASIA-HWDB2.x and ICDAR2013, and the results indicate that the method can effectively implement end-to-end handwritten text recognition without any position information of characters or text lines, and superior to the existing research methods.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2021-11-21
  • 最后修改日期:2022-04-06
  • 录用日期:2022-04-08
  • 在线发布日期:
  • 出版日期: