An On-Line Cursive Word Recognition System

Abstract

This paper presents a system for large vocabulary recognition of on-line handwritten cursive words. The system rst uses a ltering module, based onsimple letter features, to quickly reduce alarge reference dictionary to a smaller number of candidates � the reduced lexicon along with the original input is subsequently fed toarecognition module. In order to exploit the sequential nature of the temporal data, we employ a TDNNstyle network architecture which has been successfully used in the speech recognition domain. Explicit segmentation of the input words into characters is avoided by using a sliding window concept where the input word representation (a set of frames) is presented to the neural network-based recognizer sequentially. The outputs of the recognition module are collected andconverted into a string of characters that can be matched with the candidate words. A description of the complete system and its components is given. 1

Cite

Text

Seni et al. "An On-Line Cursive Word Recognition System." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 1994. doi:10.1109/CVPR.1994.323858

Markdown

[Seni et al. "An On-Line Cursive Word Recognition System." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 1994.](https://mlanthology.org/cvpr/1994/seni1994cvpr-line/) doi:10.1109/CVPR.1994.323858

BibTeX

@inproceedings{seni1994cvpr-line,
  title     = {{An On-Line Cursive Word Recognition System}},
  author    = {Seni, Giovanni and Nasrabadi, Nasser M. and Srihari, Rohini K.},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {1994},
  pages     = {404-410},
  doi       = {10.1109/CVPR.1994.323858},
  url       = {https://mlanthology.org/cvpr/1994/seni1994cvpr-line/}
}