Globally Trained Handwritten Word Recognizer Using Spatial Representation, Convolutional Neural Networks, and Hidden Markov Models

Abstract

We introduce a new approach for on-line recognition of handwrit(cid:173) ten words written in unconstrained mixed style. The preprocessor performs a word-level normalization by fitting a model of the word structure using the EM algorithm. Words are then coded into low resolution "annotated images" where each pixel contains informa(cid:173) tion about trajectory direction and curvature. The recognizer is a convolution network which can be spatially replicated. From the network output, a hidden Markov model produces word scores. The entire system is globally trained to minimize word-level errors.

Cite

Text

Bengio et al. "Globally Trained Handwritten Word Recognizer Using Spatial Representation, Convolutional Neural Networks, and Hidden Markov Models." Neural Information Processing Systems, 1993.

Markdown

[Bengio et al. "Globally Trained Handwritten Word Recognizer Using Spatial Representation, Convolutional Neural Networks, and Hidden Markov Models." Neural Information Processing Systems, 1993.](https://mlanthology.org/neurips/1993/bengio1993neurips-globally/)

BibTeX

@inproceedings{bengio1993neurips-globally,
  title     = {{Globally Trained Handwritten Word Recognizer Using Spatial Representation, Convolutional Neural Networks, and Hidden Markov Models}},
  author    = {Bengio, Yoshua and LeCun, Yann and Henderson, Donnie},
  booktitle = {Neural Information Processing Systems},
  year      = {1993},
  pages     = {937-944},
  url       = {https://mlanthology.org/neurips/1993/bengio1993neurips-globally/}
}