Writer adaptation of a HMM handwriting recognition system

@inproceedings{Senior1997WriterAO,
  title={Writer adaptation of a HMM handwriting recognition system},
  author={Andrew W. Senior and Karthik Subramanian Nathan},
  booktitle={ICASSP},
  year={1997}
}
This paper describes a scheme to adapt the parameters of a tied-mixture, hidden Markov model, on-line handwriting recognition system to improve performance on new writers' handwriting. The means and variances of the distributions are adapted using the Maximum Likelihood Linear Regression technique [1,2]. Experiments are performed with a number of new writers in both supervised and unsupervised modes. Adaptation on data quantities as small as 5 words is found to result in models with 6% lower… CONTINUE READING

From This Paper

Figures, tables, results, and topics from this paper.

Key Quantitative Results

  • Adaptation on data quantities as small as 5 words is found to result in models with 6% lower error rate than the writer independent model.

References

Publications referenced by this paper.
Showing 1-3 of 3 references

Variance compen- sation within the MLLR framework

  • M.J.F. Gales, P. C. Woodland
  • Technical report,
  • 1996
2 Excerpts

Similar Papers

Loading similar papers…