Confidence- and margin-based MMI/MPE discriminative training for off-line handwriting recognition

@article{Dreuw2011ConfidenceAM,
  title={Confidence- and margin-based MMI/MPE discriminative training for off-line handwriting recognition},
  author={Philippe Dreuw and Georg Heigold and Hermann Ney},
  journal={International Journal on Document Analysis and Recognition (IJDAR)},
  year={2011},
  volume={14},
  pages={273-288}
}
We present a novel confidence- and margin-based discriminative training approach for model adaptation of a hidden Markov model (HMM)-based handwriting recognition system to handle different handwriting styles and their variations. Most current approaches are maximum-likelihood (ML) trained HMM systems and try to adapt their models to different writing styles using writer adaptive training, unsupervised clustering, or additional writer-specific data. Here, discriminative training based on the… CONTINUE READING
Highly Cited
This paper has 29 citations. REVIEW CITATIONS