We describe an application of the minimum classification error (MCE) training criterion to online unconstrained-style word recognition. The described system uses allograph-HMMs to handle writer variability. The result, on vocabularies of 5k to 10k, shows that MCE training achieves around 17% word error rate reduction when compared to the baseline maximum likelihood system.
Minimum classification error training for online handwritten word recognition
2002-01-01
300616 byte
Conference paper
Electronic Resource
English
Minimum Classification Error Training for Online Handwritten Word Recognition
British Library Conference Proceedings | 2002
|Local minimum squared error for face and handwritten character recognition
British Library Online Contents | 2013
|Base Line Correction for Handwritten Word Recognition
British Library Conference Proceedings | 1995
|A Format-Driven Handwritten Word Recognition System
British Library Conference Proceedings | 2003
|Feature sets evaluation for handwritten word recognition
IEEE | 2002
|