A string length predictor to control the level building of HMMs for handwritten numeral recognition

/in , , /by
A string length predictor to control the level building of HMMs for handwritten numeral recognition

Britto, Alceu De S. and Sabourin, Robert and Bortolozzi, Flavio and Suen, Ching Y.

Proceedings – International Conference on Pattern Recognition 2002

Abstract : In this paper a two-stage HMM-based method for recognizing handwritten numeral strings is extended to work with handwritten numeral strings of unknown length. We have proposed a Bayesian-based string length predictor (SLP) to estimate ‘the number of digits in a string taking into account its width in pixels. The top 3 decisions of the SLP module are used to control the maximum number of levels to be searched by the Level Building (LB) algorithm. On 12,802 handwritten numeral strings and 2,069 touching digit pairs, this strategy has shown a small loss (0.91%) in terms of recognition performance compared to the results when the string length is considered as known. © 2002 IEEE.

You might also like
Intelligent oscillometric system for automatic detection of peripheral arterial disease
An ILP solver for multi-label MRFS with connectivity constraints
Toward a better noninvasive assessment of preejection period: A novel automatic algorithm for B-point detection and correction on thoracic impedance cardiogram
Texture-based descriptors for writer identification and verification
on the Applicability of Κ-Distributions
An investigation into the effects of label noise on Dynamic Selection algorithms
Sygma: Reducing symmetry in graph mining
Named entity recognition for audio de-identification