Fusion of Combination Rules of an Ensemble of MLP Classifiers for Improved Recognition Accuracy of Handprinted Bangla Numerals

Authors:
U. Bhattacharya;B. B. Chaudhuri
Affiliations:
CVPR Unit, Indian Statistical Institute, Kolkata, India;CVPR Unit, Indian Statistical Institute, Kolkata, India
Venue:
ICDAR '05 Proceedings of the Eighth International Conference on Document Analysis and Recognition
Year:
2005

Citing 7
Cited 1

A Theory for Multiresolution Signal Decomposition: The Wavelet Representation

IEEE Transactions on Pattern Analysis and Machine Intelligence
Handwritten numerical recognition based on multiple algorithms

Pattern Recognition
Decision Combination in Multiple Classifier Systems

IEEE Transactions on Pattern Analysis and Machine Intelligence
On Combining Classifiers

IEEE Transactions on Pattern Analysis and Machine Intelligence
On-Line and Off-Line Handwriting Recognition: A Comprehensive Survey

IEEE Transactions on Pattern Analysis and Machine Intelligence
On the Choice of Training Set, Architecture and Combination Rule of Multiple MLP Classifiers for Multiresolution Recognition of Handwritten Characters

IWFHR '04 Proceedings of the Ninth International Workshop on Frontiers in Handwriting Recognition
Moment normalization of handprinted characters

IBM Journal of Research and Development

On recognition of handwritten bangla characters

ICVGIP'06 Proceedings of the 5th Indian conference on Computer Vision, Graphics and Image Processing

Quantified Score

Hi-index	0.01

Visualization

Abstract

In handwritten character recognition problem, the input images are often affected by distortions and noise. Thus such images at different resolutions include different variations in the input data. In the present work, we considered wavelet transform to obtain multi-resolution representation of each input character image. At each resolution level, we considered three MLPs with different numbers of nodes in their hidden layers and combined the outputs produced by all the MLPs of the whole ensemble by using weighted sum rule, product rule and majority voting. The set of misclassified samples produced by one combination rule is neither a subset nor a superset of a similar set produced by another rule. So, majority voting has been used for the second and final round to produce final outputs after combining the results of the three combinations of the first stage. The proposed approach produced 99.10% correct recognition rate on the test set of Bangla (a major Indian script) numeral database.