Fusion of n-Tuple Based Classifiers for High Performance Handwritten Character Recognition

Authors:
Konstantinos Sirlantzis;Sanaul Hoque;Michael C. Fairhurst;Ahmad Fuad Rezaur Rahman
Affiliations:
-;-;-;-
Venue:
Proceedings of the Joint IAPR International Workshop on Structural, Syntactic, and Statistical Pattern Recognition
Year:
2002

Citing 5
Cited 0

Learning internal representations by error propagation

Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
On Combining Classifiers

IEEE Transactions on Pattern Analysis and Machine Intelligence
Computer Processing of Line-Drawing Images

ACM Computing Surveys (CSUR)
Improving Product by Moderating k-NN Classifiers

MCS '01 Proceedings of the Second International Workshop on Multiple Classifier Systems
Genetic Algorithms for Multi-classifier System Configuration: A Case Study in Character Recognition

MCS '01 Proceedings of the Second International Workshop on Multiple Classifier Systems

Quantified Score

Hi-index	0.01

Visualization

Abstract

In this paper we propose a novel system for handwritten character recognition which exploits the representational power of n- tuple based classifiers while addressing successfully the issues of extensive memory size requirements usually associated with them. To achieve this we develop a scheme based on the ideas of multiple classifier fusion in which the constituent classifiers are simplified versions of the highly successful scanning n-tuple classifier. In order to explore the behaviour and statistical properties of our architecture we perform a series of cross-validation experiments drawn from the field of handwritten character recognition. The paper concludes with a number of comparisons with results on the same data set achieved by a diverse set of classifiers. Our findings clearly demonstrate the significant gains that can be obtained, simultaneously in performance and memory space reduction, by the proposed system.