Combination of Type III Digit Recognizers using the Dempster-Shafer Theory of Evidence

Authors:
Catalin I. Tomai;Sargur N. Srihari
Affiliations:
-;-
Venue:
ICDAR '03 Proceedings of the Seventh International Conference on Document Analysis and Recognition - Volume 2
Year:
2003

Citing 4
Cited 0

Combining the results of several neural network classifiers

Neural Networks
A Method of Combining Multiple Experts for the Recognition of Unconstrained Handwritten Numerals

IEEE Transactions on Pattern Analysis and Machine Intelligence
On Combining Classifiers

IEEE Transactions on Pattern Analysis and Machine Intelligence
Application of Dempster—Shafer theory in condition monitoring applications: a case study

Pattern Recognition Letters

Quantified Score

Hi-index	0.00

Visualization

Abstract

We investigate the combination of Type-III classifiersusing the Dempster-Shafer Theory of Evidence. Variousmethods of building BPA's for each classifier usingboth "global" and "local" classifier information are explored.We propose modifications to two established BPA-computationmethods to make them better suited for combiningType-III classifiers. We also show the effectiveness ofusing compound hypotheses when a classifier cannot confidentlychoose between the top two returned classes. Experimentaltests demonstrate the superiority of some of theapproaches proposed here on the numeral recognition problemwhen combining three different character recognizerswith Type-III classification engines.