16th International Conference on Pattern Recognition (ICPR'02) - Volume 1 Rejection Strategies and Confidence Measures for a k - NN Classifier in an OCR Task Quebec City, QC, Canada August 11-August 15 ISBN: 0-7695-1695-X
In Handwritten Character Recognition, the rejection of extraneous patterns, like image noise, strokes or corrections, can improve significantly the practical usefulness of a system. In this paper, a combination of two confidence measures defi ned for a k-nearest neighbors classifier is proposed. Experiments are presented comparing the performance of the same system with and without the new rejection rules.
Citation:
Joaquim Arlandis, Juan C. Perez-Cortes, Javier Cano, "Rejection Strategies and Confidence Measures for a k - NN Classifier in an OCR Task," icpr, vol. 1, pp.10576, 16th International Conference on Pattern Recognition (ICPR'02) - Volume 1, 2002 Usage of this product signifies your acceptance of the Terms of Use. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||