Image Representation Using 2D Gabor Wavelets
IEEE Transactions on Pattern Analysis and Machine Intelligence
Machine Learning - Special issue on learning with probabilistic representations
A tutorial on learning with Bayesian networks
Proceedings of the NATO Advanced Study Institute on Learning in graphical models
A Handwritten Numeral Character Classification Using Tolerant Rough Set
IEEE Transactions on Pattern Analysis and Machine Intelligence
Neural Networks for Pattern Recognition
Neural Networks for Pattern Recognition
Probabilistic Networks and Expert Systems
Probabilistic Networks and Expert Systems
Pattern Classification (2nd Edition)
Pattern Classification (2nd Edition)
Hi-index | 0.00 |
To recognize the handwritten digit, this paper proposes a hierarchical Gabor features extraction method using a hierarchical Gabor filter scheme and hierarchical structure encoding the dependencies among the hierarchical Gabor features for a hierarchical bayesian network(HBN). Hierarchical Gabor features represent a different level of information which is structured such that, the higher the level, the more global information they represent, and the lower the level, the more localized information they represent. This is accomplished by a hierarchical Gabor filter scheme. HBN is a statistical model whose joint probability represents dependencies among the features hierarchically. A fully connected HBN may include irrelevant information which is useless for recognition. Pruning method can remove this irrelevant information so that the complexity of HBN can be reduced and the recognition can be accomplished more efficiently. In the experiments, we show the results of handwritten digit recognition by HBN with the hierarchical Gabor features and we compare with the naive bayesian classifier, the back-propagation neural network and the k-nearest neighbor classifier. Our proposed HBN outperforms all these methods in the experiments.