Reproduction and Recognition of Vowel Signals Using Single and Bagging Competitive Associative Nets

Authors:
Shuichi Kurogi;Naoko Nedachi;Yuki Funatsu
Affiliations:
Kyushu Institute of Technology, Tobata, Kitakyushu, Japan 804-8550;Kyushu Institute of Technology, Tobata, Kitakyushu, Japan 804-8550;Kyushu Institute of Technology, Tobata, Kitakyushu, Japan 804-8550
Venue:
Neural Information Processing
Year:
2008

Citing 2
Cited 3

Competitive learning algorithms for vector quantization

Neural Networks
Linear Prediction of Speech

Linear Prediction of Speech

An analysis of speaker recognition using bagging CAN2 and pole distribution of speech signals

ICONIP'10 Proceedings of the 17th international conference on Neural information processing: theory and algorithms - Volume Part I
Naive bayesian multistep speaker recognition using competitive associative nets

ICONIP'11 Proceedings of the 18th international conference on Neural Information Processing - Volume Part I
Improving bagging performance through multi-algorithm ensembles

PAKDD'11 Proceedings of the 15th international conference on New Frontiers in Applied Data Mining

Quantified Score

Hi-index	0.00

Visualization

Abstract

So far, it has been shown that the piecewise linear predictive coefficients obtained by the competitive associative net called CAN2 can provide a better performance in reproduction and recognition of vowel signals than the LPC (linear predictive coding) method which is widely used for speech processing. However, when a vowel signal involves a certain amount of observation noise, the performance becomes low. In this article, we introduce bagging CAN2 and show that it can reproduce and recognize vowel signals better than the conventional single CAN2. Furthermore, we suggest that the bagging CAN2 is useful for the analysis of vowel signals.