Novel approach for speech recognition by using self — Organized maps

Author

Venkateswarlu, R.L.K. ; Kumari, R. Vasantha

Author_Institution

Dept. of Inf. Technol., Sasi Inst. of Technol. & Eng., Tadepalligudem, India

fYear

2011

fDate

22-24 April 2011

Firstpage

215

Lastpage

222

Abstract

The method of self-organizing maps (SOM) is a method of exploratory data analysis used for clustering and projecting multi-dimensional data into a lower-dimensional space to reveal hidden structure of the data. The Self-Organizing Feature Maps (SOFMs) is a class of neural networks capable of recognizing the main features of the data they are trained on. There is extensive literature on its biological and mathematical concepts and even more on its implementation in a variety of areas including medicine, finance, chaos and data mining in general. The aim of this research is to implement a self-organizing neural network based technique for speech recognition. The Mean-SOM performance for the feature Intensity is obtained maximum as 98.17%. The Median-SOM performance for the feature Intensity is obtained maximum as 98.54%.

Keywords

data analysis; data structures; self-organising feature maps; speech recognition; data structure; exploratory data analysis; multidimensional data clustering; neural networks; self-organized maps; self-organizing feature maps; self-organizing neural network; speech recognition; Artificial neural networks; Filter banks; Mel frequency cepstral coefficient; Neurons; Speech; Vibrations; Artificial Neural Networks; Cycles; Feature; Hits; Intensity; Iterations; LPCC; MFCC; Mean-SOM performance; Median-SOM performance; Pitch; Self-organized map;

fLanguage

English

Publisher

ieee

Conference_Titel

Emerging Trends in Networks and Computer Communications (ETNCC), 2011 International Conference on

Conference_Location

Udaipur

Print_ISBN

978-1-4577-0239-6

Type

conf

DOI

10.1109/ETNCC.2011.5958519

Filename

5958519