Comparison of Different Speech Feature Extraction Techniques with and without Wavelet Transform to Kannada Speech Recognition

Abstract

Pre-processing of speech signals is considered a crucial step in the development of a robust and efficient speech or speaker recognition system. This paper deals with different speech processing techniques and the recognition accuracy with respect to wavelet transforms. It is shown that by applying wavelet transform to the conventional methods the signal recognition accuracy will be increased by using discrete wavelet transforms and the wavelet packets for clean and noisy speech signals respectively. Results presented in the tabular form, shows the advantage of pre-processing the signals with wavelet techniques gives good results over conventional methods.