Feature extraction analysis on Indonesian speech recognition system

Author

Wisesty, Untari N. ; Adiwijaya ; Astuti, Widi

Author_Institution

Telkom Univ., Bandung, Indonesia

fYear

2015

fDate

27-29 May 2015

Firstpage

54

Lastpage

58

Abstract

Speech recognition is widely applied to speech to text, speech to emotion, in order to make gadget and computer easier to use, or to help people with hearing disability. Feature extraction is one of significant step in the performance of speech recognition. Therefore, the proper selection is really needed. In this paper, we analyze feature extraction that can have good performance for Indonesian speech recognition system. The feature extraction method that will be analyzed are Linear Predictive Coding (LPC) and Mel Frequency Cepstral Coefficient (MFCC). Meanwhile, PNN is used as classification method in this study. The testing results show that MFCC is faster than LPC, but LPC can have the better accuracy. The accuracy of system is influenced by feature extraction, number of class and smoothing parameter.

Keywords

cepstral analysis; feature extraction; natural language processing; signal classification; smoothing methods; speech recognition; speech synthesis; Indonesian speech recognition system; LPC; MFCC; PNN; classification method; feature extraction analysis; hearing disability; linear predictive coding; mel frequency cepstral coefficient; smoothing parameter; speech to emotion; speech to text; Accuracy; Feature extraction; Mel frequency cepstral coefficient; Smoothing methods; Speech; Speech recognition; Training; Feature Extraction; LPC; MFCC; Speech Recognition System;

fLanguage

English

Publisher

ieee

Conference_Titel

Information and Communication Technology (ICoICT ), 2015 3rd International Conference on

Conference_Location

Nusa Dua

Type

conf

DOI

10.1109/ICoICT.2015.7231396

Filename

7231396