Summary
International Technical Conference on Circuits/Systems, Computers and Communications
2016
Session Number:W1-6
Session:
Number:W1-6-1
Speech Recognition using MFCC with Time Varying LPC for Similar Pronunciation Phrases
George Mufungulwa, Alia Asheralieva, Hiroshi Tsutsui, Yoshikazu Miyanaga ,
pp.763-766
Publication Date:2016/7/10
Online ISSN:2188-5079
Summary:
Noises pause a serious challenge in speech recognition systems. Both multiplicative and additive noises affect the performance of automatic speech recognition (ASR) systems. In this paper, a method which makes use of fast Fourier transform (FFT) based Mel-frequency cepstral coefficients (MFCCs) with time-varying linear predictive coding (TVLPC) is proposed. Evaluation results of similar pronunciation phrases and 142 words recognition experiments demonstrate that the proposed approach achieves better results at 10dB SNR than the conventional approach.