Stuttered Isolated Spoken Marathi Speech Recognition by using MFCC and LPC

Abstract

Abstract-Stuttering is a speech disorder in which the person who stutters is unable to produce sounds, syllables, words or phrases as well as involuntary silent pauses or blocks. This paper presents the stuttered speech recognition system for Marathi Regional Language We collect the stuttered speech samples i.e. zero to nine spoken in Marathi language from Marathwada region. We also calculate the speech features i.e. Mel Frequency Cepstral Coefficents (MFCCs) and linear predictive coding (LPCs). We get the 75% recognition rate for MFCC and 82% for LPC. The different statistical methods will be applied to calculate the recognition rate.