Determination Of Disfluencies Associated In Stuttered Speech Using MFCC Feature Extraction

Stuttering also known as stammering is fluency disorder in which it affects the flow of speech, an involuntary repetitions, prolongation of sounds, syllables, phrase or words, and involuntary silent pause or blocks in communication. This involuntary speech disorder involves frequent and significant problems with the normal fluency and flow of speech. The number of disfluencies present in a speech is an important factor in determining in stuttering. The objective of this paper is to determine the disfluencies in speech which helps speech language pathologist to monitor the stuttering condition and its event. The existing methods uses Artificial neural network, Hidden Markov Model etc. In this method it involves novel approach to detect the disfluencies. The proposed method comprises of segmentation, feature extraction, SVM classifier. Feature extraction is implemented using well known Mel frequency Cepstrum coefficient.