ASSESSMENT OF DYSARTHRIC SPEECH USING MFCC

Speech is the effective form of communication between human and its environment. Dysarthria is a motor speech disorder in which the person lacks the control over articulators used for speech production. Speech accuracy is the outcome of well-timed and coordinated activities of the articulators and other related neuro muscular feature. In this paper, Speech utterance is converted into a phone sequence and histograms of the pronunciation mappings are done by using Mel-frequency cepstral coefficients. Structured sparse feature selection is done using Hidden Markov Models. Prediction is done using Inverse Mel-frequency cepstral coefficients. It is a comparative study of different methodologies to improve the speech of dysarthric disabled people.