Speech Recognition System for Cerebral Palsy

This paper aims to design and implement the Speech recognition system for Cerebral Palsy using Matlab (GUI). Speech recognition system for Cerebral Palsy is a system that compares the voice from the Cerebral Palsy person as input with the reference voice in data base by using some algorithms. This work was based on the zerocross-rate (ZCR) that use for the end point detection, Mel-Frequency Cepstral Coefficients (MFCC) that use for the feature extraction, and Dynamic Time Warping (DTW) that use for the pattern classification. Basically, the system consists of two parts which are training and recognition. The simulation results show that the system cans recognition the Cerebral Palsy speech accurately. Key-Words: Cerebral Palsy, recognition, reference voice, end point detection, future extraction, pattern classification