Voice Morphing using GMM, MFCC synthesis and analysis
A method is proposed to perform Voice Morphing. Voice morphing is widely used to help humans who lost their speaking ability, also in gaming and animation. In this method the features of original speaker and target speaker is identified using MFCC parameter. Original speaker feature is mapped with target speaker using GMM and DTW. Once mapping of feature is done using the MFCC coefficient the target word is synthesized.