Automatic Detection of Pathological Voices Using GMM-SVM Method

Modern lifestyle has increased the risk of pathological voices problems. So the therapy of pathological people attracts more attention of people. Meanwhile, acoustic features have been used widely in the therapy of voice disordered people. Classification of Normal and Pathological people is also an auxiliary therapy operation. MFCC has been proved to be a useful feature with traditional classifier such as GMM or HMM. However, the precision rate of the classification can still be improved. In Pattern Recognition field, GMM-SVM has been an effective classification method. In this study, we found that this classification method is also effective in voice disorder classification. EER was improved from 8.2% of GMM to 6.0% of GMM-SVM.