Modeling and Simulation of Speech Emotion Recognition
Abstract: Four basic isolated emotions such as happy, anger, surprise and sadness are selected as the subject investigated and the modeling and simulation are performed on the speech emotion recognition system. This model consists of three parts such as preprocessing, feature extraction and pattern recognition. The related features such as duration, amplitude, fundamental frequency and formant parameters are extracted and are combined with the parameters extracted by using the MFCC model, which are used for the main parameters. The method of HMM which is combined with ANN is used for the speech emotion recognition. We can draw the conclusion that the model can effectively improve the speech emotion recognition rate through the simulation of MATLAB.
Key Words: Speech emotion recognition; Feature extraction; MFCC; ANN
目 录