Speech Emotion Recognition Using Novel HHT-TEO Based Features

Speech emotion recognition is an important issue in the development of human-computer interactions. In this paper a series of novel robust features for speech emotion recognition is proposed. Those features, which derived from the Hilbert-Huang transform (HHT) and Teager energy operator (TEO), have the characteristics of multi-resolution, self-adaptability and high precision of distinguish ability. In the experiments, seven status of emotion were selected to be recognized and the highest 85% recognition rate was achieved within the classification accuracy of boredom reached up to 100%. The numerical results indicate that the proposed features are robust and the performance of speech emotion recognition is improved substantially.