Classification of speech under stress using physical features based on two-mass model

摘要

We propose the classification methods of speech under stress based on a physical model, which characterizes the vocal folds and the vocal tract. We use physical parameters estimated by fitting a two-mass model to real speech, and two type of dynamic parameters are proposed to represent the short-term and long-term dynamic changes in physical characteristics. Experimental results show that our proposed features are effective and achieved better classification performance than features derived from traditional methods.