Expressive Speech Identifications based on Hidden Markov Model

Abstract

This paper concerns a sub-area of a larger research field of Affective Computing, focusing on the employment of affect-recognition systems using speech modality. It is proposed that speech-based affect identification systems could play an important role as next generation biometric identification systems that are aimed at determining a person’s ‘state of mind’, or psycho-physiological state. The possible areas for the deployment of voice-affect recognition technology are discussed. Additionally, the experiments and results for emotion identification in speech based on a Hidden Markov Models (HMMs) classifier are also presented. The result from experiment suggests that certain speech feature is more precise to identify certain emotional state, and that happiness is the most difficult emotion to detect.