In this paper, vocal tract characteristics related to speaking rate are explored to categorise the emotions. The emotions considered are anger, disgust, fear, happy, neutral, sadness, sarcastic and surprise. These emotions are grouped into 3 broad categories namely normal, fast and slow based on speaking rate. Mel frequency cepstral coefficients (MFCC's) are used as features and Gaussian Mixture Models are used for developing the emotion classification models. The basic hypothesis is that the sequence of vocal tract shapes in producing the speech for the given utterance is unique with respect to the speaking rate. The overall classification performance of emotions using speaking rate is observed to be 91% in case of single female utterances. © 2010 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Koolagudi, S. G., Ray, S., & Sreenivasa Rao, K. (2010). Emotion classification based on speaking rate. In Communications in Computer and Information Science (Vol. 94 CCIS, pp. 316–327). https://doi.org/10.1007/978-3-642-14834-7_30
Mendeley helps you to discover research relevant for your work.