Modelling exitement as a reaction to a virtual 3D face

This paper introduces a comparison of a linear and nonlinear one step predictive models that were used to describe the relationship between human emotional signal - excitement - as a reaction to a virtual 3D face feature - distance between eyes. An input-output model building method is proposed that allows building a stable model with the least output prediction error. Validation was performed using the recorded signals of six volunteers and the following measures: prediction error standard deviation, relative prediction error standard deviation, and average absolute relative prediction error. Validation results of the models showed that both models predict excitement signal in relatively high prediction accuracy.