No abstract

Abstract. This paper introduces a comparison of one linear and two nonlinear one-step-ahead predictive models that were used to describe the relationship between human emotional signals (excitement, frustration, and engagement/boredom) and virtual dynamic stimulus (virtual 3D face with changing distance-between-eyes). An input-output model building method is proposed that allows building a stable model with the smallest output prediction error. Validation was performed using the recorded signals of four volunteers. Validation results of the models showed that all three models predict emotional signals in relatively high prediction accuracy.

This paper introduces how predictor-based control principles are applied to the control of human excitement signal as a response to a 3D face virtual stimuli. A dynamic human 3D face is observed in a virtual reality. We use changing distance-between-eyes in a 3D face as a stimuluscontrol signal. Human responses to the stimuli are observed using EEG-based signal that characterizes excitement. A parameter identification method for predictive and stable model building with the smallest output prediction error is proposed. A predictor-based control law is synthesized by minimizing a generalized minimum variance control criterion in an admissible domain. An admissible domain is composed of control signal boundaries. Relatively high prediction and control quality of excitement signals is demonstrated by modelling results.

No abstract

This paper introduces identification results of human response to virtual 3D face stimuli. Observations of human reactions are done using preprocessed EEG (electroencephalogram) signals: excitement, meditation, frustration, engagement/boredom. Virtual 3D face features -distance between eyes, nose width, and chin width -are used as stimuli. Cross-correlation analysis demonstrated that dynamical relations between human reactions and stimuli exist. Input-output models describing relations between stimuli and corresponding human reactions are built. A new input-output model building method is proposed that allows building stable models with the least output prediction error. Models' validation results demonstrate relatively high prediction accuracy of human reactions.