PROJECT TITLE :

The Role of Nonlinear Dynamics in Affective Valence and Arousal Recognition

ABSTRACT:

This paper reports on a new methodology for the automated assessment of emotional responses. More specifically, emotions are elicited in agreement with a bidimensional spatial localization of affective states, that's, arousal and valence dimensions. A dedicated experimental protocol was designed and realized where specific affective states are suitably induced whereas three peripheral physiological signals, i.e., ElectroCardioGram (ECG), ElectroDermal Response (EDR), and ReSPiration activity (RSP), are simultaneously acquired. A group of 35 volunteers was presented with sets of pictures gathered from the International Affective Picture System (IAPS) having 5 levels of arousal and 5 levels of valence, including a neutral reference level in both. Customary strategies also nonlinear dynamic techniques were used to extract sets of features from the collected signals. The goal of this paper is to implement an automatic multiclass arousal/valence classifier comparing performance when extracted options from nonlinear methods are used as another to straightforward options. Results show that, when nonlinearly extracted options are used, the odds of successful recognition dramatically increase. A good recognition accuracy (>90 p.c) once 40-fold cross-validation steps for each arousal and valence classes was achieved by using the Quadratic Discriminant Classifier (QDC).