Petar Jerčić

Evaluating Classifiers for Emotion Recognition Using EEG

There are several ways of recording psychophysiology data from humans, for example Galvanic Skin Response (GSR), Electromyography (EMG), Electrocardiogram (ECG) and Electroencephalography (EEG). In this paper we focus on emotion detection using EEG. Various machine learning techniques can be used on the recorded EEG data to classify emotional states. K-Nearest Neighbor (KNN), Bayesian Network (BN), Artificial Neural Network (ANN) and Support Vector Machine (SVM) are some machine learning techniques that previously have been used to classify EEG data in various experiments. Five different machine learning techniques were evaluated in this paper, classifying EEG data associated with specific affective/emotional states. The emotions were elicited in the subjects using pictures from the International Affective Picture System (IAPS) database. The raw EEG data were processed to remove artifacts and a number of features were selected as input to the classifiers. The results showed that it is difficult to train a classifier to be accurate over large datasets (15 subjects) but KNN and SVM with the proposed features were reasonably accurate over smaller datasets (5 subjects) identifying the emotional states with an accuracy up to 77.78%.

Modeling cognitive load and physiological arousal through pupil diameter and heart rate

This study investigates individuals’ cognitive load processing abilities while engaged on a decision-making task in serious games, to explore how a substantial cognitive load dominates over the physiological arousal effect on pupil diameter. A serious game was presented to the participants, which displayed the on–line biofeedback based on physiological measurements of arousal. In such dynamic decision-making environment, the pupil diameter was analyzed in relation to the heart rate, to evaluate if the former could be a useful measure of cognitive abilities of individuals. As pupil might reflect both cognitive activity and physiological arousal, the pupillary response will show an arousal effect only when the cognitive demands of the situation are minimal. Evidence shows that in a situation where a substantial level of cognitive activity is required, only that activity will be observable on the pupil diameter, dominating over the physiological arousal effect indicated by the pupillary response. It is suggested that it might be possible to design serious games tailored to the cognitive abilities of an individual player, using the proposed physiological measurements to observe the moment when such dominance occurs.

The Effect of Emotions and Social Behavior on Performance in a Collaborative Serious Game Between Humans and Autonomous Robots

The aim of this paper is to investigate performance in a collaborative human–robot interaction on a shared serious game task. Furthermore, the effect of elicited emotions and perceived social behavior categories on players’ performance will be investigated. The participants collaboratively played a turn-taking version of the Tower of Hanoi serious game, together with the human and robot collaborators. The elicited emotions were analyzed in regards to the arousal and valence variables, computed from the Geneva Emotion Wheel questionnaire. Moreover, the perceived social behavior categories were obtained from analyzing and grouping replies to the Interactive Experiences and Trust and Respect questionnaires. It was found that the results did not show a statistically significant difference in participants’ performance between the human or robot collaborators. Moreover, all of the collaborators elicited similar emotions, where the human collaborator was perceived as more credible and socially present than the robot one. It is suggested that using robot collaborators might be as efficient as using human ones, in the context of serious game collaborative tasks.