Study of Human Affective Response on Multimedia Contents

The main course in human-computer interaction is focused to understand and feed back the emotion and feeling of users. The aim of this study is to develop an emotion recognition system by bio-signals measuring, feature extraction and classification. The IAPS (International Affective Picture System) and video clip is used to elicit subject's emotions which included Joy, Pleasure, Fear, Hate and Neutrality. The subject's Bio-signals, ECG, respiration and blood pulse signal, would be measured and recorded simultaneous. By signal post-processing and analysis, the KNN classifier and the Support Vector Machine are used to classify and evaluate the emotion according to the features.