EEG based Emotion Recognition of Image Stimuli

Anonymous

TL;DR: This paper presents EEG based emotion detection of a person towards an image stimuli and its applicability on neuromarketing.

Keywords: Electroencephalography (EEG), Brain computer interface (BCI), machine learning, emotion recognition, image stimuli, neuromarketingg

Abstract: Emotion is playing a great role in our daily lives. The necessity and importance of an automatic Emotion recognition system is getting increased. Traditional approaches of emotion recognition are based on facial images, measurements of heart rates, blood pressure, temperatures, tones of voice/speech, etc. However, these features can potentially be changed to fake features. So to detect hidden and real features that is not controlled by the person are data measured from brain signals. There are various ways of measuring brain waves: EEG, MEG, FMRI, etc. On the bases of cost effectiveness and performance trade-offs, EEG is chosen for emotion recognition in this work. The main aim of this study is to detect emotion based on EEG signal analysis recorded from brain in response to visual stimuli. The approaches used were the selected visual stimuli were presented to 11 healthy target subjects and EEG signal were recorded in controlled situation to minimize artefacts (muscle or/and eye movements). The signals were filtered and type of frequency band was computed and detected. The proposed method predicts an emotion type (positive/negative) in response to the presented stimuli. Finally, the performance of the proposed approach was tested. The average accuracy of machine learning algorithms (i.e. J48, Bayes Net, Adaboost and Random Forest) are 78.86, 74.76, 77.82 and 82.46 respectively. In this study, we also applied EEG applications in the context of neuro-marketing. The results empirically demonstrated detection of the favourite colour preference of customers in response to the logo colour of an organization or Service.