Title: EEG-based sympathy recognition

Authors: Muhammad Ali Sadouni; Ziba Gandomkar; Ehsan Arbabi

Addresses: School of Electrical and Computer Engineering, College of Engineering, University of Tehran, North Kargar Ave., Tehran, Iran ' School of Electrical and Computer Engineering, College of Engineering, University of Tehran, North Kargar Ave., Tehran, Iran ' School of Electrical and Computer Engineering, College of Engineering, University of Tehran, North Kargar Ave., Tehran, Iran

Abstract: Expression and recognition of different emotions play important roles in human daily life. Among different emotions, recognising sympathy is one of the most complicated ones for humans. This paper presents an EEG-based sympathy recognition system using data collected from two channels, f3 and f4, in conventional 10-20 EEG system. Different visual stimuli have been shown to two subjects. A set of features have been extracted from the recorded EEG and each interval has been labelled as neutral or stimulated. The performances of k-nearest neighbour, artificial neural network and support vector machine in classifying data into two groups have been investigated. According to the result, the best correct classification rate was in average 97.05%. It has been also found that the alpha and the delta bands are more related to sympathy, comparing to the other frequency bands.

Keywords: emotion recognition; electroencephalograms; EEG; sympathy recognition; human-computer interaction; HCI; affective computing; feature extraction; data classification; signal processing; k-nearest neighbour; kNN; artificial neural networks; ANNs; support vector machines; SVM.

DOI: 10.1504/IJMEI.2014.058529

International Journal of Medical Engineering and Informatics, 2014 Vol.6 No.1, pp.14 - 25

Received: 25 Feb 2013
Accepted: 23 Jul 2013
Published online: 24 May 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article