Two-stepped majority voting for efficient EEG-based emotion classification

Aras M Ismael,Ömer F Alçin,Karmand Hussein Abdalla,Abdulkadir Şengür

doi:10.1186/s40708-020-00111-3

Aras M Ismael, Ömer F Alçin + Show 2 more

Open Access

https://doi.org/10.1186/s40708-020-00111-3

Copy DOI

Abstract

In this paper, a novel approach that is based on two-stepped majority voting is proposed for efficient EEG-based emotion classification. Emotion recognition is important for human–machine interactions. Facial features- and body gestures-based approaches have been generally proposed for emotion recognition. Recently, EEG-based approaches become more popular in emotion recognition. In the proposed approach, the raw EEG signals are initially low-pass filtered for noise removal and band-pass filters are used for rhythms extraction. For each rhythm, the best performed EEG channels are determined based on wavelet-based entropy features and fractal dimension-based features. The k-nearest neighbor (KNN) classifier is used in classification. The best five EEG channels are used in majority voting for getting the final predictions for each EEG rhythm. In the second majority voting step, the predictions from all rhythms are used to get a final prediction. The DEAP dataset is used in experiments and classification accuracy, sensitivity and specificity are used for performance evaluation metrics. The experiments are carried out to classify the emotions into two binary classes such as high valence (HV) vs low valence (LV) and high arousal (HA) vs low arousal (LA). The experiments show that 86.3% HV vs LV discrimination accuracy and 85.0% HA vs LA discrimination accuracy is obtained. The obtained results are also compared with some of the existing methods. The comparisons show that the proposed method has potential in the use of EEG-based emotion classification.

Highlights

Emotions are defined as biological situations associated with the nervous system [1]
The DEAP dataset was used in experiments and accuracy scores 75.19% and 81.74% were reported for valence and arousal classes, respectively
The proposed method employed a two-stepped majority voting procedure to increase the performance of the emotion classification

Summary

Introduction

Emotions are defined as biological situations associated with the nervous system [1]. The authors mentioned that the efficiency of the EEG-based emotion classification could be improved using shorter time segments of the EEG signals. Rozgiç et al developed a three-stepped approach for EEG-based emotion classification [6]. Chen et al proposed deep convolutional neural networks (DCNN) for EEG-based emotion recognition [8]. Atkinson et al proposed a novel feature-based approach for EEG-based emotion recognition [10]. Zhang et al used EMD and sample entropy for classifications of emotion EEG signals into valence and arousal classes [15]. An efficient two-stepped majority voting approach is proposed for EEG-based emotion recognition. 1. Two-stepped majority voting approach is proposed for efficient EEG-based emotion recognition. Where ε symbolizes a positive threshold and must be equal or greater than two

Higuchi’s fractal dimension

Fc1 Fp2 F4 Fc2 Majority voting

Findings

Conclusions