EEG-based Emotion Recognition for Multi Channel Fast Empirical Mode Decomposition using VGG-16

Abstract

Much attention has been paid to the recognition of human emotions with the help of EEG signals based on machine learning methods. Human Emotion recognition is yet a difficult task to perform due to the non-linear property of the EEG signals. This paper presents an advanced signal processing method using the deep neural function to extract features using VGG-16 from all channels related to emotion. To reduce the computational costs of emotion recognition and achieve better results, this article presents a Fast Empirical mode decomposition (FEMD) model which significantly reduce the feature size for fast processing. In the proposed method, we convert the signal into a two-dimensional wavelet spectrogram and calculate the characteristics of each subject. An EEG-based emotional classification model using a Deep Neural Network (DNN) model is proposed on the SJTU SEED and DEAP datasets. Random Forest, SVM and k-NN are used to classify data into positive / negative / neutral dimensions for SEED data sets and Arousal/Valence dimensions for DEAP dataset. The proposed model achieves better accuracy on the SEED and DEAP datasets, as compared to other advanced methods of human emotion recognition.