Electroencephalogram based Emotion Recognition with Graph Convolutional Network Model

Abstract

Recently, researchers have focused on understanding human sentiment using mechanical devices or reactions to any machinery activity. Computerization is becoming more prevalent in today’s environment. People are unaware of the proper way of expressing their emotions to others. People are unsure how to respond in some situations. Emotional intelligence is a collection of abilities that includes emotional awareness and self-control. In 1995, Daniel Goleman’s book Emotional Intelligence popularized the term. Emotional intelligence has five components: self-awareness, motivation, self-regulation, and social abilities. Emotion indicates a broad phrase that alludes to a human being’s cognitive or intelligible and psychological comeback to the perceived circumstances of another person. Emotional response or sensitivity towards others boosts one’s chances of assisting others and displaying sentiment. Some people have been traumatized, handicapped, or have a disability that makes it difficult for them to express themselves. Our goal is to evaluate human sentiment and the factors working behind emotions using EEG signals to identify a person’s feelings. We propose a deep learning-based approach with a hybrid model for detecting emotions such as happiness, sadness, etc. The electroencephalogram, abbreviation of EEG, is a medical evaluation that computes the electrical activity of the cerebrum using electrodes or wires placed on the scalp. Using EEG-based emotion recognition, the computer can see inside the user’s head to study their mental state. To achieve this goal, our mission is to discover the cognitive stimulation that plays a crucial role in generating happiness and sadness in the human brain via brain signals using Deep learning(DL) approach and hybrid Graph Convolutional Network(GCN) model.