Psychological Detection of Emotions from EEG Brainwave Signals Using Different Deep Learning Models

Abstract

Mental anguish has become a social problem and may lead to functional impairment in routine and regular work. In this paper, an electrophysiological signal called the electroencephalogram is used to suggest a method for measuring mental stress. It analyses and records the electrical activity in the brain, and by processing it and further it may be studied to look into various mental illnesses. In order to improve human wellness, this paper presents a study that identifies stress in order to detect and improve mental health. The objective of this study is to detect stress from electroencephalograph signals (EEG) with improved accuracy by testing various Deep Learning Models. By comparing their accuracy, it can be concluded that the deep learning models Convolution Neural Network (CNN) and Model with Deep Neural Network (DNN) give Mental anguish better accuracy results that is 97.65% for CNN and 97.94% for DNN on the EEG brain-waves Dataset's for the categorization of human stress level. As a result, the use of deep learning algorithms in clinical evaluation serves as a baseline for analyzing different neurological illnesses, and a highly trustworthy system may be further used to achieve important advancements in this sector.