NeuroSyght: An AI-Powered Approach on Emotion Classification from EEG Data

The perception and interpretation of emotions are crucial to understand human behavior and cognition, which has far-reaching implications in different disciplines, including mental health diagnosis and human-computer interaction. This study is focused on the construction of a robust model capable of predicting emotions based on electroencephalogram (EEG) signals. EEG data, when harnessed through the power of artificial intelligence, has the potential to be a significant resource. The result of this research is NeuroSyght, a meticulously crafted custom model based on the architecture of a 1-dimensional (1-D) convolutional neural network. Proposed model is designed to adeptly capture the distinctive attributes of 1-D signals, particularly those derived from EEG brainwaves using the dataset SEED-V. The model effortlessly navigates the intricacies involved in mapping the nuanced features corresponding to the elicited emotions it represents. NeuroSyght stands as an exemplar to show the harmonious integration of advanced neural network design with the intricate interpretation of EEG data, presenting a refined and effective novel approach to emotion prediction. The model demonstrates an extraordinary accuracy of 99.97% when applied to the collected differential entropy features from the SEED-V dataset. This exception al precision highlights the credibility and promise of the proposed paradigm.