Recognition of emotional state using EEG signals using an adaptive deep learning model

The ability of EEG to deliver an objective and trustworthy emotion recognition for affective computing has garnered a lot of attention among the numerous emotion recognition techniques. But many current EEG emotion recognition approaches have limitations in effectively elucidating complex temporal dynamics and discriminative emotional patterns in highly nonstationary EEG data. In order to address these problems, this research proposes a unique deep learning (DL) framework for effective emotion recognition using EEG signals that combines Wavelet Scattering Transform (WST), Graph Attention Network (GAT), Adaptive Sparrow Search Optimization (ASSO), and Graph Transformer Network (GTN). During the pre-processing phase, Independent Component Analysis (ICA) is utilized to improve the quality of the EEG data by removing physiological noise. Subsequently, powerful multi-scale time-frequency representations that may extract nonlinear emotional aspects from EEG signals are obtained using the Wavelet Scattering Transform. The feature representation is successful when a Graph Attention Network is used to further enhance the feature discrimination through adaptive learning. Additionally, feature selection is optimized by adding Adaptive Sparrow Search Optimization to enhance the convergence ability of the proposed framework. Finally, a Graph Transformer Network is used to learn both local and long-range temporal dependencies in an effective way for emotion classification with attention-based graph learning mechanisms. The suggested framework’s performance metrics—accuracy, precision, recall, f1 score, and ROC-AUC—are statistically compared to the publicly accessible DEAP data set. The experimental findings demonstrate that EEG emotion detection systems employing this novel method can attain greater robustness and recognition performance, offering a useful standard for their subsequent advancement. Therefore, the suggested model can be regarded as a dependable and efficient solution for the next generation of intelligent human-computer interaction applications and affective computing.