| Deep Neural Modeling of Emotional States from Electroencephalographic Signals |
| Paper ID : 1037-ICEEM2025 (R2) |
| Authors |
|
Sameh N Attia1, Ayman Hagag1, Tamer M. Nassef *2 1Electronics Technology Department, Faculty of Technology and Education, Helwan University 2Faculty of Computer Science, October Modern University for Sciences and Arts (MSA) |
| Abstract |
| The emotion recognition through electroencephalographic (EEG) signals has emerged as a promising application within the field of Brain-Computer Interface (BCI) technology, enabling seamless interaction between the human brain and external systems. This study presents a deep neural network (DNN) architecture optimized for classifying emotional states from EEG signals using advanced feature extraction techniques and signal processing methods. Drawing from historical developments and recent advancements, the proposed model integrates statistical measures, wavelet packet transforms, and deep learning to extract temporal and frequency-based features from EEG recordings. A comprehensive experimental setup, utilizing the SEED dataset and employing a three-phase data split (training, validation, testing), demonstrates the model’s ability to achieve exceptional accuracy (98.89%) across multiple emotional classes. Performance evaluation using metrics such as accuracy, precision, recall, F1-score, and AUC confirms the model’s robustness, sensitivity, and specificity. Compared to existing models, this approach yields significant improvements in classification accuracy, computational efficiency, and generalization. The findings underscore the model’s potential for real-world applications, particularly in clinical diagnostics, neurological monitoring, and affective computing. Future directions involve expanding the dataset diversity, refining feature extraction techniques, and enhancing model generalizability across broader populations while maintaining ethical standards and data privacy. |
| Keywords |
| Emotion Recognition, Electroencephalography (EEG), Brain-Computer Interface (BCI), Deep Neural Network (DNN), Wavelet Packet Transform. |
| Status: Accepted |