Deep Learning-Based Approach for Emotion Recognition Using Electroencephalography (EEG) Signals Using Bi-Directional...

Area of Science:

• Neuroscience and Artificial Intelligence
• Computational Neuroscience
• Medical Informatics

Background:

• Emotions are crucial for human communication, and electroencephalography (EEG) signals offer insights into brain activity.
• Brain-Computer Interfaces (BCIs) leverage EEG for enhanced human-machine interaction, but accurate emotion recognition remains a challenge.
• Existing emotion recognition models using EEG signals often struggle with accuracy, impacting medical applications.

Purpose of the Study:

• To propose a deep learning-based approach for improving the accuracy of emotion recognition using EEG signals.
• To enhance the diagnostic capabilities for psychological and behavioral disorders through precise emotion analysis.
• To contribute a high-performance emotion recognition model for better medical decision-making.

Main Methods:

• Utilized the pre-processed Database for Emotion Analysis using Physiological signals (DEAP) dataset.
• Extracted statistical, wavelet, and Hurst exponent features from EEG signals.
• Employed Binary Gray Wolf Optimizer for feature selection and a stacked bi-directional Long Short-Term Memory (Bi-LSTM) model for classification of arousal, valence, and liking.

Main Results:

• The proposed deep learning model achieved high classification accuracies: 99.45% for valence, 96.87% for arousal, and 99.68% for liking.
• Demonstrated superior performance compared to previously reported methods in EEG-based emotion recognition.
• The model's accuracy provides a significant advancement for emotion recognition applications.

Conclusions:

• The developed deep learning approach offers a highly accurate method for EEG-based emotion recognition.
• This advancement has substantial implications for the medical field, particularly in diagnosing psychological and behavioral disorders.
• The research contributes a robust model for improved human-machine interaction and clinical decision support.