Accurate EEG-Based Emotion Recognition using LSTM and BiLSTM Networks

Emotion recognition is crucial for advancing human-computer interaction and mental health diagnostics. Unlike speech or image-based methods, EEG provides a direct, non-manipulable measure of neural activity, offering a more reliable insight into genuine emotional states. This study addresses the challenge of accurately recognizing emotions from EEG signals by leveraging deep learning techniques. The primary objective is to develop an efficient emotion recognition system using Long Short-Term Memory (LSTM) and Bidirectional LSTM (BiLSTM) networks. EEG data were preprocessed, and features were extracted using wavelet packet decomposition. The LSTM and BiLSTM models were then trained to classify emotions into binary and multiclass categories. Results demonstrated that the BiLSTM model achieved superior accuracy, with 91.78% for binary classification and 85.21% for multiclass classification. These findings highlight the potential of advanced deep learning models in enhancing emotion recognition accuracy, contributing significantly to fields such as affective computing and mental health monitoring.