Aug 2026· Journal of Intelligent Decision Making and Information Science· Vol 3, pp. 73-90· 0 citations· 23 references
TL;DR
The results showed that the combined CNN-LSTM model improves the accuracy of emotion detection from EEG signals and increases the accuracy of the classifications as compared to the traditional CNN and Recurrent Neural Network (RNN) models.
Abstract
This paper presents an innovative method for recognizing emotion from EEG data from the left hemisphere of the prefrontal cortex. A hybrid deep learning model using Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) is proposed. The DEAP dataset and custom dataset (MindData) used in this paper contains EEG data of emotional stimuli. The relevant features are extracted using a 1D convolutional layer to capture local temporal changes and then maxPooling1D to reduce dimension. The sequential dependencies are captured by using an LSTM layer. The final emotion is classified using a softmax layer. The feature extraction has been performed using the time domain, frequency domain, time-frequency domain, and Multi-electrode features. A hybrid approach that combines decision trees and Nearest-Neighbor classification using Random Forest (RF) and K-Nearest Neighbor (KNN) improves accuracy. Trained the model using the Adam optimizer and sparse categorical cross-entropy loss over 100 epochs. The results showed that the combined CNN-LSTM model improves the accuracy of emotion detection from EEG signals. The proposed hybrid model increases the accuracy of the classifications as compared to the traditional CNN and Recurrent Neural Network (RNN) models. This paper highlights the effectiveness of combining CNN and LSTM methods for emotion recognition and has potential applications in neurofeedback, affective computing, and real-time emotion recognition.
It is shown that gated recurrent architectures can provide strong classification performance while maintaining relatively lower architectural complexity than graph- and Transformer-based approaches and highlights the potential of GRU-based models for resource-efficient EEG emotion-recognition systems.
Naveen Saini· International journal of tec...· 0 citations
Emotion recognition based on electroencephalogram (EEG) signals is a challenging task due to the nonlinear nature and low spatial resolution of brain activity. In the context of deep learning, highly complex architectures often face the risk of hidden overfitting when dealing with noisy signals. In this work, we propos...
D. Silva, Vinicius Vitor dos Santos Dias, Marielle Jordane da Silva et al.· International Conference on...· 0 citations
Emotion dysfunction can be an indication of mental health conditions like depression, stress, anxiety, etc. Therefore, appropriate recognition of human emotion is crucial for early monitoring of mental well-being. This study aims to develop a robust multilevel feature concatenation framework that effectively integrates...
It is suggested that gated recurrent networks model the temporal structure of EEG signals more effectively than a purely convolutional model, while remaining substantially lighter than the graph-and Transformer-based architectures increasingly reported in this field.
This paper presents an explainable CNN–Transformer framework for subject-independent EEG emotion recognition, where the CNN component learns local temporal and spatial patterns, while the Transformer encoder uses self-attention to model relationships across the learned representation.
Venkatalakshmi M, B. M, Pavithra M et al.· International Research Journ...· 0 citations
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 discrimi...
M. Alagirisamy, C. Nataraj, S. Selvaperumal et al.· Discover Artificial Intellig...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.