Search

EEG-based emotion recognition is a challenging and active research area in affective computing. We used three-dimensional (arousal, valence and dominance) model of emotion to recognize the emotions induced by music videos. The participants watched a video (1 min long) while their EEG was recorded. The main objective of the study is to identify the features that can best discriminate the emotions. Power, entropy, fractal dimension, statistical features and wavelet energy are extracted from the EEG signals. The effects of these features are investigated and the best features are identified. The performance of the two feature selection methods, Relief based algorithm and principle component analysis (PCA), is compared. PCA is adopted because of its improved performance and the efficacies of the features are validated using support vector machine, K-nearest neighbors and decision tree classifiers. Our system achieves an overall best classification accuracy of 77.62%, 78.96% and 77.60% for valence, arousal and dominance respectively. Our results demonstrated that time-domain statistical characteristics of EEG signals can efficiently discriminate different emotional states. Also, the use of three-dimensional emotion model is able to classify similar emotions that were not correctly classified by two-dimensional model (e.g. anger and fear). The results of this study can be used to support the development of real-time EEG-based emotion recognition systems.

Emotion is a crucial aspect of human health, and emotion recognition systems serve important roles in the development of neurofeedback applications. Most of the emotion recognition methods proposed in previous research take predefined EEG features as input to the classification algorithms. This paper investigates the less studied method of using plain EEG signals as the classifier input, with the residual networks (ResNet) as the classifier of interest. ResNet having excelled in the automated hierarchical feature extraction in raw data domains with vast number of samples (e.g., image processing) is potentially promising in the future as the amount of publicly available EEG databases has been increasing. Architecture of the original ResNet designed for image processing is restructured for optimal performance on EEG signals. The arrangement of convolutional kernel dimension is demonstrated to largely affect the model’s performance on EEG signal processing. The study is conducted on the Shanghai Jiao Tong University Emotion EEG Dataset (SEED), with our proposed ResNet18 architecture achieving 93.42% accuracy on the 3-class emotion classification, compared to the original ResNet18 at 87.06% accuracy. Our proposed ResNet18 architecture has also achieved a model parameter reduction of 52.22% from the original ResNet18. We have also compared the importance of different subsets of EEG channels from a total of 62 channels for emotion recognition. The channels placed near the anterior pole of the temporal lobes appeared to be most emotionally relevant. This agrees with the location of emotion-processing brain structures like the insular cortex and amygdala.

Emotion recognition can be useful in various applications such as in neurofeedback training for functional enhancement. A practically realizable emotion recognition system should rely on as little physiological signals/modalities as possible. Also, emotion-related neurological activities may be vastly different from person to person. Hence, this paper presents the single-modal EEG-based personalized emotion recognition convolutional neural network (CNN) models working on the DEAP dataset. The valence and arousal level classification performance of our presented CNN classifiers have surpassed the other emotion classifiers working on the DEAP dataset based on our scope of literature reviewed. The models, which are deep CNN, rely on only plain EEG data and require no pre-extracted EEG features. The design and application of the CNN models is aimed at possible future work of identification of new emotion-related EEG features, relying on the automated feature extraction capability of the CNN. The two CNN models presented have achieved the 3-class valence classification test accuracy of 97.59% and 98.75% respectively, and the 3-class arousal classification test accuracy of 98.48% and 97.58%.

In this paper, an enhanced face recognition method is proposed to develop an efficient student attendance system. In the face recognition stage, enhanced local binary pattern (ELBP) is implemented by increasing the radius of the original local binary pattern (LBP) operator and principal component analysis (PCA) is used to extract the features from facial images. The ELBP outperforms the original LBP by suppressing the adverse influence of different illumination conditions and thus results in increased recognition rate. The test recognition rate on the Yale face database of the original LBP is 81.9%, while the ELBP with radius 2 to 5 pixels has achieved face recognition rate of 84.6% to 88.5%. The face images are recognized based on the classification results obtained from the combination of algorithms (PCA and ELBP). A user-friendly interface has been designed to improve the attendance marking process. Image pre-processing is performed to ensure better performance of the system. Lastly, the attendance of the recognized students is marked accordingly and saved in an excel sheet. Students who have not undergone pre-registration can register on the spot. Notification will be generated if a student signs in more than once. The recognition rate for high quality images in the attendance system is 100% whereas for low quality images it is 92.31%.