Driving Fatigue Detection with Three Non-Hair-Bearing EEG Channels and Modified Transformer Model

Abstract

Driving fatigue is the main cause of traffic accidents, which seriously affects people’s life and property safety. Many researchers have applied electroencephalogram (EEG) signals for driving fatigue detection to reduce negative effects. The main challenges are the practicality and accuracy of the EEG-based driving fatigue detection method when it is applied on the real road. In our previous study, we attempted to improve the practicality of fatigue detection based on the proposed non-hair-bearing (NHB) montage with fewer EEG channels, but the recognition accuracy was only 76.47% with the random forest (RF) model. In order to improve the accuracy with NHB montage, this study proposed an improved transformer architecture for one-dimensional feature vector classification based on introducing the Gated Linear Unit (GLU) in the Attention sub-block and Feed-Forward Networks (FFN) sub-block of a transformer, called GLU-Oneformer. Moreover, we constructed an NHB-EEG-based feature set, including the same EEG features (power ratio, approximate entropy, and mutual information (MI)) in our previous study, and the lateralization features of the power ratio and approximate entropy based on the strategy of brain lateralization. The results indicated that our GLU-Oneformer method significantly improved the recognition performance and achieved an accuracy of 86.97%. Our framework demonstrated that the combination of the NHB montage and the proposed GLU-Oneformer model could well support driving fatigue detection.

Neuroimaging Study of Brain Functional Differences in Generalized Anxiety Disorder and Depressive Disorder

“…Furthermore, accuracy was defined as the proportion of correctly categorized samples to the total samples, precision as the proportion of correctly predicted positive categories to all the samples predicted as positive categories, recall as the proportion of correctly predicted positive categories to all the samples that are actually positive categories, and the F1 score as the reconciled mean of precision and recall. In order to calculate these four indicators, true positive ( TP ), false negative ( FN ), true negative ( TN ) and false positive ( FP ) are described as follows [ 45 ]: TP denotes the number of samples that are actually positive and correctly predicted as positive, FP denotes the number of samples that are actually negative but incorrectly predicted as positive, FN denotes the number of samples that are actually positive but incorrectly predicted as negative, and TN denotes the number of samples that are actually negative and correctly predicted as negative. Then, these four evaluation metrics for classification performances evaluation are defined as in Formulas (8)–(11).…”

Sensor-Based Classification of Primary and Secondary Car Driver Activities Using Convolutional Neural Networks

“…When looking for examples of similar studies to compare, it should be noted that in a ratio of four to one, articles were found dedicated to searching for anomalies such as drowsiness, fatigue, lack of driver concentration, and external factors associated with vehicle damage and atmospheric factors associated with driving conditions [42][43][44].…”

SiamEEGNet: Siamese Neural Network-Based EEG Decoding for Drowsiness Detection

“…By leveraging attention mechanisms, transformer-based methods can achieve performance comparable or even superior to RNN-based approaches [32]. Consequently, researchers have started exploring the application of Transformer-based methods in drowsiness detection, aiming to overcome the limitations associated with long-term sequences often encountered by RNN-based methods [17], [18]. Furthermore, the graph convolutional network (GCN) has become a popular choice for EEG-based drowsiness detection [33].…”

“…To enhance performance of decoding drowsiness-related brain activities, many researchers have incorporated DL models into their studies. Four main DL models are used in most studies: CNN [13], [14], RNN [15], [16], Transformer [17], [18], and GCN [19], [20]. Although DL-based models have demonstrated significant improvements in decoding drowsy brain activities, most of them do not consider the relative change of brain activities, which is a crucial aspect when dealing with EEG signals.…”