An Identification Method for Road Hypnosis Based on Human EEG Data

The driver in road hypnosis has not only some external characteristics, but also some internal characteristics. External features have obvious manifestations and can be directly observed. Internal features do not have obvious manifestations and cannot be directly observed. They need to be measured with specific instruments. Electroencephalography (EEG), as an internal feature of drivers, is the golden parameter for drivers' life identification. EEG is of great significance for the identification of road hypnosis. An identification method for road hypnosis based on human EEG data is proposed in this paper. EEG data on drivers in road hypnosis can be collected through vehicle driving experiments and virtual driving experiments. The collected data are preprocessed with the PSD (power spectral density) method, and EEG characteristics are extracted. The neural networks EEGNet, RNN, and LSTM are used to train the road hypnosis identification model. It is shown from the results that the model based on EEGNet has the best performance in terms of identification for road hypnosis, with an accuracy of 93.01%. The effectiveness and accuracy of the identification for road hypnosis are improved in this study. The essential characteristics for road hypnosis are also revealed. This is of great significance for improving the safety level of intelligent vehicles and reducing the number of traffic accidents caused by road hypnosis.

EEG-based driving intuition and collision anticipation using joint temporal-frequency multi-layer dynamic brain network

Intuition plays a crucial role in human driving decision-making, and this rapid and unconscious cognitive process is essential for improving traffic safety. We used the first proposed multi-layer network analysis method, "Joint Temporal-Frequency Multi-layer Dynamic Brain Network" (JTF-MDBN), to study the EEG data from the initial and advanced phases of driving intuition training in the theta, alpha, and beta bands. Additionally, we conducted a comparative study between these two phases using multi-layer metrics as well as local and global metrics of single layers. The results show that brain region activity is more stable in the advanced phase of intuition training compared to the initial phase. Particularly in the alart state task, the JTF-MDBN demonstrated stronger connection strength. Multi-layer network analysis indicates that modularity is significantly higher for the non-alert state task than the alert state task in the alpha and beta bands. In the W4 time window (1 second before a collision), we identified significant features that can differentiate situations where a car collision is imminent from those where no collision occurs. Single-layer network analysis also revealed statistical differences in node strength and local efficiency for some EEG channels in the alpha and beta bands during the W4 and W5 time windows. Using these biomarkers to predict vehicle collision risk, the classification accuracy of a linear kernel SVM reached up to 87.5%, demonstrating the feasibility of predicting driving collisions through brain network biomarkers. These findings are important for the study of human intuition and the development of brain-computer interface-based intelligent driving hazard perception assistance systems.

Psychophysiological models of hypovigilance detection: A scoping review

Hypovigilance represents a major contributor to accidents. In operational contexts, the burden of monitoring/managing vigilance often rests on operators. Recent advances in sensing technologies allow for the development of psychophysiology-based (hypo)vigilance prediction models. Still, these models remain scarcely applied to operational situations and need better understanding. The current scoping review provides a state of knowledge regarding psychophysiological models of hypovigilance detection. Records evaluating vigilance measuring tools with gold standard comparisons and hypovigilance prediction performances were extracted from MEDLINE, PsychInfo, and Inspec. Exclusion criteria comprised aspects related to language, non-empirical papers, and sleep studies. The Quality Assessment tool for Diagnostic Accuracy Studies (QUADAS) and the Prediction model Risk Of Bias ASsessment Tool (PROBAST) were used for bias evaluation. Twenty-one records were reviewed. They were mainly characterized by participant selection and analysis biases. Papers predominantly focused on driving and employed several common psychophysiological techniques. Yet, prediction methods and gold standards varied widely. Overall, we outline the main strategies used to assess hypovigilance, their principal limitations, and we discuss applications of these models.

Self-Attentive Channel-Connectivity Capsule Network for EEG-Based Driving Fatigue Detection

Deep neural networks have recently been successfully extended to EEG-based driving fatigue detection. Nevertheless, most existing models fail to reveal the intrinsic inter-channel relations that are known to be beneficial for EEG-based classification. Additionally, these models require substantial data for training, which is often impractical due to the high cost of data collection. To simultaneously address these two issues, we propose a Self-Attentive Channel-Connectivity Capsule Network (SACC-CapsNet) for EEG-based driving fatigue detection in this paper. SACC-CapsNet starts with a temporal-channel attention module to investigate the critical temporal information and important channels for driving fatigue detection, refining the input EEG signals. Subsequently, the refined EEG data are transformed into a channel covariance matrix to capture the inter-channel relations, followed by selective kernel attention to extract the highly discriminative channel-connectivity features. Finally, a capsule neural network is employed to effectively learn the relationships between connectivity features, which is more suitable for limited data. To confirm the effectiveness of SACC-CapsNet, we collected 24-channel EEG data from 31 subjects (mean age=23.13±2.68 years, male/female=18/13) in a simulated fatigue driving environment. Extensive experiments were conducted with the acquired data, and the comparison results show that our proposed model outperforms state-of-the-art methods. Additionally, the channel covariance matrix learned from SACC-CapsNet reveals that the frontal pole is most informative for detecting driving fatigue, followed by the parietal and central regions. Intriguingly, the temporal-channel attention module can enhance the significance of these critical regions, and the reconstructed channel covariance matrix generated by the decoder network of SACC-CapsNet can effectively preserve valuable information about them.

An Electro-Oculogram (EOG) Sensor's Ability to Detect Driver Hypovigilance Using Machine Learning

Driving safely is crucial to avoid death, injuries, or financial losses that can be sustained in an accident. Thus, a driver's physical state should be monitored to prevent accidents, rather than vehicle-based or behavioral measurements, and provide reliable information in this regard. Electrocardiography (ECG), electroencephalography (EEG), electrooculography (EOG), and surface electromyography (sEMG) signals are used to monitor a driver's physical state during a drive. The purpose of this study was to detect driver hypovigilance (drowsiness, fatigue, as well as visual and cognitive inattention) using signals collected from 10 drivers while they were driving. EOG signals from the driver were preprocessed to remove noise, and 17 features were extracted. ANOVA (analysis of variance) was used to select statistically significant features that were then loaded into a machine learning algorithm. We then reduced the features by using principal component analysis (PCA) and trained three classifiers: support vector machine (SVM), k-nearest neighbor (KNN), and ensemble. A maximum accuracy of 98.7% was obtained for the classification of normal and cognitive classes under the category of two-class detection. Upon considering hypovigilance states as five-class, a maximum accuracy of 90.9% was achieved. In this case, the number of detection classes increased, resulting in a reduction in the accuracy of detecting more driver states. However, with the possibility of incorrect identification and the presence of issues, the ensemble classifier's performance produced an enhanced accuracy when compared to others.

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

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.

Landscape Perception Identification and Classification Based on Electroencephalogram (EEG) Features

This paper puts forward a new method of landscape recognition and evaluation by using aerial video and EEG technology. In this study, seven typical landscape types (forest, wetland, grassland, desert, water, farmland, and city) were selected. Different electroencephalogram (EEG) signals were generated through different inner experiences and feelings felt by people watching video stimuli of the different landscape types. The electroencephalogram (EEG) features were extracted to obtain the mean amplitude spectrum (MAS), power spectrum density (PSD), differential entropy (DE), differential asymmetry (DASM), rational asymmetry (RASM), and differential caudality (DCAU) in the five frequency bands of delta, theta, alpha, beta, and gamma. According to electroencephalogram (EEG) features, four classifiers including the back propagation (BP) neural network, k-nearest neighbor classification (KNN), random forest (RF), and support vector machine (SVM) were used to classify the landscape types. The results showed that the support vector machine (SVM) classifier and the random forest (RF) classifier had the highest accuracy of landscape recognition, which reached 98.24% and 96.72%, respectively. Among the six classification features selected, the classification accuracy of MAS, PSD, and DE with frequency domain features were higher than those of the spatial domain features of DASM, RASM and DCAU. In different wave bands, the average classification accuracy of all subjects was 98.24% in the gamma band, 94.62% in the beta band, and 97.29% in the total band. This study identifies and classifies landscape perception based on multi-channel EEG signals, which provides a new idea and method for the quantification of human perception.

A Review of EEG Signal Features and their Application in Driver Drowsiness Detection Systems

Detecting drowsiness in drivers, especially multi-level drowsiness, is a difficult problem that is often approached using neurophysiological signals as the basis for building a reliable system. In this context, electroencephalogram (EEG) signals are the most important source of data to achieve successful detection. In this paper, we first review EEG signal features used in the literature for a variety of tasks, then we focus on reviewing the applications of EEG features and deep learning approaches in driver drowsiness detection, and finally we discuss the open challenges and opportunities in improving driver drowsiness detection based on EEG. We show that the number of studies on driver drowsiness detection systems has increased in recent years and that future systems need to consider the wide variety of EEG signal features and deep learning approaches to increase the accuracy of detection.

Explaining Deep Learning-Based Driver Models

Different systems based on Artificial Intelligence (AI) techniques are currently used in relevant areas such as healthcare, cybersecurity, natural language processing, and self-driving cars. However, many of these systems are developed with “black box” AI, which makes it difficult to explain how they work. For this reason, explainability and interpretability are key factors that need to be taken into consideration in the development of AI systems in critical areas. In addition, different contexts produce different explainability needs which must be met. Against this background, Explainable Artificial Intelligence (XAI) appears to be able to address and solve this situation. In the field of automated driving, XAI is particularly needed because the level of automation is constantly increasing according to the development of AI techniques. For this reason, the field of XAI in the context of automated driving is of particular interest. In this paper, we propose the use of an explainable intelligence technique in the understanding of some of the tasks involved in the development of advanced driver-assistance systems (ADAS). Since ADAS assist drivers in driving functions, it is essential to know the reason for the decisions taken. In addition, trusted AI is the cornerstone of the confidence needed in this research area. Thus, due to the complexity and the different variables that are part of the decision-making process, this paper focuses on two specific tasks in this area: the detection of emotions and the distractions of drivers. The results obtained are promising and show the capacity of the explainable artificial techniques in the different tasks of the proposed environments.

A dynamic center and multi threshold point based stable feature extraction network for driver fatigue detection utilizing EEG signals

Driver fatigue is the one of the main reasons of the traffic accidents. The human brain is a complex structure, whose function can be evaluated with electroencephalogram (EEG). Automated driver fatigue detection utilizing EEG decreases the incidence probability of related traffic accidents. Therefore, devising an appropriate feature extraction technique and selecting a competent classification method can be considered as the crucial part of the effective driver fatigue detection. Therefore, in this study, an EEG-based intelligent system was devised for driver fatigue detection. The proposed framework includes a new feature generation network, which is implemented by using texture descriptors, for fatigue detection. The proposed scheme contains pre-processing, feature generation, informative features selection and classification with shallow classifiers phases. In the pre-processing, discrete cosine transform and fast Fourier transform are used together. Moreover, dynamic center based binary pattern and multi threshold ternary pattern are utilized together to create a new feature generation network. To improve the detection performance, we utilized discrete wavelet transform as a pooling method, in which the functional brain network-based feature describing the relationship between fatigue and brain network organization. In the feature selection phase, a hybrid three layered feature selection method is presented, and benchmark classifiers are used in the classification phase to demonstrate the strength of the proposed method. In the experiments, the proposed framework achieved 97.29% classification accuracy for fatigue detection using EEG signals. This result reveals that the proposed framework can be utilized effectively for driver fatigue detection.

Detecting Unfavorable Driving States in Electroencephalography Based on a PCA Sample Entropy Feature and Multiple Classification Algorithms

Unfavorable driving states can cause a large number of vehicle crashes and are significant factors in leading to traffic accidents. Hence, the aim of this research is to design a robust system to detect unfavorable driving states based on sample entropy feature analysis and multiple classification algorithms. Multi-channel Electroencephalography (EEG) signals are recorded from 16 participants while performing two types of driving tasks. For the purpose of selecting optimal feature sets for classification, principal component analysis (PCA) is adopted for reducing dimensionality of feature sets. Multiple classification algorithms, namely, K nearest neighbor (KNN), decision tree (DT), support vector machine (SVM) and logistic regression (LR) are employed to improve the accuracy of unfavorable driving state detection. We use 10-fold cross-validation to assess the performance of the proposed systems. It is found that the proposed detection system, based on PCA features and the cubic SVM classification algorithm, shows robustness as it obtains the highest accuracy of 97.81%, sensitivity of 96.93%, specificity of 98.73% and precision of 98.75%. Experimental results show that the system we designed can effectively monitor unfavorable driving states.

Design of electroencephalogram authentication access control to smart car

In recent years, with the development of intelligent vehicles, the demand for security will be bigger and bigger. One of the most important solutions is the use of new biometric technology. At present, there are still some areas to be improved on biometric technology. For example, diseases will destroy some biological characteristics, some detection methods are too slow, many detection methods do not need living detection, and so on. Electroencephalogram (EEG) is a new biometric tool for living identification. In this Letter, a kind of identity authentication system based on the EEG signal is presented. The overall goal of this research is to design a new authentication method and develop the corresponding application. Therefore, the authors carried out a series of EEG experiments, and analysed and discussed the experimental results. Based on these results, they build and present an access control system based on the uniqueness of their EEG signals to be capable of authenticating access control to the car. The accuracy of the authentication system is >87.3%.

Partial directed coherence based graph convolutional neural networks for driving fatigue detection

The mental state of a driver can be accurately and reliably evaluated by detecting the driver's electroencephalogram (EEG) signals. However, traditional machine learning and deep learning methods focus on the single electrode feature analysis and ignore the functional connection of the brain. In addition, the recent brain function connection network method needs to manually extract substantial brain network features, which results in cumbersome operation. For this reason, this paper introduces graph convolution combined with brain function connection theory into the study of mental fatigue and proposes a method for driving fatigue detection based on the partial directed coherence graph convolutional neural network (PDC-GCNN), which can analyze the characteristics of single electrodes while automatically extracting the topological features of the brain network. We designed a fatigue driving simulation experiment and collected the EEG signals. In the present work, the PDC method constructs the adjacency matrix to describe the relationship between EEG channels, and the GCNN combines single-electrode local brain area information and brain area connection information to further improve the performance of detecting fatigue states. Based on the features of differential entropy (DE) and power spectral density (PSD), the average recognition accuracy of ten-fold cross validation is 84.32% and 83.84%, respectively. For further experiments on each subject, the average recognition results are 95.24%/5.10% (PSD) and 96.01%/3.81% (DE). This research can be embedded in the vehicle driving fatigue detection system, which has practical application value.

Constructing Multi-scale Entropy Based on the Empirical Mode Decomposition(EMD) and its Application in Recognizing Driving Fatigue

BACKGROUND

Fatigue is one of the important factors in traffic accidents. Hence, it is necessary to devise methods to detect the fatigue and apply practical fatigue detection solutions for drivers.

NEW METHOD

This paper presents a method based on the empirical mode decomposition(EMD) of multi-scale entropy on the recorded forehead Electroencephalogram(EEG) signals. These EEG signals are decomposed to extract intrinsic mode functions(IMFs) by using the EMD technique. Then, the IMFs components are selected out by using the Pearson correlation coefficient and the best scale features on each signal are determined in multiple experiments.

RESULTS

Results indicate that the empirical mode decomposition multi-scale fuzzy entropy feature classification recognition rate is up to 87.50%, the highest is 88.74%, which is 23.88% higher than the single-scale fuzzy entropy and 5.56% higher than multi-scale fuzzy entropy.

COMPARISON WITH EXISTING METHOD

Three types of entropies measures, permutation entropy(PE), sample entropy(SE), fuzzy entropy(FE), were applied for the analysis of signal and compared by seven classifiers in 10-fold and Leave-One-Out cross-validation experiments.

CONCLUSIONS

The proposed method can be effectively applied to the detection of driving fatigue.

Research on Path Tracking Algorithm of Autopilot Vehicle Based on Image Processing

Smart cars are the result of the combination of the latest technological achievements in the fields of artificial intelligence, sensors, control science, computer, and network technology with the modern automobile industry. Intelligent cars usually have functions, such as automatic shifting, automatic driving, and automatic road condition recognition. The research of intelligent car technology involves many disciplines. This thesis focuses on the field of smart car visual navigation, focusing on image denoising, image information recognition, extraction, and pattern recognition control algorithms. The traditional trajectory tracking algorithm is mainly used in industrial computer or high-performance computer. The computational complexity leads to poor real-time control, and it is easily interfered by external complex terrain environment and internal disordered electromagnetic environment during vehicle driving. In general, on a regular basis, by the image analysis of the driver or the driver information, the image information is proposed using way trace processing technology, vehicle tracking control method and automatic driving rules. The simulation and experimental results show that the proposed control methods and rules used to carry out automatic driving vehicle are feasible. The algorithm reduces the complexity of the algorithm, improves the real-time and stability of the control and finally achieves a good trajectory tracking effect of the car on high-speed automatic driving.

Multivariate Approach for Alzheimer's Disease Detection Using Stationary Wavelet Entropy and Predator-Prey Particle Swarm Optimization

BACKGROUND

The number of patients with Alzheimer's disease is increasing rapidly every year. Scholars often use computer vision and machine learning methods to develop an automatic diagnosis system.

OBJECTIVE

In this study, we developed a novel machine learning system that can make diagnoses automatically from brain magnetic resonance images.

METHODS

First, the brain imaging was processed, including skull stripping and spatial normalization. Second, one axial slice was selected from the volumetric image, and stationary wavelet entropy (SWE) was done to extract the texture features. Third, a single-hidden-layer neural network was used as the classifier. Finally, a predator-prey particle swarm optimization was proposed to train the weights and biases of the classifier.

RESULTS

Our method used 4-level decomposition and yielded 13 SWE features. The classification yielded an overall accuracy of 92.73±1.03%, a sensitivity of 92.69±1.29%, and a specificity of 92.78±1.51%. The area under the curve is 0.95±0.02. Additionally, this method only cost 0.88 s to identify a subject in online stage, after its volumetric image is preprocessed.

CONCLUSION

In terms of classification performance, our method performs better than 10 state-of-the-art approaches and the performance of human observers. Therefore, this proposed method is effective in the detection of Alzheimer's disease.

Recognition of Emotional States Using Multiscale Information Analysis of High Frequency EEG Oscillations

Exploring the manifestation of emotion in electroencephalogram (EEG) signals is helpful for improving the accuracy of emotion recognition. This paper introduced the novel features based on the multiscale information analysis (MIA) of EEG signals for distinguishing emotional states in four dimensions based on Russell's circumplex model. The algorithms were applied to extract features on the DEAP database, which included multiscale EEG complexity index in the time domain, and ensemble empirical mode decomposition enhanced energy and fuzzy entropy in the frequency domain. The support vector machine and cross validation method were applied to assess classification accuracy. The classification performance of MIA methods (accuracy = 62.01%, precision = 62.03%, recall/sensitivity = 60.51%, and specificity = 82.80%) was much higher than classical methods (accuracy = 43.98%, precision = 43.81%, recall/sensitivity = 41.86%, and specificity = 70.50%), which extracted features contain similar energy based on a discrete wavelet transform, fractal dimension, and sample entropy. In this study, we found that emotion recognition is more associated with high frequency oscillations (51-100Hz) of EEG signals rather than low frequency oscillations (0.3-49Hz), and the significance of the frontal and temporal regions are higher than other regions. Such information has predictive power and may provide more insights into analyzing the multiscale information of high frequency oscillations in EEG signals.

Assessment of driver drowsiness using electroencephalogram signals based on multiple functional brain networks

This paper proposes a comprehensive approach to explore whether functional brain network (FBN) changes from the alert state to the drowsy state and to find out ideal neurophysiology indicators able to detect driver drowsiness in terms of FBN. A driving simulation experiment consisting of two driving tasks is designed and conducted using fifteen participant drivers. Collected EEG signals are then decomposed into multiple frequency bands by wavelet packet transform (WPT). Based on this, two novel FBN approaches, synchronization likelihood (SL) and minimum spanning tree (MST) are combined and applied to feature recognition and classification system. Unlike other methods, our approaches focus on the interaction and correlation between different brain regions. Statistical analysis of network features indicates that the difference between alert state and drowsy state are significant and further confirmed that brain network configuration should be related to drowsiness. For classification, these brain network features are selected and then fed into four classifiers considered namely Support Vector Machines (SVM), K Nearest Neighbors classifier (KNN), Logistic Regression (LR) and Decision Trees (DT). It is found that combining MST method and SL method is actually increasing the classification accuracy with all classifiers considered in this work especially the KNN classifier from 95.4% to 98.6%. Moreover, KNN classifier also gives the highest precision of 98.3%, sensitivity of 98.8% and specificity of 98.9%. Thus this kind of methodology might be a useful tool for further understanding the neurophysiology mechanisms of driver drowsiness, and as a reference work for future studies or future 'systems'.

Automated detection of driver fatigue based on EEG signals using gradient boosting decision tree model

Driver fatigue is increasingly a contributing factor for traffic accidents, so an effective method to automatically detect driver fatigue is urgently needed. In this study, in order to catch the main characteristics of the EEG signals, four types of entropies (based on the EEG signal of a single channel) were calculated as the feature sets, including sample entropy, fuzzy entropy, approximate entropy and spectral entropy. All feature sets were used as the input of a gradient boosting decision tree (GBDT), a fast and highly accurate boosting ensemble method. The output of GBDT determined whether a driver was in a fatigue state or not based on their EEG signals. Three state-of-the-art classifiers, k-nearest neighbor, support vector machine and neural network were also employed. To assess our method, several experiments including parameter setting and classification performance comparison were performed on 22 subjects. The results indicated that it is possible to use only one EEG channel to detect a driver fatigue state. The average highest recognition rate in this work was up to 94.0%, which could meet the needs of daily applications. Our GBDT-based method may assist in the detection of driver fatigue.

Real-time eye tracking for the assessment of driver fatigue

Eye-tracking is an important approach to collect evidence regarding some participants' driving fatigue. In this contribution, the authors present a non-intrusive system for evaluating driver fatigue by tracking eye movement behaviours. A real-time eye-tracker was used to monitor participants' eye state for collecting eye-movement data. These data are useful to get insights into assessing participants' fatigue state during monotonous driving. Ten healthy subjects performed continuous simulated driving for 1-2 h with eye state monitoring on a driving simulator in this study, and these measured features of the fixation time and the pupil area were recorded via using eye movement tracking device. For achieving a good cost-performance ratio and fast computation time, the fuzzy K-nearest neighbour was employed to evaluate and analyse the influence of different participants on the variations in the fixation duration and pupil area of drivers. The findings of this study indicated that there are significant differences in domain value distribution of the pupil area under the condition with normal and fatigue driving state. Result also suggests that the recognition accuracy by jackknife validation reaches to about 89% in average, implying that show a significant potential of real-time applicability of the proposed approach and is capable of detecting driver fatigue.

Automated Detection of Driver Fatigue Based on AdaBoost Classifier with EEG Signals

Purpose: Driving fatigue has become one of the important causes of road accidents, there are many researches to analyze driver fatigue. EEG is becoming increasingly useful in the measuring fatigue state. Manual interpretation of EEG signals is impossible, so an effective method for automatic detection of EEG signals is crucial needed.

Method: In order to evaluate the complex, unstable, and non-linear characteristics of EEG signals, four feature sets were computed from EEG signals, in which fuzzy entropy (FE), sample entropy (SE), approximate Entropy (AE), spectral entropy (PE), and combined entropies (FE + SE + AE + PE) were included. All these feature sets were used as the input vectors of AdaBoost classifier, a boosting method which is fast and highly accurate. To assess our method, several experiments including parameter setting and classifier comparison were conducted on 28 subjects. For comparison, Decision Trees (DT), Support Vector Machine (SVM) and Naive Bayes (NB) classifiers are used.

Results: The proposed method (combination of FE and AdaBoost) yields superior performance than other schemes. Using FE feature extractor, AdaBoost achieves improved area (AUC) under the receiver operating curve of 0.994, error rate (ERR) of 0.024, Precision of 0.969, Recall of 0.984, F1 score of 0.976, and Matthews correlation coefficient (MCC) of 0.952, compared to SVM (ERR at 0.035, Precision of 0.957, Recall of 0.974, F1 score of 0.966, and MCC of 0.930 with AUC of 0.990), DT (ERR at 0.142, Precision of 0.857, Recall of 0.859, F1 score of 0.966, and MCC of 0.716 with AUC of 0.916) and NB (ERR at 0.405, Precision of 0.646, Recall of 0.434, F1 score of 0.519, and MCC of 0.203 with AUC of 0.606). It shows that the FE feature set and combined feature set outperform other feature sets. AdaBoost seems to have better robustness against changes of ratio of test samples for all samples and number of subjects, which might therefore aid in the real-time detection of driver fatigue through the classification of EEG signals.

Conclusion: By using combination of FE features and AdaBoost classifier to detect EEG-based driver fatigue, this paper ensured confidence in exploring the inherent physiological mechanisms and wearable application.

Noise Robustness Analysis of Performance for EEG-Based Driver Fatigue Detection Using Different Entropy Feature Sets

Driver fatigue is an important factor in traffic accidents, and the development of a detection system for driver fatigue is of great significance. To estimate and prevent driver fatigue, various classifiers based on electroencephalogram (EEG) signals have been developed; however, as EEG signals have inherent non-stationary characteristics, their detection performance is often deteriorated by background noise. To investigate the effects of noise on detection performance, simulated Gaussian noise, spike noise, and electromyogram (EMG) noise were added into a raw EEG signal. Four types of entropies, including sample entropy (SE), fuzzy entropy (FE), approximate entropy (AE), and spectral entropy (PE), were deployed for feature sets. Three base classifiers (K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Decision Tree (DT)) and two ensemble methods (Bootstrap Aggregating (Bagging) and Boosting) were employed and compared. Results showed that: (1) the simulated Gaussian noise and EMG noise had an impact on accuracy, while simulated spike noise did not, which is of great significance for the future application of driver fatigue detection; (2) the influence on noise performance was different based on each classifier, for example, the robust effect of classifier DT was the best and classifier SVM was the weakest; (3) the influence on noise performance was also different with each feature set where the robustness of feature set FE and the combined feature set were the best; and (4) while the Bagging method could not significantly improve performance against noise addition, the Boosting method may significantly improve performance against superimposed Gaussian and EMG noise. The entropy feature extraction method could not only identify driver fatigue, but also effectively resist noise, which is of great significance in future applications of an EEG-based driver fatigue detection system.

Comparison of Different Features and Classifiers for Driver Fatigue Detection Based on a Single EEG Channel

Driver fatigue has become an important factor to traffic accidents worldwide, and effective detection of driver fatigue has major significance for public health. The purpose method employs entropy measures for feature extraction from a single electroencephalogram (EEG) channel. Four types of entropies measures, sample entropy (SE), fuzzy entropy (FE), approximate entropy (AE), and spectral entropy (PE), were deployed for the analysis of original EEG signal and compared by ten state-of-the-art classifiers. Results indicate that optimal performance of single channel is achieved using a combination of channel CP4, feature FE, and classifier Random Forest (RF). The highest accuracy can be up to 96.6%, which has been able to meet the needs of real applications. The best combination of channel + features + classifier is subject-specific. In this work, the accuracy of FE as the feature is far greater than the Acc of other features. The accuracy using classifier RF is the best, while that of classifier SVM with linear kernel is the worst. The impact of channel selection on the Acc is larger. The performance of various channels is very different.