Drowsy State Research Articles

Розробка алгоритма моніторингу стану водія за допомогою Аndroid-застосунку з метою підвищення рівня активної безпеки

The purpose of the research is the analysis of existing systems for monitoring the driver's condition using an Android mobile application to increase the level of active safety with the development of an algorithm for monitoring the driver's condition. At the same time, such parameters as turns and tilts of the head, duration of eyelid blinking, yawning are taken into account. As a result, the state of drowsiness, impaired attention, and drunkenness of the driver is analyzed. The article analyzes the methods of monitoring the dangerous condition of the driver during the movement of the vehicle. The list of programs used in this case is given. The Drive Safely mobile application was chosen for the research, which is based on monitoring the condition of the driver's eyes and mouth. Based on research, eye PERCLOS should not exceed 28% and mouth PERCLOS should not exceed 30%. An algorithm for recognizing emergency situations in the vehicle cabin based on the image of the driver's eyes and mouth from the front camera of a smartphone has been developed. As a result of the analysis of existing Android applications, an algorithm for recognizing emergency situations in the cabin of the vehicle was developed. The algorithm was obtained based on the image of the smartphone's front camera. This algorithm takes into account all possible dangerous conditions of the driver behind the wheel of the vehicle. As a result of the conducted research, an analysis of the existing systems for monitoring the driver's condition using the Drive Safely mobile Android application was carried out. As a result of the analysis of existing Android applications, an algorithm for recognizing emergency situations in the cabin of the vehicle was developed. The use of this algorithm in the program and the connection of its results to the cloud environment will allow it to self-improve and increase the efficiency of its operation, thus increasing the level of active security.

Smart Safety and Accident Prevention System

The primary cause of road accident results in fatalities, serious injuries and monetary losses is known to be due to drowsy or sleepy drivers, according to analysis reports on recent traffic accidents. Lack of sleep, medication, drugs, or prolonged driving contributes to drowsiness. A system that can identify a driver’s drowsy state and warn him before an accident occurs is required to avoid roadside accidents caused by distracted driving. Many researchers have recently expressed their interest in drowsiness detection. The methods essentially involve monitoring the driver’s physiological or behavioral 1summarizes some of the most recent methods put forth in this field is given. We propose an algorithm to monitor eye blinks that uses eye feature points to determine whether the eye is open or closed and sets off an alarm if the driver is drowsy. In-depth experimental results are also provided to highlight the benefits and drawbacks of the proposed method.

Isotropic Self-Supervised Learning for Driver Drowsiness Detection With Attention-Based Multimodal Fusion

Driverdrowsiness is an important cause of traffic accidents. Many studies using computer vision techniques to detect driver drowsiness states, such as slow blinking, yawning, and nodding, have demonstrated excellent potential. Although existing studies have made significant progress, the number of samples in the training corpora is small, which makes it difficult for a model to learn effective drowsiness representations from images or videos. To address this issue, we develop an isotropic self-supervised learning (IsoSSL) approach to learn powerful representations of images without relying on human-provided annotations and propose an IsoSSL-MoCo model by combining IsoSSL with momentum contrast (MoCo). To exploit the complementarity of multimodal data, an attention-based multimodal fusion model is also proposed to fuse features from the eye, mouth, and optical flow of the head. Specifically, we first use the IsoSSL-MoCo model to pretrain the image encoders for the three modalities in other datasets. Then, these encoders are fine-tuned and integrated into the proposed fusion model. The feature vectors generated by the image encoders of the three modalities are fed into the recursive layer to extract temporal information. To capture the importance degrees of the effects of temporal features from the three modalities on drowsiness detection, an attention mechanism is introduced to automatically weigh the feature vectors from the recursive layer to improve detection accuracy. Finally, a vector representation is generated by the attention layer and is used to detect driver drowsiness states. Experimental results based on two challenging datasets show that our method outperforms the baseline methods and the latest existing methods.

PERCLOS-based technologies for detecting drowsiness: current evidence and future directions.

Drowsiness associated with sleep loss and circadian misalignment is a risk factor for accidents and human error. The percentage of time that the eyes are more than 80% closed (PERCLOS) is one of the most validated indices used for the passive detection of drowsiness, which is increased with sleep deprivation, after partial sleep restriction, at nighttime, and by other drowsiness manipulations during vigilance tests, simulated driving, and on-road driving. However, some cases have been reported wherein PERCLOS was not affected by drowsiness manipulations, such as in moderate drowsiness conditions, in older adults, and during aviation-related tasks. Additionally, although PERCLOS is one of the most sensitive indices for detecting drowsiness-related performance impairments during the psychomotor vigilance test or behavioral maintenance of wakefulness test, no single index is currently available as an optimal marker for detecting drowsiness during driving or other real-world situations. Based on the current published evidence, this narrative review suggests that future studies should focus on: (1) standardization to minimize differences in the definition of PERCLOS between studies; (2) extensive validation using a single device that utilizes PERCLOS-based technology; (3) development and validation of technologies that integrate PERCLOS with other behavioral and/or physiological indices, because PERCLOS alone may not be sufficiently sensitive for detecting drowsiness caused by factors other than falling asleep, such as inattention or distraction; and (4) further validation studies and field trials targeting sleep disorders and trials in real-world environments. Through such studies, PERCLOS-based technology may contribute to preventing drowsiness-related accidents and human error.

Effects of using a snooze alarm on sleep inertia after morning awakening

BackgroundMany people use the snooze function of digital alarm clocks for morning awakening, but the effects of a snooze alarm on waking are unclear. We examined the effects of a snooze alarm on sleep inertia, which is a transitional state characterized by reduced arousal and impaired cognitive and behavioral performance immediately upon awakening.MethodsIn study 1, healthy Japanese university students responded to a sleep survey during a psychology class (study 1), and we collected 293 valid responses. In study 2, we compared a separate sample of university students (n = 10) for the effects of using or not using a snooze alarm on sleep inertia immediately after awakening from normal nocturnal sleep in a sleep laboratory.ResultsOf 293 valid respondents in study 1, 251 often used a tool to wake up in the morning (85.7%). Moreover, 70.5% reported often using the snooze function of their mobile phones, mainly to reduce anxiety about oversleeping. Study 2 indicated no differences in the sleep quality or quantity before awakening with or without the snooze alarm, except in the last 20 min. However, during the last 20 min of sleep with snooze alarm, the snooze alarm prolonged waking and stage N1 sleep. Stage N1 sleep is non-rapid eye movement sleep that is primarily defined as a drowsy state. Furthermore, Global Vigor values were enhanced after awakening compared to pre-sleep in the no-snooze condition.ConclusionsUsing a snooze alarm prolongs sleep inertia compared to a single alarm, possibly because snooze alarms induce repeated forced awakenings.

DriverSVT: Smartphone-Measured Vehicle Telemetry Data for Driver State Identification

One of the key functions of driver monitoring systems is the evaluation of the driver’s state, which is a key factor in improving driving safety. Currently, such systems heavily rely on the technology of deep learning, that in turn requires corresponding high-quality datasets to achieve the required level of accuracy. In this paper, we introduce a dataset that includes information about the driver’s state synchronized with the vehicle telemetry data. The dataset contains more than 17.56 million entries obtained from 633 drivers with the following data: the driver drowsiness and distraction states, smartphone-measured vehicle speed and acceleration, data from magnetometer and gyroscope sensors, g-force, lighting level, and smartphone battery level. The proposed dataset can be used for analyzing driver behavior and detecting aggressive driving styles, which can help to reduce accidents and increase safety on the roads. In addition, we applied the K-means clustering algorithm based on the 11 least-correlated features to label the data. The elbow method showed that the optimal number of clusters could be either two or three clusters. We chose to proceed with the three clusters to label the data into three main scenarios: parking and starting driving, driving in the city, and driving on highways. The result of the clustering was then analyzed to see what the most frequent critical actions inside the cabin in each scenario were. According to our analysis, an unfastened seat belt was the most frequent critical case in driving in the city scenario, while drowsiness was more frequent when driving on the highway.

Temporal intermittent rhythmic theta activity (TIRTA): A marker of epileptogenicity?

To describe a novel EEG rhythm, temporal intermittent rhythmic theta activity (TIRTA), and its potential association with epilepsy. We report TIRTA on scalp EEG in a series of 12 patients, all of whom were found to have epilepsy. The clinical and electroencephalographic characteristics of each patient were reviewed. In addition, features that may distinguish TIRTA from benign EEG patterns, including rhythmic temporal theta bursts of drowsiness (RTTBD), were identified. TIRTA was unilateral in all cases. For all patients, TIRTA was seen in the awake and drowsy states. Eight patients also had TIRTA observed during N2 sleep. The average frequency of TIRTA was 5.5Hz and the average duration of a train of TIRTA was 5.25s. In seven cases the morphology was notched in appearance. Temporal intermittent rhythmic delta activity (TIRDA) was seen in seven patients on the same side as TIRTA. Eleven patients also had ipsilateral temporal sharp waves. Abnormal MRI (6/12) and or PET (5/5) findings were ipsilateral to TIRTA. In this preliminary report we suggest that TIRTA may be a novel marker of potential epileptogenicity, possibly representing a higher frequency variant of TIRDA.

Driving sleepiness detection using electrooculogram analysis and grey wolf optimizer

<span lang="EN-US">In modern society, providing safe and collision-free travel is essential. Therefore, detecting the drowsiness state of the driver before its ability to drive is compromised. For this purpose, an automated hybrid sleepiness classification system that combines the artificial neural network and gray wolf optimizer is proposed to distinguish human Sleepiness and fatigue. The proposed system is tested on data collected from 15 drivers (male and female) in alert and sleep-deprived conditions where physiological signals are used as sleep markers. To evaluate the performance of the proposed algorithm, k-nearest neighbors (k-NN), support vector machines (SVM), and artificial neural networks (ANN) classifiers have been used. The results show that the proposed hybrid method provides 99.6% accuracy, while the SVM classifier provides 93.0% accuracy when the kernel is (RBF) and outlier (0.1). Furthermore, the k-NN classifier provides 96.7% accuracy, whereas the standalone ANN algorithm provides 97.7% accuracy.</span>

Misdiagnosis of Tuberculosis

Introduction: High incidence of tubercular cases in developingcountries, similar clinical presentation of tuberculosis andmalignancy, lack of adequate infrastructure, and lack ofproper follow up are the most common factors associated withmisdiagnosis.Case Summary: A 72 years old male CAME with complainsof cough with expectoration for 2 month, intermittent feverassociated with chills and rigors for 1 month, loss of appetiteand weight, intermittent pain in right side of the chest for1 month. Patient was a known smoker and alcoholic. Onauscultation bilateral air entry was decreased with occasionalrhonchi. Patient’s chest x-ray was suggestive of consolidationin right upper zone of chest. His sputum was negative forAFB from outside, patient started on ATT, 2 months prior todate of admission from private clinic on empirical basis thatwas continued till the date of admission. Patient showed noimprovement, even after 2 months of ATT. His chest X-Ray wassuggestive of right upper zone consolidation, patient had noprior radiography for comparison. He was continued on ATTas he had already taken treatment for 2 months from outside.Patient was tested negative for sputum AFB 1 and 2, CBNAATand culture.Patient presented to emergency after a gap of two monthsin state of drowsiness with complains of breathlessness andsevere right side chest pain. He was intubated in emergencyand managed on ventilator support. Patient recovered and aCECT thorax was done suggestive of likely possibility of rightlung upper lobe neoplasm with metastatic lymphadenopathywith possibility of multiple liver and bilateral adrenal metastasisand changes of chronic airways disease with fibro bullouschanges in bilateral upper lobe apical segments. The FNACwas positive for malignant cells with features suggestive of nonsmallcell carcinoma – (possibly adenocarcinoma). Patient wasfurther managed in the department of Radiation Oncology bychemotherapy. We discuss the radiographical progression andsubsequent investigations required to make a proper diagnosis.Conclusion: It is imperative to form a conclusive diagnosisusing available diagnostic modalities in smear negativetubercular cases to avoid delay in management of otherpossible life-threatening diseases like lung carcinomas.

LSTM-CNN model of drowsiness detection from multiple consciousness states acquired by EEG

This study aimed to design a deep neural network for electroencephalography (EEG)-based drowsiness detection in multiple consciousness states, i.e., “awake,” “sleep,” and “drowsiness.” Few studies have seriously considered the optimal input vector size or labeling method in classifying multiple consciousness states, which may affect classification performance. To determine the optimal input vector length, i.e., window length, three neural network models (long short-term memory [LSTM], convolutional neural network [CNN], and combined LSTM and CNN) and four feature-based models were tested with six different levels of window length. The EEG dataset was acquired from 19 participants with randomly assigned auditory stimuli and button responses. The EEG data were labeled into three classes (awake, sleep, and drowsiness) based on the defined button response pattern corresponding to the stimuli. The results demonstrated that when the input vector size exceeded 8 sec, the performance of the neural network models dropped rapidly; however, when the window size was less than 8 sec, the performance change according to the window size was small. In contrast, the performance of feature-based models increased continuously as the window size increased. The LSTM model yielded the best accuracy (86%) for a 1 sec window length, and the LSTM-CNN model yielded the best kappa index (0.77) for a 4 sec window length. In addition, the proposed model was applied to the binary classification of normal consciousness (awake) and low consciousness (drowsiness and sleep) states to determine whether this model works appropriately in actual applications such as drowsiness detection in a driving environment. For binary classification, the LSTM-CNN model resulted in 0.95 F1 scores in 4000-ms. When a short input data (500 msec) is used, the LSTM-CNN model resulted in an average accuracy of 85.6% and a kappa index of 0.77 for the three-class classification problem and 0.94 F1 scores for the binary classification problem. In conclusion, we demonstrated that the proposed model could effectively detect drowsiness. Furthermore, a significant correlation was found between reaction time and drowsiness. However, using the reaction time as an index for labeling drowsiness was challenging because of the high false-negative ratio.

Skin-to-skin SDF positioning: The key to intersubjective intimacy between mother and very preterm newborn-A pilot matched-pair case-control study.

BackgroundSkin-to-skin contact (SSC) has been widely studied in NICU and several meta-analyses have looked at its benefits, for both the baby and the parent. However, very few studies have investigated SSC’ benefits for communication, in particular in the very-preterm newborn immediately after birth.AimsTo investigate the immediate benefits of Supported Diagonal Flexion (SDF) positioning during SSC on the quality of mother—very-preterm newborn communication and to examine the coordination of the timing of communicative behaviors, just a few days after birth.Subjects and study designMonocentric prospective matched-pair case-control study. Thirty-four mothers and their very preterm infants (27 to 31 + 6 weeks GA, mean age at birth: 30 weeks GA) were assigned to one of the two SSC positioning, either the Vertical Control positioning (n = 17) or the SDF Intervention positioning (n = 17). Mother and newborn were filmed during the first 5 min of their first SSC.Outcome measuresInfants’ states of consciousness according to the Assessment of Preterm Infants’ Behavior scale (APIB). Onset and duration of newborns’ and mothers’ vocalizations and their temporal proximity within a 1-s time-window.ResultsIn comparison with the Vertical group, very preterm newborns in the SDF Intervention Group spent less time in a drowsy state and more in deep sleep. At 3.5 days of life, newborns’ vocal production in SSC did not differ significantly between the two groups. Mothers offered a denser vocal envelope in the SDF group than in the Vertical group and their vocalizations were on average significantly longer. Moreover, in a one-second time-frame, temporal proximity of mother-very preterm newborn behaviors was greater in the SDF Intervention Group.ConclusionAlthough conducted on a limited number of dyads, our study shows that SDF positioning fosters mother-very preterm newborn intimate encounter during the very first skin to skin contact after delivery. Our pioneer data sheds light on the way a mother and her very preterm vocally meet, and constitutes a pilot step in the exploration of innate intersubjectivity in the context of very preterm birth.

An intelligent contactless brake blending system with advanced driver assistance technique for electric vehicles

PurposeElectric vehicles are well known for a silent and smooth drive; however, their presence on the road is difficult to identify for road users who may be subjected to certain incidences. Although electric vehicles are free from exhaust emission gases, the wear particles coming out from disc brakes are still unresolved issues. Therefore, the purpose of the present paper is to introduce a smart eco-friendly braking system that uses signal processing and integrated technologies to eventually build a comprehensive driver assistance system.Design/methodology/approachThe parameters obstacle identification, driver drowsiness, driver alcohol situation and heart rate were all taken into account. A contactless brake blending system has been designed while upgrading a rapid response. The implemented state flow rule-based decision strategy validated with the outcomes of a novel experimental setup.FindingsThe drowsiness state of drivers was successfully identified for the proposed control map and set up vindicated with the improvement in stopping time, atmospheric environment and increase in vehicle active safety regime.Originality/valueThe present study adopted a unique approach and obtained a brake blending system for improved braking performance as well as overall safety enhancement with rapid control of the vehicle.

A new hybrid and optimized algorithm for drivers’ drowsiness detection

When the roads are monotonous, especially on the highways, the state of vigilance decreases and the state of drowsiness appears. Drowsiness is defined as the transitional phase from the awake to the sleepy state. However, In Morocco, the majority of fatal accidents on the highway are caused by drowsiness at the wheel, reaching 33.33% rate. Therefore, we proposed the conception and realization of an automatic method based on electroencephalogram (EEG) signals that can predict drowsiness in real time. The proposed work is based on time-frequency analysis of EEG signals from a single channel (FP1-Ref), and drowsiness is predicted using a personalized and optimized machine learning model (optimized decision tree classification method) under Python. The results are much significant and optimized, improving the accuracy from 95.7% to 96.4% and a time consuming from 0.065 to 0.053 seconds.

Real-Time Driver Drowsiness Detection System Using Dlib based on Driver Eye/Mouth Monitoring Technology

The driver sleepiness is common and one of the main causes of road accidents. So, there is a need for automatically detecting of this human behaviour. In any case, if the drivers feel drowsy, they still keep on driving the vehicle, and accidents occur. This study can be implemented using the CNN training model and initiating an alarm if the drowsiness condition is detected. Many of the authors suggested the process of detecting the Drowsiness (Problem Statement) of the drivers using technologies like the Internet of Things (IoT), Deep learning, and Haar cascade (to detect the coordinates of eyes and mouth, which are the target objects). However, this study contributes towards providing the real-time application in co-operating the CNN model and Dlib. Hence, this study proposes a novel embedded system with CNN technology. The CNN model is fed with inputs based on four (4) images related to eye and mouth openings and closings. This application is trained using the CNN model, which takes inputs as images and processes by identifying the features on the face using the Dlib library while representing the change in the state of coordinates of eyes and mouth as Yawning. This approach is achieved using Convolution Neural Network (CNN), pillow, Pygame, OpenCV, and the Dlib, along with providing an alarm when the position of mouth changes. The model is recorded with a maximum validation accuracy of 98% with the minimum recorded loss of less than 0.04% as areal-time application.

Drowsiness Detection System Using Deep Learning Methods

Drowsiness and driver fatigue cause a great percentage of road accidents. The number of deaths and fatalities injuries due to sleeplessness is increasing every year globally. The drowsy state, caused due to no breaks during a long-distance journey can be detected with an attention assist system, a method that can warn inactiveness and warn the driver about the current state of fatigue by detecting eyes as the region of interest. Our system will locate, track and analyze the driver’s eyes and create an alert for the navigation system. The monitoring system is a real-time system that will use machine learning and computer vision to detect driver fatigue and distraction and efficiently avoid and reduce road-accidents caused. The model is able to predict the drowsiness of driver. It is able to analyze whether the eyes are open or closed and face not detected and displayed the intended output with 99.8% accuracy.

Drowsiness Detection System Based on Machine Learning Using Eye State

Drowsiness is one of the major causes of driver-induced traffic accidents. The interactive systems developed to reduce road accidents by alerting drivers is called as Advanced Driver Assistance Systems (ADAS). The most important ADAS are Lane Departure Warning System, Front Collision Warning System and Driver Drowsiness Systems. In this study, an ADAS system based on eye state detection is presented to detect driver drowsiness. First, Viola-Jones algorithm approach is used to detect the face and eye areas in the proposed method. The detected eye region is classified as closed or open by making use of a machine learning method. Finally, the eye conditions are analyzed at time domain with PERcentage of eyelid CLOsure (PERCLOS) metric and drowsiness conditions are determined by Support Vector Machine (SVM), kNN and decision tree classifiers. The proposed methods tested on 7 real people and drowsiness states are detected at 99.77%, 94.35%, and 96.62% accuracy, respectively.

Drowsiness Detection Using Ocular Indices from EEG Signal.

Drowsiness is one of the main causes of road accidents and endangers the lives of road users. Recently, there has been considerable interest in utilizing features extracted from electroencephalography (EEG) signals to detect driver drowsiness. However, in most of the work performed in this area, the eyeblink or ocular artifacts present in EEG signals are considered noise and are removed during the preprocessing stage. In this study, we examined the possibility of extracting features from the EEG ocular artifacts themselves to perform classification between alert and drowsy states. In this study, we used the BLINKER algorithm to extract 25 blink-related features from a public dataset comprising raw EEG signals collected from 12 participants. Different machine learning classification models, including the decision tree, the support vector machine (SVM), the K-nearest neighbor (KNN) method, and the bagged and boosted tree models, were trained based on the seven selected features. These models were further optimized to improve their performance. We were able to show that features from EEG ocular artifacts are able to classify drowsy and alert states, with the optimized ensemble-boosted trees yielding the highest accuracy of 91.10% among all classic machine learning models.

Detection of Driver Dynamics with VGG16 Model

Abstract One of the most important factors triggering the occurrence of traffic accidents is that drivers continue to drive in a tired and drowsy state. It is a great opportunity to regularly control the dynamics of the driver with transfer learning methods while driving, and to warn the driver in case of possible drowsiness and to focus their attention in order to prevent traffic accidents due to drowsiness. A classification study was carried out with the aim of detecting the drowsiness of the driver by the position of the eyelids and the presence of yawning movement using the Convolutional Neural Network (CNN) architecture. The dataset used in the study includes the face shapes of drivers of different genders and different ages while driving. Accuracy and F1-score parameters were used for experimental studies. The results achieved are 91 % accuracy for the VGG16 model and an F1-score of over 90 % for each class.

Driver Drowsiness Detection System

Abstract: Recently, in addition to autonomous vehicle technology research and development, machine learning methods have been used to predict a driver's condition and emotions in order to provide information that will improve road safety. A driver's condition can be estimated not only by basic characteristics such as gender, age, and driving experience, but also by a driver's facial expressions, bio-signals, and driving behaviours. Recent developments in video processing using machine learning have enabled images obtained from cameras to be analysed with high accuracy. Therefore, based on the relationship between facial features and a driver’s drowsy state, variables that reflect facial features have been established. In this paper, we proposed a method for extracting detailed features of the eyes, the mouth, and positions of the head using OpenCV and Dlib library in order to estimate a driver’s level of drowsiness. Keywords: Drowsiness, OpenCV, Dlib, facial features, video processing

Driver drowsiness detection system using machine learning

This feature is used for the safety of the peoples from the accidents. This work shows the development of ADAS ( advance driving assistance system) this focus on the driver drowsiness detection is main aim to alert the driver from the drowsiness state to avoid road accidents. In this we use the Machine Learning to predict the condition and emotions of the driver that will improve the safety on the roads. We use the CNN for this project that will effectively detect the driver fatigue status using driver images. In this project we can use the Electrocardiogram (ECG) for the psychological system to detect drowsiness. Artificial Intelligence means system can automatically learns as well as improve without being programmed. We also use the various machine learning techniques like OpenCv, Keras, tensorflow