Calibration-Free Driver Drowsiness Classification based on Manifold-Level Augmentation

Driver fatigue is one of the most common reasons for deadly road accidents around the world. Continuous monotonous driving for long hours without rest causes drowsiness and consequently fatal road accidents. Automatic driver drowsiness detection can prevent a vast number of sleep persuaded road accidents, and hence can save precious lives. Number of techniques for driver drowsiness detection has been examined in the recent past. This paper presents a survey of these techniques. These techniques detect the driver drowsiness by observing the driving pattern. Abnormalities in driving pattern is hypothesized as a drowsiness state of driver. Various measures such as subjective, behavioral, physiological, and vehicular have been used for this purpose. The comparative analysis of these techniques indicates that behavioral measures are easy to acquire and does not disturb the driver as they are non-invasive. Among various behavioral measures, head movement measure is found to be most precise and effective.

Driver fatigue causes serious damages amongst all other road accidents. Around 20% of fatal road accidents involve driver fatigue. This paper describes a modern approach which will detect driver fatigue considering most of the fatigue symptoms. Eye closure, yawning, head tilting is the major symptoms of fatigue behavior. Inattentive vehicle movement in the road under fatigue condition is also accountable for driver fatigue. The goal of this paper is to detect these symptoms for better driving condition in the road. These symptoms are monitored by two cameras. This paper proposes a robust system where facial expressions, head tilting and lane departure for fatigue will be detected collectively. Experimental results of the proposed method are compared with the previous method. The results show good accuracy and reliable performance to avoid road accidents compared to the previous method. The proposed system is very simple and avoids any complexity.

A computer vision-based system called the Drowsiness Sensing Device using OpenCV was designed to identify driver drowsiness. The technology uses video frames from a camera positioned inside a car to identify different sleepiness indicators, including the length of eye closure and head position. The Eye Aspect Ratio (EAR), which aids in trying to assess drowsiness, is determined using the OpenCV library, which is also used to extract feature points and detect eye blinks. The system also has an alarm mechanism that sounds when a certain level of drowsiness is attained, alerting the driver to take the appropriate action. The proposed approach can be possibly employed to reduce the number of accidents occurred due to driver drowsiness. The suggested system is a real-time drowsiness sensing system that makes use of OpenCV to gauge a person's level of drowsiness. The technology employs a camera to take pictures of the driver's face, assessing the features like the mouth and eyes to determine how sleepy they are. The system can identify drowsiness by noticing changes in the eyes, such as drooping eyelids, and mouth movements, such as yawning. When the amount of drowsiness surpasses a predetermined threshold, the system informs the driver by assessing the photos using machine learning techniques. By prompting the driver to take a break, the proposed technology may help prevent accidents brought on by drowsy driving.

Safety driving performance estimation and alertness (SDPEA) has drawn the attention of researchers in preventing traffic accidents caused by drowsiness while driving. Psychophysiological measures, such as electroencephalogram (EEG), are accurately investigated to be robust candidates for drivers’ drowsiness evaluation. This paper presents an effective EEG-based driver drowsiness monitoring system by analyzing the changes of brain activities in a simulator driving environment. The proposed SDPEA system can translate EEG signals into drowsiness level. Firstly, Independent component analysis (ICA) is performed on EEG data to remove artifacts. Then, eight EEG-band powers-related features: beta, alpha, theta, delta, (alpha plus theta)/beta, alpha / beta, (alpha plus theta)/(alpha plus beta) and theta / beta are extracted from the preprocessed EEG signals by employing the Fast Fourier Transform (FFT). Subsequently, fisher score technique selects the most descriptive features for further classification. Finally, Support Vector Machine (SVM) is employed as a classifier to distinguish drowsiness level. Experimental results show that the quantitative driving performance can be correctly estimated through analyzing driver’s EEG signals by the SDPEA system.

Summary form only given. This lecture will introduce the world's first method of detecting driver's drowsiness level from the facial expression. The motivation is to realize the novel safety system which can detect the driver's slight drowsiness and keep the driver awake while driving. The brain wave, especially alpha wave, is commonly used as the drowsiness index. However, it is not suitable for the in-vehicle system since it is measured with electrodes or sensors worn over the head. We precisely investigated the relationship between the change of brain wave and other drowsiness indexes that can be measured without any contact; drowsiness level evaluated from facial expression, low-frequency power content of eye-movement, fluctuation of heart rate, and distance of lane deviation from the center. This lecture will show you a process how we found the features of drowsiness from human and developed the method of drowsiness detection without any contact sensors.

Every year thousands of people lose their life due to road accidents. One of the main reasons for these accidents is driver drowsiness. In driver drowsiness, the driver slept while driving, which causes the road accident, especially on the long routes. Driver fatigue and micro sleep while driving caused the fatal accident and death of human beings. To overcome this problem, we are implementing a technique in which we capture the image of the driver. After capturing the image of the driver, we process driver images to detect driver drowsiness. For the processing of the driver image, we are using two different techniques with each other. In the first technique, we are using the Dlib for image drowsiness detection by detecting that driver’s eyes are closed and the driver is yawning. In the second technique, we used CNN for the detection of yawning and the eyes of the driver are closed or not and predict driver drowsiness. After implementing the two techniques we combine the output of both techniques. After combining both techniques we test the system, and it gives us very good results.

As per the previous year’s report concerning to road crashes indicates that the principal cause of such a fatal road accidents is because of negligence behavior as well as drowsiness of driver. This problem reveals the requirement of such a system that can recognize drowsiness state of driver and gives alert signal to the driver before the occurrence of any accidents. Therefore, this proposed work has established drowsy detection as well as accident avoidance system based on the eye blink duration. Here, first the open and close state of eye are detected based on the eye aspect ratio (EAR). Further, the blink duration or count during the changes of eye state from open to close are analyzed. Then, it identifies the state of drowsiness, when blink duration becomes more than a certain limits and sends the alert message to the driver through the alarm. Our developed system has shown the accuracy of 92.5 % approx on yawning dataset (YawDD).

Background: Purpose of this study is to monitor the fatal road accidents (FRA) in the Region of Crete, Greece and capture their dynamics in time and space using the Geographical Information System (GIS) technology. It aims to record the FRA spatially from 2001 to 2012, predict their spatio-temporal variance, estimate the number of FRA that should be expected the next years per region and identify the high risk areas. Methods: It is a spatio-temporal study using data from the National Emergency Center’s database. The SPSS 20 and the Arc map 10 were used for the analysis. Spatio-temporal models were applied; specifically, geographical descriptive, Geary’s C, co-kriging interpolation and the Geographical Weighted regression model. Results: According to the Geary’s C, FRA follow a clustered pattern in Crete, whilst they are not randomly occurred (Geary’s C= 0.42; 95%CI= 0.029-0.873; pvalue<0.001). There was a total of 1,039 FRA cases that presented heterogeneous distribution on the island, gathering within the standard distance and ellipse. Time related factors and age were found to be significant to the risk for FRA (pvalue<0.001), [summer months: ExpB=3.43, 95%CI=1.726-5.027 and the night hours: ExpB=2.43; 1.304-4.487]. High risk areas were identified and the expected number of unrecorded FRA was found to vary from 0.0001 to 5.5 cases per 50km2. Conclusions: The present study inserts, for the first time in the Greek bibliography, a new way of monitoring and capturing the FRA dynamics and highlights the use of the GIS technology and dynamic modeling.

Background: High risk behaviors such as substance use are among the human factors that can play an important role in fatal road accidents and deaths. Objective: To determine the status of alcohol consumption based on vitreous fluid exam in cadavers sent to Tehran Legal Medicine Organization. Materials and Methods: In this descriptive analytical cross-sectional study, all fatalities sent to the Tehran Legal Medicine Organization during 2018 were studied. A vitreous sample was used to measure the amount of alcohol. Results: Out of 676 cadavers (85% male; mean age 40.6±6.45), the toxicology results for alcohol was positive in 6.51% (91% male; mean age 37.10±15.7). There was a significant difference between the two sexes (P < 0.00) and the educational level (P1< 0.00). Most fatal accidents occurred in positive toxicology cases between 10-12 pm and on Fridays followed by Thursdays. Conclusion: The policy of non-consumption of alcohol based on Islamic teachings in Iran has played an important role in significantly reducing fatal accidents caused by alcohol.

The existing indicators of the industry incident «Improving the efficiency of managing the system of measures to reduce adult mortality», which are derived from the number of fatalities in road accidents, do not reflect the overall picture of mortality from injuries and external causes in the Russian Federation. In this regard, in order to reduce mortality from manageable causes, it seems appropriate to use indicators calculated in the group of victims with combined, multiple and isolated injuries accompanied by shock. The indicators should characterize compliance with routing for the specified group of victims and the timeliness of the start of anti-shock therapy. The purpose of the study is considering the feasibility of making changes to the structure of the industry incident in terms of indicators of mortality from injuries and external causes. Materials and methods. Mortality rates from injuries and external causes in the Russian Federation for the period 2020-2023 were studied, and proposals were developed to improve their monitoring within the framework of an industry incident. The results of the study. Mortality from injuries and external causes in the Russian Federation continues to be high: in 2023, this figure was 107.1 per 100,000 population, including mortality from road accidents — 9.2 per 100,000 population (8.6%). In order to improve the efficiency of managing the system of measures to reduce mortality in the adult population, it is advisable to include in the structure of the industry Incident indicators that allow assessing compliance with predetermined routing and the timeliness of providing specialized medical care to all victims with combined, multiple and isolated injuries accompanied by shock.

A system for driver drowsiness monitoring is proposed, using multi-sensor data acquisition and investigating two decision-making algorithms, namely a fuzzy inference system (FIS) and an artificial neural network (ANN), to predict the drowsiness level of the driver. Drowsiness indicator signals are selected allowing non-intrusive measurements. The experimental set-up of a driver-drowsiness-monitoring system is designed on the basis of the sought-after indicator signals. These selected signals are the eye closure via pupil area measurement, gaze vector and head motion acquired by a monocular computer vision system, steering wheel angle, vehicle speed, and force applied to the steering wheel by the driver. It is believed that, by fusing these signals, driver drowsiness can be detected and drowsiness level can be predicted. For validation of this hypothesis, 30 subjects, in normal and sleep-deprived conditions, are involved in a standard highway simulation for 1.5h, giving a data set of 30 pairs. For designing a feature space to be used in decision making, several metrics are derived using histograms and entropies of the signals. An FIS and an ANN are used for decision making on the drowsiness level. To construct the rule base of the FIS, two different methods are employed and compared in terms of performance: first, linguistic rules from experimental studies in literature and, second, mathematically extracted rules by fuzzy subtractive clustering. The drowsiness levels belonging to each session are determined by the participants before and after the experiment, and videos of their faces are assessed to obtain the ground truth output for training the systems. The FIS is able to predict correctly 98 per cent of determined drowsiness states (training set) and 89 per cent of previously unknown test set states, while the ANN has a correct classification rate of 90 per cent for the test data. No significant difference is observed between the FIS and the ANN; however, the FIS might be considered better since the rule base can be improved on the basis of new observations.

EEG (Electroencephalogram) signals are used as effective indicators for driver fatigue evaluation. Usually, EEG signals are collected by high-cost electroencephalographs which can record multi-channels of EEG data. This paper is aimed to determine interested electrodes of the relatively optimal EEG based indicator. Firstly, 16 channels of EEG data are collected and transformed into three bands (θ, α, β). Secondly, 12 types of EEG-based parameters are established and TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) is introduced to determine the optimal indicator of driver fatigue. Thirdly, the interested electrodes (Fp1, F7) are determined by MDS (multidimensional scaling). Finally, the evaluation model of driver fatigue is established with the indicator of Fp1 and F7. Experiment results verify that the efficiency has no significant difference between evaluations with two electrodes and 16 electrodes. In addition, the established evaluation model is proven to be efficient to evaluate the driver fatigue.

SummaryAll new vehicle types within the European Union must now be equipped with a driver drowsiness and attention warning system starting from 2022. The specific requirements for the test procedure necessary for type approval are defined in the Annex of EU Regulation C/2021/2639. The objectives of this study were to: (i) investigate how sleepiness develops in professional truck drivers under real‐road driving conditions; and (ii) assess the feasibility of a test procedure for validating driver drowsiness and attention warning systems according to the EU regulation. Twenty‐four professional truck drivers participated in the test. They drove for 180 km on a dual‐lane motorway, first during daytime after a normal night's sleep and then at nighttime after being awake since early morning. The results showed higher sleepiness levels during nighttime driving compared with daytime, with a faster increase in sleepiness with distance driven, especially during the night. Psychomotor vigilance task results corroborated these findings. From a driver drowsiness and attention warning testing perspective, the study design with sleep‐deprived drivers at night was successful in inducing the targeted sleepiness level of a Karolinska Sleepiness Scale score of ≥ 8. Many drivers who reported a Karolinska Sleepiness Scale ≥ 8 during the drives also acknowledged feeling sleepy in the post‐drive questionnaire. Reaching high levels of sleepiness on real roads during daytime is more problematic, not the least from legal and ethical perspectives as higher traffic densities during the daytime lead to increased risks.

Human-machine systems designed for daily activities must balance direct user control with autonomous system behaviour to reduce cognitive effort while maintaining responsiveness. Douglas et al. have demonstrated how shared-autonomy frameworks can enable BCIs to operate across multiple levels of user involvement, from direct neural control to high-level goal specification. Rather than relying solely on low-level EEG commands, the proposed approach integrates intent inference with adaptive robotic planning, allowing multiple users to coordinate multiple robots in functional daily tasks. A key insight from this contribution is that increasing autonomy can significantly reduce user workload while preserving task accuracy and efficiency, supporting more sustainable long-term BCI use in real-world settings. This work highlights a broader shift in BCI design toward collaborative control paradigms, where machines proactively assist users instead of acting only as passive command receivers.Single-modality EEG BCIs are often limited by noise, signal ambiguity, and performance variability across users and environments. Hybrid approaches that fuse EEG with additional sensing modalities offer a promising solution, as demonstrated by Coutray and their team. Coutray et al. have proven how integrating EEG with eye-tracking improves command disambiguation, interaction speed, and overall control reliability, particularly in immersive virtual reality environments. This hybrid design reduces reliance on high-precision EEG decoding alone and enables more natural, intuitive hands-free interaction, expanding BCI applicability beyond clinical contexts into entertainment, accessibility, and smartenvironment control. This contribution reinforces the growing consensus that multimodal BCIs are more scalable, resilient, and user-friendly than EEG-only systems. Robust BCI performance depends not only on decoding algorithms but also on the user's cognitive and neural state. Mohamed et al. have provided evidence that baseline neural oscillations, particularly precue alpha activity, influence event-related desynchronisation (ERD) strength, a core signal used in motorimagery BCIs. A concrete implication of this finding is that BCIs could dynamically adjust classification thresholds, training protocols, or feedback timing based on real-time cognitive state estimates, potentially improving accuracy, consistency, and user learning rates. These insights support the development of state-aware BCIs that adapt decoding strategies in response to moment-to-moment brain dynamics, improving reliability in naturalistic, everyday environments.Beyond EEG decoding alone, combining physiological stress markers and subjective workload measures enables more comprehensive prediction of user performance in demanding tasks. Wei et al. have demonstrated that interpretable machine-learning models can forecast task outcomes while revealing which physiological factors most strongly influence performance. This is particularly relevant for realworld BCI deployment, where fatigue, stress, and cognitive overload can degrade system reliability. A key insight is that interpretable predictive models can support adaptive intervention strategies, such as adjusting task difficulty or providing rest prompts, while maintaining transparency and user trust.Across the contributions, several overarching themes emerge:1. Multimodal integration: Combining EEG with complementary sensors enhances decoding robustness and interaction flexibility. 2. Adaptive autonomy: Systems that balance user intent with automated assistance reduce workload and improve task efficiency. 3. Machine-learning innovation: Deep and interpretable models improve accuracy while supporting transparency and trust. 4. Cognitive context awareness: Accounting for neural and psychological state enables more reliable and personalised BCI performance.Together, these themes suggest a transition from static, single-user BCIs toward adaptive, context-aware, and multi-agent systems designed for long-term everyday use.

The new transport mode killed far more people than had railways and horse-drawn traffic. As motor vehicles spread, the death toll rose. By 1929 in the most thoroughly motorised country, the United States, motor accidents ranked tenth among the principal causes of death.1 Policy measures in the richer countries eventually restrained the growth of fatalities in road accidents during the 1970s, but by then the contagion had already infected poorer nations. A World Health Organization survey found, for fifteen developing countries in 1972, that road accidents accounted for almost 17 per cent of the total number of deaths studied, a value exceeded only by mortality from enteritis.2 Here we examine the various policies adopted in industrial states to control the problem and their effectiveness. We pay particular attention to the years between the World Wars in this rather anglocentric discussion because, despite the period’s poor reputation, motorisation then proceeded very rapidly and the road-accident problem assumed national importance for the first time. We focus on road deaths because the statistics are more comparable over time and between countries than those on road injuries.KeywordsMotor VehicleSpeed LimitRational TheoryRoad SafetySeat BeltThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.