Machine Learning Models for the Classification of Sleep Deprivation Induced Performance Impairment During a Psychomotor Vigilance Task Using Indices of Eye and Face Tracking (original) (raw)
Related papers
Accident Analysis & Prevention, 2013
There is currently no "gold standard" marker of cognitive performance impairment resulting from sleep loss. We utilized pattern recognition algorithms to determine which features of data collected under controlled laboratory conditions could most reliably identify cognitive performance impairment in response to sleep loss using data from only one testing session, such as would occur in the "real world" or field conditions. A training set for testing the pattern recognition algorithms was developed using objective Psychomotor Vigilance Task (PVT) and subjective Karolinska Sleepiness Scale (KSS) data collected from laboratory studies during which subjects were sleep deprived for 26-52 hours. The algorithm was then tested in data from both laboratory and field experiments. The pattern recognition algorithm was able to identify performance impairment with a single testing session in individuals studied under laboratory conditions using PVT, KSS, length of time awake and time of day information with sensitivity and specificity as high as 82%. When this algorithm was tested on data collected under real-world conditions from individuals whose data were not in the training set, accuracy of predictions for individuals categorized with low performance impairment were as high as 98%. Predictions for medium and severe performance impairment were less accurate. We conclude that pattern recognition algorithms may be a promising method for identifying performance impairment in individuals using only current information about the individual's behavior. Single testing features (e.g., number of PVT lapses) with high correlation with performance impairment in the laboratory setting may not be the best indicators of performance impairment under real-world conditions. Pattern recognition algorithms should be further tested for their ability to be used in conjunction
MACHINE LEARNING SYSTEMS FOR DETECTING DRIVER DROWSINESS
The advance of computing technology has provided the means for building intelligent vehicle systems. Drowsy driver detection system is one of the potential applications of intelligent vehicle systems. Previous approaches to drowsiness detection primarily make pre-assumptions about the relevant behavior, focusing on blink rate, eye closure, and yawning. Here we employ machine learning to datamine actual human behavior during drowsiness episodes. Automatic classifiers for 30 facial actions from the Facial Action Coding system were developed using machine learning on a separate database of spontaneous expressions. These facial actions include blinking and yawn motions, as well as a number of other facial movements. In addition, head motion was collected through automatic eye tracking and an accelerometer. These measures were passed to learningbased classifiers such as Adaboost and multinomial ridge regression. The system was able to predict sleep and crash episodes during a driving computer game with 96% accuracy within subjects and above 90% accuracy across subjects. This is the highest prediction rate reported to date for detecting real drowsiness. Moreover, the analysis revealed new information about human behavior during drowsy driving.
Using machine learning techniques to characterize sleep-deprived driving behavior
Traffic Injury Prevention
Objective: Sleep deprivation is known to affect driving behavior and may lead to serious car accidents similar to the effects from e.g., alcohol. In a previous study, we have demonstrated that the use of machine learning techniques allows adequate characterization of abnormal driving behavior after alprazolam and/or alcohol intake. In the present study, we extend this approach to sleep deprivation and test the model for characterization of new interventions. We aimed to classify abnormal driving behavior after sleep deprivation, and, by using a machine learning model, we tested if this model could also pick up abnormal driving behavior resulting from other interventions. Methods: Data were collected during a previous study, in which 24 subjects were tested after being sleep-deprived and after a well-rested night. Features were calculated from several driving parameters, such as the lateral position, speed of the car, and steering speed. In the present study, we used a gradient boosting model to classify sleep deprivation. The model was validated using a 5-fold cross validation technique. Next, probability scores were used to identify the overlap of driving behavior after sleep deprivation and driving behavior affected by other interventions. In the current study alprazolam, alcohol, and placebo are used to test/validate the approach. Results: The sleep deprivation model detected abnormal driving behavior in the simulator with an accuracy of 77 ± 9%. Abnormal driving behavior after alprazolam, and to a lesser extent also after alcohol intake, showed remarkably similar characteristics to sleep deprivation. The average probability score for alprazolam and alcohol measurements was 0.79, for alcohol 0.63, and for placebo only 0.27 and 0.30, matching the expected relative drowsiness. Conclusion: We developed a model detecting abnormal driving induced by sleep deprivation. The model shows the similarities in driving characteristics between sleep deprivation and other interventions, i.e., alcohol and alprazolam. Consequently, our model for sleep deprivation may serve as a next reference point for a driving test battery of newly developed drugs.
2013
Driver sleepiness contributes to a considerable proportion of road accidents, and a fit-forduty test able to measure a driver's sleepiness level might improve traffic safety. The aim of this study was to develop a fit-for-duty test based on eye movement measurements and on the sleep/wake predictor model (SWP, which predicts the sleepiness level) and evaluate the ability to predict severe sleepiness during real road driving. Twenty-four drivers participated in an experimental study which took place partly in the laboratory, where the fitfor-duty data were acquired, and partly on the road, where the drivers sleepiness was assessed. A series of four measurements were conducted over a 24-h period during different stages of sleepiness. Two separate analyses were performed; a variance analysis and a feature selection followed by classification analysis. In the first analysis it was found that the SWP and several eye movement features involving anti-saccades, pro-saccades, smooth pursuit, pupillometry and fixation stability varied significantly with different stages of sleep deprivation. In the second analysis, a feature set was determined based on floating forward selection. The correlation coefficient between a linear combination of the acquired features and subjective sleepiness (Karolinska sleepiness scale, KSS) was found to be R = 0.73 and the correct classification rate of drivers who reached high levels of sleepiness (KSS P 8) in the subsequent driving session was 82.4% (sensitivity = 80.0%, specificity = 84.2% and AUC = 0.86). Future improvements of a fit-for-duty test should focus on how to account for individual differences and situational/contextual factors in the test, and whether it is possible to maintain high sensitive/specificity with a shorter test that can be used in a real-life environment, e.g. on professional drivers. j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / t r c would probably be valuable in reducing driver sleepiness from an enforcement perspective but also from an educational point of view.
Detection of Driver Fatigue Caused by Sleep Deprivation
IEEE Transactions on Systems, Man, and Cybernetics - Part A: Systems and Humans, 2000
This paper aims to provide reliable indications of driver drowsiness based on the characteristics of driver-vehicle interaction. A test bed was built under a simulated driving environment, and a total of 12 subjects participated in two experiment sessions requiring different levels of sleep (partial sleepdeprivation versus no sleep-deprivation) before the experiment. The performance of the subjects was analyzed in a series of stimulus-response and routine driving tasks, which revealed the performance differences of drivers under different sleepdeprivation levels. The experiments further demonstrated that sleep deprivation had greater effect on rule-based than on skillbased cognitive functions: when drivers were sleep-deprived, their performance of responding to unexpected disturbances degraded, while they were robust enough to continue the routine driving tasks such as lane tracking, vehicle following, and lane changing. In addition, we presented both qualitative and quantitative guidelines for designing drowsy-driver detection systems in a probabilistic framework based on the paradigm of Bayesian networks. Temporal aspects of drowsiness and individual differences of subjects were addressed in the framework.
Pattern recognition methods: a novel analysis for the pupillographic sleepiness test
2010
The aim of this paper is to improve the information gained by the most commonly applied fit-for-duty sleepiness test (Pupillographic Sleepiness test, PST) by using pattern recognition approaches. The pupil diameter based sleepiness detection is enriched by several new features and machine learning methods. Using all newly computed pupil diameter features we achieved on the two-class detection problem (moderate sleepiness vs. high sleepiness) an accuracy of 83.03% on participant-dependent data with a Random Forest classifier. This result suggested that the PST-standard feature set should be enriched by the here proposed enlarged feature set.
Algorithms Comparison in Drowsiness Detection
Journal of Advanced College of Engineering and Management
Drowsy driving is a recognized leading cause of road accidents, resulting in a considerable number of fatalities and injuries. This paper presents a proposed system that leverages machine learning algorithms, specifically Convolutional Neural Network (CNN) and Support Vector Machine (SVM), to accurately detect the drowsy state of drivers by analyzing the diameter of their eyes and comparing the level of dilation or constriction. Drowsy driving poses a significant problem on roadways, as indicated by the National Highway Traffic Safety Administration's data, which reports approximately 100,000 police-reported collisions each year involving drowsy driving, leading to over 1,550 fatalities and 71,000 injuries. The proposed system demonstrates the potential to reduce accidents associated with drowsy driving. We conducted an evaluation and comparison of the effectiveness of CNN and SVM algorithms with the objective of identifying the optimal algorithm for drowsiness detection. Our al...
Industrial Health, 2007
Sleepiness is a major risk factor for serious injury and death in accidents. Although it is important to develop countermeasures to sleepiness to reduce risks, it is equally important to determine the most effective timing for these countermeasures. To determine optimum timing for necessary countermeasures, we must be able to predict performance errors. This study examined the predictability of subjective and physiological indicators of sleepiness during a vigilance task. Thirteen healthy male volunteers (mean age, 26.9 yr; SD = 5.98 yr; range 22-43 yr) participated in the study. Participants used the Karolinska sleepiness scale (KSS) to rate their subjective sleepiness every 4 min during a 40-min Mackworth clock test. Electrophysiological and performance data were divided into 10 epochs (i.e., 1 epoch lasted for 4 min). To estimate predictability, the data from the sleepiness indicators used for the correlation analysis were preceded by one epoch to the performance data. Results showed that sleepiness indicators (KSS score and electroencephalographic [EEG] alpha activity) and standard deviation of heart rate (SDNN) were significantly correlated with succeeding performance on the vigilance test. These findings suggest that the KSS score, EEG alpha activity, and SDNN could be used to predict performance errors.
Drowsiness and Lethargy Detection Using Machine Learning Techniques
2023 International Conference on Electrical, Computer and Communication Engineering (ECCE), 2023
Drowsiness has severe effects on the safety of human life. The worldwide death rate due to drowsy driving is quite alarming. As the implementation of artificial intelligence (AI) is growing faster, this paper describes an attempt to implement machine learning (ML) to detect drowsiness. 120 videos of 60 participants are collected from the Real-Life Drowsiness Video Dataset made by a research team of the Vision-Learning-Mining Lab from the University of Texas at Arlington. Then Eye Aspect Ratio, Mouth Aspect Ratio, Pupil Circularity, and Mouth Aspect Ratio Over Eye Aspect Ratio, Nose Length, Chin Length, Nose Length Over Chin Length Ratio are extracted as features of each participant using the 3D Face-Mesh 468 facial landmarks system from those videos. After that, each feature is normalized by its mean and standard deviation. Then the CSV dataset is generated using seven initial and seven normalized features. A total of 30000 instances are there in the dataset. A total of eight classification algorithms are implemented to build the model. The dataset is split such that the individual in the train set will not be in the test set to test the proposed model's ability to predict drowsiness for new faces. 5-fold cross-validation is implemented to measure performance for each algorithm. Convolutional Neural Network (CNN) yields maximum accuracy (91.63%). The state of any individual's eye closing, rapid eye blinking, yawning, putting a hand on the mouth during yawning, and head posing too much up or down can be detected as drowsiness by the proposed model.