Real time detection of driver fatigue based on CNN‐LSTM

Liu, Ming-Zhou; Xu, Xin; Hu, Jing; Jiang, Qiannan

doi:10.1049/ipr2.12373

Cited by 27 publications

(6 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It may have a detrimental effect among practical applications for nighttime detection. Meanwhile Ming-Zhou Liu et al used LSTM model to detect driver's fatigue state and obtained an accuracy of 99.78% with the fatigue detection algorithm of LSTM through a homemade dataset, i.e., 600 driver images [12].…”

Section: Long Term Memory Network (Lstm)mentioning

confidence: 99%

A review of techniques and methods for deep learning techniques in driver fatigue detection

Luo

2024

ACE

View full text Add to dashboard Cite

Road accidents in which fatigue driving is a significant cause of death are responsible for many deaths worldwide. Approximately 100,000 crashes are caused by driver fatigue each year. Also, fatigue driving is responsible for about 16% of road accidents in general and more than 20% of highway accidents, so fatigue driving accounts for a large percentage of vehicle accidents. Fatigue driving detection usually uses subjective and objective methods. Subjective methods rely on analysing the driver's psychological and facial expression information, while objective methods use external devices to extract feature parameters and apply artificial intelligence algorithms. However, these methods have limitations, such as subjectivity and individual differences. Deep learning, a promising tool inspired by neural networks, offers automatic feature learning, robust pattern recognition, and high adaptability. This review explores the application of deep learning in fatigue driving detection. It examines various deep learning feature extraction methods, classification models, prediction models, and related datasets. By leveraging deep learning techniques, fatigue driving detection can achieve higher accuracy and effectiveness, providing a reliable solution to this critical road safety problem. The review concludes with recommendations and future perspectives in this area.

show abstract

Section: Long Term Memory Network (Lstm)mentioning

confidence: 99%

A review of techniques and methods for deep learning techniques in driver fatigue detection

Luo

2024

ACE

View full text Add to dashboard Cite

show abstract

“…Moreover, in their study, LSTM eliminated the need for extensive data preprocessing and feature extraction which could have resulted in loss of useful information in EEG data. Also, Liu et al (2022) used LSTM for detecting fatigue of drivers. In construction, Qin and Bulbul (2023) used LSTM for predicting the mental workload of workers while using augmented reality head-mounted display for construction assembly.…”

Section: Machine Learning For Mental Workload Predictionmentioning

confidence: 99%

Predicting Mental Workload of Using Exoskeletons for Construction Work: A Deep Learning Approach

Akanmu,

Afolabi,

Okunola

2023

CONVR 2023 - Proceedings of the 23rd International Conference on Construction Applications of Virtual Reality

View full text Add to dashboard Cite

Exoskeletons are gaining attention as a potential solution for addressing low back injury in the construction industry. However, use of active back-support exoskeletons in construction can trigger unintended consequences which could increase mental workload of users while working with exoskeletons. Prolonged increase in mental workload could impact workers’ wellbeing and productivity. Prediction of mental workload during exoskeleton-use could inform strategies to mitigate the triggers. This study investigates a machine-learning framework for predicting mental workload of workers while using active back-support exoskeletons for construction work. Laboratory experiments were conducted wherein Electroencephalography (EEG) data were collected from participants wearing active back-support exoskeletons to perform flooring task. The EEG data underwent preprocessing, including band filtering, notch filtering, and independent component analysis, to remove artifacts and ensure data quality. A regression-based Long Short-Term Memory network was trained to forecast future time steps of the processed EEG data. The performance of the network was evaluated using root mean square error (RMSE) and r-squared (R2). A RMSE of 0.1527 and R2 of 0.9665 indicating good fit and strong correlation, respectively, were observed between the predicted and actual EEG data. Results of the comparison between the actual and predicted mental workload also show strong correction with an R2 of 0.8692. The findings motivate research directions into real-time monitoring of mental workload of workers during exoskeleton-use. The study has significant implications for stakeholders, enabling them to gain a deeper understanding of the impact of mental workload while using exoskeletons thereby providing opportunities for mitigation

show abstract

“…The above studies are only some typical application cases. There are many successful application cases of RNN in the construction of the driving behavior recognition model [116][117][118][119][120][121][122][123][124]. It can be seen that RNN can be used for all kinds of driving behavior recognition, is suitable for highdimensional and big data sample learning, and can extract deep temporal and spatial features with an accuracy rate of more than 90%.…”

Section: Recurrent Neuralmentioning

confidence: 99%

A Review for the Driving Behavior Recognition Methods Based on Vehicle Multisensor Information

Zhao

Zhong

Zhi-jun

et al. 2022

Journal of Advanced Transportation

View full text Add to dashboard Cite

The frequent traffic accidents lead to a large number of casualties and large related financial losses every year; this serious state is owed to several factors; among those, driving behavior is one of the most imperative subjects to discuss. Driving behaviors mainly include behavior characteristics such as car-following, lane change, and risky driving behavior such as distraction, fatigue, or aggressive driving, which are of great help to various tasks in traffic engineering. An accurate and reliable method of driving behavior recognition is of great significance and guidance for vehicle driving safety. In this paper, the vehicle multisensor information, vehicle CAN bus data acquisition system, and typical feature extraction methods are summarized at first. And then, several driving behavior recognition models based on machine learning and deep learning are reviewed. Through a detailed analysis of the features of random forests, support vector machines, convolutional neural networks, and recurrent neural networks used to build driving behavior recognition models, the following findings are obtained: the driving behavior model constructed by traditional machine learning model is relatively mature but it is greatly affected by feature extraction, data scale, and model structure, which affects the accuracy of the final driving behavior recognition. Deep learning model based on a neural network has achieved high accuracy in identifying driving behavior, and it may gradually become the mainstream of constructing the driving behavior model with the development of big data, artificial intelligence technology, and computer hardware. Finally, this paper points out some content that needs to be further explored, to provide reference and inspiration for scholars in this field to continue to study the driving behavior recognition model in depth.

show abstract

Real time detection of driver fatigue based on CNN‐LSTM

Cited by 27 publications

References 47 publications

A review of techniques and methods for deep learning techniques in driver fatigue detection

A review of techniques and methods for deep learning techniques in driver fatigue detection

Predicting Mental Workload of Using Exoskeletons for Construction Work: A Deep Learning Approach

A Review for the Driving Behavior Recognition Methods Based on Vehicle Multisensor Information

Contact Info

Product

Resources

About