Driver drowsiness detection using behavioral measures and machine learning techniques: A review of state-of-art techniques

Ngxande, Mkhuseli; Tapamo, Jules-Raymond; Burke, Michael G.

doi:10.1109/robomech.2017.8261140

Cited by 105 publications

(35 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…All these troubling facts motivate the need for an economical solution that can detect drowsiness in early stages. It is commonly agreed [29,20,18] that there are three types of sources of information in drowsiness detection: Performance measurements, physiological measurements, and behavioral measurements.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Realistic Dataset and Baseline Temporal Model for Early Drowsiness Detection

Ghoddoosian

Galib

Athitsos

2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

View full text Add to dashboard Cite

Drowsiness can put lives of many drivers and workers in danger. It is important to design practical and easy-todeploy real-world systems to detect the onset of drowsiness. In this paper, we address early drowsiness detection, which can provide early alerts and offer subjects ample time to react. We present a large and public real-life dataset 1 of 60 subjects, with video segments labeled as alert, low vigilant, or drowsy. This dataset consists of around 30 hours of video, with contents ranging from subtle signs of drowsiness to more obvious ones. We also benchmark a temporal model 2 for our dataset, which has low computational and storage demands. The core of our proposed method is a Hierarchical Multiscale Long Short-Term Memory (HM-LSTM) network, that is fed by detected blink features in sequence. Our experiments demonstrate the relationship between the sequential blink features and drowsiness. In the experimental results, our baseline method produces higher accuracy than human judgment.

show abstract

Section: Introductionmentioning

confidence: 99%

“…By "acted" we mean data where subjects were instructed to simulate drowsiness, compared to "realistic" data, such as ours, where subjects were indeed drowsy in the corresponding videos. The lack of large, public, and realistic datasets has been pointed out by researchers in the field [18,19,20].…”

Section: Introductionmentioning

confidence: 99%

A Realistic Dataset and Baseline Temporal Model for Early Drowsiness Detection

Ghoddoosian

Galib

Athitsos

2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

View full text Add to dashboard Cite

show abstract

“…The proposed method to calculate M-EAR can determine blinking eyes and know whether the driver is sleepy or not. From the experimental data of five drivers (each of five videos), then based on the reference [9] that the condition is sleepy if the Advances in Social Science, Education and Humanities Research, volume 203 number of blinks below 10 per minute value, it can be determined fourth driver in sleepy condition.…”

Section: Discussionmentioning

confidence: 99%

“…The features obtained were analyzed using percentage of eye closure (PERCLOS) [4] [5], blink frequency [6] [7], eye aspect ratio (EAR) [6], face position, nodding frequency. The last classified with Support Vector Machines (SVM) [8], KNN, and Hidden Markov Model (HMM) [9] to detect sleepiness of the driver.…”

Section: Introductionmentioning

confidence: 99%

Drowsiness Detection According to the Number of Blinking Eyes Specified From Eye Aspect Ratio Value Modification

Pasaribu¹,

Prijono²,

Ratnadewi³

et al. 2019

Proceedings of the 1st International Conference on Life, Innovation, Change and Knowledge (ICLICK 2018)

View full text Add to dashboard Cite

The current accidents in road increase caused by driver's negligence. There are several reasons, that is fatigue, drowsiness, not concentration, and others. One of the drowsiness detectors is through the blinking of an eye. In normal people the blinking of the eye is 10 per minute, whereas in people who are sleepy eye blinks below 10 per minute. In this research we proposed the modified algorithm to have eye aspect ratio value. Experimental data are taken from video with 30 frames per second. Experiments were performed in the laboratory with driver simulation tools. Data are taken for 105 minutes and given the load test so that the driver will experience fatigue while driving. Keywords-modified of eye aspect ratio; drowsiness; blinking rate, , = / (2)

show abstract

“…The bigger the data, the better the classification rate, efficiency rate, prediction rate and general system throughput. Solving problems which would have been rather impossible to deal with [12,15], Machine learning has impacted greatly in health, industry, transportation, marketing and other sectors of human lives through the development of robots to handle activities which are toxic or dangerous to humans, the timely detection of diseases such as cancer, glaucoma etc., the visualization of smart cars, effective web search, language translations and etc. Over time, the ever-increasing amount of data from different sources could not be stored on personal computers due to huge storage capacity needed and required millions of servers for appropriate storage These servers could only be owned by particular groups of companies or individuals who could afford for both their purchase and maintenance.…”

Section: Introductionmentioning

confidence: 99%

Big Data, Machine Learning and the BlockChain Technology: An Overview

Acheampong¹

2018

IJCA

View full text Add to dashboard Cite

The importance of big data in machine learning cannot be overemphasized in recent times. Through the evolution of big data, most scientific technologies that relied heavily on enormous data in solving complex issues in human lives gained grounds; machine learning is an instance of these technologies. Various machine learning models that yield groundbreaking throughputs with high efficiency rates in predicting, detecting, classifying, discovering and acquiring in-depth knowledge about events that would otherwise be very difficult to ascertain have been made possible due to big data. Although big data has undoubtedly helped in the field of machine learning research ,over the years, its mode of acquisition has posed great challenge in industries,education and other agencies that obtained them for various purposes. This is because these large quantities of data cannot be stored on personal computers with limited storage capabicity but required the use of high storage capacity servers for effective storage. These servers may be owned by a group of companies or individuals who had the singular priviledge to modify the data in their possession as and when deemed relevant thus the creation of a centralized data storage environment. These were mostly refered to as the Third Parties (TP) in the data acquisition process. For the services they rendered, these trusted parties priced data in their possession expensively. The adverse effect is a limitation on various researches that could help solve a number of problems in human lives. It is worth mentioning that the security of these data being purchased expensively cannot be even assured limiting various researches that thrive on secured data. In order to curb these occurrences and have better machine learning models, the incorporation of Blockchain Technology databases into machine learning. This paper discusses the concept of big data, Machine Learning and Blockchains. It further discusses how Big data has impacted the Machine learning Community, the significance of Machine Learning and how the BlockChain Technology could be used similarly impact the Machine Learning Community. The aim of this paper is to encourge further research in incoporating the BlockChain Technology into Machine Learning.

show abstract

Driver drowsiness detection using behavioral measures and machine learning techniques: A review of state-of-art techniques

Cited by 105 publications

References 28 publications

A Realistic Dataset and Baseline Temporal Model for Early Drowsiness Detection

A Realistic Dataset and Baseline Temporal Model for Early Drowsiness Detection

Drowsiness Detection According to the Number of Blinking Eyes Specified From Eye Aspect Ratio Value Modification

Big Data, Machine Learning and the BlockChain Technology: An Overview

Contact Info

Product

Resources

About