Optimally-Weighted Image-Pose Approach (OWIPA) for Distracted Driver Detection and Classification
Distracted driving is the prime factor of motor vehicle accidents. Current studies on distraction detection focus on improving distraction detection performance through various techniques, including convolutional neural networks (CNNs) and recurrent neural networks (RNNs). However, the research on d...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-07-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/21/14/4837 |
id |
doaj-8c39799a0cca412285c173bc61f6386b |
---|---|
record_format |
Article |
spelling |
doaj-8c39799a0cca412285c173bc61f6386b2021-07-23T14:05:55ZengMDPI AGSensors1424-82202021-07-01214837483710.3390/s21144837Optimally-Weighted Image-Pose Approach (OWIPA) for Distracted Driver Detection and ClassificationHong Vin Koay0Joon Huang Chuah1Chee-Onn Chow2Yang-Lang Chang3Bhuvendhraa Rudrusamy4Department of Electrical Engineering, Faculty of Engineering, University of Malaya, Kuala Lumpur 50603, MalaysiaDepartment of Electrical Engineering, Faculty of Engineering, University of Malaya, Kuala Lumpur 50603, MalaysiaDepartment of Electrical Engineering, Faculty of Engineering, University of Malaya, Kuala Lumpur 50603, MalaysiaDepartment of Electrical Engineering, National Taipei University of Technology, Taipei 10608, TaiwanSchool of Engineering and Physical Sciences, Heriot-Watt University Malaysia, Putrajaya 62200, MalaysiaDistracted driving is the prime factor of motor vehicle accidents. Current studies on distraction detection focus on improving distraction detection performance through various techniques, including convolutional neural networks (CNNs) and recurrent neural networks (RNNs). However, the research on detection of distracted drivers through pose estimation is scarce. This work introduces an ensemble of ResNets, which is named Optimally-weighted Image-Pose Approach (OWIPA), to classify the distraction through original and pose estimation images. The pose estimation images are generated from HRNet and ResNet. We use ResNet101 and ResNet50 to classify the original images and the pose estimation images, respectively. An optimum weight is determined through grid search method, and the predictions from both models are weighted through this parameter. The experimental results show that our proposed approach achieves 94.28% accuracy on AUC Distracted Driver Dataset.https://www.mdpi.com/1424-8220/21/14/4837optimally-weighted image-pose approach (OWIPA)convolutional neural network (CNN)deep learningpose estimationdistraction detectiondistraction classification |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Hong Vin Koay Joon Huang Chuah Chee-Onn Chow Yang-Lang Chang Bhuvendhraa Rudrusamy |
spellingShingle |
Hong Vin Koay Joon Huang Chuah Chee-Onn Chow Yang-Lang Chang Bhuvendhraa Rudrusamy Optimally-Weighted Image-Pose Approach (OWIPA) for Distracted Driver Detection and Classification Sensors optimally-weighted image-pose approach (OWIPA) convolutional neural network (CNN) deep learning pose estimation distraction detection distraction classification |
author_facet |
Hong Vin Koay Joon Huang Chuah Chee-Onn Chow Yang-Lang Chang Bhuvendhraa Rudrusamy |
author_sort |
Hong Vin Koay |
title |
Optimally-Weighted Image-Pose Approach (OWIPA) for Distracted Driver Detection and Classification |
title_short |
Optimally-Weighted Image-Pose Approach (OWIPA) for Distracted Driver Detection and Classification |
title_full |
Optimally-Weighted Image-Pose Approach (OWIPA) for Distracted Driver Detection and Classification |
title_fullStr |
Optimally-Weighted Image-Pose Approach (OWIPA) for Distracted Driver Detection and Classification |
title_full_unstemmed |
Optimally-Weighted Image-Pose Approach (OWIPA) for Distracted Driver Detection and Classification |
title_sort |
optimally-weighted image-pose approach (owipa) for distracted driver detection and classification |
publisher |
MDPI AG |
series |
Sensors |
issn |
1424-8220 |
publishDate |
2021-07-01 |
description |
Distracted driving is the prime factor of motor vehicle accidents. Current studies on distraction detection focus on improving distraction detection performance through various techniques, including convolutional neural networks (CNNs) and recurrent neural networks (RNNs). However, the research on detection of distracted drivers through pose estimation is scarce. This work introduces an ensemble of ResNets, which is named Optimally-weighted Image-Pose Approach (OWIPA), to classify the distraction through original and pose estimation images. The pose estimation images are generated from HRNet and ResNet. We use ResNet101 and ResNet50 to classify the original images and the pose estimation images, respectively. An optimum weight is determined through grid search method, and the predictions from both models are weighted through this parameter. The experimental results show that our proposed approach achieves 94.28% accuracy on AUC Distracted Driver Dataset. |
topic |
optimally-weighted image-pose approach (OWIPA) convolutional neural network (CNN) deep learning pose estimation distraction detection distraction classification |
url |
https://www.mdpi.com/1424-8220/21/14/4837 |
work_keys_str_mv |
AT hongvinkoay optimallyweightedimageposeapproachowipafordistracteddriverdetectionandclassification AT joonhuangchuah optimallyweightedimageposeapproachowipafordistracteddriverdetectionandclassification AT cheeonnchow optimallyweightedimageposeapproachowipafordistracteddriverdetectionandclassification AT yanglangchang optimallyweightedimageposeapproachowipafordistracteddriverdetectionandclassification AT bhuvendhraarudrusamy optimallyweightedimageposeapproachowipafordistracteddriverdetectionandclassification |
_version_ |
1721286001728421888 |