Detecting Distraction of Drivers Based on Residual Neural Network

With the popularity of mobile internet and social software, drivers are getting more and more distracted, leading to a great threat to road traffic safety. This paper put forward the ResNet to identify the distracted behaviors by analyzing the images obtained by a camera module installed above the d...

Full description

Saved in:
Bibliographic Details
Published in2019 IEEE Eurasia Conference on IOT, Communication and Engineering (ECICE) pp. 457 - 460
Main Authors ZOU, JianCheng, YAN, PeiZhou, LI, ZhengZheng
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.10.2019
Subjects
Cameras
deep learning
distraction detection
driving safety
Learning systems
Measurement
ResNet
Safety
Software
Training
Vehicles
Online AccessGet full text

Cover

More Information
Summary:With the popularity of mobile internet and social software, drivers are getting more and more distracted, leading to a great threat to road traffic safety. This paper put forward the ResNet to identify the distracted behaviors by analyzing the images obtained by a camera module installed above the dashboard in front of the driver. The network with wide residual blocks and improved metric learning method was validated in Tensorflow and showed an ideal accuracy in distraction recognition, providing a new attempt to improve the driving safety.
DOI:10.1109/ECICE47484.2019.8942692