Personal Protective Equipment Detection in Extreme Construction Conditions

Yuexiong Ding; Xiaowei Luo

doi:10.1061/9780784485248.081

Personal Protective Equipment Detection in Extreme Construction Conditions

^*Corresponding author for this work

Department of Architecture and Civil Engineering

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

3 Citations (Scopus)

Abstract

Object detection has been widely applied for construction safety management, especially personal protective equipment (PPE) detection. Though the existing PPE detection models trained on conventional datasets have achieved excellent results, their performance dramatically declines in extreme construction conditions. A robust detection model NST-YOLOv5 is developed by combining the neural style transfer (NST) and YOLOv5 technologies. Five extreme conditions are considered and simulated via the NST module to endow the detection model with excellent robustness, including low light, intense light, sand dust, fog, and rain. Experiments show that the NST has great potential as a tool for extreme data synthesis since it is better at simulating extreme conditions than other traditional image processing algorithms and helps the NST-YOLOv5 achieve 0.141 and 0.083 mAP_05:95 improvements in synthesized and real-world extreme data. This study provides a new feasible way to obtain a more robust detection model for extreme construction conditions. © 2024 American Society of Civil Engineers

Original language	English
Title of host publication	Computing in Civil Engineering 2023
Subtitle of host publication	Resilience, Safety, and Sustainability
Editors	Yelda Turkan, Joseph Louis, Fernanda Leite, Semiha Ergan
Publisher	American Society of Civil Engineers
Pages	672–679
ISBN (Electronic)	9780784485248
DOIs	https://doi.org/10.1061/9780784485248.081
Publication status	Published - 2024
Event	2023 ASCE International Computing in Civil Engineering (i3CE) Conference - Oregon State University, Corvallis, United States Duration: 25 Jun 2023 → 28 Jun 2023 https://i3ce2023.org/

Conference

Conference	2023 ASCE International Computing in Civil Engineering (i3CE) Conference
Place	United States
City	Corvallis
Period	25/06/23 → 28/06/23
Internet address	https://i3ce2023.org/

Funding

The Shenzhen Science and Technology Innovation Committee Grant #JCYJ20180507181647320 and General Research Fund from Research Grant Council of Hong Kong SAR # 11211622 jointly supported this work. The conclusions herein are those of the authors and do not necessarily reflect the views of the sponsoring agencies.

RGC Funding Information

RGC-funded

Access Output

10.1061/9780784485248.081

https://ascelibrary.org/doi/10.1061/9780784485248.081

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

GRF: Automatic Detection of Safety Violations using Vision and Knowledge
LUO, X. (Principal Investigator / Project Coordinator) & SONG, L. (Co-Investigator)
1/09/22 → …
Project: Research

Cite this

@inproceedings{f57a128c68094c59b1c7b522c63689a5,

title = "Personal Protective Equipment Detection in Extreme Construction Conditions",

abstract = "Object detection has been widely applied for construction safety management, especially personal protective equipment (PPE) detection. Though the existing PPE detection models trained on conventional datasets have achieved excellent results, their performance dramatically declines in extreme construction conditions. A robust detection model NST-YOLOv5 is developed by combining the neural style transfer (NST) and YOLOv5 technologies. Five extreme conditions are considered and simulated via the NST module to endow the detection model with excellent robustness, including low light, intense light, sand dust, fog, and rain. Experiments show that the NST has great potential as a tool for extreme data synthesis since it is better at simulating extreme conditions than other traditional image processing algorithms and helps the NST-YOLOv5 achieve 0.141 and 0.083 mAP05:95 improvements in synthesized and real-world extreme data. This study provides a new feasible way to obtain a more robust detection model for extreme construction conditions. {\textcopyright} 2024 American Society of Civil Engineers",

author = "Yuexiong Ding and Xiaowei Luo",

year = "2024",

doi = "10.1061/9780784485248.081",

language = "English",

pages = "672–679",

editor = "Yelda Turkan and Joseph Louis and Fernanda Leite and Semiha Ergan",

booktitle = "Computing in Civil Engineering 2023",

publisher = "American Society of Civil Engineers",

address = "United States",

note = "2023 ASCE International Computing in Civil Engineering (i3CE) Conference ; Conference date: 25-06-2023 Through 28-06-2023",

url = "https://i3ce2023.org/",

}

Ding, Y & Luo, X 2024, Personal Protective Equipment Detection in Extreme Construction Conditions. in Y Turkan, J Louis, F Leite & S Ergan (eds), Computing in Civil Engineering 2023: Resilience, Safety, and Sustainability. American Society of Civil Engineers, pp. 672–679, 2023 ASCE International Computing in Civil Engineering (i3CE) Conference, Corvallis, Oregon, United States, 25/06/23. https://doi.org/10.1061/9780784485248.081

Personal Protective Equipment Detection in Extreme Construction Conditions. / Ding, Yuexiong ; Luo, Xiaowei.
Computing in Civil Engineering 2023: Resilience, Safety, and Sustainability. ed. / Yelda Turkan; Joseph Louis; Fernanda Leite; Semiha Ergan. American Society of Civil Engineers, 2024. p. 672–679.

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - Personal Protective Equipment Detection in Extreme Construction Conditions

AU - Ding, Yuexiong

AU - Luo, Xiaowei

PY - 2024

Y1 - 2024

N2 - Object detection has been widely applied for construction safety management, especially personal protective equipment (PPE) detection. Though the existing PPE detection models trained on conventional datasets have achieved excellent results, their performance dramatically declines in extreme construction conditions. A robust detection model NST-YOLOv5 is developed by combining the neural style transfer (NST) and YOLOv5 technologies. Five extreme conditions are considered and simulated via the NST module to endow the detection model with excellent robustness, including low light, intense light, sand dust, fog, and rain. Experiments show that the NST has great potential as a tool for extreme data synthesis since it is better at simulating extreme conditions than other traditional image processing algorithms and helps the NST-YOLOv5 achieve 0.141 and 0.083 mAP05:95 improvements in synthesized and real-world extreme data. This study provides a new feasible way to obtain a more robust detection model for extreme construction conditions. © 2024 American Society of Civil Engineers

AB - Object detection has been widely applied for construction safety management, especially personal protective equipment (PPE) detection. Though the existing PPE detection models trained on conventional datasets have achieved excellent results, their performance dramatically declines in extreme construction conditions. A robust detection model NST-YOLOv5 is developed by combining the neural style transfer (NST) and YOLOv5 technologies. Five extreme conditions are considered and simulated via the NST module to endow the detection model with excellent robustness, including low light, intense light, sand dust, fog, and rain. Experiments show that the NST has great potential as a tool for extreme data synthesis since it is better at simulating extreme conditions than other traditional image processing algorithms and helps the NST-YOLOv5 achieve 0.141 and 0.083 mAP05:95 improvements in synthesized and real-world extreme data. This study provides a new feasible way to obtain a more robust detection model for extreme construction conditions. © 2024 American Society of Civil Engineers

U2 - 10.1061/9780784485248.081

DO - 10.1061/9780784485248.081

M3 - RGC 32 - Refereed conference paper (with host publication)

SP - 672

EP - 679

BT - Computing in Civil Engineering 2023

A2 - Turkan, Yelda

A2 - Louis, Joseph

A2 - Leite, Fernanda

A2 - Ergan, Semiha

PB - American Society of Civil Engineers

T2 - 2023 ASCE International Computing in Civil Engineering (i3CE) Conference

Y2 - 25 June 2023 through 28 June 2023

ER -

Personal Protective Equipment Detection in Extreme Construction Conditions

Abstract

Conference

Funding

RGC Funding Information

Access Output

Other Access Options

Permanent Link

Fingerprint

Projects

GRF: Automatic Detection of Safety Violations using Vision and Knowledge

Cite this