Masked Facial Region Recognition Using Human Pose Estimation and Broad Learning System

Hongli Xiao; Bingshu Wang; Jiangbin Zheng; Jin Fang; Zhulin Liu; C. L. Philip Chen

doi:10.1117/12.2643023

Masked Facial Region Recognition Using Human Pose Estimation and Broad Learning System

Hongli Xiao
, Bingshu Wang
, Jiangbin Zheng
, Jin Fang
, Zhulin Liu
, C. L. Philip Chen

School of Software

Northwestern Polytechnical University Xian
University of Macau
South China University of Technology
Guangdong Artificial Intelligence and Digital Economy Laboratory - Guangzhou

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

COVID-19 and its variants have been posing a large risk to people around the world since the outbreak of the disease. Many techniques like AI are explored to help combat epidemics. People are required or forced to wear a mask to fight against COVID-19 epidemics worldwide. It brings new challenges to the task of masked facial region recognition. When facial regions are occluded by masks, it will result in some failures of face detection algorithms. In this paper, we propose a method to recognize masked faces. It mainly includes three parts. Firstly, the human pose is estimated to produce a series of key points. It is implemented by OpenPose. Secondly, a key-points location strategy is designed to capture the masked facial regions. It can locate the positions of faces accurately. Thirdly, the broad learning system, which is also an incremental learning algorithm, is employed to recognize the classes of candidate regions. Experiments conducted on some datasets shed light on the effectiveness of the proposed method.

Original language	English
Title of host publication	Fourteenth International Conference on Digital Image Processing, ICDIP 2022
Editors	Xudong Jiang, Wenbing Tao, Deze Zeng, Yi Xie
Publisher	SPIE
ISBN (Electronic)	9781510657564
DOIs	https://doi.org/10.1117/12.2643023
State	Published - 2022
Event	14th International Conference on Digital Image Processing, ICDIP 2022 - Wuhan, China Duration: 20 May 2022 → 23 May 2022

Publication series

Name	Proceedings of SPIE - The International Society for Optical Engineering
Volume	12342
ISSN (Print)	0277-786X
ISSN (Electronic)	1996-756X

Conference

Conference	14th International Conference on Digital Image Processing, ICDIP 2022
Country/Territory	China
City	Wuhan
Period	20/05/22 → 23/05/22

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

broad learning system
masked facial recognition
pose estimation
ROI extraction

Access to Document

10.1117/12.2643023

Cite this

Xiao, H., Wang, B., Zheng, J., Fang, J., Liu, Z., & Philip Chen, C. L. (2022). Masked Facial Region Recognition Using Human Pose Estimation and Broad Learning System. In X. Jiang, W. Tao, D. Zeng, & Y. Xie (Eds.), Fourteenth International Conference on Digital Image Processing, ICDIP 2022 Article 123420S (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 12342). SPIE. https://doi.org/10.1117/12.2643023

@inproceedings{272e3cb569424cf99d1525f4b66a7aad,

title = "Masked Facial Region Recognition Using Human Pose Estimation and Broad Learning System",

abstract = "COVID-19 and its variants have been posing a large risk to people around the world since the outbreak of the disease. Many techniques like AI are explored to help combat epidemics. People are required or forced to wear a mask to fight against COVID-19 epidemics worldwide. It brings new challenges to the task of masked facial region recognition. When facial regions are occluded by masks, it will result in some failures of face detection algorithms. In this paper, we propose a method to recognize masked faces. It mainly includes three parts. Firstly, the human pose is estimated to produce a series of key points. It is implemented by OpenPose. Secondly, a key-points location strategy is designed to capture the masked facial regions. It can locate the positions of faces accurately. Thirdly, the broad learning system, which is also an incremental learning algorithm, is employed to recognize the classes of candidate regions. Experiments conducted on some datasets shed light on the effectiveness of the proposed method.",

keywords = "broad learning system, masked facial recognition, pose estimation, ROI extraction",

author = "Hongli Xiao and Bingshu Wang and Jiangbin Zheng and Jin Fang and Zhulin Liu and \{Philip Chen\}, \{C. L.\}",

note = "Publisher Copyright: {\textcopyright} 2022 SPIE.; 14th International Conference on Digital Image Processing, ICDIP 2022 ; Conference date: 20-05-2022 Through 23-05-2022",

year = "2022",

doi = "10.1117/12.2643023",

language = "英语",

series = "Proceedings of SPIE - The International Society for Optical Engineering",

publisher = "SPIE",

editor = "Xudong Jiang and Wenbing Tao and Deze Zeng and Yi Xie",

booktitle = "Fourteenth International Conference on Digital Image Processing, ICDIP 2022",

}

Xiao, H, Wang, B, Zheng, J, Fang, J, Liu, Z & Philip Chen, CL 2022, Masked Facial Region Recognition Using Human Pose Estimation and Broad Learning System. in X Jiang, W Tao, D Zeng & Y Xie (eds), Fourteenth International Conference on Digital Image Processing, ICDIP 2022., 123420S, Proceedings of SPIE - The International Society for Optical Engineering, vol. 12342, SPIE, 14th International Conference on Digital Image Processing, ICDIP 2022, Wuhan, China, 20/05/22. https://doi.org/10.1117/12.2643023

Masked Facial Region Recognition Using Human Pose Estimation and Broad Learning System. / Xiao, Hongli; Wang, Bingshu; Zheng, Jiangbin et al.
Fourteenth International Conference on Digital Image Processing, ICDIP 2022. ed. / Xudong Jiang; Wenbing Tao; Deze Zeng; Yi Xie. SPIE, 2022. 123420S (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 12342).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Masked Facial Region Recognition Using Human Pose Estimation and Broad Learning System

AU - Xiao, Hongli

AU - Wang, Bingshu

AU - Zheng, Jiangbin

AU - Fang, Jin

AU - Liu, Zhulin

AU - Philip Chen, C. L.

PY - 2022

Y1 - 2022

N2 - COVID-19 and its variants have been posing a large risk to people around the world since the outbreak of the disease. Many techniques like AI are explored to help combat epidemics. People are required or forced to wear a mask to fight against COVID-19 epidemics worldwide. It brings new challenges to the task of masked facial region recognition. When facial regions are occluded by masks, it will result in some failures of face detection algorithms. In this paper, we propose a method to recognize masked faces. It mainly includes three parts. Firstly, the human pose is estimated to produce a series of key points. It is implemented by OpenPose. Secondly, a key-points location strategy is designed to capture the masked facial regions. It can locate the positions of faces accurately. Thirdly, the broad learning system, which is also an incremental learning algorithm, is employed to recognize the classes of candidate regions. Experiments conducted on some datasets shed light on the effectiveness of the proposed method.

AB - COVID-19 and its variants have been posing a large risk to people around the world since the outbreak of the disease. Many techniques like AI are explored to help combat epidemics. People are required or forced to wear a mask to fight against COVID-19 epidemics worldwide. It brings new challenges to the task of masked facial region recognition. When facial regions are occluded by masks, it will result in some failures of face detection algorithms. In this paper, we propose a method to recognize masked faces. It mainly includes three parts. Firstly, the human pose is estimated to produce a series of key points. It is implemented by OpenPose. Secondly, a key-points location strategy is designed to capture the masked facial regions. It can locate the positions of faces accurately. Thirdly, the broad learning system, which is also an incremental learning algorithm, is employed to recognize the classes of candidate regions. Experiments conducted on some datasets shed light on the effectiveness of the proposed method.

KW - broad learning system

KW - masked facial recognition

KW - pose estimation

KW - ROI extraction

UR - https://www.scopus.com/pages/publications/85141864422

U2 - 10.1117/12.2643023

DO - 10.1117/12.2643023

M3 - 会议稿件

AN - SCOPUS:85141864422

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - Fourteenth International Conference on Digital Image Processing, ICDIP 2022

A2 - Jiang, Xudong

A2 - Tao, Wenbing

A2 - Zeng, Deze

A2 - Xie, Yi

PB - SPIE

T2 - 14th International Conference on Digital Image Processing, ICDIP 2022

Y2 - 20 May 2022 through 23 May 2022

ER -

Xiao H, Wang B, Zheng J, Fang J, Liu Z, Philip Chen CL. Masked Facial Region Recognition Using Human Pose Estimation and Broad Learning System. In Jiang X, Tao W, Zeng D, Xie Y, editors, Fourteenth International Conference on Digital Image Processing, ICDIP 2022. SPIE. 2022. 123420S. (Proceedings of SPIE - The International Society for Optical Engineering). doi: 10.1117/12.2643023

Masked Facial Region Recognition Using Human Pose Estimation and Broad Learning System

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this