106-Point Facial Landmark Localization with Mobile Networks Based on Regression

Xiangyang Zhai; Yuqing He; Qian Zhao; Yutong Ding

doi:10.1007/978-3-030-31456-9_32

106-Point Facial Landmark Localization with Mobile Networks Based on Regression

Xiangyang Zhai, Yuqing He^*, Qian Zhao, Yutong Ding

^*Corresponding author for this work

School of Optics and Photonics

Beijing Institute of Technology

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

Abstract

Sparse facial landmark localization has lower precision for face reconstruction, while more point landmarks are competent to depict the structure of facial components. In this paper, the pipeline of detecting 106-point facial landmarks with regression is proposed. Based on the convergence and practical application of multi-points regression, we design MobileNetV2-FL and VGG16-FL. Besides, an effective data preprocessing strategy and some training tricks, such as the Online Hard Example Mining algorithm and Wing loss are applied to the issue. Experimental results show that the proposed method has lower failure rate, and is an effective and robust facial landmark localization method.

Original language	English
Title of host publication	Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings
Editors	Zhenan Sun, Ran He, Shiguang Shan, Jianjiang Feng, Zhenhua Guo
Publisher	Springer
Pages	284-292
Number of pages	9
ISBN (Print)	9783030314552
DOIs	https://doi.org/10.1007/978-3-030-31456-9_32
Publication status	Published - 2019
Event	14th Chinese Conference on Biometric Recognition, CCBR 2019 - Zhuzhou, China Duration: 12 Oct 2019 → 13 Oct 2019

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11818 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	14th Chinese Conference on Biometric Recognition, CCBR 2019
Country/Territory	China
City	Zhuzhou
Period	12/10/19 → 13/10/19

Keywords

Facial landmark localization
Mobile network
Regression

Access to Document

10.1007/978-3-030-31456-9_32

Cite this

Zhai, X., He, Y., Zhao, Q., & Ding, Y. (2019). 106-Point Facial Landmark Localization with Mobile Networks Based on Regression. In Z. Sun, R. He, S. Shan, J. Feng, & Z. Guo (Eds.), Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings (pp. 284-292). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11818 LNCS). Springer. https://doi.org/10.1007/978-3-030-31456-9_32

Zhai, Xiangyang ; He, Yuqing ; Zhao, Qian et al. / 106-Point Facial Landmark Localization with Mobile Networks Based on Regression. Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. editor / Zhenan Sun ; Ran He ; Shiguang Shan ; Jianjiang Feng ; Zhenhua Guo. Springer, 2019. pp. 284-292 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{e15e7307074f4d489abd6ddcd4be6279,

title = "106-Point Facial Landmark Localization with Mobile Networks Based on Regression",

abstract = "Sparse facial landmark localization has lower precision for face reconstruction, while more point landmarks are competent to depict the structure of facial components. In this paper, the pipeline of detecting 106-point facial landmarks with regression is proposed. Based on the convergence and practical application of multi-points regression, we design MobileNetV2-FL and VGG16-FL. Besides, an effective data preprocessing strategy and some training tricks, such as the Online Hard Example Mining algorithm and Wing loss are applied to the issue. Experimental results show that the proposed method has lower failure rate, and is an effective and robust facial landmark localization method.",

keywords = "Facial landmark localization, Mobile network, Regression",

author = "Xiangyang Zhai and Yuqing He and Qian Zhao and Yutong Ding",

note = "Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 14th Chinese Conference on Biometric Recognition, CCBR 2019 ; Conference date: 12-10-2019 Through 13-10-2019",

year = "2019",

doi = "10.1007/978-3-030-31456-9_32",

language = "English",

isbn = "9783030314552",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "284--292",

editor = "Zhenan Sun and Ran He and Shiguang Shan and Jianjiang Feng and Zhenhua Guo",

booktitle = "Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings",

address = "Germany",

}

Zhai, X, He, Y, Zhao, Q & Ding, Y 2019, 106-Point Facial Landmark Localization with Mobile Networks Based on Regression. in Z Sun, R He, S Shan, J Feng & Z Guo (eds), Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11818 LNCS, Springer, pp. 284-292, 14th Chinese Conference on Biometric Recognition, CCBR 2019, Zhuzhou, China, 12/10/19. https://doi.org/10.1007/978-3-030-31456-9_32

106-Point Facial Landmark Localization with Mobile Networks Based on Regression. / Zhai, Xiangyang; He, Yuqing; Zhao, Qian et al.
Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. ed. / Zhenan Sun; Ran He; Shiguang Shan; Jianjiang Feng; Zhenhua Guo. Springer, 2019. p. 284-292 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11818 LNCS).

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

TY - GEN

T1 - 106-Point Facial Landmark Localization with Mobile Networks Based on Regression

AU - Zhai, Xiangyang

AU - He, Yuqing

AU - Zhao, Qian

AU - Ding, Yutong

PY - 2019

Y1 - 2019

N2 - Sparse facial landmark localization has lower precision for face reconstruction, while more point landmarks are competent to depict the structure of facial components. In this paper, the pipeline of detecting 106-point facial landmarks with regression is proposed. Based on the convergence and practical application of multi-points regression, we design MobileNetV2-FL and VGG16-FL. Besides, an effective data preprocessing strategy and some training tricks, such as the Online Hard Example Mining algorithm and Wing loss are applied to the issue. Experimental results show that the proposed method has lower failure rate, and is an effective and robust facial landmark localization method.

AB - Sparse facial landmark localization has lower precision for face reconstruction, while more point landmarks are competent to depict the structure of facial components. In this paper, the pipeline of detecting 106-point facial landmarks with regression is proposed. Based on the convergence and practical application of multi-points regression, we design MobileNetV2-FL and VGG16-FL. Besides, an effective data preprocessing strategy and some training tricks, such as the Online Hard Example Mining algorithm and Wing loss are applied to the issue. Experimental results show that the proposed method has lower failure rate, and is an effective and robust facial landmark localization method.

KW - Facial landmark localization

KW - Mobile network

KW - Regression

UR - http://www.scopus.com/inward/record.url?scp=85075559903&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-31456-9_32

DO - 10.1007/978-3-030-31456-9_32

M3 - Conference contribution

AN - SCOPUS:85075559903

SN - 9783030314552

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 284

EP - 292

BT - Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings

A2 - Sun, Zhenan

A2 - He, Ran

A2 - Shan, Shiguang

A2 - Feng, Jianjiang

A2 - Guo, Zhenhua

PB - Springer

T2 - 14th Chinese Conference on Biometric Recognition, CCBR 2019

Y2 - 12 October 2019 through 13 October 2019

ER -

Zhai X, He Y, Zhao Q, Ding Y. 106-Point Facial Landmark Localization with Mobile Networks Based on Regression. In Sun Z, He R, Shan S, Feng J, Guo Z, editors, Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. Springer. 2019. p. 284-292. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-31456-9_32

106-Point Facial Landmark Localization with Mobile Networks Based on Regression

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this