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

^*此作品的通讯作者

光电学院

Beijing Institute of Technology

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

摘要

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.

源语言	英语
主期刊名	Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings
编辑	Zhenan Sun, Ran He, Shiguang Shan, Jianjiang Feng, Zhenhua Guo
出版商	Springer
页	284-292
页数	9
ISBN（印刷版）	9783030314552
DOI	https://doi.org/10.1007/978-3-030-31456-9_32
出版状态	已出版 - 2019
活动	14th Chinese Conference on Biometric Recognition, CCBR 2019 - Zhuzhou, 中国期限: 12 10月 2019 → 13 10月 2019

出版系列

姓名	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
卷	11818 LNCS
ISSN（印刷版）	0302-9743
ISSN（电子版）	1611-3349

会议

会议	14th Chinese Conference on Biometric Recognition, CCBR 2019
国家/地区	中国
市	Zhuzhou
时期	12/10/19 → 13/10/19

访问文件

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

其它文件与链接

链接到 Scopus 的出版物

引用此

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

Zhai, Xiangyang ; He, Yuqing ; Zhao, Qian 等. / 106-Point Facial Landmark Localization with Mobile Networks Based on Regression. Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. 编辑 / Zhenan Sun ; Ran He ; Shiguang Shan ; Jianjiang Feng ; Zhenhua Guo. Springer, 2019. 页码 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. 在 Z Sun, R He, S Shan, J Feng & Z Guo (编辑), 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), 卷 11818 LNCS, Springer, 页码 284-292, 14th Chinese Conference on Biometric Recognition, CCBR 2019, Zhuzhou, 中国, 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 等.
Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. 编辑 / Zhenan Sun; Ran He; Shiguang Shan; Jianjiang Feng; Zhenhua Guo. Springer, 2019. 页码 284-292 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 卷 11818 LNCS).

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

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. 在 Sun Z, He R, Shan S, Feng J, Guo Z, 编辑, Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. Springer. 2019. 页码 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

摘要

出版系列

会议

访问文件

其它文件与链接

指纹

引用此