A fast and robust 3D face recognition approach based on deeply learned face representation

doi:10.1016/j.neucom.2019.07.047

	A fast and robust 3D face recognition approach based on deeply learned face representation
	Cai, Ying 2,3; Lei, Yinjie 4; Yang, Menglong 1,3; You, Zhisheng 1,3; Shan, Shiguang 5
	2019-10-21
发表期刊	NEUROCOMPUTING
ISSN	0925-2312
卷号	363 页码:375-397
摘要	With the superiority of three-dimensional (3D) scanning data, e.g., illumination invariance and pose robustness, 3D face recognition theoretically has the potential to achieve better results than two-dimensional (2D) face recognition. However, traditional 3D face recognition techniques suffer from high computational costs. This paper proposes a fast and robust 3D face recognition approach with three component technologies: a fast 3D scan preprocessing, multiple data augmentation, and a deep learning technique based on facial component patches. First, unlike the majority of the existing approaches, which require accurate facial registration, the proposed approach uses only three facial landmarks. Second, the specifical deep network with an improved supervision is designed to extract complementary features from four overlapping facial component patches. Finally, a data augmentation technique and three self-collected 3D face datasets are used to enlarge the scale of the training data. The proposed approach outperforms the state-of-the-art algorithms on four public 3D face benchmarks, i.e., 100%, 99.75%, 99.88%, and 99.07% rank-1 IRs with the standard test protocol on the FRGC v2.0, Bosphorus, BU-3DFE, and 3D-TEC datasets, respectively. Further, it requires only 0.84 seconds to identify a probe from a gallery with 466 faces. (C) 2019 Elsevier B.V. All rights reserved.
关键词	3D face recognition Deep learning Face preprocessing Multiple data augmentation
DOI	10.1016/j.neucom.2019.07.047
收录类别	SCI
语种	英语
资助项目	Chinese Universities Scientific Fund[2019NYB05] ; National Natural Science Foundation of China[61702350] ; National Natural Science Foundation of China[61402307] ; National Natural Science Foundation of China[61403265] ; National Natural Science Foundation of China[71774134] ; National Natural Science Foundation of China[71373216] ; Sichuan Science and Technology Program[18YYJC1287] ; Sichuan Science and Technology Program[2015SZ0226] ; Sichuan University[2018SCUH0042] ; National Key Research and Development Program of China[2016YFC0801100] ; National Key Scientific Instrument and Equipment Development Project of China[2013YQ49087903]
WOS研究方向	Computer Science
WOS类目	Computer Science, Artificial Intelligence
WOS记录号	WOS:000484005300034
出版者	ELSEVIER
引用统计	被引频次：31[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/4712
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Yang, Menglong
作者单位	1.Sichuan Univ, Sch Aeronaut & Astronaut, Chengdu, Sichuan, Peoples R China 2.Southwest Minzu Univ, Sch Elect & Informat Engn, Chengdu, Sichuan, Peoples R China 3.Wisesoft Software Co Ltd, Chengdu, Sichuan, Peoples R China 4.Sichuan Univ, Sch Elect & Informat Engn, Chengdu, Sichuan, Peoples R China 5.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Cai, Ying,Lei, Yinjie,Yang, Menglong,et al. A fast and robust 3D face recognition approach based on deeply learned face representation[J]. NEUROCOMPUTING,2019,363:375-397.
APA	Cai, Ying,Lei, Yinjie,Yang, Menglong,You, Zhisheng,&Shan, Shiguang.(2019).A fast and robust 3D face recognition approach based on deeply learned face representation.NEUROCOMPUTING,363,375-397.
MLA	Cai, Ying,et al."A fast and robust 3D face recognition approach based on deeply learned face representation".NEUROCOMPUTING 363(2019):375-397.