CASIA OpenIR
Large-Scale Bisample Learning on ID Versus Spot Face Recognition
Zhu, Xiangyu1,2,3; Liu, Hao1,2,3; Lei, Zhen1,2,3; Shi, Hailin1,2; Yang, Fan4; Yi, Dong5; Qi, Guojun6; Li, Stan Z.1,2,3
Source PublicationINTERNATIONAL JOURNAL OF COMPUTER VISION
ISSN0920-5691
2019-06-01
Volume127Issue:6-7Pages:684-700
Corresponding AuthorLei, Zhen(zlei@nlpr.ia.ac.cn)
AbstractIn real-world face recognition applications, there is a tremendous amount of data with two images for each person. One is an ID photo for face enrollment, and the other is a probe photo captured on spot. Most existing methods are designed for training data with limited breadth (a relatively small number of classes) and sufficient depth (many samples for each class). They would meet great challenges on ID versus Spot (IvS) data, including the under-represented intra-class variations and an excessive demand on computing devices. In this paper, we propose a deep learning based large-scale bisample learning (LBL) method for IvS face recognition. To tackle the bisample problem with only two samples for each class, a classification-verification-classification training strategy is proposed to progressively enhance the IvS performance. Besides, a dominant prototype softmax is incorporated to make the deep learning scalable on large-scale classes. We conduct LBL on a IvS face dataset with more than two million identities. Experimental results show the proposed method achieves superior performance to previous ones, validating the effectiveness of LBL on IvS face recognition.
KeywordFace recognition ID versus spot Large-scale bisample learning Dominant prototype softmax
DOI10.1007/s11263-019-01162-8
Indexed BySCI
Language英语
Funding ProjectChinese National Natural Science Foundation[61876178] ; Chinese National Natural Science Foundation[61806196] ; National Key Research and Development Plan[2016YFC080-1002] ; AuthenMetric RD Funds
Funding OrganizationChinese National Natural Science Foundation ; National Key Research and Development Plan ; AuthenMetric RD Funds
WOS Research AreaComputer Science
WOS SubjectComputer Science, Artificial Intelligence
WOS IDWOS:000468525900009
PublisherSPRINGER
Citation statistics
Document Type期刊论文
Identifierhttp://ir.ia.ac.cn/handle/173211/24237
Collection中国科学院自动化研究所
Corresponding AuthorLei, Zhen
Affiliation1.Chinese Acad Sci, Ctr Biometr & Secur Res, Beijing, Peoples R China
2.Chinese Acad Sci, Natl Lab Pattern Recognit, Inst Automat, Beijing, Peoples R China
3.Univ Chinese Acad Sci, Beijing, Peoples R China
4.Beihang Univ, Coll Software, Beijing, Peoples R China
5.DAMO Acad, Alibaba Grp, Hangzhou, Zhejiang, Peoples R China
6.HUAWEI Cloud, Boston, MA USA
First Author AffilicationChinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China
Corresponding Author AffilicationChinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China
Recommended Citation
GB/T 7714
Zhu, Xiangyu,Liu, Hao,Lei, Zhen,et al. Large-Scale Bisample Learning on ID Versus Spot Face Recognition[J]. INTERNATIONAL JOURNAL OF COMPUTER VISION,2019,127(6-7):684-700.
APA Zhu, Xiangyu.,Liu, Hao.,Lei, Zhen.,Shi, Hailin.,Yang, Fan.,...&Li, Stan Z..(2019).Large-Scale Bisample Learning on ID Versus Spot Face Recognition.INTERNATIONAL JOURNAL OF COMPUTER VISION,127(6-7),684-700.
MLA Zhu, Xiangyu,et al."Large-Scale Bisample Learning on ID Versus Spot Face Recognition".INTERNATIONAL JOURNAL OF COMPUTER VISION 127.6-7(2019):684-700.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Zhu, Xiangyu]'s Articles
[Liu, Hao]'s Articles
[Lei, Zhen]'s Articles
Baidu academic
Similar articles in Baidu academic
[Zhu, Xiangyu]'s Articles
[Liu, Hao]'s Articles
[Lei, Zhen]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Zhu, Xiangyu]'s Articles
[Liu, Hao]'s Articles
[Lei, Zhen]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.