Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing

Jingtian Xia; Yan Tang; Xi Jia; Linlin Shen; Zhihui Lai

doi:10.1007/978-3-030-31456-9_27

Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing

Jingtian Xia, Yan Tang, Xi Jia, Linlin Shen, Zhihui Lai

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

2 Citations (Scopus)

Abstract

With the wide deployment of the face recognition system, many face attacks, such as print attack, video attack and 3D face mask, have emerged. Face anti-spoofing is very important to protect face recognition system from attack. This paper proposes a structure of generative adversarial networks with skip connection for face anti-spoofing. First, we obtain the latent spatial features of faces by training generative adversarial networks to reconstruct both real and spoof faces; second, we use the convolution neural networks to detect the spoofing faces. In this paper, the proposed method is evaluated by three public databases. The results suggest that our approach achieves as high as 98% accuracy on both CASIA-FASD and REPLAY-ATTACK databases.

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	240-249
Number of pages	10
ISBN (Print)	9783030314552
DOIs	https://doi.org/10.1007/978-3-030-31456-9_27
Publication status	Published - 2019
Externally published	Yes
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

Face anti-spoofing
Generative adversarial networks
Latent spatial features

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-030-31456-9_27

Cite this

Xia, J., Tang, Y., Jia, X., Shen, L., & Lai, Z. (2019). Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing. In Z. Sun, R. He, S. Shan, J. Feng, & Z. Guo (Eds.), Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings (pp. 240-249). (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_27

Xia, Jingtian ; Tang, Yan ; Jia, Xi et al. / Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing. Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. editor / Zhenan Sun ; Ran He ; Shiguang Shan ; Jianjiang Feng ; Zhenhua Guo. Springer, 2019. pp. 240-249 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{5b8370318c4846828db91d7a9bd356dd,

title = "Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing",

abstract = "With the wide deployment of the face recognition system, many face attacks, such as print attack, video attack and 3D face mask, have emerged. Face anti-spoofing is very important to protect face recognition system from attack. This paper proposes a structure of generative adversarial networks with skip connection for face anti-spoofing. First, we obtain the latent spatial features of faces by training generative adversarial networks to reconstruct both real and spoof faces; second, we use the convolution neural networks to detect the spoofing faces. In this paper, the proposed method is evaluated by three public databases. The results suggest that our approach achieves as high as 98% accuracy on both CASIA-FASD and REPLAY-ATTACK databases.",

keywords = "Face anti-spoofing, Generative adversarial networks, Latent spatial features",

author = "Jingtian Xia and Yan Tang and Xi Jia and Linlin Shen and Zhihui Lai",

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_27",

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 = "240--249",

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

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

}

Xia, J, Tang, Y, Jia, X, Shen, L & Lai, Z 2019, Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing. 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. 240-249, 14th Chinese Conference on Biometric Recognition, CCBR 2019, Zhuzhou, China, 12/10/19. https://doi.org/10.1007/978-3-030-31456-9_27

Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing. / Xia, Jingtian; Tang, Yan; Jia, Xi et al.
Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. ed. / Zhenan Sun; Ran He; Shiguang Shan; Jianjiang Feng; Zhenhua Guo. Springer, 2019. p. 240-249 (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/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing

AU - Xia, Jingtian

AU - Tang, Yan

AU - Jia, Xi

AU - Shen, Linlin

AU - Lai, Zhihui

PY - 2019

Y1 - 2019

N2 - With the wide deployment of the face recognition system, many face attacks, such as print attack, video attack and 3D face mask, have emerged. Face anti-spoofing is very important to protect face recognition system from attack. This paper proposes a structure of generative adversarial networks with skip connection for face anti-spoofing. First, we obtain the latent spatial features of faces by training generative adversarial networks to reconstruct both real and spoof faces; second, we use the convolution neural networks to detect the spoofing faces. In this paper, the proposed method is evaluated by three public databases. The results suggest that our approach achieves as high as 98% accuracy on both CASIA-FASD and REPLAY-ATTACK databases.

AB - With the wide deployment of the face recognition system, many face attacks, such as print attack, video attack and 3D face mask, have emerged. Face anti-spoofing is very important to protect face recognition system from attack. This paper proposes a structure of generative adversarial networks with skip connection for face anti-spoofing. First, we obtain the latent spatial features of faces by training generative adversarial networks to reconstruct both real and spoof faces; second, we use the convolution neural networks to detect the spoofing faces. In this paper, the proposed method is evaluated by three public databases. The results suggest that our approach achieves as high as 98% accuracy on both CASIA-FASD and REPLAY-ATTACK databases.

KW - Face anti-spoofing

KW - Generative adversarial networks

KW - Latent spatial features

UR - http://www.scopus.com/inward/record.url?scp=85075583535&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-31456-9_27

DO - 10.1007/978-3-030-31456-9_27

M3 - Conference contribution

AN - SCOPUS:85075583535

SN - 9783030314552

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

SP - 240

EP - 249

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 -

Xia J, Tang Y, Jia X, Shen L, Lai Z. Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing. In Sun Z, He R, Shan S, Feng J, Guo Z, editors, Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. Springer. 2019. p. 240-249. (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_27

Latent Spatial Features Based on Generative Adversarial Networks for Face Anti-spoofing

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this