Novel low level local features for 3D expression invariant face recognition

Munawar Hayat; Mohammed Bennamoun; Yinjie Lei; Amar El-Sallam

doi:10.1109/ICARCV.2012.6485157

Novel low level local features for 3D expression invariant face recognition

Munawar Hayat, Mohammed Bennamoun, Yinjie Lei, Amar El-Sallam

Research output: A Conference proceeding or a Chapter in Book › Conference contribution › peer-review

1 Downloads (Pure)

Abstract

In this paper, we present a system based on novel low level local features to recognize 3D faces under varying facial expressions. Our local features are obtained by combinatorially selecting two points from expression insensitive semi-rigid portions of the face. The curve length between the two points is computed and the distribution of such curve lengths is used as a feature vector to model the geometric shape distribution of the face. Our proposed features are very simple to compute yet highly distinctive and discriminating. Kernel Fisher discriminant analysis is used for feature optimization, followed by a linear support vector machine classifier for recognition. The system is extensively tested on 2500 facial scans of BU 3DFE dataset. Our experimental results show that the proposed system achieves a very high average classification rate of 99.17% and verification rates of 99.0% and above for a false acceptance rate of 0.001.

Original language	English
Title of host publication	2012 12th international conference on control automation robotics & vision (ICARCV)
Editors	Danwei Wang, Chien Chern Cheah
Place of Publication	China
Publisher	IEEE, Institute of Electrical and Electronics Engineers
Pages	194-198
Number of pages	5
ISBN (Electronic)	9781467318723
ISBN (Print)	9781467318716
DOIs	https://doi.org/10.1109/ICARCV.2012.6485157
Publication status	Published - 5 Dec 2012
Externally published	Yes
Event	12th international conference on control automation robotics and vision ICARCV 2012 - Guangzhou, Guangzhou, China Duration: 5 Dec 2012 → 7 Dec 2012

Conference

Conference	12th international conference on control automation robotics and vision ICARCV 2012
Country/Territory	China
City	Guangzhou
Period	5/12/12 → 7/12/12

Access to Document

10.1109/ICARCV.2012.6485157

Cite this

@inproceedings{7cc04f5573a047d0a8ec2b1d9fda8b4b,

title = "Novel low level local features for 3D expression invariant face recognition",

abstract = "In this paper, we present a system based on novel low level local features to recognize 3D faces under varying facial expressions. Our local features are obtained by combinatorially selecting two points from expression insensitive semi-rigid portions of the face. The curve length between the two points is computed and the distribution of such curve lengths is used as a feature vector to model the geometric shape distribution of the face. Our proposed features are very simple to compute yet highly distinctive and discriminating. Kernel Fisher discriminant analysis is used for feature optimization, followed by a linear support vector machine classifier for recognition. The system is extensively tested on 2500 facial scans of BU 3DFE dataset. Our experimental results show that the proposed system achieves a very high average classification rate of 99.17% and verification rates of 99.0% and above for a false acceptance rate of 0.001.",

keywords = "Face-recognition, Pattern-recognition, Computer-vision",

author = "Munawar Hayat and Mohammed Bennamoun and Yinjie Lei and Amar El-Sallam",

year = "2012",

month = dec,

day = "5",

doi = "10.1109/ICARCV.2012.6485157",

language = "English",

isbn = "9781467318716",

pages = "194--198",

editor = "Danwei Wang and Cheah, {Chien Chern}",

booktitle = "2012 12th international conference on control automation robotics & vision (ICARCV)",

publisher = "IEEE, Institute of Electrical and Electronics Engineers",

address = "United States",

note = "12th international conference on control automation robotics and vision ICARCV 2012 ; Conference date: 05-12-2012 Through 07-12-2012",

}

Hayat, M, Bennamoun, M, Lei, Y & El-Sallam, A 2012, Novel low level local features for 3D expression invariant face recognition. in D Wang & CC Cheah (eds), 2012 12th international conference on control automation robotics & vision (ICARCV). IEEE, Institute of Electrical and Electronics Engineers, China, pp. 194-198, 12th international conference on control automation robotics and vision ICARCV 2012, Guangzhou, China, 5/12/12. https://doi.org/10.1109/ICARCV.2012.6485157

Novel low level local features for 3D expression invariant face recognition. / Hayat, Munawar; Bennamoun, Mohammed; Lei, Yinjie et al.
2012 12th international conference on control automation robotics & vision (ICARCV). ed. / Danwei Wang; Chien Chern Cheah. China: IEEE, Institute of Electrical and Electronics Engineers, 2012. p. 194-198.

Research output: A Conference proceeding or a Chapter in Book › Conference contribution › peer-review

TY - GEN

T1 - Novel low level local features for 3D expression invariant face recognition

AU - Hayat, Munawar

AU - Bennamoun, Mohammed

AU - Lei, Yinjie

AU - El-Sallam, Amar

PY - 2012/12/5

Y1 - 2012/12/5

N2 - In this paper, we present a system based on novel low level local features to recognize 3D faces under varying facial expressions. Our local features are obtained by combinatorially selecting two points from expression insensitive semi-rigid portions of the face. The curve length between the two points is computed and the distribution of such curve lengths is used as a feature vector to model the geometric shape distribution of the face. Our proposed features are very simple to compute yet highly distinctive and discriminating. Kernel Fisher discriminant analysis is used for feature optimization, followed by a linear support vector machine classifier for recognition. The system is extensively tested on 2500 facial scans of BU 3DFE dataset. Our experimental results show that the proposed system achieves a very high average classification rate of 99.17% and verification rates of 99.0% and above for a false acceptance rate of 0.001.

AB - In this paper, we present a system based on novel low level local features to recognize 3D faces under varying facial expressions. Our local features are obtained by combinatorially selecting two points from expression insensitive semi-rigid portions of the face. The curve length between the two points is computed and the distribution of such curve lengths is used as a feature vector to model the geometric shape distribution of the face. Our proposed features are very simple to compute yet highly distinctive and discriminating. Kernel Fisher discriminant analysis is used for feature optimization, followed by a linear support vector machine classifier for recognition. The system is extensively tested on 2500 facial scans of BU 3DFE dataset. Our experimental results show that the proposed system achieves a very high average classification rate of 99.17% and verification rates of 99.0% and above for a false acceptance rate of 0.001.

KW - Face-recognition

KW - Pattern-recognition

KW - Computer-vision

U2 - 10.1109/ICARCV.2012.6485157

DO - 10.1109/ICARCV.2012.6485157

M3 - Conference contribution

SN - 9781467318716

SP - 194

EP - 198

BT - 2012 12th international conference on control automation robotics & vision (ICARCV)

A2 - Wang, Danwei

A2 - Cheah, Chien Chern

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - China

T2 - 12th international conference on control automation robotics and vision ICARCV 2012

Y2 - 5 December 2012 through 7 December 2012

ER -

Novel low level local features for 3D expression invariant face recognition

Abstract

Conference

Access to Document

Fingerprint

Cite this