Surface Normals Based Landmarking for 3D Face Recognition Using Photometric Stereo Captures

Jiangning Gao, Mark Hansen, Melvyn Smith , Adrian Evans

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In recent decades, many 3D data acquisition methods have been developed to provide accurate and cost-effective 3D captures of the human face. An example system, which can accommodate both research and commercial applications, is the Photoface device. Photoface is based on the photometric stereo imaging technique. To improve the recognition performance using Photoface captures, a novel landmarking algorithm is first proposed by thresholding surface normals maps. The development of landmarking algorithms specifically for photometric stereo captures enables region-based feature extraction and fills a gap in the 3D face landmarking literature. Nasal curves and spherical patches are then used
respectively for recognition and are evaluated on the 3DE-VISIR database, which contains Photoface captures with expressions. The neutral vs. non-neutral matching results demonstrate high face recognition performance using spherical patches and a KFA classifier, achieving a R1RR of 97.26% when only 24 patches are selected for matching.
Original languageEnglish
Title of host publication3rd International Conference on Biometric Engineering and Applications (ICBEA 2019)
PublisherAssociation for Computing Machinery
ISBN (Print)978-1-4503-6305-1
Publication statusAccepted/In press - 5 Mar 2019
Event3rd International Conference on Biometric Engineering and Applications - Stockholm, Sweden
Duration: 29 May 201931 May 2019
http://www.icbea.org/

Conference

Conference3rd International Conference on Biometric Engineering and Applications
Abbreviated titleICBEA 2019
CountrySweden
CityStockholm
Period29/05/1931/05/19
Internet address

Cite this

Gao, J., Hansen, M., Smith , M., & Evans, A. (Accepted/In press). Surface Normals Based Landmarking for 3D Face Recognition Using Photometric Stereo Captures. In 3rd International Conference on Biometric Engineering and Applications (ICBEA 2019) Association for Computing Machinery.

Surface Normals Based Landmarking for 3D Face Recognition Using Photometric Stereo Captures. / Gao, Jiangning; Hansen, Mark; Smith , Melvyn; Evans, Adrian.

3rd International Conference on Biometric Engineering and Applications (ICBEA 2019). Association for Computing Machinery, 2019.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Gao, J, Hansen, M, Smith , M & Evans, A 2019, Surface Normals Based Landmarking for 3D Face Recognition Using Photometric Stereo Captures. in 3rd International Conference on Biometric Engineering and Applications (ICBEA 2019). Association for Computing Machinery, 3rd International Conference on Biometric Engineering and Applications, Stockholm, Sweden, 29/05/19.
Gao J, Hansen M, Smith M, Evans A. Surface Normals Based Landmarking for 3D Face Recognition Using Photometric Stereo Captures. In 3rd International Conference on Biometric Engineering and Applications (ICBEA 2019). Association for Computing Machinery. 2019
Gao, Jiangning ; Hansen, Mark ; Smith , Melvyn ; Evans, Adrian. / Surface Normals Based Landmarking for 3D Face Recognition Using Photometric Stereo Captures. 3rd International Conference on Biometric Engineering and Applications (ICBEA 2019). Association for Computing Machinery, 2019.
@inproceedings{d85dcc04bd14479b9ae9f8e3d855b425,
title = "Surface Normals Based Landmarking for 3D Face Recognition Using Photometric Stereo Captures",
abstract = "In recent decades, many 3D data acquisition methods have been developed to provide accurate and cost-effective 3D captures of the human face. An example system, which can accommodate both research and commercial applications, is the Photoface device. Photoface is based on the photometric stereo imaging technique. To improve the recognition performance using Photoface captures, a novel landmarking algorithm is first proposed by thresholding surface normals maps. The development of landmarking algorithms specifically for photometric stereo captures enables region-based feature extraction and fills a gap in the 3D face landmarking literature. Nasal curves and spherical patches are then usedrespectively for recognition and are evaluated on the 3DE-VISIR database, which contains Photoface captures with expressions. The neutral vs. non-neutral matching results demonstrate high face recognition performance using spherical patches and a KFA classifier, achieving a R1RR of 97.26{\%} when only 24 patches are selected for matching.",
author = "Jiangning Gao and Mark Hansen and Melvyn Smith and Adrian Evans",
year = "2019",
month = "3",
day = "5",
language = "English",
isbn = "978-1-4503-6305-1",
booktitle = "3rd International Conference on Biometric Engineering and Applications (ICBEA 2019)",
publisher = "Association for Computing Machinery",
address = "USA United States",

}

TY - GEN

T1 - Surface Normals Based Landmarking for 3D Face Recognition Using Photometric Stereo Captures

AU - Gao, Jiangning

AU - Hansen, Mark

AU - Smith , Melvyn

AU - Evans, Adrian

PY - 2019/3/5

Y1 - 2019/3/5

N2 - In recent decades, many 3D data acquisition methods have been developed to provide accurate and cost-effective 3D captures of the human face. An example system, which can accommodate both research and commercial applications, is the Photoface device. Photoface is based on the photometric stereo imaging technique. To improve the recognition performance using Photoface captures, a novel landmarking algorithm is first proposed by thresholding surface normals maps. The development of landmarking algorithms specifically for photometric stereo captures enables region-based feature extraction and fills a gap in the 3D face landmarking literature. Nasal curves and spherical patches are then usedrespectively for recognition and are evaluated on the 3DE-VISIR database, which contains Photoface captures with expressions. The neutral vs. non-neutral matching results demonstrate high face recognition performance using spherical patches and a KFA classifier, achieving a R1RR of 97.26% when only 24 patches are selected for matching.

AB - In recent decades, many 3D data acquisition methods have been developed to provide accurate and cost-effective 3D captures of the human face. An example system, which can accommodate both research and commercial applications, is the Photoface device. Photoface is based on the photometric stereo imaging technique. To improve the recognition performance using Photoface captures, a novel landmarking algorithm is first proposed by thresholding surface normals maps. The development of landmarking algorithms specifically for photometric stereo captures enables region-based feature extraction and fills a gap in the 3D face landmarking literature. Nasal curves and spherical patches are then usedrespectively for recognition and are evaluated on the 3DE-VISIR database, which contains Photoface captures with expressions. The neutral vs. non-neutral matching results demonstrate high face recognition performance using spherical patches and a KFA classifier, achieving a R1RR of 97.26% when only 24 patches are selected for matching.

M3 - Conference contribution

SN - 978-1-4503-6305-1

BT - 3rd International Conference on Biometric Engineering and Applications (ICBEA 2019)

PB - Association for Computing Machinery

ER -