Self-dependent 3D face rotational alignment using the nose region

Mehryar Emambakhsh, Adrian Evans

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

Abstract

One of the challenging issues for 3D face recognition is face alignment. Many alignment algorithms are computationally expensive, making them unsuitable for real-time biometrics, or not robust enough to detect large variations in pose. In this work, a novel algorithm for 3D face rotational alignment is proposed, that uses the nose region. After preprocessing and nose region identification, alignment is performed by applying two energy functions to the nose footprint, identified as the largest filled region in the inverted depth map. These functions are minimised using Simulated Annealing and the Levenberg-Marqurdt algorithm. The energy minimisation and segmentation procedures continue iteratively until a stopping criterion is met. The method has been applied to images from the Face Recognition Grand Challenge (FRGC) v2 dataset and the consistency of its alignment has been verified using the iterative closest point (ICP) algorithm. As a self-dependent algorithm, it does not require a pre-aligned image as a reference and also has a high computational speed, approximately three times faster than the brute force ICP technique.
Original languageEnglish
Title of host publication4th International Conference on Imaging for Crime Detection and Prevention 2011
Subtitle of host publicationICDP 2011
PublisherIEEE
Number of pages6
ISBN (Electronic)978-1-84919-565-2
DOIs
Publication statusPublished - 3 Nov 2011
Event4th International Conference on Imaging for Crime Detection and Prevention - London, UK United Kingdom
Duration: 3 Nov 20114 Nov 2011

Conference

Conference4th International Conference on Imaging for Crime Detection and Prevention
CountryUK United Kingdom
CityLondon
Period3/11/114/11/11

Fingerprint

Face recognition
Biometrics
Simulated annealing

Cite this

Emambakhsh, M., & Evans, A. (2011). Self-dependent 3D face rotational alignment using the nose region. In 4th International Conference on Imaging for Crime Detection and Prevention 2011: ICDP 2011 IEEE. https://doi.org/10.1049/ic.2011.0101

Self-dependent 3D face rotational alignment using the nose region. / Emambakhsh, Mehryar; Evans, Adrian.

4th International Conference on Imaging for Crime Detection and Prevention 2011: ICDP 2011. IEEE, 2011.

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

Emambakhsh, M & Evans, A 2011, Self-dependent 3D face rotational alignment using the nose region. in 4th International Conference on Imaging for Crime Detection and Prevention 2011: ICDP 2011. IEEE, 4th International Conference on Imaging for Crime Detection and Prevention, London, UK United Kingdom, 3/11/11. https://doi.org/10.1049/ic.2011.0101
Emambakhsh M, Evans A. Self-dependent 3D face rotational alignment using the nose region. In 4th International Conference on Imaging for Crime Detection and Prevention 2011: ICDP 2011. IEEE. 2011 https://doi.org/10.1049/ic.2011.0101
Emambakhsh, Mehryar ; Evans, Adrian. / Self-dependent 3D face rotational alignment using the nose region. 4th International Conference on Imaging for Crime Detection and Prevention 2011: ICDP 2011. IEEE, 2011.
@inproceedings{f36dc877c06049f9a5307dafe3e4dcc2,
title = "Self-dependent 3D face rotational alignment using the nose region",
abstract = "One of the challenging issues for 3D face recognition is face alignment. Many alignment algorithms are computationally expensive, making them unsuitable for real-time biometrics, or not robust enough to detect large variations in pose. In this work, a novel algorithm for 3D face rotational alignment is proposed, that uses the nose region. After preprocessing and nose region identification, alignment is performed by applying two energy functions to the nose footprint, identified as the largest filled region in the inverted depth map. These functions are minimised using Simulated Annealing and the Levenberg-Marqurdt algorithm. The energy minimisation and segmentation procedures continue iteratively until a stopping criterion is met. The method has been applied to images from the Face Recognition Grand Challenge (FRGC) v2 dataset and the consistency of its alignment has been verified using the iterative closest point (ICP) algorithm. As a self-dependent algorithm, it does not require a pre-aligned image as a reference and also has a high computational speed, approximately three times faster than the brute force ICP technique.",
author = "Mehryar Emambakhsh and Adrian Evans",
year = "2011",
month = "11",
day = "3",
doi = "10.1049/ic.2011.0101",
language = "English",
booktitle = "4th International Conference on Imaging for Crime Detection and Prevention 2011",
publisher = "IEEE",
address = "USA United States",

}

TY - GEN

T1 - Self-dependent 3D face rotational alignment using the nose region

AU - Emambakhsh, Mehryar

AU - Evans, Adrian

PY - 2011/11/3

Y1 - 2011/11/3

N2 - One of the challenging issues for 3D face recognition is face alignment. Many alignment algorithms are computationally expensive, making them unsuitable for real-time biometrics, or not robust enough to detect large variations in pose. In this work, a novel algorithm for 3D face rotational alignment is proposed, that uses the nose region. After preprocessing and nose region identification, alignment is performed by applying two energy functions to the nose footprint, identified as the largest filled region in the inverted depth map. These functions are minimised using Simulated Annealing and the Levenberg-Marqurdt algorithm. The energy minimisation and segmentation procedures continue iteratively until a stopping criterion is met. The method has been applied to images from the Face Recognition Grand Challenge (FRGC) v2 dataset and the consistency of its alignment has been verified using the iterative closest point (ICP) algorithm. As a self-dependent algorithm, it does not require a pre-aligned image as a reference and also has a high computational speed, approximately three times faster than the brute force ICP technique.

AB - One of the challenging issues for 3D face recognition is face alignment. Many alignment algorithms are computationally expensive, making them unsuitable for real-time biometrics, or not robust enough to detect large variations in pose. In this work, a novel algorithm for 3D face rotational alignment is proposed, that uses the nose region. After preprocessing and nose region identification, alignment is performed by applying two energy functions to the nose footprint, identified as the largest filled region in the inverted depth map. These functions are minimised using Simulated Annealing and the Levenberg-Marqurdt algorithm. The energy minimisation and segmentation procedures continue iteratively until a stopping criterion is met. The method has been applied to images from the Face Recognition Grand Challenge (FRGC) v2 dataset and the consistency of its alignment has been verified using the iterative closest point (ICP) algorithm. As a self-dependent algorithm, it does not require a pre-aligned image as a reference and also has a high computational speed, approximately three times faster than the brute force ICP technique.

UR - http://dx.doi.org/10.1049/ic.2011.0101

U2 - 10.1049/ic.2011.0101

DO - 10.1049/ic.2011.0101

M3 - Conference contribution

BT - 4th International Conference on Imaging for Crime Detection and Prevention 2011

PB - IEEE

ER -