3D facial shape reconstruction using macro- and micro-level features from high resolution facial images

Jaeik Jo, Hyunjun Kim, Jaihie Kim

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Three-dimensional (3D) facial modeling and stereo matching-based methods are widely used for 3D facial reconstruction from 2D single-view and multiple-view images. However, these methods cannot realistically reconstruct 3D faces because they use insufficient numbers of macro-level Facial Feature Points (FFPs). This paper proposes an accurate and person-specific 3D facial reconstruction method that uses ample numbers of macro- and micro-level FFPs to enable coverage of all facial regions of high resolution facial images. Comparisons of 3D facial images reconstructed using the proposed method for ground-truth 3D facial images from the Bosphorus 3D database show that the method is superior to a conventional Active Appearance Model-Structure from Motion (AAM + SfM)-based method in terms of average 3D root mean square error between the reconstructed and ground-truth 3D faces. Further, the proposed method achieved outstanding accuracy in local facial regions such as the cheek—areas where extraction of FFPs is difficult for existing methods.

Original languageEnglish
Pages (from-to)1-9
Number of pages9
JournalImage and Vision Computing
Volume64
DOIs
Publication statusPublished - 2017 Aug 1

Fingerprint

Macros
Model structures
Mean square error

All Science Journal Classification (ASJC) codes

  • Signal Processing
  • Computer Vision and Pattern Recognition

Cite this

@article{6d982710f00a4e1893d677bd16a07fe0,
title = "3D facial shape reconstruction using macro- and micro-level features from high resolution facial images",
abstract = "Three-dimensional (3D) facial modeling and stereo matching-based methods are widely used for 3D facial reconstruction from 2D single-view and multiple-view images. However, these methods cannot realistically reconstruct 3D faces because they use insufficient numbers of macro-level Facial Feature Points (FFPs). This paper proposes an accurate and person-specific 3D facial reconstruction method that uses ample numbers of macro- and micro-level FFPs to enable coverage of all facial regions of high resolution facial images. Comparisons of 3D facial images reconstructed using the proposed method for ground-truth 3D facial images from the Bosphorus 3D database show that the method is superior to a conventional Active Appearance Model-Structure from Motion (AAM + SfM)-based method in terms of average 3D root mean square error between the reconstructed and ground-truth 3D faces. Further, the proposed method achieved outstanding accuracy in local facial regions such as the cheek—areas where extraction of FFPs is difficult for existing methods.",
author = "Jaeik Jo and Hyunjun Kim and Jaihie Kim",
year = "2017",
month = "8",
day = "1",
doi = "10.1016/j.imavis.2017.05.001",
language = "English",
volume = "64",
pages = "1--9",
journal = "Image and Vision Computing",
issn = "0262-8856",
publisher = "Elsevier Limited",

}

3D facial shape reconstruction using macro- and micro-level features from high resolution facial images. / Jo, Jaeik; Kim, Hyunjun; Kim, Jaihie.

In: Image and Vision Computing, Vol. 64, 01.08.2017, p. 1-9.

Research output: Contribution to journalArticle

TY - JOUR

T1 - 3D facial shape reconstruction using macro- and micro-level features from high resolution facial images

AU - Jo, Jaeik

AU - Kim, Hyunjun

AU - Kim, Jaihie

PY - 2017/8/1

Y1 - 2017/8/1

N2 - Three-dimensional (3D) facial modeling and stereo matching-based methods are widely used for 3D facial reconstruction from 2D single-view and multiple-view images. However, these methods cannot realistically reconstruct 3D faces because they use insufficient numbers of macro-level Facial Feature Points (FFPs). This paper proposes an accurate and person-specific 3D facial reconstruction method that uses ample numbers of macro- and micro-level FFPs to enable coverage of all facial regions of high resolution facial images. Comparisons of 3D facial images reconstructed using the proposed method for ground-truth 3D facial images from the Bosphorus 3D database show that the method is superior to a conventional Active Appearance Model-Structure from Motion (AAM + SfM)-based method in terms of average 3D root mean square error between the reconstructed and ground-truth 3D faces. Further, the proposed method achieved outstanding accuracy in local facial regions such as the cheek—areas where extraction of FFPs is difficult for existing methods.

AB - Three-dimensional (3D) facial modeling and stereo matching-based methods are widely used for 3D facial reconstruction from 2D single-view and multiple-view images. However, these methods cannot realistically reconstruct 3D faces because they use insufficient numbers of macro-level Facial Feature Points (FFPs). This paper proposes an accurate and person-specific 3D facial reconstruction method that uses ample numbers of macro- and micro-level FFPs to enable coverage of all facial regions of high resolution facial images. Comparisons of 3D facial images reconstructed using the proposed method for ground-truth 3D facial images from the Bosphorus 3D database show that the method is superior to a conventional Active Appearance Model-Structure from Motion (AAM + SfM)-based method in terms of average 3D root mean square error between the reconstructed and ground-truth 3D faces. Further, the proposed method achieved outstanding accuracy in local facial regions such as the cheek—areas where extraction of FFPs is difficult for existing methods.

UR - http://www.scopus.com/inward/record.url?scp=85020184603&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020184603&partnerID=8YFLogxK

U2 - 10.1016/j.imavis.2017.05.001

DO - 10.1016/j.imavis.2017.05.001

M3 - Article

VL - 64

SP - 1

EP - 9

JO - Image and Vision Computing

JF - Image and Vision Computing

SN - 0262-8856

ER -