A landmark-based data-driven approach on 2.5D facial attractiveness computation

Shu Liu, Yang Yu Fan, Zhe Guo, Ashok K Samal, Afan Ali

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Investigating the nature and components of face attractiveness from a computational view has become an emerging topic in facial analysis research. In this paper, a multi-view (frontal and profile view, 2.5D) facial attractiveness computational model is developed to explore how face geometry affects its attractiveness. A landmark-based, data-driven method is introduced to construct a huge dimension of three kinds of geometric facial measurements, including ratios, angles, and inclinations. An incremental feature selection algorithm is proposed to systematically select the most discriminative subset of geometric features, which are finally mapped to an attractiveness score through the application of support vector regression (SVR). On a dataset of 360 facial images pre-processed from BJUT-3D Face Database and an attractiveness score dataset collected from human raters, we show that the computational model performs well with low statistic error (MSE=0.4969) and good predictability (R2=0.5756).

Original languageEnglish (US)
Pages (from-to)168-178
Number of pages11
JournalNeurocomputing
Volume238
DOIs
StatePublished - May 17 2017

Fingerprint

Error statistics
Feature extraction
Databases
Geometry
Research
Datasets

Keywords

  • 2.5 D
  • BJUT-3D
  • Data-driven
  • Facial attractiveness computation
  • Geometric features

ASJC Scopus subject areas

  • Computer Science Applications
  • Cognitive Neuroscience
  • Artificial Intelligence

Cite this

A landmark-based data-driven approach on 2.5D facial attractiveness computation. / Liu, Shu; Fan, Yang Yu; Guo, Zhe; Samal, Ashok K; Ali, Afan.

In: Neurocomputing, Vol. 238, 17.05.2017, p. 168-178.

Research output: Contribution to journalArticle

Liu, Shu ; Fan, Yang Yu ; Guo, Zhe ; Samal, Ashok K ; Ali, Afan. / A landmark-based data-driven approach on 2.5D facial attractiveness computation. In: Neurocomputing. 2017 ; Vol. 238. pp. 168-178.
@article{96284bb788714625aaf3f78562fa42ed,
title = "A landmark-based data-driven approach on 2.5D facial attractiveness computation",
abstract = "Investigating the nature and components of face attractiveness from a computational view has become an emerging topic in facial analysis research. In this paper, a multi-view (frontal and profile view, 2.5D) facial attractiveness computational model is developed to explore how face geometry affects its attractiveness. A landmark-based, data-driven method is introduced to construct a huge dimension of three kinds of geometric facial measurements, including ratios, angles, and inclinations. An incremental feature selection algorithm is proposed to systematically select the most discriminative subset of geometric features, which are finally mapped to an attractiveness score through the application of support vector regression (SVR). On a dataset of 360 facial images pre-processed from BJUT-3D Face Database and an attractiveness score dataset collected from human raters, we show that the computational model performs well with low statistic error (MSE=0.4969) and good predictability (R2=0.5756).",
keywords = "2.5 D, BJUT-3D, Data-driven, Facial attractiveness computation, Geometric features",
author = "Shu Liu and Fan, {Yang Yu} and Zhe Guo and Samal, {Ashok K} and Afan Ali",
year = "2017",
month = "5",
day = "17",
doi = "10.1016/j.neucom.2017.01.050",
language = "English (US)",
volume = "238",
pages = "168--178",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",

}

TY - JOUR

T1 - A landmark-based data-driven approach on 2.5D facial attractiveness computation

AU - Liu, Shu

AU - Fan, Yang Yu

AU - Guo, Zhe

AU - Samal, Ashok K

AU - Ali, Afan

PY - 2017/5/17

Y1 - 2017/5/17

N2 - Investigating the nature and components of face attractiveness from a computational view has become an emerging topic in facial analysis research. In this paper, a multi-view (frontal and profile view, 2.5D) facial attractiveness computational model is developed to explore how face geometry affects its attractiveness. A landmark-based, data-driven method is introduced to construct a huge dimension of three kinds of geometric facial measurements, including ratios, angles, and inclinations. An incremental feature selection algorithm is proposed to systematically select the most discriminative subset of geometric features, which are finally mapped to an attractiveness score through the application of support vector regression (SVR). On a dataset of 360 facial images pre-processed from BJUT-3D Face Database and an attractiveness score dataset collected from human raters, we show that the computational model performs well with low statistic error (MSE=0.4969) and good predictability (R2=0.5756).

AB - Investigating the nature and components of face attractiveness from a computational view has become an emerging topic in facial analysis research. In this paper, a multi-view (frontal and profile view, 2.5D) facial attractiveness computational model is developed to explore how face geometry affects its attractiveness. A landmark-based, data-driven method is introduced to construct a huge dimension of three kinds of geometric facial measurements, including ratios, angles, and inclinations. An incremental feature selection algorithm is proposed to systematically select the most discriminative subset of geometric features, which are finally mapped to an attractiveness score through the application of support vector regression (SVR). On a dataset of 360 facial images pre-processed from BJUT-3D Face Database and an attractiveness score dataset collected from human raters, we show that the computational model performs well with low statistic error (MSE=0.4969) and good predictability (R2=0.5756).

KW - 2.5 D

KW - BJUT-3D

KW - Data-driven

KW - Facial attractiveness computation

KW - Geometric features

UR - http://www.scopus.com/inward/record.url?scp=85012920504&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85012920504&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2017.01.050

DO - 10.1016/j.neucom.2017.01.050

M3 - Article

AN - SCOPUS:85012920504

VL - 238

SP - 168

EP - 178

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

ER -