3D face recognition under pose varying environments

Hwanjong Song, Ukil Yang, Kwanghoon Sohn

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

This paper describes a novel three-dimensional (3D) face recognition method when the head pose varies severely. Given an unknown 3D face, we extract several invariant facial features based on the facial geometry. We perform a Error Compensated Singular Value Decomposition (EC-SVD) for 3D face recognition. The novelty of the proposed EC-SVD procedure lies in compensating for the error for each rotation axis accurately. When the pose of a face is estimated, we propose a novel two-stage 3D face recognition algorithm. We first select face candidates based on the 3D-based nearest neighbor classifier and then the depth-based template matching is performed for final recognition. From the experimental results, less than a 0.2 degree error in average has been achieved for the 3D head pose estimation and all faces are correctly matched based on our proposed method.

Original languageEnglish
Pages (from-to)333-347
Number of pages15
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2908
Publication statusPublished - 2004 Dec 1

Fingerprint

Varying Environment
Face recognition
Face Recognition
Face
Singular value decomposition
Template matching
Template Matching
Pose Estimation
Recognition Algorithm
Nearest Neighbor
Classifiers
Classifier
Vary
Unknown
Three-dimensional
Invariant
Geometry
Experimental Results

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

@article{4d4b80d3d7ff4caabf264257e5d5e2b3,
title = "3D face recognition under pose varying environments",
abstract = "This paper describes a novel three-dimensional (3D) face recognition method when the head pose varies severely. Given an unknown 3D face, we extract several invariant facial features based on the facial geometry. We perform a Error Compensated Singular Value Decomposition (EC-SVD) for 3D face recognition. The novelty of the proposed EC-SVD procedure lies in compensating for the error for each rotation axis accurately. When the pose of a face is estimated, we propose a novel two-stage 3D face recognition algorithm. We first select face candidates based on the 3D-based nearest neighbor classifier and then the depth-based template matching is performed for final recognition. From the experimental results, less than a 0.2 degree error in average has been achieved for the 3D head pose estimation and all faces are correctly matched based on our proposed method.",
author = "Hwanjong Song and Ukil Yang and Kwanghoon Sohn",
year = "2004",
month = "12",
day = "1",
language = "English",
volume = "2908",
pages = "333--347",
journal = "Lecture Notes in Computer Science",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - 3D face recognition under pose varying environments

AU - Song, Hwanjong

AU - Yang, Ukil

AU - Sohn, Kwanghoon

PY - 2004/12/1

Y1 - 2004/12/1

N2 - This paper describes a novel three-dimensional (3D) face recognition method when the head pose varies severely. Given an unknown 3D face, we extract several invariant facial features based on the facial geometry. We perform a Error Compensated Singular Value Decomposition (EC-SVD) for 3D face recognition. The novelty of the proposed EC-SVD procedure lies in compensating for the error for each rotation axis accurately. When the pose of a face is estimated, we propose a novel two-stage 3D face recognition algorithm. We first select face candidates based on the 3D-based nearest neighbor classifier and then the depth-based template matching is performed for final recognition. From the experimental results, less than a 0.2 degree error in average has been achieved for the 3D head pose estimation and all faces are correctly matched based on our proposed method.

AB - This paper describes a novel three-dimensional (3D) face recognition method when the head pose varies severely. Given an unknown 3D face, we extract several invariant facial features based on the facial geometry. We perform a Error Compensated Singular Value Decomposition (EC-SVD) for 3D face recognition. The novelty of the proposed EC-SVD procedure lies in compensating for the error for each rotation axis accurately. When the pose of a face is estimated, we propose a novel two-stage 3D face recognition algorithm. We first select face candidates based on the 3D-based nearest neighbor classifier and then the depth-based template matching is performed for final recognition. From the experimental results, less than a 0.2 degree error in average has been achieved for the 3D head pose estimation and all faces are correctly matched based on our proposed method.

UR - http://www.scopus.com/inward/record.url?scp=15944369510&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=15944369510&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:15944369510

VL - 2908

SP - 333

EP - 347

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -