Face Recognition Using Dense SIFT Feature Alignment

This paper addresses face recognition problem in a more challenging scenario where the training and test samples are both subject to the visual variations of poses, expressions and misalignments. We employ dense Scale-invariant feature transform（SIFT） feature matching as a generic transformation to...

Full description

Saved in:
Bibliographic Details
Published inChinese Journal of Electronics Vol. 25; no. 6; pp. 1034 - 1039
Main Authors Zhou, Quan, Shafiq, ur Rehman, Zhou, Yu, Wei, Xin, Wang, Lei, Zheng, Baoyu
Format Journal Article
LanguageEnglish
Published Published by the IET on behalf of the CIE 01.11.2016
Subjects
dense scale‐invariant feature transform feature matching
Dense SIFT feature alignment
dense SIFT feature matching
Face recognition
generic transformation
image matching
image representation
learning (artificial intelligence)
Sparse representation
test samples
training samples
transforms
Face recognition
Sparse representation
Dense SIFT feature alignment
Online AccessGet full text

Cover

More Information
Summary:This paper addresses face recognition problem in a more challenging scenario where the training and test samples are both subject to the visual variations of poses, expressions and misalignments. We employ dense Scale-invariant feature transform（SIFT） feature matching as a generic transformation to roughly align training samples; and then identify input facial images via an improved sparse representation model based on the aligned training samples. Compared with previous methods, the extensive experimental results demonstrate the effectiveness of our method for the task of face recognition on three benchmark datasets.
Bibliography:This paper addresses face recognition problem in a more challenging scenario where the training and test samples are both subject to the visual variations of poses, expressions and misalignments. We employ dense Scale-invariant feature transform（SIFT） feature matching as a generic transformation to roughly align training samples; and then identify input facial images via an improved sparse representation model based on the aligned training samples. Compared with previous methods, the extensive experimental results demonstrate the effectiveness of our method for the task of face recognition on three benchmark datasets.
Face recognition Dense SIFT feature alignment Sparse representation
10-1284/TN
ISSN:1022-4653
2075-5597
2075-5597
DOI:10.1049/cje.2016.10.001