Fusing holistic and local representation for face recognition

In this paper, we propose a novel face recognition algorithm. The proposed algorithm first denotes the original and mirror face images of a same person by complex vectors and apply a holistic representation method to them. Then we extend a local representation method to the original and mirror face...

Full description

Saved in:
Bibliographic Details
Published in2016 3rd International Conference on Systems and Informatics (ICSAI) pp. 92 - 96
Main Authors Li, Zhengming, Yang, Nanyue
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.11.2016
Subjects
Algorithm design and analysis
Classification algorithms
collaborative representation
Face
Face recognition
holistic and local representation
Mirrors
Training
Online AccessGet full text

Cover

More Information
Summary:In this paper, we propose a novel face recognition algorithm. The proposed algorithm first denotes the original and mirror face images of a same person by complex vectors and apply a holistic representation method to them. Then we extend a local representation method to the original and mirror face images. Besides the above novelty, this paper has the following significant contribution: it designs an elaborated algorithm to integrate the used complex vector holistic representation method and extended local representation method and lead to very high accuracy. Actually, the proposed algorithm achieves higher recognition rates than those of all of the other algorithms.
DOI:10.1109/ICSAI.2016.7810936