Facial age range estimation with extreme learning machines

Face image based age estimation is an approach to classify face images into one of several pre-defined age-groups. It is challenging because facial aging variation is specific to a given individual and is determined by the person's gene and many external factors, such as exposure, weather, gend...

Full description

Saved in:
Bibliographic Details
Published in:Neurocomputing (Amsterdam) 2015-02, Vol.149, p.364-372
Main Authors: Sai, Phyo-Kyaw, Wang, Jian-Gang, Teoh, Eam-Khwang
Format: Article
Language:English
Subjects:
Age
Boosting
Classification
Climatology
ECOC
Extreme learning machines
Facial
Facial age range estimation
Genes
Learning
Robustness
Weather
Weathering
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Face image based age estimation is an approach to classify face images into one of several pre-defined age-groups. It is challenging because facial aging variation is specific to a given individual and is determined by the person's gene and many external factors, such as exposure, weather, gender, and living style. Age estimation is a multiclass problem and the number of classes to predict is quite large. There surely is facial aging trend and faces from closed age range have some similar facial aging features. It is difficult to say there are distinct facial aging features for an age. Facial aging features are found to be overlapped among nearby age groups along the aging life and are continuous in nature. In this paper, we emphasised our work on age range estimation with four pre-defined classes. We applied a fast and efficient machine learning method: extreme learning machines, to solve the age categorization problem. Local Gabor Binary Patterns, Biologically Inspired Feature and Gabor were adopted to represent face image. Age estimation was performed on three different aging datasets and experimental results are reported to demonstrate its effectiveness and robustness.
ISSN:0925-2312
1872-8286
DOI:10.1016/j.neucom.2014.03.074