A Two-Step Selective Region Ensemble for Facial Age Estimation

Authors:
Shenglan Ben;Guangda Su;Youshou Wu
Affiliations:
Department of Electronic Engineering, Tsinghua University, Beijing, China;Department of Electronic Engineering, Tsinghua University, Beijing, China;Department of Electronic Engineering, Tsinghua University, Beijing, China
Venue:
ICIC '08 Proceedings of the 4th international conference on Intelligent Computing: Advanced Intelligent Computing Theories and Applications - with Aspects of Artificial Intelligence
Year:
2008

Citing 6
Cited 0

Age classification from facial images

Computer Vision and Image Understanding
Face Recognition: Features Versus Templates

IEEE Transactions on Pattern Analysis and Machine Intelligence
Toward Automatic Simulation of Aging Effects on Face Images

IEEE Transactions on Pattern Analysis and Machine Intelligence
Age and Gender Estimation Based on Wrinkle Texture and Color of Facial Images

ICPR '02 Proceedings of the 16 th International Conference on Pattern Recognition (ICPR'02) Volume 1 - Volume 1
Automatic Age Estimation Based on Facial Aging Patterns

IEEE Transactions on Pattern Analysis and Machine Intelligence
Genetic algorithm based selective neural network ensemble

IJCAI'01 Proceedings of the 17th international joint conference on Artificial intelligence - Volume 2

Quantified Score

Hi-index	0.00

Visualization

Abstract

The effect of aging varies in different facial regions. The significance of regions' age related changes also differs in each age range. In this paper, an efficient subset is selected from all possible rectangle regions in the face image to form a global ensemble on the whole age range. Age range-based selective ensembles are also formed in a similar way. Based on those selective ensembles, a two-step selective region ensemble method is proposed for age estimation. In this framework, the first step is using the global ensemble to give a prediction of possible age range. The second step is to use the ensemble on the predicted age range to make a final estimation. Experiments show that using selective region ensemble can improve age estimation performance, and age range-based selective region ensemble is even superior to the global ensemble.