Verification of aging faces using local ternary patterns and Q-stack classifier

Authors:
Andrzej Drygajlo;Weifeng Li;Kewei Zhu
Affiliations:
Speech Processing and Biometrics Group, Swiss Federal Institute of Technology Lausanne, Lausanne, Switzerland;Speech Processing and Biometrics Group, Swiss Federal Institute of Technology Lausanne, Lausanne, Switzerland;Speech Processing and Biometrics Group, Swiss Federal Institute of Technology Lausanne, Lausanne, Switzerland
Venue:
BioID_MultiComm'09 Proceedings of the 2009 joint COST 2101 and 2102 international conference on Biometric ID management and multimodal communication
Year:
2009

Citing 6
Cited 1

MORPH: A Longitudinal Image Database of Normal Adult Age-Progression

FGR '06 Proceedings of the 7th International Conference on Automatic Face and Gesture Recognition
A person-specific, rigorous aging model of the human face

Pattern Recognition Letters
Q-stack: uni- and multimodal classifier stacking with quality measures

MCS'07 Proceedings of the 7th international conference on Multiple classifier systems
Enhanced local texture feature sets for face recognition under difficult lighting conditions

AMFG'07 Proceedings of the 3rd international conference on Analysis and modeling of faces and gestures
Face Verification Across Age Progression

IEEE Transactions on Image Processing
Improving classification with class-independent quality measures: Q-stack in face verification

ICB'07 Proceedings of the 2007 international conference on Advances in Biometrics

Adult face recognition in score-age-quality classification space

BioID'11 Proceedings of the COST 2101 European conference on Biometrics and ID management

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper deals with the influence of age progression on the performance of face verification systems. This is a challenging and largely open research problem that deserves more and more attention. Aging affects both the shape of the face and its texture, leading to a failure in the face verification task. In this paper, the aging influence on the face verification system using local ternary patterns is managed by a Q-stack aging model, which uses the age as a class-independent metadata quality measure together with baseline classifier scores in order to obtain better recognition rates. This allows for increased long-term class separation by a decision boundary in the score-quality measure space using a short-term enrolment model. This new method, based on the concept of classifier stacking, compares favorably with the conventional face verification approach which uses decision boundary calculated only in the score space at the time of enrolment.