Face recognition based on recursive bayesian fusion of multiple signals and results from expert classifier sets

Authors:
Michael Hild;Ryo Kuzui
Affiliations:
Graduate School of Engineering, Osaka Electro Communication University, Neyagawa, Osaka, Japan;Graduate School of Engineering, Osaka Electro Communication University, Neyagawa, Osaka, Japan
Venue:
AVBPA'05 Proceedings of the 5th international conference on Audio- and Video-Based Biometric Person Authentication
Year:
2005

Citing 7
Cited 0

The FERET Evaluation Methodology for Face-Recognition Algorithms

IEEE Transactions on Pattern Analysis and Machine Intelligence
Dynamic Vision: From Images to Face Recognition

Dynamic Vision: From Images to Face Recognition
An Experimental Comparison of Classifier Fusion Rules for Multimodal Personal Identity Verification Systems

MCS '02 Proceedings of the Third International Workshop on Multiple Classifier Systems
Seeing People in the Dark: Face Recognition in Infrared Images

BMCV '02 Proceedings of the Second International Workshop on Biologically Motivated Computer Vision
Multimodal Biometric Authentication Using Quality Signals in Mobile Communications

ICIAP '03 Proceedings of the 12th International Conference on Image Analysis and Processing
Biometrics: A Grand Challenge

ICPR '04 Proceedings of the Pattern Recognition, 17th International Conference on (ICPR'04) Volume 2 - Volume 02
Information Fusion in Face Identification

ICPR '04 Proceedings of the Pattern Recognition, 17th International Conference on (ICPR'04) Volume 3 - Volume 03

Quantified Score

Hi-index	0.00

Visualization

Abstract

We report on a system for person identification based on face images. The system uses sequences of visual wavelength intensity and thermal image pairs as input and carries out classification with a set of expert classifiers (such as ANN or SVM) for each input signal separately. The decisions of the classifiers are integrated both over the two signals and over time as new image pairs arrive, using stochastic recursive inference based on Bayes formula. Our experimental results indicate that both recognition and rejection rates are higher than those for the expert classifiers alone.