Combining fingerprint and hand-geometry verification decisions

Authors:
Kar-Ann Toh;Wei Xiong;Wei Xiong;Wei-Yun Yau;Xudong Jiang
Affiliations:
Institute for Infocomm Research, Terrace, Singapore;Institute for Infocomm Research, Terrace, Singapore;Institute for Infocomm Research, Terrace, Singapore;Institute for Infocomm Research, Terrace, Singapore;Institute for Infocomm Research, Terrace, Singapore
Venue:
AVBPA'03 Proceedings of the 4th international conference on Audio- and video-based biometric person authentication
Year:
2003

Citing 8
Cited 2

Multilayer feedforward networks are universal approximators

Neural Networks
On detecting dominant points

Pattern Recognition
On Combining Classifiers

IEEE Transactions on Pattern Analysis and Machine Intelligence
Integrating Faces and Fingerprints for Personal Identification

IEEE Transactions on Pattern Analysis and Machine Intelligence
Optimal Linear Combination of Neural Networks for Improving Classification Performance

IEEE Transactions on Pattern Analysis and Machine Intelligence
Biometric Identification through Hand Geometry Measurements

IEEE Transactions on Pattern Analysis and Machine Intelligence
A Principal Components Approach to Combining Regression Estimates

Machine Learning
Information Fusion in Biometrics

AVBPA '01 Proceedings of the Third International Conference on Audio- and Video-Based Biometric Person Authentication

Hand Geometry Verification Using Time Series Representation

KES '07 Knowledge-Based Intelligent Information and Engineering Systems and the XVII Italian Workshop on Neural Networks on Proceedings of the 11th International Conference
Model-guided deformable hand shape recognition without positioning aids

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper proposes to combine the fingerprint and hand-geometry verification decisions using a reduced multivariate polynomials model. Main advantage of this method over those neural network based methods is that only a single step is required for training and the training is optimal. Numerical experiments using a database containing over 100 identities show significant improvement of Receiver Operating Characteristics as compared to that of individual biometrics. Moreover, the result outperforms a few commonly used methods using the same database.