Study of a committee of neural networks for biometric hand-geometry recognition

TL;DR

This paper investigates the use of neural network committees for biometric hand-geometry recognition, demonstrating improved recognition rates over single networks and highlighting differences between identification and verification tasks.

Contribution

It introduces the concept of neural network committees for biometric recognition and compares their performance to single networks and multi-start algorithms.

Findings

Committee of neural networks improves recognition rates.

No strong correlation between identification and verification performance.

Neural network committees outperform single networks and multi-start algorithms.

Abstract

This Paper studies different committees of neural networks for biometric pattern recognition. We use the neural nets as classifiers for identification and verification purposes. We show that a committee of nets can improve the recognition rates when compared with a multi-start initialization algo-rithm that just picks up the neural net which offers the best performance. On the other hand, we found that there is no strong correlation between identifi-cation and verification applications using the same classifier.