In this paper, we consider the problem of template selection in biometric systems as analogous to the problem, called “editing”, of selecting representative prototypes when using the 1-Nearest Neighbour classifier (NN). Four editing algorithms are used and compared by experiments with state-of-the-art template selection algorithms. Experiments are performed on a benchmark face data set. Reported results show pros and cons of editing algorithms for template selection in biometric systems.
Template Selection by Editing Algorithms: a Case Study in Face Recognition
MARCIALIS, GIAN LUCA;ROLI, FABIO
2008-01-01
Abstract
In this paper, we consider the problem of template selection in biometric systems as analogous to the problem, called “editing”, of selecting representative prototypes when using the 1-Nearest Neighbour classifier (NN). Four editing algorithms are used and compared by experiments with state-of-the-art template selection algorithms. Experiments are performed on a benchmark face data set. Reported results show pros and cons of editing algorithms for template selection in biometric systems.
File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
