Abstract
This paper explores several ways of combining the MASKS and MKL-based classifiers which we specifically designed for the fingerprint classification task. The advantages of coupling these distinct techniques are well evident; in particular, in the case of exclusive classification, the FBI challenge requiring a classification error ≤1% at 20% rejection was broken on NIST-DB14. © Springer-Verlag Berlin Heidelberg 2000.
Cite
CITATION STYLE
Cappelli, R., Maio, D., & Maltoni, D. (2000). Combining fingerprint classifiers. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1857 LNCS, pp. 351–361). Springer Verlag. https://doi.org/10.1007/3-540-45014-9_34
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
