GEFeS: Genetic & evolutionary feature selection for periocular biometric recognition

Gerry Vernon Dozier, Kamilah Purrington, Khary Popplewell, Joseph Shelton, Tamirat T. Abegaz, Kelvin S. Bryant, Joshua Adams, Damon L. Woodard, Philip E. Miller · 2011

In this paper, we introduce the concept of genetic & evolutionary feature selection (GEFeS) for periocular biometric recognition. Our results show that GEFeS dramatically reduces the number of features needed for periocular recognition as well as increases recognition accuracy.

Read the paper · More papers on PaperTik