An Automatic Face Detection and Gender Classification

Abstract

Face detection is one of the most important issues in the identification and authentication systems that use biometric features. In this paper, we present algorithms for detecting skin color. The selection and implementation of an algorithm for an automated authentication system and face detection can significantly improve the effectiveness of such a system. The second group of algorithms is responsible for gender classification. Due to their complexity, they are not suitable for nowadays mobile devices but can be used in systems working on more demanding machines. The selection and implementation of algorithms have accuracy about 80-90%. We show the advantages and disadvantages of those methods and future challenges to the researches.