Gender classification based on fuzzy SVM

Generalization ability is an important issue in gender classification. In this paper a gender classifier based on Fuzzy SVM (FSVM) is developed to improve the generalization ability. The fuzzy membership used in FSVM indicates the relativity of one personpsilas face with female/male faces set. This paper proposes a novel method of generating fuzzy membership function automatically based on Learning Vector Quantization (LVQ) learning process. The method doesnpsilat rely on the apriori information of data and has strong robustness to variations such as illumination, expression and so on. The gender classifier based on FSVM is evaluated on the FERET, CAS-PEAL, BUAA-IRIP face databases. The results show that the gender classifier presented in this paper can tolerate more variations and show good performance in generalization ability.