Mixture of experts for classification of gender, ethnic origin, and pose of human faces

In this paper we describe the application of mixtures of experts on gender and ethnic classification of human faces, and Dose classification, and show their feasibility on the FERET database of facial images. The FERET databa se allows us to demonstrate performance on hundreds or thousands of images. The mixture of experts is implemented using the "divide and conquer" modul arity principle with respect to the granularity and/or the locality of info rmation, The mixture of experts consists of ensembles of radial basis funct ions (RBFs), Inductive decision trees (DTs) and support vector machines (SV Ms) implement the "gating network" components for deciding which of the exp erts should be used to determine the classification output and to restrict the support of the input space. Both the ensemble of RBF's (ERBF) and SVM u se the RBF kernel ("expert") for gating the inputs. Our experimental result s yield an average accuracy rate of 96% on gender classification and 92% on ethnic classification using the ERBF/DT approach from frontal face images, while the SVM yield 100% on pose classification.