EIGENFACE RECOGNITION USING DIFFERENT TRAINING DATA SIZES

paper explores the relationship between eigenface recognition performance and different training data sets. Using the Multilevel Dominant Eigenvector Estimation (MDEE) method we are able to compute eigenfaces from a large number of training samples. This allows us to compare the recognition performance using different training data sizes. Experimental results show that increasing the number of people benefits the recognition performance more than increasing the number of images per person.