Masters in Statistics Oral Defense: "Statistical Learning Methods for Facial Recognition"

Abstract: Facial recognition techniques have received a great deal of attention over the past few decades, and have been applied to daily life very often. This thesis studies the performance of several statistical facial recognition methods applied to two different databases, which contain more diverse poses and illumination conditions compared to standard databases pose and illumination than standard database. The methods, PCA, LDA and KNN, are compared in terms of recognition accuracy and processing time.

Host: Todd Kuffner