Exploring Open Set Classification for Facial Recognition Applications

Abstract

In this work we have proposed a system to perform open set classification for facial recognition applications and using incremental learning furthermore we have studied and explored all the different algorithms that enable us to perform that task. The algorithms included different computer vision and machine learning algorithms, where each played an important role in the final results. Different classifiers were tuned and tested using k fold cross validation techniques for different scenarios on the choke point dataset in order to find the classification algorithm that performs best for our task.