In this thesis, we make an experiment on and analyze face recognition models using unsupervised learning. Unsupervised learning which don`t require any supervision signal, is a preferable learning model of human visual system. In the face recognition, unsupervised learning methods of PCA and ICA are highlighted because of its high performance and similarity to brain mechanism.
To compare and analyze PCA and ICA, we implement and test unsupervised learning face recognition model. And also, we perform cognitive psychological experiment to compare computational recognition model and human recognition.