Dimensionality Reduction - Supervised and `unsupervised Approaches for Facial Recognition

Abstract

Nowadays Authentication of a person using biometric traits is very common. Number of Biometric features is used for authenticating person like fingerprint, iris, face etc. Among all of them facial recognition are very much popular for authenticating person. Number of algorithm is present but all have certain pros and cons. As facial image consists of number of features so dimensionality Reduction is important steps in any facial recognition algorithm. The main focus of    this paper to apply principle component analysis techniques   for face recognition process. PCA gives very good results in reducing dimension. Principle Component Analysis produces eigenvectors. One combines those eigenvectors into images and then visualizes the eigenfaces.