Automated Face Recognition

Abstract

Machine based face recognition can never surpass human facial recognition system due to obvious reasons. There is, and will remain a room for improvement in terms of efficiency or optimal recognition results. One of the major dilemmas of facial recognition is its computational complexity. The problem is aggravated further for increased data base even using state of the art machines.In this project Discrete Wavelet Transform will be applied prior to Automated Face Recognition (AFR) algorithm that will reduce its computational complexity by a factor of four with negligible degradation in results. The family of Daubechies will be investigated and good bases in terms of efficiency with in the family of Daubechies will be proposed for AFR.