Real-time face verification system on a cell-phone using advanced correlation filters

We present a real-time face recognition system that we implemented on a cell-phone using computationally efficient minimum average correlation energy filters. We examine the challenges faced in realizing such a practical system that can perform both enrollment and verification in real-time on the cell-phone using its limited memory and computational resources. Due to the speed and memory constraints of the cellphone, we optimize the correlation filters using a fast fixed-point implementation of 2D Fourier transform optimal for the cell-phone platform. We also detail an efficient algorithm for synthesizing these filters efficiently in order to minimize computation load and making the system practical with real-time enrollment of users.