Face Recognition Using Neural Network Based Fourier Gabor Filters & Random Projection

Face detection and recognition has many applications in a variety of fields such as authentication,security, video surveillance and human interaction systems. In this paper, we present a neuralnetwork system for face recognition. Feature vector based on Fourier Gabor filters is used asinput of our classifier, which is a Back Propagation Neural Network (BPNN). The input vector ofthe network will have large dimension, to reduce its feature subspace we investigate the use ofthe Random Projection as method of dimensionality reduction. Theory and experiment indicatesthe robustness of our solution.