Preliminary Study of Dimension Reduction Methods in Facial Recognition
Wei Chen · Journal of Gansu Lianhe University · 2011
In this paper,a manifold learning algorithm for classification problems is proposed.Experiments based on selected face database showed that primary linear algorithms and nonlinear manifold learning algorithms are inadequate for classification on nonlinear data set.An approach is proposed to recognize face with a kind of idea of Supervised Local Linear Embedding.Experiments results show the efficiency of this algorithm in facial recognition.