Dimensionality Reduction Method Applied for Multi-view Face Database Recognition

Abstract

Faces under varying illumination, pose and non-rigid deformation are empirically thought of as a highly nonlinear manifold in the observation space. How to discover intrinsic low-dimensional manifold is important to characterize meaningful face distributions and classify them using some classifiers. In this paper, we use the Locally Linear Embedding (LLE) algorithm to reduce the dimensionality of face image. The LLE algorithm is the fast dimensionality reduction algorithm that finds local geometry in high dimensional space, and produces a projection to low dimensional space which preserves the original geometry. So, we use the Locally Linear Embedding (LLE) algorithm to reduce the dimensionality of face image for face recognition. Both frontal head images and rotated head images are investigated. Experiments on The UMIST Face Database that is a multi-view database show that the advantages of our proposed approach.KeywordsDimensionality ReductionFace RecognitionFace ImageLocally Linear EmbeddingNonlinear Dimensionality ReductionThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.