Face Recognition Using Binary Locality Preserving Projections

This paper presents a technique for face recognition based on Binary Locality Preserving Projections. Locality preserving projections is a method that preserves the local structure of an image, and it use nearest neighbor search. But nearest neighbor search decreases its capability in terms of time. In the proposed approach the Laplacian is used but prior knowledge of each face to its class is used, which removes the Nearest Neighbor search, removal of this step enhances its capability in terms of time. So it is a combination of Laplacian and LDA in which LPP tries to preserve the local structure and LDA tries to match the face using Euclidean Distance. Locality Preserving Projections (LPP) is non-orthogonal, and this makes it difficult to reconstruct the data. In BLPP orthogonal basis functions are used which helps in better reconstruction of data.