Orthogonal Laplacianfaces for 3D face recognition

In this paper, a novel framework for 3D face recognition based on depth information, is proposed. The core of our framework is Orthogonal Laplacianfaces (OLPP), a method for utilizing a Euclid space into which data points are mapped. In order to overcome facial expression variation, we first utilize curvature information projected onto the moving least-squares (MLS) surface to segment a face rigid area, which is insensitive to expression variation. Data noise is caused by distorted meshes and misalignment. To solve these problems, we introduce an OLPP strategy, which can make good use of the robustness and efficiency of accurate local statistical information. The OLPP method produces orthogonal basis functions and can have more locality preserving power than popular methods. Since the locality preserving power is potentially related to the discriminating power, the OLPP is expected to have more discriminating power. Our experiments are based on CASIA 3D face database which contain large pose and expression variations. Experimental results show OLPP performs better than many commonly used appearance-based methods.