3D Face Recognition with Multiple Kernel Learning

A novel 3D face recognition framework based on Multiple Kernel Learning (MKL) is proposed in this work. As a first step, preprocessing is applied in order to extract relevant information and remove noise from 3D face scans. Next, a surface normals and Locally Adaptive Regression Kernels (LARK) features are extracted and a kernel function is associated with them. Finally, the corresponding kernel matrices are used in SimpleMKL algorithm with Support Vector Machines (SVM) classifier. The experiments using a publicly available dataset delivered promising results and lead us to propose this framework as alternative to other 3D face recognition frameworks. Key–Words: 3D Face Recognition, Surface Normals, Locally Adaptive Regression Kernels, Multiple Kernel Learning, SimpleMKL, Support Vector Machines