Expression invariant 3D face recognition based on GMDS
Yuehui Sun · 2015
In this paper, we propose an efficient expression-invariant 3D face recognition algorithm to compute the minimum-distortion mapping between two 3D face by the Generalized MultiDimensional Scaling (GMDS). Both full and partial parts matching are computed for finding the least distortion embedding of one 3D face into another during GMDS. The problem of expression-invariant three-dimensional face recognition turns to be isometry-invariant matching of surfaces by the means of GMDS. The FRGC v2.0 dataset is conducted for covering face recognition under expression changes. The experimental results demonstrate that the proposed method provides a very encouraging new solution for 3D face recognition.