Facial Expression Generic Elastic Models from a Single 2D Image for large Pose Face Synthesis and Recognition

In this paper, we propose a novel and efficient method to reconstruct the 3D models of a human face from a single 2D face image with variety in fa-cial expression using the Facial Expression Generic Elastic Model (FE-GEM). We employed three generic models for modeling facial expression in the Gener-ic Elastic Model (GEM) framework and a mixture of these three models by us-ing computing distance around face lips. Particularly, we tested our method on an available 2D face databases with/without facial expression images and a new synthesized large pose from gallery images and compared our synthesized re-sults with target face images by performing face recognition using the rank-one recognition rate with the smallest cosine distance. We acquired promising re-sults for handling pose and expression changes based on the proposed method compared to the GEM approach.