Hallucinating 3D faces

Super resolution technique could produce a higher resolution image than the originally captured ones. However, a few works have been done in 3D models. In this paper, we focus on generating the 3D face model of higher resolution from one input of 3D face model. In our method, the 3D face models including the training samples and input model are remeshed to construct consistent mesh to find out the correspondence among the models. The super resolution then is performed in the remeshed models to reconstruct the high resolution model. The experiments using USF HumanID 3D face database of 100 3D face models are carried out, and demonstrate the presented algorithm is promising.