Improving Facial Depth Data by Exemplar-based Comparisons

3D face models are widely used for several purposes, such as biometric systems, face verification, facial expression recognition, 3D visualization, etc. They can be captured by using different types of devices, like plenop- tic cameras, structured light cameras, time of flight, etc. Nevertheless, the model generated by all of these consumers devices are very noisy. In this work, we present a filtering method for meshes of faces preserving their intrinsic features. It is based in an exemplar-based neighborhood matching where all models are in a frontal position avoiding rotation and perspective drawbacks. Moreover, the model is invariant to depth transla- tion and scale. The obtained results showed that this method is robust and promising.