The accuracy of a 3D face recognition system depends on a correct registration that aligns the facial surfaces and makes a comparison possible. The best results obtained so far use a costly one-to-all registration approach, which requires the registration of each facial surface to all faces in the gallery. We explore the approach of registering the new facial surface to an average face model (A...

Average faces are used in multiple areas including biometric and social studies as well as facial recognition software. In the past average faces were computed in 2D images but these are limited if a 3D representation is needed. 3D representations usually were not textured, therefore we decided to create a fully colored 3D Mesh. We divided the averaging process into averaging mesh and averaging...

Face recognition is still an active pattern analysis topic. Faces have already been treated as objects or textures, but human face recognition system takes a different approach in face recognition. People refer to faces by their most discriminant features. People usually describe faces in sentences like ``She's snub-nosed'' or ``he's got long nose'' or ``he's got round eyes'' and so like. These...

The past few years have seen a marked increase in our understanding of how faces are represented in the brain, with the discovery of new anatomical structures and new algorithms for representing faces. Still, the basic computational mechanism used by the primate visual system to identify faces remains a topic of intense debate. Are faces represented by matching to a set of stored exemplars, or ...

The convolutional neural networks (CNN), including AlexNet, GoogleNet, VGGNet, etc. extract features for many computer vision problems which are very discriminative. trained CNN model over one dataset performs reasonably well whereas on another of similar type the hand-designed feature descriptor outperforms same model. Rectified Linear Unit (ReLU) layer discards some values in order to introdu...

The convolutional neural networks (CNN) like AlexNet, GoogleNet, VGGNet, etc. have been proven as the very discriminative feature descriptor for many computer vision problems. The trained CNN model over one dataset performs reasonably well over another dataset of similar type and outperforms the hand-designed feature descriptor. The Rectified Linear Unit (ReLU) layer discards some information i...