Fast generation of realistic virtual humans

Abstract

Figure 1: We generate realistic virtual humans from real persons through multi-view stereo scanning. The resulting characters are ready to be animated through a skeletal rig and facial blendshapes, and are compatible with standard graphics and VR engines. The whole reconstruction process requires only minimal user input and takes less than ten minutes. ABSTRACTIn this paper we present a complete pipeline to create ready-toanimate virtual humans by fitting a template character to a point set obtained by scanning a real person using multi-view stereo reconstruction. Our virtual humans are built upon a holistic character model and feature a detailed skeleton, fingers, eyes, teeth, and a rich set of facial blendshapes. Furthermore, due to the careful selection of techniques and technology, our reconstructed humans are quite realistic in terms of both geometry and texture. Since we represent our models as single-layer triangle meshes and animate them through standard skeleton-based skinning and facial blendshapes, our characters can be used in standard VR engines out of the box. By optimizing for computation time and minimizing manual intervention, our reconstruction pipeline is capable of processing whole characters in less than ten minutes.

Construction of the Virtual Embodiment Questionnaire (VEQ)

“…9). These insights would be specifically interesting, considering that avatar appearance realism as well as avatar personalization [1] were argued to have a moderating to substantial impact on embodiment [3,45,54,86]. In turn, this may suggest that alterations (such as used for a proteus effect) would act as antipole.…”

The Virtual Caliper: Rapid Creation of Metrically Accurate Avatars from 3D Measurements

“…Several approaches describe a pipeline to rapidly produce an avatar from 3D body and face scan data [8,60]. These approaches include facial and many features of the avatar and have proven to be very useful, as they allow for instance the study of the impact of the avatar personalization on virtual body ownership [65].…”

A method for automatic forensic facial reconstruction based on dense statistics of soft tissue thickness

“…Here, A e is the area associated to edge e, and R e are per-edge rotations to best-fit deformed and undeformed Laplacians (see [20] for details). In the spirit of non-rigid ICP [19] we alternatingly compute correspondences and minimize (3), starting with a rather stiff surface that is subsequently softened (by reducing λ reg ) to allow for more and more accurate fits. Whenever λ reg is decreased, we also update the rest state H prev by the current deformed state H. From the 39 fits to the complete optical surface scans we construct a preliminary parametric head model.…”