Extracting face bump maps from video

Accurate, realistic and efficient rendering of human faces is of great importance in graphics applications. For high-speed rendering, a low-fidelity 3D model may be effectively enhanced by the use of bump mapping techniques which superimpose high-resolution surface normal maps on low-resolution geometry. Traditionally, obtaining high-quality bump maps for human faces has required highresolution 3D models, or complex studio-bound lighting rigs. In this sketch we describe progress towards the extraction of highquality bump maps from a single video camera and a static light source.

[1]  Andrew W. Fitzgibbon,et al.  Flexible bump map capture from video , 2002, Eurographics.

[2]  Andrew W. Fitzgibbon,et al.  3D head tracking using non-linear optimization , 2003, BMVC.

[3]  Paul E. Debevec,et al.  Acquiring the reflectance field of a human face , 2000, SIGGRAPH.

[4]  Luc Van Gool,et al.  Wide-baseline multiple-view correspondences , 2003, 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings..

[5]  Gabriel Taubin,et al.  Appying Shape from Lighting Variation to Bump Map Capture , 1997, Rendering Techniques.