|
Tim Weyrich, Bernd Bickel, Markus Gross
In the past, computer graphics showed a large improvement in rendering quality. However, rendering
human faces realistically still remains a demanding task. While recent feature films already show
authentic artificial renderings of character faces, at present, reproducing the visual appearance
of a real person's face is not solved in a satisfactory manner.
One of the larger challenges is the realistic modeling of human skin. Skin shows complex reflection
properties, varying across different subjects, genders, races, age classes, etc. In order to capture
skin with its full spectrum of reflection properties, we developed a new kind of facial scanner.
The scanner is able to acquire geometry and texture data, as well as information on the BRDF
(Bi-directional Radiance Distribution Function) for every single point in the face. In addition,
we developed a device to measure subsurface scattering properties of human skin.
The scanning facility has been built in collaboration with the Mitsubishi Electric Research
Laboratories (MERL). In the context of this collaboration, we are continuously scanning subjects in
order to collect a representative skin reflectance data base.
Our project deals with the careful analysis of the acquired reflectance data. An individual spatially
varying BRDF is fitted to each subject's scan, and machine learning techniques are engaged to derive
systematic appearance variations across subjects. Ultimately, this leads to a reflectance model for human
skin, that allows for realistic renderings of human faces and provides intuitive trait parameters to change
the subject's appearance along the variational directions between subjects derived from our data base.
In an additional study of appearance modeling, we investigated in the potential of appearance preserving
deformations of image-based reflectance representations. We developed a method to render Surface Reflectance
Fields under arbitrary deformations while preserving the spatially varying BRDF of the acquired objects.
- H.-J. Kim, B. Bickel, M. Gross, S.-M. Choi, Subsurface Scattering in Point-Based Rendering, Poster Proceedings of the Pacific Conference on Computer Graphics and Applications (Jeju, Korea, October 7-9, 2009), pp. 63-66
[Abstract]
[PDF]
- T. Weyrich, W. Matusik, H. Pfister, B. Bickel, C. Donner, C. Tu, J. McAndless, J. Lee, A. Ngan, W. Jensen, M. Gross, Analysis of Human Faces using a Measurement-Based Skin Reflectance Model, Proceedings of ACM SIGGRAPH (Boston, USA, 30 July - 3 August, 2006), ACM Transactions on Graphics, vol. 25, no. 3, pp. 1013-1024
[Abstract]
[PDF] [Video]
- B. Bickel, T. Weyrich, W. Matusik, H. Pfister, C. Donner, C. Tu, J. McAndless, J. Lee, A. Ngan, H. W. Jensen, M. Gross, Processing and Editing of Faces using a Measurement-Based Skin Reflectance Model, Proceedings of the ACM SIGGRAPH Sketches (Boston, USA, July 30 - August 3, 2006), pp. 168
[Abstract]
[PDF] [Video] [Video] [Video] [Video]
- T. Weyrich, H. Pfister, M. Gross, Rendering Deformable Surface Reflectance Fields, IEEE Transactions on Computer Graphics and Visualization (11), pp. 48-58, 2005
[Abstract]
[PDF]
|
inf.ethz.ch