Under the direction of Thomas Vetter, the Gravis Research Group has earned international recognition for its pioneering development of the Morphable Model technology for the analysis and synthesis of images of human faces since 2002. The group's approach unifies Computer Graphics and Computer Vision technology in an analysis by synthesis loop for the analysis of images, enabling a 3D interpretation of 2D images highly invariant to pose and illumination variations.
The second research area is the segmentation and modeling of human bones from CT or MRI data, in particular the statistical modeling of human bones. From example data of bones (e.g. skull or femur) statistical models are derived that reflect the variability of these structures between different humans. Awareness of this variability is useful for the construction and prediction of implants. Further, these statistical models are used as prior knowledge for the analysis and interpretation of medical images from CT and MRI.
In a third area of research, we focus on the analysis of serial block face scanning electron microscopy data. This imaging technology results in 3D volumetric images, allowing the application of different techniques from medical image analysis.