The research covers the entire image acquisition and processing pipeline from capture, analysis to efficient synthesis and display of photorealistic 3D models including reflection properties.  

Another focus is set on the semantic analysis of images and videos in order to transform natural language into images, animations and 3D models.

The applied tools combine techniques from computer graphics and computer vision: Monte Carlo simulation for global illumination, hardware accelerated rendering and image synthesis, computational photography, and general, massively parallel algorithms on GPUs.