This visit was a continuation of my previous visit to MPII in July 2004, when I worked on real-time algorithms for combination of multiple depth maps for rendering a captured 3D scene. This time the goal was a direct interaction with the MPII researchers in order to work out a suitable representation of a captured time-varying 3D scene. Some aspects of this work are related to parameterization of mesh data, as treated in Task 6.7. The resulting interaction can be divided into two types. First is working out concrete technical details, for example, figuring out the correct way to interpret and use the 3D sequences which were reconstructed and kindly provided by the IRG3. The second type of interaction was a direct learning about the current research conducted at the MPII and general discussions with the researchers in order to work out new ideas.