Dynamic body transformation and matching from scanned data

Nadia Magnenat-Thalmann

Hyewon Seo


Appropriate combinations of statistical and geometric methods permit modeling and estimation of complex objects. This talk investigates such combinations in the modeling of human body models in computer graphics. By concentrating on such a particular class of objects, we show how example based techniques together with interpolation ones can exploit existing data and knowledge about body shape and size modelling. Across the human population, the bodies of individuals exhibit a great deal of variation in their appearance, but they still have a good deal of structure in common. The work presented here takes advantage of this commonality - information concerning the appearance of the bodies is used either by a   numeric form or directly fed into the model. This results in more successful and robust systems. Aside from being a good testbed for our example-based techniques, human bodies are interesting on their own, playing an important role in a wide range of applications. Graphical modeling of human bodies obviously has applications in the entertainment industry for character animation and simulations involving people. Other applications range from virtual reality to ergonomic and manufacturing studies.