Section: New Results

Progressive Shape Models

In this work we address the problem of recovering both the topology and the geometry of a deformable shape using temporal mesh sequences (Figure 4 ). The interest arises in multi-camera applications when unknown natural dynamic scenes are captured. While several approaches allow recovery of shape models from static scenes, few consider dynamic scenes with evolving topology and without prior knowledge. In this nonetheless generic situation, a single time observation is not necessarily enough to infer the correct topology of the observed shape and evidences must be accumulated over time in order to learn this topology and to enable temporally consistent modelling. This appears to be a new problem for which no formal solution exists. We have proposed a principled approach based on the assumption that the observed objects have a fixed topology. Under this assumption, the topology can be progressively learned during the capture of a dynamic scene evolutions. The approach has been successfully experimented on several standard 4D datasets and we believe that it paves the way to more general multi-view scene capture and analysis[8] .

Figure 4. Progressive Shape Models : the balloon can be separated from humans