Date Added: Mar 2010
In this paper, the authors present a new framework for Non-Rigid Structure From Motion (NRSFM) that simultaneously addresses three significant challenges: severe occlusion, perspective camera projection, and large non-linear deformation. They introduce a concept called a model graph, which greatly reduces the computational cost of discovering groups of input images that depict consistent 3D shapes. A 3D model is constructed for each input image by traversing the model graph along multiple evolutionary paths. A compressive shape representation is constructed, which consolidates the multiple 3D models for each image reconstructed during model evolution and reduces the number of models needed to represent the input image set.