Learning pose variations within shape population by constrained mixtures of factor analyzers
This work addresses pose variation learning in shape modeling for domains like 3D animation and image segmentation, representing an incremental improvement over existing statistical shape methods.
The paper tackled the problem of learning pose variations within shape populations by formulating it as constrained mixtures of factor analyzers, resulting in a method that automatically learns rotations and generates smooth, realistic new poses for applications like motion animation.
Mining and learning the shape variability of underlying population has benefited the applications including parametric shape modeling, 3D animation, and image segmentation. The current statistical shape modeling method works well on learning unstructured shape variations without obvious pose changes (relative rotations of the body parts). Studying the pose variations within a shape population involves segmenting the shapes into different articulated parts and learning the transformations of the segmented parts. This paper formulates the pose learning problem as mixtures of factor analyzers. The segmentation is obtained by components posterior probabilities and the rotations in pose variations are learned by the factor loading matrices. To guarantee that the factor loading matrices are composed by rotation matrices, constraints are imposed and the corresponding closed form optimal solution is derived. Based on the proposed method, the pose variations are automatically learned from the given shape populations. The method is applied in motion animation where new poses are generated by interpolating the existing poses in the training set. The obtained results are smooth and realistic.