This paper presents an algorithm for generating statistical shape models by addressing it as a correspondence problem of finding the parameterization of each shape in the training set, instead of ...