There have been great strides in shape analysis in this decade. Pattern recognition, image analysis, and morphometrics have been the major contributors to this area but now bioinformatics is driving the subject as well, and new challenges are emerging; also the methods of pattern recognition are evolving for bioinformatics. Shape analysis for labelled landmarks is now moving to the new challenges of unlabelled landmarks motivated by these applications. ICP, EM algorithms, etc. are well used in image analysis but now Bayesian methods are coming into the arena. Dynamic Bayesian networks are other developments. We will discuss the problem of averaging, image deformation, projective shape and Bayesian alignment. The aim of this talk will be to convince the scientists that statistical shape analysis is pivotal to the modern pattern recognition.