Active Shape Models

Active Shape Models are statistical models of the shapes of objects which iteratively deform to fit to an example of the object in a new image. The shapes are constrained by a Statistical Shape Model to vary only in ways seen in a training set of labelled examples.

In addition to the shape model, we require models of the image appearance around each model point. The simplest is to assume the points lie on strong edges. More complex models can be built to represent the statistical variation of the gradient along profiles through the points, normal to the boundary curve at that point.

We assume we have an initial estimate for the pose and shape parameters (eg the mean shape). This is iteratively updated as follows:

The performance can be significantly improved using a multi-resolution implementation, in which we start searching on a coarse level of a gaussian image pyramid, and progressively refine. This leads to much faster, more accurate and more robust search.

Face Example

Example of a multi-resolution search (displayed at highest resolution):

Initial Pose After 5 iterations Convergence
This takes less than one second on a Sun Sparc20.

A detailed report about Active Shape Models and Active Appearance Models Postscript (>1.6Mb)

You can now download a set of tools to build and play with Appearance Models and AAMs here. Enjoy.

Tim Cootes