PDF version of this entire document

next up previous
Next: Validation Up: A Generic Method for Previous: Measuring Distances Between Images

Experimental Evaluation

We demonstrate the proposed approach to model evaluation in two stages. Firstly, a set of validation experiments are performed in which the behaviour of specificity and generalisation ability are observed for a deliberate and controlled degradation of a set of appearance models. The approach is then applied to the problem of choosing an optimal non-rigid registration algorithm for automatic construction of appearance models.


Roy Schestowitz 2007-03-11