PDF version of this document

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

Experimental Evaluation

The proposed model evaluation approach is demonstrated in two stages. In the first instance, a set of validation experiments are performed where the behaviour of the metrics is observed for a deliberate and controlled degradation of set of appearance models. The approach is then practically demonstrated on the problem of choosing an optimal non-rigid registration algorithm for automatic construction of appearance models.


Roy Schestowitz 2005-11-17