LearnParameters Properties

Contrast trigger is the minimum gray value difference the two regions of one pair feature must have to be eligible to become a classifier feature. Valid contrast triggers must be greater than 0, the default value is 8.

Indifference radius is the radius (L1 norm!) around a positive instance from which no counter sample is to be extracted. Valid indifference radii must be greater than 0. The default value is 6 here.

This value defines how many features must be extracted at the very least per class in a classifier.

Negatives density is the density at which counter samples are extracted from the training set images during the learning phase. Valid negatives densities range from 0.0 to 1.0. The default value is 1.0.