Click or drag to resize

FacemarkLBFParams Properties

The FacemarkLBFParams type exposes the following members.

Public propertyBaggingOverlap
overlap ratio for training the LBF feature
Public propertyCascadeFace
filename of the face detector model
Public propertyInitShapeN
multiplier for augment the training data
Public propertyModelFile
filename where the trained model will be saved
Public propertyNLandmarks
number of landmark points
Public propertyPtr
Pointer to the unmanaged object
(Inherited from UnmanagedObject.)
Public propertySaveModel
flag to save the trained model or not
Public propertyShapeOffset
offset for the loaded face landmark points
Public propertyStagesN
number of refinement stages
Public propertyTreeDepth
the depth of decision tree, defines the size of feature
Public propertyTreeN
number of tree in the model for each landmark point refinement
Public propertyVerbose
show the training print-out
See Also