X-Boost  2.3.8
Public Types | Public Member Functions | Static Public Member Functions | Public Attributes | Static Public Attributes | List of all members
BoostableClassifier< WeakClassifier > Struct Template Reference

#include <BoostableClassifier.h>

Inherits WeakClassifier.

Inherited by SoftCascadeStage< WeakClassifier >.

Public Types

typedef WeakClassifier WeakClassifierType
 

Public Member Functions

 BoostableClassifier (const WeakClassifier &weak, float a)
 
template<class R , class P1 , class P2 >
 BoostableClassifier (const R &src, P1 p1, P2 p2)
 build and convert costructor with 2 parameters
 
template<class R , class P0 , class P1 , class P2 >
 BoostableClassifier (const R &src, P0 p0, P1 p1, P2 p2)
 build and convert costructor with 3 parameters
 
template<class DataType >
float operator() (const DataType data) const
 
template<class DataType , class TParam1 >
float operator() (const DataType data, TParam1 stride) const
 
float getAlpha () const
 

Static Public Member Functions

static std::string signature ()
 

Public Attributes

float alpha
 Weight associated to this classifier.
 

Static Public Attributes

static const ClassifierType Type = RealClassifier
 BoostableClassifier convert a DiscreteClassifier in a RealClassifier.
 

Detailed Description

template<class WeakClassifier>
struct BoostableClassifier< WeakClassifier >

Expand a binary weak classifier in order to become an ensemble learner. An ensemple learner with a weight associated to classifier. WeakClassifier must implement a method operator() that returns {-1|+1}

Note
WeakClassifier must be a Discrete Weak Classifier, Single or MultiClass TODO: change name in BoostedClassifier

Member Function Documentation

template<class WeakClassifier>
float BoostableClassifier< WeakClassifier >::getAlpha ( ) const
inline

return the maxima response (in absolute value) of this classifier.

Returns
the inner alpha (used only for debug purpose usually)
template<class WeakClassifier >
template<class DataType >
float BoostableClassifier< WeakClassifier >::operator() ( const DataType  data) const
inline

evaluate integral image and return alpha/-alpha according to binary class

template<class WeakClassifier >
template<class DataType , class TParam1 >
float BoostableClassifier< WeakClassifier >::operator() ( const DataType  data,
TParam1  stride 
) const
inline

evaluate integral image and return alpha/-alpha according to binary class


The documentation for this struct was generated from the following file: