Statismo  0.10.1
 All Classes Namespaces Functions Typedefs
Public Types | Public Member Functions | Static Public Member Functions | List of all members
statismo::LowRankGPModelBuilder< T > Class Template Reference

#include <LowRankGPModelBuilder.h>

Inheritance diagram for statismo::LowRankGPModelBuilder< T >:
Inheritance graph

Public Types

typedef Representer< T > RepresenterType
 
typedef RepresenterType::PointType PointType
 
typedef ModelBuilder< T > Superclass
 
typedef
Superclass::StatisticalModelType 
StatisticalModelType
 
typedef Domain< PointType > DomainType
 
typedef
DomainType::DomainPointsListType 
DomainPointsListType
 
typedef MatrixValuedKernel
< PointType > 
MatrixValuedKernelType
 
- Public Types inherited from statismo::ModelBuilder< T >
typedef Representer< T > RepresenterType
 
typedef StatisticalModel< T > StatisticalModelType
 
typedef DataManager< T > DataManagerType
 
typedef
DataManagerType::DataItemListType 
DataItemListType
 

Public Member Functions

void Delete ()
 
virtual ~LowRankGPModelBuilder ()
 
StatisticalModelTypeBuildNewZeroMeanModel (const MatrixValuedKernelType &kernel, unsigned numComponents, unsigned numPointsForNystrom=500) const
 
StatisticalModelTypeBuildNewModel (typename RepresenterType::DatasetConstPointerType mean, const MatrixValuedKernelType &kernel, unsigned numComponents, unsigned numPointsForNystrom=500) const
 

Static Public Member Functions

static LowRankGPModelBuilderCreate (const RepresenterType *representer)
 

Additional Inherited Members

- Static Public Attributes inherited from statismo::ModelBuilder< T >
static const double TOLERANCE = 1e-5
 
- Protected Member Functions inherited from statismo::ModelBuilder< T >
MatrixType ComputeScores (const MatrixType &X, const StatisticalModelType *model) const
 
ModelInfo CollectModelInfo () const
 

Detailed Description

template<typename T>
class statismo::LowRankGPModelBuilder< T >

A model builder for building statistical models that are specified by an arbitrary Gaussian Process. For details on the theoretical basis for this type of model builder, see the paper

A unified approach to shape model fitting and non-rigid registration Marcel Lüthi, Christoph Jud and Thomas Vetter IN: Proceedings of the 4th International Workshop on Machine Learning in Medical Imaging, LNCS 8184, pp.66-73 Nagoya, Japan, September 2013

Constructor & Destructor Documentation

template<typename T >
virtual statismo::LowRankGPModelBuilder< T >::~LowRankGPModelBuilder ( )
inlinevirtual

The desctructor

Member Function Documentation

template<typename T >
StatisticalModelType* statismo::LowRankGPModelBuilder< T >::BuildNewModel ( typename RepresenterType::DatasetConstPointerType  mean,
const MatrixValuedKernelType kernel,
unsigned  numComponents,
unsigned  numPointsForNystrom = 500 
) const
inline

Build a new model using a Gaussian process with given mean and kernel.

Parameters
meanA dataset that represents the mean (shape or deformation)
kernelA kernel (or covariance) function
numComponentsThe number of components used for the low rank approximation.
numPointsForNystromThe number of points used for the Nystrom approximation
Returns
a new statistical model representing the given Gaussian process
template<typename T >
StatisticalModelType* statismo::LowRankGPModelBuilder< T >::BuildNewZeroMeanModel ( const MatrixValuedKernelType kernel,
unsigned  numComponents,
unsigned  numPointsForNystrom = 500 
) const
inline

Build a new model using a zero-mean Gaussian process with given kernel.

Parameters
kernelA kernel (or covariance) function
numComponentsThe number of components used for the low rank approximation.
numPointsForNystromThe number of points used for the Nystrom approximation
Returns
a new statistical model representing the given Gaussian process
template<typename T >
static LowRankGPModelBuilder* statismo::LowRankGPModelBuilder< T >::Create ( const RepresenterType *  representer)
inlinestatic

Factory method to create a new ModelBuilder

template<typename T >
void statismo::LowRankGPModelBuilder< T >::Delete ( )
inline

Destroy the object. The same effect can be achieved by deleting the object in the usual way using the c++ delete keyword.


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