Public Member Functions | Public Attributes | Private Member Functions | Private Attributes | List of all members
ISM::Trainer Class Reference

#include <Trainer.hpp>

Public Member Functions

void setClusterForManualDefHeuristic (std::vector< std::pair< std::vector< ManuallyDefPseudoHeuristic::ClusterObject >, uint16_t >>)
 
void setPredefinedRefs (std::map< std::string, std::string > &refs)
 
void setSkipsPerCycle (const int skips)
 
void setUseClustering (const bool useClustering)
 
 Trainer (std::string dbfilename="record.sqlite", bool dropOldModelTables=false)
 
void trainPattern ()
 
void trainPattern (const std::string &patternName)
 

Public Attributes

double maxAngleDeviation
 
double staticBreakRatio
 
double togetherRatio
 

Private Member Functions

TrackPtr doTraining (const std::vector< ObjectSetPtr > sets, std::string patternName)
 
HeuristicPtr findHeuristicMatch (const TracksPtr &tracks)
 
void learn ()
 

Private Attributes

PointPtr absoluteReferencePoint
 
std::vector< std::pair< std::vector< ManuallyDefPseudoHeuristic::ClusterObject >, uint16_t > > mClusterForManualDefHeuristic
 
std::map< std::string, std::string > mPatternToTypesOfPredefinedRefs
 
bool mUseManualDefHeuristic
 
bool mUsePredefinedRefs
 
RecordedPatternPtr recordedPattern
 
int skips
 
TableHelperPtr tableHelper
 
bool useClustering
 

Detailed Description

Trainer class. Learns scene models from training data. See Meissner et al. 2013 in Section IV and V (A).

Definition at line 36 of file Trainer.hpp.

Constructor & Destructor Documentation

ISM::Trainer::Trainer ( std::string  dbfilename = "record.sqlite",
bool  dropOldModelTables = false 
)

Create training interface to an sqlite db.

Parameters
dbfilenameDb from which training data is taken and into which scene models are written.

Definition at line 37 of file Trainer.cpp.

Member Function Documentation

TrackPtr ISM::Trainer::doTraining ( const std::vector< ObjectSetPtr sets,
std::string  patternName 
)
private

Definition at line 163 of file Trainer.cpp.

HeuristicPtr ISM::Trainer::findHeuristicMatch ( const TracksPtr tracks)
private

Definition at line 136 of file Trainer.cpp.

void ISM::Trainer::learn ( )
private

Definition at line 79 of file Trainer.cpp.

void ISM::Trainer::setClusterForManualDefHeuristic ( std::vector< std::pair< std::vector< ManuallyDefPseudoHeuristic::ClusterObject >, uint16_t >>  cluster)

Definition at line 125 of file Trainer.cpp.

void ISM::Trainer::setPredefinedRefs ( std::map< std::string, std::string > &  refs)

Definition at line 131 of file Trainer.cpp.

void ISM::Trainer::setSkipsPerCycle ( const int  skips)

Definition at line 52 of file Trainer.cpp.

void ISM::Trainer::setUseClustering ( const bool  useClustering)

Whether to learn one ism on training data or rather a tree of isms.

Parameters
useClusteringDecides whether to subdivide scene elements into clusters based on heuristics or the leave them all in one set before learning of isms.

Definition at line 56 of file Trainer.cpp.

void ISM::Trainer::trainPattern ( )

Perform scene model learning on data from sqlite db loaded beforehand.

Definition at line 60 of file Trainer.cpp.

void ISM::Trainer::trainPattern ( const std::string &  patternName)

Definition at line 68 of file Trainer.cpp.

Member Data Documentation

PointPtr ISM::Trainer::absoluteReferencePoint
private

Definition at line 39 of file Trainer.hpp.

double ISM::Trainer::maxAngleDeviation

Definition at line 51 of file Trainer.hpp.

std::vector<std::pair<std::vector<ManuallyDefPseudoHeuristic::ClusterObject>, uint16_t> > ISM::Trainer::mClusterForManualDefHeuristic
private

Definition at line 45 of file Trainer.hpp.

std::map<std::string, std::string> ISM::Trainer::mPatternToTypesOfPredefinedRefs
private

Definition at line 46 of file Trainer.hpp.

bool ISM::Trainer::mUseManualDefHeuristic
private

Definition at line 42 of file Trainer.hpp.

bool ISM::Trainer::mUsePredefinedRefs
private

Definition at line 43 of file Trainer.hpp.

RecordedPatternPtr ISM::Trainer::recordedPattern
private

Definition at line 38 of file Trainer.hpp.

int ISM::Trainer::skips
private

Definition at line 40 of file Trainer.hpp.

double ISM::Trainer::staticBreakRatio

Definition at line 51 of file Trainer.hpp.

TableHelperPtr ISM::Trainer::tableHelper
private

Definition at line 37 of file Trainer.hpp.

double ISM::Trainer::togetherRatio

Definition at line 51 of file Trainer.hpp.

bool ISM::Trainer::useClustering
private

Definition at line 41 of file Trainer.hpp.


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


asr_lib_ism
Author(s): Hanselmann Fabian, Heller Florian, Heizmann Heinrich, Kübler Marcel, Mehlhaus Jonas, Meißner Pascal, Qattan Mohamad, Reckling Reno, Stroh Daniel
autogenerated on Wed Jan 8 2020 04:02:41