GestureRecognitionToolkit  Version: 1.0 Revision: 04-03-15
The Gesture Recognition Toolkit (GRT) is a cross-platform, open-source, c++ machine learning library for real-time gesture recognition.
GRT::FFTFeatures Class Reference
Inheritance diagram for GRT::FFTFeatures:
GRT::FeatureExtraction GRT::MLBase GRT::GRTBase GRT::Observer< TrainingResult > GRT::Observer< TestInstanceResult >

Public Member Functions

 FFTFeatures (UINT fftWindowSize=512, UINT numChannelsInFFTSignal=1, bool computeMaxFreqFeature=true, bool computeMaxFreqSpectrumRatio=true, bool computeCentroidFeature=true, bool computeTopNFreqFeatures=true, UINT N=10)
 
 FFTFeatures (const FFTFeatures &rhs)
 
virtual ~FFTFeatures (void)
 
FFTFeaturesoperator= (const FFTFeatures &rhs)
 
virtual bool deepCopyFrom (const FeatureExtraction *featureExtraction)
 
virtual bool computeFeatures (const VectorDouble &inputVector)
 
virtual bool reset ()
 
virtual bool saveModelToFile (string filename) const
 
virtual bool loadModelFromFile (string filename)
 
virtual bool saveModelToFile (fstream &file) const
 
virtual bool loadModelFromFile (fstream &file)
 
bool init (UINT fftWindowSize, UINT numChannelsInFFTSignal, bool computeMaxFreqFeature, bool computeMaxFreqSpectrumRatio, bool computeCentroidFeature, bool computeTopNFreqFeatures, UINT N)
 
- Public Member Functions inherited from GRT::FeatureExtraction
 FeatureExtraction ()
 
virtual ~FeatureExtraction ()
 
bool copyBaseVariables (const FeatureExtraction *featureExtractionModule)
 
virtual bool clear ()
 
string getFeatureExtractionType () const
 
UINT getNumInputDimensions () const
 
UINT getNumOutputDimensions () const
 
bool getInitialized () const
 
bool getFeatureDataReady () const
 
VectorDouble getFeatureVector () const
 
FeatureExtractioncreateNewInstance () const
 
- Public Member Functions inherited from GRT::MLBase
 MLBase (void)
 
virtual ~MLBase (void)
 
bool copyMLBaseVariables (const MLBase *mlBase)
 
virtual bool train (ClassificationData trainingData)
 
virtual bool train_ (ClassificationData &trainingData)
 
virtual bool train (RegressionData trainingData)
 
virtual bool train_ (RegressionData &trainingData)
 
virtual bool train (TimeSeriesClassificationData trainingData)
 
virtual bool train_ (TimeSeriesClassificationData &trainingData)
 
virtual bool train (TimeSeriesClassificationDataStream trainingData)
 
virtual bool train_ (TimeSeriesClassificationDataStream &trainingData)
 
virtual bool train (UnlabelledData trainingData)
 
virtual bool train_ (UnlabelledData &trainingData)
 
virtual bool train (MatrixDouble data)
 
virtual bool train_ (MatrixDouble &data)
 
virtual bool predict (VectorDouble inputVector)
 
virtual bool predict_ (VectorDouble &inputVector)
 
virtual bool predict (MatrixDouble inputMatrix)
 
virtual bool predict_ (MatrixDouble &inputMatrix)
 
virtual bool map (VectorDouble inputVector)
 
virtual bool map_ (VectorDouble &inputVector)
 
virtual bool print () const
 
virtual bool save (const string filename) const
 
virtual bool load (const string filename)
 
virtual bool getModel (ostream &stream) const
 
double scale (const double &x, const double &minSource, const double &maxSource, const double &minTarget, const double &maxTarget, const bool constrain=false)
 
virtual string getModelAsString () const
 
UINT getBaseType () const
 
UINT getNumInputFeatures () const
 
UINT getNumInputDimensions () const
 
UINT getNumOutputDimensions () const
 
UINT getMinNumEpochs () const
 
UINT getMaxNumEpochs () const
 
UINT getValidationSetSize () const
 
UINT getNumTrainingIterationsToConverge () const
 
double getMinChange () const
 
double getLearningRate () const
 
double getRootMeanSquaredTrainingError () const
 
double getTotalSquaredTrainingError () const
 
bool getUseValidationSet () const
 
bool getRandomiseTrainingOrder () const
 
bool getTrained () const
 
bool getModelTrained () const
 
bool getScalingEnabled () const
 
bool getIsBaseTypeClassifier () const
 
bool getIsBaseTypeRegressifier () const
 
bool getIsBaseTypeClusterer () const
 
bool enableScaling (bool useScaling)
 
bool setMaxNumEpochs (const UINT maxNumEpochs)
 
bool setMinNumEpochs (const UINT minNumEpochs)
 
bool setMinChange (const double minChange)
 
bool setLearningRate (double learningRate)
 
bool setUseValidationSet (const bool useValidationSet)
 
bool setValidationSetSize (const UINT validationSetSize)
 
bool setRandomiseTrainingOrder (const bool randomiseTrainingOrder)
 
bool registerTrainingResultsObserver (Observer< TrainingResult > &observer)
 
bool registerTestResultsObserver (Observer< TestInstanceResult > &observer)
 
bool removeTrainingResultsObserver (const Observer< TrainingResult > &observer)
 
bool removeTestResultsObserver (const Observer< TestInstanceResult > &observer)
 
bool removeAllTrainingObservers ()
 
bool removeAllTestObservers ()
 
bool notifyTrainingResultsObservers (const TrainingResult &data)
 
bool notifyTestResultsObservers (const TestInstanceResult &data)
 
MLBasegetMLBasePointer ()
 
const MLBasegetMLBasePointer () const
 
vector< TrainingResultgetTrainingResults () const
 
- Public Member Functions inherited from GRT::GRTBase
 GRTBase (void)
 
virtual ~GRTBase (void)
 
bool copyGRTBaseVariables (const GRTBase *GRTBase)
 
string getClassType () const
 
string getLastWarningMessage () const
 
string getLastErrorMessage () const
 
string getLastInfoMessage () const
 
GRTBasegetGRTBasePointer ()
 
const GRTBasegetGRTBasePointer () const
 
- Public Member Functions inherited from GRT::Observer< TrainingResult >
virtual void notify (const TrainingResult &data)
 
- Public Member Functions inherited from GRT::Observer< TestInstanceResult >
virtual void notify (const TestInstanceResult &data)
 

Protected Attributes

UINT fftWindowSize
 
UINT numChannelsInFFTSignal
 
bool computeMaxFreqFeature
 
bool computeMaxFreqSpectrumRatio
 
bool computeCentroidFeature
 
bool computeTopNFreqFeatures
 
UINT N
 
double maxFreqFeature
 
double maxFreqSpectrumRatio
 
double centroidFeature
 
VectorDouble topNFreqFeatures
 
- Protected Attributes inherited from GRT::FeatureExtraction
string featureExtractionType
 
bool initialized
 
bool featureDataReady
 
VectorDouble featureVector
 
- Protected Attributes inherited from GRT::MLBase
bool trained
 
bool useScaling
 
UINT baseType
 
UINT numInputDimensions
 
UINT numOutputDimensions
 
UINT numTrainingIterationsToConverge
 
UINT minNumEpochs
 
UINT maxNumEpochs
 
UINT validationSetSize
 
double learningRate
 
double minChange
 
double rootMeanSquaredTrainingError
 
double totalSquaredTrainingError
 
bool useValidationSet
 
bool randomiseTrainingOrder
 
Random random
 
vector< TrainingResulttrainingResults
 
TrainingResultsObserverManager trainingResultsObserverManager
 
TestResultsObserverManager testResultsObserverManager
 
- Protected Attributes inherited from GRT::GRTBase
string classType
 
DebugLog debugLog
 
ErrorLog errorLog
 
InfoLog infoLog
 
TrainingLog trainingLog
 
TestingLog testingLog
 
WarningLog warningLog
 

Static Protected Attributes

static RegisterFeatureExtractionModule< FFTFeaturesregisterModule
 

Additional Inherited Members

- Public Types inherited from GRT::FeatureExtraction
typedef std::map< string, FeatureExtraction *(*)() > StringFeatureExtractionMap
 
- Public Types inherited from GRT::MLBase
enum  BaseTypes { BASE_TYPE_NOT_SET =0, CLASSIFIER, REGRESSIFIER, CLUSTERER }
 
- Static Public Member Functions inherited from GRT::FeatureExtraction
static FeatureExtractioncreateInstanceFromString (string const &featureExtractionType)
 
- Static Public Member Functions inherited from GRT::GRTBase
static string getGRTVersion (bool returnRevision=true)
 
static string getGRTRevison ()
 
- Protected Member Functions inherited from GRT::FeatureExtraction
bool init ()
 
bool saveFeatureExtractionSettingsToFile (fstream &file) const
 
bool loadFeatureExtractionSettingsFromFile (fstream &file)
 
- Protected Member Functions inherited from GRT::MLBase
bool saveBaseSettingsToFile (fstream &file) const
 
bool loadBaseSettingsFromFile (fstream &file)
 
- Protected Member Functions inherited from GRT::GRTBase
double SQR (const double &x) const
 
- Static Protected Member Functions inherited from GRT::FeatureExtraction
static StringFeatureExtractionMapgetMap ()
 

Detailed Description

Definition at line 37 of file FFTFeatures.h.

Constructor & Destructor Documentation

GRT::FFTFeatures::FFTFeatures ( UINT  fftWindowSize = 512,
UINT  numChannelsInFFTSignal = 1,
bool  computeMaxFreqFeature = true,
bool  computeMaxFreqSpectrumRatio = true,
bool  computeCentroidFeature = true,
bool  computeTopNFreqFeatures = true,
UINT  N = 10 
)

Default Constructor, sets the default settings for the FFTFeatures module. The fftWindowSize and numChannelsInFFTSignal parameters should match the settings of the FFT module that will be used as input to this module.

Parameters
UINTfftWindowSize: the window size of the FFT that will be used as input to this module. Default value = 512
UINTnumChannelsInFFTSignal: this is the number of channels (i.e. input dimensions) to the FFT module. Default value = 1
boolcomputeMaxFreqFeature: sets if the maximum frequency feature will be included in the feature vector. Default value = true
boolcomputeMaxFreqSpectrumRatio: sets if the maximum-frequency spectrum-frequency ratio feature will be included in the feature vector. Default value = true
boolcomputeCentroidFeature: sets if the centroid frequency feature will be included in the feature vector. Default value = true
boolcomputeTopNFreqFeatures: sets if the top N frequency feature will be included in the feature vector. Default value = true
boolN: sets if size of N for the top N frequency features. Default value = 10

Definition at line 30 of file FFTFeatures.cpp.

GRT::FFTFeatures::FFTFeatures ( const FFTFeatures rhs)

Copy Constructor, copies the FFTFeatures from the rhs instance to this instance

Parameters
constFFTFeatures &rhs: another instance of the FFTFeatures class from which the data will be copied to this instance

Definition at line 44 of file FFTFeatures.cpp.

GRT::FFTFeatures::~FFTFeatures ( void  )
virtual

Default Destructor

Definition at line 57 of file FFTFeatures.cpp.

Member Function Documentation

bool GRT::FFTFeatures::computeFeatures ( const VectorDouble &  inputVector)
virtual

Sets the FeatureExtraction computeFeatures function, overwriting the base FeatureExtraction function. This function is called by the GestureRecognitionPipeline when any new input data needs to be processed (during the prediction phase for example).

Parameters
constVectorDouble &inputVector: the inputVector that should be processed. Must have the same dimensionality as the FeatureExtraction module
Returns
true if the data was processed, false otherwise

Reimplemented from GRT::FeatureExtraction.

Definition at line 269 of file FFTFeatures.cpp.

bool GRT::FFTFeatures::deepCopyFrom ( const FeatureExtraction featureExtraction)
virtual

Sets the FeatureExtraction deepCopyFrom function, overwriting the base FeatureExtraction function. This function is used to deep copy the values from the input pointer to this instance of the FeatureExtraction module. This function is called by the GestureRecognitionPipeline when the user adds a new FeatureExtraction module to the pipeline.

Parameters
FeatureExtraction*featureExtraction: a pointer to another instance of an FFTFeatures, the values of that instance will be cloned to this instance
Returns
true if the deep copy was successful, false otherwise

Reimplemented from GRT::FeatureExtraction.

Definition at line 80 of file FFTFeatures.cpp.

bool GRT::FFTFeatures::init ( UINT  fftWindowSize,
UINT  numChannelsInFFTSignal,
bool  computeMaxFreqFeature,
bool  computeMaxFreqSpectrumRatio,
bool  computeCentroidFeature,
bool  computeTopNFreqFeatures,
UINT  N 
)

Initializes the FFTFeatures. Should be called before calling the computeFFT(...) or computeFeatures(...) methods. This function is automatically called by the constructor.

Parameters
UINTfftWindowSize: the window size of the FFT that will be used as input to this module. Default value = FFT::FFT_WINDOW_SIZE_512
UINTnumChannelsInFFTSignal: the size of the FFT feature vector that will be used as input to this module. Default value = 1
boolcomputeMaxFreqFeature: sets if the maximum frequency feature will be included in the feature vector. Default value = true
boolcomputeMaxFreqSpectrumRatio: sets if the maximum-frequency spectrum-frequency ratio feature will be included in the feature vector. Default value = true
boolcomputeCentroidFeature: sets if the centroid frequency feature will be included in the feature vector. Default value = true
boolcomputeTopNFreqFeatures: sets if the top N frequency feature will be included in the feature vector. Default value = true
boolN: sets if size of N for the top N frequency features. Default value = 10
Returns
true if the FTT was initialized, false otherwise

Definition at line 231 of file FFTFeatures.cpp.

bool GRT::FFTFeatures::loadModelFromFile ( string  filename)
virtual

This saves the feature extraction settings to a file.

Parameters
fstream&file: a reference to the file to save the settings to
Returns
returns true if the settings were saved successfully, false otherwise

Reimplemented from GRT::MLBase.

Definition at line 111 of file FFTFeatures.cpp.

bool GRT::FFTFeatures::loadModelFromFile ( fstream &  file)
virtual

This loads the feature extraction settings from a file. This overrides the loadSettingsFromFile function in the FeatureExtraction base class.

Parameters
fstream&file: a reference to the file to load the settings from
Returns
returns true if the settings were loaded successfully, false otherwise

Reimplemented from GRT::FeatureExtraction.

Definition at line 154 of file FFTFeatures.cpp.

FFTFeatures & GRT::FFTFeatures::operator= ( const FFTFeatures rhs)

Sets the equals operator, copies the data from the rhs instance to this instance

Parameters
constFFTFeatures &rhs: another instance of the FFTFeatures class from which the data will be copied to this instance
Returns
a reference to this instance of FFT

Definition at line 61 of file FFTFeatures.cpp.

bool GRT::FFTFeatures::reset ( )
virtual

Sets the FeatureExtraction reset function, overwriting the base FeatureExtraction function. This function is called by the GestureRecognitionPipeline when the pipelines main reset() function is called. This function resets the FFTFeatures by re-initiliazing the instance.

Returns
true if the FFTFeatures was reset, false otherwise

Reimplemented from GRT::FeatureExtraction.

Definition at line 346 of file FFTFeatures.cpp.

bool GRT::FFTFeatures::saveModelToFile ( string  filename) const
virtual

This saves the feature extraction settings to a file.

Parameters
conststring filename: the filename to save the settings to
Returns
returns true if the settings were saved successfully, false otherwise

Reimplemented from GRT::MLBase.

Definition at line 97 of file FFTFeatures.cpp.

bool GRT::FFTFeatures::saveModelToFile ( fstream &  file) const
virtual

This saves the feature extraction settings to a file. This overrides the saveSettingsToFile function in the FeatureExtraction base class.

Parameters
fstream&file: a reference to the file to save the settings to
Returns
returns true if the settings were saved successfully, false otherwise

Reimplemented from GRT::FeatureExtraction.

Definition at line 126 of file FFTFeatures.cpp.


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