eVision::EasyDeepLearning::ELocator Class Reference

EasyLocate tool - axis aligned bounding box version. More...

#include <eVision/EasyDeepLearning/EasyLocate/ELocator.h>

Inheritance diagram for eVision::EasyDeepLearning::ELocator:

Collaboration diagram for eVision::EasyDeepLearning::ELocator:

Public Member Functions
	ELocator ()
	Constructs a ELocator object.
	ELocator (const eVision::EasyDeepLearning::ELocator &other)
	Copy constructor of a ELocator object.
virtual	~ELocator () throw ()
eVision::EasyDeepLearning::ELocator &	operator= (const eVision::EasyDeepLearning::ELocator &other)
	Assignment operator.
virtual void	SerializeSettings (eVision::ESerializer *serializer)
	Serializes the settings of the locator.
virtual eVision::EasyDeepLearning::EDeepLearningToolType	GetToolType () const
	Type of the deep learning tool.
std::vector< eVision::ESize >	GenerateAnchors (const eVision::EasyDeepLearning::EDataset &dataset) const
	Generates anchors based on the objects in the dataset or a formal specification.
std::vector< eVision::ESize >	GenerateAnchors (float minDimension, float maxDimension, int numSubScales, std::vector< float > aspectRatios) const
	Generates anchors based on the objects in the dataset or a formal specification.
void	SetSameLabelMaxOverlap (float val)
	Maximum overlap (intersection over union) between two predicted objects with the same label.
float	GetSameLabelMaxOverlap () const
	Maximum overlap (intersection over union) between two predicted objects with the same label.
void	SetAbsoluteMaxOverlap (float val)
	Maximum overlap (intersection over union) between two predicted objects regardless of their label.
float	GetAbsoluteMaxOverlap () const
	Maximum overlap (intersection over union) between two predicted objects regardless of their label.
	ELocator (CppWrapperInternal::InternalConstructor)
Public Member Functions inherited from eVision::EasyDeepLearning::ELocatorBase
virtual	~ELocatorBase () throw ()
void	SetCapacity (eVision::EasyDeepLearning::ELocatorCapacity capacity)
	Capacity of the ELocatorBase.
eVision::EasyDeepLearning::ELocatorCapacity	GetCapacity () const
	Capacity of the ELocatorBase.
void	SetWidth (uint32_t width)
	Width of input images.
uint32_t	GetWidth () const
	Width of input images.
void	SetHeight (uint32_t height)
	Height of input images.
uint32_t	GetHeight () const
	Height of input images.
void	SetChannels (uint32_t channels)
	Number of channels of input images. It can either be 1 for grayscale images or 3 for color images.
uint32_t	GetChannels () const
	Number of channels of input images. It can either be 1 for grayscale images or 3 for color images.
void	SetPredictionAnchors (const std::vector< eVision::ESize > &anchors)
	Prediction anchors.
std::vector< eVision::ESize >	GetPredictionAnchors () const
	Prediction anchors.
eVision::EasyDeepLearning::ELocatorMetrics	GetTrainingMetrics (int iteration) const
	Training metrics at the given iteration.
eVision::EasyDeepLearning::ELocatorMetrics	GetValidationMetrics (int iteration) const
	Validation metrics at the given iteration.
void	Train (eVision::EasyDeepLearning::EDataset &dataset)
	Trains the ELocatorBase with the given dataset for the specified number of iterations.
void	Train (eVision::EasyDeepLearning::EDataset &dataset, const eVision::EasyDeepLearning::EDataAugmentation &dataAugmentation)
	Trains the ELocatorBase with the given dataset for the specified number of iterations.
void	Train (eVision::EasyDeepLearning::EDataset &trainingDataset, eVision::EasyDeepLearning::EDataset &validationDataset)
	Trains the ELocatorBase with the given dataset for the specified number of iterations.
void	Train (eVision::EasyDeepLearning::EDataset &trainingDataset, eVision::EasyDeepLearning::EDataset &validationDataset, const eVision::EasyDeepLearning::EDataAugmentation &dataAugmentation)
	Trains the ELocatorBase with the given dataset for the specified number of iterations.
void	Train (eVision::EasyDeepLearning::EDataset &dataset, int iterations)
	Trains the ELocatorBase with the given dataset for the specified number of iterations.
void	Train (eVision::EasyDeepLearning::EDataset &dataset, const eVision::EasyDeepLearning::EDataAugmentation &dataAugmentation, int iterations)
	Trains the ELocatorBase with the given dataset for the specified number of iterations.
void	Train (eVision::EasyDeepLearning::EDataset &trainingDataset, eVision::EasyDeepLearning::EDataset &validationDataset, int iterations)
	Trains the ELocatorBase with the given dataset for the specified number of iterations.
void	Train (eVision::EasyDeepLearning::EDataset &trainingDataset, eVision::EasyDeepLearning::EDataset &validationDataset, const eVision::EasyDeepLearning::EDataAugmentation &dataAugmentation, int iterations)
	Trains the ELocatorBase with the given dataset for the specified number of iterations.
eVision::EasyDeepLearning::ELocatorResult	Apply (const eVision::EBaseROI &img)
	Applies the tool on the given image and its mask region.
eVision::EasyDeepLearning::ELocatorResult	Apply (const eVision::EBaseROI &img, const eVision::ERegion &mask)
	Applies the tool on the given image and its mask region.
eVision::EasyDeepLearning::ELocatorResult	Apply (const eVision::EBaseROI &img, const eVision::EFrame &frame)
	Applies the tool on the given image and its mask region.
eVision::EasyDeepLearning::ELocatorResult	Apply (const eVision::EBaseROI &img, const eVision::EFrame &frame, const eVision::ERegion &mask)
	Applies the tool on the given image and its mask region.
std::vector< eVision::EasyDeepLearning::ELocatorResult >	Apply (const std::vector< eVision::EImageBW8 > &imgs)
	Applies the tool on the given image and its mask region.
std::vector< eVision::EasyDeepLearning::ELocatorResult >	Apply (const std::vector< eVision::EImageBW8 > &imgs, const std::vector< eVision::ERegion > &masks)
	Applies the tool on the given image and its mask region.
std::vector< eVision::EasyDeepLearning::ELocatorResult >	Apply (const std::vector< eVision::EImageBW16 > &imgs)
	Applies the tool on the given image and its mask region.
std::vector< eVision::EasyDeepLearning::ELocatorResult >	Apply (const std::vector< eVision::EImageBW16 > &imgs, const std::vector< eVision::ERegion > &masks)
	Applies the tool on the given image and its mask region.
std::vector< eVision::EasyDeepLearning::ELocatorResult >	Apply (const std::vector< eVision::EImageC24 > &imgs)
	Applies the tool on the given image and its mask region.
std::vector< eVision::EasyDeepLearning::ELocatorResult >	Apply (const std::vector< eVision::EImageC24 > &imgs, const std::vector< eVision::ERegion > &masks)
	Applies the tool on the given image and its mask region.
std::vector< eVision::EasyDeepLearning::ELocatorResult >	Apply (const std::vector< const eVision::EBaseROI * > &imgs)
	Applies the tool on the given image and its mask region.
std::vector< eVision::EasyDeepLearning::ELocatorResult >	Apply (const std::vector< const eVision::EBaseROI * > &imgs, const std::vector< eVision::ERegion > &masks)
	Applies the tool on the given image and its mask region.
eVision::EasyDeepLearning::ELocatorMetrics	Evaluate (const eVision::EasyDeepLearning::EDataset &dataset)
	Evaluates the dataset.
float	GetDetectionThreshold () const
	Detection threshold. The detection threshold is set automatically during training to maximize accuracy. It can also be changed after training to obtain a different true positive/false positive tradeoff.
void	SetDetectionThreshold (float threshold)
	Detection threshold. The detection threshold is set automatically during training to maximize accuracy. It can also be changed after training to obtain a different true positive/false positive tradeoff.
void	SetMaxNumberOfObjects (int val)
	Maximum number of objects in an image (default value: 100).
int	GetMaxNumberOfObjects () const
	Maximum number of objects in an image (default value: 100).
void	SetSameLabelMaxObjectProximity (float val)
	Maximum proximity between two predicted objects with the same label.
float	GetSameLabelMaxObjectProximity () const
	Maximum proximity between two predicted objects with the same label.
void	SetAbsoluteMaxObjectProximity (float val)
	Maximum proximity between two predicted objects regardless of their label.
float	GetAbsoluteMaxObjectProximity () const
	Maximum proximity between two predicted objects regardless of their label.
int	GetLocatorFeatures () const
	The features supported by the locator.
bool	HasFeature (eVision::EasyDeepLearning::ELocatorFeature feature)
	Whether the given feature is enabled.
	ELocatorBase (CppWrapperInternal::InternalConstructor)
Public Member Functions inherited from eVision::EasyDeepLearning::EDeepLearningTool
virtual	~EDeepLearningTool () throw ()
virtual void	Load (const std::string &filePath)
	Loads a deep learning tool. The given ESerializer must have been created for reading.
virtual void	Load (eVision::ESerializer *serializer)
	Loads a deep learning tool. The given ESerializer must have been created for reading.
std::unique_ptr< eVision::EasyDeepLearning::EDeepLearningTool >	Clone () const
	Clones this tool.
virtual void	Save (const std::string &filePath, bool includeTrainingModel)
	Saves the settings, inference and training model of the deep learning tool.
virtual void	Save (const std::string &filePath)
	Saves the settings, inference and training model of the deep learning tool.
virtual void	Save (eVision::ESerializer *serializer, bool includeTrainingModel)
	Saves the settings, inference and training model of the deep learning tool.
virtual void	Save (eVision::ESerializer *serializer)
	Saves the settings, inference and training model of the deep learning tool.
void	SerializeInferenceModel (eVision::ESerializer *serializer)
	Serializes the inference model of the tool.
void	SerializeTrainingModel (eVision::ESerializer *serializer)
	Serializes the training model of the tool.
void	SaveSettings (const std::string &filePath)
	Saves the settings of the tool to a file.
void	SaveInferenceModel (const std::string &filePath)
	Saves the tool's inference model to a file.
void	SaveTrainingModel (const std::string &filePath)
	Saves the tool's training model to a file.
void	LoadSettings (const std::string &filePath)
	Loads the settings of a tool.
void	LoadInferenceModel (const std::string &filePath)
	Loads the tool's inference model. The inference model is required to apply the tool on new images.
void	LoadTrainingModel (const std::string &filePath)
	Loads the tool's training model. The training model is required to continue a training.
void	UnloadInferenceModel ()
	Unloads the inference model.
void	UnloadTrainingModel ()
	Unloads the training model.
void	UnloadModels ()
	Unloads the inference and training model.
bool	HasInferenceModel () const
	Whether the inference model is loaded or created when the tool is not yet trained.
bool	HasTrainingModel () const
	Whether the training model is loaded or created when the tool is not yet trained.
bool	IsTrained () const
	Tells whether the deep learning tool has been trained.
void	SetEnableGPU (bool enable)
	Enable the use of a GPU for training and inference of the deep learning tool. If the active engine is not the default engine, EDeepLearningTool::SetEnableGPU will be false even if the active device has a GPU device type. Setting EDeepLearningTool::SetEnableGPU to any value when the active engine is not the default one will throw an exception.
bool	GetEnableGPU () const
	Enable the use of a GPU for training and inference of the deep learning tool. If the active engine is not the default engine, EDeepLearningTool::SetEnableGPU will be false even if the active device has a GPU device type. Setting EDeepLearningTool::SetEnableGPU to any value when the active engine is not the default one will throw an exception.
void	SetGPUIndexes (const std::vector< uint32_t > &indexes)
	Indexes of the GPUs to use for computations.
std::vector< uint32_t >	GetGPUIndexes () const
	Indexes of the GPUs to use for computations.
uint32_t	GetNumGPUs () const
	Number of detected NVIDIA GPU.
void	SetEngine (const std::string &engineName)
	Sets the engine with which to execute or train the neural network.
std::string	GetEngine () const
	Sets the engine with which to execute or train the neural network.
eVision::EasyDeepLearning::EDeepLearningInferencePrecision	GetInferencePrecision () const
	Inference precision.
void	SetInferencePrecision (eVision::EasyDeepLearning::EDeepLearningInferencePrecision prec)
	Inference precision.
int	GetNumAvailableDevices () const
	Available devices for the current EDeepLearningTool::SetEngine.
std::string	GetAvailableDeviceName (int i) const
	Name of the i-th device of the current engine.
eVision::EasyDeepLearning::EDeepLearningDevice	GetAvailableDevice (int i) const
	i-th EDeepLearningDevice of the current engine.
std::vector< eVision::EasyDeepLearning::EDeepLearningDevice >	GetAvailableDevices () const
	Get all available device for the current engine.
int	GetNumActiveDevices () const
	Gets the number of devices with which to run or train the model.
void	SetActiveDevice (const eVision::EasyDeepLearning::EDeepLearningDevice &device)
	Sets the device with which to run or train the model.
eVision::EasyDeepLearning::EDeepLearningDevice	GetActiveDevice (int i) const
	Gets the active device. By default, it returns the first active device.
eVision::EasyDeepLearning::EDeepLearningDevice	GetActiveDevice () const
	Gets the active device. By default, it returns the first active device.
void	SetActiveDeviceByName (const std::string &name)
	Sets the device with which to run the model using its name.
void	SetActiveDeviceById (int id)
	Sets the device with which to run the model using its index in the list returned by EDeepLearningTool::GetAvailableDevices.
std::vector< eVision::EasyDeepLearning::EDeepLearningDevice >	GetActiveDevices () const
	Devices with which to run or train the model.
void	SetActiveDevices (const std::vector< eVision::EasyDeepLearning::EDeepLearningDevice > &devices)
	Devices with which to run or train the model.
void	SetActiveDevicesById (const std::vector< int > &ids)
	Sets the devices with which to run the model using their index in the list returned by EDeepLearningTool::GetAvailableDevices.
eVision::EasyDeepLearning::EDeepLearningExecutionSettings	GetExecutionSettings () const
	Execution settings of the tool (engine, device(s), inference precision, batch size settings).
void	SetExecutionSettings (const eVision::EasyDeepLearning::EDeepLearningExecutionSettings &settings)
	Execution settings of the tool (engine, device(s), inference precision, batch size settings).
bool	HasEngineSetupData () const
	Whether the inference model has engine setup data.
void	ClearEngineSetupData ()
	Clears the engine data saved in the inference model. Use this method when your engine fails.
eVision::EasyDeepLearning::EDeepLearningDevice	GetEngineSetupDataDevice () const
	Device corresponding to the engine data contained in the tool.
eVision::EasyDeepLearning::EDeepLearningInferencePrecision	GetEngineSetupDataInferencePrecision () const
	Inference precision corresponding to the engine data contained in the tool.
int32_t	GetEngineSetupDataBatchSize () const
	Batch size corresponding to the engine data contained in the tool.
bool	GetOptimizeBatchSize () const
	Indicates whether to optimize the batch size (see EDeepLearningTool::SetBatchSize) to maximize the training and inference speed according to the engine, the device and the available memory.
void	SetOptimizeBatchSize (bool optimize)
	Indicates whether to optimize the batch size (see EDeepLearningTool::SetBatchSize) to maximize the training and inference speed according to the engine, the device and the available memory.
int32_t	GetBatchSizeForMaximumInferenceSpeed () const
	Computes the batch size that will maximize the inference speed on NVIDIA GPU. For all other devices, it will return 1.
void	SetBatchSize (int32_t size)
	Batch size. The batch size is the number of images that are processed together during training and batch inference.
int32_t	GetBatchSize () const
	Batch size. The batch size is the number of images that are processed together during training and batch inference.
bool	IsTraining () const
	Indicates whether the object is currently training.
int	WaitForTrainingCompletion (int timeout)
	Waits until the training is complete or the timeout is expired. A call to this method will block the calling thread for the shortest time between the timeout and the time it takes for the training to complete.
int	WaitForTrainingCompletion ()
	Waits until the training is complete or the timeout is expired. A call to this method will block the calling thread for the shortest time between the timeout and the time it takes for the training to complete.
int	WaitForIterationCompletion ()
	Waits until an iteration is complete. A call to this method will block the calling thread until a training iteration in the training thread is finished. This method returns the number of trained iterations.
int	StopTraining (bool wait)
	Stops training and returns the last completed iteration. If the parameter 'wait' is set to true, the method will wait for the training thread to completely stop. Otherwise, the method will return immediately.
float	GetCurrentTrainingProgression () const
	Current training progression as a percentage (between 0 and 1).
uint32_t	GetCurrentTrainingFinishedIterations () const
	Number of iterations that are already finished in the current training.
uint32_t	GetCurrentTrainingNumIterations () const
	Total number of training iterations that will be performed during the current training of the deep learning tool.
uint32_t	GetNumTrainedIterations () const
	Number of iterations that were performed to train this deep learning tool.
int	GetBestIteration () const
	Iteration at which the minimum validation error was reached. After training, the tool is in the state it was at this best iteration.
void	SetImageCacheSize (uint64_t size)
	Size in byte of the image cache. The cache is used during training to store reformatted and normalized images. A correctly sized cache can reduce the hard drive accesses and the preprocessing time for each image.
uint64_t	GetImageCacheSize () const
	Size in byte of the image cache. The cache is used during training to store reformatted and normalized images. A correctly sized cache can reduce the hard drive accesses and the preprocessing time for each image.
std::string	GetPath () const
	Path of the tool when loaded or saved using EDeepLearningTool::Save or EDeepLearningTool::Load.
std::string	GetInferenceModelPath () const
	Path of the inference model when loaded or saved using EDeepLearningTool::SaveInferenceModel, EDeepLearningTool::LoadInferenceModel or EDeepLearningTool::SerializeInferenceModel.
std::string	GetTrainingModelPath () const
	Path of the training model when loaded or saved using EDeepLearningTool::SaveTrainingModel, EDeepLearningTool::LoadTrainingModel or EDeepLearningTool::SerializeTrainingModel.
std::string	GetSettingsPath () const
	Path of the settings when loaded or saved using EDeepLearningTool::SaveSettings, EDeepLearningTool::LoadSettings or EDeepLearningTool::SerializeSettings.
void	SetAutoSavePeriod (int period)
	Auto-save period. The auto-save period is the number of training iterations between each automatic save of the model.
int	GetAutoSavePeriod () const
	Auto-save period. The auto-save period is the number of training iterations between each automatic save of the model.
void	SetLearningRate (eVision::EasyDeepLearning::EDeepLearningLearningRateType type, const std::vector< float > &parameters)
	Specifies how the value of the learning rate and how it evolves during training.
eVision::EasyDeepLearning::EDeepLearningLearningRateType	GetLearningRateType () const
	Learning rate type.
std::vector< float >	GetLearningRateParameters () const
	Parameters of the learning rate.
void	SetEnableDeterministicTraining (bool enable)
	Whether to use deterministic training algorithm that allows to repeatable training results. Default value: false.
bool	GetEnableDeterministicTraining () const
	Whether to use deterministic training algorithm that allows to repeatable training results. Default value: false.
void	SetDeterministicTrainingRandomSeed (int32_t seed)
	Random seed to use when EDeepLearningTool::SetEnableDeterministicTraining is set to true. The default value is 42.
int32_t	GetDeterministicTrainingRandomSeed () const
	Random seed to use when EDeepLearningTool::SetEnableDeterministicTraining is set to true. The default value is 42.
void	InitializeInference (int width, int height, eVision::EImageType imageType, int numImages)
	Initializes the neural network for inference (i.e. making predictions with the neural network).
void	InitializeInference (const eVision::EBaseROI &img)
	Initializes the neural network for inference (i.e. making predictions with the neural network).
void	InitializeInference (std::vector< const eVision::EBaseROI * > &imgList)
	Initializes the neural network for inference (i.e. making predictions with the neural network).
void	InitializeInference (std::vector< eVision::EImageBW8 > &imgList)
	Initializes the neural network for inference (i.e. making predictions with the neural network).
void	InitializeInference (std::vector< eVision::EImageBW16 > &imgList)
	Initializes the neural network for inference (i.e. making predictions with the neural network).
void	InitializeInference (std::vector< eVision::EImageC24 > &imgList)
	Initializes the neural network for inference (i.e. making predictions with the neural network).
uint32_t	GetNumLabels () const
	Number of labels of this tool. If the tool is not trained, the method will throw an exception.
std::string	GetLabel (uint32_t index) const
	Gets the label from its index. If the tool is not trained, the method may throw an exception.
virtual void	SetLabel (uint32_t index, const std::string &label)
	Changes a label predicted by the tool.
void	SetLabelWeight (uint32_t index, float weight)
	Sets the label weight. If the tool is not trained, the label weights are not defined. After training, they are set to the value of the training dataset label weights.
float	GetLabelWeight (uint32_t index) const
	Gets the label weight. If the tool is not trained, the label weights are not defined. After training, they are set to the value of the training dataset label weights.
void	SetLabelColor (uint32_t index, const eVision::ERGBColor &c)
	Changes a label color.
eVision::ERGBColor	GetLabelColor (uint32_t index) const
	Label color. The label colors affect how the results will be displayed.
void	SetLabelOpacity (uint32_t index, uint8_t opacity)
	Changes a label opacity (between 0 for fully transparent and 255 for opaque color).
uint8_t	GetLabelOpacity (uint32_t index) const
	Gets the label opacity (between 0 for fully transparent and 255 for opaque color). The label opacities affect with the label colors how the results will be displayed.
int	GetOptimalNumImagesForBatchSize (const eVision::EBaseROI &sampleImage) const
	Optimal number of images to process together for inference given the batch size and EDeepLearningTool::GetNumPatchesForImage.
int	GetOptimalNumImagesForBatchSize (int imageWidth, int imageHeight) const
	Optimal number of images to process together for inference given the batch size and EDeepLearningTool::GetNumPatchesForImage.
int	GetNumPatchesForImage (const eVision::EBaseROI &sampleImage) const
	Number of patches that will be extracted from an input image to perform inference.
virtual int	GetNumPatchesForImage (int imageWidth, int imageHeight) const
	Number of patches that will be extracted from an input image to perform inference.
virtual void	Clean ()
	Reset the engine and clean CPU and GPU memory allocated for running the model.
	EDeepLearningTool (CppWrapperInternal::InternalConstructor)

Additional Inherited Members
Static Public Member Functions inherited from eVision::EasyDeepLearning::EDeepLearningTool
static std::unique_ptr< eVision::EasyDeepLearning::EDeepLearningTool >	Create (const std::string &path)
	Factory method to open a deep learning tool.
static std::unique_ptr< eVision::EasyDeepLearning::EDeepLearningTool >	Create (eVision::ESerializer *serializer)
	Factory method to open a deep learning tool.
static std::unique_ptr< eVision::EasyDeepLearning::EDeepLearningTool >	Create (eVision::EasyDeepLearning::EDeepLearningToolType type)
	Factory method to open a deep learning tool.
static int	GetNumAvailableEngines ()
	Number of available engines.
static std::string	GetAvailableEngineName (int id)
	Gets the name of the engine given its id.
static std::vector< std::string >	GetAvailableEngines ()
	Available engines.
static std::vector< eVision::EasyDeepLearning::EDeepLearningDevice >	GetAvailableDevicesForEngine (const std::string &engine)
	Available devices for the specified engine.
Public Attributes inherited from eVision::EasyDeepLearning::EDeepLearningTool
CppWrapperInternal::WrappedClass	wrappedClass_

Detailed Description

EasyLocate tool - axis aligned bounding box version.

The ELocator locates and classifies objects (or defects). With this class, an object is defined by the axis-aligned bounding box that surrounds it and a label (see ELocatorObject class). The tool must be trained using a dataset with annotated objects. The minimum size resolution supported by the tool is 128. The maximum size of the image supported is 500 000 pixels.

The tool has 5 main parameters:

• A set of anchors, i.e. pre-defined object size that it must be able to detect. The anchors can be automatically determined from the training dataset, specified manually using ELocatorBase::SetPredictionAnchors, or generated according to size information about the objects (see ELocator::GenerateAnchors).
• A detection threshold (ELocatorBase::GetDetectionThreshold) to accept or reject predicted objects based on their score.
• The maximum number of objects in an image (ELocatorBase::SetMaxNumberOfObjects)
• The maximum overlap between predicted objects with the same label (ELocator::SetSameLabelMaxOverlap, ELocatorBase::SetSameLabelMaxObjectProximity)
• The maximum overlap between predicted objects regardless of their label (ELocator::SetAbsoluteMaxOverlap, ELocatorBase::SetAbsoluteMaxObjectProximity). Except for the anchors, all the parameters can be changed before and after training the tool.

License(s): EasyLocate or EasyLocateInference

Constructor & Destructor Documentation

ELocator() [1/3]

eVision::EasyDeepLearning::ELocator::ELocator

(

)

Constructs a ELocator object.

ELocator() [2/3]

eVision::EasyDeepLearning::ELocator::ELocator

(

const eVision::EasyDeepLearning::ELocator &

other

)

Copy constructor of a ELocator object.

Parameters

other

Reference to the ELocator object that should be copied

~ELocator()

eVision::EasyDeepLearning::ELocator::~ELocator ( ) throw (  )

virtual

ELocator() [3/3]

eVision::EasyDeepLearning::ELocator::ELocator

(

CppWrapperInternal::InternalConstructor

)

Member Function Documentation

GenerateAnchors() [1/2]

std::vector< eVision::ESize > eVision::EasyDeepLearning::ELocator::GenerateAnchors

(

const eVision::EasyDeepLearning::EDataset &

dataset

)

const

Generates anchors based on the objects in the dataset or a formal specification.

For the version that takes a dataset, the properties ELocatorBase::SetWidth and ELocatorBase::SetHeight must be set.

Parameters

dataset

Dataset

The dimension of an object corresponds to the square root of its area.

GenerateAnchors() [2/2]

std::vector< eVision::ESize > eVision::EasyDeepLearning::ELocator::GenerateAnchors	(	float	minDimension,
		float	maxDimension,
		int	numSubScales,
		std::vector< float >	aspectRatios ) const

Generates anchors based on the objects in the dataset or a formal specification.

For the version that takes a dataset, the properties ELocatorBase::SetWidth and ELocatorBase::SetHeight must be set.

Parameters

minDimension	Minimum dimension of objects (minimum value: 16)
maxDimension	Maximum dimension of objects.
numSubScales	Number of sub-scales to produce for each scale covered by the given dimensions.
aspectRatios	Aspect ratios to generate at each scale. The aspect ratios are the ratios between the width and height of the anchor. Recommended value: {1.0f, 0.5f, 2.0f}.

The dimension of an object corresponds to the square root of its area.

GetAbsoluteMaxOverlap()

float eVision::EasyDeepLearning::ELocator::GetAbsoluteMaxOverlap

(

)

const

Maximum overlap (intersection over union) between two predicted objects regardless of their label.

The value of the parameter must be between 0 (no overlap allowed between two predicted objects with different labels) and 1 (full overlap allowed between two predicted objects with different labels).

Default value: 1.

This parameter can be changed before and after training.

GetSameLabelMaxOverlap()

float eVision::EasyDeepLearning::ELocator::GetSameLabelMaxOverlap

(

)

const

Maximum overlap (intersection over union) between two predicted objects with the same label.

The value of the parameter must be between 0 (no overlap allowed between two predicted objects with the same label) and 1 (full overlap allowed between two predicted objects with the same label). A low value will reduce the amount of falsely detected objects but may increase the number of missed objects. A high value will increase the number of falsely detected objects and reduce the number of missed objects.

Default value: 0.5

This parameter is also used to compute the matching between ground truth and predicted objects during evaluation.

This parameter can be changed before and after training.

GetToolType()

eVision::EasyDeepLearning::EDeepLearningToolType eVision::EasyDeepLearning::ELocator::GetToolType ( ) const

virtual

Type of the deep learning tool.

Implements eVision::EasyDeepLearning::EDeepLearningTool.

operator=()

eVision::EasyDeepLearning::ELocator & eVision::EasyDeepLearning::ELocator::operator=

(

const eVision::EasyDeepLearning::ELocator &

other

)

Assignment operator.

Parameters

other

Reference to the ELocator object that should be copied

SerializeSettings()

void eVision::EasyDeepLearning::ELocator::SerializeSettings ( eVision::ESerializer * serializer )

virtual

Serializes the settings of the locator.

Parameters

serializer

Pointer to ESerializer

Reimplemented from eVision::EasyDeepLearning::ELocatorBase.

SetAbsoluteMaxOverlap()

void eVision::EasyDeepLearning::ELocator::SetAbsoluteMaxOverlap

(

float

val

)

Maximum overlap (intersection over union) between two predicted objects regardless of their label.

The value of the parameter must be between 0 (no overlap allowed between two predicted objects with different labels) and 1 (full overlap allowed between two predicted objects with different labels).

Default value: 1.

Parameters

val	Maximum overlap

This parameter can be changed before and after training.

SetSameLabelMaxOverlap()

void eVision::EasyDeepLearning::ELocator::SetSameLabelMaxOverlap

(

float

val

)

Maximum overlap (intersection over union) between two predicted objects with the same label.

The value of the parameter must be between 0 (no overlap allowed between two predicted objects with the same label) and 1 (full overlap allowed between two predicted objects with the same label). A low value will reduce the amount of falsely detected objects but may increase the number of missed objects. A high value will increase the number of falsely detected objects and reduce the number of missed objects.

Default value: 0.5

Parameters

val	Maximum overlap

This parameter is also used to compute the matching between ground truth and predicted objects during evaluation.

This parameter can be changed before and after training.