tesseract::LSTMTrainer Class Reference

#include <lstmtrainer.h>

Inheritance diagram for tesseract::LSTMTrainer:

Collaboration diagram for tesseract::LSTMTrainer:

Public Member Functions
	LSTMTrainer ()
	LSTMTrainer (FileReader file_reader, FileWriter file_writer, CheckPointReader checkpoint_reader, CheckPointWriter checkpoint_writer, const char *model_base, const char *checkpoint_name, int debug_interval, int64_t max_memory)
virtual	~LSTMTrainer ()
bool	TryLoadingCheckpoint (const char *filename, const char *old_traineddata)
void	InitCharSet (const std::string &traineddata_path)
void	InitCharSet (const TessdataManager &mgr)
bool	InitNetwork (const STRING &network_spec, int append_index, int net_flags, float weight_range, float learning_rate, float momentum, float adam_beta)
int	InitTensorFlowNetwork (const std::string &tf_proto)
void	InitIterations ()
double	ActivationError () const
double	CharError () const
const double *	error_rates () const
double	best_error_rate () const
int	best_iteration () const
int	learning_iteration () const
int32_t	improvement_steps () const
void	set_perfect_delay (int delay)
const GenericVector< char > &	best_trainer () const
double	NewSingleError (ErrorTypes type) const
double	LastSingleError (ErrorTypes type) const
const DocumentCache &	training_data () const
DocumentCache *	mutable_training_data ()
Trainability	GridSearchDictParams (const ImageData *trainingdata, int iteration, double min_dict_ratio, double dict_ratio_step, double max_dict_ratio, double min_cert_offset, double cert_offset_step, double max_cert_offset, STRING *results)
void	DebugNetwork ()
bool	LoadAllTrainingData (const GenericVector< STRING > &filenames, CachingStrategy cache_strategy, bool randomly_rotate)
bool	MaintainCheckpoints (TestCallback tester, STRING *log_msg)
bool	MaintainCheckpointsSpecific (int iteration, const GenericVector< char > *train_model, const GenericVector< char > *rec_model, TestCallback tester, STRING *log_msg)
void	PrepareLogMsg (STRING *log_msg) const
void	LogIterations (const char *intro_str, STRING *log_msg) const
bool	TransitionTrainingStage (float error_threshold)
int	CurrentTrainingStage () const
bool	Serialize (SerializeAmount serialize_amount, const TessdataManager *mgr, TFile *fp) const
bool	DeSerialize (const TessdataManager *mgr, TFile *fp)
void	StartSubtrainer (STRING *log_msg)
SubTrainerResult	UpdateSubtrainer (STRING *log_msg)
void	ReduceLearningRates (LSTMTrainer *samples_trainer, STRING *log_msg)
int	ReduceLayerLearningRates (double factor, int num_samples, LSTMTrainer *samples_trainer)
bool	EncodeString (const STRING &str, GenericVector< int > *labels) const
const ImageData *	TrainOnLine (LSTMTrainer *samples_trainer, bool batch)
Trainability	TrainOnLine (const ImageData *trainingdata, bool batch)
Trainability	PrepareForBackward (const ImageData *trainingdata, NetworkIO *fwd_outputs, NetworkIO *targets)
bool	SaveTrainingDump (SerializeAmount serialize_amount, const LSTMTrainer *trainer, GenericVector< char > *data) const
bool	ReadTrainingDump (const GenericVector< char > &data, LSTMTrainer *trainer) const
bool	ReadSizedTrainingDump (const char *data, int size, LSTMTrainer *trainer) const
bool	ReadLocalTrainingDump (const TessdataManager *mgr, const char *data, int size)
void	SetupCheckpointInfo ()
bool	SaveTraineddata (const STRING &filename)
void	SaveRecognitionDump (GenericVector< char > *data) const
STRING	DumpFilename () const
void	FillErrorBuffer (double new_error, ErrorTypes type)
std::vector< int >	MapRecoder (const UNICHARSET &old_chset, const UnicharCompress &old_recoder) const
Public Member Functions inherited from tesseract::LSTMRecognizer
	LSTMRecognizer ()
	~LSTMRecognizer ()
int	NumOutputs () const
int	training_iteration () const
int	sample_iteration () const
double	learning_rate () const
LossType	OutputLossType () const
bool	SimpleTextOutput () const
bool	IsIntMode () const
bool	IsRecoding () const
bool	IsTensorFlow () const
GenericVector< STRING >	EnumerateLayers () const
Network *	GetLayer (const STRING &id) const
float	GetLayerLearningRate (const STRING &id) const
void	ScaleLearningRate (double factor)
void	ScaleLayerLearningRate (const STRING &id, double factor)
void	ConvertToInt ()
const UNICHARSET &	GetUnicharset () const
const UnicharCompress &	GetRecoder () const
const Dict *	GetDict () const
void	SetIteration (int iteration)
int	NumInputs () const
int	null_char () const
bool	Load (const char *lang, TessdataManager *mgr)
bool	Serialize (const TessdataManager *mgr, TFile *fp) const
bool	DeSerialize (const TessdataManager *mgr, TFile *fp)
bool	LoadCharsets (const TessdataManager *mgr)
bool	LoadRecoder (TFile *fp)
bool	LoadDictionary (const char *lang, TessdataManager *mgr)
void	RecognizeLine (const ImageData &image_data, bool invert, bool debug, double worst_dict_cert, const TBOX &line_box, PointerVector< WERD_RES > *words, int lstm_choice_mode=0)
void	OutputStats (const NetworkIO &outputs, float *min_output, float *mean_output, float *sd)
bool	RecognizeLine (const ImageData &image_data, bool invert, bool debug, bool re_invert, bool upside_down, float *scale_factor, NetworkIO *inputs, NetworkIO *outputs)
STRING	DecodeLabels (const GenericVector< int > &labels)
void	DisplayForward (const NetworkIO &inputs, const GenericVector< int > &labels, const GenericVector< int > &label_coords, const char *window_name, ScrollView **window)
void	LabelsFromOutputs (const NetworkIO &outputs, GenericVector< int > *labels, GenericVector< int > *xcoords)

Static Public Member Functions
static bool	EncodeString (const STRING &str, const UNICHARSET &unicharset, const UnicharCompress *recoder, bool simple_text, int null_char, GenericVector< int > *labels)

Protected Member Functions
void	InitCharSet ()
void	SetNullChar ()
void	EmptyConstructor ()
bool	DebugLSTMTraining (const NetworkIO &inputs, const ImageData &trainingdata, const NetworkIO &fwd_outputs, const GenericVector< int > &truth_labels, const NetworkIO &outputs)
void	DisplayTargets (const NetworkIO &targets, const char *window_name, ScrollView **window)
bool	ComputeTextTargets (const NetworkIO &outputs, const GenericVector< int > &truth_labels, NetworkIO *targets)
bool	ComputeCTCTargets (const GenericVector< int > &truth_labels, NetworkIO *outputs, NetworkIO *targets)
double	ComputeErrorRates (const NetworkIO &deltas, double char_error, double word_error)
double	ComputeRMSError (const NetworkIO &deltas)
double	ComputeWinnerError (const NetworkIO &deltas)
double	ComputeCharError (const GenericVector< int > &truth_str, const GenericVector< int > &ocr_str)
double	ComputeWordError (STRING *truth_str, STRING *ocr_str)
void	UpdateErrorBuffer (double new_error, ErrorTypes type)
void	RollErrorBuffers ()
STRING	UpdateErrorGraph (int iteration, double error_rate, const GenericVector< char > &model_data, TestCallback tester)
Protected Member Functions inherited from tesseract::LSTMRecognizer
void	SetRandomSeed ()
void	DisplayLSTMOutput (const GenericVector< int > &labels, const GenericVector< int > &xcoords, int height, ScrollView *window)
void	DebugActivationPath (const NetworkIO &outputs, const GenericVector< int > &labels, const GenericVector< int > &xcoords)
void	DebugActivationRange (const NetworkIO &outputs, const char *label, int best_choice, int x_start, int x_end)
void	LabelsViaReEncode (const NetworkIO &output, GenericVector< int > *labels, GenericVector< int > *xcoords)
void	LabelsViaSimpleText (const NetworkIO &output, GenericVector< int > *labels, GenericVector< int > *xcoords)
const char *	DecodeLabel (const GenericVector< int > &labels, int start, int *end, int *decoded)
const char *	DecodeSingleLabel (int label)

Protected Attributes
ScrollView *	align_win_
ScrollView *	target_win_
ScrollView *	ctc_win_
ScrollView *	recon_win_
int	debug_interval_
int	checkpoint_iteration_
STRING	model_base_
STRING	checkpoint_name_
bool	randomly_rotate_
DocumentCache	training_data_
STRING	best_model_name_
int	num_training_stages_
FileReader	file_reader_
FileWriter	file_writer_
CheckPointReader	checkpoint_reader_
CheckPointWriter	checkpoint_writer_
double	best_error_rate_
double	best_error_rates_ [ET_COUNT]
int	best_iteration_
double	worst_error_rate_
double	worst_error_rates_ [ET_COUNT]
int	worst_iteration_
int	stall_iteration_
GenericVector< char >	best_model_data_
GenericVector< char >	worst_model_data_
GenericVector< char >	best_trainer_
LSTMTrainer *	sub_trainer_
float	error_rate_of_last_saved_best_
int	training_stage_
GenericVector< double >	best_error_history_
GenericVector< int >	best_error_iterations_
int32_t	improvement_steps_
int	learning_iteration_
int	prev_sample_iteration_
int	perfect_delay_
int	last_perfect_training_iteration_
GenericVector< double >	error_buffers_ [ET_COUNT]
double	error_rates_ [ET_COUNT]
TessdataManager	mgr_
Protected Attributes inherited from tesseract::LSTMRecognizer
Network *	network_
CCUtil	ccutil_
UnicharCompress	recoder_
STRING	network_str_
int32_t	training_flags_
int32_t	training_iteration_
int32_t	sample_iteration_
int32_t	null_char_
float	learning_rate_
float	momentum_
float	adam_beta_
TRand	randomizer_
NetworkScratch	scratch_space_
Dict *	dict_
RecodeBeamSearch *	search_
ScrollView *	debug_win_

Static Protected Attributes
static const int	kRollingBufferSize_ = 1000

Constructor & Destructor Documentation

LSTMTrainer() [1/2]

tesseract::LSTMTrainer::LSTMTrainer

(

)

LSTMTrainer() [2/2]

tesseract::LSTMTrainer::LSTMTrainer	(	FileReader	file_reader,
		FileWriter	file_writer,
		CheckPointReader	checkpoint_reader,
		CheckPointWriter	checkpoint_writer,
		const char *	model_base,
		const char *	checkpoint_name,
		int	debug_interval,
		int64_t	max_memory
	)

~LSTMTrainer()

tesseract::LSTMTrainer::~LSTMTrainer ( )

virtual

Member Function Documentation

ActivationError()

double tesseract::LSTMTrainer::ActivationError ( ) const

inline

best_error_rate()

double tesseract::LSTMTrainer::best_error_rate ( ) const

inline

best_iteration()

int tesseract::LSTMTrainer::best_iteration ( ) const

inline

best_trainer()

const GenericVector<char>& tesseract::LSTMTrainer::best_trainer ( ) const

inline

CharError()

double tesseract::LSTMTrainer::CharError ( ) const

inline

ComputeCharError()

double tesseract::LSTMTrainer::ComputeCharError ( const GenericVector< int > &  truth_str, const GenericVector< int > &  ocr_str  )

protected

ComputeCTCTargets()

bool tesseract::LSTMTrainer::ComputeCTCTargets ( const GenericVector< int > &  truth_labels, NetworkIO *  outputs, NetworkIO *  targets  )

protected

ComputeErrorRates()

double tesseract::LSTMTrainer::ComputeErrorRates ( const NetworkIO &  deltas, double  char_error, double  word_error  )

protected

ComputeRMSError()

double tesseract::LSTMTrainer::ComputeRMSError ( const NetworkIO &  deltas )

protected

ComputeTextTargets()

bool tesseract::LSTMTrainer::ComputeTextTargets ( const NetworkIO &  outputs, const GenericVector< int > &  truth_labels, NetworkIO *  targets  )

protected

ComputeWinnerError()

double tesseract::LSTMTrainer::ComputeWinnerError ( const NetworkIO &  deltas )

protected

ComputeWordError()

double tesseract::LSTMTrainer::ComputeWordError ( STRING *  truth_str, STRING *  ocr_str  )

protected

CurrentTrainingStage()

int tesseract::LSTMTrainer::CurrentTrainingStage ( ) const

inline

DebugLSTMTraining()

bool tesseract::LSTMTrainer::DebugLSTMTraining ( const NetworkIO &  inputs, const ImageData &  trainingdata, const NetworkIO &  fwd_outputs, const GenericVector< int > &  truth_labels, const NetworkIO &  outputs  )

protected

DebugNetwork()

void tesseract::LSTMTrainer::DebugNetwork

(

)

DeSerialize()

bool tesseract::LSTMTrainer::DeSerialize	(	const TessdataManager *	mgr,
		TFile *	fp
	)

DisplayTargets()

void tesseract::LSTMTrainer::DisplayTargets ( const NetworkIO &  targets, const char *  window_name, ScrollView **  window  )

protected

DumpFilename()

STRING tesseract::LSTMTrainer::DumpFilename

(

)

const

EmptyConstructor()

void tesseract::LSTMTrainer::EmptyConstructor ( )

protected

EncodeString() [1/2]

bool tesseract::LSTMTrainer::EncodeString ( const STRING &  str, GenericVector< int > *  labels  ) const

inline

EncodeString() [2/2]

bool tesseract::LSTMTrainer::EncodeString ( const STRING &  str, const UNICHARSET &  unicharset, const UnicharCompress *  recoder, bool  simple_text, int  null_char, GenericVector< int > *  labels  )

static

error_rates()

const double* tesseract::LSTMTrainer::error_rates ( ) const

inline

FillErrorBuffer()

void tesseract::LSTMTrainer::FillErrorBuffer	(	double	new_error,
		ErrorTypes	type
	)

GridSearchDictParams()

Trainability tesseract::LSTMTrainer::GridSearchDictParams	(	const ImageData *	trainingdata,
		int	iteration,
		double	min_dict_ratio,
		double	dict_ratio_step,
		double	max_dict_ratio,
		double	min_cert_offset,
		double	cert_offset_step,
		double	max_cert_offset,
		STRING *	results
	)

improvement_steps()

int32_t tesseract::LSTMTrainer::improvement_steps ( ) const

inline

InitCharSet() [1/3]

void tesseract::LSTMTrainer::InitCharSet ( const std::string &  traineddata_path )

inline

InitCharSet() [2/3]

void tesseract::LSTMTrainer::InitCharSet ( const TessdataManager &  mgr )

inline

InitCharSet() [3/3]

void tesseract::LSTMTrainer::InitCharSet ( )

protected

InitIterations()

void tesseract::LSTMTrainer::InitIterations

(

)

InitNetwork()

bool tesseract::LSTMTrainer::InitNetwork	(	const STRING &	network_spec,
		int	append_index,
		int	net_flags,
		float	weight_range,
		float	learning_rate,
		float	momentum,
		float	adam_beta
	)

InitTensorFlowNetwork()

int tesseract::LSTMTrainer::InitTensorFlowNetwork

(

const std::string &

tf_proto

)

LastSingleError()

double tesseract::LSTMTrainer::LastSingleError ( ErrorTypes  type ) const

inline

learning_iteration()

int tesseract::LSTMTrainer::learning_iteration ( ) const

inline

LoadAllTrainingData()

bool tesseract::LSTMTrainer::LoadAllTrainingData	(	const GenericVector< STRING > &	filenames,
		CachingStrategy	cache_strategy,
		bool	randomly_rotate
	)

LogIterations()

void tesseract::LSTMTrainer::LogIterations	(	const char *	intro_str,
		STRING *	log_msg
	)		const

MaintainCheckpoints()

bool tesseract::LSTMTrainer::MaintainCheckpoints	(	TestCallback	tester,
		STRING *	log_msg
	)

MaintainCheckpointsSpecific()

bool tesseract::LSTMTrainer::MaintainCheckpointsSpecific	(	int	iteration,
		const GenericVector< char > *	train_model,
		const GenericVector< char > *	rec_model,
		TestCallback	tester,
		STRING *	log_msg
	)

MapRecoder()

std::vector< int > tesseract::LSTMTrainer::MapRecoder	(	const UNICHARSET &	old_chset,
		const UnicharCompress &	old_recoder
	)		const

mutable_training_data()

DocumentCache* tesseract::LSTMTrainer::mutable_training_data ( )

inline

NewSingleError()

double tesseract::LSTMTrainer::NewSingleError ( ErrorTypes  type ) const

inline

PrepareForBackward()

Trainability tesseract::LSTMTrainer::PrepareForBackward	(	const ImageData *	trainingdata,
		NetworkIO *	fwd_outputs,
		NetworkIO *	targets
	)

PrepareLogMsg()

void tesseract::LSTMTrainer::PrepareLogMsg

(

STRING *

log_msg

)

const

ReadLocalTrainingDump()

bool tesseract::LSTMTrainer::ReadLocalTrainingDump	(	const TessdataManager *	mgr,
		const char *	data,
		int	size
	)

ReadSizedTrainingDump()

bool tesseract::LSTMTrainer::ReadSizedTrainingDump ( const char *  data, int  size, LSTMTrainer *  trainer  ) const

inline

ReadTrainingDump()

bool tesseract::LSTMTrainer::ReadTrainingDump ( const GenericVector< char > &  data, LSTMTrainer *  trainer  ) const

inline

ReduceLayerLearningRates()

int tesseract::LSTMTrainer::ReduceLayerLearningRates	(	double	factor,
		int	num_samples,
		LSTMTrainer *	samples_trainer
	)

ReduceLearningRates()

void tesseract::LSTMTrainer::ReduceLearningRates	(	LSTMTrainer *	samples_trainer,
		STRING *	log_msg
	)

RollErrorBuffers()

void tesseract::LSTMTrainer::RollErrorBuffers ( )

protected

SaveRecognitionDump()

void tesseract::LSTMTrainer::SaveRecognitionDump

(

GenericVector< char > *

data

)

const

SaveTraineddata()

bool tesseract::LSTMTrainer::SaveTraineddata

(

const STRING &

filename

)

SaveTrainingDump()

bool tesseract::LSTMTrainer::SaveTrainingDump	(	SerializeAmount	serialize_amount,
		const LSTMTrainer *	trainer,
		GenericVector< char > *	data
	)		const

Serialize()

bool tesseract::LSTMTrainer::Serialize	(	SerializeAmount	serialize_amount,
		const TessdataManager *	mgr,
		TFile *	fp
	)		const

set_perfect_delay()

void tesseract::LSTMTrainer::set_perfect_delay ( int  delay )

inline

SetNullChar()

void tesseract::LSTMTrainer::SetNullChar ( )

protected

SetupCheckpointInfo()

void tesseract::LSTMTrainer::SetupCheckpointInfo

(

)

StartSubtrainer()

void tesseract::LSTMTrainer::StartSubtrainer

(

STRING *

log_msg

)

training_data()

const DocumentCache& tesseract::LSTMTrainer::training_data ( ) const

inline

TrainOnLine() [1/2]

const ImageData* tesseract::LSTMTrainer::TrainOnLine ( LSTMTrainer *  samples_trainer, bool  batch  )

inline

TrainOnLine() [2/2]

Trainability tesseract::LSTMTrainer::TrainOnLine	(	const ImageData *	trainingdata,
		bool	batch
	)

TransitionTrainingStage()

bool tesseract::LSTMTrainer::TransitionTrainingStage

(

float

error_threshold

)

TryLoadingCheckpoint()

bool tesseract::LSTMTrainer::TryLoadingCheckpoint	(	const char *	filename,
		const char *	old_traineddata
	)

UpdateErrorBuffer()

void tesseract::LSTMTrainer::UpdateErrorBuffer ( double  new_error, ErrorTypes  type  )

protected

UpdateErrorGraph()

STRING tesseract::LSTMTrainer::UpdateErrorGraph ( int  iteration, double  error_rate, const GenericVector< char > &  model_data, TestCallback  tester  )

protected

UpdateSubtrainer()

SubTrainerResult tesseract::LSTMTrainer::UpdateSubtrainer

(

STRING *

log_msg

)

Member Data Documentation

align_win_

ScrollView* tesseract::LSTMTrainer::align_win_

protected

best_error_history_

GenericVector<double> tesseract::LSTMTrainer::best_error_history_

protected

best_error_iterations_

GenericVector<int> tesseract::LSTMTrainer::best_error_iterations_

protected

best_error_rate_

double tesseract::LSTMTrainer::best_error_rate_

protected

best_error_rates_

double tesseract::LSTMTrainer::best_error_rates_[ET_COUNT]

protected

best_iteration_

int tesseract::LSTMTrainer::best_iteration_

protected

best_model_data_

GenericVector<char> tesseract::LSTMTrainer::best_model_data_

protected

best_model_name_

STRING tesseract::LSTMTrainer::best_model_name_

protected

best_trainer_

GenericVector<char> tesseract::LSTMTrainer::best_trainer_

protected

checkpoint_iteration_

int tesseract::LSTMTrainer::checkpoint_iteration_

protected

checkpoint_name_

STRING tesseract::LSTMTrainer::checkpoint_name_

protected

checkpoint_reader_

CheckPointReader tesseract::LSTMTrainer::checkpoint_reader_

protected

checkpoint_writer_

CheckPointWriter tesseract::LSTMTrainer::checkpoint_writer_

protected

ctc_win_

ScrollView* tesseract::LSTMTrainer::ctc_win_

protected

debug_interval_

int tesseract::LSTMTrainer::debug_interval_

protected

error_buffers_

GenericVector<double> tesseract::LSTMTrainer::error_buffers_[ET_COUNT]

protected

error_rate_of_last_saved_best_

float tesseract::LSTMTrainer::error_rate_of_last_saved_best_

protected

error_rates_

double tesseract::LSTMTrainer::error_rates_[ET_COUNT]

protected

file_reader_

FileReader tesseract::LSTMTrainer::file_reader_

protected

file_writer_

FileWriter tesseract::LSTMTrainer::file_writer_

protected

improvement_steps_

int32_t tesseract::LSTMTrainer::improvement_steps_

protected

kRollingBufferSize_

const int tesseract::LSTMTrainer::kRollingBufferSize_ = 1000

staticprotected

last_perfect_training_iteration_

int tesseract::LSTMTrainer::last_perfect_training_iteration_

protected

learning_iteration_

int tesseract::LSTMTrainer::learning_iteration_

protected

mgr_

TessdataManager tesseract::LSTMTrainer::mgr_

protected

model_base_

STRING tesseract::LSTMTrainer::model_base_

protected

num_training_stages_

int tesseract::LSTMTrainer::num_training_stages_

protected

perfect_delay_

int tesseract::LSTMTrainer::perfect_delay_

protected

prev_sample_iteration_

int tesseract::LSTMTrainer::prev_sample_iteration_

protected

randomly_rotate_

bool tesseract::LSTMTrainer::randomly_rotate_

protected

recon_win_

ScrollView* tesseract::LSTMTrainer::recon_win_

protected

stall_iteration_

int tesseract::LSTMTrainer::stall_iteration_

protected

sub_trainer_

LSTMTrainer* tesseract::LSTMTrainer::sub_trainer_

protected

target_win_

ScrollView* tesseract::LSTMTrainer::target_win_

protected

training_data_

DocumentCache tesseract::LSTMTrainer::training_data_

protected

training_stage_

int tesseract::LSTMTrainer::training_stage_

protected

worst_error_rate_

double tesseract::LSTMTrainer::worst_error_rate_

protected

worst_error_rates_

double tesseract::LSTMTrainer::worst_error_rates_[ET_COUNT]

protected

worst_iteration_

int tesseract::LSTMTrainer::worst_iteration_

protected

worst_model_data_

GenericVector<char> tesseract::LSTMTrainer::worst_model_data_

protected

---

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

• /home/stephane/src/tesseract/src/lstm/lstmtrainer.h
• /home/stephane/src/tesseract/src/lstm/lstmtrainer.cpp