Types & classes — github.com/CSTR-Edinburgh/merlin

CodeHub MCPcopy Index your code

hub / github.com/CSTR-Edinburgh/merlin / types & classes

Types & classes119 in github.com/CSTR-Edinburgh/merlin

⨍ Functions 1,134◇ Types & classes 119

↓ 86 callersClassSilenceRemover

src/frontend/silence_remover.py:46

↓ 68 callersClassBinaryIOCollection

src/io_funcs/binary_io.py:44

↓ 57 callersClassMinMaxNormalisation

src/frontend/min_max_norm.py:45

↓ 44 callersClassLabelComposer

src/frontend/label_composer.py:59

↓ 31 callersClassHTSLabelNormalisation

This class is to convert HTS format labels into continous or binary values, and store as binary format with float32 precision. The class supports

src/frontend/label_normalisation.py:40

↓ 21 callersClassListDataProvider

This class provides an interface to load data into CPU/GPU memory utterance by utterance or block by block. In speech synthesis, usually we are

src/utils/providers.py:46

↓ 20 callersClassMeanVarianceNorm

plan: 1: support normal MVN and denormalisation for both input and output 2: support stream-based operation: for example, some streams

src/frontend/mean_variance_norm.py:47

↓ 19 callersClassIndividualDistortionComp

src/utils/compute_distortion.py:149

↓ 17 callersClassParameterGeneration

src/frontend/parameter_generation.py:55

↓ 14 callersClassAcousticComposition

variables inheritate from AcousticBase: self.compute_dynamic = {} self.file_number = 0 self.data_stream_number = 0

src/frontend/acoustic_composition.py:47

↓ 12 callersClassListDataProviderWithProjectionIndex

Added kwarg index_to_project to __init__

src/utils/providers.py:804

↓ 9 callersClassBinaryIOCollection

misc/scripts/voice_conversion/binary_io.py:5

↓ 7 callersClassFilePaths

src/utils/file_paths.py:26

↓ 6 callersClassDNN

src/models/dnn.py:62

↓ 5 callersClassDeepRecurrentNetwork

This class is to assemble various neural network architectures. From basic feedforward neural network to bidirectional gated recurrent neural net

src/models/deep_rnn.py:21

↓ 5 callersClassGeneralLayer

src/layers/layers.py:182

↓ 5 callersClassMeanVarianceNorm

plan: 1: support normal MVN and denormalisation for both input and output 2: support stream-based operation: for example, some streams

misc/scripts/alignment/state_align/mean_variance_norm.py:7

↓ 4 callersClassAlignFeats

misc/scripts/voice_conversion/align_feats.py:6

↓ 4 callersClassHTKFeat_read

Read HTK format feature files

misc/scripts/alignment/state_align/htkmfc.py:55

↓ 4 callersClassHTSLabelModification

This class is to modify HTS format labels with predicted duration. Time alignments are expected in the HTS labels. Here is an example of the HTS

src/frontend/label_modifier.py:9

↓ 4 callersClassMLParameterGenerationFast

src/frontend/mlpg_fast.py:56

↓ 4 callersClassSimplifiedLstm

This class implements a simplified LSTM block which only keeps the forget gate, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/gating.py:753

↓ 4 callersClassVanillaLstm

This class implements the standard LSTM block, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/gating.py:461

↓ 3 callersClassHTK_Parm_IO

For details look at the HTK book, Chapter 5.10 Storage of Parameter Files

src/io_funcs/htk_io.py:24

↓ 3 callersClassKerasClass

src/run_keras_with_merlin_io.py:48

↓ 3 callersClassLinearLayer

src/layers/layers.py:86

↓ 3 callersClassTensorflowClass

src/run_tensorflow_with_merlin_io.py:48

↓ 2 callersClassBidirectionLstm

src/layers/gating.py:846

↓ 2 callersClassBidirectionSLstm

src/layers/gating.py:835

↓ 2 callersClassColouredFormatter

src/logplot/logging_plotting.py:268

↓ 2 callersClassDistributedSequenceEncoder

src/models/seq2seq.py:36

↓ 2 callersClassGatedRecurrentUnit

This class implements a gated recurrent unit (GRU), as proposed in Cho et al 2014 (http://arxiv.org/pdf/1406.1078.pdf).

src/layers/gating.py:870

↓ 2 callersClassHTKFeat_write

Write Sphinx-II format feature files

misc/scripts/alignment/state_align/htkmfc.py:135

↓ 2 callersClassHTK_Parm_IO

For details look at the HTK book, Chapter 5.10 Storage of Parameter Files

misc/scripts/alignment/state_align/htk_io.py:24

↓ 2 callersClassHTSDurationLabelNormalisation

Unlike HTSLabelNormalisation, HTSDurationLabelNormalisation does not accept timings. One line of labels is converted into 1 datapoint, that i

src/frontend/label_normalisation.py:920

↓ 2 callersClassHiddenLayer

src/layers/layers.py:264

↓ 2 callersClassMergeFeat

src/frontend/merge_features.py:7

↓ 2 callersClassMyDropoutWrapper

src/tensorflow_lib/model.py:185

↓ 2 callersClassRecurrentOutputLayer

This class implements a standard recurrent output layer: y_{t} = g(h_{t}W^{hy} + y_{t}W^{yy} + b_{y})

src/layers/recurrent_output_layer.py:8

↓ 2 callersClassSigmoidLayer_LHUC

src/layers/lhuc_layer.py:11

↓ 2 callersClassSimplifiedGRU

This class implements a simplified GRU block which only keeps the forget gate, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/gating.py:791

↓ 2 callersClassVanillaLstm_LHUC

This class implements the standard LSTM block, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/lhuc_layer.py:194

↓ 2 callersClassVanillaRNN

This class implements a standard recurrent neural network: h_{t} = f(W^{hx}x_{t} + W^{hh}h_{t-1}+b_{h})

src/layers/gating.py:11

↓ 1 callersClassBinaryIOCollection

misc/scripts/alignment/state_align/binary_io.py:5

↓ 1 callersClassCMPNormalisation

src/frontend/acoustic_normalisation.py:46

↓ 1 callersClassDNN

src/models/dnn_cm.py:19

↓ 1 callersClassDeepEncoderDecoderNetwork

This class is to assemble various neural network architectures. From basic feedforward neural network to bidirectional gated recurrent neural net

src/models/hed_rnn.py:23

↓ 1 callersClassDistributionInfo

misc/scripts/hybrid_voice/processHybridInfo.py:64

↓ 1 callersClassForcedAlignment

misc/scripts/alignment/state_align/forced_alignment.py:26

↓ 1 callersClassLayerNormGRUCell

Gated Recurrent Unit cell (cf. http://arxiv.org/abs/1406.1078).

src/tensorflow_lib/model.py:330

↓ 1 callersClassLstmNFG

This class implements a LSTM block without the forget gate, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/gating.py:592

↓ 1 callersClassLstmNIG

This class implements a LSTM block without the input gate, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/gating.py:630

↓ 1 callersClassLstmNOG

This class implements a LSTM block without the output gate, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/gating.py:670

↓ 1 callersClassLstmNoPeepholes

This class implements a LSTM block without the peephole connections, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/gating.py:710

↓ 1 callersClassMLParameterGeneration

src/frontend/mlpg.py:50

↓ 1 callersClassMixtureDensityNetwork

src/models/mdn.py:63

↓ 1 callersClassMixtureDensityOutputLayer

src/layers/layers.py:51

↓ 1 callersClassPlotWithData

src/logplot/logging_plotting.py:77

↓ 1 callersClassSequentialDNN

src/models/st_dnn_cm.py:19

↓ 1 callersClassSigmoidLayer

src/layers/layers.py:128

↓ 1 callersClassSimplifiedLstmDecoder

This class implements a simplified LSTM block which only keeps the forget gate, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/gating.py:550

↓ 1 callersClassTrainKerasModels

src/keras_lib/train.py:49

↓ 1 callersClassTrainTensorflowModels

src/tensorflow_lib/train.py:49

↓ 1 callersClassTrain_Encoder_Decoder_Models

src/tensorflow_lib/train.py:236

↓ 1 callersClassValidation

Runs Merlin validations

src/validation.py:33

↓ 1 callersClassVanillaLstmDecoder

This class implements the standard LSTM block, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/gating.py:503

↓ 1 callersClassVanillaRNNDecoder

This class implements a standard recurrent neural network decoder: h_{t} = f(W^{hx}x_{t} + W^{hh}h_{t-1}+ W^{yh}y_{t-1} + b_{h}) y_{t

src/layers/gating.py:92

↓ 1 callersClassstat

tools/WORLD_v2/test/synth.cpp:146

↓ 1 callersClassstat

tools/WORLD/test/synth.cpp:298

ClassAcousticBase

src/frontend/acoustic_base.py:45

ClassContextInfo

misc/scripts/hybrid_voice/processHybridInfo.py:9

ClassContextLstm

This class implements the standard LSTM block, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/recurrent_decoders.py:593

ClassContextRNNDecoder

This class implements a standard recurrent neural network decoder: h_{t} = f(W^{hx}x_{t} + W^{hh}h_{t-1}+ W^{yh}y_{t-1} + b_{h}) y_{t

src/layers/recurrent_decoders.py:124

ClassDistortionComputation

src/utils/compute_distortion.py:46

ClassEncoder_Decoder_Models

src/tensorflow_lib/model.py:202

ClassExpDecreaseLearningRate

src/utils/learn_rates.py:165

ClassFeatureNormBase

to normalise feature into specific range to de-normalise feature back support min-max norm, MVN, this is a genetic class

src/frontend/feature_normalisation_base.py:46

ClassGeneralLayer

src/layers/mdn_layers.py:162

ClassHiddenLayer

src/layers/mdn_layers.py:228

ClassLabelNormalisation

src/frontend/label_normalisation.py:14

ClassLearningMinLrate

src/utils/learn_rates.py:122

ClassLearningRate

src/utils/learn_rates.py:43

ClassLearningRateConstant

src/utils/learn_rates.py:54

ClassLearningRateExpDecay

src/utils/learn_rates.py:76

ClassLinearLayer

src/layers/mdn_layers.py:79

ClassLinguisticBase

src/frontend/linguistic_base.py:48

ClassLoggerPlotter

Based on the built-in logging class, with added capabilities including plotting

src/logplot/logging_plotting.py:210

This class provides as a base for all long short-term memory (LSTM) related classes. Several variants of LSTM were investigated in (Wu & King, IC

src/layers/gating.py:194

This class provides as a base for all long short-term memory (LSTM) related classes. Several variants of LSTM were investigated in (Wu & King, IC

src/layers/recurrent_decoders.py:463

ClassLstmBase_LHUC

Very similar to the LSTM layer in the gating file Extra parameters are 'C' for scaling the hidden value

src/layers/lhuc_layer.py:71

ClassLstmDecoderBase

This class provides as a base for all long short-term memory (LSTM) related classes. Several variants of LSTM were investigated in (Wu & King, IC

src/layers/gating.py:323

ClassLstmDecoderBase

This class provides as a base for all long short-term memory (LSTM) related classes. Several variants of LSTM were investigated in (Wu & King, IC

src/layers/recurrent_decoders.py:236

ClassMixtureDensityOutputLayer

src/layers/mdn_layers.py:50

ClassMultipleSeriesPlot

src/logplot/logging_plotting.py:143

ClassRecurrentOutput

src/layers/gating.py:858

ClassSigmoidLayer

src/layers/mdn_layers.py:119

ClassSimplifiedLstmDecoder

This class implements a simplified LSTM block which only keeps the forget gate, inheriting the genetic class :class:`layers.gating.LstmBase`.

src/layers/recurrent_decoders.py:421

ClassSingleWeightMatrixPlot

src/logplot/logging_plotting.py:187

ClassSplitHiddenLayer

The nin x nout matrix is vertically split into 2 portions which can be updated independently. n_in1 -- by convention, use this part for

src/layers/layers.py:330

ClassTensorflowModels

src/tensorflow_lib/model.py:53

next →1–100 of 119, ranked by callers