SingleUtteranceNnet3DecoderTpl< FST > Class Template Reference

You will instantiate this class when you want to decode a single utterance using the online-decoding setup for neural nets. More...

#include <online-nnet3-decoding.h>

Collaboration diagram for SingleUtteranceNnet3DecoderTpl< FST >:

Public Member Functions
	SingleUtteranceNnet3DecoderTpl (const LatticeFasterDecoderConfig &decoder_opts, const TransitionModel &trans_model, const nnet3::DecodableNnetSimpleLoopedInfo &info, const FST &fst, OnlineNnet2FeaturePipeline *features)
void	InitDecoding (int32 frame_offset=0)
	Initializes the decoding and sets the frame offset of the underlying decodable object. More...
void	AdvanceDecoding ()
	Advances the decoding as far as we can. More...
void	FinalizeDecoding ()
	Finalizes the decoding. More...
int32	NumFramesDecoded () const
void	GetLattice (bool end_of_utterance, CompactLattice *clat) const
	Gets the lattice. More...
void	GetBestPath (bool end_of_utterance, Lattice *best_path) const
	Outputs an FST corresponding to the single best path through the current lattice. More...
bool	EndpointDetected (const OnlineEndpointConfig &config)
	This function calls EndpointDetected from online-endpoint.h, with the required arguments. More...
const LatticeFasterOnlineDecoderTpl< FST > &	Decoder () const
	~SingleUtteranceNnet3DecoderTpl ()

Private Attributes
const LatticeFasterDecoderConfig &	decoder_opts_
BaseFloat	input_feature_frame_shift_in_seconds_
const TransitionModel &	trans_model_
nnet3::DecodableAmNnetLoopedOnline	decodable_
LatticeFasterOnlineDecoderTpl< FST >	decoder_

Detailed Description

template<typename FST>
class kaldi::SingleUtteranceNnet3DecoderTpl< FST >

You will instantiate this class when you want to decode a single utterance using the online-decoding setup for neural nets.

The template will be instantiated only for FST = fst::Fst<fst::StdArc> and FST = fst::GrammarFst.

Definition at line 52 of file online-nnet3-decoding.h.

Constructor & Destructor Documentation

SingleUtteranceNnet3DecoderTpl()

SingleUtteranceNnet3DecoderTpl	(	const LatticeFasterDecoderConfig &	decoder_opts,
		const TransitionModel &	trans_model,
		const nnet3::DecodableNnetSimpleLoopedInfo &	info,
		const FST &	fst,
		OnlineNnet2FeaturePipeline *	features
	)

Definition at line 29 of file online-nnet3-decoding.cc.

References SingleUtteranceNnet3DecoderTpl< FST >::decoder_.

                                         :
    decoder_opts_(decoder_opts),
    input_feature_frame_shift_in_seconds_(features->FrameShiftInSeconds()),
    trans_model_(trans_model),
    decodable_(trans_model_, info,
               features->InputFeature(), features->IvectorFeature()),
    decoder_(fst, decoder_opts_) {
  decoder_.InitDecoding();
}

~SingleUtteranceNnet3DecoderTpl()

~SingleUtteranceNnet3DecoderTpl ( )

inline

Definition at line 101 of file online-nnet3-decoding.h.

{ }

Member Function Documentation

AdvanceDecoding()

void AdvanceDecoding

(

)

Advances the decoding as far as we can.

Definition at line 51 of file online-nnet3-decoding.cc.

References SingleUtteranceNnet3DecoderTpl< FST >::decodable_, and SingleUtteranceNnet3DecoderTpl< FST >::decoder_.

Referenced by main().

                                                          {
  decoder_.AdvanceDecoding(&decodable_);
}

Decoder()

const LatticeFasterOnlineDecoderTpl<FST>& Decoder ( ) const

inline

Definition at line 99 of file online-nnet3-decoding.h.

References SingleUtteranceNnet3DecoderTpl< FST >::decoder_.

Referenced by main().

{ return decoder_; }

EndpointDetected()

bool EndpointDetected

(

const OnlineEndpointConfig &

config

)

This function calls EndpointDetected from online-endpoint.h, with the required arguments.

Definition at line 88 of file online-nnet3-decoding.cc.

References SingleUtteranceNnet3DecoderTpl< FST >::decodable_, SingleUtteranceNnet3DecoderTpl< FST >::decoder_, kaldi::EndpointDetected(), DecodableNnetLoopedOnlineBase::FrameSubsamplingFactor(), SingleUtteranceNnet3DecoderTpl< FST >::input_feature_frame_shift_in_seconds_, and SingleUtteranceNnet3DecoderTpl< FST >::trans_model_.

Referenced by main().

                                        {
  BaseFloat output_frame_shift =
      input_feature_frame_shift_in_seconds_ *
      decodable_.FrameSubsamplingFactor();
  return kaldi::EndpointDetected(config, trans_model_,
                                 output_frame_shift, decoder_);
}

FinalizeDecoding()

void FinalizeDecoding

(

)

Finalizes the decoding.

Cleans up and prunes remaining tokens, so the GetLattice() call will return faster. You must not call this before calling (TerminateDecoding() or InputIsFinished()) and then Wait().

Definition at line 56 of file online-nnet3-decoding.cc.

References SingleUtteranceNnet3DecoderTpl< FST >::decoder_.

Referenced by main().

                                                           {
  decoder_.FinalizeDecoding();
}

GetBestPath()

void GetBestPath	(	bool	end_of_utterance,
		Lattice *	best_path
	)		const

Outputs an FST corresponding to the single best path through the current lattice.

If "use_final_probs" is true AND we reached the final-state of the graph then it will include those as final-probs, else it will treat all final-probs as one.

Definition at line 82 of file online-nnet3-decoding.cc.

References SingleUtteranceNnet3DecoderTpl< FST >::decoder_.

Referenced by main().

                                                                        {
  decoder_.GetBestPath(best_path, end_of_utterance);
}

GetLattice()

void GetLattice	(	bool	end_of_utterance,
		CompactLattice *	clat
	)		const

Gets the lattice.

The output lattice has any acoustic scaling in it (which will typically be desirable in an online-decoding context); if you want an un-scaled lattice, scale it using ScaleLattice() with the inverse of the acoustic weight. "end_of_utterance" will be true if you want the final-probs to be included.

Definition at line 66 of file online-nnet3-decoding.cc.

References SingleUtteranceNnet3DecoderTpl< FST >::decoder_, SingleUtteranceNnet3DecoderTpl< FST >::decoder_opts_, LatticeFasterDecoderConfig::det_opts, LatticeFasterDecoderConfig::determinize_lattice, fst::DeterminizeLatticePhonePrunedWrapper(), KALDI_ERR, LatticeFasterDecoderConfig::lattice_beam, SingleUtteranceNnet3DecoderTpl< FST >::NumFramesDecoded(), and SingleUtteranceNnet3DecoderTpl< FST >::trans_model_.

Referenced by main().

                                                                         {
  if (NumFramesDecoded() == 0)
    KALDI_ERR << "You cannot get a lattice if you decoded no frames.";
  Lattice raw_lat;
  decoder_.GetRawLattice(&raw_lat, end_of_utterance);

  if (!decoder_opts_.determinize_lattice)
    KALDI_ERR << "--determinize-lattice=false option is not supported at the moment";

  BaseFloat lat_beam = decoder_opts_.lattice_beam;
  DeterminizeLatticePhonePrunedWrapper(
      trans_model_, &raw_lat, lat_beam, clat, decoder_opts_.det_opts);
}

InitDecoding()

void InitDecoding

(

int32

frame_offset = 0

)

Initializes the decoding and sets the frame offset of the underlying decodable object.

This method is called by the constructor. You can also call this method when you want to reset the decoder state, but want to keep using the same decodable object, e.g. in case of an endpoint.

Definition at line 45 of file online-nnet3-decoding.cc.

References SingleUtteranceNnet3DecoderTpl< FST >::decodable_, SingleUtteranceNnet3DecoderTpl< FST >::decoder_, and DecodableNnetLoopedOnlineBase::SetFrameOffset().

Referenced by main().

                                                                         {
  decoder_.InitDecoding();
  decodable_.SetFrameOffset(frame_offset);
}

NumFramesDecoded()

int32 NumFramesDecoded

(

)

const

Definition at line 61 of file online-nnet3-decoding.cc.

References SingleUtteranceNnet3DecoderTpl< FST >::decoder_.

Referenced by SingleUtteranceNnet3DecoderTpl< FST >::GetLattice(), and main().

                                                                  {
  return decoder_.NumFramesDecoded();
}

Member Data Documentation

decodable_

nnet3::DecodableAmNnetLoopedOnline decodable_

private

Definition at line 114 of file online-nnet3-decoding.h.

Referenced by SingleUtteranceNnet3DecoderTpl< FST >::AdvanceDecoding(), SingleUtteranceNnet3DecoderTpl< FST >::EndpointDetected(), and SingleUtteranceNnet3DecoderTpl< FST >::InitDecoding().

decoder_

LatticeFasterOnlineDecoderTpl<FST> decoder_

private

Definition at line 116 of file online-nnet3-decoding.h.

Referenced by SingleUtteranceNnet3DecoderTpl< FST >::AdvanceDecoding(), SingleUtteranceNnet3DecoderTpl< FST >::Decoder(), SingleUtteranceNnet3DecoderTpl< FST >::EndpointDetected(), SingleUtteranceNnet3DecoderTpl< FST >::FinalizeDecoding(), SingleUtteranceNnet3DecoderTpl< FST >::GetBestPath(), SingleUtteranceNnet3DecoderTpl< FST >::GetLattice(), SingleUtteranceNnet3DecoderTpl< FST >::InitDecoding(), SingleUtteranceNnet3DecoderTpl< FST >::NumFramesDecoded(), and SingleUtteranceNnet3DecoderTpl< FST >::SingleUtteranceNnet3DecoderTpl().

decoder_opts_

const LatticeFasterDecoderConfig& decoder_opts_

private

Definition at line 104 of file online-nnet3-decoding.h.

Referenced by SingleUtteranceNnet3DecoderTpl< FST >::GetLattice().

input_feature_frame_shift_in_seconds_

BaseFloat input_feature_frame_shift_in_seconds_

private

Definition at line 108 of file online-nnet3-decoding.h.

Referenced by SingleUtteranceNnet3DecoderTpl< FST >::EndpointDetected().

trans_model_

const TransitionModel& trans_model_

private

Definition at line 112 of file online-nnet3-decoding.h.

Referenced by SingleUtteranceNnet3DecoderTpl< FST >::EndpointDetected(), and SingleUtteranceNnet3DecoderTpl< FST >::GetLattice().

---

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

• online2/online-nnet3-decoding.h
• online2/online-nnet3-decoding.cc