Package saui_pr4 :: Module latticeui :: Class LatticeModel

Class LatticeModel

object --+
         |
        LatticeModel

Model of a word lattice obtained from speech recognition, along with a possibly user-specified correct hypothesis.

It provides alternate words with associated probabilities for visualization, and tracks the user's corrected hypothesis.

Instance Methods

[hide private]

__init__(self, dag, lm=None, lw=9.5, wip=0.5)
Create a LatticeModel object.

source code

set_lm(self, lm=None, lw=9.5, wip=0.5)
Set the language model and update the best hypothesis and posterior probabilities with the new model.

source code

set_dag(self, dag, lm=None, lw=9.5, wip=0.5)
Set the DAG and update the best hypothesis from its best path.

source code

list of LatticeWord

get_hyp(self)
Retrieve the best hypothesis from a LatticeModel

source code

set_hyp(self, hyp)
Set the best hypothesis

source code

string

get_hyp_text(self)
Retrieve the text of the top hypothesis

source code

LatticeCloud

get_cloud(self, word)
Retrieve a cloud around a hypothesis word.

source code

Inherited from object: __delattr__, __getattribute__, __hash__, __new__, __reduce__, __reduce_ex__, __repr__, __setattr__, __str__

Properties

[hide private]

Inherited from object: __class__

Method Details

[hide private]

init(self, dag, lm=None, lw=9.5, wip=0.5)
(Constructor)

source code

Create a LatticeModel object.

Parameters:

dag (sphinx.lattice.Dag) - Word lattice to initialize this object with.

Overrides: object.__init__

set_lm(self, lm=None, lw=9.5, wip=0.5)

source code

Set the language model and update the best hypothesis and posterior probabilities with the new model.

Parameters:

lm (sphinx.arpalm.ArpaLM (or equivalent)) - Language model to use in computation
lw (float) - Language model weight
wip (float) - Word insertion penalty

set_dag(self, dag, lm=None, lw=9.5, wip=0.5)

source code

Set the DAG and update the best hypothesis from its best path.

Parameters:

dag (sphinx.lattice.Dag (or equivalent)) - DAG to use.
lm (sphinx.arpalm.ArpaLM (or equivalent)) - Language model to use in computation of best path
lw (float) - Language model weight
wip (float) - Word insertion penalty

get_hyp(self)

source code

Retrieve the best hypothesis from a LatticeModel

Returns: list of LatticeWord: The top hypothesis, annotated with start and end frames and posterior probabilities

set_hyp(self, hyp)

source code

Set the best hypothesis

Parameters:

hyp (list of LatticeWord) - The new top hypothesis

get_hyp_text(self)

source code

Retrieve the text of the top hypothesis

Returns: string: The top hypothesis text

get_cloud(self, word)

source code

Retrieve a cloud around a hypothesis word.

Parameters:

word (LatticeWord) - Word in a hypothesis

Returns: LatticeCloud

Cloud around this word with beam width 0

Class LatticeModel

__init__(self, dag, lm=None, lw=9.5, wip=0.5) (Constructor)

set_lm(self, lm=None, lw=9.5, wip=0.5)

set_dag(self, dag, lm=None, lw=9.5, wip=0.5)

get_hyp(self)

set_hyp(self, hyp)

get_hyp_text(self)

get_cloud(self, word)

init(self, dag, lm=None, lw=9.5, wip=0.5)
(Constructor)