HTKLoader (sphinx4-core 5prealpha-SNAPSHOT API)

java.lang.Object
- edu.cmu.sphinx.linguist.acoustic.tiedstate.HTKLoader

All Implemented Interfaces:

Loader, Configurable
```
public class HTKLoader
extends java.lang.Object
implements Loader
```
Remark1: S4 does not use HMM tying: the HTK "tiedlist" is not loaded nor user for now Remark2: HTK does nearly never backoff to monophones, whereas S4 might do it. Hence, all 1ph must be present, which is not always the case in HTK models. In this file, we have simply tied all 1phs to the first 3ph found with the same base phone. This is clearly not the best option, this may impact the performances. A better alternative should be proposed. Remark3: HTK is case-sensitive, while S4 requires upper-case names. A good option would be to modify S4 to make it case-sensitive, but in this version, I have rather built a class "NamesConversion" to transform the models, lexicons and grammar. Remark4: HTK does not have separate filler file. It should be quite easy to define the fillers in the configuration of HTKLoader, but I have not yet done it. For now, a separate filler file must be built. Remark5: The HTK/Julius lexicon has different fields for the word name and the printed output. We simply delete and don't use this field ("[...]") in this version. This modify the way scoring must be done. WARNING ! Remark6: Diphones are not allowed in S4 (?), and when they occur in the HTK MMF, the 2ph is transformed into a 3ph with "SIL" context. When both the transformed 2ph and the 3ph coexists in the MMF file, only the first one is kept (!) Remark7: HTKloader does not support yet all the HTK format. However it should work for standard 3ph models (it supports state/transition tying).

Field Summary

Fields
Modifier and Type	Field and Description
`protected static int`	`BYTE_ORDER_MAGIC`
`protected static int`	`CONTEXT_SIZE`
`protected static java.lang.String`	`DENSITY_FILE_VERSION`
`protected static java.lang.String`	`FILLER`
`protected static java.lang.String`	`MIXW_FILE_VERSION`
`static java.lang.String`	`MODEL_VERSION` Supports this version of the acoustic model
`static java.lang.String`	`PROP_IS_BINARY` Specifies whether the model to be loaded is in ASCII or binary format
`static java.lang.String`	`PROP_MC_FLOOR` Mixture component score floor.
`static java.lang.String`	`PROP_MODEL` The name of the model definition file (contains the HMM data)
`static java.lang.String`	`PROP_MW_FLOOR` Mixture weight floor.
`static java.lang.String`	`PROP_PROPERTIES_FILE` The property for the name of the acoustic properties file.
`static java.lang.String`	`PROP_TIE_1PH` Shall we use the real 1ph or tie them from the 3ph ?
`static java.lang.String`	`PROP_UNIT_MANAGER` The unit manager
`static java.lang.String`	`PROP_VARIANCE_FLOOR` Variance floor.
`static java.lang.String`	`PROP_VECTOR_LENGTH` The property for the length of feature vectors.
`protected static java.lang.String`	`SILENCE_CIPHONE`
`protected static java.lang.String`	`TMAT_FILE_VERSION`

Constructor Summary

Constructors
Constructor and Description
`HTKLoader()`
`HTKLoader(java.lang.String propsFile, UnitManager unitManager, boolean isBinary, int vectorLength, java.lang.String model, boolean tie1ph, float distFloor, float mixtureWeightFloor, float varianceFloor)`

Method Summary

All Methods Instance Methods Concrete Methods
Modifier and Type	Method and Description
`java.util.Map<java.lang.String,Unit>`	`getContextIndependentUnits()` Returns the map of context indepent units.
`HMMManager`	`getHMMManager()` Returns the HMM Manager for this loader.
`int`	`getLeftContextSize()` Returns the size of the left context for context dependent units.
`protected java.lang.String`	`getLocation()` Return the location.
`Pool<float[]>`	`getMeansPool()` Gets the pool of means for this loader.
`Pool<float[][]>`	`getMeansTransformationMatrixPool()` Gets the means transformation matrix pool.
`Pool<float[]>`	`getMeansTransformationVectorPool()` Gets the means transformation vectors pool.
`GaussianWeights`	`getMixtureWeights()` Gets the mixture weight pool.
`java.lang.String`	`getName()`
`java.util.Properties`	`getProperties()` Return the acoustic model properties.
`int`	`getRightContextSize()` Returns the size of the right context for context dependent units.
`Pool<Senone>`	`getSenonePool()` Gets the senone pool for this loader.
`protected SenoneSequence`	`getSenoneSequence(int[] stateId)` Gets the senone sequence representing the given senones.
`float[][]`	`getTransformMatrix()` Gets the transformation matrix.
`Pool<float[][]>`	`getTransitionMatrixPool()` Gets the transition matrix pool.
`Pool<float[]>`	`getVariancePool()` Gets the variance pool.
`Pool<float[][]>`	`getVarianceTransformationMatrixPool()` Gets the variance transformation matrix pool.
`Pool<float[]>`	`getVarianceTransformationVectorPool()` Gets the variance transformation vectors pool.
`void`	`load()` Loads the acoustic model.
`protected void`	`loadHMMPool(boolean useCDUnits, edu.cmu.sphinx.linguist.acoustic.tiedstate.HTKLoader.HTKStruct htkModels, java.lang.String path)` Loads the model file.
`void`	`logInfo()` logs information about this loader
`void`	`newProperties(PropertySheet ps)` This method is called when this configurable component needs to be reconfigured.
`protected float`	`readFloat(java.io.DataInputStream dis)` Read a float from the input stream, byte-swapping as necessary. the input stream
`protected float[]`	`readFloatArray(java.io.DataInputStream dis, int size)` Reads the given number of floats from the stream and returns them in an array of floats.
`protected int`	`readInt(java.io.DataInputStream dis)` Read an integer from the input stream, byte-swapping as necessary.
`protected boolean`	`sameSenoneSequence(int[] ssid1, int[] ssid2)` Returns true if the given senone sequence IDs are the same.
`void`	`update(Transform transform, ClusteredDensityFileData clusters)` Apply the transform

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

- Field Detail
  - PROP_UNIT_MANAGER
```
@S4Component(type=UnitManager.class)
public static final java.lang.String PROP_UNIT_MANAGER
```
    The unit manager
    
    See Also:
    
    Constant Field Values
  - PROP_IS_BINARY
```
@S4Boolean(defaultValue=true)
public static final java.lang.String PROP_IS_BINARY
```
    Specifies whether the model to be loaded is in ASCII or binary format
    
    See Also:
    
    Constant Field Values
  - PROP_MODEL
```
@S4String(mandatory=true,
          defaultValue="hmmdefs")
public static final java.lang.String PROP_MODEL
```
    The name of the model definition file (contains the HMM data)
    
    See Also:
    
    Constant Field Values
  - PROP_TIE_1PH
```
@S4Boolean(defaultValue=true)
public static final java.lang.String PROP_TIE_1PH
```
    Shall we use the real 1ph or tie them from the 3ph ?
    
    See Also:
    
    Constant Field Values
  - PROP_PROPERTIES_FILE
```
@S4String(defaultValue="model.props")
public static final java.lang.String PROP_PROPERTIES_FILE
```
    The property for the name of the acoustic properties file.
    
    See Also:
    
    Constant Field Values
  - PROP_VECTOR_LENGTH
```
@S4Integer(defaultValue=39)
public static final java.lang.String PROP_VECTOR_LENGTH
```
    The property for the length of feature vectors.
    
    See Also:
    
    Constant Field Values
  - PROP_MC_FLOOR
```
@S4Double(defaultValue=0.0)
public static final java.lang.String PROP_MC_FLOOR
```
    Mixture component score floor.
    
    See Also:
    
    Constant Field Values
  - PROP_VARIANCE_FLOOR
```
@S4Double(defaultValue=9.999999747378752E-5)
public static final java.lang.String PROP_VARIANCE_FLOOR
```
    Variance floor.
    
    See Also:
    
    Constant Field Values
  - PROP_MW_FLOOR
```
@S4Double(defaultValue=1.0000000116860974E-7)
public static final java.lang.String PROP_MW_FLOOR
```
    Mixture weight floor.
    
    See Also:
    
    Constant Field Values
  - FILLER
```
protected static final java.lang.String FILLER
```
    See Also:
    
    Constant Field Values
  - SILENCE_CIPHONE
```
protected static final java.lang.String SILENCE_CIPHONE
```
    See Also:
    
    Constant Field Values
  - BYTE_ORDER_MAGIC
```
protected static final int BYTE_ORDER_MAGIC
```
    See Also:
    
    Constant Field Values
  - MODEL_VERSION
```
public static final java.lang.String MODEL_VERSION
```
    Supports this version of the acoustic model
    
    See Also:
    
    Constant Field Values
  - CONTEXT_SIZE
```
protected static final int CONTEXT_SIZE
```
    See Also:
    
    Constant Field Values
  - DENSITY_FILE_VERSION
```
protected static final java.lang.String DENSITY_FILE_VERSION
```
    See Also:
    
    Constant Field Values
  - MIXW_FILE_VERSION
```
protected static final java.lang.String MIXW_FILE_VERSION
```
    See Also:
    
    Constant Field Values
  - TMAT_FILE_VERSION
```
protected static final java.lang.String TMAT_FILE_VERSION
```
    See Also:
    
    Constant Field Values
- Constructor Detail
  - HTKLoader
```
public HTKLoader(java.lang.String propsFile,
                 UnitManager unitManager,
                 boolean isBinary,
                 int vectorLength,
                 java.lang.String model,
                 boolean tie1ph,
                 float distFloor,
                 float mixtureWeightFloor,
                 float varianceFloor)
```
  - HTKLoader
```
public HTKLoader()
```
- Method Detail
  - newProperties
```
public void newProperties(PropertySheet ps)
                   throws PropertyException
```
    Description copied from interface: Configurable
    
    This method is called when this configurable component needs to be reconfigured.
    
    Specified by:
    
    newProperties in interface Configurable
    
    Parameters:
    
    ps - a property sheet holding the new data
    
    Throws:
    
    PropertyException - if there is a problem with the properties.
  - load
```
public void load()
          throws java.io.IOException
```
    Description copied from interface: Loader
    
    Loads the acoustic model.
    
    Specified by:
    
    load in interface Loader
    
    Throws:
    
    java.io.IOException - if an error occurs while loading the model
  - getName
```
public java.lang.String getName()
```
  - getProperties
```
public java.util.Properties getProperties()
```
    Return the acoustic model properties.
    
    Specified by:
    
    getProperties in interface Loader
    
    Returns:
    
    the acoustic model properties
  - getLocation
```
protected java.lang.String getLocation()
```
    Return the location.
    
    Returns:
    
    the location
  - getContextIndependentUnits
```
public java.util.Map<java.lang.String,Unit> getContextIndependentUnits()
```
    Description copied from interface: Loader
    
    Returns the map of context indepent units. The map can be accessed by unit name.
    
    Specified by:
    
    getContextIndependentUnits in interface Loader
    
    Returns:
    
    the map of context independent units
  - readInt
```
protected int readInt(java.io.DataInputStream dis)
               throws java.io.IOException
```
    Read an integer from the input stream, byte-swapping as necessary.
    
    Parameters:
    
    dis - the input stream
    
    Returns:
    
    an integer value
    
    Throws:
    
    java.io.IOException - on error
  - readFloat
```
protected float readFloat(java.io.DataInputStream dis)
                   throws java.io.IOException
```
    Read a float from the input stream, byte-swapping as necessary. the input stream
    
    Parameters:
    
    dis - the input stream
    
    Returns:
    
    a floating value
    
    Throws:
    
    java.io.IOException - on error
  - readFloatArray
```
protected float[] readFloatArray(java.io.DataInputStream dis,
                                 int size)
                          throws java.io.IOException
```
    Reads the given number of floats from the stream and returns them in an array of floats.
    
    Parameters:
    
    dis - the stream to read data from the number of floats to read
    
    size - size of the array
    
    Returns:
    
    an array of size float elements
    
    Throws:
    
    java.io.IOException - if an exception occurs
  - loadHMMPool
```
protected void loadHMMPool(boolean useCDUnits,
                           edu.cmu.sphinx.linguist.acoustic.tiedstate.HTKLoader.HTKStruct htkModels,
                           java.lang.String path)
                    throws java.io.IOException
```
    Loads the model file. A set of density arrays are created and placed in the given pool.
    
    Parameters:
    
    useCDUnits - if true, loads also the context dependent units
    
    htkModels - the set of HTK models
    
    path - the path to a density file
    
    Throws:
    
    java.io.IOException - if an error occurs while loading the data
  - sameSenoneSequence
```
protected boolean sameSenoneSequence(int[] ssid1,
                                     int[] ssid2)
```
    Returns true if the given senone sequence IDs are the same.
    
    Parameters:
    
    ssid1 - first sequence of senones
    
    ssid2 - second sequence of senones
    
    Returns:
    
    true if the given senone sequence IDs are the same, false otherwise
  - getSenoneSequence
```
protected SenoneSequence getSenoneSequence(int[] stateId)
```
    Gets the senone sequence representing the given senones.
    
    Parameters:
    
    stateId - is the array of senone state ids
    
    Returns:
    
    the senone sequence associated with the states
  - getMeansPool
```
public Pool<float[]> getMeansPool()
```
    Description copied from interface: Loader
    
    Gets the pool of means for this loader.
    
    Specified by:
    
    getMeansPool in interface Loader
    
    Returns:
    
    the pool
  - getMeansTransformationMatrixPool
```
public Pool<float[][]> getMeansTransformationMatrixPool()
```
    Description copied from interface: Loader
    
    Gets the means transformation matrix pool.
    
    Specified by:
    
    getMeansTransformationMatrixPool in interface Loader
    
    Returns:
    
    the pool
  - getMeansTransformationVectorPool
```
public Pool<float[]> getMeansTransformationVectorPool()
```
    Description copied from interface: Loader
    
    Gets the means transformation vectors pool.
    
    Specified by:
    
    getMeansTransformationVectorPool in interface Loader
    
    Returns:
    
    the pool
  - getVariancePool
```
public Pool<float[]> getVariancePool()
```
    Description copied from interface: Loader
    
    Gets the variance pool.
    
    Specified by:
    
    getVariancePool in interface Loader
    
    Returns:
    
    the pool
  - getVarianceTransformationMatrixPool
```
public Pool<float[][]> getVarianceTransformationMatrixPool()
```
    Description copied from interface: Loader
    
    Gets the variance transformation matrix pool.
    
    Specified by:
    
    getVarianceTransformationMatrixPool in interface Loader
    
    Returns:
    
    the pool
  - getVarianceTransformationVectorPool
```
public Pool<float[]> getVarianceTransformationVectorPool()
```
    Description copied from interface: Loader
    
    Gets the variance transformation vectors pool.
    
    Specified by:
    
    getVarianceTransformationVectorPool in interface Loader
    
    Returns:
    
    the pool
  - getMixtureWeights
```
public GaussianWeights getMixtureWeights()
```
    Description copied from interface: Loader
    
    Gets the mixture weight pool.
    
    Specified by:
    
    getMixtureWeights in interface Loader
    
    Returns:
    
    the pool
  - getTransitionMatrixPool
```
public Pool<float[][]> getTransitionMatrixPool()
```
    Description copied from interface: Loader
    
    Gets the transition matrix pool.
    
    Specified by:
    
    getTransitionMatrixPool in interface Loader
    
    Returns:
    
    the pool
  - getTransformMatrix
```
public float[][] getTransformMatrix()
```
    Description copied from interface: Loader
    
    Gets the transformation matrix.
    
    Specified by:
    
    getTransformMatrix in interface Loader
    
    Returns:
    
    the matrix
  - getSenonePool
```
public Pool<Senone> getSenonePool()
```
    Description copied from interface: Loader
    
    Gets the senone pool for this loader.
    
    Specified by:
    
    getSenonePool in interface Loader
    
    Returns:
    
    the pool
  - getLeftContextSize
```
public int getLeftContextSize()
```
    Description copied from interface: Loader
    
    Returns the size of the left context for context dependent units.
    
    Specified by:
    
    getLeftContextSize in interface Loader
    
    Returns:
    
    the left context size
  - getRightContextSize
```
public int getRightContextSize()
```
    Description copied from interface: Loader
    
    Returns the size of the right context for context dependent units.
    
    Specified by:
    
    getRightContextSize in interface Loader
    
    Returns:
    
    the left context size
  - getHMMManager
```
public HMMManager getHMMManager()
```
    Description copied from interface: Loader
    
    Returns the HMM Manager for this loader.
    
    Specified by:
    
    getHMMManager in interface Loader
    
    Returns:
    
    the HMM Manager
  - logInfo
```
public void logInfo()
```
    Description copied from interface: Loader
    
    logs information about this loader
    
    Specified by:
    
    logInfo in interface Loader
  - update
```
public void update(Transform transform,
                   ClusteredDensityFileData clusters)
```
    Description copied from interface: Loader
    
    Apply the transform
    
    Specified by:
    
    update in interface Loader
    
    Parameters:
    
    transform - transform to apply to the model
    
    clusters - transform clusters

Class HTKLoader

Field Summary

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Field Detail

PROP_UNIT_MANAGER

PROP_IS_BINARY

PROP_MODEL

PROP_TIE_1PH

PROP_PROPERTIES_FILE

PROP_VECTOR_LENGTH

PROP_MC_FLOOR

PROP_VARIANCE_FLOOR

PROP_MW_FLOOR

FILLER

SILENCE_CIPHONE

BYTE_ORDER_MAGIC

MODEL_VERSION

CONTEXT_SIZE

DENSITY_FILE_VERSION

MIXW_FILE_VERSION

TMAT_FILE_VERSION

Constructor Detail

HTKLoader

HTKLoader

Method Detail

newProperties

load

getName

getProperties

getLocation

getContextIndependentUnits

readInt

readFloat

readFloatArray

loadHMMPool

sameSenoneSequence

getSenoneSequence

getMeansPool

getMeansTransformationMatrixPool

getMeansTransformationVectorPool

getVariancePool

getVarianceTransformationMatrixPool

getVarianceTransformationVectorPool

getMixtureWeights

getTransitionMatrixPool

getTransformMatrix

getSenonePool

getLeftContextSize

getRightContextSize

getHMMManager

logInfo

update