edu.cmu.sphinx.linguist.g2p

Class G2PConverter

java.lang.Object

edu.cmu.sphinx.linguist.g2p.G2PConverter

public class G2PConverter
extends java.lang.Object

The grapheme-to-phoneme (g2p) decoder

Constructor Summary

Constructors

Constructor and Description
G2PConverter(java.lang.String g2pmodel_file) Create a decoder by loading the serialized model from a specified filename
G2PConverter(java.net.URL g2pModelUrl) Create a decoder by loading the serialized model from a specified URL

Method Summary

Modifier and Type	Method and Description
java.util.ArrayList<Path>	phoneticize(java.util.ArrayList<java.lang.String> entry, int nbest) Phoneticize a word
java.util.ArrayList<Path>	phoneticize(java.lang.String word, int nbest) Phoneticize a word

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

G2PConverter

public G2PConverter(java.net.URL g2pModelUrl)
             throws java.io.IOException

Create a decoder by loading the serialized model from a specified URL

Parameters:

g2pModelUrl - the URL of the serialized model

Throws:

java.io.IOException - when model is not loaded

G2PConverter

public G2PConverter(java.lang.String g2pmodel_file)

Create a decoder by loading the serialized model from a specified filename

Parameters:

g2pmodel_file - the filename of the serialized model

Method Detail

phoneticize

public java.util.ArrayList<Path> phoneticize(java.util.ArrayList<java.lang.String> entry,
                                             int nbest)

Phoneticize a word

Parameters:

entry - the word to phoneticize transformed to an ArrayList of Strings (each element hold a single character)

nbest - the number of distinct pronunciations to return

Returns:

the pronunciation(s) of the input word

phoneticize

public java.util.ArrayList<Path> phoneticize(java.lang.String word,
                                             int nbest)

Phoneticize a word

Parameters:

word - the word to phoneticize

nbest - the number of distinct pronunciations to return

Returns:

the pronunciation(s) of the input word