This class is essentially an array of DecodingHMM instances, representing the HMM's for each pronunciation we can recognise.
Public Members
-
Uses the phoneme DecodingHMM instances in 'phone_models_',
-
and the phonetic transcription info for the pronunciations
-
as defined in 'lex_info_' and creates a complete DecodingHMM
-
Extracts LexiconInfo pointer from 'speech_hmm' then proceeds
-
Internal function.
-
Returns the number of states in a particular model including
-
Calculates the emission probability for a state in a particular
-
word model using the current input vector.
-
Returns the number of successor states, the successor states
-
themselves and the associated log transition probabilities
Documentation
This class is essentially an array of DecodingHMM instances,
representing the HMM's for each pronunciation we can recognise.
The ordering of this array conforms to the ordering the
LexInfo instance that is passed as a parameter to the
constructor.
Uses the phoneme DecodingHMM instances in 'phone_models_',
- Uses the phoneme DecodingHMM instances in 'phone_models_',
and the phonetic transcription info for the pronunciations
- and the phonetic transcription info for the pronunciations
as defined in 'lex_info_' and creates a complete DecodingHMM
- as defined in 'lex_info_' and creates a complete DecodingHMM
Extracts LexiconInfo pointer from 'speech_hmm' then proceeds
- Extracts LexiconInfo pointer from 'speech_hmm' then proceeds
Internal function.
- Internal function. Both constructors call this to create the
Returns the number of states in a particular model including
- Returns the number of states in a particular model including
Calculates the emission probability for a state in a particular
- Calculates the emission probability for a state in a particular
word model using the current input vector.
- word model using the current input vector. Only checks for
Returns the number of successor states, the successor states
- Returns the number of successor states, the successor states
themselves and the associated log transition probabilities
- themselves and the associated log transition probabilities
- This class has no child classes.
- Author:
- Darren Moore (moore@idiap.ch)
Alphabetic index HTML hierarchy of classes or Java
This page was generated with the help of DOC++.