Hello,
if i understand right the silent models in the SpjinxTrain Scripts have always the same size like the phonem or word models. Is this right ???
I ask because i try to build an acoustic WORD model for the aurora task and ask me if i have to put SILENT models into the transcription files for the trainer.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Anonymous
-
2006-03-23
The Silent Models are handled as special Words with a special count of States from the decoder ???
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello,
if i understand right the silent models in the SpjinxTrain Scripts have always the same size like the phonem or word models. Is this right ???
I ask because i try to build an acoustic WORD model for the aurora task and ask me if i have to put SILENT models into the transcription files for the trainer.
The Silent Models are handled as special Words with a special count of States from the decoder ???
Is this still a point of confusion?