The Lexicon button seems to be the only one which allows training
Last reviewed: 3/1/1999
PRB Article ID: P119802
The information in this article applies to:
- SpeechKit 2.0
On the Chant Speech Recognition engine selection dialog, the Lexicon button seems to be the only one that allows training of any kind. Although other training buttons are active, they do nothing. The Lexicon training method is far beyond the ability of our users. Do you have any suggestions as to how to train the engines in a more intuitive way?
The training is part of the engine features.
The ShowDialog (ShowSRDialog, ShowTTSDialog) offers a way to invoke an engine dialog for adding words and phrases to the speaker(user) personal lexicon. You will need to set this speaker (user) as the active speaker to take advantage of this training.
For continuous dictation, you must use the engine training facility to create a speaker profile. This is usually provided as a separate utility. For MS engines, you can use the training that comes with MS Dictate to create speaker profiles.