If your application requires users to train speaker-dependent speech commands, write your application to access the training function. Use the interface IVCmdUserWord to enable user training for speech commands.
Call IVCmdUserWord::ModifyTraining to delete a speaker-dependent template or to enable an application to specify templates to enable for a particular phrase.
The speaker-independent and speaker-dependent templates operate independently of each other. If your application enables both templates for a word or phrase, you may achieve better recognition accuracy.
The IVCmdUserWord::AddSIFile and IVCmdUserWord::RemoveSIFile functions are called by the APCLoad to add or remove application-specific, speaker-independent templates. For more information on APCLoad, see Preparing an Application for Distribution.