Post by Rick Price on Apr 26, 2006 11:30:04 GMT -5
Voice tags are an advanced feature of speaker-independent ASR. They are used to customize the phonetic transcriptions in word-spotting vocabularies. The recognizer creates additional phonetic information from a user's voice, which it uses along with a word's phonetic transcription. This is important when a word is pronounced different from what is allowed by the phonetic transcription. For example, voice tags can increase recognition if the speaker has a heavy accent.