| Popis: |
This research examined voice recognition performance as a function of time and showed no decrement in performance after 21 weeks. In addition, vocabulary sizes up to 240 utterances showed stable performance. Two people also combined their voice reference patterns and were then able to achieve an error rate of less than 2% when either person spoke to the speaker-dependent voice recognition unit. (Author) sponsored by Mr. Frank Deckelman, NAVELEX, Code 330, The work was performed by the author at the Naval Postgraduate School http://archive.org/details/longitudinalstud00pooc N0003981WROX008 NA |
