Speaker-independent large-vocabulary continuous speech recognition is now in use in a closed-captioned broadcasting service for certain NHK news, music, sports, and lifestyle programming. The system ...