Currently sufficient to create features for speech recognition
Joe Woelfel
February 25, 2010 6:01pm
無