Vocabulary independent acoustic-phonetic modeling for continuous speech recognition | IEEE Conference Publication | IEEE Xplore