Generation of nasalized speech sounds based on branched tube models obtained from separate mouth and nose outputs | IEEE Conference Publication | IEEE Xplore