A new model-based speech analysis/Synthesis system | IEEE Conference Publication | IEEE Xplore