Modelling the global acoustic correlates of expressivity for Chinese text-to-speech synthesis | IEEE Conference Publication | IEEE Xplore