Multipopulation genetic learning of midsagittal articulatory models for speech synthesis | IEEE Conference Publication | IEEE Xplore