Multilingual acoustic models for speech recognition and synthesis | IEEE Conference Publication | IEEE Xplore