Text-to-speech (TTS) systems are is used to transform written text in a given language into its spoken equivalent. That is, a TTS system converts an input sentence, which is usually a sequence of letters (as in English) or symbols (as in Chinese), into machine-generated speech. The input sentence may include details of linguistic representation such as phrasing, intonation, and stress. Unlike limited vocabulary systems, which produce only a small predefined set of possible utterances, open-domain TTS systems must be able to intelligently handle any input text with an unlimited set of vocabularies.