Speech synthesis method, device and system and storage medium

The embodiment of the invention provides a speech synthesis method, device and system and a storage medium. The method comprises the steps that a text analysis model is utilized to analyze a to-be-processed text, and the to-be-processed text is converted into a first editing text which contains one or more control elements; a first speech corresponding to the first editing text is generated; an editing instruction from a user for editing the to-be-processed text is received; the control elements in the first editing text are modified according to the editing instruction to generate a second editing text; a second speech corresponding to the second editing text is generated; a confirmation instruction from the user on the second speech is received; the to-be-processed text and the second editing text are used as training samples to train the text analysis model; and the trained text analysis model is utilized to perform subsequent speech synthesis. According to the embodiment, a speechwhich better meets the de