Method and device for forecasting duration of speech synthesis unit

Abstract

The invention provides a method and device for forecasting duration of a speech synthesis unit. The method comprises the steps of: aiming at context environmental parameters, carrying out initial forecasting on the duration of the speech synthesis unit by utilizing a stepwise linear regression duration forecasting model so as to obtain an initial duration forecasting result; and distributing the initial duration forecasting result by utilizing a decision tree-Gaussian mixture model so as to obtain a distributed duration forecasting result. According to the method and device which are providedby the invention, the accuracy of the duration forecasting result can be increased to ensure that a speech synthesized in a speech synthesis system has a real sense of rhythm.

Description

Technical field

[0001] The present invention relates to the technical field of information processing, in particular to a method and device for training a duration prediction model of stepwise linear regression, and a method and device for predicting the duration of a speech synthesis unit. Background technique

[0002] In a speech synthesis system (Text-to-Speech, TTS), the prediction and generation of the duration of a speech synthesis unit is an indispensable step, which plays a vital role in the prosody hearing of synthesized speech.

[0003] According to the theories of phonetics and phonology, the duration and other characteristics of the speech synthesis unit are determined by its context. The prediction of speech duration is essentially a mapping from the value space of the context environment parameter to the duration value space. For the analysis and modeling method of this kind of mapping relationship, the existing time length prediction method usually adopts the decisi...

Images