Acoustic model building, speech synthesis method, device, equipment and storage medium

Abstract

The embodiment of the present invention discloses an acoustic model establishment, speech synthesis method, device, equipment and storage medium, wherein the acoustic model establishment method comprises: obtaining phoneme sequence samples of a plurality of training samples from a corpus, and obtaining the phoneme sequence samples The context feature of each phoneme and the duration of each phoneme in the phoneme sequence sample; wherein, the childish phoneme in the phoneme sequence sample is split into two phonemes; Acoustic features are extracted from the training sample; with the phoneme sequence sample, The context feature and duration of each phoneme in the phoneme sequence sample are used as the input of the acoustic model, and the acoustic feature is used as the output of the acoustic model, and the acoustic model is trained to obtain a pre-trained acoustic model, which can make The modelling performance of Erhuayin is better, the synthesis of Erhuayin can be better realized, the Erhuayin that does not appear in the corpus can be synthesized, and the recording cost of the corpus in the corpus can be reduced.

Description

technical field [0001] Embodiments of the present invention relate to the field of information-to-speech synthesis, and in particular, relate to an acoustic model establishment, speech synthesis method, device, equipment, and storage medium. Background technique [0002] With the continuous development of multimedia communication technology, speech synthesis technology, which is one of the important ways of human-computer interaction, has attracted extensive attention of researchers because of its convenience and speed. Speech synthesis is a technology that generates artificial voice through mechanical and electronic methods. It is a technology that converts text information generated by the computer itself or externally input into intelligible and fluent spoken language output. The purpose of speech synthesis is to convert text into speech and play it to users, and the goal is to achieve the effect of live text broadcasting. [0003] Speech synthesis technology has been wi...

AI Technical Summary

Problems solved by technology

During the specific implementation process, the inventor found the following problems in the prior art: if the common application scenarios are covered, more corpus needs to be recorded to establish an acoustic model with better effect of synthesizing tones, but the cost of corpus recording is relatively high ; If there are fewer recordings of Erhuayin, it is easy to cause the problem of poor Erhuayin modeling in the acoustic model; it is also impossible to borrow the existing final phonemes in the corpus to model Erhuayin, and it is impossible to synthesize the speech synthesis library. Er Hua Yin

Images