Title:
SPEECH SYNTHESIS METHOD AND APPARATUS, READABLE MEDIUM, AND ELECTRONIC DEVICE
Document Type and Number:
WIPO Patent Application WO/2022/156464
Kind Code:
A1
Abstract:
A speech synthesis method and apparatus (200), a readable medium, and an electronic device (300), relating to the technical field of electronic information processing. The method comprises: obtaining a text to be synthesized and a specified acoustic feature (101), the specified acoustic feature being used for indicating a prosodic feature of an audio; extracting a phoneme sequence corresponding to the text to be synthesized (102); extending the specified acoustic feature according to the phoneme sequence to obtain an acoustic feature sequence (103); and inputting the phoneme sequence and the acoustic feature sequence into a pre-trained speech synthesis model to obtain a target audio output by the speech synthesis model and corresponding to the text to be synthesized (104), an acoustic feature of the target audio matching the specified acoustic feature. Speech synthesis of a text is controlled by means of the specified acoustic feature, so that the target audio output by the speech synthesis model can correspond to the specified acoustic feature, explicit control of the acoustic feature during the process of speech synthesis can be implemented, and the expressiveness of the target audio is improved.
More Like This:
Inventors:
WU PENGFEI (CN)
WU LIN (CN)
PAN JUNJIE (CN)
WU LIN (CN)
PAN JUNJIE (CN)
Application Number:
PCT/CN2021/139987
Publication Date:
July 28, 2022
Filing Date:
December 21, 2021
Export Citation:
Assignee:
BEIJING YOUZHUJU NETWORK TECH CO LTD (CN)
International Classes:
G10L13/10; G10L13/04; G10L25/30
Foreign References:
| CN112786008A | 2021-05-11 | |||
| CN102385858A | 2012-03-21 | |||
| US20050071163A1 | 2005-03-31 | |||
| CN111199724A | 2020-05-26 | |||
| US20030078780A1 | 2003-04-24 | |||
| CN110992927A | 2020-04-10 |
Attorney, Agent or Firm:
CCPIT PATENT AND TRADEMARK LAW OFFICE (CN)
Download PDF: