Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Voiceless sound and voiced sound judging method and device and voice synthesizing system

A judging method and unvoiced sound technology, which is applied in speech synthesis, speech analysis, instruments, etc., can solve the problems of affecting the synthesis effect and inaccurate judgment of unvoiced and voiced sounds, and achieve the effect of improving the success rate of judgment and improving the quality

Active Publication Date: 2014-11-12
TENCENT TECH (SHENZHEN) CO LTD +1
View PDF9 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] However, the set of questions designed to train the Hidden Markov (HMM) model is not specifically for unvoiced judgment, and during the prediction process, the questions in the decision tree may not be related to unvoiced sound at all, but are used to judge speech voicing, which naturally leads to inaccurate determination of voicing
When the accuracy of unvoiced and voiced sound determination is not high enough and errors occur, the unvoiced voiced sound and unvoiced voiced sound of the synthesized speech will seriously affect the synthesis effect

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Voiceless sound and voiced sound judging method and device and voice synthesizing system
  • Voiceless sound and voiced sound judging method and device and voice synthesizing system
  • Voiceless sound and voiced sound judging method and device and voice synthesizing system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0029] In the trainable speech synthesis system (Trainable TTS) based on the hidden Markov model, the speech signal is converted into an excitation parameter (Excitation parameter) and a spectral parameter (Spectral parameter) by frame. The excitation parameters and spectral parameters are trained as HMM models (training part) respectively. Then, on the speech synthesis side (synthesis part), based on the voicing judgment, voiced sound fundamental frequency and spectral parameters predicted by the HMM model, it is synthesized into speech by a vocoder.

[0030] In the synthesis stage, if a frame is judged as voiced, the excitation signal is assumed to be an impulse response sequence; if it is judged to be unvoiced, the excitation signal is assumed to be white noise...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The embodiment of the invention provides a voiceless sound and voiced sound judging method and device and a voice synthesizing system. The method comprises the steps of setting a voiceless sound and voiced sound judging problem set, utilizing speech training data and the voiceless sound and voiced sound judging problem set to train a voiceless sound and voiced sound judging model of a dichotomia decision tree structure, receiving voice test data and utilizing the trained voiceless sound and voiced sound judging model to judge whether the voice test data are voiceless sound or voiced sound. In addition, non-leaf nodes in the dichotomia decision tree structure are problems in the voiceless sound and voiced sound judging problem set, and leaf nodes in the dichotomia decision tree structure are voiceless sound and voiced sound judging results. The embodiment of the voiceless sound and voiced sound judging method improves the voiceless sound and voiced sound judging success rate and the voice synthesizing quality.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of speech processing, and more specifically, relate to a method, device, and speech synthesis system for judging unvoiced and voiced sounds. Background technique [0002] In today's information age, various information devices have emerged as the times require: fixed telephones and mobile phones for voice transmission; servers and personal computers for information resource sharing and processing; various TV sets for video data display, etc. Wait. These devices are produced to solve actual needs in specific fields. With the advent of the integration of electronic consumption, computer, and communication (3C), people are paying more and more attention to the research on the comprehensive utilization of information equipment in various fields, so as to make full use of existing resources and equipment to serve people. better service. [0003] Speech synthesis is the technology of p...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G10L25/93G10L13/02
CPCG10L13/02G10L25/93
Inventor 唐宗尧
Owner TENCENT TECH (SHENZHEN) CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products