Ill-conditioned voice evaluation method based on Chinese voice

A speech and Chinese technology, used in speech analysis, speech recognition, instruments, etc.

Active Publication Date: 2019-05-07
SHENZHEN RES INST THE CHINESE UNIV OF HONG KONG
View PDF8 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0009] In order to solve the problem that there is no method in the prior art that can carefully evaluate patholog

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
  • Ill-conditioned voice evaluation method based on Chinese voice
  • Ill-conditioned voice evaluation method based on Chinese voice
  • Ill-conditioned voice evaluation method based on Chinese voice

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0036]In the field of voice signal processing and speech therapy, the objective analysis of pathological voices has attracted many attentions. For example, MDVP is a voice signal analysis software system commonly used by speech therapists. MDVP provides 22 acoustic signal parameters to describe voice quality. These parameters are used by speech therapists as a basis for voice assessment. LingWAVES is another commercial software system used to assist physicians in clinical diagnosis of voice problems. LingWAVES can perform basic acoustic signal analysis, such as spectrum feature analysis, tone analysis, volume analysis and so on. In recent years, many researchers have adopted a free software, Praat, for acoustic signal analysis, but Praat is not a tool specifically for voice analysis of lesions, nor can it judge the type and severity of lesions.

[0037] From the perspective of signal processing, related research mainly focuses on how to extract effective feature parameters ...

Embodiment 2

[0096] The specific implementation of adopting the method of the present invention to carry out pathological voice evaluation is as follows:

[0097] (1) Establish a database of pathological voices: this database is jointly completed by hospital speech therapists, scholars and volunteer patients. The scale of patients is 230 native speakers of Chinese, and the gender and age are balanced. In a relatively fixed environment, each patient recorded voice signals including long vowels, short texts read aloud, and question answers. The recorded audio is a two-channel, 16bit, wav file with a sampling rate of 44.1kHz. After simple pre-processing of the collected speech signals (using speaker diarization technology to delete speech therapist's speech content, fixed multiple amplification and noise reduction), a total of 48 professional speech therapists were given for subjective scoring. Subjective scoring of 10 voice problems was performed on each patient's recording. In order to i...

Embodiment 3

[0104] Such as Figure 5 As shown, the present invention also provides a pathological voice evaluation system based on Chinese speech, including a speech input module, a speech evaluation module using the method described in Embodiment 1, and an evaluation result output module.

[0105] The voice input module is used to accept real-time recording or audio, and transmits the real-time recording or audio to the voice evaluation module; the voice evaluation module is used to evaluate the real-time recording or audio, and transmits the evaluation result to An evaluation result output module; the result output module is used to output the evaluation result.

[0106] The method and system provided by the present invention have the following beneficial effects:

[0107] (1) Carry out phoneme segmentation based on automatic speech recognition technology to input continuous speech signals of pathological voices, and classify the segmented speech sequences according to the vocalization...

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 invention provides an ill-conditioned voice evaluation method based on Chinese voice, which comprises the following steps: completing automatic alignment of input continuous voice signals and corresponding texts by using an automatic Chinese voice recognition system, and completing phoneme segmentation; segmenting the continuous voice signal into vowels, clear consonants and turbid consonantsaccording to the time sequence of phoneme segmentation, and then extracting features; wherein the extracted features form a feature parameter set, and the feature parameter set is screened by a feature selector and then input into a classifier to obtain a fine score of the voice problem. According to the method, a large-scale normal voice database is used for training an automatic voice recognition system; the segmented voice sequences are classified according to the sounding characteristics of different phonemes, the characteristic parameters are designed for different phoneme types, finally,the objective overall score and each detailed score for the voice problem are obtained, and an important reference is provided for clinical diagnosis and rehabilitation treatment.

Description

technical field [0001] The invention relates to the technical field of voice detection, in particular to a method for evaluating pathological voices based on Chinese phonetics. Background technique [0002] The voice is the carrier of human language communication. The vocal system produces sound driven by the vibration of the vocal cords, and spreads out through the passage formed by the throat and oral cavity. Sound carries different information and is an indispensable means of communication between people. The sound emitted by the vocal system can be described by a waveform signal, which is called a voice signal. When the vocal organs are in a normal state, the vibration of the vocal cords has obvious periodicity, and the transmission channels formed by the throat and oral cavity also change regularly, so the voice signals generated are also very regular. [0003] In real life, the voice is not only used to exchange information with each other, but also used for singing...

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): G10L25/51G10L25/93G10L15/16G10L15/14G10L15/04G10L15/06G10L15/08
Inventor 李丹刘媛媛
Owner SHENZHEN RES INST THE CHINESE UNIV OF HONG KONG
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
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap