Speech error detection method by front-end processing using artificial neural network (ANN)

An artificial neural network and front-end processing technology, applied in the field of speech recognition evaluation, can solve problems such as universality and error detection performance to be improved, and achieve the effect of improving error detection performance

Inactive Publication Date: 2011-07-13
龚澍
View PDF0 Cites 36 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] At present, most of the systems for language learning in the world are evaluation and learning systems based on computer-assisted language learning (CALL). Although there are also ap

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
  • Speech error detection method by front-end processing using artificial neural network (ANN)
  • Speech error detection method by front-end processing using artificial neural network (ANN)
  • Speech error detection method by front-end processing using artificial neural network (ANN)

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0031] A voice error detection method using artificial neural network for front-end processing, including the use of multi-layer perceptron MLP to extract 64-dimensional new features, machine recognition of test data, generating error detection metric score GOP, and pointing out the pronunciation in accordance with the set threshold The error and its degree, the specific steps are:

[0032] 1. Establish a standard database of phoneme balance for pronunciation error detection, including standard pronunciation of words, phrases and continuous speech streams:

[0033] 1) Design the recording text according to the phoneme balance principle required by Putonghua error detection;

[0034] 2) Find a group of suitable standard speakers according to gender and age;

[0035] 3) Arrange standard speakers for recording.

[0036] 2. Collect the corpus to be checked for errors and establish a voice database of the test corpus:

[0037] 1) At the site of the Putonghua proficiency test, select a group ...

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 a speech error detection method by front-end processing using an artificial neural network (ANN). The method comprises the following steps: extracting new 64-dimensional features with strong pattern recognition capability and good discrimination property from 39-dimensional mel-frequency cepstrum coefficient (MFCC) parameters by utilizing a multilayer perceptron (MLP); performing speech recognition on test data by a machine, and generating a goodness of pronunciation (GOP) score; and pointing out pronunciation errors and error degree according to a set threshold, and performing directed-learning for pronunciation errors.

Description

Technical field [0001] The invention relates to the field of speech recognition evaluation, in particular, comprehensively using speech recognition methods, phonetics knowledge and artificial neural network knowledge to effectively improve the error detection performance of a speech recognition system when using a computer to evaluate the level of a speaker. Background technique [0002] Putonghua proficiency test is an important method for the promotion of Putonghua, and an important measure for the use of Putonghua to be gradually scientific, standardized and institutionalized. According to the Law of the People's Republic of China on the National Common Language and Characters, announcers, program hosts, film and television actors, teachers, and staff of state agencies who use Putonghua as their working language must take the Putonghua proficiency test and meet the national standards. [0003] The current Putonghua proficiency test work is done by manual scoring. Generally, each...

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): G10L15/16G10L15/14
Inventor 龚澍
Owner 龚澍
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