Method and device for identifying Chinese and English speech signal

A speech signal, Chinese and English technology, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of multi-system resources, occupation, inaccurate recognition performance of Chinese and English bilingual speech recognition system, etc., to improve the recognition rate.

Inactive Publication Date: 2010-09-08
GLOBAL INNOVATION AGGREGATORS LLC
View PDF6 Cites 37 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] In the process of realizing the embodiment of the present invention, the inventors found that the above-mentioned scheme has at least the following problems: the implementation scheme of the first Chinese and English bilingual speech recognition system needs a large amount of marked speech data for the training of the acoustic model, and takes up more system resources
The implementation of the second C

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
  • Method and device for identifying Chinese and English speech signal
  • Method and device for identifying Chinese and English speech signal
  • Method and device for identifying Chinese and English speech signal

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0027] The processing flow of a method for recognizing Chinese and English speech signals provided by this embodiment is as follows: figure 1 , including the following processing steps:

[0028] Step 11. Extract the features of the Chinese and English speech signals to be recognized by using the search algorithm, and obtain the feature information of the speech signals to be recognized.

[0029] After the Chinese and English speech signals to be recognized are obtained, feature extraction is performed on the Chinese and English speech signals to be recognized through a search algorithm to obtain feature information of the speech signals to be recognized. The above-mentioned search algorithm may be a frame synchronization beam search algorithm, an N-best stack decoding search algorithm based on a backward ternary grammar (3-gram), and the like.

[0030] Step 12, comparing the characteristic information with the acoustic model corresponding to each phoneme sequence in the prese...

Embodiment 2

[0038] Embodiment 2 of the present invention provides a Chinese phonetic model and English phoneme model method based on Chinese and English parameter sharing. The processing flow of the method is as follows figure 2 shown, including the following steps:

[0039] Step 21: Create a single-state model comparison table for Chinese consonants and English phonemes.

[0040] In order to avoid confusion caused by the same Chinese phoneme symbols and English phoneme symbols, the prefix ch_ is added before the Chinese phoneme symbols, and the prefix eng_ is added before the English phonemes. For example, the Chinese initial consonant f is written as ch_f, and the English phoneme f is written as eng_f: Chinese There are 64 consonants and finals (including zero consonants), 45 English phonemes (British English phonemes are selected), and 109 Chinese consonants and English phonemes.

[0041] Split each Chinese consonant and English phoneme into multiple (for example, 3) single-state mod...

Embodiment 3

[0097] Based on the above-mentioned Chinese phoneme model and English phoneme model, the structural block diagram of a Chinese-English speech signal recognition device provided by this embodiment is as follows: image 3 As shown, the following modules are included:

[0098] The feature information extraction module 33 is configured to perform feature extraction of the Chinese and English speech signals to be recognized through a search algorithm, and obtain feature information of the speech signals to be recognized. The above feature information extraction module can be realized by a language-independent speech recognition decoder,

[0099]The identification and comparison module 35 is used to compare the feature information with the acoustic model corresponding to each phoneme sequence in the preset mixed pronunciation database;

[0100] The processing module 36 is used to determine the phoneme sequence corresponding to the feature information according to the comparison res...

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 method and a device for identifying a Chinese and English speech signal. The method mainly comprises the following steps of: carrying out feature extraction on a Chinese and English speech signal to be identified by a searching algorithm to acquire the feature information of the speech signal to be identified; and comparing the feature information with an acoustic model corresponding to each phoneme sequence in a mixed speech database, determining a phoneme sequence corresponding to the feature information based on the comparative result, and acquiring a Chinese and English mixed phrase corresponding to the phoneme sequence, wherein the Chinese and English mixed phrase is taken as an identification result of the Chinese and English speech signal to be identified. The invention can establish the acoustic model with less confusion, and does not need a large amount of labeled speech training data, thereby saving system resources. The invention can effectively raise the identification rate of the Chinese and English speech signal.

Description

technical field [0001] The invention relates to the field of speech recognition, in particular to a method and device for recognizing Chinese and English speech signals. Background technique [0002] With the development of information globalization, multilingualism and multilingual communication have become more and more common. A single speech recognition system cannot effectively recognize multilingual communications, and it is a new task for speech recognition technology to establish a speech recognition system that can recognize multiple languages ​​and speech signals. [0003] Chinese is currently the language with the most users, and English is the language with the widest distribution of users. Therefore, establishing a Chinese-English bilingual recognition system has a good application prospect. [0004] The implementation scheme of the first Chinese and English bilingual speech recognition system in the prior art is: integrate the Chinese speech recognizer and the...

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): G10L15/02
Inventor 刘轶詹五洲王东琦
Owner GLOBAL INNOVATION AGGREGATORS LLC
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