Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Switching method and device for multiple voice identification models

A speech recognition model and language recognition technology, applied in speech recognition, speech analysis, instruments, etc., can solve problems such as low efficiency and insufficient intelligence, and achieve the effect of improving switching efficiency

Active Publication Date: 2016-09-21
BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD
View PDF8 Cites 29 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Obviously, this language switching method is relatively inefficient and not intelligent enough

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
  • Switching method and device for multiple voice identification models
  • Switching method and device for multiple voice identification models
  • Switching method and device for multiple voice identification models

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0022] figure 1 This is a flowchart of a method for switching between multiple voice recognition models provided in the first embodiment of the present invention. This embodiment is applicable to the case of switching under multiple voice recognition models. This method can be implemented by the multiple voice recognition methods provided by the embodiments of the present invention. Switch the device to execute, the device can be integrated in a mobile terminal, a fixed terminal or a server, such as figure 1 As shown, specifically including:

[0023] S101. Acquire at least one piece of voice information in the voice input by the user.

[0024] Wherein, the voice information may be part of the voice information in the intercepted input voice, or may be a complete voice information of the user. The voice information may include one or more voice sentences.

[0025] Specifically, the voice can be collected through the microphone of the terminal. For example, a voice input button is pr...

Embodiment 2

[0067] figure 2 This is a flowchart of a method for switching multiple speech recognition models provided in the second embodiment of the present invention. Based on the above-mentioned embodiment, this embodiment will recognize the voice information and match the language category to determine the degree of matching The corresponding target language category is optimized to recognize the voice information based on the characteristics of at least two language categories, and obtain the degree of similarity between the voice information and each language category, and use the degree of similarity as the matching degree of the language category. Such as figure 2 As shown, specifically including:

[0068] S201: Acquire at least one piece of voice information in the voice input by the user.

[0069] S202: Recognizing the voice information based on the characteristics of the at least two language categories, and obtaining the degree of similarity between the voice information and each...

Embodiment 3

[0074] image 3 This is a flowchart of a method for switching multiple speech recognition models provided in the third embodiment of the present invention. On the basis of the above-mentioned embodiment, this embodiment performs recognition and language category matching on the voice information to determine the corresponding The target language category is optimized to recognize at least two voice sentences contained in the voice information to obtain the matching degree of each voice sentence with the language category; determine the initial language category according to the matching degree, and determine the initial language category according to the matching degree of each voice sentence and The matching degree of the initial language category determines the corresponding target language category. Such as image 3 As shown, specifically including:

[0075] S301. Acquire at least two voice sentences in the voice input by the user.

[0076] S302. Recognize the at least two voic...

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 discloses a switching method and device for multiple voice identification models. The method comprises: at least one piece of voice information inputted into a voice by a user is obtained; identification and voice type matching are carried out on the voice information, so that a corresponding target language type is determined based on the matching degree; and a currently used voice identification model is switched to a voice identification model corresponding to the target language type. According to the embodiment of the invention, identification and voice type matching are carried out on the obtained voice information and the corresponding target language type is determined based on the matching degree; and the currently used voice identification model is switched to the voice identification model corresponding to the target language type, so that automatic switching of voice identification models among different languages can be realized. Therefore, the voice identification model switching frequency is improved and the voice identification becomes intelligent.

Description

Technical field [0001] The embodiment of the present invention relates to the technical field of speech recognition, and in particular to a method and device for switching multiple speech recognition models. Background technique [0002] With the development of science and technology, speech input technology is less restricted by the scene because of its use, and it is faster and more convenient than handwriting input, so it has gradually been widely used. For example, existing search engines have added voice search functions. [0003] Although Putonghua has become the main language of communication among the people, there is still a great demand for local dialect communication in some areas. The existing speech recognition engine only supports a specific language, and the speech recognition performance outside the language is basically unavailable. Therefore, the user generally needs to select a speech recognition engine in a specific language in advance before using it. [0004] ...

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): G10L15/00
CPCG10L15/00G10L15/005G10L15/32G10L15/02G10L15/197G10L15/22G10L25/78
Inventor 蒋兵李先刚丁科
Owner BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com