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

Speech recognition method, device, electronic device and storage medium

A speech recognition and speech technology, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of inability to use information data, poor low-frequency word recognition effect, etc., and achieve the effect of improving the accuracy.

Active Publication Date: 2022-07-12
出门问问创新科技有限公司 +1
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

During the specific implementation process, the inventor found that the personalized voice recognition in the prior art mainly uses the user's voice data to adapt the acoustic model. This method requires the user to actively provide voice data, but cannot use other information data
At the same time, the inventor also found that the high-frequency words and low-frequency words are different when everyone speaks. The user's unique high-frequency words may be low-frequency words in the big data statistical model, and the current speech recognition is mainly based on big data. Statistical model, which uniformly analyzes the high-frequency words of all users, which leads to a significantly worse recognition effect on low-frequency words (or user-specific high-frequency words) than high-frequency words

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 recognition method, device, electronic device and storage medium
  • Speech recognition method, device, electronic device and storage medium
  • Speech recognition method, device, electronic device and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0027] figure 1 This is a flow chart of a speech recognition method provided in Embodiment 1 of the present invention, which can be applied to personalized speech recognition for different users. and / or hardware, and can generally be integrated in a processor, such as a processor of a mobile terminal. like figure 1 As shown, the method of the embodiment of the present invention specifically includes:

[0028] S110. Acquire language model reference data associated with the user according to the user's voice recognition request.

[0029] For example, in the application scenario of man-machine dialogue, a user sends a voice signal to an electronic device such as a mobile terminal, and the electronic device will perform voice recognition on the user's voice after receiving the voice. Among them, the user will first trigger the physical button or virtual button on the electronic device to receive the voice, and then send out the voice signal.

[0030] Furthermore, the voice rec...

Embodiment 2

[0042] figure 2 This is a flowchart of a speech recognition method provided in Embodiment 2 of the present invention. On the basis of the above technical solution, the embodiment of the present invention concretizes the language model reference data:

[0043] A specific implementation is that the language model reference data is specifically personal information reference data, and the personal language model is specifically a personal information language model; further, the user's personal language model is constructed according to the language model reference data, specifically: according to Personal information reference data builds a personal information language model.

[0044] Another specific implementation is that the language model reference data is specifically the personal information reference data and the personal dialogue reference data; correspondingly, the personal language model is specifically the personal information language model and the personal dialog...

Embodiment 3

[0066] image 3 This is a flowchart of a speech recognition method provided in Embodiment 3 of the present invention. On the basis of the above technical solutions, the embodiment of the present invention will perform speech recognition on the to-be-recognized speech input by the user according to the general language model and the individual language model, specifically:

[0067] Perform real-time word segmentation recognition on the received speech to be recognized, and obtain at least one basic candidate word of the current word segmentation;

[0068] Using the general language model and the personal language model to respectively score the basic candidate word under at least one recognition path;

[0069] According to the scoring results of the basic candidate words by the general language model and the personal language model, the standard candidate words of the current word segmentation under at least one recognition path and the comprehensive score corresponding to the...

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

Embodiments of the present invention disclose a speech recognition method, device, electronic device and storage medium. The method includes: obtaining language model reference data associated with the user according to a user's voice recognition request; constructing a personal language model of the user according to the language model reference data; and according to the general language model and the personal language model, Perform speech recognition on the speech to be recognized entered by the user. The above method solves the problem of identifying high-frequency words unique to users in speech recognition, and improves the accuracy of personalized speech recognition.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of speech recognition, and in particular, to a speech recognition method, apparatus, electronic device, and storage medium. Background technique [0002] With the development of big data, machine learning, cloud computing, artificial intelligence and other technologies, speech recognition is gradually liberating users' hands, and the speech input box is also likely to replace the mouse and keyboard. With the popularization of intelligent mobile devices, voice interaction, as a new type of human-computer interaction method, has attracted more and more attention of the entire IT (Information Technology, information technology) industry. [0003] In view of the fact that speakers often come from different dialect areas, have different accents, and have different habits and emotions when speaking, personalized speech recognition based on deep learning emerges as the times require. During t...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G10L17/04G10L15/30
CPCG10L15/30G10L17/04
Inventor 邹明
Owner 出门问问创新科技有限公司
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