Error correction method and device for voice recognition text

A technology of speech recognition and error correction methods, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of lack of recognition error correction means, affect the promotion of speech recognition technology, etc., and achieve the effect of accurate error correction

Active Publication Date: 2015-03-25
BEIJING BAIDU NETCOM SCI & TECH CO LTD
View PDF5 Cites 28 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Especially in everyday spoken language scenarios, speech recognition errors are more common
However, there is a lack of error correction means for recognition errors in the prior art, thus affecting the further promotion of speech recognition technology.

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
  • Error correction method and device for voice recognition text
  • Error correction method and device for voice recognition text
  • Error correction method and device for voice recognition text

Examples

Experimental program
Comparison scheme
Effect test

no. 1 example

[0030] figure 1 It is a flow chart of the error correction method for speech recognition text provided by the first embodiment of the present invention. see figure 1 , the error correction method of the speech recognition text comprises:

[0031] S110. According to the multi-level K-Gram index of the speech recognition result text, extract at least one candidate error correction text for error correction of the result text.

[0032] Before error correction is performed on the result text of the speech recognition, a multi-level K-Gram index of the result text is established first. After the multi-level K-Gram index of the result text is established, the candidate error correction text most similar to the result text is pulled from a preset corpus according to the multi-level K-Gram index.

[0033] Specifically, the multi-level K-Gram index includes the K-Gram index at the Chinese character level, the K-Gram index at the Pinyin syllable level, the K-Gram index at the Quanpin...

no. 2 example

[0054] figure 2 It is a flow chart of the error correction method for speech recognition text provided by the second embodiment of the present invention. The error correction method of the speech recognition text is based on the first embodiment of the present invention. Further, the error correction method of the speech recognition text includes: according to the multi-level K-Gram index of the speech recognition result text, the At least one non-template candidate error-correcting text for performing error correction on the result text; determine the fuzzy sound editing distance matrix between the at least one non-template candidate error-correcting text and the result text respectively; according to the determined fuzzy sound The edit distance matrix obtains the fuzzy sound editing distance and the candidate error correction boundary between the at least one non-template candidate error correction text and the result text respectively; according to the fuzzy sound editing ...

no. 3 example

[0087] Figure 5 It is a flow chart of the error correction method for speech recognition text provided by the third embodiment of the present invention. The error correction method of the speech recognition text is based on the first embodiment of the present invention. Further, the error correction method of the speech recognition text includes: according to the multi-level K-Gram index of the speech recognition result text, the At least one template candidate error-correcting text for error-correcting the result text; determine the fuzzy sound editing distance matrix between the at least one template candidate error-correcting text and the result text respectively; according to the determined fuzzy sound editing distance The matrix obtains the fuzzy sound edit distance and the candidate error correction boundary between the at least one template candidate error correction text and the result text respectively; selects the error correction according to the fuzzy sound edit d...

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 an error correction method and device for a voice recognition text. The error correction method for the voice recognition text comprises the steps that at least one candidate error correction text used for error correction of a result text is obtained according to the multi-layer K-Gram index of the voice recognition result text, the fuzzy sound editing distance matrix between the at least one candidate error correction text and the result text is determined, the fuzzy sound editing distance between the at least one candidate error correction text and the result text and a candidate error correction boundary are obtained according to the determined fuzzy sound editing distance matrix, an error correction text is selected according to the fuzzy sound editing distance corresponding to the at least one candidate error correction text, and error correction is conducted on the result text according to the candidate error correction boundary corresponding to the error correction text. By the adoption of the error correction method and device for the voice recognition text, accurate error correction of the voice recognition result text is achieved.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of speech recognition, and in particular, to a method and device for correcting errors in speech-recognized text. Background technique [0002] With the maturity of speech recognition technology, the application fields of speech recognition are becoming more and more extensive. Compared with other text input methods, the voice input method implemented by speech recognition is more in line with people's daily habits, and also makes the input process more efficient. It can be expected that speech recognition technology will be widely used in many fields such as industrial production, communication, medical treatment, and housekeeping services. [0003] In the practical application of speech recognition technology, due to the influence of surrounding noise, dialect and other factors, the recognition result of speech recognition is often inconsistent with the user's input. Especially in e...

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 Applications(China)
IPC IPC(8): G10L15/26
Inventor 时迎超周晓张海雷
Owner BEIJING BAIDU NETCOM SCI & TECH CO LTD
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