Keyword detection method and system, mobile terminal and storage medium

A keyword detection and keyword technology, which is applied in the field of keyword detection, can solve the problems of low detection accuracy and achieve the effects of improving accuracy, speeding up detection speed and detection efficiency

Active Publication Date: 2020-07-17
XIAMEN KUAISHANGTONG TECH CORP LTD
View PDF11 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the embodiment of the present invention is to provide a keyword detection method, system, mobile terminal and storage medium, aiming to solve the problem of accurate detection of existing keyword detection methods low sex problem

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
  • Keyword detection method and system, mobile terminal and storage medium
  • Keyword detection method and system, mobile terminal and storage medium
  • Keyword detection method and system, mobile terminal and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0049] see figure 1 , is a flowchart of the keyword detection method provided in the first embodiment of the present invention, including steps:

[0050] Step S10, obtaining the text corpus and the transcribed text corresponding to the text corpus in the training set, and performing model training on the language model according to the text corpus and the transcribed text;

[0051] Wherein, the text corpus is the language to be recognized by the speech recognition model, and the text corpus can be selected according to requirements. For example, the text corpus can be Cantonese or Hokkien and other languages. In this step, the expression of Putonghua is used in the transcribed text , and there is a one-to-one correspondence between the text corpus and the transcribed text;

[0052] Preferably, by dividing the local pre-stored data set, the data is divided into a training set, a verification set and a test set, the training set is used to provide training data for the language...

Embodiment 2

[0067] see figure 2 , is a flowchart of the keyword detection method provided in the second embodiment of the present invention, including steps:

[0068] Step S11, obtaining the text corpus and the transcribed text corresponding to the text corpus in the training set, and performing model training on the language model according to the text corpus and the transcribed text;

[0069] Among them, after the text corpus is acquired, noise and reverberation processing can be performed on the text corpus, so as to effectively expand the data, and improve the robustness of the language model, so that the model can adapt to more complex environments ;

[0070] Preferably, by dividing the local pre-stored data set, the data is divided into a training set, a verification set and a test set, the training set is used to provide training data for the language model and the acoustic model in the speech recognition model, the verification The set and test set are used for verification and...

Embodiment 3

[0099] see image 3 , is a schematic structural diagram of the keyword detection system 100 provided by the third embodiment of the present invention, including: a language model training module 10, a model combination module 11, a word graph index module 12, a keyword retrieval module 13 and an occurrence probability calculation module 14, in:

[0100] The language model training module 10 is used to obtain the transcribed text corresponding to the text corpus in the text corpus and the training set, and perform model training on the language model according to the text corpus and the transcribed text;

[0101] The model combination module 11 is configured to perform model training on the chain model according to the acoustic features in the training set, and combine the chain model and the language model to obtain a speech recognition model.

[0102] Wherein, the model combination module 11 is also used to: train a monophone acoustic model according to the acoustic features...

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 keyword detection method and system, a mobile terminal and a storage medium. The method comprises the steps of: acquiring text corpus and a transliteration text to perform model training on a language model; performing model training on a chain model according to acoustic features in a training set, and combining the chain model with the language model to obtain a speech recognition model; inputting a to-be-detected speech segment into the speech recognition model for analysis to obtain a word graph, and performing inverted indexing on the word graph; converting the index result into a factor converter, and inputting a preset keyword into the factor converter for retrieval to obtain a keyword retrieval result; and calculating the occurrence probability of the preset keyword according to the keyword retrieval result, and when the occurrence probability is greater than a probability threshold, judging that the preset keyword occurs in the to-be-detected speech segment. According to the invention, the speech recognition model is controlled to decode the to-be-detected speech segment to generate the word graph, so that the situation of keyword detection errorscaused by speech recognition errors is avoided, and the accuracy of keyword detection is improved.

Description

technical field [0001] The invention belongs to the technical field of keyword detection, and in particular relates to a keyword detection method, system, mobile terminal and storage medium. Background technique [0002] Keyword detection is a technology to detect keywords from continuous speech of interest. Keyword detection has important applications in smart home, telephone monitoring, voice data mining and other fields. Keyword detection has a research history of more than 40 years, but keyword detection in low resource, low power consumption, and low computational complexity environments is still a research hotspot. Keyword detection can be divided into two categories in terms of application scenarios. One category is that the number of keywords is small and fixed. It continuously detects whether there are keywords in the keyword list in the voice stream from the continuous voice stream. The typical application is in the smart home. The other type is that the number of...

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): G10L15/26G10L15/08G10L15/06G10L15/04G06N20/00G06F16/61
CPCG06F16/61G06N20/00G10L15/063G10L15/04G10L15/08G10L2015/0631
Inventor 徐敏肖龙源李稀敏蔡振华刘晓葳谭玉坤
Owner XIAMEN KUAISHANGTONG TECH CORP 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