Unlock instant, AI-driven research and patent intelligence for your innovation.

Keyword sample determination method, speech recognition method, device, equipment and medium

A technology of speech recognition and determination method, applied in speech recognition, speech analysis, instruments, etc., can solve problems such as high cost, and achieve the effect of improving comprehensiveness, improving accuracy and reducing acquisition cost

Active Publication Date: 2021-05-11
GUANGZHOU NETSTAR INFORMATION TECH CO LTD
View PDF15 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In the existing scheme, a large amount of audio data is collected to obtain a keyword training set by manually recording the corresponding keyword speech, which requires a relatively high cost, and requires the recording environment of the collected audio data to be consistent with the pre-defined keywords. The actual environment is consistent, which leads to certain limitations in the generation of various keywords

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 sample determination method, speech recognition method, device, equipment and medium
  • Keyword sample determination method, speech recognition method, device, equipment and medium
  • Keyword sample determination method, speech recognition method, device, equipment and medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0042] Figure 1A It is a flow chart of a method for determining a keyword sample provided by Embodiment 1 of the present invention. This embodiment can be applied to any situation where it is necessary to determine a keyword sample for model training. The solutions in the embodiments of the present invention can be applied to how to solve the problem of high cost and limitations in acquiring keyword samples. A method for determining a keyword sample provided in this embodiment can be executed by the device for determining a keyword sample provided in an embodiment of the present invention. The device can be implemented by means of software and / or hardware, and integrated into the device that executes the method. , the device can be any smart terminal device, such as a laptop, tablet or desktop.

[0043] Specifically, refer to Figure 1A , the method may include the following steps:

[0044] S110, acquiring keywords.

[0045] Among them, the keyword refers to any word that t...

Embodiment 2

[0059] Figure 2A It is a flow chart of a keyword sample determination method provided in Embodiment 2 of the present invention, Figure 2B It is a schematic diagram of the principle of a keyword sample determination process provided by Embodiment 2 of the present invention. In this embodiment, optimization is performed on the basis of the technical solutions provided in the foregoing embodiments. Specifically, in this embodiment, a detailed explanation is given on the specific determination process of the keyword speech segment in the target speech sample.

[0060] optional, such as Figure 2A As shown, the following steps may be included in this embodiment:

[0061] S210, acquiring keywords.

[0062] S220. Obtain target speech samples including keywords from an existing speech recognition sample database.

[0063] S230. Determine the start time point and end time point of the phoneme of the keyword in the audio data phoneme of the target speech sample.

[0064] Wherein...

Embodiment 3

[0074] Figure 3A It is a flow chart of a voice recognition method provided by Embodiment 3 of the present invention, and this embodiment can be applied to any situation of recognizing keywords contained in a user's voice instruction. The solution of the embodiment of the present invention can be applied to how to solve the problem of cumbersome training process of the keyword recognition model. A speech recognition method provided in this embodiment can be executed by the speech recognition device provided in the embodiment of the present invention. The device can be realized by means of software and / or hardware, and can be integrated into a device that executes the method. The device can It is any kind of intelligent terminal equipment, such as laptop, tablet or desktop.

[0075] Specifically, refer to Figure 3A , this embodiment may include the following steps:

[0076] S310. Acquire a user's voice instruction.

[0077] Specifically, when the user needs to perform a ce...

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 discloses a keyword sample determination method, a speech recognition method, a device, a device and a medium. Wherein, the method for determining the keyword sample includes: obtaining a keyword; obtaining a target speech sample including the keyword in an existing speech recognition sample library; determining a speech segment of the keyword in the target speech sample, and obtaining the keyword sample. The technical solution provided by the embodiments of the present invention does not need to generate keyword samples by specially recording the keyword speech of each user in each scene, but obtains target speech samples containing keywords from the existing speech recognition sample library, and intercepts Keyword speech fragments in the target speech sample are extracted to obtain keyword samples, which reduces the acquisition cost of keyword samples and improves the comprehensiveness of keyword sample determination. The keyword recognition model obtained through the keyword sample training is used to identify corresponding The keywords contained in the user's voice improve the accuracy of voice recognition.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of speech recognition, and in particular, to a keyword sample determination method, speech recognition method, device, equipment and medium. Background technique [0002] With the increasing number of smart speakers on the market, related technologies in the field of speech recognition have been greatly developed and applied, among which Key Word Spotting (KWS) technology has also been widely used as the basis of voice interactive control. [0003] At present, KWS technology mainly adopts methods based on various neural networks to identify keywords carried in speech. At this time, it is necessary to collect a large amount of audio data containing predefined keywords and non-keywords. The parameters in the network are trained, verified and tested, so that the constructed neural network can accurately recognize the keyword information in the user's voice. [0004] In the existing scheme...

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): G10L15/08G10L15/04G10L15/22
CPCG10L15/04G10L15/08G10L15/22G10L2015/088G10L2015/223
Inventor 李敬
Owner GUANGZHOU NETSTAR INFORMATION TECH CO LTD