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

Wake-up word recognition method, device, electronic device, and computer-readable storage medium

A recognition method and wake-up word technology, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of inconvenient use, inability to fix, increase design complexity and R&D cost, and reduce complexity, reduce R&D cost, The effect of improving recognition efficiency and response speed

Active Publication Date: 2021-04-02
MOBVOI INFORMATION TECH CO LTD
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] During the specific implementation process, the inventor found the following defects in the prior art: words or words are stored in each node of the speech recognition model, for example, when the wake-up word is "Wang Mimi", each node in the speech recognition model "Wang" and "Mimi" are stored, and when performing speech recognition, the recognition result is also to output the corresponding wake-up word itself, that is, output "Wang Mimi", but when changing the wake-up word, for example, the wake-up word " Wang Mimi" is changed to "Xiao Jia", and the speech recognition model also needs to be changed accordingly, that is, the wake-up word model that stores "Wang" and "Mimi" in the node is changed to store "Xiao" and "Jia" in the node The wake-up word model leads to the speech recognition model being changed with the change of the wake-up word, which cannot be fixed, which not only causes inconvenience in use, but also greatly increases the complexity of the design and the cost of research and development

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
  • Wake-up word recognition method, device, electronic device, and computer-readable storage medium
  • Wake-up word recognition method, device, electronic device, and computer-readable storage medium
  • Wake-up word recognition method, device, electronic device, and computer-readable storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] The embodiment of the present invention provides a wake-up word recognition method, which can be executed by an intelligent voice device, such as figure 1 As shown, it includes: step S110, obtaining the speech information to be recognized input by the user; step S120, based on the preset speech recognition model, determining the first syllable sequence corresponding to the speech information; step S130: determining whether the first syllable sequence includes The second syllable sequence of the preset wake-up word; Step S140: If yes, determine that the voice information includes the preset wake-up word, and perform a corresponding wake-up operation.

[0034] Specifically, the embodiment of the present invention may be executed by an intelligent voice device, or may be executed by a server, which is not limited by the embodiment of the present invention. Among them, the wake-up word recognition process of the intelligent voice device as the execution subject is roughly t...

Embodiment 2

[0049] The embodiment of the present invention provides another possible implementation manner. On the basis of the first embodiment, the method shown in the second embodiment is also included, wherein,

[0050] The speech recognition model includes a list of syllables, which includes either non-pitched syllables or pitched syllables.

[0051] Step S120 includes step S1201 (not marked in the figure) and step S1202 (not marked in the figure), wherein,

[0052] Step S1201: Divide the voice information according to the length of the voice segment of the preset wake-up word to obtain multiple voice information segments.

[0053] Step S1202: Based on the preset speech recognition model, determine the third syllable sequences respectively corresponding to the plurality of speech information segments.

[0054] Step S130 includes step S1301 (not marked in the figure) and step S1302 (not marked in the figure), wherein,

[0055] Step S1301: Determine whether any third syllable sequenc...

Embodiment 3

[0066] The embodiment of the present invention provides another possible implementation manner. On the basis of the second embodiment, the method shown in the third embodiment is also included, wherein,

[0067] Step S100 (not marked in the figure) and step S101 (not marked in the figure) are also included before step S110, wherein,

[0068] Step S100: Receive an input audio signal.

[0069] Step S101: Perform noise filtering processing on the audio signal to obtain voice information in the audio signal.

[0070]Specifically, when the smart voice device is turned on, it usually receives the input audio signal in real time through its built-in high-performance audio acquisition device (such as a microphone, microphone array), wherein, in a very quiet scene, the audio signal can be Only the speech of the user is included, and in a slightly noisy environment, the audio signal may be speech including various noises.

[0071] Further, after receiving the input audio signal, the a...

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, which relates to the field of voice processing, provides a wake-up word recognition method and device, an electronic device, and a computer-readable storage medium. The wake-up word recognition method comprises: acquiring to-be-recognized voice information inputted by a user; on the basis of a preset voice recognition model, determining first syllable sequence corresponding to the voice information; determining whether the first syllable sequence includes a second syllable sequence of a preset wake-up word; and if so, determining that the voice information includes the preset wake-up word and performing a corresponding wake-up operation. According to the method provided by the embodiment of the invention, the user only needs to identify whether the wake-upword is included by the voice information based on the syllable sequence instead of identifying whether the voice information includes a wake-up word or phrase, so that the voice recognition model does not need to change with changing of the wake-up word and can be kept to be unchanged and thus the design complexity and the research and development costs are reduced substantially.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of voice processing, and in particular, relate to a wake-up word recognition method, device, electronic equipment, and computer-readable storage medium. Background technique [0002] With the continuous development of terminal equipment, the application of intelligent voice hardware equipment is becoming more and more extensive, such as intelligent speakers, robots, etc., users can input a sound signal in the intelligent voice hardware device, and then the intelligent voice hardware device or intelligent voice hardware The background server of the device can carry out semantic recognition of the segment of sound signal, and perform corresponding operations according to the semantic recognition results, and in some cases, return corresponding operation results to the user. [0003] At present, after the intelligent voice device acquires the voice signal input by the user, it needs to ide...

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): G10L15/22G10L15/02
CPCG10L15/02G10L15/22G10L2015/027G10L2015/223
Inventor 胡亚光
Owner MOBVOI INFORMATION TECH CO LTD