Voice recognition method and device and device for voice recognition

A speech recognition and speech technology, applied in speech recognition, speech analysis, instruments, etc., can solve problems such as low speech recognition efficiency

Active Publication Date: 2020-04-10
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO LTD
View PDF4 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In practical applications, due to the dependence between different frames corresponding to the recurrent neural network structure, the operation of the recurrent neural network structure is serial, and the serial operation leads to low efficiency of speech recognition

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
  • Voice recognition method and device and device for voice recognition
  • Voice recognition method and device and device for voice recognition
  • Voice recognition method and device and device for voice recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0069] refer to Figure 4 , shows a flow chart of the steps of an embodiment of a training method for a neural network model according to an embodiment of the present invention. The neural network model is applied to speech recognition, and the above-mentioned acoustic model for speech recognition specifically includes: the above-mentioned neural network model and hidden Mark Husband model; the method specifically includes the following steps:

[0070] Step 401. Align the training data to obtain alignment information;

[0071] Step 402, according to the above-mentioned alignment information, divide the above-mentioned training data into data blocks of a preset length;

[0072] Step 403, according to the data blocks corresponding to the above training data, the above neural network model is trained; the above neural network model may include: an input layer, a hidden layer and an output layer; the above hidden layer may include: a feedforward neural network layer and a self-at...

Embodiment 2

[0088] refer to Figure 5 , which shows a flow chart of the steps of an embodiment of a voice recognition method in an embodiment of the present invention, the method specifically includes the following steps:

[0089] Step 501, determining the speech characteristics of the speech to be recognized;

[0090] Step 502, using an acoustic model to determine the speech recognition result corresponding to the above-mentioned speech features; the above-mentioned acoustic model may include: a neural network model and a hidden Markov model; the above-mentioned neural network model may include: an input layer, a hidden layer, and an output layer; the above-mentioned The hidden layer may include: a feedforward neural network layer and a self-attention neural network layer;

[0091] Step 503 , outputting the above speech recognition result.

[0092] Embodiments of the present invention can Figure 4 The resulting neural network model is used in the speech recognition process.

[0093] I...

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 provides a voice recognition method and device and a device for voice recognition. The method specifically comprises the steps: determining voice features of to-be-recognized voice; determining a voice recognition result corresponding to the voice feature by using an acoustic model, wherein the acoustic model comprises a neural network model and a hidden Markov model, the neural network model comprises an input layer, a hidden layer and an output layer, and the hidden layer comprises a feedforward neural network layer and a self-attention neural network layer; and outputting the voice recognition result. The voice recognition method and device and the device for voice recognition can improve the voice recognition efficiency.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of voice processing, and in particular, to a voice recognition method, device, and device for voice recognition. Background technique [0002] Speech recognition technology is a technology that converts speech into text. With the development of computer technology, the application scenarios of voice recognition are gradually increasing, such as voice input scenarios, smart chat scenarios, voice translation scenarios, and the like. [0003] The current speech recognition technology is based on a deep neural network (DNN, Deep Neural Networks) and a hidden Markov model (HMM, Hidden Markov Model) to establish an acoustic model. The acoustic model uses a deep neural network model to model the mapping relationship between acoustic pronunciation and basic acoustic units. Since language is context-dependent, the current deep neural network usually adopts a recurrent neural network structure ...

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): G10L15/16G10L15/14
CPCG10L15/16G10L15/142
Inventor 王智超王佳文刘忠亮
Owner BEIJING SOGOU TECHNOLOGY DEVELOPMENT 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