Voice recognition test environment control system and method

A speech recognition and test environment technology, applied in speech recognition, speech analysis, measuring devices, etc., can solve the problems of poor stability of the decibel value of the test speech, poor reliability of speech recognition accuracy data, etc., and achieve the effect of increasing speed

Inactive Publication Date: 2020-09-11
江苏清微智能科技有限公司
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The recognition accuracy of the recognition system is affected by the test environment. The test environment includes the decibel value of the test voice at the test position, that is, the sound intensity when the test voice played by the player of the playback device reaches the test position, and the test voice at the test position. The stability of the decibel value is poor, and the reliability of the measured speech recognition accuracy data is poor

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 test environment control system and method
  • Voice recognition test environment control system and method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0035] Such as figure 1 As shown, the present invention provides a test environment control system for speech recognition, including:

[0036] A playback device 10 includes: a player, which has a volume adjustment terminal, and the player can play a test voice with a current volume value. The volume adjustment terminal can receive a volume-up driving message and a volume-down driving message. The player can drive the player to play the corresponding volume according to the volume-up driving information or the volume-down driving information.

[0037] A control unit has an input terminal and a control information output terminal, and the control information output terminal is connected with the volume adjustment terminal of the player. The control information output terminal is capable of outputting volume up driving information and volume down driving information. The input terminal can acquire the target decibel value, the first threshold and the second threshold input by ...

Embodiment 2

[0055] Such as figure 2 As shown, this embodiment provides a test environment control system for speech recognition on the basis of Embodiment 1, including:

[0056] The control unit of the playback device 10 receives a set time length, the control unit obtains multiple decibel values ​​collected by the decibel sensor 20 in real time, and calculates the decibel average value of the multiple decibel values ​​collected by the single time length. Get the difference between the decibel average and the target decibel value.

[0057] The decibel sensor 20 collects the decibel value of the test position in real time. For example, if the set time length is 3s, all the decibel values ​​collected by the decibel sensor for the first time length of 3s are first stored. The six sampling values ​​are 37dB, 42dB, 41dB, 38dB, 44dB, 39dB, calculate the average decibel value of the six decibel values ​​collected in the first time length of 3s is 40dB, and the difference between the obtained a...

Embodiment 3

[0071] This embodiment provides a test environment control method for speech recognition, including:

[0072] Configure a playback device, which includes:

[0073] A player has a volume adjustment terminal, and the player can play a test voice with a current volume value. The volume adjustment terminal can receive a volume-up driving message and a volume-down driving message. The player can drive the player to play the corresponding volume according to the volume-up driving information or the volume-down driving information.

[0074] A control unit has an input terminal and a control information output terminal, and the control information output terminal is connected with the volume adjustment terminal of the player. The control information output terminal is capable of outputting volume up driving information and volume down driving information. The input terminal can acquire the target decibel value, the first threshold and the second threshold input by the user. The fi...

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

According to a voice recognition test environment control system and method, a player can be driven to play the corresponding volume according to the volume increase driving information or the volumedecrease driving information, and a decibel sensor can collect the current decibel value of the test voice played by the player at the test position. And the input end of the control unit acquires thecurrent decibel value output by the decibel information output end. The control unit obtains the difference value between the current decibel value and the target decibel value. The control unit judges whether the difference value is smaller than a first threshold value or not, and if yes, the volume increase driving information is output to the player. The control unit judges whether the difference value is larger than a second threshold value or not, and if yes, the volume decrease driving information is output to the player. Not only is the speed of adjusting the test environment to the expected test environment improved, but also the test environment of the recognition system can be controlled within the expected decibel value range.

Description

technical field [0001] The invention relates to the field of speech recognition testing, in particular to a speech recognition testing environment control system and method. Background technique [0002] During the speech recognition test process, the test speech is played by the player of the playback device, and the recognition system calculates the test speech recognition accuracy rate by recognizing the test speech. For example, the test voice is a recorded Chinese voice. The recognition system recognizes a Chinese voice and converts it into text. The recognition system calculates the recognition accuracy of the recognition system based on the converted text and the accurate text. [0003] The recognition accuracy of the recognition system is affected by the test environment. The test environment includes the decibel value of the test voice at the test position, that is, the sound intensity when the test voice played by the player of the playback device reaches the test ...

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): G10L15/01G06F3/16G01H17/00
CPCG01H17/00G06F3/165G10L15/01
Inventor 孙昊杨广俊欧阳鹏
Owner 江苏清微智能科技有限公司
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