Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Pronunciation detection method and device, computer equipment and storage medium

A detection method and a technology for correct pronunciation, which are applied in speech analysis, teaching aids and instruments for electrical operation, etc., can solve the problems of error in judgment results, the generalization ability of classification models with limited segmentation accuracy, and the interpretation of pronunciation characteristics. The effect of improving accuracy

Pending Publication Date: 2021-01-05
北京乐学帮网络技术有限公司
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, this judgment of right and wrong based on the pronunciation characteristics of a single speech is limited by the segmentation accuracy and the generalization ability of the classification model, which leads to certain errors in the judgment results and reduces the accuracy of the detection results.

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
  • Pronunciation detection method and device, computer equipment and storage medium
  • Pronunciation detection method and device, computer equipment and storage medium
  • Pronunciation detection method and device, computer equipment and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0082] Aiming at the two aspects of the user's individual pronunciation characteristics and the feedback of the user's pronunciation errors, the embodiment of the present disclosure provides a pronunciation detection method, see Figure 1b As shown, it is a flowchart of a pronunciation detection method provided by an embodiment of the present disclosure, the method includes steps S101-S106, wherein:

[0083] S101: For any target user, acquire audio data of the target user.

[0084] In this step, the device receiving the audio data may be the terminal device 11 mentioned above, such as a computer, mobile phone, tablet computer and other devices installed with an evaluation client. During specific implementation, the client collects the audio data of the text read by the target user by calling the microphone of the terminal device. The audio data includes phonemes, and the client sends the audio data to the server to detect whether the reading is accurate or not.

[0085] Of co...

Embodiment 2

[0134] refer to Figure 6 As shown, it is a schematic diagram of a pronunciation detection device provided by an embodiment of the present disclosure, which includes: an extraction unit 601, a decoding unit 602, a first determination unit 603, a second determination unit 604, and a detection unit 605; wherein,

[0135] An extraction unit 601, configured to acquire audio data of the target user for any target user, the audio data including phonemes;

[0136] The decoding unit 602 is configured to use a pre-built network to decode each phoneme included in the audio data to obtain a time boundary corresponding to the phoneme, and the network is constructed using text information corresponding to the audio data of;

[0137] The first determining unit 603 is configured to use a phoneme coding model to encode the phonemes with determined time boundaries, and determine a first phoneme vector corresponding to each phoneme, wherein the phoneme coding model is based on the audio sample...

Embodiment 3

[0151] Based on the same technical idea, the embodiment of the present application also provides a computer device. refer to Figure 7 As shown, it is a schematic structural diagram of a computer device provided by the embodiment of the present application, including a processor 701 , a memory 702 , and a bus 703 . Among them, the memory 702 is used to store execution instructions, including a memory 7021 and an external memory 7022; the memory 7021 here is also called an internal memory, and is used to temporarily store calculation data in the processor 701 and exchange data with an external memory 7022 such as a hard disk. The processor 701 exchanges data with the external memory 7022 through the memory 7021. When the computer device is running, the processor 701 communicates with the memory 702 through the bus 703, so that the processor 701 executes the execution instructions mentioned in the above method embodiments .

[0152] Embodiments of the present disclosure furthe...

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 pronunciation detection method and device, computer equipment and a storage medium; and the method comprises the steps: obtaining audio data of a target user for any target user; for each phoneme contained in the audio data, decoding the phoneme by using a pre-constructed network to obtain a time boundary corresponding to the phoneme; respectively encoding each phoneme with the determined time boundary by using a phoneme encoding model, and determining a first phoneme vector corresponding to each phoneme; for each phoneme, determining the distance between a first phoneme vector and a second phoneme vector corresponding to the phoneme, the second phoneme vector being a vector corresponding to the phoneme obtained in the phoneme coding model training process; and detecting the audio data according to the distance between the first phoneme vector and the second phoneme vector corresponding to each phoneme. According to the embodiment of the disclosure, personalized detection is carried out according to the pronunciation characteristics of each user, so that the accuracy of the pronunciation detection result is improved.

Description

technical field [0001] The present disclosure relates to the technical field of audio detection, and in particular, to a pronunciation detection method, device, computer equipment and storage medium. Background technique [0002] With the rise of online education services, users read texts online, and the client encodes the corresponding audio data, and the server detects the received audio data to determine whether the user's reading is accurate. [0003] At present, when a user reads English or Chinese text aloud, the server usually extracts features that characterize the pronunciation characteristics from the user's voice, scores the pronunciation or classifies it as correct or incorrect, and sets a threshold based on the pronunciation score or according to the classification. As a result, it can be judged whether the pronunciation is correct. [0004] However, this judgment of right and wrong based on the pronunciation characteristics of a single speech is limited by th...

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
IPC IPC(8): G10L25/60G10L25/69G09B5/04
CPCG09B5/04G10L25/60G10L25/69
Inventor 蒋成林梁球斌其他发明人请求不公开姓名
Owner 北京乐学帮网络技术有限公司
Features
  • Generate Ideas
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More