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

Speaker recognition system with high recognition degree

A technology of speaker recognition and speaker, applied in the field of speaker recognition system, can solve the problem of poor accuracy of speaker recognition system, achieve the effect of reducing false detection and improving efficiency

Pending Publication Date: 2022-04-01
江苏清微智能科技有限公司
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, restricted by various uncertain factors, the accuracy of the current speaker recognition system is still not satisfactory

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
  • Speaker recognition system with high recognition degree
  • Speaker recognition system with high recognition degree
  • Speaker recognition system with high recognition degree

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0052] In order to make the purpose and features of the present invention more obvious and understandable, the technical solution will be described in detail below through embodiments and in conjunction with the accompanying drawings.

[0053] Such as figure 1 As shown, a speaker recognition system with a high degree of recognition includes: a speech collection module, a feature extraction module, a model training module and a speaker recognition module, wherein the speech collection module is used to collect the speaker The speech, the feature extraction module is used to extract the speaker features in the speech, the model training module is based on the improved TDNN model, and uses data enhancement to train the speaker recognition model, the speaker recognition The module uses similarity comparison and speaker recognition based on the similarity comparison score.

[0054] The working process of the system includes:

[0055] Step 1. First, the voice collection module col...

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 relates to the field of neural networks, machine learning and information processing, and discloses a speaker recognition system with high recognition degree. According to the system, firstly, a feature extraction module receives a voice signal output by a voice acquisition module, speaker features in voice are extracted through an improved feature extraction algorithm, then, a model training module receives the extracted speaker features to train an improved TDNN model, and finally, a speaker recognition module receives an output model to recognize the speaker. And speaker recognition is carried out on the input voice. According to the system, by improving a feature extraction algorithm and a speaker recognition algorithm, the problem that speaker features are difficult to extract in a complex noise environment is well solved, and the accuracy of speaker recognition is remarkably improved.

Description

technical field [0001] The invention relates to the fields of neural network, machine learning and information processing, in particular to a speaker recognition system with high recognition degree. Background technique [0002] Speaker recognition systems are widely used in different fields such as security, finance, and social security. However, restricted by various uncertain factors, the accuracy of current speaker recognition systems is still unsatisfactory. [0003] These uncertain factors include complex and changeable background noise, physiological fluctuations of the speaker itself, etc. Therefore, research on more effective feature extraction algorithms and more robust speaker recognition algorithms in complex backgrounds is of great importance to improving speech The accuracy of the human recognition system is of great significance. Contents of the invention [0004] Aiming at the difficulties and deficiencies existing in the current speaker recognition syste...

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): G10L17/02G10L17/04G10L17/18G10L25/24G06K9/62
Inventor 孔天龙宋丹丹欧阳鹏
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 Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com