Twin network voiceprint recognition method based on 3D convolution

A twin network, voiceprint recognition technology, applied in biological neural network models, neural learning methods, speech analysis, etc., can solve the problems of low recognition rate, ignore the spatial and temporal characteristics of speech information, etc., to improve the accuracy rate, Relevance-enhancing effects

Active Publication Date: 2020-04-21
中国人民解放军空军研究院通信与导航研究所 +1
View PDF9 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the one-dimensional convolution and two-dimensional convolution methods ignore the spatial and temporal features of speech information, and the recognition rate is not high.

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
  • Twin network voiceprint recognition method based on 3D convolution
  • Twin network voiceprint recognition method based on 3D convolution
  • Twin network voiceprint recognition method based on 3D convolution

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0034] The present invention will be described in detail below with reference to the drawings and specific embodiments.

[0035] The embodiment of the invention discloses a twin network based on 3D convolution for voiceprint recognition. The Siamese-Net network is abbreviated as Sia-Net network, and includes: a feature extraction unit: used to convert audio data into a three-dimensional tensor, The three-dimensional tensor is the MFLC feature.

[0036] Sia-Net network: used to process the MFLC features, shorten the feature distance of data between the same speaker, and increase the feature distance of data between different speakers. This distance is the Euclidean distance. CNN network: used to build a model library for each speaker. Prediction unit: used to test and determine the speaker identity of audio data.

[0037] The Sia-Net network: There are two, each of the Sia-Net networks includes: three 3D convolutional layers, one pooling layer, four 3D convolutional layers, one connec

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 discloses a voiceprint recognition twin network based on 3D convolution, and the network comprises a feature extraction unit, an Sia-Net network, a CNN network and a prediction unit. Thefeature extraction unit is used for converting audio data into a three-dimensional tensor, and the three-dimensional tensor is an MFLC feature. And the Sia-Net network is used for processing the MFLCfeature, shortening the feature distance of the data between the same speaker and increasing the feature distance of the data between different speakers. And the CNN network is used for establishinga model library of each speaker. And the prediction unit is used for testing the speaker identity of the audio data. When the network is used for voiceprint recognition, voice information can be fullysupervised and learned, time domain information of the voice information can also be considered, and the accuracy of voiceprint recognition is further improved.

Description

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

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
Owner 中国人民解放军空军研究院通信与导航研究所
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