CNN-DNN hybrid neural network based noise reduction method

A hybrid neural network and noise reduction technology, applied in the field of noise reduction, can solve problems such as difficult and unsatisfactory effects

Active Publication Date: 2019-12-24
XI'AN POLYTECHNIC UNIVERSITY
View PDF9 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

These traditional algorithms have their limitations and compl

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
  • CNN-DNN hybrid neural network based noise reduction method
  • CNN-DNN hybrid neural network based noise reduction method
  • CNN-DNN hybrid neural network based noise reduction method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0057] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0058] A kind of noise reduction method based on CNN-DNN mixed neural network of the present invention, such as figure 1 Shown, specifically follow the steps below:

[0059] Step 1, establish a CNN-DNN hybrid neural network noise reduction model;

[0060] The front section of the CNN-DNN hybrid neural network noise reduction model is composed of a ResNet residual network. The ResNet residual network includes 10 residual units connected in sequence. The residual units are connected by a jump method, starting from the first residual At the beginning of the unit, each residual unit is activated by the ReLU nonlinear activation function; the 10 residual units are divided into two groups in order, and a shortcut is added between the two residual units in each group to form a residual module; the number of nodes in the first residual module is 64...

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 CNN-DNN hybrid neural network based noise reduction method. The method is implemented by the following steps that 1, a CNN-DNN hybrid neural network noise reduction model is established; 2, a training set is established for training the CNN-DNN hybrid neural network noise reduction model established in the step 1; and 3, a speech signal needing to be subjected to noise reduction is input to the trained CNN-DNN hybrid neural network noise reduction model in the step 3, and a clean speech signal spectrum is output. The CNN-DNN hybrid neural network based noise reductionmethod has better automatic identification separation and removal capabilities on transient noise and non-transient noise.

Description

technical field [0001] The invention belongs to the technical field of noise reduction methods, and relates to a noise reduction method based on a CNN-DNN hybrid neural network. Background technique [0002] In recent years, the problem of human aging has become more and more prominent, and the hearing impairment of the elderly has also received widespread attention. The research and development of hearing aids has begun to receive widespread attention. The hearing-impaired can hear high-definition audio information and improve the quality of hearing, and the removal of noise is particularly important for the hearing-impaired. [0003] At present, there are two types of noise for hearing aids, one is transient noise and the other is non-transient noise. There are two research methods for transient noise suppression (Transient Noise Suppression, TNS). The Optimally Modified-Log Spectral Amplitude (OM-LSA) algorithm, by considering the prior signal-to-noise ratio and the unce...

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): G10L21/0208G10L21/0216G10L21/0232G10L25/30
CPCG10L21/0208G10L21/0216G10L21/0232G10L25/30
Inventor 李云红穆兴张秋铭刘旭东何琛
Owner XI'AN POLYTECHNIC UNIVERSITY
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