Supercharge Your Innovation With Domain-Expert AI Agents!

Time domain flexible vibration sensor speech enhancement method and system

A vibration sensor and voice enhancement technology, applied in the field of vibration sensors and flexible vibration sensors, can solve the problem of the upper limit of enhancement performance improvement, and achieve the effect of improving voice quality, strong interpretability, and enhanced performance improvement.

Pending Publication Date: 2022-02-18
中国人民解放军火箭军士官学校
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Since the phase spectrum used is suboptimal, there is an upper limit to the overall enhancement performance

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
  • Time domain flexible vibration sensor speech enhancement method and system
  • Time domain flexible vibration sensor speech enhancement method and system
  • Time domain flexible vibration sensor speech enhancement method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0058] In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application, and are not intended to limit the present application.

[0059] Assume that the voice of the flexible vibration sensor is x(t)∈R T , the corresponding air-conducted speech is y(t)∈R T , to enhance it, the goal is to find an enhanced function f( ) so that the following formula holds:

[0060] y(t)≈f(x(t))

[0061] Where T is the length of speech time.

[0062] figure 2 Magnitude spectra of flexible vibration sensor speech and corresponding air-conducted speech are shown. It can be seen that the low frequencies of the air-conducted voice and the flexible vibration sensor voice have a high similarity, but not exactly the same...

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 neural network (DPT-EGNet) based on dual-path Transformer and in combination with equalization-generation component prediction. The neural network is used for speech enhancement of a flexible vibration sensor in a time domain and comprises a preprocessing module, a dual-path Transformer module, an equalization module, a generation module and a post-processing module. The dual-path Transformer module is used for learning local and global context relationships of a long-term voice sequence; and based on analysis of a voice loss mechanism of the flexible vibration sensor, the equalization module and the generation module are provided, and learned context information is used for prediction of an equalization coefficient and a loss component. The equalization coefficient and the output of the preprocessing module are subjected to point multiplication, a lost component is added, and finally the result is sent to the post-processing module to realize reconstruction from voice features to voice waveforms. Experiments prove that compared with other popular time-frequency domain models, the provided network can obtain better enhancement performance and has very low model complexity.

Description

technical field [0001] The invention belongs to the field of vibration sensors, in particular to the field of flexible vibration sensors, and in particular relates to a time-domain flexible vibration sensor speech enhancement method and system combining dual-path Transformer and equalization-generating component prediction. Background technique [0002] Different types of flexible vibration sensors can be realized based on piezoelectric effect, piezoresistive effect, piezoelectric capacitive effect, electromagnetic effect, etc. This sensor has been studied to capture the vibration of the throat to obtain speech signals. like figure 1 Shown is a flexible vibration sensor based on electromagnetic effects. Like the throat microphone of rigid material, this sensor can also block the noise transmitted in the air and has natural anti-noise performance, but because the material is soft and comfortable to wear, it is a potential future wearable device solution. However, due to the...

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): G10L21/02G10L21/0232G10L21/0264G10L21/057G10L21/0224
CPCG10L21/0232G10L21/057G10L21/0264G10L21/0224
Inventor 郑昌艳杨吉斌徐利国崔洪亮刘宁葛含黄显
Owner 中国人民解放军火箭军士官学校
Features
  • R&D
  • 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

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com