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

Phase-sensitive gated multi-scale dilated convolutional network speech enhancing method and system

A convolutional network, phase-sensitive technology, applied in speech analysis, instruments, etc., can solve the problems of voice trembling, voice distortion, etc., to achieve the effect of improving the effect, avoiding voice distortion, and good voice intelligibility

Pending Publication Date: 2021-02-02
SHENZHEN INSTITUTE OF INFORMATION TECHNOLOGY
View PDF0 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, when our voice products face harsher acoustic scenarios, such as the signal-to-noise ratio is lower than 0dB, or the noise signal completely submerges the voice signal in a local time, it is not possible to only enhance the amplitude of the voice signal. Ensure that the enhanced voice has good speech intelligibility, and there may even be some voice distortion problems such as trembling and humming

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
  • Phase-sensitive gated multi-scale dilated convolutional network speech enhancing method and system
  • Phase-sensitive gated multi-scale dilated convolutional network speech enhancing method and system
  • Phase-sensitive gated multi-scale dilated convolutional network speech enhancing method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] The present invention will be further described below in conjunction with the description of the drawings and specific embodiments.

[0020] A phase-sensitive gated multi-scale atrous convolutional network speech enhancement method, which aims to use the neural network model to construct the mapping relationship between the complex spectrum of the speech signal, and process the real and imaginary part of the noisy speech spectrum after time-frequency analysis. Mapping, the enhanced real and imaginary part spectrum is obtained, and restored to the enhanced time-domain speech signal. The processing flow of the whole algorithm is as follows figure 1 As shown, the dotted line part is the gated multi-scale atrous convolutional network structure designed in the present invention, which is the core module of the whole algorithm. Noise reduction processing of real and imaginary part spectrum.

[0021] Such as figure 1 As shown, the noisy speech signal is first processed by f...

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 phase-sensitive gated multi-scale dilated convolutional network speech enhancing method. The method comprises the following steps: constructing a mapping relationship betweencomplex frequency spectrums of speech signals by using a neural network model, mapping a real and imaginary part frequency spectrum of noisy speech subjected to time-frequency analysis processing to obtain an enhanced real and imaginary part frequency spectrum, and recovering the spectrum into an enhanced time domain voice signal. The invention also provides a phase-sensitive gating multi-scale dilated convolutional network speech enhancing system. The method has the beneficial effects that: the method improves the speech enhancement effect, guarantees that the enhanced speech has good speechintelligibility, and better avoids the problem of speech distortion.

Description

technical field [0001] The invention relates to a speech enhancement method, in particular to a phase-sensitive gated multi-scale hole convolution network speech enhancement method and system. Background technique [0002] Early auditory experiments have shown that when the signal-to-noise ratio is higher than 6dB, phase distortion has little effect on speech quality and intelligibility. Therefore, most current single-channel speech enhancement methods mainly perform noise reduction in the amplitude domain of the speech signal. processing, and directly use the noisy phase to reconstruct the speech signal. However, when our voice products are faced with harsher acoustic scenarios, such as signal-to-noise ratios below 0dB, or when noise signals completely submerge voice signals in a local time, it is not possible to only enhance the amplitude of voice signals. Ensure that the enhanced voice has good speech intelligibility, and there may even be some voice distortion problems ...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G10L21/02G10L25/30
CPCG10L25/30Y02T10/40
Inventor 刘明周彦兵唐飞周小明赵学华
Owner SHENZHEN INSTITUTE OF INFORMATION TECHNOLOGY
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