Noise reduction method for multi-channel information transmission of enterprise multi-party meeting

An information transmission, multi-channel technology, applied in the field of communication noise reduction processing, can solve the problems of suppression, poor meeting experience, inability to distinguish the voice of the target speaker, and the effect of improving the experience and recovery.

Inactive Publication Date: 2021-11-09
辰风策划(深圳)有限公司
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the above-mentioned shortcomings existing in the prior art, the present invention provides a noise reduction method for multi-channel information transmission for multi-party meetings in enterprises, which solves the problem that the noise reduction method is processed by the front-end of the voice signal, and cannot distinguish the voice of the target speaker from the speech Therefore, the voices other than the speaker cannot be effectively suppressed, and the existing noise reduction technology can only reduce the human voice noise interference by prohibiting other people from speaking when the speaker is speaking. The progress of the meeting had an impact, and it was necessary to frequently prohibit other people from speaking, resulting in a very bad meeting experience

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
  • Noise reduction method for multi-channel information transmission of enterprise multi-party meeting
  • Noise reduction method for multi-channel information transmission of enterprise multi-party meeting
  • Noise reduction method for multi-channel information transmission of enterprise multi-party meeting

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0040] The noise reduction method for the multi-channel information transmission of the enterprise multi-party meeting in this embodiment, such as figure 1 shown, including the following steps:

[0041] S1: Determine the speaker through face recognition, and then collect the speaker's voice data through the microphone array;

[0042] S2: Use the neural network to calculate the sound interval divided by the preset duration of the sound data, and calculate the noise covariance matrix according to the noise probability in the sound interval;

[0043] S3: Calculate the weight of merging multiple channels according to the covariance matrix of the noise in S2 and the eigenvector of the covariance matrix of the noise;

[0044] S4: Then the microphone array in the conference machine collects the voice of the speaker, and inputs the collected voice into the voice noise reduction module;

[0045] S5: According to the weight value in S3, it is judged whether the voice segment is the sp...

Embodiment 2

[0051] like Figure 1 to Figure 2 As shown, the microphone array collects the voice data of the speaker, and uses a single filter or multiple filters to pass through the fast Fourier transform, and then converts the time domain signal collected in the multi-channel signal into a frequency domain signal to obtain each frequency band signal of.

[0052] In this embodiment, before collecting the voice data of the speaker, the speaker reads a paragraph of text through the microphone, and then the enterprise login device receives the voice data of the speaker.

[0053] The voice data of this embodiment includes voice data and background sound, and the voice data of the speaker is obtained from the voice data through the voiceprint information; wherein, the voice features are extracted from the voice data to form a recognition set in the voiceprint information, and then the enterprise terminal The device recognizes the speech data in the sound data through the recognition set.

[...

Embodiment 3

[0064] The noise reduction method for multi-channel information transmission of enterprise multi-party meeting in this embodiment is used to assist the noise reduction method of embodiment 1 and embodiment 2 for multi-channel information transmission of enterprise multi-party meeting, which is used for registration And enter the identity information of the speaker, such as figure 2 As shown, the method of registering and entering the identity information of the speaker includes the following sub-steps:

[0065] S11: The speaker takes an image through the camera on the enterprise terminal device and registers with his real name;

[0066] S12: The registration database generated according to the registration in S11 includes the speaker's image and user information;

[0067] S13: The speaker logs in the personal user information through the enterprise login device, and performs face recognition authentication to obtain the speaking login authority.

[0068] The enterprise term...

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 technical field of communication noise reduction processing, in particular to a noise reduction method for multi-channel information transmission of enterprise multi-party meeting. The noise reduction method comprises the following steps: determining a speaker through a face recognition mode, and then collecting sound data of the speaker through a microphone array; calculating sound intervals divided by the sound data in a preset duration by using a neural network, and calculating a noise covariance matrix according to noise probabilities in the sound intervals; and according to the covariance matrix of the middle noise and the feature vector of the noise covariance matrix, calculating and combining the weight of multiple channels. Multi-channel signals are collected through the microphone array, time domain signals are converted into frequency domain signals through a single filter or a plurality of filter banks, so that signals of all frequency bands in a sound interval can be effectively reserved, noise probabilities existing on all the frequency bands are calculated through a neural network, the noise covariance matrix can be calculated more quickly and effectively, and the recovery of the original sound is improved.

Description

technical field [0001] The invention relates to the technical field of communication noise reduction processing, in particular to a noise reduction method for multi-channel information transmission used in multi-party meetings of enterprises. Background technique [0002] With the increasing frequency of business activities, the development of multi-party meetings in enterprises is becoming more and more extensive. In multi-party meetings, different people speak at different positions based on the microphone in the conference machine. Gather together for a multi-party meeting. As the name suggests, there will be many people speaking, and the speech needs to be transmitted at a certain distance and broadcast from the speaker. When the speaker is speaking, other people's voices, tables, chairs, and tables that have nothing to do with the meeting will be heard. All will bring noise, which is not conducive to the meeting personnel to obtain accurate conference voice information,...

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 & AuthorityApplications(China)
IPC IPC(8): G10L21/0208G10L21/0232H04R1/32H04R1/40H04L12/18
CPCG10L21/0208G10L21/0232H04R1/342H04R1/406H04L12/1813
Inventor徐员梅
Owner辰风策划(深圳)有限公司