TOPK-based multi-channel sound source effective signal screening system and TOPK-based multi-channel sound source effective signal screening method

An effective signal and screening method technology, applied in the field of communication, can solve problems such as inability to separate and screen out effective signals

Pending Publication Date: 2021-03-26
CHONGQING COLLEGE OF ELECTRONICS ENG
View PDF11 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a multi-channel sound source effective signal screening system based on TOPK, which solves the technical problem that the prior art cannot separate and screen effective signals from the mixed sound formed by multi-channel voice and multi-channel background sound

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
  • TOPK-based multi-channel sound source effective signal screening system and TOPK-based multi-channel sound source effective signal screening method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0042] The embodiment is basically as attached figure 1 Shown: includes:

[0043] The input unit is used to input the mixed sound formed by N channels of voice and N channels of background sound;

[0044] The pre-judgment unit is used to use the VAD algorithm to predict each voice one by one: if the voice is normal, the VAD value is 1; if the voice output state is uncertain, the VAD value is 0; if there is no voice output, the VAD value is -1;

[0045] The grading unit is used to divide the voice signal with a VAD value of 1 into grades 1 to 10 by using the AMDF algorithm, and assign values;

[0046] The screening unit is used to receive N channels of voice signals, and filter out M channels of the strongest signals for the N channels of buffered signals at each time according to the set buffer amount;

[0047] The error correction unit is used to make use of the signal correlation and use the FEC algorithm to supplement the front-end voice signal lost due to the time delay of...

Embodiment 2

[0064] The only difference from Embodiment 1 is that in S5, the error correction unit first performs pre-processing on the strongest signals of M channels, including pre-emphasis processing, windowing processing and endpoint detection, and then performs sound processing on the strongest signals of M channels one by one. Fingerprint recognition, retain the strongest signal that matches the preset voiceprint characteristics, and delete the strongest signal that does not match the preset voiceprint characteristics, thereby removing noise.

[0065] Finally, wavelet decomposes the strongest signals of M channels one by one to obtain the wavelet signal sequence, and obtains the effective speech signal according to the wavelet signal sequence. Specifically, for the M strongest signals, wavelet decomposition is performed on the audio frame signals one by one, so as to obtain multiple wavelet decomposition signals corresponding to each audio frame signal, each wavelet decomposition sign...

Embodiment 3

[0067] The only difference from Embodiment 2 is that, before classifying the multi-channel sound sources, the multi-channel speech is firstly complemented. Specifically, the text corpus related to classroom live teaching is pre-stored on the server. When the network signal is not good, the voice signal may be intermittent, and part of the voice signal is missing. At this time, it is necessary to correct the missing voice The signal is complemented.

[0068] First of all, when the network signal is not good, extract the speech signal of the front and back parts of the intermittent speech signal, and convert it into text, and use the semantic recognition algorithm combined with the text corpus to make corresponding text for the missing speech signal content filling. That is, fill in the text content corresponding to the missing speech signal according to the semantic understanding, and convert the text content into a speech signal, so as to realize the completion of the discont...

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, in particular to a TOPK-based multi-channel sound source effective signal screening system, which comprises an input unit used for inputting mixed sound formed by N channels of voice and N channels of background sound; a pre-judging unit used for pre-judging each path of voice one by one by adopting a VAD algorithm; a grading unit usedfor sequentially dividing the voice signals with the VAD value of 1 into 1-10 grades by adopting an AMDF algorithm and assigning values; a screening unit used for screening M paths of strongest signals for the N paths of cache signals at each moment according to a set cache amount; an error correction unit used for complementing the screened M paths of strongest signals into lost front-end voicesignals due to time delay by using signal correlation and adopting an FEC algorithm; and an output module used for outputting the supplemented M paths of strongest signals. According to the method, aVAD algorithm, an AMDF algorithm and an FEC algorithm are combined, and the technical problem that effective signals cannot be separated and screened from mixed sound formed by multiple paths of voiceand multiple paths of background sound in the prior art is solved.

Description

technical field [0001] The invention relates to the field of communication technology, in particular to a TOPK-based multi-channel sound source effective signal screening system and method. Background technique [0002] With the wide application of mobile smart devices and the continuous development of technology, voice has gradually become a means of human-computer interaction. However, there are various sound sources in an actual environment, and the sounds from different sound sources will interfere with each other to form multiple sound sources, thereby affecting user experience. Therefore, it is necessary to take necessary measures to screen multi-channel audio sources and select effective sound signals, such as TOPK, that is, sorting algorithm for selection. [0003] For example, the patent CN106484833A discloses a sound source screening method, including the steps of: obtaining at least one search information for searching audio files transmitted by an audio playback...

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): G10L21/0272G10L17/00G10L25/78
CPCG10L21/0272G10L25/78
Inventor 陶亚雄王彬
Owner CHONGQING COLLEGE OF ELECTRONICS ENG
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