Echo cancellation in online conference systems

Patent metadata

Field	Value
Publication number	US-11516579-B2
Application number	US-202117301677-A
Country	US
Kind code	B2
Filing date	Apr 12, 2021
Priority date	Apr 12, 2021
Publication date	Nov 29, 2022
Grant date	Nov 29, 2022

Abstract

An embodiment for cancelling echo in online conference systems is provided. According to some embodiments of the present disclosure, the computer-implemented method comprises, in response to an update of devices of participants in an online conference, dividing, by one or more processors, the devices in an online conference into a plurality of groups, wherein the devices located in a same physical location are divided into a same group. The method also comprises, in response to an update of the devices in an online conference, selecting at least one speaker of the devices in each of the plurality of groups as a representative speaker for each of the plurality of groups. The method further comprises forwarding audio data received from microphones of the devices in one of the plurality of groups to the respective representative speaker of other groups of the plurality of groups.

First claim

What is claimed is:

1. A computer-implemented method performed by one or more processors, the method comprising:
in response to an update of devices of participants in an online conference, dividing the devices in the online conference into a plurality of groups and selecting at least one speaker of the devices in each of the plurality of groups as a representative speaker for each of the plurality of groups, wherein the devices located in a same physical location are divided into a same group, wherein the selecting at least one speaker of the devices in each of the plurality of groups further comprises:
identifying an intensity of the audio data received from the devices in each group;
estimating relative distances between each participant and any other participant in each group based on the intensity;
obtaining a volume percentage of the speaker of each device;
calculating a weight for each speaker in the group based on the volume percentage of the speaker and the relative distances; and
selecting at least one speaker with a largest weight as the representative speaker for respective groups;

and
forwarding audio data received from the devices in one of the plurality of groups to the respective representative speaker of other groups of the plurality of groups.

2. The method of claim 1 , wherein the dividing of the devices in the online conference into the plurality of groups further comprises:
checking whether there are devices which are neighbors to each other;
in response to there being devices which are neighbors to each other, including the devices which are neighbors to each other into a group; and
in response to there being one or more devices which are not included in the group, including each of the one or more devices into an individual group.

3. The method of claim 2 , further comprising:
in response to receiving the audio data from each of the devices, obtaining a spectrum of the audio data for each device;
in response to there being two or more kinds of spectrums within a group, splitting the group into two or more groups correspondingly, each group having the same spectrum; and
in response to there being groups having the same spectrum, combining the groups into one group.

4. The method of claim 1 , wherein the dividing of the devices in the online conference into the plurality of groups further comprises:
in response to receiving the audio data from each of the devices, obtaining a spectrum of the audio data for each device; and
including one or more devices with the same spectrum into one group.

5. The method of claim 1 , wherein the forwarding of audio data received from the devices in one of the plurality of groups to the respective representative speaker of the other groups further comprises:
receiving the audio data from the devices in one group;
identifying a spectrum of the received audio data;
in response to there being audio data with a repetitive spectrum, removing the audio data with the repetitive spectrum;
composing the audio data without the repetitive spectrum as output audio data for the group; and
forwarding the output audio data to the representative speaker of the other groups.

6. The method of claim 1 , wherein the forwarding of the audio data received from the devices in one of the plurality of groups to the respective representative speaker of the other groups further comprises:
receiving the audio data from the devices in one group;
detecting whether there is a signal of a flag together with the received audio data;
in response to detecting there is no signal of the flag, adding the signal of the flag to the received audio data;
forwarding the received audio data with the signal of the flag to the representative speaker of the other groups; and
in response to detecting there is the signal of the flag, removing the received audio data.

7. A computer-implemented system, comprising:
at least one processing unit; and
a memory coupled to the at least one processing unit and storing instructions thereon, the instructions, when executed by the at least one processing unit, performing actions comprising:
in response to an update of devices of participants in an online conference, dividing the devices in the online conference into a plurality of groups and selecting at least one speaker of the devices in each of the plurality of groups as a representative speaker for each of the plurality of groups, wherein the devices located in a same physical location are divided into a same group, wherein the selecting at least one speaker of the devices in each of the plurality of groups further comprises:
identifying an intensity of the audio data received from the devices in each group;
estimating relative distances between each participant and any other participant in each group based on the intensity;
obtaining a volume percentage of the speaker of each device;
calculating a weight for each speaker in the group based on the volume percentage of the speaker and the relative distances; and
selecting at least one speaker with a largest weight as the representative speaker for respective groups;

and
forwarding audio data received from the devices in one of the plurality of groups to the respective representative speaker of other groups of the plurality of groups.

8. The computer-implemented system of claim 7 , wherein the dividing the participants devices in the online conference into a plurality of groups comprising:
checking whether there are devices which are neighbors to each other;
in response to there being devices which are neighbors to each other, including the devices which are neighbors to each other into a group; and
in response to there being one or more devices which are not included in the group, including each of the one or more devices into an individual group.

9. The computer-implemented system of claim 8 , further comprising:
in response to receiving the audio data from each of the devices, obtaining a spectrum of the audio data for each device;
in response to there being two or more kinds of spectrums within a group, splitting the group into two or more groups correspondingly, each group having the same spectrum; and
in response to there being groups having the same spectrum, combining the groups into one group.

10. The computer-implemented system of claim 7 , wherein the dividing of the devices in the online conference into the plurality of groups further comprises:
in response to receiving the audio data from each of the devices, obtaining a spectrum of the audio data for each device; and
including one or more devices with the same spectrum into one group.

11. The computer-implemented system of claim 7 , wherein the forwarding of audio data received from the devices in one of the plurality of groups to the respective representative speaker of the other groups further comprises:
receiving the audio data from the devices in one group;
identifying a spectrum of the received audio data;
in response to there being audio data with a repetitive spectrum, removing the audio data with the repetitive spectrum;
composing the audio data without the repetitive spectrum as output audio data for the group; and
forwarding the output audio data to the representative speaker of the other groups.

12. The computer-implemented system of claim 7 , wherein the forwarding of the audio data received from the devices in one of the plurality of groups to the respective representative speaker of the other groups further comprises:
receiving the audio data from the devices in one group;
detecting whether there is a signal of a flag together with the received audio data;
in response to d

Assignees

• IBM

Inventors

• Feng Kai
• Yang yun juan
• Zhu Xiao Zhen
• Dong Ming
• Cao Yang
• Jiang Jing Hua
• Wang yan hui
• Rui jia lei
• Li Qi

Classifications

• H04M3/002

  Applications of echo suppressors or cancellers in telephonic connections (in two-way loud-speaking telephone systems H04M9/08, echo suppressors or cancellers per se H04B3/20) · CPC title

• H04L12/1827

  Network arrangements for conference optimisation or adaptation · CPC title

• H04M9/08

  Two-way loud-speaking telephone systems with means for conditioning the signal, e.g. for suppressing echoes for one or both directions of traffic · CPC title

• H04M3/568

  audio processing specific to telephonic conferencing, e.g. spatial distribution, mixing of participants (echo suppression in two-way loud-speaking telephone systems H04M9/02; sound field processing per se H04S7/30) · CPC title

• G10L25/18

  the extracted parameters being spectral information of each sub-band · CPC title