Method and system for controlling speaker tracking in a video conferencing system

1. A video conferencing device ( 250 , 350 ) for video conferencing between at least one local participant ( 192 ) and a remote participant ( 196 ), the video conferencing device comprising:

a video camera ( 120 ) providing a local video input signal ( 122 );

a microphone array ( 124 ) providing a local audio input signal ( 126 ),

wherein the video conference device ( 250 , 350 ) excludes a loudspeaker ( 128 ) of a computing system ( 210 ) that outputs audio based on a remote audio input signal ( 116 ); and

a speaker tracker ( 260 , 360 ) configured to identify a local speaker ( 194 ) from the at least one local participant using a sound source localizer ( 270 ),

wherein the video conferencing device processes the local video input signal without audio from the loudspeaker ( 128 ), based on the video conferencing device receiving a signal from the computing system ( 210 ), the signal dependent on a loopback audio output signal ( 332 ) indicating that the remote participant is speaking, wherein the loopback audio output signal ( 332 ) comprises a copy of a local output audio signal ( 130 ) provided to the loudspeaker ( 128 ) of the computing system ( 210 ).

2. The video conferencing device of claim 1 , wherein, based at least in part on an absence of the signal from the computing system:

the video conferencing device activates the sound source localizer,

the speaker tracker identifies the local speaker for the at least one local participant, and

the video conferencing device processes the local video input signal using the local speaker.

3. The video conferencing device of claim 2 , wherein the video conferencing device processes the local video input signal using the local speaker by performing operations comprising:

generating a remote video output signal ( 112 ), based at least in part on the local video input signal, and

identifying the local speaker, in the remote video output signal.

4. The video conferencing device of claim 1 , wherein the signal from the computing system is a remote speaker detected flag ( 242 ), generated by the computing system.

5. The video conferencing device of claim 1 , further comprising:

a communication interface for receiving the loopback audio output signal ( 332 ) generated by the computing system as the signal from the computing system, and

a remote speaker detector ( 340 ) configured to:

analyze the loopback audio output signal for a speech signal, and

based at least in part on a detection of the speech signal, provide a remote speaker detected flag ( 342 ) to the sound source localizer to cause deactivation of the sound source localizer.

6. The video conferencing device of claim 1 , further comprising:

a remote speaker detector ( 340 ) configured to:

compare a signal energy of the loopback audio output signal against a predetermined threshold, in a speech frequency band.

7. A method for operating a video conferencing system, the method comprising:

during a video conference between at least one local participant ( 192 ) and a remote participant ( 196 ), wherein the video conferencing system excludes a loudspeaker ( 128 ) of a computing system ( 210 ) that outputs audio based on a remote audio input signal ( 116 ) received from a remote device of the remote participant:

receiving ( 402 ) a local audio input signal from a microphone array of the video conferencing system and a local video input signal from a video camera of the video conferencing system;

receiving ( 404 ) a remote audio input signal from the remote participant;

generating ( 406 ), from the remote audio input signal, a local audio output signal to drive the loudspeaker of the video conferencing system;

generating ( 408 ) a loopback audio output signal from the local audio output signal using an audio loopback interface, wherein the loopback audio output signal ( 332 ) comprises a copy of a local output audio signal ( 130 ) provided to the loudspeaker ( 128 ) of the computing system ( 210 );

analyzing ( 410 ) the loopback audio output signal for a speech signal;

based at least in part on a detection of the loopback audio output signal, processing ( 418 ) the local video input signal without audio from the loudspeaker.

8. The method of claim 7 , wherein the processing the local video input signal comprises:

generating a remote video output signal, based at least in part on the local video input signal; and

identifying a local speaker, in the remote video output signal.

9. The method of claim 7 ,

wherein the generating the loopback audio output signal from the local audio output signal is performed by a computing system,

wherein the computing system comprises the audio loopback interface,

wherein the computing system is separate from a video conferencing device, and interfacing with the video conferencing device using a communication interface,

wherein the video conferencing device comprises the video camera and the microphone array, and

wherein the remote audio input signal is inaccessible by the video conferencing device.

10. The method of claim 9 , further comprising:

generating, by the computing system, a remote speaker-detected flag, and

deactivating a sound source localization based at least in part on the remote speaker-detected flag.

11. The method of claim 9 , further comprising:

providing, by the computing system, the loopback audio output signal to the video conferencing device;

wherein the video conferencing device is configured to:

perform the analyzing the loopback audio output signal for the speech signal;

perform deactivating of the sound source localization based at least in part on the detection of the speech signal on the loopback audio output signal.

12. The method of claim 9 , wherein the audio loopback interface is provided by an operating system of the computing system.

13. The method of claim 7 , wherein analyzing the loopback audio output signal for the speech signal comprises:

comparing a signal energy of the loopback audio output signal against a predetermined threshold, in a speech frequency band.

14. A method for operating a computing device ( 210 , 310 ) of a video conferencing system ( 110 ), the method comprising, during a video conference between at least one local participant ( 192 ) and a remote participant ( 196 ), wherein the video conferencing system ( 110 ) excludes a loudspeaker ( 128 ) of a computing device ( 210 , 310 ) that outputs audio based on a remote audio input signal ( 116 ) received from a remote device of the remote participant:

receiving ( 404 ) a remote audio input signal ( 116 ) from the remote participant;

generating ( 406 ), from the remote audio input signal, a local audio output signal;

generating ( 408 ) a loopback audio output signal ( 232 , 332 ) from the local audio output signal ( 130 ) using an audio loopback interface ( 230 , 330 ), wherein the loopback audio output signal ( 332 ) comprises a copy of the local output audio signal ( 130 ) provided to the loudspeaker ( 128 ) of the computing device ( 210 , 310 );

depending on the loopback audio output signal ( 232 ), providing a signal ( 242 , 332 ) to the video conferencing system ( 110 ), wherein the signal controls a sound source localization by the video conferencing system ( 110 ) without audio from the loudspeaker ( 128 ).

15. The method of claim 14 , further comprising:

analyzing ( 410 ) the loopback audio output signal for a speech signal.

16. The method of claim 15 , further comprising:

based at least in part on a detection of the speech signal, ( 412 - 416 ), providing a remote speaker detected flag ( 242 ) to the video conferencing system.

17. The method of claim 14 , further comprising: wherein analyzing the loopback audio output signal for the speech signal comprises:

comparing a signal energy of the loopback audio output signal against a predetermined threshold, in a speech frequency band.

18. The method of claim 14 , wherein the audio loopback interface is provided by an operating system of the computing device ( 210 , 310 ).

19. The method of claim 14 , further comprising:

driving a loudspeaker ( 128 ) of the computing device, using the local audio output signal.