Satellite microphone array for video conferencing
Speakers are identified based on sound origination detection through use of infrared detection of satellite microphones, estimation of distance between satellite microphones and base unit utilizing captured audio, and/or estimation of satellite microphone orientation utilizing captured audio. Multiple sound source localization results are combined to enhance sound source localization and/or active speaker detection accuracy.
1. A method to be executed at least in part in a computing device for enhanced detection of an active human speaker through sound source localization in a video conference system, the method comprising:
estimating a first orientation and a first distance of a satellite microphone module based on detecting a sound emitted from the base unit at the satellite microphone module;
estimating a second orientation and a second distance of the satellite microphone module based on detecting an infrared light emitted from the satellite microphone module at the base unit; and
determining a location of the satellite microphone module by comparing the estimated first and second orientation and the first and second distance of the satellite microphone module.
2. The method of claim 1 , further comprising:
estimating a third orientation and a third distance of the satellite microphone module based on detecting another sound emitted from the satellite microphone module at the base unit; and
determining the location of the satellite microphone module by comparing the estimated first and third orientation and the first and third distance of the satellite microphone module.
3. The method of claim 1 , further comprising:
causing the infrared light to be emitted from the satellite microphone module by one of:
an infrared Light Emitting Diode (LED) located at a top of the satellite microphone module; and
an infrared LED located inside the satellite microphone module and a light pipe coupled to the infrared LED, wherein the light pipe is configured to emit the infrared light from the satellite microphone module; and
detecting the emitted infrared light through one of a dedicated infrared detection device and a video capture device for capturing video images of participants on the base unit.
4. The method of claim 2 , wherein the third distance is estimated based on at least one of:
a geometric computation based on base unit height and field of view parameters; and
a time of arrival of sound generated by a speaker on the satellite microphone module unit.
5. The method of claim 1 , wherein the first distance is estimated based on a time of arrival of sound at the satellite microphone module generated by a loudspeaker on the base unit by:
synchronizing the loudspeaker on the base unit with the satellite microphone module; and
computing the distance of the satellite microphone module d as c*T, where c is the speed of sound in air and T is the time of arrival.
6. The method of claim 5 , further comprising:
employing one of: an audible tone, an inaudible tone, and regular speech transmitted from the speaker at the base unit for detecting at each satellite microphone module and computing the distance d.
7. The method of claim 1 , wherein estimating the third distance of the satellite microphone module comprises:
computing the distance d for the satellite microphone module detected through infrared emission as d=h*f/u, where h is a height of a video capture device detecting the infrared light, f is a focal length of the video capture device, and u is an image plane width.
8. The method of claim 1 , wherein estimating the first orientation of the satellite microphone module comprises:
computing a probability distribution function for the satellite microphone module based on detection of sound from a loudspeaker of the base unit by the satellite microphone module; and
determining a peak of the probability distribution function for the satellite microphone module, the peak indicating a direction toward the base unit.
9. The method of claim 8 , wherein the orientation of the satellite microphone module is estimated by θ=v*360/x max , where θ is an orientation angle with respect to a predetermined reference axis, v is an x image coordinate, and x max is a maximum x coordinate in an image sensor of the video capture device on the base unit.
10. The method of claim 1 , wherein each satellite microphone module of a plurality of satellite microphone modules connected to the base unit is configured to transmit an infrared signal with a distinct frequency, and wherein each of the distinct frequencies are about a cut-off frequency of the video capture device.
11. The method of claim 1 , wherein a microphone on the base unit is used for estimating a location and an orientation of the satellite microphone modules only, and microphones on the satellite microphone modules are used for capturing audio from participants of the video conference.