Adjusting audio volume and quality of near end and far end talkers
An orchestrator associated with a collaboration application client executed by a near end device dynamically adapts volume level and/or other audio quality parameters to deliver a consistent voice experience to a collaboration participant. The orchestrator is informed by multiple machine learning engines collecting and analyzing inputs from one or more existing sensor-based functions embedded in the near end device. The orchestrator determine an audio configuration of the device and audio preferences of the user. Identities of far end participants are determined and their profiles are mapped against the user volume preferences. The orchestrator functions as an informing agent to the audio signal processing engine of the device, managing speaker output volume and microphone gain settings, based upon the machine learning engines and the sensor-based functions. The sensor based functions may detect proximity, head pose, gaze point, eye position, facial identities, mood, and so forth.
1. An audio control method, comprising:
detecting, by an orchestrator resource associated with a collaboration application, an audio configuration of a near end device associated with a near end participant of a collaboration session wherein the audio configuration includes a near end audio speaker and a near end microphone;
dynamically characterizing one Or more audio parameters associated with one or more far end participants at one or more far end locations; and
dynamically adjusting one or more audio parameters of the near end device during the collaboration session, wherein said adjusting is informed by:
a first output from a first machine learning engine maintaining an audio preference history of the near end participant, wherein the first output is responsive to a first set of one or more indicators sensed by the near end device; and
a second output from a second machine learning engine configured to map profiles of the one or more far end participants at the one or more far end locations against audio preference of the near end participant.
2. The audio control method of claim 1 , wherein dynamically adjusting one or more audio parameters comprises dynamically adjusting a gain level of at least one of: the audio speaker and the microphone.
3. The audio control method of claim 1 , wherein the audio preference history reflects audio parameter changes initiated by the near end participant during one or more previous collaboration sessions.
4. The audio control method of claim 3 , wherein the first machine learning engine collects and analyzes data indicative of the first set of one or more indicators corresponding to one or more of the audio parameter changes initiated by the near end participant.
5. The audio control method of claim 1 , wherein the first set of one or more indicators includes a proximity of the near end participant sensed by the near end device.
6. The audio control method of claim 1 , wherein the first set of one or more indicators includes at least one eye tracker indicator sensed by an eye tracker of the near end device.
7. The audio control method of claim 6 , wherein the at least one eye tracker indicator is selected from: a head pose, a gaze point, and an eye position of the near end participant.
8. The audio control method of claim 1 , wherein the second output from the second machine learning engine reflects at least one of: a number of far end participants at a particular far end location and a room type of a particular far end location.
9. The audio control method of claim 1 , further comprising: determining identities of the one or more far end participants includes invoking a participant list function exposed by an application program interface (API) of the collaboration application.
10. The audio control method of claim 1 , wherein the orchestrator dynamically manages adjustments of the one or more audio parameters based on a composite of the first input from the first machine learning engine and a second input from the second machine learning engine.
11. An information handling system, comprising:
a central processing unit coupled to a memory resource including instructions, executable by the processor, for performing audio control operations, wherein the audio control operations include:
detecting, by an orchestrator resource associated with a collaboration application, an audio configuration of a near end device associated with a near end participant of a collaboration session wherein the audio configuration includes a near end audio speaker and a near end microphone;
dynamically characterizing one or more audio parameters associated with one or more far end participants at one or more far end locations; and
dynamically adjusting one or more audio parameters of the near end device during the collaboration session, wherein said adjusting is informed by:
a first output from a first machine learning engine maintaining an audio preference history of the near end participant, wherein the first output is responsive to a first set of one or more indicators sensed by the near end device; and
a second output from a second machine learning engine configured to map profiles of one or more far end participants at one or more far end locations against audio preference of the near end participant.
12. The information handling system of claim 11 , wherein dynamically adjusting one or more audio parameters comprises dynamically adjusting a gain level of at least one of: the audio speaker and the microphone.
13. The information handling system of claim 11 , wherein the audio preference history reflects audio parameter changes initiated by the near end participant during one or more previous collaboration sessions.
14. The information handling system of claim 13 , wherein the first machine learning engine collects and analyzes data indicative of the first set of one or more indicators corresponding to one or more of the audio parameter changes initiated by the near end participant.
15. The information handling system of claim 11 , wherein the first set of one or more indicators includes a proximity of the near end participant sensed by the near end device.
16. The information handling system of claim 11 , wherein the first set of one or more indicators includes at least one eye tracker indicator sensed by an eye tracker of the near end device.
17. The information handling system of claim 16 , wherein the at least one eye tracker indicator is selected from: a head pose, a gaze point, and an eye position of the near end participant.
18. The information handling system of claim 11 , wherein the second output from the second machine learning engine reflects at least one of: a number of far end participants at a particular far end location and a room type of a particular far end location.
19. The information handling system of claim 11 , wherein the operations include:
determining identities of the one or more far end participants includes invoking a participant list function exposed by an application program interface (API) of the collaboration application.
20. The information handling system of claim 11 , wherein the orchestrator dynamically manages adjustments of the one or more audio parameters based on a composite of the first input from the first machine learning engine and a second input from the second machine learning engine.