Detection of a lecturer in a videoconference

• US 8,436,888 B1
• Filed: 02/20/2008
• Issued: 05/07/2013
• Est. Priority Date: 02/20/2008
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Hubenthal, Michael Brian, Baldino, Brian James
• Primary Examiner(s): MILLER, BRANDON J
• Application Number:
  US12/033,950
• Time in Patent Office: 1,903 Days
• Field of Search: 348/14.1, 348 1401- 1403, 348 1406- 1409, 348/14.16
• US Class Current: 348/14.08
• CPC Class Codes: H04N 7/152 Multipoint control units therefor

Abstract

In one embodiment, a system detects a lecturer within a videoconference. The system includes two or more participant systems and a media switch coupled to each of the participant systems. The media switch includes a switch processor to receive and process audio and video information from the participant systems. The switch processor analyzes loudness metric values of active speakers and designates a particular speaker as a lecturer. In addition, the switch processor provides audio and video information of the remaining conference participants to the lecturer participant system to rotate through conference participants with each participant being displayed to the lecturer for a predetermined interval (e.g., in a round robin fashion), thereby enabling the lecturer to view the entire audience.

Claims

1. A system comprising:

a plurality of participant systems each associated with at least one corresponding participant of a videoconference and including a plurality of audio devices to capture and present audio information of videoconference participants, and a plurality of video devices to capture and present video information of videoconference participants, wherein each participant system produces a loudness metric for each corresponding videoconference participant indicating an amplitude of said captured audio information; and

a switching device coupled to each participant system to receive said loudness metrics and said audio and video information of said videoconference participants, wherein said switching device includes a switch processor to process and distribute said received audio and video information, said switch processor including;

a lecture module to analyze said received loudness metrics and designate a videoconference participant as a lecturer; and

an audience module to cyclically distribute to one of said participant systems associated with said designated lecturer said received audio and video information of videoconference participants from others of said participant systems to facilitate presentation of each of said videoconference participants for a predetermined audience time interval to said designated lecturer.

2. The system of claim 1, wherein said lecture module includes:

a voice module to periodically determine average values of received loudness metrics for videoconference participants and designate said lecturer based on said average values in response to expiration of a predetermined lecture time interval.

3. The system of claim 2, wherein said voice module includes:

a speaker module to analyze and process said received loudness metrics and designate videoconference participants as speakers;

an average module to determine an average value for each video conference participant of said received loudness metrics for that videoconference participant, and to periodically capture said average values for each speaker and produce a lecture value for each speaker by determining an average value of said captured average values for that speaker; and

a lecturer module to designate said lecturer based on said lecture values for said speakers in response to expiration of said predetermined lecture time interval.

4. The system of claim 3, wherein said lecturer module includes:

a margin module to compare said lecture values of said speakers and designate said lecturer in response to a corresponding lecture value exceeding a lecture threshold and exceeding lecture values of remaining speakers by a threshold value.

5. The system of claim 3, wherein said voice module further includes:

a reset module to reset said lecture values and said predetermined lecture time interval in response to detection of a new speaker during said predetermined lecture time interval.

6. The system of claim 2, wherein said predetermined lecture time interval is in the range of five to fifteen seconds.

7. The system of claim 1, wherein said switch processor further includes:

a view module to distribute to said others of said participant systems said received audio and video information of said designated lecturer to facilitate presentation of said designated lecturer to videoconference participants,

wherein said view module comprises at least a first display for displaying one of said videoconference participants and a second display for displaying another of said videoconference participants.

8. The system of claim 1, wherein said switch processor further includes:

a speaker module to analyze and process said received loudness metrics and to remove said lecturer designation in response to detection of a speaker other than said designated lecturer.

9. The system of claim 8, wherein said switch processor further includes:

a view module to distribute to participant systems associated with videoconference participants other than said detected speaker said received audio and video information of said detected speaker to facilitate presentation of said detected speaker to those videoconference participants.

10. The system of claim 8, wherein said speaker module detects said speaker in response to said loudness metrics of a videoconference participant exceeding a threshold for a predetermined speaker time interval, wherein said predetermined speaker time interval is equal to or less than said predetermined lecture time interval.

11. The system of claim 1, wherein said predetermined audience time interval is in the range of one to twenty seconds.

12. A method comprising:

capturing audio and video information of each of a plurality of videoconference participants and producing a loudness metric for each videoconference participant indicating an amplitude of said captured audio information, the capturing being performed using a plurality of participant systems respectively associated with said plurality of video conference participants;

analyzing said loudness metrics and designating a videoconference participant as a lecturer; and

cyclically distributing to one of said participant systems associated with said designated lecturer said audio and video information of videoconference participants from others of said participant systems to facilitate presentation of each of those videoconference participants for a predetermined audience time interval to said designated lecturer.

13. The method of claim 12 further including:

periodically determining average values of said loudness metrics for videoconference participants and designating said lecturer based on said average values in response to expiration of a predetermined lecture time interval.

14. The method of claim 13 further including:

analyzing and processing said loudness metrics and designating videoconference participants as speakers;

determining an average value for each video conference participant of said loudness metrics for that videoconference participant, and periodically capturing said average values for each speaker and producing a lecture value for each speaker by determining an average value of said captured average values for that speaker; and

designating said lecturer based on said lecture values for said speakers in response to expiration of said predetermined lecture time interval.

15. The method of claim 14 further including:

comparing said lecture values of said speakers and designating said lecturer in response to a corresponding lecture value exceeding a lecture threshold and exceeding lecture values of remaining speakers by a threshold value.

16. The method of claim 14 further including:

resetting said lecture values and said predetermined lecture time interval in response to detection of a new speaker during said predetermined lecture time interval.

17. The method of claim 13, wherein said predetermined lecture time interval is in the range of five to fifteen seconds.

18. The method of claim 12 further including:

distributing to said others of said participant systems said audio and video information of said designated lecturer to facilitate presentation of said designated lecturer to videoconference participants,

wherein said distributing comprises;

displaying, in a first display, one of said videoconference participants; and

displaying, in a second display, another of said videoconference participants.

19. The method of claim 12 further including:

analyzing and processing said loudness metrics and removing said lecturer designation in response to detection of a speaker other than said designated lecturer.

20. The method of claim 19 further including:

distributing to participant systems associated with videoconference participants other than said detected speaker said audio and video information of said detected speaker to facilitate presentation of said detected speaker to those videoconference participants.

21. The method of claim 19 further including:

detecting said speaker in response to said loudness metrics of a videoconference participant exceeding a threshold for a predetermined speaker time interval, wherein said predetermined speaker time interval is equal to or less than said predetermined lecture time interval.

22. The method of claim 12, wherein said predetermined audience time interval is in the range of one to twenty seconds.