Abstract: In one embodiment, an apparatus includes two modules. A first module receives a request from a first endpoint to subscribe to a voice activated multicast group and causes the first endpoint to receive a current speaker's video stream if the first endpoint is not the current speaker and to receive a last speaker's video stream if the first endpoint is the current speaker. A second module receives a request from a second endpoint to subscribe to a continuous presence multicast group and causes the second endpoint to receive a continuous presence, current speaker video stream if the second endpoint is not the current speaker and to receive a continuous presence, last speaker video stream if the second endpoint is the current speaker. The continuous presence, current speaker video stream includes a composition or two or more video streams, one of which includes at least a portion of the current speaker's video stream.

Abstract: A speech processing system receives multiple streams of speech frames. The system selects among concurrent ones of the frames a subset of those frames that are the most relevant, based on pre-assigned stream priorities and energy content of the frames. The selected frames are then decoded and rendered. The resulting signals are mixed. This architecture provides bandwidth scalability and/or processing power scalability.

Abstract: A method by which video compression parameters of different video streams are changed to optimize overall bandwidth utilization according to rules related to the varying interactions among group members. Different video streams are allocated different portions of the overall bandwidth based on a priority scheme. The priority scheme is based upon group dynamics. A controller (central or distributed) receives a group dynamics monitoring signal which monitors the various activities of group members. The controller (central or distributed) uses the network congestion control feedback signal and the priorities assigned to the different video streams to send feedback signals to the different video transmitters requesting the transmitters to adhere to, increase, or decrease their video data rates by changing parameters controlling the video encoding process.