The combination of acoustical and video processing to achieve a smart audio and video feed from a set of microphones and cameras is a task that might conventionally be accomplished by camerapersons and control room staff. However, in the context of videoconferencing this process needs to be performed by control software. We discuss the use of a multi-camera multi-microphone set up for unattended videoconferencing, and present details of a prototype implementation being developed.
Dmitry N. Zotkin, Ramani Duraiswami, Vasanth Philo