Smart Video Orchestration for Immersive Communication
21 February 2013
In the context of immersive communication, and in order to enrich sensorial and attentional immersion in videoconferences for a remote attendant, the problem of camera orchestration has been evoked. It consists of selecting and displaying the most relevant view or camera. To model the different video events we choose HMMs due to its numerous advantages comparing with Rule based-Models. Then, as video events, we adopted high level observations that allowed us to develop a new algorithm for the learning process, an algorithm that enable the training of the system by a non-expert final user. The rate of good detection of the screen to display is 0.87, so besides of lightening the configuration procedure, the model present good technical results. In the future, we intent to organize usability testing sessions in order to evaluate the new learning module from a final user perspective
