Deducing the visual focus of attention from head pose estimation in dynamic multi-view meeting scenarios | Synapse