Most nodes have an optional camera. The camera can be configured to follow moving objects and focus on possibly different moving objects.
When an object in the tree is selected, the camera and level of visual detail is taken from the node at or above the selected one, where the visual-detail and/or camera are enabled.