Incrementally resolving references in order to identify visually present objects in a situated dialogue setting