A generative spectral model for semantic mapping of buildings

Consider a mobile robot exploring an initially unknown school building and assume that it has already discovered some classrooms, offices, and bathrooms. What can the robot infer about the presence and the locations of other classrooms and offices in the school building? This paper makes a step toward providing an answer to the above question by proposing a system based on a generative model that is able to represent the topological structures and the semantic labeling schemas of buildings and to predict the structure and the schema for unexplored portions of these environments. We represent the buildings as undirected graphs, whose nodes are rooms and edges are physical connections between them. Given an initial knowledge base of graphs, our approach, relying on a spectral analysis of these graphs, segments each graph for finding significant subgraphs and clusters them according to their similarity. A graph representing a new building or an unvisited part of a building is eventually generated by sampling subgraphs from clusters and connecting them.