Title :
Large-scale semantic mapping and reasoning with heterogeneous modalities
Author :
Pronobis, Andrzej ; Jensfelt, Patric
Author_Institution :
Centre for Autonomous Syst., KTH R. Inst. of Technol., Stockholm, Sweden
Abstract :
This paper presents a probabilistic framework combining heterogeneous, uncertain, information such as object observations, shape, size, appearance of rooms and human input for semantic mapping. It abstracts multi-modal sensory information and integrates it with conceptual common-sense knowledge in a fully probabilistic fashion. It relies on the concept of spatial properties which make the semantic map more descriptive, and the system more scalable and better adapted for human interaction. A probabilistic graphical model, a chaingraph, is used to represent the conceptual information and perform spatial reasoning. Experimental results from online system tests in a large unstructured office environment highlight the system´s ability to infer semantic room categories, predict existence of objects and values of other spatial properties as well as reason about unexplored space.
Keywords :
graph theory; human-robot interaction; inference mechanisms; knowledge representation; mobile robots; chain-graph; conceptual common-sense knowledge; heterogeneous modalities; human interaction; large-scale semantic mapping; large-scale semantic reasoning; multimodal sensory information; objects existence; online system tests; probabilistic framework; probabilistic graphical model; semantic room categories; spatial properties; spatial reasoning; system ability; Computational modeling; Humans; Probabilistic logic; Robot sensing systems; Semantics; Shape;
Conference_Titel :
Robotics and Automation (ICRA), 2012 IEEE International Conference on
Conference_Location :
Saint Paul, MN
Print_ISBN :
978-1-4673-1403-9
Electronic_ISBN :
1050-4729
DOI :
10.1109/ICRA.2012.6224637