Abstract
Autonomous robots are unsuccessful at operating in complex, unconstrained environments. They lack the ability to learn about the physical behaviour of different objects. We examine the viability of using qualitative spatial representations to learn general physical behaviour by visual observation. We combine Bayesian networks with the spatial representations to test them. We input training scenarios that allow the system to observe and learn normal physical behaviour. The position and velocity of the visible objects are represented as discrete states. Transitions between these states over time are entered as evidence into a Bayesian network. The network provides probabilities of future transitions to produce predictions of future physical behaviour. We use test scenarios to determine how well the approach discriminates between normal and abnormal physical behaviour and actively predicts future behaviour. We examine the ability of the system to learn three naive physical concepts, ‘no action at a distance’, ‘solidity’ and ‘movement on continuous paths’. We conclude that the combination of qualitative spatial representations and Bayesian network techniques is capable of learning these three rules of naive physics.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
J. F. Allen. Maintaining knowledge about temporal intervals. Communications of the ACM, 26(11):832–843, November 1983.
Peter Bryant. Perception and understanding in young children. Methuen and Co Ltd, London, 1974.
Peter Cheeseman. An inquiry into computer understanding. Computational Intelligence, 4(1):58–66, 1988.
A. G. Cohn. The challenge of qualitative spatial reasoning. ACM Computing Surveys, 27(3):323–325, September 1995.
Anthony G. Cohn. Qualitative spatial representation and reasoning techniques. In Gerhard Brewka, Christopher Habel, and Bernhard Nebel, editors, KI-97, Advances in Artificial Intelligence, pages 1–30. Springer-Verlag, Berlin, 1997.
Bruce D’Ambrosio. Inference in bayesian networks. AI Magazine, 1999.
E. Davis. The naive physics perplex. AI Magazine, 19(1):51–79, 1998.
H. L. Dreyfus. What Computers Still Can’t Do: A Critique of Artificial Reason. MIT Press, 1992.
J Fernyhough, AG Cohn, and DC Hogg. Building qualitative event models automatically from visual input. In Sixth International Conference on Computer Vision, pages 350–5. Narosa Publishing House, New Delhi, India, Jan 1998.
Francesco Gardin and Bernard Meltzer. Analogical representations of naive physics. Artificial Intelligence, 38:139–159, 1989.
Patrick J. Hayes. The naive physics manifesto. In D. Michie, editor, Expert Systems in the Microelectronic Age. Edinburgh University Press, Edinburgh, Scotland, 1978.
Patrick J. Hayes. The second naive physics manifesto. In Formal Theories of the Commonsense World. Ablex Publishing Corp., Norwood, New Jersey, 1984.
Patrick J. Hayes. Naive physics I: ontology for liquids. In Jerry R. Hobbs and Robert C. Moore, editors, Formal Theories of the Commonsense World, chapter 3, pages 71–107. Ablex, Norwood, New Jersey, 1985.
Daniel Hernandez. Qualitative Representation of Spatial Knowledge. Number 804 in Lecture Notes in Artificial Intelligence. Springer-Verlag, 1994.
F. V. Jensen. An introduction to Bayesian Networks. UCL Press (Taylor & Francis Ltd), London, 1996.
Douglas B. Lenat and R. V. Guha. Building Large Knowledge-Based Systems: Representation and Inference in the CYC Project. Addison-Wesley, Reading, Massachusetts, 1990.
D. B. Lenat, R. V. Guha, K. Pittman, D. Pratt, and M. Shepherd. CYC: towards programs with common sense. CACM, August 1990, 33(8):30–49, 1990.
Douglas B. Lenat. CYC: A large-scale investment in knowledge infrastructure. Communications of the ACM, 38(11):33–38, November 1995.
Richard Mann, Allan Jepson, and Jeffrey Mark Siskind. Computational perception of scene dynamics. In Roberto Cipolla Bernard Buxton, editor, Computer Vision-ECCV’ 96 vol 2, number 1065 in Lecture Notes in Computer Science, pages 528–39. Springer-Verlag, 1996.
J. McCarthy. Programs with common sense. In M. Minsky, editor, Semantic Information Processing, pages 403–418. MIT Press, Cambridge, MA, 1968. Part of this article is a reprint from an an article by the same title, in Proc. Conf. on the Mechanization of Thought Processes, National Physical Laboratory, Teddington, England, Vol. 1, pp. 77-84, 1958.
Amitabha Mukerjee. Neat vs scruffy: A review of computational models for spatial expressions. 1997.
Philippe Muller. A qualitative theory of motion based on spatio-temporal primitives. In Anthony G. Cohn, Lenhart Schubert, and Stuart C. Shapiro, editors, KR’98: Principles of Knowledge Representation and Reasoning, pages 131–141. Morgan Kaufmann, San Francisco, California, 1998.
A. Musto, K. Stein, A. Eisenkolb, K. Schill, and W. Brauer. Generalization, segmentation and classification of qualitative motion data. In Henri Prade, editor, Proceedings of the 13th European Conference on Artificial Intelligence (ECAI-98), pages 180–184, Chichester, August 23–28 1998. John Wiley & Sons.
Jean Piaget and Baerbel Inhelder. The Psychology of the Child. Routledge and Kegan Paul, London, 1969.
R. Rajagopalan and B. Kuipers. Qualitative spatial reasoning about objects in motion: Application to physics problem solving. In Dan O’Leary and Peter Selfridge, editors, Proceedings of the 10th Conference on Artificial Intelligence for Applications, pages 238–245, San Antonio, TX, March 1994. IEEE Computer Society Press.
David A. Randell, Zhan Cui, and Anthony G. Cohn. A spatial logic based on regions and connection. In William Nebel, Bernhard; Rich, Charles; Swartout, editor, Proceedings of the 3rd International Conference on Principles of Knowledge Representation and Reasoning, pages 165–176, Cambridge, MA, October 1992. Morgan Kaufmann.
Jochen Renz and Bernhard Nebel. On the complexity of qualitative spatial reasoning: A maximal tractable fragment of the region connection calculus. Artificial Intelligence, 108(1–2):69–123, 1999.
Stuart Russell and Peter Norvig. Artifical Intelligence: A Modern Approach,. Prentice-Hall, Englewood Cliffs, NJ, ISBN 0-13-103805-2, 912 pp., 1995, 1995.
C. Schlieder. Qualitative shape representation. 1994.
J. M. Siskind and Q. Morris. A maximum likelihood approach to visual event classification. In Roberto Cipolla Bernard Buxton, editor, Computer Vision-ECCV’ 96, number 1065 in Lecture Notes in Computer Science, pages 347–62. Springer-Verlag, 1996.
J. K. Tsotsos, J. Myopoulos, H. D. Covvey, and S. W. Zucker. A framework for visual motion understanding. IEEE-T PAMI, 2(6):563–573, 1980.
Elaine Vurpillot. The Visual World of the Child. George Allen and Unwin Ltd, London, 1976.
Daniel Weld and J. DeKleer. Qualitative Reasoning about Physical Systems, ed. Morgan Kaufmann, 1990, 1990.
Kai Zimmermann. Enhancing qualitative spatial reasoning—combining orientation and distance. In Andrew U. Frank and Irene Campari, editors, Spatial Information Theory. A Theoretical Basis for GIS. European Conference, COSIT’93, volume 716 of Lecture Notes in Computer Science, pages 69–76, Berlin, September 1993. Springer.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2001 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Boxer, P.A. (2001). Towards Learning Naive Physics by Visual Observation: Qualitative Spatial Representations. In: Stumptner, M., Corbett, D., Brooks, M. (eds) AI 2001: Advances in Artificial Intelligence. AI 2001. Lecture Notes in Computer Science(), vol 2256. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-45656-2_6
Download citation
DOI: https://doi.org/10.1007/3-540-45656-2_6
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-42960-9
Online ISBN: 978-3-540-45656-8
eBook Packages: Springer Book Archive